SCIENTIFIC ESTABLISHMENTS AND HIERARCHIES

SOCIOLOGY OF THE SCIENCES

A YEARBOOK

Editorial Board:
G. Bohme, Technische Hochschule, Dannstadt
N. Elias, Universities of Leicester and Bielefeld
Y. Elkana, The Van Leer Jerusalem Foundation, Jerusalem
L. Graham, Massachusetts Institute of Technology
R. Krohn, McGill University, Montreal
W. Lepenies, Free University of Berlin
H. Martins, University of Oxford
E. Mendelsohn, Harvard University
H. Nowotny, European Centre for Social Welfare Training

and Research, Vienna

H. Rose, University of Bradford
Claire Salomon-Bayet, University of Paris
P. Weingart, University of Bielefeld
R. D. Whitley, Manchester Business School, University ofManchester

Managing Editor: R. D. Whitley

VOLUME VI - 1982

SCIENTIFIC
EST ABLISHMENTS
AND
HIERARCHIES
Edited by
NORBERT ELIAS
Universities of Leicester and Bielefeld

HERMINIO MARTINS
University of Oxford

and
RICHARD WHITLEY
Manchester Business School, University of Manchester, England

D. REIDEL PUBLISHING COMPANY

DORDRECHT: HOLLAND / BOSTON: U.S.A.
LONDON: ENGLAND

Library of Congress Cataloging in Publication Data

Main entry under title:
Scientific establishments and hierarchies.
(Sociology of the sciences: a yearbook; v. 6)
Includes index.
1. Research institutes. 2.
aspects.
I. Elias, Norbert. II. Martins, Herminio. III. Whitley,
Richard (Richard D.) IV. Series: Sociology of the sciences;
v.6.
Q180.AIS345
306'.45
82-308
ISBN-13: 978-90-277-1323-0
001: 10.1 007/978-94-009-7729-7

e-ISBN-13: 978-94-009-7729-7

Published by D. Reidel Publishing Company,

P.O. Box 17,3300 AA Dordrecht, Holland
Sold and distributed in the U.S.A. and Canada
by Kluwer Boston Inc.,
190 Old Derby Street, Hingham, MA 02043, U.S.A.
In all other countries, sold and distributed
by Kluwer Academic Publishers Group,
P.O. Box 322, 3300 AH Dordrecht, Holland
D. Reidel Publishing Company is a member of the Kluwer Group

All Rights Reserved

Copyright 1982 by D. Reidel Publishing Company, Dordrecht, Holland
and copyrightholders as specified on the appropriate pages within
Sotlcover reprint of the hardcover 1sl edition 1982

No part of the material protected by this copyright notice may be reproduced or

utilized in any form or by any means, electronic or mechanical,
including photocopying, recording or by any information storage and
retrieval system, without written permission from the copyright owner

TABLE OF CONTENTS

Introduction

vii

xiii

Contributors to this Volume

PART I
Scientific and Other Establishments
N. ELIAS - Scientific Establishments
P. WEINGART - The Scientific Power Elite - a Chimera; The De-

institutionalization and Politicization of Science

3
71

H. KATOUZIAN - The Hallmarks of Science and Scholasticism: A

Historical Analysis

89

C. HAY - Advice from a Scientific Establishment: the National

Academy of Sciences

111

PART II
Establishments and Hierarchies in the Development of
Scientific Knowledge
E. YOXEN - Giving Life a New Meaning: The Rise of the Molecular

Biology Establishment
HOHLFELD - Two Scientific Establishments which Shape
the Pattern of Cancer Research in Germany: Basic Science and
Medicine
J. FLECK - Development and Establishment in Artificial Intelligence
A. RIP - The Development of Restrictedness in the Sciences
T. SHINN - Scientific Disciplines and Organizational Specificity:
the Social and Cognitive Configuration of Laboratory Activities

123

R.

145
169
219
239

vi

Table of Contents

PART III
Establishing Boundaries and Hierarchies in the Sciences
R. G. A. DOLBY - On the Autonomy of Pure Science: The Con-

struction and Maintenance of Barriers between Scientific Establishments and Popular Culture
D. CHUBIN and T. CONNOLLY - Research Trails and Science
Policies: Local and Extra-Local Negotiation of Scientific Work
R. D. WHITLEY - The Establishment and Structure of the Sciences
as Reputational Organizations

313

Index

359

267
293

INTRODUCTION

In recent years sociologists of sciences have become more interested in scientific elites, in the way they direct and control the development of sciences
and, beyond that, in which the organization of research facilities and resources
generally affects research strategies and goals. In this volume we focus on
scientific establishments and hierarchies as a means of bringing aspects of
these concerns together in their historical and comparative contexts. These
terms draw attention to the fact that much scientific work has been pursued
within a highly specific organizational setting, that of universities and academic research institutes. The effects of this organizational setting as well as
its power relations, and its resources in relation to governmental and other
non-scientific establishments in society at large, deserve closer attention.
One significant aspect of scientific establishments and hierarchies and of
the power relations impinging upon scientific research, is the fact that the
bulk of leading scientists have the professional career, qualifications and
status of a professor. As heads or senior members of departments, institutes
and laboratories, professors form the ruling groups of scientific work.
They are the main defenders of scientific - or departmental - autonomy,
accept or resist innovations in their field, play a leading part in fighting
scientific controversies or establishing consensus. Even where research units
are not directly controlled by professors, authority structures usually remain
strongly hierarchical. These hierarchies too deserve attention in any exploration of the social characteristics of scientific knowledge and its production.
Not only do heads of institutes and departments often direct and control the
careers of their scientific staff, but they also wield considerable power in the
evaluation of scientific research and its practitioners. As most scientific work
is conducted in a hierarchical setting whose apex is formed by a more or less
authoritative scientific establishment, the significance of this social structure
for the development of scientific knowledge is considerable.
Scientific establishments and hierarchies within departments and research
institutes need to be considered in the wider context of hierarchies between
departments and scientific disciplines. Different departments vary in the
vii
Norbert Elias, Herminia Martins and Richard Whitley reds.), Scientific Establishments
and Hierarchies. Sociology a/the Sciences, Volume VI, 1982. vii-xi.
Copyright 1982 by D. Reidel Publishing Company.

viii

N. Elias and R. D. Whitley

amount of resources they control and different scientific disciplines vary in

power resources, in status and prestige. These power and status hierarchies
too directly affect production and development of scientific knowledge, for
instance, through the differential status associated with certain concepts,
techniques and types of explanation by comparison with others. The relative
prestige of scientific fields is not only an outcome of intra-scientific pressures
and debates, but it also results from extra-mural pressures and successes.
Especially since the Second World War relations between scientific and nonscientific establishments have become a significant aspect of the organization
and development of the sciences.
The development of establishments in the sciences producing and controlling distinct bodies of knowledge obviOUsly depends on the cognitive value
of the knowledge produced by them. Moreover, it requires a considerable
measure of autonomy in its relationships with other powerful groups. The
way in which this autonomy has been attained and subsequently maintained
and enlarged, is by no means fully understood. The relative autonomy of
sciences certainly affects conceptions of sciences and these in tum are connected to hierarchies of scientific fields. It has been suggested that the compromises made by scientists in 17th-century England in their efforts to
legitimize science as a mode of cognitive orientation, affected the particular
conception of scientific knowledge which then became institutionalized. In
the same way it can be argued that the sort of science which led to the vast
expansion of sciences since the Second World War became so dominant by
demonstrating its success and usefulness in terms of the goals of non-scientific
establishments. As a result its own aims and structures have altered. Current
conceptions of scientific knowledge and scientific method reflect these
changes. They have reinforced the tendency to see one specific type of
science as the only valid form of scientific work, as the authoritative model
for all other sciences.
ln his theme-paper for this volume Norbert Elias provided an introductory
account of aspects of scientific establishments and of their connections with
other social establishments. This paper served as the basis for inviting possible
contributors to prepare drafts for a conference held at St Antony's College,
Oxford in July 1980. FollOwing this meeting, papers were revised for this
book in the light of comments and suggestions.
Collaboration between scientific establishments and powerful non-scientific

Introduction

ix

groups, as well as the emergence of a trans-science establishment, has led

some commentators to discern a major change in the political organization
of society itself - a king of "scientification" of politics. As Weingart points
out, though, these processes have been accompanied more by the politicization of science and the decline in its institutional specificity and independence
than by the accession of scientists to dominant positions. The case of the
u.s. National Academy of Sciences, discussed by Hay, illustrates the limited
control exerted by one establishment organization over national policies.
Nonetheless, the increasing dominance of the physical sciences as the
icon of scientificity, and science in general as the only legitimate form of
knowledge, has had major effects on the organization of the sciences and the
development of new areas. Katouzian considers the "professionalization" of
academic science as having turned much science into scholasticism. While
this is partly a consequence of directly linking occupational statuses to international intellectual reputations so that jobs and promotions, especially in
universities, became dependent on following particular intellectual goals and
procedures, the expansion of the sciences and the dominance of a particular
conception of scientific research have greatly reinforced this tendency. The
pattern of competition and increasing specialization which is so marked in
much contemporary science is not only due to jobs and resources becoming
controlled by reputational systems in the manner sketched by Whitley, but
also an outcome of the hierarchization of sciences according to their closeness
in approach and organization to parts of physics. Furthermore, while such an
ordering of scientific fields has perhaps always been evident to some extent,
at least since science became a legitimate form of knowledge production and
validation and, hence, an honorific adjective, its entrenchment in bureaucratic
structures and policy apparatuses makes it much more rigid and powerful.
Some of the effects of particular establishments and hierarchies are demonstrated in the case studies by Yoxen, Hohlfeld, and Fleck. They point to
hierarchies of conceptual approaches and ontologies influencing the development of molecular biology, cancer research and artificial intelligence. At the
same time they show how extra scientific establishments directly intervene
in knowledge development and reinforce such hierarchies. They also illustrate
the loss of autonomy of academic science and the greater fluidity of organizational arrangements since the First World War. The Rockefeller Foundation
initiative in biology not only change the way scientists thought of "life", it

N. Elias and R. D. Whitley

also heralded major shifts in the way research was organized, administered,
and funded .. State direction of research has, of course, now become commonplace but its organizational implications are not always realized. In particular,
the development of full-time research laboratories which are partly oriented
to non-intellectual goals has led to the formation of new fields which transcend traditional academic boundaries and values. Furthermore, funding
patterns now affect intellectual development to a greater extent and more
directly. They are often short term, leading to greater variety of intellectual
goals and a strong emphasis on "efficiency" in the administration of research.
This in tum increases the division of labour and specialization as scientists
strive to acquire reputations in a narrow, highly specific area.
A major part of the dominant belief system in science is, of course, the
need for greater precision and control over phenomena. This is discussed by
Rip in terms of the "restrictedness" of objects in Chemistry which enabled
that science to develop standardized samples and techniques. Such standardization encouraged the extensive division of labour in modem chemistry
which is described by Shinn in his comparison of laboratories in different
sciences. As he shows, research goals differed in this fields, as did the degree
of control over materials - or tasks uncertainty - and these variations were
clearly connected to the authority and communication structures of the work
organizations.
The connections between differences in laboratory organization and
differences in intellectual structures are not always easy to discern - and in
many way have become more complex since the development of multi-goals,
multiple-funded, full time research laboratories. A number of studies have
emphasized the importance of "local organizational" factors in the development of research strategies, especially in the biological sciences, but few have
directly addressed the issue of how such strategies coalesce - or do not
coalesce - into coherent social groups around particular intellectual goals and
procedures. Chubin and Connolly suggest the metaphor of research "trails"
bunching to form specialties in particular periods and the utility of different
evolutionary models in dealing with this problem.
The dangers of over reifying social boundaries have, of course, been well
rehearsed in sociology as a whole, and this had led some writers to an extreme
voluntarism. However, the development and maintenance of social and
intellectual boundaries are themselves social strategies which have intellectual

Introduction

xi

consequences as Dolby shows in his discussion of Darwinism and sociobiology.

The formation of social groups based on expertise represents an attempt to
exclude the lay audience and other "expert" groups from influence and
competition; it is a bid for autonomy and independence which, if successful,
leads to increased mutual dependence among "professionals" and increased
standardization of craft skills and techniques.
The complex relations between local organizational establishments,
national and international establishments in particular fields, interdisciplinary
establishments and extra-scientific establishments are an important area for
further research. They necessitate an approach which recognizes the plurality
of ways of acquiring and organizing scientific knowledge. This recognition
underlies all the papers in this volume which together represent an attempt to
link the study of scientific work at an organizational level with patterns of
intellectual and social change at a more general level.
N. ELIAS
ZiF

Bielefeld

R. D. WHITLEY
Manchester Business School

CONTRIBUTORS TO THIS VOLUME

Dr C. Chubin, Georgia Institute of Technology, Atlanta

Dr T. Connolly, Georgia Institute of Technology, Atlanta
Dr A. Dolby, University of Kent at Canterbury
Prof. N. Elias, Zentrum [iir Interdisziplinare Forschung, Bielefeld
Mr J. Fleck,Aston University, Birmingham
Dr C. Hay,Brunei University, Uxbridge
Dr R. Hohlfeld, Universitiit Erlangen-Nurnberg, Erlangen
Dr H. Katouzian, University of Kent at Canterbury
Dr A. Rip, University of Leiden
Dr T. Shinn, CNRS, Paris
Prof. P. Weingart, Universitiit Bielefeld
Mr R. D. Whitley, Manchester Business School, University of Manchester
Dr E. Yoxen, University of Manchester

PART I

SCIENTIFIC AND OTHER ESTABLISHMENTS

SCIENTIFIC ESTABLISHMENTS*

NORBERT ELIAS
Zentrum fiir Interdisziplinare Forschung, Bielefeld

1. Introduction
The social characteristics of scientific establishments are bound up with those
of the social institutions where most of them are located - with those of
universities. Occasionally, though, in the development of sciences, extra
mural establishments have arisen, especially during the earlier phases. Among
the best known examples are the Paracelsians and the Freudians - both
representatives of medical sciences which may be significant. Of course,
the interplay, and particularly the battles, between intra mural and extra
mural establishments, like their distinguishing characteristics, deserve more
attention: But the simple reference to extra mural establishments alone is
enough to put the problem of the relationship between universities and
sciences into better perspective. Whatever the contributions of extra mural
establishments to the development of sciences have been in the past and may
be at present, university-related groups have .gained the ascendancy - in the
development of almost all higher branches of learning they play the dominant
part.
During the Middle Ages universities in Europe formed centres of higher
learning at its pre-scientific, mainly theological and philosophical, stage where
it was linked to the authority of revealed knowledge. In the Renaissance
humanist groups, still tied to authority - that of Greek and Roman antiquity
-- established themselves in some European universities as powerful, model
setting groups of learned men, thus replacing the Church controlled by
court- and state- controlled establishments. Then, with the take off into
science during the 15th, 16th, and 17th centuries, natural scientists, together
with science-oriented philosophers and mathematicians, slowly made their
way into the universities and established themselves there. Finally, university3
Norbert Elias, Herminio Martins and Richard Whitley reds.), Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 3-69.
Copyright 1982 by Norbert Elias.

Norbert Elias

trained and centred scientific professionals in professorial positions became

dominant representatives of scientific work in each of its branches. Moreover,
during the 19th and 20th centuries the division into more and more specialized branches of all the major fields of scientific knowledge taught and
explored at universities accelerated at a rate unknown before (as part of
an unplanned process whose dynamics need not concern us here). Hence,
departments or institutes related to universities as the institutional setting of
scientific specialisms multiplied too.
A!; a rule, each of these academic departments or institutes consist of a
hierarchy of offices headed by one or more people with the title, qualifications, power resources, and emoluments of a professor. He or, as the case may
be, she or they, form a local scientific establishment, perhaps together with
senior members of their staff who head the middle rung of an institute or a
department followed at the lower rungs of the ladder by junior members
hoping to ascend.
At first glance, the hierarchic order of offices, rank, salaries etc. in these
academic institutions may not seem so very different from that in economic,
military, administrative, or other organizations with a firmly instituted office
hierarchy. There are indeed, similarities which should not be overlooked. In
contemporary societies scientific knowledge is taught and advanced in an
increasingly bureaucratized setting. Bureaucratic control favours routinization.
A!; an advance in knowledge, though, scientific innovation cannot be easily
routinized. If it is, it may get stale, may stagnate, follow a set track, and be
tied to a prescribed method; its representatives may miss unusual openings,
and may shun radical innovations. This then, is one of the Significant differences between scientific and other bureaucratized institutions. Innovatory
research requires a fair measure of independence for its practitioners. If
institutionalized research is to be productive, a fairly wide margin for the
autonomy of its practitioners has to be built into the institutions. One of
the central problems of scientific establishments financed and controlled
by extraneous agencies thus becomes that of the balance between dependence and independence. The power tensions and conflicts, potential or
actual, which are generated by balance problems of this kind, form part
of a wider set of standing tensions and conflicts which have their roots in
the social situation of scientific establishments and in the figuration of
interdependencies within which they have their place. One way of advancing

Scientific Establishments

the understanding of these problems may be a systematic comparison between

Eastern and Western scientific establishments,not only in the field of physical
and biological but also in that of human sciences. That could serve as a
testing field of the wider problems resulting from the encounter between
bureaucratic state control and the intrinsic need for relative autonomy of
scientific establishments.
There are other distinguishing characteristics of these establishments. A
junior member may surpass in inventiveness, imagination, and power of
scientific discovery some or all of the higher ranking members of a department - he or she may even succeed in publishing his (her) fmdings under
his or her own name, thus avoiding its appropriation by higher-ranking and
more powerful members of the institution.
All in all, the nature of scientific work gives rise to special problems
with regard to the bureaucratic relationship of people bound together as
members of the same or related institutions. It has a strongly individualiZing
effect, counter balanced - and often countered - by the need for scientific
collaborated and team work. Also, scientific institutions acquire rank and
status collectively within their own field or beyond it. The standing of an
institution within this highly competitive, wider field strongly affects that of
the institution's individual members and vice versa. The internal figuration of
university departments and institutes with its inbuilt pressures and constraints,
its specific potential for cooperation and strife, and its self-perpetuating
tendencies, has unplanned effects upon scientific production. A comprehensive
enquiry could hardly bypass them.
However, in this context attention is focussed mainly on the professors.
They form the core of local, national, and international scientific establishments. As the main power holders of scientific enterprises, they playa key
part in the development of scientific knowledge in each of its numerous
branches. At present, they are, as it were, the principal human agents of
scientific processes. In the last resort, it is they whose consensus, dissenSions,
and power struggles on the national or the international level, determine
whether and which scientific discoveries and innovations are recognized
as advances in human knowledge, are received into the common fund of
knowledge of one or several branches of higher learning and thus as part of
the teaching programmes, are handed on to the next generations. Ruling
opinion among the professors of a scientific field can paralyze or stimulate

Norbert Elias

productivity in that branch of learning. How and why the reception of

innovation and discoveries into the common fund of a scientific field is - or
is not - achieved, or how innovation is blocked or given leeway, along with
many other related problems, need no longer be a matter of philosophical
or historical surmise. There is a great deal of factual evidence available for
enquiries into problems of this kind; they are open to systematic scientific
enquiries.
2. Metaphysics and the Phllosophers' Establishment
It is no longer necessary or particularly useful to discuss the problem of the
relationship between the human agents of scientific knowledge and their
changing institutional setting on the one hand, and changes in human knowledge itself on the other hand, in those terms to which we have become
habituated through an old-established, respected, and powerful philosophical
tradition. No doubt, the type of abstraction cultivated by representatives .of
that tradition carries high prestige but its cognitive value is rather doubtful.
It is still quite common in learned discourse to refer to the persons who carry
on scientific research as the "subject" of science or of knowledge, or else one
may refer to them by means of sverely reducing and de-personalizing concepts
such as "reason" or "consciousness". Lack of a theory of concept-formation
processes makes it difficult to indicate briefly the peculiar nature - and the
defects - of the type of concept which forms the stock-in-trade of philosophical epistemology and metaphysics. But one of the common underlying
characteristics of such leading concepts, which serve as tools of a philosophical
theory of scientific knowledge, is the tacit assumption that science is an
eternal given of mankind. In fact, it is the late result of a long development.
These conceptual tools represent the philosopher's attempt at approaching
science as if it were a piece of nature. Without stating it explicitly, philosophers approach science as an object of their reflection in the same manner in
which the great pioneering scientists approached natural objects. like the
latter, they take it for granted that their effort of discovery must be directed
towards the discovery of unchanging and universal general regularities or laws.
As a matter of course, philosophical concepts are cast in a physicalist mould.
The concept "metaphysics" reflects this direction of the philosopher;s aim
and the method of their abstraction. Not abstraction as such, is at fault, but

Scientific Establishments

the type of abstraction which dominates the thinking of metaphysical philosophers about science and human knowledge. From the philosophers'
perspective these two appear as pieces of nature to be caught like all others
in a conceptual net of static and law-like generalizations beyond time. Yet
sciences are not part of physical nature. To pretend that they are is a futile
exercise. One cannot hope to discover universal laws, similar to those discovered by classical physicists, of something that is in no way universal, that
is in fact the result of a long diachronic social process. Philosophical sciencetheory with its law-like generalizations obscures the fact that sciences form
part, not of physical nature, but of the human level, of what one often calls
"culture", of the symbolic universe of human beings. Its firmly-structured
development is as inaccessible to static abstractions such as "eternal laws",
"validity", or "truth", as to unstructured historical descriptions. It lies
beyond the reach of both philosophical absolutism and historicist relativism.
When Galileo discovered regularities of falling bodies by studying the
movement of downward rolling balls with the help of a clepsydra, he could
rely on a vast fund of human knowledge, which gave him the certainty that
the regularities of falling bodies, which he observed there and then and which
he concisely represented as a general rule with the help of a few mathematical
symbols, would be the same the world over. When Descartes or, for that
matter, Kant tried to examine fundamentals of human knowledge by introspectively examining their own knowledge, they made the same assumption.
They assumed, as a matter of course, that concepts at a very high level of
generality, which they found in their own possession - concepts such as
"reason", "natural law", "mechanical causation" or "substance" - must be
universal properties of human beings everywhere. They assumed, in short,
that these concepts formed part of the nature of men. This assumption,
however, was fictitious. Philosophers had learned these concepts, and others
of the same type, with their language as part of the common conceptual
stock -in -trade of their society or, at least, of the stock-in-trade oflearned men
of their time. Had they lived a couple of hundred years earlier, they could not
have used, and not reflected upon, the same concepts because they were not
available in their society. Even if the words "reason" "nature", "natural law",
or "cause" were in use, their meaning in that age - and thus the concepts were different. The assumption that immutable laws discovered in a particular
case are the same universally, which was reasonably well founded in the case

Norbert Elias

of physical events, was unfounded in the case of men's changeable linguistic

and conceptual symbols, the vehicles of human knowledge generally, and of
scientific knowledge in particular. For sciences are not only processes themselves, they also represent a comparatively late stage in the wider knowledge
process of humanity. As such they require theories of a different type from
those of classical physics - process-theories representing diachronic sequences,
not law-like theories abstracted from all sequences of this kind. Transcendental philosophy has led us astray. Its representatives try to persuade us that they
can discover eternal and unchanging conditions or forms of whatever it is - reflection, science, or experience - while, in fact, they merely abstract selectively
and, often enough in a highly arbitrary manner, law-like generalizations from
a given phase in the development of human knowledge, thus adding their mite
to the history of philosophy. The case of Kant, father of the concept ''transcendental", can serve as an exemplary illustration. A very simple and unequivocal summary of his basic position can be found in one of his last works,
in his Essay of the Progress of Metaphysics since l.eibniz and Wolff. He states
quite clearly (1) that, in his view, the form of the objects, as men perceive
them, is not founded in the properties of the objects as such, but in the
natural properties of the subjects. Transcendentalism and metaphysics, in
fact, are based on the hypothesis that specific forms of connection of knowledge, reason or experience, are given to men by nature, are, in other words,
innate. The crux of the matter is that philosophical statements, such as those
made by Kant about the ''natural properties of men" and all transcendental
philosophies after his, at the present stage in the development of knowledge,
require examination by competent biologists. For Kant developed the discarded concept of speCific innate ideas into that of specific innate forms of
connection, such as categories or forms of apperception, which he believed
to be built into people by nature and therefore, unlearned. Today that can
only mean Kant believed them to be innate.
The connection of events as mechanical and impersonal causes and effects
can serve as an example. It is indeed a highly specific form of connecting
events. Kant recognized - rightly - that no human being could conceive the
idea of such a connection as a result of his own experiences of nature. He
concluded - wrongly - that it must, therefore, be built into people's faculty of
connecting events prior to any experience as part of their natural make-up. He
did not consider the possibility that humans could learn concepts such as this

Scientific Establishments

in their society as part of its conceptual and linguistic heritage. Kant, in fact,
had learned the concept of cause in this manner, together with many others of
the same kind. It was prior to his individual experience because it existed in
his society prior to his own existence. In his society he learned the concept of
mechanical and impersonal causation from his elders. In his reflections, however, he disregarded the fact that he had learned from others the word "cause"
and its meaning in the specific sense it had attained at the time he was born.
As learning forms part of a person's experience, he disregarded the fact that
concepts such as "cause" came to him through experience in that sense.
According to current knowledge, the metaphysical idea that the concept
of mechanical causation and other similar concepts of connection are built
into people's unchanging nature prior to all experience, can only mean that
this type of relationship-concept is genetically implanted in man and thus
transmitted from one generation to another without learning as part of man's
biological constitution. In actual fact, these concepts developed in a society
over time as part of its developing fund of knowledge. They came into being
through a continuing process of intergenerational conceptualization of
experiences. They were transmitted from one generation to another, not
genetically but through learning, and changed - advanced or declined in
scope, precision, and fitness - according to their double function as a means
of communication and a means of orientation; they changed as a result of
changes in knowledge and collective experiences of such societies which, in
turn, were dependent on antecedent conceptualization.
Systematic enquiries into long-term processes of conceptualization as
well as theoretical models of such processes, are still in their infancy. That
the Greeks used a word with a meaning akin to our "guilt" (aida) as term
with a meaning akin to our "cause", no doubt, is suggestive of a process 'of
coceptualization leading from a more personalized and involved to a more
impersonal and detached stage (2), from the quest for someone who is to
blame to the quest for something that acts without purpose, according to
recurrent rules. There are many 'other examples of long-term processes of
conceptualization all pointing in the same direction. To study and to explain
processes of this kind would certainly fill a significant gap in our knowledge.
Philosophy, however, tends to block rather than to encourage enquiries of
this kind. The philosopher's assumption that humans have, as it were, innately
specific rules of experience prior to any experience, makes it impossible for

10

Norbert Elias

them to escape from the notion of such a barrier interposing itself between
"object" and "subject". This conception of rules is the barrier. As long as one
believes in transcendental rules or forms, and regards their discovery as the
principal task of a philosophy of knowledge in general, and of sciences in
particular, the latter, human knowledge - non-scientific and scientific must appear as permanently patterned by people's own constitution. On a
transcendentalist view, human experience, for all its variety, seems to be set
for ever into a rigid and unvarying mould.
The strong solipsistic tendency underlying that kind of philosophy has its
intellectual roots in this assumption. Kant gave expression to this tendency
when he spoke of the significant doubt as to whether the object "which
we suppose to be outside of us may not perhaps be within us" (3). Similar
solipsistic doubts can be found in the writings of other transcendental philosophers, among them Husserl and Popper. However much they differ in other
respects, sooner or later they sound a note of despair.
One may well ask why this odd kind of philosophy, from Descartes to
Husseri and beyond, has gained so strong a position in the world of learning
which is, perhaps, even stronger in the 20th century than ever before. The
question deserves a more extensive enquiry. Off-hand one can refer to a few
of the interdependent conditions which contribute to the ascendancy of
transcendental philosophy in the wQrld of academic learning and to the firm
hold which metaphysical beliefs have gained over science theories in our time.
One of these conditions, as I have indicated elsewhere (4), is the strong
civilizing spurt which set in during the 16th and 17th centuries and which,
with many cross- and counter- currents and with many swings of the pendulum, continues to this day. It made for a stronger armour of individual selfrestraint interposing itself, all round, more evenly and stably between people's
spontaneous impulses to act and the execution of any action. The continued
growth and the growing effectiveness of state control and specific changes in
the mode of control, such as the change from almost exclusive reliance on
direct command and constraint to greater reliance on indirect methods of
control, contributed to this civilizing spurt. So did the increasing, lengthening,
differentiation and stability of interdependence-chains binding individuals
to each other. A complex of social changes in this direction offered social
rewards in terms of status, career, power chances or income to men capable
of containing their short-term impulses in the pursuit of long-term aims.

Scientific Establishments

11

The ascent of sciences and the scientification of society, too, were closely
connected with this civilizing spurt.
Take the transition from a geocentric to a heliocentric view of the universe,
which was one of the most momentous achievements of early science. The
geocentric view, even in its most sophisticated form, was characteristic of
people's unreflected self-centredness which dominated the long sequence of
pre scientific stages. The heliocentric view marked the ascent to a level of
consciousness at which humans were able, in their reflections, to look at themselves as it were, from a distance. The Copernican innovation implied, not
merely an advance in people's substantive knowledge, but also the ascent to a
higher level of restraint and self-distanciation; in short, a specific change in
people's personality structure. It required a lessening of their involvement in
the experience of natural events or, expressed differently, a greater capacity
for detachment. They had to forgo the pleasure of experiencing themselves,
humans on this earth, as the centre of the universe and to accept instead, an
emotionally far less appealing, although better orientating, i.e., more realistic,
world view.
A higher degree of individualization was another aspect of the same
transformation. Increasing differentiation of society and the stronger social
pressure for individual self-restraint, threw people back on their own resources. At the same time, the greater measure of self-restraint in more and
more situations, required of people by the larger and increasingly complex
figurations they formed with each other, was, in self-reflection, experienced
by many of them as an invisible wall separating them from other objects and
persons.
Here lay some of the strongest social and emotional roots of the homo
clausus feeling and the solipsistic tendencies that went with it. Transcendental
philosophy reflected this experience. It gave expression to it in the form of
the paradigmatic model underlying its knowledge theory - of the model
according to which an existential barrier interposed itself between the "subject" of knowledge and its "objects", between the ideas ''within'' and the socalled "external" world'.
The intellectual construction giving expression to this feeling of a barrier
and a distance between the "world within" and the "world without", the
assumption of an unlearned and immovable panel of rules and forms, built
into men prior to all experience, was all the more inflexible as it had a very

12

Norbert Elias

high value for philosophers. Their whole social and professional existence as
philosophers was bound up with it. With few exceptions, philosophy to this
day embodies a specific hierarchy of values which dominates the cast of its
problems as well as of solutions that appear philosophically relevant. It is a
value scheme which philosophers share with theologians and classical physicists. Transcendental philosophy, in particular, is "set" at discovering behind
the changing knowledge and experience of humans, something that is not
changing - something eternal and immutable. The value scheme underlying
this design of philosophical problems and of the forms of thinking that go with
it, may appear simple and perhaps self-evident. According to it, that which is
unchanging and eternal, as an object of enquiries has a much higher value, than
that which is changeable and impermanent. The implied assumption, however,
is far from self-evident. If people in quest of knowledge, allow themselves to
be guided by this value scheme, they implicitly assume that whatever they try
to explore, is itself arranged in a specific way. The value statement, on closer
inspection, reveals itself as an ontological statement. According to it, the
understanding and explanation of all that is changing can and must be found
in something behind it that is timeless, immobile, and immutable. The value
scheme is linked to a substantive view about the nature of things; it has, in
other words, ontological significance. Concealed in it is a statement - or a
belief - about the existing structure of the world. The classical philosophers'
and the classical physicists' concept of explanations in terms of eternal laws
or lawlike statements corresponds to this implied ontological belief.
There is no need to assume that this belief and the implied value scheme
are the result of a clear cognitive decision. It is more likely that this value
scheme reflects specific wishes and experiences of people themselves, not the
structure of their objects. Pride of place among these wishes probably takes
men's desire to find relief from the awareness of their own transience, the
wish to discover behind the impermanence of their lives, symbols of something that is unchanging and imperishable. It is an old tradition that among
the groups who specialize in the management and production of means of
orientation, those who cater for this desire are most honoured among humans
and enjoy the highest prestige. Transcendental philosophers belong to these
groups. They aim at providing people with an orientation which can give them
the assurance of something death-defying and eternal. As physicists reveal
eternal laws behind the observable flux of natural events, so meta-physicists

Scientific Establishmentr

13

try to show the unchanging conditions behind people's changing experiences,

the eternal forms of reasoning behind their variable reasoning, and universal
rules of knowledge and of science behind their advancing scientific knowledge. Whatever the cognitive value of this bent for the eternal may be, appealing as it does to people's wishes for symbols of eternity, it rates high in social
esteem.
Moreover, by means of this process-reducing code, transcendental philosophers not orily seek to discover something immobile and immutable behind
all that is moving and changing; they also set out on this venture in the belief
that what they discover in this manner as immobile and immutable is fundamental and basic to the processes reduced in this way. The results of an
operation of change-reducing abstraction is thus adorned with praisewords
which gives high marks to that which is immobile and low marks to that
which is mobile. They imply that the latter rests on the former as on its
fundament and that it, thus, in some undeclared way, provides the explanation for it. Many of our most respected research strategies, and most certainly
those of transcendental philosophy, have their raison d'etre in this value
scheme and its ontological premisses. Epistemologically, not only are timeless
and immutable conditions perceived as fundamental to all knowledge and as
basic to reason and experience, but, ontologically too, the moving world is
seen to rest on a non-moving fundament. Knowledge of that which is devoid
of all movements which can be repreSented by timeless abstractions, appears
not only as more appropriate, but also as more profound than knowledge of
change-contlflua themselves and of their immanent sequential order.
The assumptions underlying the epistemological use of this value scheme
and its ontological counterpart are reminiscent of an assumption one encounters in Kant's presentation of his pioneering hypothesis about the genesis of
stars which he published five decades before Laplace. His nebular hypothesis
was the first coherent attempt to present a model of the process during which
the many sunlike stars of a galaxy might have formed themselves. At this
stage of his life Kant, much in advance of his time, was concerned with the
problem of a formative process itself, not of the eternals behind it. However,
at the time, such a process model was still too novel and too daring to be left,
as it were, hanging in the air. Kant felt that the whirling mass of stars required
a non-moving support. He, therefore, postulated the existence of a body
larger than any other and, thus of greater force of gravity, towards which all

14

Norbert Elias

the other star systems gravitated. He postulated, in other words, as frame of

reference of the moving universe an absolute centre of gravity. Similar views
were still expressed in the 19th century .
It is as if people had difficulties in imagining free-floating movements without reference to a non-moving resting place. Although not without predecessors, it was mainly people like Einstein and Hubble who, in the 20th century,
sufficiently jolted people's imagination, to make them accept, at least with
regard to the physical level, the possibility of a universe without centre where
everything was on the move and nothing was ever absolutely at rest. Space
travels, a little later, powerfully reinforced this change.
Yet, the difficulty of such a conception is understandable. Prior to the
20th century people's immediate and dominant experience on earth was that
of movements which eventually caIlle to rest on something non-moving.
their own movements needed the support of the firm earth. As expressions
such as "fundamental" and "basic" show, their code of thinking was not unrelated to these experiences. They combined with people's wish for something
absolutely unchanging as counterfoil to their changeable and impermanent
lives, so as to give this code of thinking, this scheme of values the position of
unchallenged dominance which they have held for a long time.
The irony of the matter is that theories of knowledge and of science,
which abide by this code of thinking and values, will prove to be very impermanent. The implication of this search for specific immutables built into
the mental frame of a person without learning and prior to all experiences,
are obvious. Such an assumption means that human cognitive or experiential
capacity is, by nature, limited. It means that people, whatever the object of
their search for knowledge, are forced to experience them, to perceive them,
and to reason about them in rigidly preordained and innate ways. In that case
they are not capable of freely attuning their forms of thinking and their
symbolic representations to the world which they set out to discover, where
they have to live and to act and where, in order to act more adequately
according to their needs, they can search for more adequate means of orientation. Transcendental philosophy implies that this search is in vain (5). Its representatives promising to lead people to the death-defying fundamentals of their
quest for knowledge, only lead them into an impass, namely to the imaginary
barrier, supposedly set up by a natural pre-disposition between a person and
its world. The very heart of transcendental metaphysics, the "transcendental

Scientific Establishments

15

reduction", a reduction of that which is changeable to something unchangeable and eternal, reveals itself on closer inspection as a philosophical trap.
If such unchanging forms of people's intellectual activities really existed, they
would form part of their natural equipment. Their existence would cast
doubt on one's ability to advance the congruence of one's symbolic representations with the "objects", the relationships they represent. One thus gets a
clearer view of the basic paradox of. transcendental metaphysics. Idealistic in
appearance, it is naturalistic in fact. Kant still could, in all innocence, identify
nature and reason, for the concept of human nature, in his time, was not as
firmly linked as it is today to the research programme of natural sciences.
Biology in particular was not as advanced as it is today. In our time statements
about aspects of humans which are unchanging, common to all and part of
their nature, can only be understood as something rooted in one's biological
constitution. A philosophical quest for immutable and universal conditions
of human experience or knowledge, today, inevitably has a biologistic ring.
The philosophers' dilemma, as one may see, is that they cannot easily 'revise
their basic aim without stepping outside of what they themselves regard, and
what is regarded by others, as specifically philosophical.
There is an obvious way of escape from the impass where, for centuries,
transcendental philosophers have found themselves trapped. That way,
however, is closed to them. They cannot use it without losing their identity.
They are like people enclosed in a room from which they try to escape.
They try to unlock the windows, but the windows resist. They climb up
the chimney, but the chimney is blocked. Yet, the door is not locked; it is
open all the time. If they only knew it, they could easily leave the room. But
they cannot open the door, because to do so would disagree with the rules of
the game which they as philosophers have set themselves. They cannot open
the door, because that would not be philosophical.
The scientific mode of acquiring knowledge as a philosophers' objective
can serve as an illustration. Science is -a fairly recent stage in the development
of human knowledge. To attempt the construction of a science theory by
abstracting from so short a period in the development of human knowledge
what pretend to be immutable universals of the human intellect, is an
interesting intellectual game - but not more. Inevitably it leads into the
solipsistic impass. If one implies that rules of scientific rationality or of
experience have been implanted into human independently of all experience,

16

Norbert Elias

one cannot escape the conclusion that they form a barrier between "subject"
and "object" and limit once and for all human adaptive capacity - the
capacity for developing perceptions and observations of detail in the light of
new integrating concepts and theories, and integrating concepts and theories
in the light of new perceptions and observations of details.
One might be less sure of that possibility if past developments, including
the take-off into science itself - first of the physical, then of the biological,
and now of the human sciences - would not provide telling examples of
such double-harness advances in the development of conceptual symbols
and experiences, of theoretical syntheses and explorations of details. So
far, no limits of these developments are in sight. There are only manmade blockages, among them those produced by process-reducing theories
of knowledge and of science. No adequate orientation for those engaged in
scientific work can be expected from any science-theory which does not
account for the continued advance which sciences can make over long periods
and which is, after all, the centre-piece of all scientific work. In fact, the
difference between a sociological and a philosophical theory of science can
be summed up by saying that the latter is geared to untestable conceptual
symbols beyond time such as "eternal truth", while the former is focussed on
testable concepts such as "advance in knowledge", whose frame of reference
is a diachronic long-term process, a long chain of generations where specific
advances made by earlier generations are the condition of those of later
generations. The model of a diachronic sequential order thus replaces, as the
heart of a science-theory, the abstraction of timeless law-like universals which,
at the most, become auxiliaries. Nor is it possible to confine attention to
advances in knowledge made during the last three or four hundred years, i.e.,
to scientific advances. One can go further and say: No adequate orientation
for those engaged in scientific work can be expected from any science-theory
which does not explain how and why the production of human knowledge
from being non-, pre- and proto-scientific has become scientific over a widening range of problem-fields and why it has taken the form of a production by
an increasing number of specialized science establishments. Only by working
out theoretical models of this long knowledge process and, as part of it, of
the long take-off into science with the following multiplication of science
branches can one escape from the solipsistic trap. But in order to do that,
one has to give up the fictitious assumption that any scientist goes about

Scientific Establishments

17

his business starting from scratch. One has to consider as relevant to one's
theory the demonstrable fact that all scientists start their work with attempts
to solve problems resulting from previous advances in knowledge, from a
structured sequence of earlier problems and problem solutions.
If knowledge, non-scientific and scientific, is seen as an intergenerational
process, if every individual problem-solver is seen to stand on the shoulders
of others, instead of being considered as an individual in a vacuum, then the
door of the philosophers' closed room can be easily opened. Then it becomes
possible to approach the central problem, which has occupied metaphysical
philosophers at least since Descartes, in a different manner - the problem
of how human symbolic representations "within" can fit, or can be made to
fit better, events in the world "without". In that case this problem loses
the character of an infernal trap where human beings run around in circles from
"subjectivism" to "objectivism" and back; it ceases to be a matter of philosophical speculation and becomes a matter of theoretical and empirical
research in the field of the human i.e. social sciences.
3. Philosophy and the Problem of Time
Reference to one example may be enough, briefly to illustrate the change.
If one studies the problem of time in the philosophical manner, it remains
insoluble. Opinions have changed from an objectivist view of time (Newton)
to a subjectivist (Leibniz, Kant) and back. By and large, the philosophical
subjectivists appear to have gained the upper hand. Accordingly, time is
widely regarded as a form of apperception a priori, as a mode of experience
laid on in the human intellect prior to all experience and thus unlearned. One
cannot imagine, in other words, that oneself could ever experience events
without reference to a tightly knit framework of time measurement, such as
hours of the day or the sequence of calendar years. In actual fact, however,
this specific time experience is bound up with a stage of social development
at which societies could not function without a differentiated and firmly
institutionalized framework of time measurements - a society with long
interdependence-chains binding the social functions of many thousands of
people to each other and thus requiring very close coordination of their
activities in terms of time. It is well known that people ofless differentiated
societies neither possess nor need timing devices of our kind. Most of them,
as long as they are left to their own resources, lack time concepts at the high

18

Norbert Elias

level of generality and synthesis to which we are used, among them the high
level concept of time itself. Among the greatest difficulties one encounters, if
one ever attempts to fit members of less differentiated societies into the
temporal rhythm of an industrial society, is the former's incomprehension of
the social demands made on them in terms of punctuality and the impersonal
time regulation of their watches and clocks. Members of simpler societies are
often not able to adjust to the regulatory demands which time makes, as a
matter of course on members of more complex societies. The latter in tum
are hardly aware how profoundly their own personality is structured in
accordance with the regulatory demands of their society's institutionalized
timing devices. The self-regulation in terms of the regulatory demands of their
social time becomes an almost inescapable habit, a kind of second nature.
That is the reason why many people structured in this manner, transcendental
philosophers among them, are inclined to think that their own time experience
must be, in fact, part of human nature, an unlearned property of humanity
and independent of all experience.
A theory of civilizing processes is necessary to gain enough distance in
relation to one's own personality structure to recognize its connection with
the structure of the society where one grew up, and to perceive both, not
simply as something which exists, but as something which, in the course of a
lengthy development, has become what it is. A theory of the development of
the kind of concept represented today by the concept "time" (6) and of the
wider social development related to it, forms part of the theory of civilizing
processes. In Simpler societies, no differentiated and ubiquitous self-regulation
in terms of time is required. The early forms of what we call timing are
discontinuous and intermittent, mainly passive in character, and geared to
pointlike events, such as the appearance of the sun at a particular spot in the
sky, and to events which one can directly observe here and now, such as the
coming of a new moon. The ancestral forerunners of our concept "time" are
cncepts representing a low-level synthesis such as: "We start sowing when this
kind of bird appears." They are characteristic of pre-state societies such
as bands or self-ruling villages. Most early-state societies develop unifying
timing devices and corresponding time concepts at a higher level of synthesis.
As social units increase in size, differentiation, complexity and levels of
integration, timing devices, and concepts, change accordingly. In the more
developed industrial nation states, timing devices have the character of a

Scientific Establishments

19

highly differentiated ever-present and inescapable regulatory time grid. Time

concepts represent a very high level synthesis. They are one of the symbolic
representations of the functional interdependence of millions of people all
exposed to what they feel to be "the pressure of social time" which, in reality,
is the pressure of so many functionally interdependent people upon each
other; they feel the pressure of biological time, which is in reality the pressure
of themselves getting older, and their knowledge of it.
A theoretical model of the long-term development leading from the earlier
to our forms of time concepts, thus shows, as one of its core aspects, the
patterns of an advancing synthesis. Many other concepts which are philosophically regarded as high-level abstractions or unchanging human universals,
concepts such as "cause", "substance", "natural laws" , or "nature" show a
similar developmental pattern. The direction of the development to which
the concept of an emerging synthesis refers, can be summed up fairly easily:
experienced events which people at an earlier stage cannot connect or can only
connect with each other in the form of collective fantasies, and for which
they, theroefre, have no connecting concepts or only magical concepts, can
be connected by people at a later stage; and these known connections are then
represented in their communication with each other by conceptual symbols
which embody a more comprehensive, more realistic testable synthesis.
Advancing synthesis, as represented by concepts at different stages of
social development, forms part of a process of intergenerational learning
which depends on several closely interwoven and inseparable evolutionary
strands. Examples are: a social evolution from small scale, functionally less
differentiated, social units with short interdependence chains to very large,
functionally more differentiated and more populous social units with very
long and closely knit interdependence chains, a social evolution of the means
of orientation leading from a relatively small fund of knowledge, which may
be within its small compass more highly differentiated compared with ours,
to a very much larger and, all in all, vastly more differentiated fund of knowledge, a development from concepts representing a small-scale, highly groupcentred, personalized and affective synthesis to concepts representing a largescale, more object-centred, more impersonal, and more detached synthesis.
Time is a concept of the latter type. To call it "abstraction" or "generalization" is not entirely satisfactory. From what is the concept of "time"
supposed to be abstracted and what particular cases are, in the form of this

20

Norbert Elias

concept, supposed to be generalized? One can see that a theoretical enquiry

into the problem of concept formation is long overdue. It is one step in
that direction to make it clear that the human ''subject'' of the formation of
concepts is not, as traditional theories of knowledge often seem to imply, an
isolated individual abstracting on its own common properties from a number
of "concrete" objects by comparing them, but an intergenerational process
formed by many individuals - an unplanned process whose direction is
functionally related to that of societies providing conditions of continuity
for this knowledge process.
One possible direction is that of an expanding synthesis. Our concept of
time is a symbol of this kind of synthesis on a very high synthesis level. The
small-scale, discontinuous timing of small groups which is centred on social
activities such as hunting, food collecting, or sowing and harvesting, and for
which at an early stage not even continuous movements, but fixed pointlike
positions serve as yardstick, has given way to a minutely differentiated and
unified time grid, reaching from seconds and minutes to calendar months and
years, as a tightly-knit framework of all social activities, now, in some cases,
spanning the whole globe. Moreover, this social time-grid has expanded; it
now includes the evolution of stars, of galaxies, and, indeed, of the whole
physical universe. The social time unit of a year, perhaps first conceived and
institutionalized by Egyptian priest, has been expanded into that of a lightyear, a human-made unit referring to the distance light travels during a
socially agreed form of the earth year. By means of this unified and vastly
expanded time grid, one cannot only coordinate social activities, such as the
departure and arrival of aeroplanes over the whole earth, one can also time
travels to the moon, to other planets and, sooner or later, perhaps to other
galaxies. One can determine with a fairly high degree of accuracy the number
of years it took humans of our type to develop from other types, or mammals
and birds from reptilian ancestors and, somewhat less accurately, living from
non-living things. Present-day timing and the more or less unified time-grid
of human-made clocks and calenders reflect the contemporary capacity to
connect aspects of "people in society" and "nature" in a more extensive,
more object-adequate, and more differentiated way than those of earlier generations. They reflect, in a word, the expanding synthesis characteristic of
humanity's long knowledge process as we de facto observe it (which does not
mean it must necessarily and will always go in that direction).

Scientific Establishments

21

Broadly speaking, one can say that the function of timing is that of
coordinating and connecting the one-after-another aspects of two or more
continuous sequences of change, one of which usually serves as a reference
sequence and as a means of determining the relative positions or intervals
between them within the sequential changes of the other or others. Thus,
a recurrent sequence of social activities may be timed in terms of recurrent
positions of the roving sun on his way through the skies; the change continuum of a living person may be timed in terms of the sequence of years in
the reign of kings; and the one-after-another sequence of a person's work and
leisure may be timed in terms of the little mechanical change-continua which
we call watches.
However, at an earlier stage of social development, people have few means
of coordinating and connecting change-continua in this manner and, by and
large, fewer needs to do so. Hence, the actual social practices of timing and
the corresponding concepts change over time - they change in the specific
direction I have outlined before. A small-scale, point-like, relatively loose and
intermittent ad hoc synthesis transforms itself into a highly differentiated
continuous, permanent and very wide ranging synthesis, embracing what we
call "nature" as well as "society" and represented by our concept of time.
The difference between a philosophical and a sociological theory of time
illuminates the difference between a philosophical and a sociological theory
of knowledge. It also indicates why it is pOSsible, by means of a sociological
approach, to fmd a testable solution of problems which philosophers, for
centuries, have tried to solve in vain. In transcendental philosophical terms
time is made to appear, explicitly or not, as representative of a synthesis a
priori. That is a speculative hypothesis not grounded on evidence and easy
to refute. According to it, one is compelled by nature to perform a synthesis
of events, i.e., to connect them, in terms of time. As a rule, philosophers do
not explain why human nature should have evolved in that specific way. The
mystery which they set out to solve becomes only heightened by the solution
they offer. The sociological approach, sketched out above, provides a testable
answer to the apparent mystery. It requires a higher degree of self-distanciation - it only makes sense if one is able to emancipate oneself from the
assumption that one's own way of perceiving events in terms of time must
be a universal way of all humans. Once one is able to free oneself from this
illusion, time ceases to be a riddle. The great mass of material available from

22

Norbert Elias

earlier ages as well as from contemporary societies, shows a clear developmental pattern in the direction I have briefly outlined above. The naturalistic
metaphysical hypothesis of a synthesis a priori is replaced by the testable
sociological hypothesis of time as representative of a developing conceptual
synthesis. The former is beyond the reach of evidence unless one discovers
biological evidence in its favour. There is ample evidence for the latter.
Transcendental philosophers are apt to argue that a sociological theory
can have no relevance to their own, to a philosophical problem. That is a
good example of an academic establishment's defence of its own autonomy.
The insistence on the necessity to use arguments, which are "philosophically
relevant", deflects attention from the question of whether or not an argument
is intrinsically relevant to the solution of the problem under discussion. If
it is relevant to that solution, it is quite irrelevant whether or not it is philosophically relevant, especially if the relevance of philosophy itself for any
theory of knowledge is in doubt. The demand for philosophical relevance thus
illustrates the tendency of institutionalized academic establishments to claim
a monopoly in the production of knowledge in their field and to fashion the
knowledge produced accordingly. Establishments of this kind form power- and
status-hierarchies with each other. Within these hierarchies establishments of
different disciplines hold different positions - at any given time some rank
higher, some lower. The fact is known by most of the people concerned. But
one rarely refers to these status- and power-differential in these terms; they
are discussed at a lower level of synthesis with a high degree of involvement
and mostly in personal terms: as tensions between professors, departments or,
more rarely, disciplines. They are seldom conceptualized at a higher level of
synthesis. Sociologically the power- and status- differentials between different
scientific establishments and their reasons still are a largely unexplored field.
So are the tensions and struggles engendered by them. Compared with many
other tensions and struggles, power- and status- rivalries between academic
establishments are usually fought out in a more civilized manner. As far as is
known, no-one has ever physically mained or killed in their course. Nevertheless, these interdisciplinary establishment struggles, often long lasting and
unresolved, can be fierce; they can be highly injurious to lower-status establishments, to the outsiders, and to the defeated. Though they do not normally
lead to physical injury, mental injuries are frequent and often severe.
At stake in these rivalries are the relative autonomy and independence

Scientific Establishments

23

of establishments as well as their power-chances, their status, and their

prestige. What these chances are need not be more explicitly discussed in
this context. As a rule, higher ranking and more powerful disciplines can
impose upon those who rank lower their own method and categories of
thinking as a model to be imitated. They can effectively stigmatize deviants,
who do not or cannot comply with their' prescriptions, as "non-scientific",
"not philosophical", "not professional", and in many other ways. They can
use fact-related terms simply as praise words as indicators of their own claim
to a high position in the status hierarchy and of the lower position of other
disciplines. The term "basic" is an example. The more "basic" a diScipline can
effectively claim to be in relation to others, the higher is usually its prestige,
and the greater its relative power.
Establishments of metaphysical and transcendental philosophy rank
relatively high in the present academic status hierarchy. Their representatives
claim for their field to be "basic" to all others. The rebuttal of an outsider'S
argument as "not philosophically relevant" can therefore serve as reproof
of lower status persons by a person of higher status. It is an example of a
relatively high ranking establishment's attack and defence strategy. It means:
"We need not discuss the substance of your argument. If you mean to enter
the philosophers' territory, you have to dress up and to speak like a philosopher. Otherwise beware - trespassers will be prosecuted (or at least
stigmatized)."
There are many stigmatizing strategies in the metaphysicists'armoury.
Weapons of an old tradition, ready made, they save the need for new appraisals. In the metaphysicists' world, asjt seems, there can be nothing really
new. Though the actors change, there is little change in the roles which they
play in relation to each other. There are always subjectivists and objectivists
or those who tryout intermediary positions and compromises. They appear
in different guises - as rationalists and empiricists, as apriorists and positvists,
as phenomenalists and realists, as deductivists and inductivists, and as others
like them. There is no end to it, nothing can ever reconcile the polar 'views
and solve the problems arising from the fictitious assumption of an existential
gulf between human beings and the world they set out to discover and
to control - the world of which they themselves form part. This assumption is the stumbling block. Nothing new, no advances in the theory of
knowledge and of sciences are possible as long as the assumption of an

24

Norbert Elias

ontological gulf between "subject" and "object", explicitly or not, remains

the basis of these theories.
It also forms the frame of reference of the quest for universals. That, too,
is an heirloom of classical metaphysics. The objects exist beyond the reach of
the marble statue on the other bank of the river - how they are "an sich",
unftltered through the subject's perception, experience, intuition, reason
or language, no one can say. The filter seems irremovable. That the imaginary
filter, if it really existed, would also apply to this view itself, does not appear
to perturb transcendentalists or only very little. They have got used to the
solipsistic doubt. Undeterred, they go on looking behind all changes for
unchanging universals in the only place where, according to the scenario of
the marble statue, they can look for it, within the statue itself. There, in the
constitution of the subject, in the unchanging condition of humans, they try
to discover the unchanging universals of science or of knowledge generally.
Already Kant was unable to escape the conclusion that objectivity of judgement was based on the unvarying conditions of the subject, i.e., of "pure
reason". Nowadays transcendental philosophers are perhaps more inclined to
substitute for reason as symbols of eternity ''logic'' or ''language'', but the
scenario is the same.
That is not to cast doubt on the integrity of metaphysical and transcendental philosophers. They are honestly engaged in enquiries which they themselves believe to have a high cognitive value. They themselves are caught up in
the snares of their tradition and get used to living with them. They have made
their own the value scheme, according to which the clue to the fundamental
questions of knowledge and of science can only be found if one leaves behind,
as transient encrustations,all that is changeable in knowledge and in science,
and penetrates within the "subject" - the maker of science, the knower - to
the immutable conditions, to a layer of unalterable universals and eternal laws
of knowledge, just as physicists penetrate through changing events to the laws
of nature.
Having made their own this value scheme, transcendental philosophers
cannot help considering enquiries into diachronic sequences such as the
development of concepts, of forms of thinking, or of knowledge generally, as
enquiries of lesser cognitive value, less "basic" than their own. Thus, when
told that they cannot possibly escape from the snares and pitfalls of the
transcendental tradition as long as they fail to take into account the long

Scientific Establishments

25

development of knowledge in the course of which concepts such as "cause"

or "time" and "space" have developed to their present meaning, they cannot
respond .. Such a recasting of the problem of knowledge would mean to them
stepping down from a highly valued approach to a lower status approach. It
would imply the recasting of their whole scenario. The philosophical subject
of knowledge,the marble statue which thinks for itself, untaught by others,
would have to be abandoned for something that is alien to them - for a chain
of human generations who can learn from each other, for an intergenerational
process in the course of which knowledge can grow and decay. It would
mean, in short, a change of their image of human beings as well as their value
scheme. Instead of searching for timeless universals they would have to direct
their search to a development, to a directional order of changes in the sequence of time, p'roduced and kept going by a long line of generations. Instead
of starting from the lonely thinker they would have to start from groups of
humans in the sequence of time.
4. Towards a Theory of Scientific Establishments
This point throws light not only on characteristics of the philosophers'
establishment, but beyond it on the relationship between different academic
establishments and their departments generally. According to a widespread
view, the existence of a variety of specialized academic departments simply
represents the division of labour which is necessary in the exploration of this
world because the target of scientific research, the universe with its manifold
differentiations and integration levels, is itself too variegated and complex
for a single group of scientists to explore in its entirety. Up to a point, that
is certainly correct. However, the relationship between different academic
specialisms has not quite the character which one might expect if it were
determined by the instrumental need for a division of labour alone. If that
were the case, one would expect efforts at an interdisciplinary co-operation
to be easy; the special work of different departments of knowledge would
neatly dovetail into each other like the pieces of a puzzle. In reality that is
rarely the case. The departments of scientific knowledge, as constituted today,
have some of the characteristics of sovereign states. Their relations with each
other resemble, in some respects, interstate relations. Like states, some departments of knowledge, some disCiplines rank higher than others and their establishments compared with other are more powerful. Like state establishments

26

Norbert Elias

those of scientific disciplines and their departments differ in their traditional

ideology and their scheme of values. These beliefs, these value sehemes deeply
permeate the knowledge they produce. That is why the pieces of the puzzle
do not easily fit together - why interdisciplinary collaboration, at present,
is exceedingly difficult and almost impossible in many cases. The quest for
knowledge, in other words, is greatly affected by the social organization set
up for it.
This organiZation is to some extent determined by the fact, that effective
scientific work - both, teaching and research - is not possible without a
fairly high measure of self-rule of those who undertake this work. A relatively
high autonomy is indispensable; for scientists are those specialists whose task
it is to provide their societies with more adequate, more realistic, and extensive means of orientation without which social practices cannot become more
realistic and less fantasy-ridden. Domination of scientific research by external
orders and interests almost inevitably leads to a deformation of research
results and to a disorientation of any practices which may follow from them.
However, the social organization of scientific work in the academic form
has an unplanned dynamics of its own. Two of its most significant features
are, frrstly, an unplanned long-term trend towards increasing specialization
and, secondly, unplanned power- and status- differentials between the various
specialized disciplines. A third connected with these two, is the tendency of
scientific establishments to develop professional ideologies, a kind of scientific
folklore as for instance, an intra-disciplinary ancestor worship, special beliefs
about selected "great men and women" belonging to this discipline and to no
other, or beliefs about the unique value of one's own field of work, compared
with that of others - a folkore which, though perhaps of little cognitive value,
does add to the sense of belongingness, to the pride in their own work of
members of a discipline which, within reason, people may need. However, all
too often these professional ideologies of scientific establishments disguise as
theories, thus becoming responsible for the sterility of research efforts. Even
cases of deliberate falsification of research results for the greater glory of
one's discipline or, may be, of one's own theory, one's own value scheme and,
thus, of oneself, are not unknown.
Needless to say, research always entails the risk of errors and mistakes.
They form an integral part of a scientific process. And although some disciplines, such as physics, appear to have a highly effective built-in organization

Scientific Establishments

27

of self-criticism and self-correction, other disciplines seem to lack these

means. In their case knowledge of little or no cognitive value may be carried
along unchecked from one generation to another. In such cases, interdisciplinary status- and power-rivalries and battles may sometimes have a corrective
function. Generally, however, only members of lower status disciplines are
likely to listen to interdisciplinary criticism; and they, like other outsider
groups, tend to listen only to the criticisms of higher status disciplines. By
using them as a model, and characteristics of theirs as a status symbol, they
try to gain kudos for their own field or perhaps for themselves. In reality,
they achieve little by this kind of mimicry. They inevitably, fall between
two stools and perpetuate the ineffectualness and the lowly status of their
own field. Little it to be gained by attempts at earning kudos through the
imitation of a higher status group.
There are many other examples of status-differentials between scientists
and the pressures engendered by them leading to interdisciplinary miscegenation - to the adoption of methods of research, schemes of values or
forms of thinking of a discipline where they may be appropriate and bear
convincing results, by another discipline perhaps concerned with quite a
different level of integration, where they are less appropriate and thus produce sham-results.
Philosophical theories of knowledge and of science have been instrumental
in fostering this kind of mimicry. Their traditional direction towards timeless
universals has resulted in the fata morgana of a timeless science, an ideal
image which makes it possible to describe characteristics of science in a highly
formalistic manner, but which, as a rule, is nothing more than an idealized
picture of physics. In that way philosophers reinforce and propagate the
physicist's own claim that their own mathematizing type of method and
theory alone can be regarded as scientific. The physicists' claim to a monopoly in science-characteristics, which may have been justified in the 17th and
18th centuries, and perhaps even in the early 19th century, is thus carried
over into the late 20th century where it is no longer justified. As before,
philosophers tend to speak of science in the singular in spite of the multiplication of sciences and the development of several types of science. In that way
philosophical theories of science have, until now, almost completely blocked
the development of theories of sciences which take account of the plurality
of sciences that have evolved since the 18th century. Philosophical theories of

28

Norbert Elias

science, in short, have persistently omitted to take into account such aspects
of scientific work as the process of increasing scientific specialization and the
problems connected with it, among them the problem of interdisciplinary
relationships and their effect on the production and advance of knowledge
itself.
The exclusive fixation of transcendental science theories on the search
for timeless universals and the congruent value scheme has prevented their
representatives from perceiving the relevance for any science theory of
diachronic processes, not only of an individual science but also of the increasingly differentiated and expanding network of sciences. Today, problems
of interdisciplinary relationships in the form of cooperation as in that of
rivalries and conflicts can be regarded as a normal part of scientific development. That they, like some of the other aspects of sciences mentioned before,
are almost completely disregarded by these traditional philosophical theories
is symptomatic of the severely restricted horizon their tradition imposes upon
them. It indicates how remote from the developing sciences themselves these
science theories are, and how large is the discrepancy between their claim to
be basic to all forms of knowledge and their actual cognitive value, though
the idiosyncratic language of philosophers may disguise it.
Even more important, it may disguise that the problem of human knowledge, as framed by transcendental philosophers, admits of no solution. If the
contention were correct that special forms of intellectual activities are built
into every member of the human race prior to experience, independently of
learning, there would be no way in which humans could convince themselves
that the world, as represented by their social symbols, and thus, as experienced by them, can ever be an accurate representation of the world as it is
independently of their experience - except perhaps on Leibniz's assumption
of a pre-established harmony, instituted by God, between people's "internal"
images and the "external" world.
Nor does the opposite view carry greater conviction - the view that man's
symbols representing objects in connection with each other, are simply mirror
images of these objects. "Open the eye and you'll get the right idea", is not
a productive prescription. To break the deadlock one has to acknowledge the
fact that humans learn social symbols of the world from childhood on, which
pattern their experience and which they may, in their lifetime, develop further,
What has been lacking so far, are model of the unique relationship between

Scientific Establishments

29

these human-made symbols with their double function as means of orientation

and of communication, and the world of connected events they represent. The
static and spatial imagery characteristic. of the old philosophical epistemology, the imagery which opposes an "internal" to an "external" world, is quite
inadequate as a means of conceptualizing this relationship. The concept 'table'
is not a projection into a person of the table "outside"; the concept 'cause'
not a projection from within a person on to the world "outside". Concepts are
symbols (means of communication and orientation) which develop over time
in a given society to a given stage of fitness, higher or lower than before, as the
case may be. In the course of this development, the meaning of these symbols
can become more or, alternatively, less adequate to whatever they represent.
The crucial point is that the old transcendental epistemology imputes
severe limits to the malleability and adaptability of human symbolic representations. The basic postulate of transcendental philosophy is the existence of a
number of universal concepts set unalterably into the human mind. On the
other hand, their opponents imply that these concepts can be formed by a
person on her own simply by looking at things. Both are at fault because they
reduce a dynamic relationship without beginning - was there ever a subject
without objects, human beings without a world? - into a static polarity between subject and object, a conception which, in fact, implies that humans
can exist without a world, that the subject enters only a posteriori and accidentally into a relationship with objects. In the last resort it is this process
reduction which makes philosophical epistemology a futile enterprise, a wild
goose chase without end. There are many other examples of process reduction
leading to endless battles between representatives of two static positions. In
pre-Darwinian biology for instance, representatives of two highly regarded
establishments were engaged in a fierce dispute, one led by Cuvier, maintaining that the way of life of an animal determines the structure of its organs,
the other led by Geoffrey Saint-Hilaire, maintaining that the organic structure
of an animal determines its manner of life. There are other examples showing
how and why a reduction to static positions of what can only be understood
as a continuous process of change leaves man with two polar alternatives,
both equally untenable. Few other instances, however, have led to disputes so
long drawn out in time, so wasteful of intellectual energies as the disputes
between subjectivists and objectivists of various shades in the field of philosophical theories of knowledge and of science.

30

Norbert Elias

The scenario changes if one re-frames one's questions in terms of processes

such as that leading from pre-scientific conceptual representations of simpler
people with their limited control of natural events to the, all round, better
fitting scientific symbols of industrial societies which, because they are better
fitting, facilitate a much more extensive control. Processes of this kind lend
support to the assumption of an unlimited adaptability of human symbols.
Philosophers may say we cannot consider such processes because that is not
philosophical; and indeed, one can understand their difficulties. To adopt a
process paradigm means re-casting the central problem of traditional theories
of knowledge and of science; it means casting out the old and adopting a
new paradigm. That, given the present ethos of academic establishments, is
difficult to do. It is a sobering thought that a very great intellectual effort of
learned men and women has been directed for centuries and who knows, may
be directed for more centuries towards the exploration of a problem that is
based on a misconceived assumption and admits of no solution. As a case
study in the self-perpetuating and inertial power of academic establishments,
the example of philosophical epistemology and science theory is instructive.
This should not be understood as a wholesale denial of innate perception
and behaviour tendencies in humans. On present evidence, a good case can be
made for the assumption that certain, rather elementary types of perception
and the reaction tendencies connected with them, are innate and thus prior
to experience, although they undoubtedly grow and mature in continuous
interdependence with experiences. In a relatively pure form one encounters
them above all in young babies. Their reactions to a human face, particularly
to its eye portion, are tendencies of this kind; so is the primary smile which,
under appropriate conditions, may be elicited by an adult's face.
One could point to a good number of other human perception and behaviour tendencies which require no learning, are common to all human
beings and thus innate. But there is a far cry from these elementary unlearned
tendencies in humans to the forms of reasoning, of intuition and of other high
level syntheses which transcendental philosophers claim are prior to human
experience and thus unlearned. All the evidence known at present points in
the opposite direction. As far as can be seen, the assumption of an a priori
character of such syntheses has no cognitive function at all. But it certainly
has ideological functions for those who make this assumption. It serves as
justification of the philosophers' claim to be able to discover fundamentals

Scientific Establishments

31

of all other fields of knowledge. In a discipline whose representatives have

some difficulty in explaining unequivocally the specific nature of their
common problem field, the belief in the existence of a layer of unchanging
universals which, although the condition of all experience is wholly indepen.
dent of it, has a very important function for transcendental philosophers
themselves: it forms a field of studies which is entirely their own. No one has
access to it but philosophers themselves. To explore it requires particular
skills, especially the skill of arguing and reflecting on a very high level of abstraction in a highly specialized language. In societies such as ours, this kind
of skill enjoys great prestige; philosophical establishments of this kind are
therefore able to recruit for their field young men and women of great intellectual aptitude, capable of reflecting on a high level of abstraction without
any need for checking their reflections against observable facts. Thus, the hypothesis of a layer of immutable universals of knowledge and of science helps
to define the special task of metaphysics and transcendental philosophy. Arid
as it is transmitted from one generation to another, the belief in the existence
of this layer becomes self-perpetuating as the core of an establishment's status
and power chances. It is a supreme example of a self-perpetuating fallacy.
Hence, if one opens the door of the philosophers' self-made prison and says
to them: You will be free, you can escape from your solipsistic trap if you
will only consider that no person's knowledge has its beginning in himself;
that each of us, with all his reflections, perceptions, intuitions, or experiences,
stands on the shoulders of others, and that in order to understand the pattern
of these intellectual activities, as they are today, you have to re-trace the long
intergenerational process, in the course of which they have become what they
are, if that is said to them, these philosophers cannot respond. If runs counter
to the whole tenor of their beliefs and to the value scheme they cherish. What
is demanded of them is the re-framing of their problem and its solution in
terms which, in their eyes, have a lower cognitive status than their own. All
that, they may feel, is too high a price to pay for an escape from an intellectual impasse. So they prefer to stay in the homely trap of their insoluble
problem and to carry along from generation to generation the flag of a
tradition which, though it has little intrinsic cognitive value has, as one can
well understand, a high value for their representatives.
Moreover, their tradition offers them ready-made conceptual weapons as
a means of defending their own values against precisely that kind of argument

32

Norbert Elias

which has been put forward here. Take, for instance, the philosophical
distinction between "systematic" and "historical" enquiries. It is another way
of referring to the distinction between enquiries aimed at timeless universals
and others aimed at diachronic sequences of change. In these terms, philosophical epistemology is "systematic" and sociological epistemology, as
understood here, is "historical". On the face of it, this distinction may appear
to be purely fact-related; on closer inspection one can recognize that it
represents a specific scheme of values. It means "only" historical, namely
directed towards the layer of changing events, and thus "not systematic",
not directed towards timeless universals and, therefore, not philosophically
relevant, not participating in the high value of the discovery of eternals. In
other words, this too is an example of an academic establishment trying
to defend the independence of its tradition, its sovereignty, and its interests
by means of a conceptual polarization implicitly reflecting its own axiomatic
hierarchy of values, buttressing its own superior status.
The re-casting of the problem of knowledge suggested here, thus not only
affects the traditional concept of the "subject" of knowledge and the whole
scenario of the thinking statues, it also entails a change in what is regarded
as "basic" in man's quest for knowledge; it entails a change in the balance
between the cognitive value attributed to two distinct aims of this quest,
between the aim of discovering timeless, lawlike universals and that of discovering the structure and explanation of continuous processes of change in
the sequence of time, the equivalent, as one may see, of the transition from a
Newtonian to an Einsteinian approach.
In the age which put its stamp on the secular philosophy tradition of
Western Europe, in the 17th and 18th centuries, philosophers as well as
scientists would have hardly hesitated, had they ever been asked to decide
which of these two aims, in their view, endowed research with a higher
cognitive value. Most probably, they would have said that the discovery of
eternal laws, such as those of Newton, had a very much higher cognitive value
than the chronicling of historical events or discoveries about the development
of humanity. The philosophical distinction between "systematic" and "historical" enquiries carries this view into the 20th century.
However, the development of sciences in the late 19th and the 20th centuries itself casts some doubt upon the traditional assessment of these value
differentials between two aims of research. Thus, in biology a process model
the theory of evolution, has long gained acceptance as a theoretical reference

Scientific Establishments

33

frame of research. The high cognitive value of this diachronic theory of an

unplanned directional long-term process has a bearing on the theory of theories. The quest for biological universals has not disappeared, but the balance
between the two complementary research aims has shifted; the discovery of an
evolutionary order of changes has gained by comparison with the discovery of
unchanging universals. Although, in terms of the philosophical distinction
between systematic and historical types of enquiries, the biological theory of
evolution might have to be classified with the latter type, one would probably
think twice before calling it an "historical" theory. For the term "history" is,
at present, associated with a specific type of writing history, with the type
sanctioned and practised by the historians' establishments. However, their
perspective of diachronic changes in human societies represents a relatively
low level, pointlike synthesis - a much lower level synthesis than that represented, for instance, by models of civilizing or state-formation processes, a
much lower level, too, than that represented by the biological theory of
evolution. But I cannot explore here the science-theoretical significance of
the distinction between different levels of synthesis.
There is, thus, a good reason why one hesitates to call the theoretical model
of a biological evolution an "historical" model. The level of synthesis it represents is much higher history is still largely a-theoretical. Yet the theory of
evolution has certainly the character and functions of a theory, even though
it differs from the law-like type of theory dominant in classical physics.
Philosophical theories of scientific theories usually proceed as if all theories
which deserve to be called scientific had a law-like character. Process theories
still lie beyond their horizon.
Yet if one is attentive enough, one can observe today under one's very
eyes a comparable shift in the balance between law-like and process-theories
in one of the physical sciences itself. In contemporary cosmology, the quest
for law-like and timeless regularities has certainly not disappeared, but the
quest for regularities of long-term processes, for evolutionary changes has
gained ground. For Newton, the movements in the sky were still largely
recurrent movements. The planets moved again and again around the sun
with the same precision as clockwork wheels. The genetic question, above
all that of the genesis of the universe, was partly answered by his belief in
God's creation. The problem of the origin of the sun system in purely physical
terms was not entirely absent from his mind. But even had he wished to

34

Norbert Elias

explore it, at that stage, people had hardly the means to do so. Thus,he was
quite content with discovering recurrent law-like regularities in the movements
of the planets. It would need a closer study of the development of physical
sciences in order to show why it became possible, in the course of the 20th
century, not only to raise, but also to start solving the problem of the evolution of stars, of galaxies and, indeed, of the whole universe. But whatever the
reasons, in that field, too, the recognition has gained ground that law-like
theories alone cannot satisfy one's scientific curiosity. They have their
cognitive value as largely descriptive means of orientation or as rules of thumb
for practical purposes. But on their own they do not provide any explanation
as to how things came to be what they are. Newton's laws and, still more,
Einstein's law-like theories of relativity are indispensable as an aid in the
human search for an answer to the question of how stars, galaxies, or the whole
universe evolved. But the theoretical vehicles of genetic explanations are
evolutionary theories, models of processes in the course of which stars or
galaxies form themselves or decay. Process models embody a different type
of explanation from law-like theories. They seek to explain why one thing
happens after another in a long sequence of directional, though not necessarily irreversible events. It is the demand for this kind of explanation, for
genetic explanations which accounts for the greater weight, which, on balance,
process-theories have gained by comparison with law-like theories in cosmology.
In using this formula, as one may see, I have been careful to avoid giving
the impression of a polarization - the impression that law-like theories
and process theories exclude one another. In many cases, they complement
each other. Their cognitive weight, however, as instruments of research their
weight in relation to each other, has changed and can change quite drastically.
The cognitive value of the quest for animal universals has not entirely disappeared, but it greatly diminished when Darwin presented his genetic theory
of biological evolution, compared with the value attributed to it before.
A similar shift may be expected with regard to the cognitive value of lawlike and of process theories of knowledge, of science and of the human level
generally. At the human level, unchanging universals do not play the same part
as they play at the level of physical nature. Humans can learn not only from
their own individual experience, but by means of an intergenerational transfer
of symbols, also from the experience and knowledge acquired by previous
generations. By means of linguistic and other symbols, human beings can

Scientific Establishments

35

acquire a fund of knowledge based on both the experience of small-scale

details and the labour of conceptual synthesis of antecedent generations. That
fact alone makes it impossible to work out a theory of knowledge without
regard to the long process in the course of which the fund of human knowledge, and as part of it, the fund of conceptual symbols, has developed up to a
given stage. It is as little possible to penetrate from the study of a given stage
in the development of knowledge directly to unchanging human universals as
it is, say, to penetrate alone from the study of wolves to the universals of
mammals or of animal organisms generally. In that case, too, a theoretical
model of evolution, of the diachronic process of change is required in order
to discover what is the same and unchanging in all animals and what is characteristic of a particular stage in the evolution of animals. In the same way, a
model of the evolution of human knowledge. of people's visible and audible
symbols, is needed in order to find out what are the unchanging universals
of human knowledge, of people's intellectual activities generally and what is
characteristic of a particular stage in their intergenerational development. In
fact, the need for an enquiry into their evolutionary credentials is all the
greater because the malleability of human symbols and thus the tempo of
change, in the case of human knowledge, by and large, is so much greater
than the overall tempo of evolutionary change at the level of stars and
galaxies or at that of living organisms. Human knowledge and above all
scientific knowledge, is changeable to an extent and in a manner which has no
counterpart on the physical or the biological level. And its unique features such
as the capacity, mentioned before, for growth and advance, both in an individual's lifetime and in the course of generations, accentuates the difference.
Hence, by comparison with the generic, the genetic question has greater
significance. Models of law-like abstraction and generalization and the quest
for unchanging universals, though they have a high cognitive value in the exploration of the physical level, cannot have the same high value as models of an
evolutionary synthesis, models of genesis, have in the exploration of the
human level, and particularly of human knowledge with its greater variability,
its higher tempo of change and its unique forms of advance and decay.
With the rise of human sciences, transcendental philosophers of science
who by tradition had strong links with phYSical sciences, were confronted
with a task to which they, so far, have hardly faced up. In the exploration
of their subject matter, these sciences have to pay more attention to

36

Norbert Elias

sequences of change, to the diachronic order in which structures become

what they are, than either the physical or the biological sciences. Philosophers
used their not inconsiderable authority and their claim to be able to legislate
for all kind of sciences, in order to reinforce the physicists' claim that their
science should serve as a model for all other sciences. In that way, philosophical theories of science made it difficult for the representatives of social
sciences to come to terms with the distinguishing conditions of their taks and
to develop scientific models of their own.
In away, this situation illustrates what has been said before about the
relationship between scientific establishments of different power and status.
It is an example of scientific establishments with a high status and a relatively
high power ratio prescribing for others of lower rank. There is no room here
to explore the kind of stigmatization which is used in this type of power
relationship. It is often outwardly inoffensive, yet in its effects oppressive.
The philosophers distinction between "systematic" and "historical" enquiries, mentioned before, can serve as example. It represents a very inoffensive way of asserting the superiority of enquiries directed towards the
discovery of timeless generalizations, compared with genetic enquiries into
sequences of change. I have tried to show that this value scheme accounts
for the fixation of philosophical theories of knowledge on a central problem
cast in a manner which admits of no solution. It also accounts for the fact
that philosophical theories of science, today even more than at the beginning
of this century, are unable to come to grips with the multiplication of sciences,
with their increasing differentiation, and with the emergence of different
types of sciences.
To do justice to this development requires a paradigm change. It requires,
for theories of knowledge, too, a shift in the balance between law-like and
process theories. Traditional philosophical epistemology, particularly in its
transcendental form, it seems to me, has come to the end of its road. Without
understanding genesis one cannot understand structure. Without knowing
how and why human knowledge has become scientific; one cannot know how
and why science works. The genetic question, often de-valued by philosophers
and others as historical still remains unanswered. We have yet more fully to
understand the astonishing efflorescence of sciences in our age. We have yet
to explain how and why human beings, over a widening area of their world,
learned to develop their conceptual symbols, whose meaning is knowledge, to

Scientific Establishments

37

such a close fit to the observable connections of events, that they now can, in
some areas, control and manipulate these connections according to their
social aims, with a fair measure of certainty. What is needed, therefore, is a
sociological type of enquiry, capable of working out process models of the
development of knowledge, fitting into, but not reducible to, models of the
long-term development of human societies. In this case, the paradigm change
does not concern one particular discipline. It concerns the whole tree of
knowledge with all its branches. Inevitably, it will get entangled in interdiSciplinary tensions and disputes. It cannot fail to be affected by power and
status differentials between different disciplines. Thus, one cannot hope to
bring about such a shift without including into one's theory of knowledge
and of science the role played by interdisciplinary relationships in scientific
developments and especially by the relationship between different scientific
establishments. Transcendental philosophers often claim that they can
prescribe for sciences generally. Their claim ought to be firmly rejected.
Theirs is an esoteric enterprise of no relevance to the work of social scientists,
and probably not to that of natural scientists either.
5. Scientific Establishments and Control of the Means of Orientation
The central social function of knowledge is that as a means of orientation. As
the individual orientation of every member of a society depends on the means
of orientation available there, groups of people who are able to monopolize
the guardianship, transmission, and development of a society's means of
orientation, hold in their hands very considerable power chances, especially
if the monopoly is centrally organized. The mediaeval Church was a large-scale
organization of this type. The very heart of its high power ratio was its
monopoly of the basic means of orientation, of revealed knowledge in large
parts of Europe. One of the principal conditions of the emergence of a scientific type of knowledge production was the breakdown of the Church's monopoly of the basic means of orientation. At least since the Crusade against
the Albigensians, this monopoly had been carefully guarded against dissenters
partly by persuasion, but increasingly by fire and sword. All attempts to explain the rise of a scientific type of knowledge which do not take into account
the armed monopolization of the European fund of knowledge by the mediaeval Church and do not link it to the partial break-up of this monopoly, are

38

Norbert Elias

bound to fail. If it is taken into account the problem ofthe emergence of a new
type of monopolization of the existing fund of knowledge and of the production of new knowledge by a new type of establishment, by scientific establishments, stands out in fuller relief.
The increasing professionalization of scientific work during the 19th and
20th centuries and its concentration in the hands of highly placed professional
groups, has led to the formation of a type of scientific establishment which in
certain respects differs from that of previous ages. Even though, through
integration into the universities it still stands in the line of succession of the
long tradition of higher learning in Europe, there are a number of differences
which require investigation. How, for instance, did universities transform
themselves from a stage where they were largely Church-controlled clerical
institutions to a stage of - mostly - State-controlled scientific institutions?
How did it come about that the rising sciences became attached to this
particular type of institution, to universities? Perhaps one should not take it
for granted that they did. How did this scientification affect the institutional
development of universities? And how did the development of universities
affect that of the sciences?
An example may help to see this interplay in better perspective. European
universities traditionally had a double function. As part of their mediaeval
heritage, they were, at the same time, agencies of teaching and of enquiry.
Their members were simultaneously engaged in advancing knowledge, in
widening the scope of human orientation and in handing on knowledge,
old and new, to the rising generations. This double function of universities
has not been entirely without influnece upon the development of sciences;
in fact the influence may be greater than the construction of a purely rational
model of science indicates. The role of textbooks is symptomatic of this
double function. On the one hand, textbooks are devices linked to the
teaching function of universities. Yet, on the other hand, they are also
vehicles of the struggle for, or the achievement of, consensus among leading
representatives of a particular academic field or, in other words, of scientific
establishments. Textbooks indicate what the establishment of a particular
academic field or sections of it, at a given time, regard as the essentials of
the received knowledge in that field. Competing textbooks may point to
dissensions and rivalries. Changes in textbooks often, though not always,
indicate that, and when, a scientific innovation at the theoretical or the

Scientific Establishments

39

empirical level has become canonized as part of the standard knowledge

in that field. In fact, a long-term study of the development of scientific
knowledge in terms of the succession and changes of textbooks, with due
regard to their factual use in teaching programmes and to their author's
position, vis-a-vis given establishments, could serve as a much needed supplement to the prevailing mode of studying the developI1J.ent of scientific
knowledge in terms of a sequence of great individuals. It could also open up
an avenue to the investigation of scientific establishments themselves. For
the selection or approval of textbooks for the use of students is one of the
privileges of scientific establishments. Their members can choose between
competing textbooks. They can influence the selection and presentation of
topics in textbooks. In these and other ways, studies in the development of
textbooks, provided a long-term perspective is used, can offer a good deal
of information both on the development of sciences and on that of their
establishments. They provide a link between the transmission of old and the
discovery or reception of new knowledge.
Yet another feature of the development of scientific knowledge which
owes much to the structure of universities, is the competition between
scientists as individuals and as groups. At a given time, every department
and every laboratory has its place within the academic status hierarchy at
a variety of levels - locally at its own university, nationally and, in some
case internationally, among representatives of its own and related fields.
The whole figuration is animated by a continuous competitive struggle for
preservation, avoidance of loss or rise of status and power chances. It is
a controlled form of competition subject to certain, mostly unwritten,
rules and often, if not always, compatible with a measure of cooperation.
Not all these competitive relationships have a direct bearing on scientific
processes, but some do. However, in order to explore their significance for
the development of sciences, a few words have to be said about the nature of
this competition and of the stakes for which scientists compete with each
other.
The strong impact which competition between economic establishments
as a model of competitive relationship generally has on present-day thinking
(whether one approves or disapproves of it) may focus attention on the
economic aspects of competitive relationships of academics and their establishments. Indeed, they are not to be neglected. As scientists - in competition

40

Norbert Elias

with others - advance in their career, their income is likely to rise. However, in
the case of economic establishments, such as those of commercial enterprises
or factories, their standing within the status hierarchy of competitors is, to a
large extent, determined by the size of their income, i.e., by the establishment's
economic success or failure. In the case of academic establishments, inversely,
the economic resources accruing to them largely depend upon the status and
the power resources they can command within the academic network.
These resources are, in a general way, of the same kind as those at the
disposal of other professional establishments. Scientific establishments are
groups of people who collectively are able to exercise a monopolistic control
over resources needed by others. They control, and engage in, the production
of a particular type of knowledge. In many cases they combine this activity
with that of administering a fund of knowledge handed on to them, in their
particular field, from previous generations and with that of controlling the
transmission of that fund, and whatever advances they themselves have
contributed, to the following generations. By virtue of their monopolistic
control of an existing fund of knowledge and of the skills needed for developing it, for producing new knowledge, they can exclude others from access
to these resources or admit them to their use selectively. That is to say, they
are establishments only in so far as there are groups of not-established people,
. of outsiders, who need the resources monopolized by them and who depend
on them for access to, or use of, them. However, the need is never, or only in
marginal cases, entirely one-sided. Those who are outsiders, in relation to
a given establishment, as a rule, have on their part resources needed by the
establishments' members (which does not imply that the outsider groups
have a centralized or even a monopolistic control over these resources).
Established and outsiders, in other words; have specific functions for each
other. No
relationship is likely to maintain itself for long
without some reciprocity of dependence. However, the dependence of the
established on the outsiders is smaller and, as a rule, very much smaller, than
that of the outsiders on the established. Members of an establishment usually
are very careful to maintain and, if possible, to increase the high dependence
ratio of their outsider groups and thus the power differentials between these
and themselves. If this ratio diminishes, the power surplus of the established
group vanishes and the group itself sooner or later ceases to have the character
of an establishment, as indeed happens again and again. Thus established and

Scientific Establishments

41

outsider groups together form a highly variable figuration with an uneven

balance of power relationship as its main axis.
Scientific establishments derive their power surplus from the monopolization of knowledge of a particular type, of scientific knowledge. Yet, at
present neither the nature nor the distinguishing characteristics of scientific,
as compared with non "Scientific knowledge, are particularly clear. A reminder
of a few basic facts, therefore, may be of help.
What we call knowledge is a nexus of human-made symbols which serve at
the same time as a means of orientation and of communication. A fund of
these human-made symbols has to be acquired by all human beings as they
grow up through learning from other humans, usually, of a particular group.
Without acquiring knowledge in that manner, children of the species homo
sapiens lack the essential characteristics of human beings. They cannot
orientate themselves in their world and cannot communicate with other
human beings without acquiring a fund of symbolic representations from
their elders. In fact, without learning social symbols - without knowledge human being remains helpless and cannot survive.
Subhuman organisms are able, for the satisfaction of their elementary
needs within a multiform world, to rely to a very much greater extent than
humans on inborn directives attuned to a specific setting of that world. As
long as that setting does not change too much, a species of living things has
a good chance of survival in its existing form. However, inborn steering
mechanisms have specific disadvantages when regarded simply as a technical
device for fitting a species of organisms into its niche within the universe.
They are extremely rigid and severely limit the adaptability of the conduct
of organisms to major changes in their surroundings. Biological equipment
which enables every member of a species to acquire individually through
learning from elders, means of orientation as steering devices of conduct,
represents an evolutionary advance which greatly enhances the survival
chances of a species (unless its members use it blindly for its selfdestruction.)
In fact, it has been instrumental in the rise of humans to a dominant position
among their fellow creatures. Organisms equipped in this manner have a
chance of steering their conduct much more flexibly, of adjusting it to a
much greater variety of changing conditions and of adjusting it more fmely,
than organisms which are wholly or mainly dependent on inborn automatic
steering devices attuned once and for all to biologically pre.determined reo
leasers (or symbols).

42

Norbert Elias

Moreover, greater dependence on learned rather than on un-learned

symbols has led to some unique features of human as compared with nonhuman societies. One of them is the ability of human societies to undergo
enormous changes without any concomitant evolutionary changes, i.e., while
the biological equipment remains unchanged. Additionally, the members of
any given generation of human are able to orientate themselves and to steer
their conduct not merely in the light of experiences and reflections accumulated within the short span of their own lives, but also with the help of
symbols, which embody experiences and reflections of a long line of ancestors.
The fact itself is known well enough. But the recognition of its significance
for people's self-image - and thus for the image of a hUman being underlying
philosophy, science theory and most of the human sciences, is still largely
blocked. The notion of a human being as a wholly autonomous entity, as a
homo clausus whose "inside" is hermetically sealed from the "external"
world bars the way. Indeed, many aspects of traditional philosophy and
science theory would fall by the wayside, if intergenerational processes of
concept formation were allowed to enter their makers' fields of vision.
An old cat may learn quite a bit from experience in its lifetime. But it
does not possess the biological equipment for making symbols in which, as
it were, its own experience and that of earlier generations of cats are deposited, thus enabling later generations of cats to fashion their perception
and their conduct in the light of knowledge evolved by a long line of ancestors. Humans, on the other hand, are biologically equipped for the reproduction and learning of symbols of this kind. They possess a biological potential
for the production and reception of symbols such as words which are not
biologically fixated, which, like these groups themselves, are changeable and
which, in relation to that which they symbolize, can be improved and extended. Hence, it is possible for later generations to possess more extensive
and better fitting means of orientation than earlier generations. One has to
add that a human being who is born with fewer innate directives of conduct
than other higher organisms, is also more helpless than other organisms if
he or she does not acquire, through learning, any lingo-conceptual means of
orientation patterning perception, cognition, and conduct.
Knowledge acquired SOcially, in other words, is as indispensable to humans as food, and means of orientation as indispensable as means of production. Any attempt at attributing to one of them priority over the other,

Scientific Establishments

43

of declaring the development of the means of production as "basic" and

that of the means of orientation as a "superstructure" or inversely the latter,
as basic and the former as superstructure, leads into an endless and utterly
sterile chicken-and-egg-dispute.
The significance of a monopolization of the means of production as a
power-resource has been widely recognized, though the detailed exploration
of the sociological problems that arise here lags behind. That the monopolization of the means of orientation, too, is one of the major power resources of
human groups, has perhaps not yet found the attention it deserves. Scientific
establishments are based on a monopoly of this kind. However, in accordance
with the nature of their work they are among the least centralized and unified
types of establishment. As monopolists of means of orientation they are latecomers. Pre-scientific means of orientation lend themselves to a much more thorough monopolization. The greatest and most illustrious predecessors of scientific were priestly establishments. They show, in the form of churches, how great
people's need for means of orientation is and how great accordingly are the
power chances of groups who can satisfy and monopolize these requirements.
I have already referred to the fact that scientific establishments derive
their high power ratio from the monopolization of a particular type of
knowledge. They are, in other words, groups of specialists, whose social
function it is to administer a specific fund of symbolic representations which
can serve people as a means of orientation. This fund has been handed down
to them by previous generations and they set out to develop it further in a
variety of ways, e.g., by extending the range of knowledge within their own
field, by developing concepts (symbolic representations) for a nexus of
observables previously not within reach of concepts, by discovering and
conceptualizing connections between events which previously could not be
connected, by developing concepts which fit the observables better than the
existing ones, and in many other ways.
A degree of autonomy is characteristic of all specialisms. But in many
other cases, the limitations of the autonomy of a specialized occupation and
the dependence of each group of specialists on others is so obvious, the
reciprocal balance of dependencies so firmly institutionalized that the groups
concerned can maintain a measure of autonomy while at the same time
cooperating with interdependent groups of specialists according to a set code.
At universities this is rarely the case. Cooperation between different groups

44

Norbert Elias

of scientific specialists (interdisciplinary cooperation) is still the exception

rather than the rule, and a code of interdisciplinary cooperation hardly exists.
Even when interdepartmental cooperation occurs fairly regularly, as for
example in the case of representatives of different physical specialisms, one
can usually observe hidden or open status struggles and contradictory strivings
for autonomy and dominance. Elsewhere, especially at the level of the human
sciences, the fortress walls between different specialisms are often still so
high, and the status insecurity of each group behind its wall is still so great,
that most specialized establishments concerned with scientific exploration
at this level prefer to stay at home and avoid the risks that lie ahead in
the great problem areas between the closed fortresses of the established
disciplines.
The bulk of economists care little about history or political science, let
alone sociology and vice versa. There are, indeed, hybrids comparable to
bio-chemistry, such as economic or social history, but far from forming a
bridge between. history on the one hand and economics or sociology on the
other hand, they develop, in accordance with the normal pattern of a university department, into a scientific specialism of their own they evolve their
own standards, conventions and rituals and try to attain the greatest possible
autonomy in relation to all neighbouring fields. Cooperation between, say
historians and sociologists or sociologists and Departments of French, English,
German literature, or of linguistiCS is still tenuous and rather exceptional,
even though many research problems, very obviously, require it. As for the
relationship between the main divisions of science, between physical, biological and social or human sciences, the physicists' advance into biology is fairly
obvious and, so far, rather successful. For the rest, apartheid prevails in its
most extreme form. Some sociologists do indeed try to scale the fortress wall
of physicists' establishments, for the need for a sociological approach to
physical and other sciences, replacing the anachronistic Science-philosophical
approach, is obvious. But almost all these enterprises encounter the difficulties
inherent in any attempt made by representatives of a "lower status science"
to seek the cooperation of a "higher status establishment".
Thus, by and large, the striving for complete autonomy of one's own
discipline and, if possible, for domination of other disciplines within the
"groves of Academe" still outweighs' by far the capacity for systematic cooperation. Needless to say, the striving for absolute autonomy of scientific

Scientific Establishments

4S

establishments, their status rivalry and the differentials of their power resources are not without influence on the construction of theories, the framing
of problems and the character of the techniques used for solving
6. Political and Scientific Establishments
But before getting on to some of the problems generated by the struggle
for dominance and autonomy among scientific establishments themselves,
I feel I must say something about the relationship between scientific and
establishments, in particular political establishments such as
governments, their high-ranking bureaucratic agencies and the spectrum of
parties. The problem, too little studied, does not lack urgency. An instructive
example are the similarities and differences between the relationship of
scientific establishments and state governments or party and military establishments in communist and non-communist countries, in one-party and
multi-party societies.
In these one-party states scientific establishments at universities form part
of a career pattern and of a power hierarchy whose highest point lies outside
the universities. The highest point in the career pattern of a professor is
membership in the central scientific institution of communist states, the
Academy of Sciences. In Russia a professor who can call himself Academician
has a higher status and greater power resources than one who cannot. This
centralization of control over scientific work corresponds to the - organizationally - monolithic character of a one-party state. The Academy of Sciences
has close links with, and is closely controlled by, the Central Committee of
the Communist party. Thus "Academician" is, in fact, the highest rung a
scientist can hope to reach in his professional career.
In East Germany, for example, one encounters a career pattern according
to which a university professor can be promoted first to the position of a
corresponding member of the Academy of Sciences. He may also be promoted
to a position at one of several research institutes, of the AoS (which is not
identical with membership of the Academy itself) or, alternatively, to a
position in one of several Research Institutes of the Central Committee of the
ruling party. All these are possible stepping stones on the way to membership
of the AoS itself. (That there may be a hidden overlapping of the functions
of some of these high-level research institutes, and thus a measure of rivalry,

46

Norbert Elias

need not concern us here.) Tight central control thus limits rather severely
the autonomy of scientific establishments below that of the Academy level.
The Academicians themselves are subject to a fairly strict party and state
control. Ideologically, the need for a measure ofautonomy of scientists at all
levels is still recognized. It is frequently alluded to in official publications
under names such as "self responsibility". In actual fact, the working out of
research projects is firmly controlled from above. Formally, they appear
to require approval of members of the Academy. The same goes for the
evaluation of research results.
In Western countries the evaluation still has a rather informal character.
One speaks rather vaguely of a "scientific community", as the agency which
accepts or rejects research results and decides on their integration into the
common fund of knowledge of a scientific discipline. But that is not a firmly
organized body. In fact, the formation of an opinion among the various scientific establishments about the work done by, or under the aegis of, one of
them is not a formalized social process - and, by and large, an open problem.
In communist countries it is firmly institutionalized. It lies in the hand of
special bodies, not necessarily of members of the AoS themselves.
In short, universities in these countries have become closely integrated into
the hierarchic structure of a one-party state with regard to research as well as
to teaching. It would be interesting to learn more about the consequences of
this type of organization for the development of the various sciences. Maybe
the Lysenko affair has served as a warning against the possibility that scientists
produce research results which they know will please the powers that be.
In multi-party states trends pointing in the same direction are by no means
entirely lacking. The dependence of scientific establishments on state fmance
and thus on state agencies of one kind or the other is on the increase. The
two great wars of the 20th century were among the levers accelerating the
interdependence between scientific and state establishments. Already during
the First World War some of the status and power rivalries between scientific
and non-scientific establishments flared up. They are not unusual in the case
of increasing interdependence between previously independent groups. In the
Second World War, again, the steadily increasing scientification of warfare
contributed to the frequent blend of collaboration and rivalry between
scientific and non-scientific, particularly military, establishments of which
many stories are told. In fact, it has been said, somewhat hyperbolically, that

Scientific Establishments

47

the 1914-1918 war was a chemists' war, the 1939-1945 war a physicists'
war.
In away, thus, the wars heightened the differentials in social power and
status of different scientific establishments. Evidently physicists and chemists
were much more capable of contributing to the war efforts than, say, sociologists, historians, or students of language and literature. During the greater
part of the 19th century, the social usefulness of physics was hardly recognized. In America, practical men, such as Edison who had little mathemlftics,
gained very high esteem as a result of their tangible successes both as inventors
and businessmen. Mathematizing physicists were still regarded as people in an
ivory tower whose social usefulness was doubtful. Late in the 19th century
and in the course of the 20th century all that changed. The advances made by
physical scientists in theory and practice were felt almost everywhere in
society at large. Thus, the wars, and especially the Second World War whose
outcome was in some situations decisively influenced by the inventiveness
of physicists, powerfully reinforced a trend in the hierarchic relationship
between different academic specialisms, which, perhaps less strongly, made
itself felt in peacetime too. Within the status hierarchy of scientific establishments, the physicists' claim to the dominant position as trend setters of
scientific work generally gained conviction as a result of the staggering
transformations in peace and war they had helped to bring about in society
at large. The demonstration of their social usefulness enhanced their status
and their power resources within, no less than outside, the academic world.
However, this was in no way the achievement of a clearly envisaged goal.
The descriptive historical literature about the relationship of scientific and
state establishments, especially in America, France and England is growing.
But sociological enquiries into the science-theoretical significance of these
developments lag behind; research into the impact which social developments,
such as the changing relationships between scientific establishments and
governmental or military establishments, have on the development of scientific knowledge itself, is still in its infancy. That the emancipation from an
extraneous authority was one of the principal conditions under which the
acquisition of knowledge gained its specifically scientific character, is fairly
well known - as far as it concerns the past. Galileo stands out as a pioneer of
the emancipation of science from the control of a powerful non-scientific
establishment, which claimed a monopoly of the means of orientation.

48

Norbert Elias

However, the question of why and how far scientists were able, even in
the world of the old state absolutism, to gain and to preserve a measure of
autonomy for their new mode of producing means of orientation, has hardly
been raised and explored. Nor has the complementary question found much
attention - the question whether and how far the relative independence of
scientific establishments can be maintained in the face of growing dependence
of their work on non-scientific establishments, bureaucratic, military, industrial or whatever. That scientific enquiries can be successfully carried out, at
least in some fields, under the very tight state control practised in the Eastern
one-party states, seems to point to the fact that the scope of autonomy
regarded as normal for scientific work in the Western world can, be whittled
down without seriously impairing the scientists: powers of discovery. But
whether that is really the case and, if so, how it is achieved, for instance
in the Soviet Union, these and other related questions can hardly be answered
without more systematic sociological enquiries in the developmental trends of
the relationship between scientific and non-scientific establishments and their
impact on the knowledge produced by the former. There may be serious
flaws in an authoritarian set-up that will come to light only in the long run.
7. The Development of Scientific Establishments
Scientific establishments derive their high power ratio from the monopolization of a particular type of knowledge. They are, in other words, groups of
specialists whose social function it is to administer and to augment a specific
fund of symbolic representations which has been handed on to them by
previous generations - a fund of symbolic representations which can serve
people as means of orientation. Scientists are not the first and not the only
groups who collectively monopolize the administration and production
of a society's central fund of knowledge, of its basic means of orientation.
Their principal predecessors in that capacity and, in more recent times
often their competitors, were priests of all kinds. However, the type of
knowledge monopolized by the two groups is different, even though there
are hybrids, and many transitional forms, of which philosophy is one.
The type of knowledge which has been administered and produced by
priests through the ages is, in the last resort, always revealed knowledge,
knowledge held to be received by people through a revelation of non-

Scientific Establishments

49

human forces, either indirectly through an oral or written tradition that

is handed on from previous generations, together with prescriptions for
invoking and communicating with these forces; or through a revelation
experienced directly by particular persons here and now. This type of knowledge, therefore, binds people to an authoritative tradition as the source of
their fund of knowledge or to the authority of communications from the
spirit world itself.
Scientists, too, are intrinsically dependent on knowledge transmitted to
them as authoritative from previous generations. Although some philosophical
doctrines present science as a quest for knowledge which an individual scientist starts as it were, from scratch. IIi fact, every scientist learns her or his
craft, including knowledge of how to acquire new knowledge, from previous
generations. Scientists' always start in the middle of a stream at the point it
has reached when they enter it. Even the greatest scientific innovators stand
on the shoulders of others, their discoveries presuppose a specific sequence of
antecedent discoveries, the problem they solve emerge in a form which makes
their solutions possible only at a particular stage of a sequence of problems
and problem solutions. like priests, scientists stand in a line of succession of
others of their kind, although never of scientists alone but also of problem
posing and solving activities in society at large. In one way or the other,
scientists continue the work of other people. However, in contrast to priests,
scientists can critically examine every piece of knowledge handed on to them
from previous generations in the light of their own observations and reflections. They are dependent on a fund of knowledge acquired by others in a
long inter-generational process - knowledge which they need as means of
orientation, as a point of departure. But they learn as part of their own
craft methods of testing what they have learned. They need not accept any
findings of their predecessors as authoritative, not even those regarded' as the
most basic and fundamental. That is one of the most crucial differences
between the type of knowledge whose monopolization forms the mainstay of
the power of priestly establishments and the type of knowledge whose
monopolistic possession and production forms the power chance of scientific
establishments.
But in practice, this anti-authoritarian character of scientific work is
curtailed, countered and circumvented in a great variety of ways, both from
without and within the scientific establishments. For one thing, the scientists'

50

Norbert Elias

endeavour to extend and to improve human means of orientation is carried

on within a wider social setting where differentials of power and status and
the concomitant conflicts playa central part. Scientific work presupposes
the existence of fairly differentiated state-societies which are internally
reasonably pacified. Scientists themselves, therefore, take their place within
a complex power and status hierarchy. They are thus often enough motivated
by the need to prevent a decline, and to attempt an increase of their power
ratio and to protect or to widen the scope of their autonomy in relation to
non-scientific establishments. The tensions between physicists and naval
officers in world war II, is one of numerous figurations of this kind.
At the same time, there are recurrent tendencies of some groups, or of some
individuals among scientists themselves, to set themselves up as authorities
whose views cannot be doubted or criticized. Again and again the organization
of scientific work creates conditions which allow some scientists, collectively
or individually, to claim and to establish effectively a hegemonic position
either within their own field or in relation to other fields. They achieve, in
other words, the position of a firmly entrenched scientific establishment.
Kuhn has seen something of this peculiar structure of scientific processes
in the course of which an intrinsically anti-authoritarian spurt in the development of a science, a "scientific revolution" may transform itself and harden
into an authoritative scientific orthodoxy. The interplay between these two
tendencies within a scientific process, between thawing and freezing trends,
between innovatorY rebellion and authoritarian orthodoxy has no simple
pattern and many aspects; only two shall be mentioned here, however, and
only one of them examined more closely.
As scientific developments are intergenerational processes, the interplay
between thawing and freezing (which may be partial or total) very much
concerns the relationship between scientific generations. Representatives of
a scientific view who have gained recognition in their own field or in society
at large for their work, can establish themselves as a hegemonic generation.
As high status persons they can assume the attitude of an established authority, discourage criticism, however well founded, and no longer teach the
rising generations those scientific techniques and strategies which would
enable them to examine effectively, and eventually to revise or to go beyond,
the "paradigm" of their teachers, the basic standard knowledge reached
within the scientific process by the hegemonic generation.

Scientific Establishments

51

Moreover, a hegemonic generation quite often can enlist the willing

cooperation of following generations by convincing them that the particular
virtue, and thus status and prestige, of their own scientific specialisms depends
on the acceptance of one or the other aspect of their orthodoxy which may
be represented by a particular type of theory or perhaps by a type of method.
The generations in that case make common cause because they have come to
believe that the status and prestige of their own field, in its status struggle
with other scientific specialisms within the academic universe, depends on
their acceptance of the traditional orthodoxy of their chosen field of academic work.
The problem one encounters here is of some importance for the understanding of the dynamics of scientific processes. The question is, what influence has the involvement of the steadily growing number of scientific
specialisms in an intramural status and power struggle on the products of
their labour, the knowledge produced by them? In other fields, it is easy
to recognize that occupational specialisms are engaged in a competitive status
and power struggle with each other and develop a professional ideology as
a weapon of defence and attack in that struggle. The significance of class and
party ideologies in highly industrialized nation-states need not make us forget
that of occupational ideologies. Not only do industrial enterpreneurs and
industrial workers, employers associations and trade unions develop professional ideologies, but so do journalists, printers, lawyers, boilermen, trainconductors, civil servants and, in fact, every possible variety of occupation.
Scientists are no exception. However, while in all these cases the professional ideology can profoundly affect their occupational conduct and
their work itself, in the case of scientists the relationship between their
profeSSional products and their profeSSional ideology is particularly close.
Both have the character of knowledge; both are means of orientation. The
question evidently is to what extent does the knowledge produced by scientists suffer a professional deformation because it also serves as a means of
preserving and enhancing their power ratio and their status intramurally as
well as extramurally, because, in other words, it also serves as a weapon of
defence against the encroachment of others upon their own field of work or,
alternatively, as a weapon of attack in attempts at imposing their own model
of scientific work as authoritative on other fields.
In a way it is surprising that sociologists, so far, have not paid more atten-

52

Norbert Elias

tion to the problem of scientific establishments, to their status and power

struggles and to the professional deformation their products may suffer in
connection with these struggles. It is surprising not merely because the
sociology of sciences has become a prominent area of sociological research
in recent years, but also because sociology as a scientific discipline and
sociologists in their work are almost everywhere very obviously involved,
intramurally as well as extramurally, in a long drawn-out power and status
struggle. Uncertainties about the standing of their own professional field and
concomitant status anxieties are probably stronger and more obvious in the
case of sociologists than they are in the case, say, of economists, historians
and even of psychologists. We may be inclined to think that physicists and
mathematizing scientists in general enjoy such an assured status, are so
certain of their high power ratio that they stand as it were, above these status
struggles and are not afflicted by uncertainties about their own goals and
their own standing. But that is probably an illusion; it remains, in any case, to
be examined.
However, the actuality of this problem in the case of sociologists is so
much in evidence that their failure to lift it from the prescientific level to
the scientific level of reflection and observation may well be due to the
feeling that it is too delicate and perhaps too painful a problem area to be
subjected to a reasonably detached scientific scrutiny. However, repression
of a painful problem does not help. towards its solution. It will prove in the
long run, much more promising to bring this problem in the open and to
work out the conceptual tools needed for enquiring into problems of this
kind regardless as to whether they concern sociologists themselves or any
other group of professional scientists. Nor is there any need to avoid using
the sociologists' experience and situation as evidence for the exploration of
such problems.
If one were to sum up the situation of sociologists today more pointedly
than is probably justified, one might say that they are aware of working in
a problem field of great promise and great importance. Their aspirations,
therefore, are high, but compared with them their performance on the average
is low, and low therefore, is comparatively speaking, their present status
within the hierarchy of scientific disciplines. That performance falls below
expectation and aspiration, however, is not entirely their own fault. It is not
a question of finding fault or of meting out blame. The problem to be ex-

Scientific Establishments

S3

plored is that of the sons for the sociologists' difficulties. Sociology being
itself a science can serve as an example. It enables one to probe into some
aspects of the wider social setting within which sociological establishments
are working today and into the pressures to which they are exposed, intramurally as well as extramurally.
8. Sociology in The Shadows of Two Stronger Power-Blocks
Intramurally Sociologists work within a setting where one of the most powerful and most successful groups of scientific establishments, whose members
rank high in the academic status hierarchy, that of physiCists, have formed
an idea of science in their own image; they have succeeded in propagating
the belief that their own method of setting and solving problems is the only
scientific method and that research not undertaken in accordance with their
prescription ought not to be accorded the status of scientific research. It
would perhaps be easier to recognize this as a hegemonic claim put forward
by one highly respected and successful set of scientific establishments in
its status and power struggles with others, were it not for the fact that many
members of another type of academic establishment, also traditionally
endowed with a high academic status, members of philosophical establishments, plough the same furrow. They reinforce the belief that physics is
the science par excellence, that other disciplines in order to be recognized
as scientific should use as nearly as possible the same method as physics
and, generally, should proceed along the same line. They have, indeed,
succeeded in giving many an academic a bad conscience or a feeling of
inferiority if they do not succeed in following the example of the physicists.
Yet, philosophers never bother to explain the diversity of sciences, their
science theories are generally based on the axiomatic belief that all sciences
must be cut according to the same pattern. Some of them go so far as to
present a purely phYSicalist ideal type model of science as a universal "theory
of science". By constructing highly formalized models of a science abstracted
from physics they claim to establish norms and to make prescriptions for
sciences of all kinds. The differences which exist between different types of
sciences lie beyond their field of vision. They neither account for these
differences; nor do they ask why many non-physical sciences, among them
most human sciences, cannot be fitted into the Procrustean model of a physicalistic science theory.

54

Norbert Elias

As far as sociology is concerned, one can say that measuring, quantification, mostly in statistical form, and the use of mathematics, has a legitimate
place in its proceedings. But it is a very limited place. The central areas of
the sociological problem field cannot be reached by means of quantifying
methods. They require other methods of enquiry and other types of theory
of which process and figurational theories are two, but by no means the only
examples. They can be constructed with as high an accuracy and reliability
as the law-like mathematical formulae of physics. Power balances and conflicts play a central part in these models and so do problems of conflictcontrol especially of war-control. Almost all social institutions are born from
a structural social conflict; they contain the birth-conflict, as it were, in a
frozen form which, in some cases, can thaw, thus bringing the conflict once
more into the open. In others it transforms itself into a stabilized, uneven
balance of power, a hierarchic order of superior and subordinate ranks. None
of these or of other long-term social processes, of figurational changes and
their dynamics can be adequately represented by means of mathematical
symbols and operations. If a quantifying reduction ala physics is attempted,
the results, inevitably, are barren and vast tracts of the problem field remain
unexplored.
As an example of the still unrecorded power- and status-struggles between
academic and other establishments, the case of sociology is rather instructive.
If we were to draw up a sketch of the situation in which sociological establishmen ts find themselves today, we would have to say that they are hemmed
in between at least two blocks - two types of establishment with much
higher power- and status-resources than they themselves have. In one way or
another, therefore, sociologists, uncertain of their own task as well as of their
standing, feel impelled, as weaker establishments often do, to seek greater
certainty and to gain kudos for themselves by a kind of mimicry; they take
over concepts, methods, values and forms of thinking from these more powerful establishments around them. But by doing so, they achieve the opposite
from what they hope to achieve. They increase their own uncertainty, impress only each other and perpetuate their malaise.
I have already referred to one of these blocks, the establishments of
physicists and physicalistic philosophers. The other type of establishment
from which prescriptive categories, concepts and values are taken over by
sociologists are political groupings, especially party establishments. This is

Scientific Establishments

55

made possible because the party establishments of our age legitimize themselves not simply as flag bearers of specific personal or group interests, but
also as harbingers of a general way of ordering human societies. This blend
of sectional self-interest and generalized social programme, represented by
the standard party ideologies of our time, is at present very much taken
for granted. One is hardly aware of the intellectual untidiness of a type of
ideology which tempers and half conceals a programme for the furtherance
of a group's self-interest by means of a programme for the general good of
society. Moreover, these untidy doctrines of an intrastate class conflict have
now become battlecrics of an interstate conflict threatening the destruction
of humanity.
However, although to-day social ideologies, social beliefs and ideals of this
type must be perceived as antagonistic to sociological theory, they were one
of its roots, a stepping stone towards the perception of human societies at a
higher level of synthesis. As this is one of the characteristics which party
belief and sociological theory have in common, it is not surprising that one
can observe, ever since the inception of sociology early in the 19th century, a
continuous interplay between them. Both have come into their own at about
the same time and sociological theories, so far, have always remained tied to
shades of party beliefs. The social establishments which form the primary
representatives of these beliefs, and their fol!owers in society at large, form
a much more powerful grouping of people than the groups of sociologists,
the makers of sociological theories, who, since Durkheim's days, came to
establish themselves slowly at the world's universities among the older scien-
tific establishments. It would be a rewarding task to follow in greater detail
the long-term development of the relationship between these two types of
establishment as well as that between social beliefs and ideals in society at
large and sociological theories in a univerSity setting. Perhaps the future
sociology of sciences will take care of such problems. Curiously enough,
sociology is among its own objects of study. One could probably show that
social belief and sociological theory at present form different poles of the
same parameter of knowledge, the one characteristic of greater involvement,
the other of greater detachment, different in their emphasis; they are represented by different professional groups.
This, then, is the second set of establishments whose concepts, values and
forms of thinking impinge upon those of sociolOgists. Again, the balance of

56

Norbert Elias

power between these two types of establishments is very uneven. Political

establishments and their followers, some of whom in our days occupy governmental positions and control the central monopolies of the state, command
very much greater power-resources than sociologists and social scientists. It is
not surprising, therefore, that in that respect, too, the sociologists' autonomy
in their work is limited, is in some cases almost extinguished, and that shades
of sociological theories, at the present stage of development, can always be
arranged broadly speaking, in accordance with the shades of the political
spectrum of social beliefs and ideals in society at large.
The situation in which human scientists of all kinds and among them
sociologists find themselves at the present juncture, the dependence of theory
on social ideology retarding the emancipation of the former from the latter, is
not entirely without precedent. One encounters an analogous situation in
the early development of the.natural sciences. In that case, the Church and
other religious institutions were the powerful political establishments which
monopolized and controlled the central beliefs of people. These were the
beliefs and ideals in relation to which the pioneers of the nascent science of
nature had to struggle for greater autonomy of the new concepts and theories,
the new categories and values and altogether the new type of knowledge they
were trying to work out. The emancipation of the rising natural sciences,
from super-natural belief was, if anything, even more difficult than that of
the rising social sciences from dogmatic social belief. The pioneers of the
natural sciences in the early days risked their lives or long imprisonment in
their emancipatory struggle for a new type of knowledge and a novel way of
acquiring it.
Moreover, in their case, as in that of representatives of human sciences
today, the emancipatory struggle was not only directed against forms of
knowledge monopolized and propagated by powerful establishments outside.
It was almost without exception also a struggle of the pioneers of natural
science within themselves. From Galileo to Newton and beyond early physicists and philosophers were almost without exception believing Christians.
Historians of science often try; in retrospect, to separate aspects of the early
scientists' work which appear to them as scientifically and particularly as
theoretically, correct from aspects connected with their religious beliefs. But
this separation of theory and belief in the work of the early natural scientists
gives a thoroughly misleading picture of this early stage in the development of

Scientific Establishments

57

physics - as misleading as the separation of theory and belief in the development of early sociology. A fuller understanding of the development of science
in its early stages is only possible if one enquires how and why a scientific
mode of approach emerges from non-scientific beliefs. In the early stages of
the emancipatory process, people still take it for granted that belief as a form
of knowledge takes precedence over theory; they do not perceive clearly
the difference, nor do they clearly recognize contradictions between belief
and theory. The emancipation of the scientific form of knowledge and of
acquiring knowledge from the matrix of beliefs, usually, is a slow intergenerational process. Beliefs, at that stage are usually backed by powerful
social establishments. By contrast, early scientific forms of enquiry and of
knowledge, as a matter of course, lack any institutional shell of their own,
they are often represented by circles of people without other than personal
bonds, or, at the most, with loose institutional bonds whose power ratio,
compared with that of the establishments representing the beliefs of the
moment, is very small; and the same is, of course, true of the rise of a new
type of science in its relation with old established sciences or with competing
social beliefs. At first, emerging theory or, for that matter, a new type of
science, is likely to be weak in terms of the power chances of its promoters.
Representatives of established beliefs are apt to use a variety of intellectual
devices in order to counter the challenge implicit in the emergence of a
novel type of knowledge. They may at first try to disregard its existence,
to kill it through silence. They may argue that the novel type of knowledge
does not contain much information beyond that which has been known
before. They may adopt some peripheral aspects of the new knowledge and
disregard those crucial innovations which run counter to their own tradition. In accordance with the power differentials between the representatives
of an emergent theory and the old established beliefs, representatives of
the former themselves may be accommodating vis-a-vis those of the latter;
they may be in all sincerity, unable to cur themselves loose from the older
beliefs.
There are, thus, a great variety of ways in which an emergent scientific
theory can be influenced by the ruling types of belief. The existence of a
centralized church enforCing the belief in a single god was probably quite
an important determining factor in the emergence of a type of theory embodying the concept of unified nature ordered according to eternal laws
(7).

58

Norbert Elias

One of the most significant characteristics of contemporary sociological

establishments is the diffuseness of their theories as well as the relatively
small measure of their autonomy. I have pointed out before to what extent
the Sociologists' theories and, indeed, their whole conceptual equipment is
determined today by extraneous models, in particular models set by the
physical sciences and by their academic establishments on the one hand and,
on the other hand, by models set by political establishments of all hues, red,
blue, black, green, brown or whatever. Their autonomy is limited and uncertain, heteronomy prevails. One cannot doubt that this heteronomy of
orientation contributes greatly to the inability of sociological establishments
to work out a convincing central theory of human societies, comparable in
unity and object adequacy to the central theories of nature which are the
pride of phYSicists and also a condition of their high power-ratio and their
high status within the academy hierarchy. The diffuseness and the high degree
of heteronomy of sociological theories affects, the power-ratio and the standing
of sociologists and related professional groups. Their relatively low standing
in tum makes them inclined to look for prestige-giving high status models
elsewhere. Their intellectual subservience to other established groups, whether
scientific or political, reinforces their outsider position and their inability
to fulfil their central function as explorers of human societies; this, in tum,
reinforces their subservience and the heteronomy of their procedures, and
so the process goes on.
The fact that the relationship between the relative autonomy of a particular group of scientists and the relative autonomy of their theories, so far
has not been brought sufficiently into focus, is in no small degree due to a
misunderstanding about the nature of the high relative autonomy of physical
theories and of physical establishments. Both are often directly or indirectly,
attributed to the use of a particular method of research, based on a high
degree of mathematization, of reduction, of "qualities" to "quantities". But
this is an incomplete science-theoretical analysis with some ideological undertones. The effectiveness of a research method depends on its instrumental
appropriateness, its appropriateness as an instrument of research into problems of a particular type. To put it in a somewhat heretical form, the relative
autonomy of a field of scientific studies and thus of the human groups
engaged in these studies in relation to other groups of scientists is, in the last
resort, a function of the relative autonomy of their subject matter, or, to

Scientific Establishments

59

avoid the somewhat confusing idiom, of their object, in relation to the problem field, the object of other academic groups. The question, in short, with
which one is confronted here admits of a perfectly neat and clear formulation.
It is simply the question of whether and to what extent the nexus of events
to which we refer as human societies,
a structure of its own, which
is not entirely reducible to the structure and regularity of that nexus, which
we call ''nature'' and which is the subject matter of the research of natural
scientists, particularly of that field whose representatives claim today that they
are able to provide models for all other research enterprises and that their
method can and must be used for the exploration of all other fields of study.
The hegemOnic claim of the physical sciences is to some extent supported
by the widely held idea that the recognition of ''nature'' as a relatively autonomous nexus of events with structures and regularities of its own, is obvious.
One need only open one's eye, so it seems, in order to be aware of the fact
that the sun and the moon move in regular circles, that falling bodies or gases
obey immutable laws; in short, that the nexus of physical events possesses
recurrent regularities of its own has, a fairly high degree of autonomy in
relation to the wishes and hopes of human beings. Many people would
probably argue it is much more difficult to maintain that the societies formed
by human beings have structures and regularities of their own which are
relatively independent of human wishes and aims. That indeed is the core
of the problem. If human societies in their structure and their course are
wholly determined by wishes and aims of the people who form them; if they
are, for instance, determined by the wishes and aims embodied in political
party programmes, then indeed, the chance that sociologists and other social
scientists may gain a measure of autonomy in relation to political establishments is small. The aspiration of sociologists for a higher measure, though
of course not for complete, autonomy provided the power structure of a
state allows it, would be a vain hope. Again, if the structure of societies and
the regularities of social processes could be adequately expressed and explained as derivatives of the structure of atoms and molecules of which humans are undoubtedly composed; if in other words, the structure of societies
is reducible to the structure of their smallest component parts, then indeed
the claim of physicists to set models for social scientists would be justified.
In that case social sciences and, in a wider sense, human sciences, could have
no degree of autonomy in relation to physics, as their subject matter -

60

Norbert Elias

human societies - would have no relative autonomy in relation to the subject

matter of physics.
I believe that putting the question of the relative autonomy of the subject
matter of the various sciences clearly into the centre of the discussion of their
relationship, can help to open the way to a better assessment of the hegemonic claim of the physical sciences. If sociologists can convince themselves
as well as others that human societies, as a field of studies, have a measure
of autonomy in relation to the levels of nature explored by physicists, the
level of ions, atoms and molecules, then indeed, one can, with greater confidence, go on one's work at emancipating research into human societies from
heteronomous models and at framing problems, at experimenting with
methods, commensurate with the specific nature of one's subject/matter, of
human societies.
But before one can tum to this matter, some of the rubble left over from
previous discussions has to be cleared away. There is the notion that physital
nature can be recognized as an autonomous nexus of events quite easily, much
easier in any case than societies. That is a piece of ideological fiction well
worth debunking. It was enormously difficult for humans to envisage what
we call "nature" as an autonomous nexus of events, and to devise a method
of exploration commensurate with its peculiar characteristics as a semiautonomous nexus. The difficulty is disguised, especially by philosophers,
because they lack a contrast picture - a picture of the pre-scientific vision of
the universe and of the corresponding mode of gaining knowledge about it.
That is why I am using as a terminus technicus for the world image of
people at the pre-scientific, the magic-mythical stage of development, the
concept of a spirit world. It is a distortion of one's perspective to speak
only of "pre-scientific forms of thinking" without regard for the world
picture, for the substantive vision of the nexus of events within which people,
through the activity we call "thinking", orientate themselves. The prevailing
subjectivistic or phenomenalistic trend of our theories of knowledge make
us inclined to separate the form of thinking from peoples' vision of the
object of thinking. The two go together and cannot be separated. It is in
the encounter with the world of objects and in constant connection with
their experience within a world that forms of thinking pattern themselves
and change in a specific order of succession. The magic-mythical forms of
thinking are the earliest, the most spontaneous forms of thinking, as I have

Scientific Establishments

61

already said, because they correspond to that kind of experience that is

primary both in the social, as in the individual experience of human beings
- the experience of one-self in relation to other human beings with one's
emotions and passions moving relatively freely from one-self to others, from
others to one-self. This is the primary human experience of the world - the
experience as a universe of potentially or actually living things, a world of
spirits. The development from there to the experience of the world as "nature"
was long, slow and arduous - how arduous, is certainly difficult to imagine
for the beati possedentes - for those who inherit this concept as a fait
accompli. As long as one follows the tradition according to which the
categories or forms of thinking are shuffled into one compartment, the objects
of thinking into a second and separate compartment, human experience into
a third, and the social groups within which standards of orientation or of
thinking develop into a fourth, it is hardly possible to understand the development of any of them. For the actual development - the observable development - comprises all these facets closely interwoven without priority of
anyone of them.
For a theory of scientific establishments the awareness of this interwovenness is of particular importance. It means, among other things, that the
terminological separation to which we are used, between the structure of
knowledge, the genesis of knowledge, and the structure and development of
the human groups, which store, administer and produce knowledge, is unreal
and misleading.
One can easily observe this interdependence in action. The specificity of magicmythicalcfu-rms of thinking and of the experience of the world as a spirit world
finds expression in the specificity of its most representative groups as groups
of magicians, soothsayers or priests. The emergence of physical science and the
vision of the world as a causal nexus, as "nature': is inseparable from the emergence of groups of physicists and finally, ofestablishmentsof physicists. Because,
up to a point, "knowledge" and "the groups of knowers" are today often treated
as objects of different and perhaps of rival academic disciplines, that difference
and the related rivalries project themselves into the conception of the objects
themselves. As a result some people may assert that the structure and even the
development of knowledge is an autonomous fact and can be understood quite
independently of
structure and development of the human groups where this
knowledge is administered, transmitted and produced. Others will assert that

62

Norbert Elias

the development of knowledge has no intrinsic structure of its own and can be
fully explained in terms of the structure and development of the groups of
knowers. However, neither of these two approaches does work, least of all in
the case of scientific knowledge. In that case it is particularly clear how
inextricably interwoven and interdependent are the social value and thus the
status and power ratio of groups of scientists and the cognitive value of the
scientific knowledge produced by them. The same can be said with regard to
the relative autonomy of both. The autonomy of the scientific knowledge
and particularly of the theories of a specialized branch of sciences and that
of the group of scientists who administer and develop a particular type of
theory, again, are inextricably interwoven and interdependent.
The fact determines to a very high degree the relationship between different academic establishments. Most of them are busily engaged in the task
of proving to themselves and to all the world, the autonomy of their own
field of studies in relation to other fields thus trying to ensure their own
autonomy as a professional group. Thus psychologists are eagerly building up
theories of their own which protect them from the threat of being subdued by
physiologists and, more generally, biologists, while some biologists, especially
etholOgists like Lorenz, on their part are on the warpath engaged in some
highly successful forays into the psychologists' and even the sociologists'
territories and have begun to settle down there. On the other hand, biologists
themselves are on the defensive against a vigorous expansionary move of the
physicists who, in the form of microbiology, have begun to colonize some
branches of biology such as genetics, to transform them into provinces of the
great physics empire and to subject them to the rule of what is called the
"scientific method" , which is in effect, the method of the physiCists.
All this may give some insight into the magnitude of the task before us.
Some of the basic theories, some of the forms of thinking, with which we
have grown up, may have to be revised. The theory of abstraction is only
one of many examples. If I were asked to say briefly and superficially what
I regard as the linchpin of the reformation to which I have alluded, I would
say it is the re-thinking of our theory of knowledge in terms of evolving
figurations of people, of developing groups of interdependent individuals
as the subject of knowledge rather than of an isolated individual of the
homo clausus type. It is the break with the habit of discussing knowledge
without saying what it is, without stating clearly that knowledge is the

Seicntijic Establishments

63

meaning of human-made social symbols which are intergenerationally transmissable. It is further the re-thinking of all our problems and theories of
knowledge in terms of long-term processes of change, replacing the present
tendency towards reflections, concepts and theories which aim at reducing
all processes of change symbolically to static conditions.
That such a research programme challenges many existing habits of thinking I have already said. These habits of thinking are backed by the authority
of two types of high ranking academic establishments which show strong
affinities to each other, the establishment of theoretical physicists and the
establishment of transcendental philosophers. The latter are firmly wedded
to the concept of a subject-object divide which casts doubt either upon the
existence of objects or the human capacity to grasp them as they really are.
Kant's brief statement that each of us may assume the existence of other
minds operating in a similar way as ours, but, of course, we can never be sure
of it, is revealing. Some contemporary transcendental philosophers with the
same solipsistic bent as Kant, speak of inter-subjectivity, apparently without
any awareness that the very concept contradicts the basic assumption of
transcendental metaphysics, which reduces cognition not only of objects,
but inevitably also of other subjects, to the transcendental operations of an
individual mind.
The blockages imposed on many of our reflections by the establishments
of theoretical phYSicists are of a different kind. An example is the assumption
that knowledge of the structures and regularities of the smallest constituent
parts of a composite unit provide the ultimate key to the knowledge of
the structure and regularity of that composite unit itself. The conclusion
to be drawn from this assumption for the relative cognitive value of the
knowledge provided by all the various non-physical groups of scientists, is
obvious. Physicists are the specialized providers of knowledge about the
smallest constituent parts of which everything else in the universe, humans
included, consists. On the assumption that the properties of composite units
can be fully deduced and explained from the exploration of their constituent
elements, physicists hold in fact, the key to all other sciences. There are
many variations and permutations of the physicists' claim that their science
provides the key to all others. In some cases, their hegemonic claim is based
on the idea that their method of research is the only scientific method and
ought to be imitated by all other scientists. In other cases, physicists put

64

Norbert Elias

forward the claim that the forms of thinking, the categories, which they
develop, in the exploration of the constituent parts of matter are applicable
to all realms including the realm of life.
Thus, Werner Heisenberg in a series of lectures, published in 1973 said
(8):
the forms of thinking evolved with the development of atomic physics are wide enough
to provide scope for the various aspects of the problem of life and the direction of
research connected with them.

This kind of statement exemplifies the hegemonic claim of physicists

quite well. If Heisenberg goes so far as to claim that the forms of thinking,
the categorial apparatus of modem atomic theories, are applicable to all
problems of life, he has, of course, in mind, problems of biology. It is
characteristic of the present climate of opinion among physicists that we
have reached a stage where biology is recognized as a legitimate field for
scientific work, but human sciences in the field we call human society, still
lie beyond their borizon. They recognize humans only as subjects, as their
object only non-human nature, but they do not recognize human beings
themselves, and the societies they form, as a relatively autonomous object
of studies. On the spiral staircase of knowing physiCists can only claim to see
the nexus of non-human events as a relatively autonomous field of research
with regularities of their own. They cannot yet ascend to the higher platform
from which they can see themselves as groups of humans forming fluid figurations of various kind, among them scientific establishments and, in a wider
sense, societies which also have structures and regularities of their own.
But is that correct? Do the groupings that humans form with each other,
have structures and regularities of their own? One can see that this is the
question which has to be answered if the claim of sociologists to have a
field of studies of their own can be seen to have a legitimate basis. But is
there such a basis? If sociologists talk of social structure, is it true that
there are such structures which have to be investigated as such, which cannot
be reduced to biological structures and, beyond that, to the structures of
matter? I am not sure whether one can recognize this as a fairly vital question
for sociologists. One cannot easily come to grips with this problem, if one
does not, at the same time, bring into the open the half hidden fact that
this is a struggle between different scientific establishments, if one does

Scientific Establishments

65

not investigate more closely the power and status hierarchy of the different
academic disciplines and raise the question: What determines, what
legitimizes the claim of different academic disciplines to a relative autonomy
in relation to each other? Or, alternatively, what justifies the claim of some
disciplines such as physics, that none of the others can have any relative
autonomy and that they all, sooner or later, can be explained and investigated
in terms of one of them, in this case, in terms of physics?
If one removes all disguises, the questions before us are these: Can we
assume that sooner or later all the structures and processes investigated by
sociologists will be explainable in terms of atomic structures or, for that
matter, of the structures of the ultimate particles of the universe whose
secrets contemporary physiCists try to unravel more eagerly than any of
their ancestors? Or, alternatively, shall we assume that social structures and
processes, because of their greater complexity, do not admit of any reliable
scientific investigation and must for ever remain a mystery?
It is not uninteresting in this context that Heisenberg in the same series of
lectures, stated even with regard to physics that (9)
In the great majority of cases, the complete mathematical calculation of a set problem
will technically not be possible for all too great complications can no longer be mastered
in mathematical terms.

Human societies are, in terms of their functional interdependencies, the most

complex structures known. They are certainly very much more complex
than the most complex structures with which physicists are concerned. That,
no doubt, is one of the reasons - one among others - which helps to explain
why sociology is a late-comer among the developing sciences.
Complexity alone, however, cannot explain the difficulties of our task.
Above all, it cannot help us to decide the crucial question of whether or not
social structures possess, in fact, a relative autonomy in relation to biological
and physical structures in the same way in which biological structures, too,
possess a relative autonomy in relation to physical structures. In order to
answer this question convincingly, more space and time would be needed
than is available to me in this context. The essential point, though, is simple
and its gist is this.
The assumption, based on the physicists' experience, that the properties of
all composite units can be explained in terms of the properties of the smallest

66

Norbert Elias

component parts holds good only at those relatively simple levels of integration with which physicists are concerned. That is to say, with objects reaching
from the smallest subatomic particles of matter to the larger molecules. At
that level, in most cases, though not in all, processes of integration are
reversible. The properties of the component parts do not fundamentally
change when a given composite unit at the higher level disintegrates. When
a water molecule disintegrates into it atoms, the properties of the atoms
are not different from their properties as part, of a molecule. As soon as one
ascends to levels of integration which we call biological; represented, for
instance, by an organism consisting of a single cell, the rule established at
the physical levels no longer holds good. If a cell disintegrates the constituent
parts of the cell at the next lower level, e.g., its nucleus, its membrane, etc.,
disintegrate, too. They disintegrate, broadly speaking, into physical units,
into atoms and molecules. The difficulty for many people is to think in terms
of not only three or four but, perhaps, 20 or 30 levels of integration superimposed upon each other like the part-units of a Chinese box.
Such units require, in fact, different forms of thinking. Here there are, at
every higher level of integration, regularities which cannot be explained alone
in terms of the properties of the lower constituent parts. Thus, for example,
at a certain level of integration, one encounters a type of events for which
no parallel can be found at a lower level. Concepts such as life and death,
birth and heredity, for example, are characteristic of the biological but not of
the physical levels. We have separate words for these distinct properties
of different levels, one is often inclined to explain the events to which they
refer in terms of an addendum, a kind of substance added to or taken away
from the other component parts. Thus, some people try to explain life in
terms of an invisible force or substance added to the visible forces or substances. But that is only because a reifying physicalistic mode of thinking
blocks the awareness of the fact that forms of integration and the configuration of component units have explanatory functions, in addition to the
explanations to be derived from the properties of the component parts.
But perhaps I have said enough to indicate what that means with reference
to human societies themselves. In their case, too, the figuration of constituent
units, of human beings has an explanatory function of its own in conjunction
with the explanatory function to be derived from the properties of the constituent parts. In this case of the biological properties of individual human
beings.

Scientific Establishments

67

I had to cover my ground rather rapidly, but for the time being it may be
enough to put in fuller relief the basic problems with which we are concerned.
I am trying to state the fact that social structures have a relative autonomy in
relation to biological structures, that they represent a nexus of their own
different from the nexus which we call nature, though derived from it and
related to it. Present reflections of physicists and philosophers are still tied
to a scheme where the universe consists of physical eVents into which humans
are set inexplicably as explorers of objects. Neither physicists nor philosophers
so far recognize the distinct order of human beings, which we call societies,
as an order with structures and regularities of its own, as a semi-autonomous
level of the Universe. To establish this fact requires a struggle against many
established views and against the groups of people who are the holders of
these established views, of the older scientific establishments. Sociologists
should, I think, examine the basis of their own field of studies much more
thoroughly than they have done so far. They should know that a rising
science cannot assert itself unless it is able convincingly to free itself from
the inappropriate models of older establishments, to develop theoretical
models and forms of thinking commensurate with the relative autonomy
of its subject matter, and to fight for its own relative autonomy in relation
to the older sciences and the old philosophy which even when seemingly
hostile to them, is in terms of its categories and its de-humanizing mode of
abstraction their companion-in-arms.

Notes and References

'" This paper has been written at the Center for Interdisciplinary Research, University
of Bielefeld. I wish to thank its directors for putting at my disposal all its facilities. I
am grateful to Mrs Petra Kunze who with great patience has helped me producing the
manuscript. I am greatly indebted to Richard Kilminster, University of Leeds, for his
very helpful and stimulating comments on this paper and I wish to thank my coeditor, Richard Withley, University of Manchester, for his helpful advise.
1. Immanuel Kant, 'Uber die von der koniglichen Akademie der Wissenschaften zu
Berlin fUr das Jahr 1791 ausgesetzte Preisfrage: Welches sind die wirklichen Fortscnitte, die die Metaphysik seit Leibniz'ens und Wolff's Zeiten in Deutschland gemacht

68

2.
3.

4.
5.

Norbert Elias

hat?' Hrsg. von D Friedrich Theodor Ring, Konigsberg 1804 in: Imm. Kant's Werke,
Bd. viii, Cassirer Berlin 1,!22, S. 245:
"Die Form des Objekts, wie es allein in einer Anschauung vorgestellt werden kann,
griindet, sich also nicht auf der Beschaffenheit des Objekts an sich, sondern auf
der Naturbeschaffenheit des Subjekts ... "
See N. Elias, Problems ofInvolvement and Detachment.
Imm. Kant, ibid. p. 256 ..... Ob das Objekt, welches wir ausser uns ansetzen,
nicht vielleicht in uns sein konne und es wohl ganz unmoglich sei, etwas ausser
uns als ein solches mit Gewissheit anzusetzen." One is again and again surprised
at Kant's repressive innocence which allowed him not to notice that the very use
of the personal pronoun "us" referring as it does to a plurality of people "outside
of himself" vitiated his whole argument. The same goes for his use of a common
language directed at a plurality of human beings. A language, i.e., a specific pattern
of learnt sound signals symbolizing, and at the same time serving as a store of, the
common knowledge of a human group, presupposes a plurality of human beings. It
shows a remarkable lack of consistency in thinking - or, to use the fashionable word,
of "logic" - to try fitting the intrinsically social datum of language into the framework of a transcendental philosophy of whatever kind unless one postulates that the
specific language which "language communities" speak is inborn and biologically
inherited by each of the members of that community individually.
The essence of transcendentalism is the assumption of a highly specific
pattern either of thinking or of speaking which is present in every individual human
person prior to all learning. The concept of a human being, underlying transcendental
philosophy is that of an isolated individual, a homo clauses, endowed with specific
ways of connecting events which act as an impenetrable wall between his own consciousness and whatever is outside of it - natural objects or other persons. Every
specific language on the other hand presupposes learning from other persons, presupposes in fact the existence of other persons. Languages change over time. Every
individual person learns the language in the form which it has assumed at the time he
or she enters the "language community". Languages, like the societies where they
flourish, are in a condition of flux. Transcendental philosophy makes assumptions
about specific aspects of people's mental capacities which are unchanging or, in other
words, innate. These assumptions can hardly be reconciled with the observable changing languages spoken in observable changing societies.
In fact, expressions such as "language community" or "communication community" are a good example of one of the principal characteristics of transcendental
philosophy. In a large measure it creates its own objects. The fact, to which this
expression refers, is simple: as children human beings learn a communal language
from other human beings, and without learning it, they can neither argue nor communicate with others. By conceptualizing this observable datum in the form of a
noun at a very high level of abstraction, a philosopher gives it the appearance of a
datum whose discovery is of the very highest cognitive value. A new technical term
suggests in some way a reflection upon an "object", a condition of humans existing
prior to all experience. That is a good example of a philosophically invented object.
N. Elias, The Civilizing Process, Vol. 1 Preface, Blackwell, Oxford, 1979.
"Science does not rest upon solid bedrock. The bold structure of its theories rises,
as it were, above a swamp ... ", Karl Popper, The Logic of Scientific Discovery,
London,1968,p.59.

Scientific Establishments

69

6. Norbert Elias: 'Een essay over tijd' (An Essay on Time), De Gids, CXXXVII
CXXXVIII Amsterdam, 1974/75.
7. Joseph Needham has drawn attention to the possibility that the Chinese did not
develop the concept of a unified natural science partly because they did not have the
concept of a single god ruling the whole universe, see. "Why didn't China develop
modern science?" International Herald Tribune, Oct. 24, 1979, p. 98.
8. Werner Heisenberg, Wandlungen in den Grundlagen der Naturwissenschaft, Stuttgart,
1933, p. lxx.
9. " .. , dasz in den allermeisten Fiillen die vollstiindige mathematische Durchrechnung
eines gestellten problems technisch nicht mBglich sein wird, denn allzu grobe Komplikationen konnen wir eben mathematisch nicht mehr bewiiltigen" (Heisenberg, ibid.,
p.156).

THE SCIENTIFIC POWER ELITE - A CHIMERA; THE

DE-INSTITUTIONALIZA TlON AND POLlTICIZA TION
OF SCIENCE

PETER WEINGART
Universitiit Bielefeld

1. Knowledge and the Power of Problem Defmition

The notion of "scientific establishments" almost inevitably raises associations

of the plethora of writings on the emergence of a "scientific power elite", a
"new priesthood", the "scientific-estate", "new mandarins" and the like (1).
These concepts, the latest example of which is Gouldner's "new class" of
intellectuals, have some common assumptions.
that theoretical knowledge assumes centrality as a source of innovation
and policy formulation and that it serves as a power resource;
that the scientists (or professional and technical communities) exert
their power as a fairly coherent group which, often implicitly, can be
identified as the academic community;
that the scientists represent common values and interests (only this
makes their rise to "power" noteworthy).
A somewhat simplified image is one of a scientific community based in
universities, representing academic values and the vested interests of pure
research, appearing to be homogeneous with respect to political opinions and
the ability to represent uniform goals.
What, one has to ask next, constitutes the power of 'scientific establishments' or power elites, or how can knowledge be a resource of power? Two
answers, it seems are possible. Knowledge is power because it provides the
means to determine problem definitions, and because it enables those who
hold knowledge to determine the solutions of problems. (The distinction is
71
Norbert Elias. Herminio Martins and Richard Whitley reds.). Scientific Establishments
and Hierarchies. Sociology of the Sciences. Volume VI. 1982. 71-87.
Copyright 1982 by D. Reidel Publishing Company.

72

p",

,f

Weingart

analytical, of course, as often the definition of a problem implies its solution.)

While there is a widespread consensus that the role of scientific and technical
knowledge in politics is growing, it is rarely specified what this means in
terms of the resource of power (2). Lane, in his thoughtful analysis of what
he calls the 'knowledgeable society' points out "that the political domain
is shrinking and the knowledge domain is growing, in terms of criteria for
decisions, kinds of counsel sought, evidence adduced, and nature of the
'rationality' employed" (3). In a similar vein but more precisely, Don Price
observed that
the process of responsible policy making is ... not something that begins with the
definition of a political ideal according to some partisan doctrine, and concludes by
using administrative and scientific means to attain that end. It is a process of interaction
among the scientists, professional leaders, administrators, and politicians; ultimate
authority is with the politicians but the initiative is quite likely to rest with others,
including the scientists in or out of government (4).

If it is true then, that the "knowledge domain is growing", that scientifictechnical and professional knowledge has assumed considerable functional
importance and legitimating power, it will also have a growing impact on the
structure and contents of political problems. It may, therefore, be inferred
that knowledge conveys political power insofar, and only insofar, as it becomes a major ingredient in the definition of political problems. It does so
because a certain body of knowledge will then assume an orienting function
which entails the acceptance of implicit assumptions, the pre-determination
of the possible range of solutions, the exclusion of other knowledge and thus
alternative problem perceptions and solutions, and, as a result of all these,
the distribution of life chances in a certain pattern rather than another. On
the other hand, the instrumental function of knowledge alone, i.e., its role in
determining problem solutions, does not convey power because it leaves the
scientists in a state of dependency on those who do or do not call on them
for advice. This is the classical image of the scientific advisor, which, of
course, depends on the assumption, that politics and science, values and
knowledge, can be neatly separated.
Looking at the implicit and explicit assumptions of the writings on the
rise of the scientists to power and referring them to that notion of power
which alone can be meaningfully connected with knowledge as a resource,
I want to show that these assumptions are flawed. It can be argued that the

The Scientific Power Elite - a Chimera

73

scientists have managed to achieve a unified representation of interests and

very strong influence, if not virtual power, in the area of science policy, or
policy (-making) for science, as Brooks has termed it (5). But in the area of
general policy-making, Le., science for politics, such influence or even power
of a 'scientific establishment' or power elite is science-fiction, and so are the
hopes and fears that are connected with it: hopes for the "end of ideology"
and a higher rationality, fears of irresponsible technocrats, Frankensteins
and the like (6). The general reason for making this claim is that while much
attention has been given to changes in the structure of the political system
brought about by the impact of science and technology, little if any attention
was given to the impact that the increased involvement of science in politics
has had on science itself. The changes in the institutional structure of science
went by unnoticed or, if they are the focus of analysis, they are not linked to
the role of science in politics (7).
Here I want to deal with two particular reasons why scientists do not
form a new establishment or power elite: (a) The institutional locus of the
production of knowledge can no longer be identified with academic science.
Systematic knowledge of the kind which is instrumental for policy-making
and the crucial resource for the definition of political problems is produced
in government science installations, industrial research laboratories and
management staffs, by experts in political administration and probably to a
small extent only in academic institutions. Science, thus, loses its institutional
identity which used to be identified with universities and academic settings.
The notion of the 'scientific establishment' therefore becomes misleading.
(b) The diversification of the institutional basis of science has its correlate
in the loss of a common frame of value-orientations and beliefs as well as
a common basis of interests among scientific and technical experts. Their
involvement in politics which has been interpreted as a 'scientification' of
politics, turns out to be the 'politicization' of science at the same time. The
professional status of science with its sharp delineation from other social
institutions, its self-governance with respect to quality standards, criteria of
relevance and a code of ethics becomes subject to political conflicts. Alliances
and fractions emerge which run along the lines of political convictions rather
than of systems of knowledge.
In the following sections, I will illustrate these two points. First, I will
show that the impact of scientists and scientific knowledge on policy-making

74

Peter Weingart

is limited, that the power to defme problems rests with what we have called
'hybrid communities', experts from different institutional backgrounds rather
than with the scientists alone. Secondly, I will show that the very condition
for the scientists' rise to political power, the growing instrumental importance
of scientific and technical knowledge leads to a politicization and consequent
de-professionalization of academic science.
2. Who Dermes Policy-Problems? - The Emergence of 'Hybrid Communities'
(a) The Institutional Basis of 'Experts'
In order to show that experts from a diversity of institutional backgrounds,
rather than scientists alone, are involved in the process of problem defmition
two types of indicators will be used: the formal representation of scientists
and other 'experts' in policy-defming bodies, and the differential impact of
types of knowledge held by these groups on the defmition of problems.
Regarding the first indicator the theoretically possible continuum of cases
ranges from the scientists exclusively initiating and structuring policy to their
being completely excluded from any influence. The latter case implies that
'experts' from industry, interest groups or public administration itself take
the leading role, if, as one can assume nowadays, there are always 'experts'
involved in policy-making. Examples for the first case can usually be found in
the area of science policy where a strong impact of scientists is least surprising
(8). I will deal with an 'intermediate' case, however, which seems to be more
representative: the formulation of the German government's programme
for environmental protection of 1970. The scientific-technical component
in this case is still very important but just one among others: aspects of
economic and fiscal policy as well as problems of regional planning and
political administration in general playa major role.
First of all, the initiative for the formulation of the programme emerged
within the government and was motivated both by objective pressures (international efforts coordinated by the U.N.) to reach agreements on environmental protection which would affect the competitiveness of industry and
internal political considerations (the administrative reorganization of old and
new competencies for environmental protection as part of a 'strategy of the
smaller coalition partner to gain more influence).

The Scientific Power Elite - a Chimera

75

Outside advice was first obtained after the schedule for the programme
formulation had been fixed, when the Battelle-Institute, a commercial thinktank, received a grant from the Ministry of the Interior. After negotiations,
Battelle offered to work out an 'indicator matrix' for the evaluation of
objectives and scientific-technical considerations concerning environmental
protection. The research was supposed to support the formulation of objectives in the area of environmental protection. At this time a crucial decision
had already been made in the administration: formulation of the programme
was to be divided between 'project groups' organized "vertically", i.e.,
according to medial sectors of the environment such as air, water, open seas,
noise etc. Only three "horizontal" groups were planned, one on economic
problems, another one on problems of public information and a third on the
organization of scientific advice for environmental protection. Only the first
and the third were eventually established.
It is not difficult to see that this structure of the apparatus which was to
formulate the programme was a result of the jurisdictional structure of the
administration. Neither the cabinet committee nor the steering committee
set up in the Ministry of the Interior had the necessary competence for the
identification and processing of problems which would have allowed a more
encompassing approach to the problem of environmental protection.
The task of the project groups was supposed to be defined in detail by
the steering committee set up for this purpose. Their function, in general
terms, was to establish programmes for their respective subject matters which
allowed integration into the general programme. These sub-programmes
should contain an analysis of the status quo and a projection of likely developments for the next five years as well as a concept for protective and
counter-measures, the definition of threshold values, budgetary requirements
and economic implications. It was at the level of the project groups that
scientific, legal and economic expertise had to be relied upon, they were the
central mechanism with which to organize scientific advice to government in
establishing the programme on environmental protection.
The pattern of recruitment into the project groups was diverse, to say the
least, and no clear picture emerges. The group on 'environment oriented
technology', and its sub-groups, was composed almost exclusively of experts
from industry. Its work was virtually delegated to industry, in order to tap
the experience assembled in it, but obviously also for legitimation reasons,

76

Peter Weingart

i.e., to assure acceptance of regulatory measures which would affect industry

(9). Another extreme example is the project-group on 'measures against noise'.
In this group the share of specialists from public administration (both State
and Federal) was exceptionally high. Of 17 members, only one was from a
university. Of the 27 additional members and guests of the subgroups, only
one was a representative of science.
The group on 'environmental chemicals and biocides', the one closest to
ecological research, was almost equally divided between experts from Federal
ministries and from government research laboratories. There, too, with one
exception, university scientists were not called in and only two experts for
ecological chemistry were represented. Another five experts from chemical
industry served as outside advisors. .
Whether or not recruitment of advisors into the project groups was often
ruled by chance, as observers claim, remains an open question. It can be said,
though, that the administration primarily resorted to the expertise available
in its subordinate government research labs, state administrations, interest
groups and industry. Thus, a communication network was re-activated which
existed already beforehand, between ministerial departments, research departments in government laboratories, industry and technical associations such as
the German Association of Engineers (VDI).
The result of the work done by the project groups was a thick volume
containing the "state of the art" in environmental pollution and measures
against it, but not the transformation of political goals into science policy, let
alone the transmission of scientifically-based interests into policy-making.
Where it assumed that function at all, it was at best on the level of very
detailed problems. In other words, a directing function of the project groups
in terms of derming and structuring the overall programme is not apparent.
This conclusion, which is also supported by interviews with experts during
the study, may be corroborated by another circumstance. The work of the
groups was under extreme time pressure. They had less than three months
for what sometimes amounted to extensive research. More important than the
fact that sometimes the groups had to operate with rough estimates is that
the formulation of the actual 'programme for the environment' took place
parallel to the work of the project groups. Although it is not possible to
reconstruct in detail the utilization and influence of the project work on the
programme, it is indicative that the academic experts had the impression

The Scientific Power Elite - a Chimera

77

that their work had no influence at all; the formulation of the programme
was completed before the report of their project groups. If, nevertheless, it
can be said that parts of the work had an impact, it is probably due to the
high proportion of administration experts in the project groups.
Looking at the composition of the project groups it is worth noticing
the different institutional backgrounds apart from the academic scientists.
One prominent group are the specialists from public administration. These
"experts" participate in policy formulation whenever the issue entails problems of regulation, jurisdiction, and implementation. Here, the Federal and
State administrations are directly concerned and their staff alone commands
the pertinent knowledge. The important function they serve is the integration
of that knowledge with that of the outside experts but their's is strongly
affected by both their departments' jurisdiction, their immediate area of
responsibility and the political strategy of their respective department heads.
A further group involved in policy formulation are scientists who do
not work in an academic setting, but either in government laboratories, big
science installations, or in industry. There is no systematic information to
my knowledge on how the different institutional settings in which they
work affect their perception of problems and the expertise they provide
in advisory capacities other than that they are known to internalize their
institutions' objectives (10). With the few exceptions of basic researchoriented laboratories, it is a common characteristic of the industrial and
government scientists, that their institutional base is non-academic. This
shows that even strictly scientific advice in policy-making is increasingly
provided by a part of the scientific community which has its institutional
base outside the universities and which works for objectives not set by itself.
Obviously, membership in advisory, planning and other ad-hoc groups
employed in policy formulation is only an indicator of the type of knowledge
that serves as an input and of the (institutional) interests that determine its
biases and selectivity. It is possible, however, to generalize on the
of the
material presented and evidence cited (11) regarding the role of scientists as
compared to other professional groups.
Not surprisingly, the degree to which scientists are coopted, and are being
delegated part of the function of problem definition and policy formulation,
depends on the nature of the 'political problem' or 'issue' at stake. In other
words, the more limited the problem, and the more focused on science policy

78

Peter Weingart

aspects, the more important is the role granted to the scientists. As the
scope of the problem extends to other areas of policy-making, and thus
other institutionalized interests are affected (in particular industrial), the
scientists have to share their position with "experts" from industry and
political administration, if they are called in at all. This means that judging
on the basis of institutional affiliation, the influence of scientists on policymaking beyond the boundaries of their competence and, more importantly,
areas of legitimate interests may occur as an exception but can hardly be
claimed to be a rule.
What we observe is rather a configuration of "expert groups" with different
institutional bases participating in the policy-making process which we have
called "hybrid community". (The term was chosen to differentiate it from
the scientific community although it is probably even less a "community"
than the academic scientists.) "Hybrid communities" represent the institutional expression of the increased communication pressures between the
differentiated systems of politics, science and the economy. Their function
is to help defme policy problems in terms of systematic knowledge, to translate (operationalize) them into technical goals, to tum them into research,
strategies, development programmes and correlate policy measures, all of
which feed back into the perception and defmition of the policy problems
themselves. The significance of the "hybrid communities", therefore, lies in
their cognitive function as brokers of expert knowledge and political values.

(b) The Impact ofScientific, Industrial and Administrative Knowledge

To corroborate the conclusion that experts from diverse institutional
grounds shape the defmitions of political problems rather than just scientific
experts, I want to take a direct look at the different types of expertise and
their differential impact on policy-making. Although this is much more
difficult to do and the evidence is more sketchy, of course, it can be expected
that it correlates with the previous result. This expectation is justified because
different institutional settings (academic science, industrial research, public
administration) shape interests, evaluative frames of reference, perceptions of
problems and knowledge. Also, it is safe to assume that the continuum of
possible configurations of types of knowledge having an impact on problem
defmition will closely resemble the arrangements of expert groups. Thus,

The Scientific Power Elite - a Chimera

79

at one end of the continuum strictly scientific knowledge alone defines a

political proble!l1. At the other end, political problems emerge in the political
arena and if scientific knowledge is involved at all, it is instrumentalized
according to political criteria of relevance.
By going back far enough in time we can say that the issue of environmental protection was initiated by alarming reports from scientists (12). It
took a time lag of about ten years and particular political circumstances
before the environment issue became politicized. Once that had happened the
scientists lost control of the course of events. This is very likely due to the
issue being one with a wide range of policy implications. Not only was the
issue politicized by the government but the administration retained the power
of problem definition throughout both by providing much of the necessary
expertise itself and by defming the objectives of the advice sought. The
experts were thus reduced to the classic advisory role: providing knowledge
according to goals defined by policy-makers, or rather translating these goals
into scientific-technical information and research and development strategies.
It can also be shown most clearly how the prevalence of administrative
expertise (under the pressure of political exigencies) affects the definition of
problems. In the earliest phase of internal discussions, the ecology-oriented
'integrated, systemic view' competed with the 'pragmatic, departmental
view'. The latter one quickly gained prominence both because environmental
protection was revealed as a vehicle for the expansionist aspirations of the
Minister for the Interior which could not be realized against the resistance
of the other departments and because it corresponded to the prevailing
structures of communication, jurisdiction and administrative organization of
policy-making. Once the decision had been made on the fragmented structure
of the ad-hoc groups, a systemic view of the problem was ruled out. It was to
be salvaged, half-heartedly, by the integration of the individual group reports
in the end. An attempt to introduce a wider problem perspective, to change
the definition of the problem from one based on administrative departmentalizations and jurisdictional categorizations to a cross-disciplinary and crossdepartmental view based on a systemic ecological approach was made by
the "horizontal" ad-hoc group on the "organization of scientific advice for
environmental protection". This most explicit attempt on the part of (social)
scientists to change the problem perspective failed largely because of the very
conditions it criticized.

80

Peter Weingart

On the level of the project groups, similar examples can be cited. For

instance, the composition of the project group on 'measures against noise'

mentioned above had a specific implication. The combat of noise was not
considered a scientific but a technical-administrative problem, a view partly
explained by the fact that research on noise concentrates on the improvement
of measuring instruments. Medical-psychosomatic noise research is hardly
developed at all, so that existing physical-technical knowledge provides the
more easily accessible base for legislative measures.
The least that can be said safely is this: scientific knowledge may, in
certain cases, be the crucial determinant of the initial perception of political
problems. However, it is extremely unlikely that this 'power of problem
defmition' remains with the scientists throughout the process during which
the problem is being placed on the political agenda and solutions are being
devised. Very soon political and/or economic objectives assume a selective
function vis-a-vis, scientific knowledge and the latter also has to compete
with, and be adapted to, knowledge from other institutional frames of
reference, the administrative-political and the economic-technical. Also, the
impact of scientific knowledge correlates inversely with the political scope of
the issue at hand. As mentioned before, the wider the policy implications of a
particular issue and the farther away from being limited to the science-policy
arena, the less influential is scientific knowledge in shaping the definition of
the problem.
3. The Politicization of Science and its Consequences
Having shown that the scientists are not necessarily powerful when involved
in policy-making I now tum to the second complex: the repercussions of that
involvement, i.e., the politicization of science.
For some time the advent of scientists in the executive strata of governments suggested to political analysts. that politics had become the object of a
general 'scientification'. 'Scientification' may be defmed as the penetration
and eventual disSipation of the guiding images of social institutions by scientific knowledge, theories, concepts, and modes of thought. An example of
this point is law.
institution of law traditionally interprets crime in moral
terms. The social sciences, foremost sociology and psychology, have penetrated the institutionalized link of crime and punishment by providing a

The Scientific Power Elite - a Chimera

81

picture of the social and psychological conditioning of criminal, i.e., deviant,

behaviour. Crime is 'explained' in terms of an individual's biography, which
means that traditional moral convictions are no longer wholly valid in explaining and judging a crime. Problems of social integration, for example,
are treated as scientific problems and have thus been removed from the
mechanisms of the institutions which have dealt with them so far. In this
case, the decision of punishment is transferred from the judge to the scientific
expert or, one could say, is determined less by legal norms than by considerations based on the social sciences. A similar process has taken place with
respect to mental illnesses. As a result of the advances of modem psychiatry,
psychoanalysis and psychology, the once firmly established, borderline
between the normal and the pathological has been shifted (13).
Scientification defmed in this sense has a de-institutionalizing effect.
Mental illness, crime, quality of life, humane conditions of work are not, per
se, subject matters of the traditional organization of science. They become
so only after the old institutions lose their orientating power due to the
diffusion of pertinent certified knowledge and/or a generalized anticipation
of the explanatory and orientating potential of science regardless of whether
this anticipation is justified. Scientification and de-institutionalization thus go
hand in hand, the extension of the boundaries of science and the dissolution
of traditional institutions are mutually implicative. However, the story is
incomplete. It does not account for the repercussions that the conflicts over
the extension of the boundaries of science have for science itself.
It must be remembered that science cannot fully substitute the orientating
functions of traditional institutions. It may clarify issues by reducing their
problematics offact to the underlying value problema tics, but with respect to
value decisions science remains ambiguous. These value issues which are
implicit in the scientized subject matters, are suddenly within the boundaries
of science and have a de-institutionalizing effect on it. To give an example,
research on the carcinogenity of various types of particles in emissions, serves
to clarify the issue of emission control but cannot decide it because of the
risk and benefit assessments involved. But it is precisely because of the latter
that research becomes subject to the political debates of which these assessments are a part. Therefore, the scientification of social institutions, i.e.,
the de-institutionalizing effect of the extension of science, is inextricably
connected with the politicization of science.

82

Peter Weingart

There are some other aspects of scientification besides the ambiguity

of science, vis-a-vis, values which account for politicization. By becoming
closely linked to the pursuit of 'interests' (Le., in policy-making), scientific knowledge is revealed as tentative (whenever judgements are - and
have to be- given prematurely), relative or 'interested' (when alternatives are shown to exist and/or to be excluded). It cannot escape these
verdicts and claim neutrality when the problems to which it is directed,
are obviously political in nature, value laden and sensitive to public debate.
Thus, the application of nuclear physics and the development of a vast
technology has become inherently political; likewise, the development of
genetics and its rapid industrialization as well as its 'promises' have suddenly
turned into a political and ethical issue; the so called 'electronic revolution'
brought about by the advances in solid-state physics and their applications
to the data processing and communications technologies has proved its
inherently social and political impact by forcing governments to pass regulative legislation to cope with fundamental structural changes in industry
and to support (if not launch) massive attempts to improve the educational
level (and capacity to control) of those who have to handle the new
technologies.
Another aspect of the duality of scientification and politicization is
the development of scientific knowledge into areas where it conflicts with
socially-held values and ethical convictions. Research subjects over which
public debate has arisen are almost all in the biological, biochemical and
behavioural fields of science (leaving aside technological areas such as nuclear
energy, which can, of course, be traced back to nuclear physics, but is a
different issue). And within these fields the advancement of knowledge has
irreversibly reached man as its subject. Behaviour modification through
chemical, electrical and psychological intervention, manipulation of genetic
properties, and experiments on human fetuses are cases in point. Although it
is not altogether clear yet, which values are all involved and where exactly the
line of conflict is located, it is evident, for example, that the subjective rights
of human participants are a defmitive boundary. Nobody can be subjected to
tests without his voluntary and informed consent. Pursuit of knowledge
cannot justify an intervention into the life and health of test persons. And
although existing legal provisions are more unequivocal than the boundaries
set by general moral principles, the extension of research into areas involving

The Scientific Power Elite - a Chimera

83

human subjects eliminates the very value-neutrality that is the basis of the
legitimation of free inquiry (14).
The same phenomenon occurs with the extension of research into areas
which are considered to be risky to human life. While this is, of course, a
familiar experience with respect to technologies (whether derived from
scientific research or not) and has led to prolonged attacks particularly on
technology in the case of nuclear energy, it is more recent in relation to
scientific research proper. So far, very little is known, how a general awareness of risks of certain types or research crystallized into a political issue. In
the case of the DNA-controversy it has been shown that it originated among
biologists in a technical discussion over safety, only then developing in to
a debate over values and was finally extended to a political discussion of
authority and trust (15).
There are numerous indications that the politicization of science is becoming a prevalent phenomenon. One is the recurring appearance of "political"
movements within the scientific community which can be summarily identified with a deepening division between the establishment and the dissenters.
The first incident of this kind, perhaps, was the conflict among the "atomic
scientists" over the use of the A-bomb (16). Since then other issues have led
to a politicization of science in the same manner: the involvement in the
Vietnam war, Fluoridation, environmental protection, nuclear power and
genetic engineering, to name the most well-known.
The pattern of the emergence and eventual disappearance of the political
movements within science, together with their respective political issues, is
fairly regular. As an issue begins to crystallize, two factions of scientific
experts emerge, aligning themselves with the 'non-scientific' groups taking
part in the debate. There are those who take sides with the 'establishment'
- usually government, industry, and military or all three of them - and
there are the dissenting, anti-establishment scientists who challenge the
establishment's position. The latter usually tum to the general public, seeking
to become advocates of its interests and mobilizing it for support. What
follows is an extended and often heated public debate which reveals an often
unbridgeable dissensus among the scientific experts over the interpretation
of the pertinent knowledge and of the interests and implicit political and
ethical convictions guiding these interpretations.
What, then, are the consequences of this politicization of science? Within

84

Peter Weingart

the scientific profession politicization is identical with the loss of professional

consensus, which may develop into a veritable threat to the profession as
such. If political convictions become explicitly an input into the internal
discourse over choice of problems, legitimation of research projects, basic
assumptions, interpretation of results and definition of standards, science
runs the risk of losing the very characteristics which constitute it as a special
sort of profession: its autonomy in determining the legitimation of problems,
the assumptions guiding research, the interpretation of results and the setting
of standards. Taken altogether, it runs the risk of losing the chief rationale
for being granted this autonomy, namely its claim to providing objective
knowledge.
The consequences for the relation of the scientific profession to society
are apparent in the diminished authority of scientific experts which, of course,
hinges on the public's acceptance of the claim to objectivity. Science, which
is special among the professions because it does not have any institutionalized
clientele, becomes identified with 'political clienteles' (proponents and
opponents) and, by defmition, loses objectivity. The most pertinent case in
this respect, the debate over nuclear energy and related issues such as riskassessment, reveals the mechanisms involved. On the basis of identical data
concerning safety, profitability and control, experts reach divergent conclusions. The lay public and political decision-makers who depend on expert
judgement, have nowhere to turn. The experts lose their orienting function.
They may even tum to the public for support of their convictions just as their
political division makes them vulnerable to being utilized for political ends.
Experts become selected according to their known political convictions,
rather than their expertise, to serve on hearings which are designed to prepare
and legitimate political decisions (17). This not uncommon pattern leaves
hardly any doubt as to where the power to define problems resides.
4. Conclusions
The two developments in the relation of science and politics which I have
illustrated: the diversification of the institutional base of the production and
diffusion of systematic knowledge (to distinguish it from scientific knowledge)
and the politicization of the scientific profession, are both aspects of what I
want to call the de-institutionalization of (academic) science. It is this process

The Scientific Power Elite - a Chimera

85

which is being overlooked by all those analysts who believe that the scientists
as a group, an establishment or elite or as a "new class" rise to political
power. They fail to recognize that by virtue of its very success, science is
being transformed. In providing knowledge that is instrumentally superior to
all other kinds of knowledge (e.g., experience, amateur knowledge, tradition
and belief) and in becoming involved in the defmition and solution of societal
and political problems, as an institution it loses its boundaries and, thus, its
identity. While, indeed, theoretical (or systematic) knowledge has become the
strategic resource of the post.industrial society, as Bell claims, this implies
neither an undue political power of the scientists, nor the disappearance of
value conflicts, political strife and conflicts of interest, let alone the achieve
ment of a higher rationality based on scientific knowledge.
Notes and References
1. Among others of the titles resounding this theme more than a decade ago were:
D. K. Price, The Scientific Estate, Cambridge, 1967; R. Gilpin, R. Wright (eds.),
Scientists and National Policy Making, New York, 1965; R. E. Lapp, The New
Priesthood, New York, 1965; S. A. Lakoff (ed.), Knowledge and Power, New York,
1966. That this tradition of thought continues to attract sociologists and social
critics is documented in A. Gouldner's The Future of Intellectuals and the Rise of
the New Class, London, 1979.
'
2. Dan Greenberg aptly observed: "The question ... isn't whether the scientists have
found a place in the upper councils of government ... Rather, the question is
whether the scientists' ... new role has swamped the traditional political process,
or whether we have been afflicted by some confusions between' presence and
power" (D. S. Greenberg, 'The Myth of the Scientific Elite', The Public Interest, I,
1965, p. 53).
Considerable confusion also results from the failure to differentiate between the
scientists' influence in 'policy for science' and 'science for politics'. Examples are
abundant, cf. among others, R. C. Wood, 'Scientists and Politics: The Rise of an
Apolitical Elite', in: R. Gilpin, R. Wright (eds.), op. cit., p. 55; A. M. Weinberg,
'Criteria for Scientific Choice', in: Knowledge and Power, S. A. Lakoff (ed.), op.
cit., p. 409; D. K. Price, op. cit., pp.12 and 97.
I also miss the needed clarity in N. Elias' notion of the power of scientific
establishments being derived from the monopolization of the means of orientation
- as he acknowledges himself - but, of course, the basic idea is the same as my
own. Cf. N. Elias, in. this volume (pp.41, 43, 45 ff.).
3. R. E. Lane, 'The Decline of Politics and Ideology in a Knowledgeable Society',
American Sociological Review 31, 658 (1966).
4. D. K. Price, op. cit., pp. 67, 68.

86

Peter Weingart

5. As to the truth content of the scientists' representation of their own interests

cf. P. Weingart, Die amerikanische Wissenschaftslobby, DUsseldorf, 1970. D. S.
Greenberg, The Politics of Pure Science, New York, 1967.
6. Concerning the hopes, cf. D. Bell, The End of Ideology, Glencoe, 1960; D. Bell,
'The Post-Industrial Society: A Speculative View', in: Scientific Progresg and
Human Values, E. and E. Hutchins (eds.), 1967, pp. 154-170; The Meallurement
of Knowledge and Technology, Indicators of Social Change, E. J. Sheldon and
W. E. Moore (eds.), New York, pp. 145-246; R. E. Lane,op. cit., pp. 649-62; H.
Schelsky, 'Der Mensch in der wissenschaftlichen Zivilisation', in: Auf der Suche nach
Wirklichkeit, H. Schelsky (ed.), DUsseldorf, 1965, pp. 439-80. Concerning the fears,
cf. Note 1; cf. also L. Winner, Autonomous Technology, Cambridge, 1977, p. 150.
7. One apt example is J. K. Galbraith, who analyzed the emergence of the 'technostructure' and then took recourse to the 'scientific estate' in its traditional form as
providing a set of correcting values. Cf. J. K. Galbraith, The New Industrial State,
Boston, 1967. The literature which does analyze changes in the institutional fabric
of science, i.e., its ..industrialization.... has a different thrust. Cf. Jerome Ravetz,
Scientific Knowledge and its Social Problems, Oxford, 1971.
8. For the variety of cases, cf. the case studies compiled by W.. van den Daele, W.
Krohn and P. Weingart (eds.), Geplante Forschung. Vergleichende Studien aber den
Einfluss politischer Programme auf die Wissenschaftsentwicklung, Frankfurt, 1979.
Although, as the subtitle states, the goal of these studies was to determine the
effect of government programmes on scientific development it was in their context
that we Illst coined the term 'hybrid communities'. On the influence of scientific
knowledge on governmental problem perception, cf. the preliminary results of the
case on environmental protection, P. Weingart, 'Science Policy and the Development
of Science', in: S. Blume (ed.), Perspectives in the Sociology of Science, Chichester,
1977, pp. 51-70. An abbreviated version of that case is also in van den Daele et al.,
op. cit. The full length study to which I refer, appeared as G. KUppers, P. Lundgreen
and P. Weingart, Umweltforschung - die gesteuerte Wissenschaft?, Frankfurt, 1978.
9. This consideration has an enormous importance for the selection of experts. One
example among many is the composition of the German Atomic Energy Commission, an advisory body which until 1971 determined the government's nuclear
energy policy. Conceived as a 'self-administrative' body it comprised the clientele
of nuclear energy policy: the reactor industry and the nuclear physics community
both from government laboratories and academia. Internally both groups neatly
divided the pie: formally and de facto industry dominated the commissions concerned with technical, economic, financial and social (!) problems, the scientists
took charge of the commissions concerned with research, training and radiation
protection. Cf. H. Kitschelt, Kemenergiepolitik - Arena eines gesellschaftlichen
Konflikts, Frankfurt, 1980. Kitschelt's very detailed analysis of nuclear energy
policy in Germany, although pursuing another objective, provides further evidence
to this point.
10. Evidence of this is reported in several studies, cf. N. D. Ellis, 'The Occupation of
Science', in Sociology of Science, B. Barnes (ed.), Harmondsworth, 1972. pp. 88105; R. C. Krohn, The Social Shaping of Science, Westport, 1971; R. W. Avery,
'Enculturation in Industrial Research', IRE Transactions on Engineering Management, vol. 7,1960.

The Scientific Power Elite - a Chimera

87

11. Cf.note8,above.
12. For the involved story about the crystallization of this issue, cf. Kiippers et al.,
op. cit.
13. Science, 181 (Sept. 23, 1973), p. 113.
14. Cf. also J. Katz, Experimentation with Human Beingr: The Authority of the
Investigator, Subject, Professions, and State in the Human Experimentation Process,
New York, 1972, esp. p. 312.
15. Cf. D. Nelkin, 'Threats and Promises: Negotiating the Control of Research',
Daedalus 107,2 (1978),200.
16. Cf. A. K. Smith, A Peril and a Hope, The Scientists' Movement in America 194547, Chicago, 1965.
17. For a very detailed analysis of the structural patterns of this process, cf. H.
Nowotny, Kemenergie: Gefahr oder Notwendigkeit?, Frankfurt, 1979. Also B. J.
Culliton, 'Science's Restive Public', Daedalus 107,2 (1978), 147-56.

THE HALLMARKS OF SCIENCE AND SCHOLASTICISM:

A HISTORICAL ANALYSIS*

HOMA KATOUZIAN

University of Kent at Canterbury

1. The Self-Centric View of SCience

Our views of science, scientific communities, and their methods and results
are strongly influenced by the experience of the last few centuries, that is, by
the scientific, industrial, social and technological revolutions of this period of
human history. We therefore search for their characteristics, methods and
results from the vantage point of this experience (1). The fact that, in general,
science and scientific knowledge tend to be strongly associated, even identified, with modern technological progress and products is - I think - an
aspect of this broader view of science: there is only one specific philosophy
of science - appearing in the (marginally different) guises of pragmatism,
instrumentalism and conventionalism - which explicitly relates the truth
of scientific theories purely to their 'success', that is, with the growth of
technological (and, probably, social) utility; and, what is more significant, it
implicitly refers not to a vague and general but to the specifically modern
notions of social and technical usefulness, the specific types of technology
and social utility which are associated with modern living. Yet, although - as
a theory of scientific knowledge - this view has had many influential rivals, it
is, in practice, the most popular vision of science, even among those scientists
and academics who, in theory, might subscribe to other views of scientific
knowledge.
This, however, is only one aspect of the much wider phenomenon which
I describe as scientific self-centricism: the fact that our vision of science its logic, its methodology and its sociology - tends to be restricted to our
own contemporary understanding of science, scientific institutions as well as
their technological consequences. It is a self-centric view of science both in

89
Norbert Elias, Herminio Martins and Richard Whitley (eds.) , Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 89-109.
Copyright 1982 by D. Reidel Publishing Company.

90

Homa Katouzian

time and in space: in time, because - notwithstanding (sometimes) significant

differences on detail - it tends to associate all aspects of science with the
recent and contemporary experience, which it further assumes to be 'the
same as', or 'the mere continuation of' the Renaissance and post-Renaissance
scientific tradition; in space, because it tends to be almost exclusively based
on the scientific history of Europe.
This scientific self-centricism, I believe, has had some eventful consequences both for science and for society (the latter of which I cannot take
up here), including the danger of an effective loss of the most fundamental
characteristics - the hallmarks - of science from our scientific efforts and
our theories of scientific development. Hence, the usual references to 'the'
scientific method, 'the' method of the natural sciences, 'the' sociology of
science, even 'the' structure of scientific revolutions: perhaps even more than
describing the uniqueness of a rule or theory, the definite article implies the
uniqueness of 'the' science which we have in mind. And the fact that, while
there are many such 'unique' (yet conflicting) rules and theories, the point of
departure as well as reference of many of them is the recent and contemporary science, which we believe to be essentially the same as the Renaissance
and post-Renaissance scientific phenomenon. To explain the point by the aid
of a simple metaphor, even though we may have different theories about the
elephant, we are nevertheless thinking of the same animal. The question
which I am posing, however, is whether or not the animal which we all have
in mind is, in fact, the elephant.
2. Science and Scholasticism in Time
It is commonly agreed that there is a basic conflict between 'scholasticism'
and 'science'; but it is not clear precisely what factors, characteristics, etc.,
account for this fundamental conflict. There have been scattered references
to monks and 'schoolmen', religious beliefs and metaphysical systems, methodological obscurantism and technological backwardness, and even, occasionally, specific problems tackled by medieval thinkers which we regard
as being simply ridiculous. Yet these and similar observations do not, in
themselves, provide :us with a basic line of demarcation between science and
scholasticism: in late medieval Europe when scholasticism was experiencing
its heyday, technology was in many ways more advanced than in ancient

The Hallmarks of Science and Scholasticism

91

Greece or Rome; religious beliefs and metaphysical systems have not been
peculiar to medieval Europe or European scholasticism (we can once again
think of classical Greece and Rome, among other times and places); and
'ridiculous' or 'irrelevant' topics of research and discourse are familiar from
all other ages, not least our own, even though their specific features, formulations and 'solutions' have inevitably varied from one to another.
The absence of a clear, explicit and precise line of demarcation between
scholasticism and science is, I believe, due to the scientific self-centricism
which I mentioned in the previous section. For, once we identify science
with the Renaissance and post-Renaissance experience which we think has
essentially remained the same until this day, it will be automatically assumed
that 'scholasticism' - i.e., that which was confronted by, fought against, and
eventually overthrown by the Renaissance and its aftermath - was also an
essentially homogenous phenomenon, a unique (even though long) historical
experience, and a thing of the past. In other words, the implicit line of
demarcation between science and scholasticism is neither analytical nor sociohistorical; it is, in fact, chronological: scholasticism is medieval, and science is
classical and modem! This, however, begs the original question.
I have already pointed out that our references to the specific methodological, technological, or even substantial characteristics of medieval scholasticism
do not, in themselves, make up a clear mark of distinction between scholasticism and science. As for the purely chronological line of demarcation, it will
be sufficient to pose the following questions to indicate its lack of sufficiency:
if we believe in a progressive theory of history, then it is surely odd that a
thousand years of social development (however gradual as well as discontinuous it may have been) should have been accompanied with a thousand years
of non-science, even anti-science; if, on the other hand, we hold a cyclical
theory of history (or, for that matter, no theory of history at all), then there
is no reason why we should think of scholasticism as a uniquely time-bound
phenomenon.
The key to the problem is I think to be found in our implicit, even unconscious, counterposition not of the experience of a thousand years of medieval
against several centuries of modem learning, but of the late medieval scholasticism against the early modem (i.e., the Renaissance and post-Renaissance)
knowledge and science, which - as a matter of historical fact - co-existed
over a long stretch of time: the dawn of modem science predated the

92

Homa Katouzian

Copernician Revolution; the dusk of medieval scholasticism fell at some time

around the French Revolution. The hallmarks of both science and scholasticism are to be found in the process of the struggle between them: they are to
be found not so much in what they achieved, but in how they achieved it, not
in what they grasped, but how they groped for it, not in their static success
or failure, but in their dYrIIlmic progress or decay.
The quality of medieval scholasticism does not date back to the early
medieval scholars who are credited - or discredited - with laying its foundation stones. For (regardless of their ideas and beliefs) these sages and scholars
had themselves been involved in a struggle against a statically rich and developed, but dynamically decadent and declining scientific framework. The
reality of classical knowledge and science at the time was no longer the same
as we have in mind when we think of the moments of its dynamic glory. On
the contrary, it was being plagued with many of those very characteristics
which we tend to attribute to scholasticism itself: it was divided, even dismembered, demoralized, and - in spite of the body of past knowledge on
which it could draw (or perhaps because of it) - it was increasingly involved
in intellectually meaningless and socially irrelevant 'puzzle solving'. Together
with its socio-political counterpart, it lost the game to forces which were less
developed, less sophisticated, but more dynamic, mobile, bold and daring
(2). They both fell not to a more sophisticated internal force which is likely
to have strengthened their basic foundations and broader frameworks, but
to external forces which were, at great risks, determined to uproot those
foundations and break these frameworks. Adam Smith describes the fall of
the Roman Empire and - especially - the succession of European feudalism
as a 'great catastrophe'. But 'there was logic in the madness', and - I believe
- it was avoidable.
Those scholars, sages and even evangelists who together provided the static
base - the body of knowledge - of what later became the medieval scholastic
framework, belonged to this relentlessly critical movement. Their collective,
though not entirely organized, effort led to the creation of a synthetic body
of knowledge (put together from various
cultural and religious
traditions, and consisting of a matrix of physical, metaphysical, social,
political and ontological ideas) which - correct or incorrect - supplied
answers to social and intellectual questions and riddles about life and labour
on earth and in heaven - the kind of answers to the kind of questions which

The Hallmarks of Science and Scholasticism

93

had been once supplied by classical science, the 'heirs' to which were merely
repeating or mystifying, rather than adapting and developing it in order to
meet the challenge of their own time.
Early medieval knowledge was - regardless of its relative truth content or
'technological success' - by no means scholastic: 'it posed serious questions,
even including those about the relationship between man and God (as well as
man and man); it had brought God down to earth (or sent man up to heaven)
but - unlike its scholastic 'heir' - it was not keeping man, as well as each and
every man, 'to his station'; it did tackle many questions about the heavens
above, but it seldom asked 'how many angels could collectively stand on top
of a pin'. And, likewise, its founders and propagators were committed, even
dedicated men seeking and promoting such knowledge without fear or favour;
they, like many of those who were to lead the scientific revolution against
their decadent 'heirs' and degenerated 'legacy', forewent moral and material
comfort, and suffered moral and material persecution for the sake of their
theories, their methods and their principles. 8t Thomas Aquinas may have
been a very knowledgeable, certainly a very clever man; but 8t Augustine is
likely to have been a greater and more original thinker.
3. Science and Scholasticism in Space
I said earlier that our self-centric view of science is both temporal and spatial.
I will now make a few remaks on the spatial aspects of the problem, with a
brief reference to Islamic science and Islamic scholasticism. As it happens,
Islamic science began to rise and (in its own context) attain maturity, at
about the same time as (European) medieval scholarship began to decline and
decay; while, on the other hand, its decline and degeneration into Islamic
scholasticism took place in the successive periods of the gradual emergence
and rapid development of modern European science.
Islamic science and philosophy was, rather like the early (European)
medieval scholarship, a predominantly semetico-Hellenistic synthesis of its
own kind. It extensively drew on the neo-Platonics of Alexandria, GraecoRoman gnosticism and stoicism, Graeco-Persian medical and allied sciences,
as well as a both direct and indirect knowledge of Hebraic, Arabian, Christian
and Iranian metaphysics, politics, ethics and mysticism. It was 'Islamic' both
in so far as it was a direct result of the Islamic social revolution, and in as

94

Homa Katouzian

much as it encompassed various cultures and civilizations between India and

Spain. A distinctive feature of Islamic science and learning (rather in contrast
to early Christian scholarship) was that it was not only - perhaps not so
much - a product of the rise of a new religious doctrine, but an immediate
result of the rapid triumph of the Islamic social revolution, and its establishment over vast territories of land. Indeed, the Islamic Empire reached
the pinnacles of its political, economic and military power at about the
same time as Islamic knowledge and science approached the heights of its
achievements.
After the first three centuries, however, there appear the early signs of
Islamic dogmatism and scholasticism, which coincide with the establishment,
growth and predominance of (intellectually) orthodox schools and colleges,
with a growing emphasis on the static absorption of the existing knowledge,
and concentration on obscurantist puzzles and irrelevant issues. And it was
in this way that it sowed and cultivated the seeds of a self-destruction from
which it has yet to recover: the last great Islamic states - the Ottoman
Turkish, the Safavid Iranian and the 'Moghul' [in fact, Turkic] Indian, which
were founded almost simultaneously - though benefiting from the remnants
of the earlier scientific achievements, failed to recapture their levels or spirit,
let alone surpassing them.
The rise of Islamic scholasticism coincides with the growth of Islamic
mysticism, and an increasing tendency for great men of learning to turn their
backs to scholastic institutions: at first, to leave them, and in later periods,
never to join them at all. For example, the Imam Muhammad Ghazzalf (AD
11 th century) - entitled the Proof of Islam, and generally regarded as the
greatest (committed) Islamic philosopher of all times -left the great Nizamfyi College in Baghdad, repudiated the established orthodox ideas, went home
(in Eastern Persia) to live on agriculture and learn by himself, and refused to
return to a scholastic life even when Nizam al-Mulk, the powerful Vazier and
founder of the College, respectfully asked him to do so. He wrote in reply
that "on visiting the shrine of Abraham ... I took three vows: one was never
to accept any money from any sultan, the other, never to have an audience
with a sultan, and, third, not to partake in scholastic debates." And he went
on to add: "Suppose that as soon as Ghazzalf reaches Baghdad, he receives
the summons of his death. Would it then not be necessary to appoint a [new]
lecturer? Suppose that this is [now] so" (3) Ghazzalfs historical fame is, in

The Hallmarks of Science and Scholasticism

95

fact, due to the works which he produced in this period of self-banishment. It

would be illuminating to reflect on the three practices which he had decided
to repudiate: accepting money and patronage from the political authority,
and being a member of a scholastic institution.
Mawlavi Rumi, the great Persian mystic poet (AD 12th/13th. century)
relinquished and renounced his important theological-academic position at
great material and social costs before producing the mangificent works by
which he is known. Sa'di Shirazi, his outstanding contemporary poet and
writer, and a distinguished scholar and graduate of the Niziimi-yi College,
became a traveller and, later, a renowned poet, writer and thinker who
combined elements of realism, cynicism as well as mysticism, without ever
again assuming an academic position. In his scattered autobiographical
references, he says (among many similar instances) that he was once giving a
lecture to the scholars at the University of Damascus: it was not the scholars,
but a by-stander who was touched by the depth and the novelty of his
analysis! (4)
Islamic science and philosophy was comparable with Renaissance and
post-Renaissance science (even though it failed to reach the same heights of
achievement as did the latter) in as much as it was realistic, speculative and
experimental, and it was socially relevant and technologically useful. For
example, E. G. Browne's Arabian Medicine (so called because nearly all the
great works were written in the Arabic language) contains a considerable
amount of evidence for this in connection with Islamic medical knowledge
and practice (5). My observation about the comparability of the two experiences is, however, related to a much more fundamental point: the fact that
their spirit and outlooks are similar, or, in other words, that they both bear
the hallmarks of science.
Let me take up a single example from Abu Ali ibn Sina, known to his
European readers as Avicenria, the Persian physician and philosopher (AD
10th/11th century) entitled the Chief Master. He is, more than any other
Islamic scientist before and after, the very personification of Islamic science
and philosophy. And his life and times mark (as well as reveal the symptoms
of) the social and intellectual atmosphere which led to the decline of science
and the rise of scholasticism. He was perennially on the move (usually on the
run) and - rather like Voltaire - managed to survive with dignity by crossing
various frontiers at the crucial moment. He once wrote in a poem: When I

96

Homa Katouzian

became great, no country had enough space for me; as my value increased, I
lacked a purchaser.
I refer to Avicenna's voluminous, but relatively unknown, writings on
'scientific method', where he says:
When you are told that an 'Ari! [Islamic gnostic] refrains from having his daily food for
a long time, you should easily accept it and realise that such a thing is a well known
possibility in nature ...

And further
You may hear reports from the [practices] of the 'Ari! community which, being extraordinary, you may reject; for example, such claims as: an 'Ari! asked for it to rain ...
and it did rain ... Reflect, and do not rush into judgement! For such things are part of
the secrets of nature ....

He then proceeds to bring out the significance of these remarks by telling his
reader not to base his attitude on an automatic rejection of whatever he fmds
to be extraordinary, because, he says, he who rejects everything which lacks
an immediate basis in reason or observation is not very different from he who
accepts everything without argument or evidence:
It is therefore your duty to reflect upon all such [extraordinary 1 cases, and tend to

regard them as being within the realm of possibility, until you have better arguments
[than those offered in their favour] for rejecting them ... (6).

The great scientific spirit contained in these simple and clear precepts has,
I believe, never been surpassed; and if it is re-Iearned it will be to the great
advantage of our contemporary science and society. Avicenna was a scientist
and philosopher; he was neither a mystic nor (very probably) a committed
Muslim. And it is in that role and capacity that he is warning his fellow scientists against dogmatism - even including 'scientific' dogmatism. This is an
11 th-century rational philosopher who regards nature as being capable of
phenomena which are not immediately testable by human senses and reason,
emphasizes argument and evidence as the most important means of scientific
investigation, and tolerates metaphysical theories and claims as being 'within
the realm of possibility', not to be dismissed without patient and deliberate
reflection. Compare with the methodological criteria of logical positivism, the
most influential among 20th-century scientists, which regard 'unverifiable'
statements as 'meaningless noises', etc., etc. No wonder that when, a few

The Hallmarks of Science and Scholasticism

97

years ago, it was claimed that a performer could bend forks and spoons by
the touch of his fingers, many natural scientists were shocked into declaring
(for example, in the correspondence columns of The Times) that if this was
true, then the whole basis of their scientific beliefs would have to be false.
Indeed, this small and apparently insignificant event should give one a lot of
food for thought, for a science that is easily frightened by an alleged fork
bender must be in a very peculiar state.
4. Hallmarks of Science and Scholasticism
Medieval scholasticism emerged with the growing 'success', that is, the increasing social and intellectual respectability and fashionability of the new
synthesis. And it became firmly established as a social and intellectual institution, concurrently with the total establishment of the received doctrine
as a body of dogma. This was a social institution both in the concrete and
in the abstract senses of the term: in its concrete sense, it consisted of a
community of scholars in various seats and institutions of learning; in its
abstract sense, it was a rigid framework for the maintenance as well as defence
of the established dogma. The two faces of this institution together made
up the most distinctive hallmark of medieval scholasticism (in fact of any
scholasticism), that is, dogmatism. There can be a significant distinction
between dogmatism and mere belief in a given body of dogma, because (a)
dogmatism is an outlook which may be applied to any body of knowledge,
and (b) it dermes an active commitment to a body of dogma, and it involves
not only a rejection of all alternative systems of thought (both old and new),
but also resistance against any real addition to, or extension of the established
body of knowledge: it fights the 'infidel', hounds the 'heretic' and hunts the
'witch' all alike.
It was this (concrete and abstract) social and intellectual institution which
led to the more specific and more familiar characteristics of medieval scholasticism: a general preoccupation with dynamically useless and irrelevant, but
statically safe and 'solvable' puzzles; a rejection of every source, method and
topic of inquiry other than those contained within the established dogmatic
framework; a dominant (if implicit) view of knowledge and learning in terms
of their 'success' and 'usefulness' - both for the individual scholars and for
those whom they served; the excommunication, damnation and persecution

98

Homa Katouzian

of anyone who - in the very spirit of the accredited founders of the established dogma - dared to break the scholastic rules, and propose new methods
and ideas (not necessarily in conflict with the basic tenets of the established
dogma), or by pointing out the basic differences between the outlooks of the
acclaimed founders of the dogma and the established patriarchs of scholastic
dogmatism. The cases ofWycliffe, Jan Hus and Martin Luther merely provide
well-known 'religious' examples in an otherwise much wider movement both
in quantity and in quality. They generally fought from without the established
framework, and they eventually destroyed it.
The most basic hallmark of science in history is its anti-dogmatism, its opposition not merely to specific faiths, but to faith in general. And for this very
reason, there is no basic conflict between science and religion: anti-dogmatic
science does not have to reject all religious beliefs as nonsense because it has
no claim to total universal knowledge. On the contrary, it is the science which
itself is based on a faith - even if that is the faith in reason - which is essentially in conflict with religious beliefs, for the faith in any religious system
automatically excludes the faith in another. And this is an important reason
why some recent and contemporary scientists would tend to suffer from a
conscious intellectual schizophrenia if they happen to have any religious
beliefs; for they would like to hold on to both their scientific and their
religious beliefs as matters offaith, which are, by definition, contradictory (7).
It may be argued that I have merely produced a 'definition' of science
according to my own value system, and that my 'defmition' has no validity
in historical fact: that 'normal science' is bound to be a dogmatic system of
thought, that it is precisely this dogmatism which accounts for its 'success',
and that the science of which I am speaking is an abnormal, rare, extraordinary or revolutionary phenomenon (8). This argument which, in a different
form, has been very influential among modern academics can be put in a
different way, namely, that scholasticism is a much more deeply entrenched
and tenacious phenomenon than is science. I readily admit that dogmatism
(regardless of the dogma involved) has, up to now, been a more 'normal'
phenomenon than its opposite, just as well as tyranny has been historically
more 'normal' - i.e., much more frequently experienced - than democracy.
But this does not mean that tyranny is necessary for social progress, and that
does not prove that the growth of scientific knowledge is inevitably bound up
with dogmatism (9).

The Hallmarks of Science and Scholasticism

99

The reason why I regard that 'rare and extraordinary' phenomenon as

science, and this 'normal' situation as scholasticism, is that science - the
undogmatic search for knowledge - has invariably led to significant discoveries, to solutions to social and intellectual problems, to real (and sometimes
rapid) additions to knowledge, by means of simple and clear theories, arguments and observations; whereas, scholasticism, at its best, has reaped where
scientists have previously sown, and at its worst invented little irrelevant
puzzles to which it has then supplied complex, mystifying, at times even
incomprehensible 'solutions'. It looks as if scholasticism has been re-named
normal science.
I believe that the confusion is (to say the least) partly due to what I have
previously described as scientific self-centricism; that is, (a) our (more-or-Iess
exclusive) association of science with the modern European science; (b) our
view of contemporary science as the legitimate heir of this modern - Le.,
Renaissance and post-Renaissance - science; and, consequently, (c) our
generalization of the institutional (as well as theoretical, technological and
methodological) characteristics of contemporary scientific frameworks to the
whole of the modern age. It is not science itself, but modern science (together
with its specific theories, methods, etc.) which is a modern phenomenon;
just as it is not scholasticism itself, but medieval scholasticism (including its
specific contents) which is a thing of the past. Throughout history, science
has evolved in terms of its methods, contents, and results; and so has scholasticism. But the basic social and intellectual hallmarks of each of them have
remained the same. This is not only true of science and scholasticism, but of
many other social values and phenomena. For example, freedom has had
various specific implications in different times and places; yet we can easily
identify its defenders and promoters - as also those of tyranny, bigotry,
etc. - throughout time and space. And when we talk of tyranny in a contemporary society, we do not mean that it has exactly the same social,
institutional or technological frameworks as in ancient Syracuse; we are
merely recognizing the basic hallmarks of tyranny itself.
There is nothing 'impersonal' and 'objective' about science itself any more
than there is any impersonality and objectivity in the motive to acquire
knowledge, to promote private and public education, to build business
empires, to achieve economic efficiency and growth, to eradicate poverty, or
to attain 'the greatest happiness of the greatest number'. In fact, it is rather

100

Homa Katouzian

surprising that those who put a lot of emphasis on the social and technological
usefUlness of science, also tend to be the strongest advocates of 'the neutrality
of science'.
However, the argument can be analytically split into two parts. First, that
a critical and undogmatic search for natural and social knowledge is likely to
be the surest and most efficient method of solving technological and social
problems. This is a statement of fact (not value) although it may be incorrect.
Second, that this is what we should do, which is a statement of value, analogous to those concerning the promotion of freedom, democracy, welfare etc.
It is in this sense that the hallmarks of science are unchangeable social norms
and values, even though their specific implications and applications do change
through time and space.
The basic problem with some recent historical generalizations about the
process of scientific research - and, in particular, those of Kuhn and Lakatos
- is that they generalize not so much about the (timeless) scientific vision
and approach which has always been critical, non-conformist and, if you
will, revolutionary, but about professional, institutionalized and established
activities which make up a well-known trade in history: some people used
to become Masters of Arts in order to gain admission into the scholastic
establishments (or receive permission from them for independent teaching),
just as well as others had to become master craftsmen before being accepted
by their relevant guilds. They now become university lecturers and professors,
while the others become skilled workers, managers, etc. Normal science tends
to be a description of the institutional behaviour and 'rules of the game' of
this particular trade - a trade which (as I will try to show in the following
section) made very little contribution to the development of modern European
knowledge and science until the 19th century. Indeed, it would be very
interesting to try and discover how much of the work carried out in those
institutions has stood the test of the time, especially in comparison with what
was produced outside and often in defiance of them.
It is sometimes believed that great scientific achievements must await the
arrival of extraordinarily intelligent and hardworking individuals. This at best,
is a half truth, however. There are always such men everywhere, and - in
particular - it is very unlikely that the European thinkers and scientists of
the past few centuries had greater native abilities than their contemporary
official scholastics. The reason why even the most learned, intelligent and

The Hallmarks of Science and Scholasticism

101

contemporaneously respectable and successful of these scholars did not (in

general) survive beyond their own times, must be the fact that they were
caught within their dogmatic intellectual, social and institutional frameworks.
To go back to our discussion of the concept of normal science, one can
equally plausibly argue that democracy or social justice (regardless of their
specific types) have been rare in history. But this could only mean that
'normal society' - not normal democracy - is undemocratic or unjust. Otherwise, democracy itself cannot be undemocratic, just as well as science itself
cannot be dogmatic, and red cannot be blue. A social revolution becomes
inevitable when all the main gates to social progress are securely locked. The
position of scientific change is somewhat analogous to this: science is not
necessarily involved in a process of permanent revolution in the total and
epoch-making sense of this term; but once the avenues for its real progress are
effectively blocked, then it tends to reassert itself through total revolutions.
Note, however, that my reference is to great revolutions, not just piecemeal
and partial (but real) advances which, if permitted, may make total scientific
revolutions avoidable. Whereas Kuhn's argument is that, not only great scientific revolutions, but any real progress in science (all of which he describes as
revolutions) would be resisted by the institutional framework, the 'scientific
community' and its 'invisible college'. This looks like a reasonable description
of our contemporary situation, and of scholasticism anywhere any time.

s.

Science and Scholasticism in Modem Times

Contemporary science is formally and 'chronologically' the heir to modern

European science. But this is not sufficient for a legitimate claim of descent
from that tradition, if only because formal and chronological right of descent
had also been claimed by Islamic scholasticism from Islamic science, and by
medieval scholasticism from early medieval scholarship. In other words (and
partly on the basis of the foregoing arguments), the fact that an intellectual
framework seemingly 'continues' in the tradition of a historic scientific
achievement - that it apparently accepts its basic tenets and methods and
ceremonially applauds its founders - is not sufficent for it to claim legitimate
descent from that achievement. This, indeed, is a well-known fact, if not
principle, in social theory and history.
The reason why the claim of contemporary science and learning to be the

102

Homa Katouzian

legitimate heir to the Renaissance and post-Renaissance scientific tradition is

normally taken for granted is twofold: (a) the self-centric view of science,
together with its tendency to read history backwards; and (b) the impressive
technological progress of our time. Before discussing these issues, however,
it will be necessary to make a number of observations on the institutional
features and the established mode of behaviour of the contemporary academic
profession. (This will be done as briefly as possible, because I have already
discussed the subject in greater detail elsewhere (10).
There are a number of social and institutional features which give a distinct identity to the modem academic profession: it is a full-time occupation
and the principal (if not the sole) means of the livelihood and social placement
of its members; it is dominated by a small, powerful (academic) elite Thomas Kuhn's 'invisible college' - who dictate the disciplinary 'research
agenda' by virtue of their control over the academic journals, appointments,
promotions, etc.; it contains established bodies of knowledge, in its various
disciplines, which tend to be held and defended uncritically; its main research
activity is based on specialization of a kind which is analogous to that of the
mass-producing modem factory system; its permitted, or respectable, debates
and arguments are characteristically restricted to issues, or puzzles, arising
from such specialized works; and it, therefore, inhibits a wholesome disciplinary approach to social and scientific problems, not to mention effective
inter-disciplinary studies.
It follows that the majority of modem academics, who (very understandably) wish to 'succeed' - that is, to be given moral and material recognition
for their efforts - sooner or later 'lose their illusions', succumb to real
pressures, and end up as more or less 'successful' joints in this machine. And,
in any case, there is no need to enter into earnest scientific debates with those
who try to hold out against all odds: they can be 'disciplined' in various ways,
and their ideas are dismissed with contempt. The modem academic profession
and institutions are a distinct phenomenon of our age; they seem to bear
some of the (timeless) hallmarks of scholasticism, nevertheless.
Scholasticism has two basic hallmarks which are directly related: first,
dogmatism, regardless of the nature, content and history of the dogma in
question; second, institutionalism, irrespective of the formal characteristics
of the institutions involved. By institutionalism I do not mean the mere
existence of scientific institutions in a social and physical sense, but an

The Hallmarks of Science and Scholasticism

103

institutional framework which - in ways similar to what I have described

above - selects its members and their ideas in the interest of its established
dogma. This, of course, is what many social institutions, certainly all private
clubs, do; but they neither claim to be based on freedom of thought and
discourse, nor pretend to exist for 'the advancement of learning'. A closed
scientific institution is one in which power and authority are the most effective determinants as well as deterrents in the pursuit of knowledge. This is
more obviously the case where there is direct political intervention in the
institutions of higher learning. But their greater limitations on intellectual
freedom, and the cruder and uglier punishments which they inflict on the
dissenters have (in an inverted way) a redeeming feature: they tend to make
intellectual dissent socially more respectable and psychologically more
tolerable than is the case when - in the name of 'the advancement of learning'
- seemingly autonomous academic institutions apply their subtle sanctions
against it. My concern here is with these seemingly autonomous academic
institutions, not those which - in many parts of the world - are directly
interfered with by the organs of the state. I would merely observe in passing
that, in this and other cases, injustice is likely to be less effective when 'it is
seen to be done'.
These institutions are a 20th-century phenomenon. They are (directly
or indirectly) the results of the social and intellectual evolution of postRenaissance universities which, in their turn, had evolved out of the medieval
colleges. Modern European science and knowledge grew by and large outside
the established institutions, and was only very gradually assimilated by them.
Those sciences which experienced their revolutionary transformations earlier
than others were predictably the first to be accepted by the universities,
especially as it was easier to emphasize their technological usefulness and
underplay their epistemic implications. There were also differences in the
response of established universities in different parts of Europe to the new
ideas and methods.
Yet, even in the 19th-century, Farady and Pasteur worked outside universities, Darwin, a relatively unsuccessful student at Cambridge, never again
joined a university, and most of the inventive and innovative engineers the Watts, the Telfords, the Stephensons, the Emersons - were not even
university graduates let alone academics. The case of the philosophers and
social thinkers - of the Lockes, the Voltaires, the Ricardos, the J. S. Mills,

104

Homa Katouzian

the Marxes, etc. - is too well known to merit a detailed discussion: modern
philosophy and social science posed a more direct threat to the scholastic
dogma (and its social base) than many of the natural sciences; and their social
relevance as intellectual channels for social change was thought to be not
'useful', but positively dangerous. Adam Smith -- a Glasgow graduate and
professor, and a scholar at Balliol College, Oxford - might appear to be a
prominent exception to this list. Yet, his even-tempered approach to life
and learning notwithstanding, he left Oxford after six years of unhappiness
(without the slightest prospect for the future), and later resigned his chair at
Glasgow (which was an unusually progressive university for its time) when he
was forty (11).
In fact, it is from the 19th-century (albeit with some variations among
different European countries), that universities begin to become centres of
modern learning; from the end of the First World War when they (together
with similar research institutions) begin to become so, more or less uniquely;
and from the end of the Second World War when they (and the other institutions) became the established and 'mass-based' academic power centres as we
know them (12). And although there are no statistics available, there now
seems to be a growing tendency for the departure of some of their actual or
potential members who are ill at ease with their 'rules of the game', either
into other profeSSions, or into alternative (formal and informal) intellectual
institutions. Such a tendency would, in the short run, consolidate their
internal position; but, if there are any lessons to be learnt from history, its
long-run consequences could be of a very different kind.
Up to now, I have been discussing the institutionalist aspects of contemporary science and scientific frameworks. I will now try and explain, and
trace the developments of what I mean by their dogmatism. By dogmatism I
do not mean a belief in any given body of knowledge (or dogma), but an
uncritical commitment to any body of dogma, regardless of its relative truth
content, technological products, or - for that matter - ideological implications. That is to say, I regard dogmatism not as a mere belief in a set of ideas,
which would be different in different ages and among different societies as
well as individuals; but as a unique, timeless, attitude which may be shared
by all of them regardless of time, place or content. Medieval scholasticism
was dogmatic not because of what it believed in, but how and on what basis
it believed in it. To put the point in a different way, I am proposing an

The Hallmarks of Science and Scholasticism

105

analytical (as well as social) distinction between the beliefin something, even
though it may be 'false', and the faith in anything, even though it may be
'true': a belief in 'falsehood' is, by itself, not dogmatic, because it implies the
possibility of modification, correction and even rejection; the faith in 'truth'
is, however, dogmatic, because - by definition - it is final and unalterable. I
have purposefully put the case as strongly and categorically as possible in
order to make my understanding of the concept of dogmatism absolutely
clear. Otherwise, belief and faith are not entirely separable; and if we ever
knew 'the truth', the distinction between them would (and should) completely
disappear.
I now assert that - apart from their institutionalist aspects which I have
already discussed, and in spite of all their obvious and significant differences
which I have repeatedly emphasized - what makes the contemporary approach to knowledge and science (or the new scholasticism) fundamentally
comparable to the old scholasticism is that they are both based on a faith: the
one in God, the other in Reason. The historical process which led to the
decline of the old faith (though not necessarily its associated beliefs) and the
rise of the new, also describes the process of the rise of modern European
science and its subsequent decline (though not necessarily its associated body
of knowledge): the rise and decline of an open and critical search for greater
knowledge of reality which had been in direct conflict with the dogmatism of
its own time.
The first part of this historical process, that is, the war of attrition (although ina series of winning battles) against the old faith, the old dogmatism,
is well known and beyond general dispute; but its second part, that is, the
gradual emergence and establishment of the new faith, has received very little
attention: our scientific self-centricism has automatically excluded it from
explicit discussion; and the avalanche of detailed and extensive debates on
'the scientific method' has effectively buried it under its great heap.
The faith in Reason began to emerge at about the end of the 18th-century,
when the triumphant (social and intellectual) war against the old faith was
nearing its completion. The French philosophes and their sympathizers arid
fellow travellers everywhere, who had played a great role in the closing stages
of this victory, also sowed the seeds of the new faith which, however, can
have been no part of their intention. Voltaire's implicit hints that Christ may
have been a complete fiction, his explicit claim that Muhariunad, the prophet

106

Homa Katouzian

of Islam, must have been a cunning charlatan, and his pragmatic interpretation of religious beliefs - 'if God did not exist, we would have had to invent
him' - are probably the simplest and clearest examples of the approach to
knowledge which would automatically dismiss whatever is not immediately
available to reason and sense-perception as pure lies, inventions and fabrications (13).
To be fair, there was no science of psychology at the time, and theories of
history such as Vico's and Montesquieu's had not managed (and they have
not even yet managed) to temper universalist views of science and society.
Yet we know that, seven centuries earlier, Avicenna had succeeded in piercing
through the veil of dogmatic rationalism, of the faith in reason. There is,
however, a stronger historical explanation (even justification) for the attitude
of Voltaire and the others, namely, the fact that they were still in opposition,
still engaged in an ongoing struggle against the established dogmatism. But
this justification cannot be extended beyond its time-spatial context, for
example, to those contemporary scientists who would be seriously disturbed
by the performance of an alleged fork bender. For Voltaire's intention
been to defeat, not to defend, the established dogmatism.
However, the active and positive cult of worship of reason begins towards
the end of the 18th, and the beginning of the 19th-centuries. Voltaire's (and
the others') faith in reason had been implicit; and it had served as a weapon
in the fight against the old faith. Whereas the faith in reason (regardless
of otherwise significant differences within itself on 'the correct scientific
method' - on Objectivism, Subjectivism, and the like) nurtured and developed
by Condorcet, Laplace, de Tracy, St Simon, Comte, Bentham and others was
active, positive and assertive: it promised - in various, sometimes conflicting
ways - the coming of the Kingdom of Reason; and, both then and later,
it used similar methods to those of the old scholasticism in castigating its
critics. Yet the process was bound to be incomplete so long as the upholders
of this new faith (who, I emphasize, had and still have serious quarrels among
themselves, just as this was true of the adherents of the scholastic faith) were
still 'free floating'; that is, as long as the new faith had not been completely
internalized by scholastic institutionalism. I have already discussed the
process of its completion.
Science and scholasticism are timeless values and, in practice, established
academic frameworks may be based on either of them. The relative truth

The Hallmarks of Science and Scholosticism

107

content and/or the social and technological utility of any body of knowledge
does not, by itself, determine its scientific or scholastic status. The crucial
test is the relative prevalence (or absence) of dogmatism, including an effective machinery for the uncritical defence and preservation of established
doctrines and methods. Science promotes and scholasticism retards the
progress of knowledge.
Notes and References

*
1.

2.

3.

4.
5.
6.
7.

I am grateful to Alex Dolby, Herminio Martins, Helga Nowotny and Richard Whitley
for helpful comments on various drafts of this paper. The views and the mistakes
are entirely my responsibility, of course.
For example, even our observations on the Graeco-Roman scientific experience and
legacy, (which in any case we tend to regard as a preparatory exercise for modern
developments), are usually coloured by our own contemporary experience: ourown
participation in, and identification with contemporary science. What I have in mind
is something like Vico's (and, perhaps Montesquieu's) view that historical events
can be properly understood in their full contexts, and that a mere 'knowledge' of
them is often misleading. See Isaiah Berlin, Vico and Herder, London: Hogarth
Press, 1976: and Against the Cu"ent; London: Hogarth Press, 1979.
Adam Smith once observed that alth<.>ugh China was a richer country than England,
the economic and social characteristics of the English society in the 18th-century
provided it with much better prospects than China. Clarence Ayres, founder of
the American neo-institutionalist school of economics, has associated economic
progress in the whole of human history with a 'frontier' existence and mentality.
By analogy, scientific progress may be the result of a dynamic process of groping
for new knowledge, rather than a static act of grasping it.
See 'Aqilf, AtMr al-Wuzara, Teheran: Muhaddis, n.d., pp. 230-1. The form and
content of both these letters leave little doubt about their authenticity, although
(for a variety of detailed historical reasons) they may, in fact, have been exchanged
between Gazzali and Niziim al-Mulk's son, or his successor Taj al-Mulk.
See his Gulistan for this particular reference, and the whole of his CoOected Works
for the broader argument.
E. G. Browne, Arabian Medicine (The FitzPatrick Lectures delivered at the College
of Physicians, November 1919 and November 1920), Cambridge: Cambridge
University Press, 1962.
See al-Ishamt wa al-Tanbihat (Hints and Warnings], Cairo: Sulayman Duniya, n.d.,
pp. 850-902.
Bertrand Russell's contrast between religion and science implies much the same
thing (i.e., science as dynamic discovery in contrast to religion as static faith)
because he implicitly identifies all religion as dogmatic and all science as critical.
That is, his concept of religion is exclusively that of medieval scholasticism and his
concept of science is exclusively that of the classical as well as early modern science.
He therefore overlooks the fact that official science may become dogmatic, and

108

8.
9.
10.
11.

12.

13.

Homa Katouzian

unofficial religion could be critical. In other words, the wider and more basic
contrast is between dogmatism and criticism, almost regardless of their formal
frameworks. See his Religion and Science, London: Oxford University Press, 1960.
Berlin's reference to 'Averroist lines' - i.e., a complete separability of scientific and
religious beliefs held by the same individual - is of some relevance to my argument,
although I would advocate the possibility of a lack of basic conflict only as long as
the body of knowledge contained within a religious (institutional) framework is not
held dogmatically (there would, of course, have to be a few basic principles which
must be so held, or there would be no religion nor science; but this is not what I
have in mmd). 'Averroist' is probably an adjective of Averros (Ibn al-Rushd) the
great Islamic philosopher, but in any case, many of the great Islamic thinkers held
such a view, which was not usually or predominantly a matter of expediency; for,
in taking this view, they opened up Islamic doctrines themselves to wide interpretation and extensions in the service of intellectual development. See Isaiah Berlin,
Vicoand Herder, op. cit., p. 79.
I am obviously alluding to Thomas Kuhn's The Structure of Scientific Revolutions
(Chicago: The University of Chicago Press, second edition, 1970).
See my critique of Kuhn's thesis in H. Katouzian, Ideology and Method in Economics, London: Macmillan, and New York: New York University Press, 1980,
Chapters 4 and 5.
See further, ibid., Chapter 5.
Adam Smith gave up his eleven-year scholarship at Balliol College, Oxford, after
six years, and returned to his home town in Scotland. His letters of the Oxford
period are highly critical of that experience, and betray symptoms of alienation
and depression. Many years later, he resigned his chair in Glasgow in order to take
up the tutorship of a young nobleman with the immediate prospect of a long
journey to France, and the companionship of the leading French intellectuals, none
of whom were university scholars. He has aired some of his critical views of the
contemporary British universities in the Wealth ofNations. See John Rae, The Life
of Adam Smith, Jacob Viner (ed.), pub. A. M. Kelley, New York, 1965; and, Edwin
Cannan (ed.), The Wealth of Nations, London: University Paperbacks, 1961, Vol. 2.
Book V., Chapter 1, Part III.
It should be clear that the reference here is not to universities as opposed to
other (complementary or competitive) institutions such as the state, independent
and publicly-endowed research institutes, but to all of them taken together as
opposed to unprofessional learning and research by individuals and voluntary
associations through which most of the foundation stones of modern science
had been laid. And to forestall misunderstanding, it should perhaps be emphasized that no part of this paper's argument is intended to mean that knowledge
can be properly acquired via what is usually dubbed as 'pseudo-science' - i.e.,
an undisciplined 'search' in the 'world of reality', often leading to a grand vision.
On the contrary, the point being made is that precisely that which we recognize
as science and scientific progress has seldom flourished in the formal institutions
of learning, and that we appear to be living in an age when, once more in history,
these institutions
to be failing in their primary task of advancing knowledge.
These and other examples concerning 'religious matters' are intended purely for
short-hand illustrative purposes, and carry no other significance. For all we know,

The HaHmurks of Science and Scholallticism

109

Voltaire (and others) may have been right on these specific issues; but, although a
good deal of rational argument and evidence can be brought against them, my
purpose is simply to point out the dogmatism of their outlook and approach, rather
than the falsehood (or truth) of their views.

ADVICE FROM A SCIENTIFIC ESTABLISHMENT: THE

NA TIONAL ACADEMY OF SCIENCES

CYNTHIA HAY

Brunei University, Uxbridge

Much of the literature on scientific advice and on the National Academy of

Sciences takes for granted that scientists have a contribution to make to
policy (1). Critics of the Academy question its suitability for performing this
role, but do not doubt that some such role is incumbent upon scientists. The
characteristic ways in which the Academy proceeds in providing scientific
advice have been described in some detail (2); this essay seeks to account
for these features in the light of the analyses of scientific establishments
developed elsewhere in this volume, in the essays by Elias, Whitley and
Weingart (3). Many of the recognized features of scientific advice, which are
displayed by the Academy, can be subsumed under and explained by the
theory of scientific establishments.
I

The origins of academies of science have been located in the 17th and 18th
centuries, as scientific institutions linked with science as the pursuit of gentlemen and amateurs. The development of science as a profession, which
gathered pace throughout the latter half of the 19th century, was accompanied by the formation of an appropriate institutional basis, in the form of
research laboratories, and the acquisition and expansion of a foothold for
science in universities (4). In this context, the survival and adaptation of
academies of science in the 20th century, as more than vestigial scientific
institutions, attracts attention.
The history of the National Academy of Sciences is part of a broader
history of American science, which has been extensively explored by historians "acutely conscious of the importance of the problem of an elite in a
111
Norbert Elias, Herminio Martins and Richard Whitley (eds.), Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 111-119.
Copyright 1982 by D. Reidel Publishing Company.

112

Cynthia Hay

democracy" (5). Much of the history of 19th-century American science

reflects two themes that relate to this antithesis. First, there were efforts to
provide an institutional framework for science congruent with "best science
elitism", the doctrine that science would best flourish if recognition and
preferential treatment were provided for an exemplary elite, who were the
most productive scientists, in both quality and quantity of scientific work
(6). Second, there were efforts to justify and gain support for science in a
broader democratic context: grounds for justification were frequently stated
in terms of the utility of science. Both these themes can be identified in the
origins and early years of the Academy.
The detailed events surrounding the founding of the Academy in 1863
have been extensively explored (7). There had been considerable discussion of
the pros and cons of establishing a national scientific institution, among a
small group of scientists involved in scientific work for the government. The
advocates of such an institution modelled their aspirations on an idealization
of institutions such as the Prussian Academy of Sciences or the Royal Society,
then in a state of decline (8). The main argument against establishing such
an institution was that it would be politically unacceptable: it would be
"opposed as something at variance with ... democratic institutions" (9). The
argument was broken off when the advocates of an academy of science gained
a senator's support and obtained the passage of a bill which established the
Academy in the closing hours of Congress; the Academy was founded almost
unnoticed (10). The political justification provided for the Academy, and
encapsulated in its charter, was in terms of utility: its availability to provide
scientific advice to the government (11).
The earliest years of the Academy were a struggle for survival. That it did
survive has been attributed" in large part to Joseph Henry's efforts on its
behalf, which had the effect of establishing the Academy as an institutional
base for a highly regarded elite among American scientists (12). The circumstances of the Academy's founding, then, meant that an institution, whose
members were a recognized elite among American scientists, was available to
the government.
In the 1950s, there was an international development of institutions,
sponsored by governments and concerned with science for policy; governments required scientific advice of a different order from that provided by
specialist institutions concerned with aspects of the development of science

Advice from a Scientific Establishment

113

and technology (13). The expansion of the Academy was part of this development. It was fostered by a sequence of energetic officials who were officials
in the Academy, as well as being involved in other institutions in the government and science network.
Academies of science have adapted to the 20th century in a variety of
ways. West German academies of science have undergone a shift, from their
18th-century origins as honourary discussion societies, to undertaking longterm, collective efforts on the behalf of scientists (14): tasks which might be
described as those of a gentlemen's trade union. The Soviet Academy of
Sciences is closely linked with the government; it has survived at the pinnacle
of a hierarchy of institutions for implementing policies for science, subject to
occasional sudden changes in its organization and powers, at the government's
decision (15).
Some of the particular indigenous factors which have contributed to the
adaptation of the National Academy to science for policy can be identified
by comparison with efforts to establish an institution analogous to the
Academy in Japan. Occupation authorities, in setting up the Science Council
of Japan in 1949, "envisaged" a role for it "equivalent" to the Academy in
s'cience and government relations (16). The Science Council never achieved
such a position, for a number of reasons. These have to do with differing
political ambiances, and the related availability of scientists to take an active
part in science policy.
The Japanese scientific community traditionally has been distant from the
government; academics have been largely opposed to the ruling political party
in Japan. Consequently the Science Council took on the role of critic of the
government, rather than that of scientific adviser acting at the behest of the
government. Scientific statesmen have been thin on the ground in Japan, not
only because of the political attitudes of the scientific community, but also
because "collegial" decision-making procedures are customary in Japan,
which reduce the importance of particular individuals (17). The intransigent
Science Council was effectively kicked upstairs, with the establishment of a
more accommodating body, the Council for Science and Technology,in 1959,
with this body the government has maintained its traditional dominance in
science policy decisions.
Contrasts with the political environment of the Academy are marked. The
willingness of some highly-regarded American scientists to seek and maintain

114

Cynthia Hay

close links with government has long been a feature of American science, and
contributed to the founding of the Academy. The political views of American
scientists are various, and not, on the whole, as sharply opposed to the
government as in the Japanese situation. Individual scientists who have been
opposed to particular policies of a government have nonetheless been willing
on occasion to be linked with the government through science advisory
institutions (18).
The make-up of the Academy fosters its position of political cooperation
in government service. The backgrounds of elected members reflect the
characteristics of a scientific elite. The membership is heavily weighted
towards individuals from premier academic institutions, who have flourished
professionally, starting from their early recognition as especially able scientists (19).
The Academy is widely recognized as a conservative body of scientists,
both in its membership, and in the general tenor of its recommendations.
The lengthy and complicated procedures of election to membership tend to
favour non-abrasive scientists (20). The Academy, then, may be described
as an establishment in a classic sense: that is, a body of people integrated into
a network of major institutions, whose attitudes reflect this position. This
description stops short; the Academy as a scientific advisor, can more fruitfully be analyzed as a scientific establishment.
II

The Academy as a scientific advisor illustrates many of the features of scientific establishments which have been analyzed by Norbert Elias, Richard
Whitley, and Peter Weingart,in their papers for this volume (21). It commands
and controls resources for studies of the uses of science and technology,
through its privileged position as a quasi-official body, with a right to private
deliberations and with access to information through its relationship with
government, it commands resources in a different sense, by means of its
prestige, which enables the Academy to attract and to draw on the services
and expertise of scientists.
The position of the Academy as a "Supreme Court of science" whose
activities are largely financed by the government agencies who bring cases for
judgment reflects the way in which the visibility of a scientific elite may

Advice from a Scientific Establishment

115

exaggerate its political power. The Academy provides advice on issues whose
identification as policy problems stems from elsewhere.
The second set of processes which are reflected in the workings of the
Academy as a scientific adviser derive from the development of scientific
establishments. Weingart's analysis of scientific establishments links them
closely with the growth of science for policy, which is the direction in which
the Academy has deliberately sought to develop in recent years; a major aim
of the reforms of the Academy's structure in the early 1970s was to equip the
Academy better to contribute to science for policy discussions.
A number of writers have discussed the characteristic features and problems of science for policy: they describe the locus of scientific advice of this
kind, by terms such as Saloman's technonature and Weinberg's trans-science,
which identify its issues as being of a different order from those of science,
and not susceptible to resolution by scientific means (22). Weingart argues
that issues of this order have emerged in political arenas as a result of two
correlative processes, which he calls the de-institutionalization of science and
the politicization of science. The de-institutionalization of science refers to
the emergence of science as a free-floating approach, which has encroached
upon areas and issues traditionally the province of other institutions, such
as religion or the family; science is seen as a source of prescriptions and
judgments on topics such as child-rearing or healthy diets. The range of
scientific advice has been vastly extended, but the other side of the coin is the
politicization of science: with issues of this order, the grounds for decision
and the powers of decision are not matters of science. The location and
incorporation of scientific advice in policy processes reflects this ambiguous
position, which might be said to be one of authority without power.
The notion that scientists with particular competences have especial
qualities to provide advice on broad questions illustrates the emergence of
the scientific establishment as a locus of science above and beyond the
achievements of particular sciences: the hypostatization of science is its title
to authority. Science has acquired an authority surplus to the achievements
of particular sciences; the aura of the scientific establishment overshadows, in
more than a figurative sense, the lack of adequate criteria or institutions to
resolve cross-disciplinary questions on a level with the procedures adopted
within disciplines (23). The dearth of such procedures is critical in science
for policy. A recent controversy about an Academy report suggests how

116

Cynthia Hay

cross-disciplinary differences may, in practice, be handled in a scientific

establishment.
The report, Towards HealthfUl Diets, was, unusually for the Academy,
aimed at the general public rather than a specialist audience in government
circles. The aim of the report was to provide consistent and scientifically
based guidelines about nutrition to counter public confusion in the face of
conflicting advice from numerous sources (24). The centre of the furore
which its publication aroused was the report's position on dietary cholesterol.
The consumption of foods which are high in cholesterol, such as eggs, butter
and cream, has often been linked, not least by advertising, with heart disease.
The report argued that scientific studies did not warrant any reduction of
dietary cholesterol in the average diet of people in normal health. Another
recommendation in the report was that salt consumption should be sharply
reduced, because of its contribution to high blood pressure and. possible
heart disease. The evidence cited in support of this recommendation might
well impress a lay reader as parallel to that used in connection with the
recommendation that cholesterol consumption need not be reduced. A
notable silence in the report was the omission of any discussion of sugar
consumption and dental health.
Several themes emerged in this subsequent controversy. Much was made of
the fact that financial support for the report had come from organizations
whose food products were known to be high in cholesterol, such as the
National Dairy Council, the United Egg Producers, and the National Livestock
and Meat Board. The connections here were taken to indicate the influence
of industry on an Academy report. Another issue, which is perhaps equally
Significant, was less discussed: the problem of producing scientific advice on
cross-disciplinary topics. Much of the evidence in support of the correlation
of dietary cholesterol with heart disease is epidemiological. The report was
largely written by biochemists who discounted the epidemiological evidence;
biochemists give little weight to evidence establishing correlations, because no
biochemical process has been shown to exist, and have an impact, in connection with dietary cholesterol and heart disease (25). The report represented
the position of biochemists on dietary cholesterol, perhaps even more than it
represented the influence of industrial interests on Academy deliberations;
the biochemists' view was presented as representing scientific advice on diet
and health.

117

Advice from a Scientific Establishment

The report and the surrounding controversy illustrate the cross-cutting of

the correlative processes of the de-institutionalization and politicization of
science. The aim of the report, providing scientific guidance on nutrition
in the face of public confusion, reflects the encroachment of science as an
authority on issues traditionally the provinces of other institutions. Both the
selection of issues for consideration in the report, and the approach to the
issues considered, suggest that the problem for scientific advice, or arbitrating
between different diSCiplines, may, in practice, be dealt with by means of
existing hierarchies among the sciences. In default of scientific procedures for
resolving cross-disciplinary disputes, judgment is awarded to the dominant
discipline, and the results presented in the name of science. The politicization
of science is both internal to scientific establishments associated with science
for policy, and endemic to science for policy.
III

The analysis of the Academy as a scientific establishment illustrates the

disjunction between authority and power which Weingart argues is more
generally a feature of scientific establishments involved in science for policy.
Such scientific establishments differ from the academic scientific establishments which Elias has analyzed, due to this discrepancy. The authority which
an institution such as the Academy has, emanates from the scientific establishment in Whitley's sense. The internal workings of this authority reflect the
structure of the hierarchy of sciences.
The Academy's relative lack of power derives from the enmeshing of
scientific establishments in policy processes. The mirage of the scientific
power elite now appears as a smoke screen, beclouding discussions of science
for policy. The alternative lies in the analysis of the ambiguities of authority
and power associated with scientific establishments.
Notes and References
1. Harvey Brooks, The Scientific Advisor', in R. Gilpin and C. Wright (eds.), Scientists
and National Policy-Making, New York: Columbia University Press, 1964, pp. 7396; Philip Boffey, The Brain Bank of America: an Inquiry into the Politics of
Science, New York: McGraw-Hill, 1975.

118

Cynthia Hay

2. Boffey, op. cit., 1975, Note 1.

3. Norbert Elias, 'Scientific Establishments'; Richard Whitley, 'On the Emergence
of the Scientific Establishment' unpublished paper, 1980; Peter Weingart, 'The
Scientific Power Elite - A Chimera'.
4. Everett Mendelsohn, 'The Emergence of Science as a Profession in 19th century
Europe', in Karl Hill (ed.), The Management of Scientists, Boston: Beacon Press,
1964, pp. 3-48. J. Ben-David, 'The Profession of Science and its Powers', Minerva
10 (1972) 362-83. G. Daniels, 'The Progress of Professionalization in American
Science: The Emergent Period, 1840-1860', Isis 58 (1967) 151-60.
S. Nathan Reingold, 'DefInitions and Speculations: the Professionalization of Science
in America in the Nineteenth Century', in A. Oleson and S. C. Brown (eds.), The
Pursuit of Knowledge in the Early American Republic, Baltimore: Johns Hopkins
Press, 1976,pp. 33-69.
6. Daniel J. Kevles, The Physicists: the History of a Scientific Community in Modem
America, New York: Alfred A. Knopf, 1978, pp. 41,--4.
7. A Hunter Dupree, Science in the Federal Government: A History of Politics and
Activities to 1940, Cambridge, Mass.: Harvard University Press, 1957, pp. 138-48;
Nathan Reingold (ed.), Science in Nineteenth Century America: A Documentary
History, London: MacMillan, 1966, pp. 200-25.
8. Kevles,op. cit., 1978, Note 6, p. 41.
9. Letter of Joseph Henry to Louis Agassiz, 13 August, 1864, in Reingold, op. cit.,
1966, Note 7, p. 213.
10. Reingold,op. cit., 1966, Note 7, pp. 201-4.
11. An Act to Incorporate the National Academy of Sciences, reproduced in The
National Academy of Sciences, Organization and Members, 1979-80, Washington
D. C.: The National Academy of Sciences, 1979, p. x.
12. Dupree,op. cit., 1957, Note 7, pp. 147-8.
13. Ronald Brickman and Arie Rip, 'Science Policy Advisory Councils in France, the
Netherlands and the United States, 1955-77: A Comparative Analysis', Social
Studies of Science 9 (1975) 167-98.
14. OECD, Country Reports on the Organisation of Scientific Research: Germany,
Paris: OECD, 1963.
15. Zhores A. Medvedev, Soviet Science, Oxford: Oxford University Press: 1979,
pp. 72-3,108.
16. T. Dixon Long, 'Policy and Politics in Japanese Science: The Persistence of a
Tradition', Minerva 7 (1969) 426-53; T. Dixon Long, 'The Dynamics of Japanese
Science Policy', in T. D. Long and C. Wright (eds.), Science Policies of Industrial
Nations, New York, Praeger, 1975, pp. 133-68.
17. /bid.
18. E.g., Matthew S. Meselson, in connection with the use of herbicides in Vietnam.
19. Harriet Zuckerman, Scientific Elite, New York: Free Press, 1977, pp. 144-62.
20. Daniel Greenberg, "The National Academy of Sciences: Prome of an Institution
(I)" Science 156 (1967) 222-9; Philip Boffey, 'National Academy of Sciences:
How the Elite Choose their Peers', Science 196 (1977) 738-41.
21. Elias, op. cit., Whitley, op. cit., Weingart, op. cit., Note 3.
22. J. J. Salomon, Science and Politics, London: MacMillan, 1973; Alvin Weinberg,
'Science and Trans-Science',Minerva 10 (1972) 209-22.

Advice from a Scientific Establishment

119

23. Don Price, 'Money and Influence: the Links of Science to Public Policy', Daedalus
103 (1974) 97-114.
24. Food and Nutrition Board, Division of Biological Sciences, Assembly of Life Sciences, Towards HealthfUl Diets, Washington D. C.: National Academy of Sciences,
1980.
25. Jane E. Brody, 'When Scientists Disagree, Cholesterol is in Fat City', New York
Times 17 June 1980.

PART II

ESTABLISHMENTS AND HIERARCHIES IN THE

DEVELOPMENT OF SCIENTIFIC KNOWLEDGE

GIVING LIFE A NEW MEANING: THE RISE OF THE

MOLECULAR BIOLOGY ESTABLISHMENT

EDWARD YOXEN
University of Manchester

1. Introduction
1980 saw the appointment of the distinguished physiologist, Sir Andrew
Huxley, to the Presidency of the Royal Society in Britain and the joint award
of the Nobel prize for physiology or medicine to three molecular biologists,
Frederick Sanger, Paul Berg and Walter Gilbert for their various contributions
to the practice of genetic manipulation. Huxley's selection as the titular head
of the British scientific establishment is evidence of its deep conservatism (1).
The Nobel award illustrates where the newer fields of biology are now moving,
towards lucrative and contentious industrial involvement, and the ruthless
competition of corporate research. Even amongst this tiny group of rewarded
scientists there are some striking contrasts. Huxley comes from an older
discipline than molecular biology and is the scion of an established intellectual
family. As a pure scientist at the head of an- elite institution, he can draw
upon a rich cultural vocabulary to reaffirm a traditional, seemingly autonomous role for science in resolute opposition to the economic imperative for
change. Sanger, a Cambridge biochemist without a public identity except
as the winner of a second Nobel prize, epitomizes productive research in
a patrician academic context. On the other hand, Berg has shown more
Zivilcourage and innovative skill in the politics of science by organizing
an unprecedented moratorium on their new research field as an exercise
in the management of public concern. Finally, Gilbert, who crossed from
theoretical physics into molecular biology, has shown a desire for power and
fmancia! gain from the applications of his research. His involvement with
the newly formed research corporation, Biogen, has led to controversy at
Harvard (2) and reveals an attitude to science quite different in spirit from

123
Norbert Elias, Herminio Martins and Richard Whitley (eds.), Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 123-143.
Copyright 1982 by D. Reidel Publishing Company.

124

Edward Yoxen

Sanger's cloistered curiosity, or Huxley's hereditary self-assurance in the

defence of tradition, or Berg's flexible academic liberalism that will surely
be unable to contain the forces generated by venture capital in the research
lab.
Here then are classical, pragmatic and entrepreneurial forms of scientific
practice and they represent contradictory tendencies within contemporary
scientific establishments. These men are figures in a masque presented to
the general public in which symbols of wealth, authority, arcane knowledge
and power over nature are being deployed to sustain the consensus that
the industrial exploitation of molecular genetics is being carried on in an
acceptable way. At one level the emergence of biotechnology is a political
drama in which some life scientists are learning to assume new public roles,
whilst some remain committed to the more traditional parts.
These remarks are intended to convey the general sense of my analysis of
molecular biology. I want to understand the causes of its growth as a research
field, the nature of its appeal as a mode of analysis, the evolution of its role
within the changing sciences of life in the 20th century and the strength of its
hold over popular conceptions of the natural world. For present purposes I
focus on the conditions that led to the consolidation of molecular biology
and the power of molecular biologists within national scientific establishments. My usage ofthis term follows that of Whitley.
Establishments in the sciences acquire, control and allocate resources for the production,
validation and extension of particular knowledges by awarding reputations to individuals,
groups and employment organizations on the basis of their contributions to the intellectual goals of the organization as interpreted by the establishment (3).

But I also want to stress the importance of the activity which Norbert Elias
describes as "the administration of a specific fund of symbolic representations, which can serve man as a means of orientation" (4). It is precisely
the claim to be able to conceptualize and act upon the most fundamental,
constitutive processes of life, a claim that itself has a history, that has conferred upon molecular biology its power in particular social conditions. The
formation of its establishment has come about, I shall suggest, through a
certain type of interaction between specific conceptions of life and a process
of institutional change in biomedical research (5).

Giving Life a New Meaning

125

2. Changes in the Material Conditions of Research Prior to the Second World

War
By convention molecular biology has a pre-history and a history. I want to

trace the roots of the former back to the second and third decades of the
20th century, to a period of expansion and institutional change in science
and medicine, as the dominant classes of the major industrialized nations
sought to generate the technical expertise required to run the new business
corporations, to administer new government institutions and to supply new,
more scientifically-oriented hospital-based forms of medical care. In America,
because of the opportunities and challenges of its industrialization and the
relative weakness of constraining political and cultural traditions, by contrast
with Europe, this process of educational and profeSSional development was
pursued with particular vigour (6). To this end, industrialists like John D.
Rockefeller, and Andrew Carnegie, created evolving, managed philanthropic
institutions, which were to direct resources into secondary and higher education, public health, scientific and medical research, medical education and
aesthetic culture, according to some kind of very general strategy (7). Their
aim as members of a newly-emergent corporate ruling class with enormous
private fortunes drawn from the oil and steel industries was to transcend mere
charity by creating a scientific philanthropy under the centralized control of
a group of Trustees and Foundation officials. The resulting disbursement
of funds was to produce institutions, skills and policies that would sustain
commitment to the changing social relations of urban and industrial life.
The managed, scientific programme of donation br such Foundations was
intended to stabilize and consolidate the corporate state. In Germany, similar
initiatives were more closely tied to the provision of government funds, as in
the founding of the Kaiser-Wilhelm-Gesellschaft in 1911 and the subsequent
appearance of its various research institutes. In Britain the relative lack of
enthusiasm of the long established ruling elites for science and technology
weakened the effect of any such reforms with the result that when the
Research Councils were set up after the First World War, they were dominated
by a very traditional, academic views of science and medicine, and had not
the money to effect major changes in the structure of research.
American foundations were therefore the expression of a particular view
of society and politics, namely the ideology of progressivism. They have

126

Edward Yoxen

enormous funds at their disposal and from their inception produced changes
in public health, medical education, and higher learning on an international
scale. As organizations they developed through time, passing through phases
of rationalization and reorientation and were organized around particular
management principles. Very soon the expertise on which they drew, and the
resources they developed were such that specific policies could be forced
upon recalcitrant disciplinary or professional groups as, for example, in the
reform of American medical education (8).
In the 1920s, the Rockefeller philanthropic trusts disbursed large sums of
money to build up American universities (9). By 1929 this policy came under
review, and the structure of the enterprise was itself re-cast by Raymond
Fosdick, J. D. Rockefeller's chief counsel, and a single Foundation created
with five separate divisions. Under Fosdick's influence, the strategy of general
support for scientific and professional education was shifted to one of more
concentrated support for specific areas of science, conceived as an increasingly
highly specialized, differentiated and professionalized activity (10).
The Rockefeller exercise in philanthropy had since its inception placed a
major emphasis on scientific medicine and public health. There can be no
doubt that the accumulated experience in the Foundation in the medical
domain, in dealing with eminent researchers as advisers, in setting up major
public health programs in America and elsewhere, in promoting educational
and clinical reforms, in building up medical schools and confronting dissenting
medical groups was an important influence on the outcome of the policy
deliberations of the late 1920s and early 1930s. In 1931, a Natural Sciences
Division was set up, to which a young mathematical physicist, Warren Weaver,
was brought in as Director, who in due course set up a programme of research
into 'vital processes', which subsequently became known as 'molecular biology'. When Weaver arrived, it was a period of tension between the Trustees
and Foundation Officials.
To consolidate his reforms, Robert Kohler has suggested, Fosdick established a Committee of Appraisal in 1934 to review the scientific rationale of
the Natural Sciences Division (11). Simon Flexner, the influential director of
the Rockefeller Institute for Medical Research in New York and chairman
of the committee, though not a Trustee, was critical of the directions and
policies. But Weaver's programme, albeit with some criticism of his support
for work in psychobiology, was endorsed by other committee members who

Giving Life a New Meaning

127

included F. R. Ullie, Dean of the newly formed Biology Division at the

University of Chicago from 1931, and President of the National Academy
of Sciences from 1935 to 1939, W. B. Cannon, professor of physiology
at Harvard, and David Edsall, Dean of Harvard Medical School from 1918
to 1935, a trustee of the Foundation from 1927 to 1935 and a lifelong
supporter of scientific medicine, educational reform and reductionist biology.
As a result of this negotiation, Weaver's plans were approved in a slightly
revised form; and as Kohler has noted,his ability to exert a directing influence
in biology increased (12).
What was being contested here were different conceptions of the nature
and social relations of research, a contest that was fought out in the terms of
what one can call different 'ideologies of practice' (13). Flexner, a supporter
himself of an earlier phase of reform in the formation of medical expertise
was in favour of a high degree of academic autonomy and opposed to a strategy of concentration and selection of effort into particular areas. Universitybased scientists like Edsall and others, allowed a degree of professional
autonomy to be given up, so that managerial power over science could then
be exerted by officials like Warren Weaver. This shift in power relations and
change in funding strategy was crucial, since, I suggest, the emergence of
molecular biology was in effect a consequence of these developments.
The concern was to fmd an area where significant developments in the
medium term were only likely to occur with external support, where grant
provision could be arranged programmatically, specifically in ways that
reflected the principles of the Foundation and not those implied by university-based discipline boundaries, and where physicalist principles of reliable
knowledge and reductionist methodology could be applied. By the early
1930s the support of research was itself somewhat suspect ideologically, as
an irrelevance in a time of economic crisis and correspondingly the Trustees
were inclined to choose a biological field where help with highly esoteric
investigations could be more easily justified in humanitarian terms. By the
autumn of 1932, Weaver had proposed the construction of a program of work
on 'vital processes'. It was presented as the contribution of his Division to the
overall Foundation concern with a new science of man, based on a kind of
corporalist rationalism and an interest in the principles of general human
behaviour and its aberrations in capitalist society. He clearly had in mind
something different from conventional academic biology.

128

Edward Yoxen

The phrase 'vital process' ... was the label which I wanted to have tied to the programme
because I wanted to get it away from ancient limitations adhering to the word 'biology'
... But for some reason or other, some of my biological friends found this phrase 'vital
process' a little precious. The trustees never quite understood whether they were being
sold a bill of goods or not ... But it was biology, in its purpose; it was biology in its
ultimate orientation. It was not by any manner of means biology in terms of its classical
technique (14).

The aim was to defme a set of phenomena common to all organisms, which
could be analyzed experimentally in quantitative physico-chemical terms, the
understanding of which would throw light on more specific issues like human
behaviour, reproduction or the causes of disease. As Weaver's project gained
legitimacy and coherence, his attitudes and methods changed, and the contribution of his programme to an intellectual transformation in biology became
significant in its own right. As the 1930s progressed, he was able to rely
upon the general support of the Rockefeller Foundation Trustees and a
confidence in his skill and sensitivity granted to him by a network of leading
scientists, whose advice he sought, and to whom he channelled research
funds, often on a scale that no other agency, governmental or private, would
have been able to match. This'is not the place to consider in detail how
Weaver fulfilled this role in science, but it is important to emphasize that
he should be seen as a kind of 'product champion', transmitting ideas and
rhetorics and negotiating between two influential groups, the Trustees and
university scientists (15). Furthermore, he was forced to be selective in his
patronage and became able to enforce and consolidate his intervention in
biology.
So far I have considered the economic, political and institutional preconditions for this kind of development. In the next section I want to consider the
state of technical and theoretical developments in the life sciences by the late
1920s and early 1930s, and why molecular biology specifically should have
resulted from this kind of initiative. We then go on to consider what kinds of
institutions molecular biologists built up in the post-war period, and the kinds
of actions they pursued, having 'established' their science, given its rhetorical
and technical potential as research organized around a new conception of life.
3. Conceptions of Life and a New Strategy for Biology

One of the ideas I want to emphasize is the shifting view of what counts as

Giving Life a New Meaning

129

'biology'. The Rockefeller programme is clearly an attempt to select problems,

experimental systems, methods and research technologies and to co-ordinate
the resulting research; in short to constitute it as a programme. To put this
another way, one could say that the programme represented an attempt to
organize or select certain phenomena in order to do 'biology' on them, in a
new, and somewhat contentious way.
Underlying this change in research goals is a change in conceptions of what
living matter is thought to be, how it operates and manifests the phenomena
characteristic of the living state, how these phenomena are interrelated and
what kinds of investigation can best reveal the organizing relations of living
matter. In this section I want briefly to explore the connections between
these institutional changes and a new conception of life, which accorded the
gene a central role.
Implicit in the theory and practice of any kind of biology is a conception
of life: in the case of molecular biology it has often been made explicit. In the
development of physiology scientists have concerned themselves, sometimes
rather indirectly, with the nature of the processes occurring in living matter,
which maintain it, in specific forms, as living (16). With the development of
theories of evolution in the 19th century, the problem of life has come to be
seen as concerned also with the question, how is it that living organisms are
related to another in ways that imply a common descent through increasingly
complex and differentiated forms and historical emergence from a state
of matter which was not 'living'? Against the background of these kinds
of questions, biological analysis has proceeded to fmer and finer levels of
structure, and, as Francois Jacob has shown, has concentrated successively
upon different modes of integration (17). In the 20th century, most scientists
espouse a primitive form of materialism, or a quasi-materialism which may be
interpreted idealistically, which suggests that the 'problem' of life is located
or immanent in the nature, organization and structure of living matter.
If the particular order of living matter holds the secret of its constitution
as living, then one can ask whether it sufficiently resembles the internal structure or activity of inanimate matter, such that it can be apprehended using
the methods and assumptions of the physical sciences. The so-called 'medical
materialists' of the mid-19th century certainly thought that this was the case:
in the early 20th century mechanists like Jacques Loeb, at the Rockefeller
Institute in New York, also believed very fervently that all the purposive,

130

Edward Yoxen

systemic and self-directing properties of living organisms could be understood

analytically in physico-chemical terms. By the late 1920s, despite Loeb's
influence in creating a reductionist tradition at the Rockefeller Institute,
which had its effect on early work in molecular biology {I 8), some biologists
were sceptical of the value of such drastic simplification {I 9).
Nonetheless, molecular biology was built on the belief that the nature of
biological organization was such that sufficiently sophisticated reductionist
analysis of certain kinds of organizing relations could succeed. The question
to ask, therefore, is what factors promoted the acceptance of a reductionist
programme in the life sciences? I shall suggest that the form of one such
programme originating in genetics was such that it fitted particularly well
into the institutional context described in the last section. In other words
a conception of life - and a strategy for biology - facilitated the kind
of institutional development that Weaver was employed to promote. This
conception of life appeared in the 1920s, and accorded to the gene a central
role in biological processes and a central role for the studies of gene structure
in the life sciences.
By the second decade of the 20th century geneticists were able to offer a
partial but useful account of how it is that successive generations of organisms
from the same species can resemble in each other in law-governed ways.
Formal studies of heredity, began by Mendel and taken up again by Bateson
and others in 1900, sought to explain the transmission of hereditary characteristics between generations, such that specific qualities or types recur and
given patterns of resemblance are created. By adding the idea of mutation to
the original hypothesis of the existence of 'hereditary factors' or genes, one
could attempt an explanation of phenomena such as variation, the divergence
of species and evolution. Indeed the analysis of heredity promised much
needed support for the Darwinian account of evolution by natural selection.
By 1910 some geneticists came to the view that genes could be viewed as
particulate entities, with a concrete existence, linearly arranged on bodies
within the cell nucleus called chromosomes (20). The fruit fly, Drosophila,
offered Thomas Morgan and his colleagues at Columbia University an experimental organism with which to test this idea, and in the course of the next
five years or so they created the so-called chromosome theory of heredity.
From one point of view, however, this highly suggestive line of investigation
was unsatisfactory, it could say little about how genes acted so as to produce,

Giving Life a New Meaning

131

moderate or inhibit specific biochemical processes. Sexually reproducing

organisms had to be considered as machines, which began their existence
through the recombination of two sets of genes present in two germ cells, and
which somehow developed from this basis into adult forms. and somehow
maintained themselves throughout their life cycle. This silence on questions
of development and physiological function led Morgan himself initially to
doubt the value of genetics in biology and other scientists to maintain their
sceptical attitudes (21). But for other geneticists these were not significant
objections, precisely because theoretically informed studies of the gene,
offered insight into crucial biological questions, such as the nature of life.
In several papers from the 1920s, the geneticist Herman Muller argued that
one should consider why it is that genes can mutate, and yet still retain the
property of being able to transmit specific characteristics across generations
(22). Without this ability mutation would be a disastrous process, and could
not be the continuing source of potentially beneficial variation on which
natural selection can act. Muller suggested, therefore, that the structure of the
gene which,in his view,held the key to this property of 'heritable mutability',
could account not only for the phenomenon of hereditary transmission of
characteristics, but for the role of that phenomenon in the entire history of
the evolutionary process. Muller thus reduced the question of 'life' to the
problem of gene structure; those properties, patterns and relations which are
characteristic of the evolutionary process of life coalesce in or are derivable
from the structure of the gene. What is particularly interesting about this
highly reductive idea is that it places a specific problem in a fundamental
position in biology. It organizes biological thought such that one problem can
take on the role of being the key question or central issue, the resolution of
which will open up the secret of life.
Such an important and difficult question could, Muller suggested, be
attacked in drastically simplified instances of living organisms such as viruses,
through interdisciplinary work, by virologists, geneticists, chemists and
physicists, each contributing their insights to the central task (23). This we
might call the Mullerian strategy in biology, and here I follow Carlson and
Roll-Hansen in asserting its importance for the development of molecular
biology (24). Let us note the isolation of one general feature out of the
totality of biological functions and processes as characterising 'life'. Organisms
are effectively stripped bare of everything but their genes. Muller's 'life'

132

Edward Yoxen

science is through its concentration on essence remarkably unbiological. It

represented a conceptual re-ordering of the field of biology in a reductionist
way. By concentrating on general features of living matter and ignoring the
detail and subtlety of organismic function it allowed 'outsiders' either to comment on major questions in biology, as did the physicist Erwin Schr6dinger
in What is Life? (25), to work at a distance on biological problems, as did the
chemist linus Pauling or to become biologists as did Max Delbriick, Francis
Crick and many others (26). It amounted, in effect, to a re-classification of
skills, techniques, methods and problems as 'biological' or as falling within
the domain of biological research. This process necessitated continual negotiation with classical biologists, whose tolerance or permissiveness clearly had its
limits. The crystallographer, W. T. Astbury, for example, despite his world
reputation for work on the molecular structure of a wide range of biological
structures, was not allowed to call his department one of 'Molecular Biology'
but rather one of 'Biomolecular Structure', since the relevant committee
at the University of Leeds felt that Astbury could not be described as a
'biologist' (27).
Let us now go back to Weaver's task in the early 1930s of building up a
co-ordinated set of projects on 'vital processes'. Given the state of biology
at the time it is reasonable to suggest that he could have concentrated his
attention on any of the following areas, as each of them could have been
described as of fundamental significance to biology and medicine: genetics,
biochemistry, microbiology, physiology, evolutionary theory, and embryology. Except for the last two, each of them received significant attention, yet
none of them provided the organizing principles for Weaver's programme.
Although funds went to geneticists, for example at C31tech, its scope in the
1930s seemed very narrow except to those like Muller with an extraordinary
commitment to genetics. Moreover, Morgan, a Nobel prizewinner in 1933,
America's most distinguished biologist and one of Weaver's advisers, was more
interested in neurobiology than genetics by the mid 1930s. Muller himself
went into
exile in Europe in 1932 and soon left for the Soviet Union.
Biochemistry was also supported, but Kohler has argued that in America its
institutionalization as a service discipline in medical schools inhibited imaginative research and contracted the intellectual horizons of its practitioners
(28). Physiology had the necessary depth and its leading practitioners the
vision, but its very solidity as an established diSCipline inhibited Weaver, with

Giving Life a New Meaning

133

his desire to stimulate new fields and new methods (29). Microbiology also
had a close connection with medicine, but in the 1930s could scarcely have
seemed a likely source of general biological understanding. But, at least, in
the very special context of the Institut Pasteur in Paris, an important school
of molecular biologists emerged from this disciplinary base (30). Finally,
evolutionary theory and systematics seemed too classical in approach to
Weaver, too loosely related to experiment and analysis and too distant from
physics. Embryology had flowered in the 1890s, but the 1930s was in difficulties. In any case, Weaver's task as a patron was to evade or transcend the
divisions of existing disciplines, not least because it would shift the balance
of power in his favour, and to strengthen the analytic and methodological
rigour of biological research, according to criteria operating in physics and
chemistry. What he chose to do was to cluster together a set of projects of a
transdisciplinary character amongst them Pauling's structural chemistry of
biological macromolecules, Astbury's crystallography of DNA and protein,
Schoenheimer's use of radioisotopes in biochemistry, Beadle's work in
biochemical genetics, Svedberg's development of the ultracentrifuge, Perutz's
X-ray studies of haemoglobin. The intention was to intensify research, to
allow people to travel to the more progressive labs, to promote the use of
physical theories and methods in biology and to interest physical scientists
in the life sciences (31). The effect was to organize interest and resources
around specific fundamental problems, particularly those concerned with
molecular structure, and to alter the standards of what counted as an explanation in biology away from questions of adaptation, behaviour, macroscopic
form and the process of evolution, and towards functional questions that
were considered in terms of molecular structure. It co-ordinated a set of
approaches to the general question of how to derive biological insights from
knowledge of the structure of biological macromolecules. As the 1930s wore
on, increasing attention was paid to the structure of deoxyribonucleic acid
(DNA), although the bulk of the evidence suggested that the main constituent
of the hereditary material was protein and not DNA. By the late 1930s one of
the physicists sponsored by Weaver, Max Delbrtick, had begun to use bacterial
viruses as the simplest possible system in which to study the mechanism of
gene replication (32). By the late 1940s, Delbrtick had refmed his ideas to a
point where he and his collaborator Salvador Luria, who worked in Muiler's
department in Indiana, were able to establish a research network and training

134

Edward Yoxen

centre that took up the strategy proposed by Muller in the 1920s. In so

doing, the so-called phage group transformed the rationale for studies of
biomolecular structure, by conceiving of genes as informational molecules,
the structure of which held the key to the mechanism of heredity (33).
Weaver's selection of projects and individuals included those who were
directly influenced by Muller's strategy and their work within his programme
had the effect of developing the power of molecular genetics to the extent
that it was able to alter the nature of that programme, so as to bring it closer
to what Muller had in mind. The Rockefeller programme and the developments it set in motion provided the conditions for molecular biologists to
acquire power and status in the post-war research environment. Under those
conditions it allowed them, by utilizing their conception of life, to become
an establishment.

4. Programming a Managed Biology

At the beginning of the Second World War, Foundation officials, like Weaver,
and astute research administrators, like Vannevar Bush, were able to offer the
American government an institutional system for the planning and support of
targeted basic research relevant to military needs. Throughout the war Bush
ran the Office of Scientific Research and Development from his office in
Washington at the Carnegie Institution. The OSRD soon came to include
the Committee on Medical Research (CMR) chaired by the distinguished
physiologist A. Newton Richards. In the course of the war, the Committee
disbursed the unprecendented sum of $25 million, through 593 contracts,
placed in universities, research institutes and medical schools, with the help
of the National Research Council of the National Academy of Sciences, and
51 committees and panels (34).
As the war came to its end various agencies in America began to plan a
post-war research policy, that would be based on a much expanded level of
State support (35). Similarly in Britain, the Royal Society set up a series of
committees to review research needs in a number of areas and influential
figures within it, such as the physiologist A. V. Hill, Biological Secretary
from 1935 to 1945, began to press for more funding to allow new areas like
biophysics and molecular biology to be opened up (36). Not all leading

Giving Life a New Meaning

135

scientists were in favour of expansion and institutional change however. As

Stephen Strickland has suggested:
In the universities, soon to house great laboratories ... research before the war was but a
fraction of normal campus activity. It was an activity, usually small and personal, often
disjointed and sometimes, by some academicians, disdained. If there was a research
establishment ... it was by and large one which was perfectly content with the pre-war
sty Ie and pace (37).

This situation began to change as the more ambitious lobbyists for science
began to exploit the opportunity afforded by the impact of vastly increased
wartime expenditures on science, and the levels of funding increased. In 1944
the statutes regulating the U.S. Public Health Service (PHS) were revised,
which allowed the service to carry out the role played by the CMR. In 1945
the plans for a National Science Foundation in America were drawn up by
Bush and his advisers. In the event, it took a further five years of political
debates for this institution to be established. Stric1dand has suggested therefore that in such a situation expansionist officials in the PHS saw the need to
conduct a campaign of offensive appropriation of research monies, by taking
over in early 1946 the medical contracts placed by the OSRD, using money
made available by the fall in the world price of penicillin (38). Very quickly
the money allocated to the National Institute of Health (NIH) as the research
arm of the Public Health Service began to increase. In fiscal 1947 the figure
was almost $8 million, ten times the level in 1939; in fiscal 1948 it was $26
million. But despite this kind of Congressional endorsement and despite the
involvement of leading academic figures in medical research in the allocation
of State funds, disagreement continued over how research should be organized
and conducted strategically. This was linked to a political struggle over the
plan for a nationalized health care system in the United States. In particular,
the American Medical Association (AMA) used its power to obstruct governmental reforms that would subsidize hospitals, health insurance, and expand
medical education. Given this situation, Strickland has suggested that the
rapid growth of biomedical research support in the 1940s and 1950s can be
explained in terms of the interaction of a number of political lobbies and
factions (39). They included the AMA, which was prepared to accept the
State support of biomedical research whilst vigorously resisting State subsidy
to the hospital system; a private, very wealthy, well-organized research lobby,
with close links with leading NIH officials and influence on Capitol Hill; a

136

Edward Yoxen

bureaucratic elite within the NIH, which was able to expand its own power by
working in alliance for a long while with the research lobby; and a group of
Senators and Congressmen, whose political influence and status derived from
a particular kind of interest in health and its promotion via scientific research.
Thus one can view the expansion of research support in the United States
at the expense of any government intervention into the market for health
care as a resolution of forces, as a policy founded on compromise. One con"
sequence was the formation of a network of laboratory directors controlling
the flow of funds into universities and medical schools (40); they became in
effect a research establishment. Another consequence was the recurrent call
for greater centralized control over expenditure, and the introduction of a
management structure within the granting agencies like NIH and NSF to
achieve more and more precise regulation of specific programmes and policies
(41). Indeed the academic advisers - what Brown calls the 'superacademic
general staff' - have increasingly found themselves constrained by centrally
imposed research strategies (42). This situation has also appeared in Britain
with the initiatives and policies symbolized by the Trend and Rothschild
reports and the continuing attempts to restructure and redirect State-supported scientific research. Throughout the world, the control by research
elites over money, status, and access to research opportunities within an
increasingly specialized set of biological disCiplines has intensified the degree
of competition and assisted the formation of a very esoteric, introspective
research culture. Heirich has described the institutional consequences of these
economic relations upon university research and the fmancial necessity to
form big labs in which hierarchically-ordered researchers produce results
that are continually exchanged for new grants (43). Pickvance has portrayed
the psychological and emotional effects of undergoing and reacting against
socialization into these sorts of roles as a biologist (44). The effect of this
increasingly intense drive for results was to enforce an exclusive concentration
on just these problems, phenomena, or aspects of an organism that ftlled the
needs of specific research programmes. Molecular biologists learnt to take an
increasingly instrumental attitude to the living material with which they
worked and were forced through the pressures of international competition
to intensIfy the degree of specialization in the problem-solving skills required
to stay in the field. As a result, their relationship with, and conception of,
nature changed. Ufe came to be viewed in informational terms.

Giving Life a New Meaning

137

As the crystallographer J. D. Bernal put it in 1967:

'Life is beginning to cease to be a mystery and becoming practically a cryptogram, a
puzzle, a code that can be broken, a working model that sooner or later can be made'
(45).

My suggestion is that it was the mobilization of just such a conception of

life, exemplified by the important successes of the molecular biological
programme in the 1950s and early 1960s - the double-helical model of DNA
structure, the sequence analysis of a protein, the nature and structure of the
genetic code, the structural analysis of protein, the general outlines'of the
mechanism of protein synthesis, and the regulatory function of the operon that allowed some molecular biologists to come together as an establishment
and seek to re-organize the life-science. Indeed the importance of these
achievements is not only that they allowed the extension of a set of theories
in cell biology and genetics, but that they also added symbolic weight to a
specific way of thinking about life. They facilitated a transformation in the
life sciences through a change in the meaning of the term 'life' itself, a change
in the phenomena to which the word is agreed to refer. Against the background of the structural changes described above, I think we can say that
molecular biology played the role of a metabiology, organized around the
idea of a genetic programme. In my analysis so far I have tried to present the
development of molecular biology as the consequence of deliberate attempts
to intensify research on certain questions defined as central to contemporary
biology. Certainly molecular biologists present themselves as occupying a
central role, as being an avant-garde, and as possessing a style and vigour
eminently worthy of imitation by those who are able to do so (46). It may
be, however, that I have simply projected on to the history of biology the
propaganda of a group of arrivistes (47). Nonetheless, one can perhaps explain
for example the tenacity and dogmatism with which certain apparently
general principles, such as the universality of the genetic code or the double
helical model of DNA structure, have been defended, in terms of the insecurity of an initially marginal group attempting to re-organize the conceptual
map of biology, and thereby claim a superior status in the life sciences. The
interesting thing is how hard it is now to dispense with a theoretical idiom
based on the unifying concept of a genetic programme. As Franc;:ois Jacob
has described, it unites two orders or dimensions of biology, the evolutionary

138

Edward Yoxen

and the functional (48). But its power within a theoretical rationale is also
a means to power on a social or professional level. Molecular biology offers
a general language for thinking about disparate biological problems in a
historical context where increasingly there has been a need for research
bureaucracies to order research strategically, to compare different highly
specialized lines of analysis, and manage the production of results. In this
sense a unifying, informational, systems biology serves a managerial research
system. This is not to say that the concept of a genetic programme is only a
social construction, but that its organizing role within a system of metaphors
has been developed through the interaction of the State, the biomedical
research establishment and an interdisciplinary alliance in biology in the
post-war era.

s.

Cultural Interventions: Using the New Media for a New Biology

Finally, I want to return to the ideas of spectable and display with which I
began, since I believe that the rise of molecular biology was, in one sense, a
peculiarly public process. In other words, one part of the negotiation for
status by molecular biologiSts was a more general cultural intervention,
mobilizing public sentiments in support of their conception of life. Indeed it
is interesting to reflect that the post-war period in which television rapidly
emerged as a major medium of mass communication, was also the period in
which molecular biology emerged as a new speciality in the life sciences.
Television offered a new cultural resource to be used in the process of institution building, by playing upon the theme of an attack on the 'secrets of life'
and molecular biologists were quick to seize upon this opportunity. Two
examples are presented here. There are just two instances of a whole genre of
popularization and media reporting about 'The Biological Revolution', that
appeared in the late 19508 and is still around today, in forms that increasingly
stress the technological possibilities (49). Its influence on how the general
public thinks about biology, evolution, genetics and the nature of human
existence has been profound. It amounts to an attempt to inculcate a new
sensibility about the power and role of science.
In the 1950s the BBC began to experiment with science programmes
on British television. The success of some of the early series on science,
such as Your Life in Their Hands which was devoted to surgery, encouraged

Giving Life a New Meaning

139

enterprising producers within the BBC to make more expensive and spectacular programmes at a time when its patrician styles were being challenged
by the new commercial television stations. The fIrst such Science Spectacular,
What is Life?, was shown on the 1st December 1959, presented by Raymond
Baxter and Professor Michael Swann. One of the main themes of the programme that appears in the drafts and planning discussions is the idea that
biologists are facing a redefmition of 'life' through their work in the cell
biology and genetics (SO). This was portrayed by a vast hemispherical model
of the cell, through which Baxter and Swann walk at one point, pointing to
ribosomes and mitochondria and so on. It was a carefully guided tour around
the new world of the cell, and various sections of the journey are overseen by
other scientists, mmed in their labs or with their equipment into the studio.
One has the strong impression, watching the programme twenty years later,
of a kind of cultural alliance behind the programme, between ambitious
young producers in the BBC exploiting the visual possibilities of a new
medium and a new topic for broadcasting and a group of scientists, of junior
professorial rank, recently admitted to the Establishment, doing biology in
a new way at the molecular level and seeing the need to mobilize cultural
support for their work to assist the consolidation of their research programme.
Some of these people went on with these exercises in popularization. John
Kendrew (now Sir John and until recently secretary-general of the European
Molecular Biology Organization) presented a series on molecular biology
in the mid-6Os called The Thread of Life. Lord Swann was until recently
Chairman of the BBC.
In the United States, one can fmd similar instances of this use of a new
resource to develop a new public sensitivity to the power of contemporary
science to analyze and control the fundamental features of living matter,
indeed to the extent that 'life' could be created in a test tube. In 1967 a
group of Californian scientists succeeded in creating an artificial copy of a
virus on a natural template. Their institutional sponsors, Stanford University
and the National Institutes of Health, carefully orchestrated the publicity of
this scientific result so as to maximize its coverage as 'the creation of life in
a test-tube'. President Johnson, also seeking to redeem his political image,
added to the publicity with the result that this item reached the front page
of virtually every U.S. daily paper, in the rather ambiguous terms of 'the
synthesis of life'. In this instance, the deliberate playing on a new conception

140

Edward Yoxen

of life in the media almost jeopardized the results it was supposed to achieve,
namely the increased legitimacy of biomedical research institutions. One can
see the same kind of attempts to exploit the possibilities of media attention
with the 1974 moratorium on recombinant DNA research and the same kind
of ambivalent response. I have elsewhere discussed this incident in greater
detail (51). It provides an interesting example of the highly complex and
culturally variable responses to the deployment of what Elias has called 'a
specific fund of symbolic representations' as part of a strategy for professional
or institutional legitimation.
6. Conclusion

Molecular biology came into existence through the interaction, on the one
hand, of a specific promotional initiative in research, by a private philanthropic foundation seeking to sponsor new developments in biomedical
research and to extend a system of control over that research by transcending
the constraints of existing discipline boundaries, and on the other, a reductionist strategy for biology drawn from genetics, that re-ordered the
conceptual field of the life sciences, so as to place the problem of gene
structure in a central position. This interaction produced a cluster of projects
concerned with the structural analysis of biological macromolecules, which
formed a programme labelled 'molecular biology'. As the programme developed, the genetic elements in it acquired an organizing role at the research
front in biology. In the post-war conditions of rapid institutional growth
the field of molecular biology produced highly significant technical results.
The formation of a research establishment controlling governmental funds
followed from the political and economic conditions that made such unprecedented sums of money available and the ability of molecular biologiSts
to offer a rationale for thinking about and ordering the life sciences and to
mobilize various kinds of support for their conception of life.
Acknowledgements
This paper has been through a number of drafts and in its revision I have been
helped by discussions with a large number of people. I am particularly grateful to Graham Cox, to the organizers of the conference in June 1980 on 'The

Giving Life a New Meaning

141

Recasting of the Sciences between the two World Wars' in Florence and
Rome, to Robert Seidel, Pnina Abir-Am, Richard Whitley, Jon Harwood,
John Pickstone and Barbara Wilkinson.
Notes and References
1. Anon, 'Democracy and the "Royal" (editorial)" New Scientist 88 (4 December
1980) 618; A. Huxley, 'Evidence, Clues and Motives in Science', Times Higher
Education Supplement (2 September 1977) 4-6; R. M. Young, 'Can We Really
Distinguish Fact from Value in Science?' ibid. (23 September 1977) 6;A. Huxley,
'Fact and Value Must Not be Confused', ibid. (7 October 1977) 27.
2. Anon, 'Harvard backs off recombinant DNA' (editorial), Nature 288 (4 December
1980) 423-4.
3. R. D. Whitley, this volume.
4. N. Elias, this volume.
5. These ideas have also been developed in E. J. Yoxen, 'The Social Impact of Molecular Biology' (unpublished Ph.D Thesis, University of Cambridge, 1978); .and in
E. J. Yoxen, 'Life as a Productive Force: Capitalising the Science and Technology
of Molecular Biology', in R. M. Young and L. Levidow (eds.), Studies in the Labour
Process, Vol. 1 (London: CSE Books, 1981), pp. 66-122.
6. N. Reingold (ed.), The Sciences in the American Context: New Perspectives (Washington: Smithsonian Institution Press, 1979); D. J. Kevles, The Physicists: the
History of a Scientific Community in Modern America (New York: Random House,
1979); D. Noble,America by Design: Science, Technology and the Rise of Corporate
Capitalism (London: Knopf, 1977).
7. J. Weinstein, The Corporate Ideal in the Liberal State, 1900-1918 (Boston: Beacon
Press, 1968).
8. E. R. Brown, Rockefeller Medicine Men: Medicine and Capitalism in America
(London: University of California Press, 1979).
9. R. B. Fosdick, Adventure in Giving: the Story of the General Education Board
(New York: Harper, 1962).
10. R. E. Kohler, 'The Management of Science: the Experience of Warren Weaver and
the Rockefeller Foundation Programme in Molecular Biology', Minerva 14 (1976)
279-306; R. S. Seidel, 'The Evolution of Science Policy in the Foundation: The
Rockefeller and Carnegie Philanthropies' Support of the Physical Sciences' (paper
given to the conference on 'Recasting Science between the Wars', Rome, 1980).
11. Kohler,op. cit., pp. 291-6.
12. /bid.
13. L. Hodgkin, 'The Politics of the Physical Sciences', Radical Science Journal 4
(1976) 29-60; see also Yoxen, 1981 (note 5 above).
14. Warren Weaver, Transcript of Oral History Memoir, Oral History Office, Butler
Library, Columbia University (Record No. 343, 3 vols., 1961), pp. 333-4.
15. The term 'product champion' was suggested by Robert Seidel; see note 10 above.
16. T. S. Hall, Ideas of Life and Matter: Studies in the History of General Physiology,
600 BC. -AD 1900 (London: University Of Chicago Press, 1969).

142

Edward Yoxen

17. F. Jacob, The Logic of Living Systems: a History of Heredity (London: Allen Lane,
1974).
18. R. Dubos, The Professor, the Institute and DNA (New York: Rockefeller University
Press, 1976).
19. D. J. Haraway, o-ystals, Fabrics and Fields: Metaphors of Organicism in TwentiethCentury Developmental Biology (London: Yale University Press, 1976).
20. G. Allen, Life Science in the Twentieth Century (Cambridge: Cambridge University
Press, 1978).
21. G. Allen, Thomas Hunt Morgan: the Man and his Science (Princeton: Princeton
University Press, 1978).
22. H. J. Muller, 'The Gene as the Basis of life', Proceedings of the International
Congress of Plant Science 1 (1926) 897-921; reprinted in H. J. Muller, Studies
in Genetics (Bloomington: University ofIndiana Press, 1964) pp. 188-204.
23. H. J. Muller, 'The Need of Physics in the Attack on the Fundamental Problems of
Genetics', Scientific Monthly 44 (1936) 210-14.
24. E. A. Carlson, 'An Unacknowledged Founding of Molecular Biology: H. J. Muller's
Contribution to Gene Theory, 1910-1936', Journal of the Hisotry of Biology 4
(1971) 149-70; N. Roll-Hansen, 'Drosophila Genetics: a Reductionist Research
Program', Journal of the History ofBiology 11 (1978) 159-210.
25. E. J. Yoxen, 'Where Does SchrOdinger's What is Life? Belong in the History of
Molecular Biology', History of Science 17 (1979) 17-52.
26. R. C. Olby, The Path to the Double Helix (London: Macmillan, 1974).
27. J. D. Bernal, 'W. T. Astbury', Biographical Memoirs of FeUows of the Royal Society
9 (1963) 1-36.
28. R. E. Kohler, 'Medical Reform and Biomedical Science: Biochemistry - a Case
Study', in M. J. Vogel, C. E. Rosenberg (eds.), The Therapeutic Revolution
(Philadelphia: University of Pennsylvania Press, 1979), pp. 27-66.
29. Interestingly, in the British research council system physiologists acted as sponsors
of molecular biological projects, which seemed likely to push physiological understanding down to the molecular level.
30. A. Lwoff, A. Ullmann, Origins of Molecular Biology: a Tribute to Jacques Monod
(London: Academic Press, 1979).
31. C. H. Waddington, 'Some European Contributions to the Prehistory of Molecular
Biology', Nature 221 (25 January 1969) 318-21.
32. J. Cairns, G. S. Stent, and J. D. Watson, (eds.),Phage and the OriginsofMoiecular
Biology (New York: Cold Spring Harbor Laboratory 1966).
33. N. C. Mullins, 'The Development of a Scientific Speciality: The Phage Group and
the Origins of Molecular Biology',Minerva 10 (1972) 51-82.
34. J. P. Baxter, Scientists Against Time (1946: Cambridge: MIT Press, 1968).
35. C. Pursell, 'Science Agencies in World War II: The OSRD and its Challengers', in
Reingold,op. cit. (note 6 above).
36. Olby,op. cit. (note 26 above), p. 328.
37. S. P. Strickland, Politics, Science and Dread Disease: a Short History of u.s.
Medical Research Policy (Cambridge: Harvard University Press, 1972), p. 35.
38. Op. cit., pp. 25-6.
39. Op. cit., Chapters 3-7.
40. M. Heirich, 'Why We Avoid the Key Questions: How Shifts in Funding of Scientific

Giving Life a New Meaning

41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.

143

Inquiries Affect Decision-Making About Science', in S. Stich and D. Jackson (eds.),

The recombinant DNA Debate (Ann Arbor: University of Michigan Press, 1977),
pp.234-60.
R. A. Rettig, Cancer CruBllde: The Story of the National Cancer Act, 1971.
Brown,op. cit. (note 8 above).
Op. cit. (note 39 above).
S. Pickvance, '''Ufe'' in a Molecular Biology Lab', &dical Science Journal No.4
(1976) 11-28.
J. D. Bernal, 'Defmitions of Ufe', New Scientist 33 (5 January 1967) 12-14, at
p.13.
H. F. Judson, The Eighth Day of Creation: Makers of the Revolution in Biology
(London: Jonathan Cape, 1979).
E. Chargaff, 'A Quick Climb up Mount Olympus', Science 159 (29 March 1968),
1448-9.
Jacob,op. cit. (note 17 above) ..
E. J. Yoxen, 'The Meanings of Life', Trends in Biochemical Sciences 3 (February
1978), N29-30; for greater detail see Yoxen 1978 (note 5 above), chapter 5.
G. R. Taylor, Outline for What is Life? (Document TEL 1 D421/692/ 23509,
17 June 1958, 3 pages) BBC Archives.
E. J. Yoxen, 1978 (see note 5above), ch. 6; 'Regulating the exploitation of recombinant genetics' in R. Johnston and P. Gummett, (eds.), Directing Technology:
Policies for Promotion and Control (London: Croon Helm, 1979), 225-44; 'Review
of June Goodfield's Playing God', Radical Science Journal, No. 10 (1981),75-84.

TWO SCIENTIFIC ESTABLISHMENTS WHICH SHAPE THE

PATTERN OF CANCER RESEARCH IN GERMANY:
BASIC SCIENCE AND MEDICINE*

RAINER HOHLFELD
Institut {iir Gesellschaft und Wissenschaft an der Universitiit ErlangenNiimberg, Erlangen

I. Introduction
During the last five years, the war against cancer has become a serious issue in
political debates among health politicians, science policy-makers, the medical
profession, scientists and the general public in the Federal Republic of
Germany. The German debate started with a time lag of at least ten years
compared to the U.s. cancer crusade and cancer debates which culminated
in the biggest research programme ever known in the biosciences - the U.S.
National Cancer Program Plan of 1974 (1). The questions in the German
discussion were: are we lagging behind the American effort? What remains to
be done in view of this American crash programme? Does it make any sense
to deal with the cancer problem as if it were a 'biological' moonshot? How
can German cancer efforts best be coordinated? (2) The German government
and science administration's response to these questions was to set the priorities and goals within the frame of the "Programme of the Federal Government for the Promotion of Research and Development in the Service of
Health Care" (3).
By derming the health policy demand for research, this programme gave
rise to intervention in the self-regulation operation within health research and
related areas. Cancer as a chronic disease won high priority within the frame
of the programme. Both basic scientists in biological research and medical
scientists issued strong warnings against planning cancer research at alL
But at the same time, the scientific communities concerned did make some
proposals and recommendations. In view of the requirements to be met by
research oriented towards so complex a goal as reducing the incidence of
145
Norbert Elias, Herminio Martins and Richard Whitley
Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982.145-168.
Copyright 1982 by D. Reidel Publishing Company.

146

Rainer Hohlfeld

cancer in the population, the scientists proposed that comprehensive cancer

centres be set up according to the American model to integrate the entire
spectrum of basic research, clinical research and clinical practice under the
'same roof' (4), as well as the establishment of a German National Cancer
Advisory Board. One leading tumor virologist coined the phrase of 'pluralism
with organization' for the approach that might help to overcome the desolate
state of German cancer research efforts (5)
Discussions about the German programme for health research, and debates about the proposals of the scientists, brought to light those traditional
cognitive and social patterns in the communities and disciplines concerned
which may be held to determine the success or failure of mission-orientation
in the domain of cancer research. Here, as in other fields, the established
structures within the scientific institutions acted as permissive and nonpermissive fIlters which shaped the perception of the problems aM the
defmition of research strategies (6).
To facilitate evaluation of German cancer research in this sense, the various
cancer research fields - different in range of phenomena, in theoretical
maturity, and in degree of complexity - can best be ordered systematically
if the research programme of molecular biology is taken as the frame of
reference.
I will therefore start my argument with a short presentation of the basic
features of this research programme; in the subsequent section, I will describe
the social and cognitive patterns characteristic of the different research areas
which I have investigated, mainly by means of interviews with experts (7),
and show how these impinge on the problem of directing research to achieve
health policy goals, taking the molecular biology research programme as a
systematizing criterion. Finally, I will outline the thesis that the present
German cancer research scene and the interpretation given to the complex
phenomenon 'tumor diseases', are the result of the operation of two centers
of intellectual and institutional power: the molecular biology establishment
and the medical profession.
2. The Scientific Revolution in the 20th Century's Biosciences: the Research
Programme of Molecular Biology

Classical biology in its theoretical terms, conceived structures and functions

The Pattern of Cancer Research in Germany

147

of organisms in the smallest units which could be identified by light microscopy: cells and cell organelles, e.g., chromosomes. For hereditary phenomena,
the Mendel-Morgan chromosomal theory of heredity is still a firm base for
animal and plant breeding today (8).
The theoretical revolution of biological research in our century consisted
in the replacement of the terms of classical biological theories by the terms of
physics and chemistry: atoms and molecules. The goal of this new way of
conceiving living matter is to reduce biological units and entities such as genes
or membranes or cells to chemical terms. Thereby, biological phenomena are
inserted within the range of chemical subjects and biological systems are
conceived as chemical systems. Biology as 'molecular biology' (9) becomes a
particular type of chemistry. In relation to traditional biology, the classical
macrotheories which conceive living matter in units of cells and cell elements
are 'reduced' to or 'replaced' by molecular microtheories (10). The reductionist concept constitutes the paradigm of molecular biology and defines a
researchprogramllle.
The dynamics of the reductionist programme can be characterized by the
intention of molecular biologists to theorize the particular subject areas of
biology, which can be phenomenologically distingUished from one another, in
molecular terms. The various ranges of phenomena can be ordered according
to a
of biological structures: bacterial cell, higher cell
(eucaryotic cell), cell system, organism, population. According to the biological functions, this hierarchy can be defmed as follows: metabolism, heredity,
cellular differentiation and development (11) (see Figure 1).
The reductionist programme started with an explanation of cell composition and the molecular mechanisms underlying metabolism in biochemistry
(12). With the enzyme theory, biochemistry reached the level of theoretical
maturity (13). Historically, this programme was continued by the molecular
theory of the gene - molecular genetics - which started in the thirties and
was accomplished for the molecular genetics of bacteria about thirty years
later (14).
At present, the strategy of molecular biologists is to extend the validity
of the molecular genetics paradigm to biological phenomena of higher complexity than bacteria, i.e., to the molecular genetics of the higher cell and
the processes of cell differentiation and cell development, which take place
in cellular systems (15). I have tried to localize the research frontiers of

I.

contemporary research fronts

cell (e.g.,
sporulation)

differentiation

Integration of
different genetic
systems
Ontogenesis

Morphogenesis,
embryology

Biochemical
integration
(e.g. humoral
regulation of
glucose-metabolism)

Organism

Developmental
genetics

BioChemi!uy of
cellular systems
(e.g. blood cells)

Cellular systems

Phylogenesis

Population
genetics

Biochemical
ecology

Population

Fig. 1. Diagram of Research Fronts in Molecular Biology in Relation to Biological Subject Areas. The Scheme is ordered by levels of
growing complexity of biological structure (from left to right) and of biological function (from top to bottom).

development

Cytogenetics

Genetics of
bacteria

heredity

Cellular
/
biochemistry

Higher cell
(eucaryotes)

Bacterial
biochemistry

Bacterial cell
(procaryotes)

Function

metabolism

Structure

Degree of
Complexity

5:

g:

00

The Pattern of Cancer Research in Germany

149

molecular biology in relation to the distinguishable subject areas of biology

in Figure 1.
The state of the art as embodied in the research programme has different
consequences for the problem of directing research towards biotechnological
or biomedical goals. The research areas in the wake of the research fronts are
reaching theoretical maturity and thus providing the base for developing
molecular 'high' technologies by 'fmalization' of basic research (16). The new
products of gene-technology such as hormones, vaccines, and new possible
anticancer drugs like interferon are the most intriguing examples of this
technological potential (17).
With the advances made by molecular biology during the last decade,
cell systems and human biology have now been brought within the
compass of the molecular biology research fronts. The human genome,
hormonal control of metabolism and behaviour, the molecular mechanism
underlying immunological reactions, the regulatory proteins controlling the
differentiation-systems of skin-, nerves-, and blood cells, are currently key
problems of biological research.
Deviations or disorders within these systems produce phenomena of
disease, "pathologies". According to this concept, cancer can be interpreted
as an error in cell differentiation (18). By molecular biology, the study of
pathological phenomena in humans and their physiological correlates - the
object of medical research - is raised to a high-technological dimension, as
was previously the case for biotechnology. This convergence of molecular
biology with medical research can be defmed as 'biomedical research' (19).
Even the researchers who investigate disease phenomena of high complexity,
like the growth of human cancer cells within the body, or the interplay of
the organism with the social and physical environment, are now starting to
conceive their phenomena in molecular terms, although prematurely and not
yet within the orbit of the research fronts of molecular biology.
3. The Scientific Communities Confronted with the Problem of Research
Demand in the War on Cancer
In this section I will discuss how the scientific communities concerned with
the cancer problem dealt with political research planning on the basis of
their community values and belief systems, as well as their specific theory

150

Rainer Hohlfeld

dynamics and research strategies to the extent that these components are
involved in the cancer issue (20).
3.1. The Ideology of Basic Research: Science Cannot be Planned Towards
Political Goals
Scientists in molecular biology still hold to the idea of science for its own
sake. In their eyes, the laws of scientific progress are determined by the
structure of nature itself, the results cannot be anticipated. Science is an
enterprise of experts and research can be regulated by no one but the scientists themselves. For all these arguments science policy planning is viewed
with considerable scepticism:
Directing research, whatever kind of method used, results in bad research performed by
people who will let themselves be influenced by science policy because they lack any
true scientific motivation (21).

Nonetheless, the belief system of autonomous science offers an interpretation of the guide-line for the utilization of knowledge: First, since the success
of research ca1inot be programmed, society must wait for theories to reach
maturity, whereby a new base is built for technological applications, and,
second, a clear distinction must be made between basic and applied science.
The 'supply-model' of science and with it the dichotomy of basic and applied
science result from this interpretation.
Combining both these elements - the claim that science must be directed
by the experts themselves who are to maintain their authority in the planning
of research goals and that the dichotomy of basic and applied science be
taken account of - the German Research Association, as the spokesman of
the republic of science in Germany, concluded:
Targeted research can only begin when the requisite knowledge has been established on a
flIm basis. Accordingly, the relationship between basic research and targeted research
must not evolve to the detriment of the former. Basic research must be promoted on a
wider scale. This requires additional funding but certainly not research programming
(22).

Cell biology is the field which is thought to produce the knowledge essential to understanding the cancer problem. In the opinion of the biologists, a
precondition for the solution of the cancer problem is the clarification of the
laws of cell biology.

The Pattern of Cancer Research in Germany

151

Basic research, in particular cell biology, must generate the necessary knowledge before
any real breakthrough can be expected to occur. Impatience, no matter how justified
and understandable on the part of millions of cancer patients, should not make either
scientific organizations or the politicians adopt measures which ultimately swallow up
vast amounts of money without bringing any real success (23).

This way of thinking reduces the cancer problem to key events in biological
processes, which must be explained by molecular theories. Only the theoretical breakthrough is seen to provide the rational base for any real improvement
in fighting cancer as well as for a transition from a half-way technology to a
medical high technology. In the eyes of basic scientists the theoretical strategy
in cancer is supported by the fact that 'steel, beam and drug' as the conventional tools in fighting the disease have left research marking time instead of
moving ahead.
3.2. The First Step Towards Application: Experimental Cancer Research

Experimental cancer research is a type of research located behind the frontiers of 'true' science, structured by the still unsolved fundamental theoretical
questions. Scientific activity in experimental cancer research is not determined by the internal dynamics of scientific advance but by goal orientation.
Compared to basic research, experimental systems and research problems are
selected from the perspective of solving problems of practical relevance rather
than that of accumulating pure knowledge. Scientists working in this field,
however do share with basic scientists the idea that health problems of this
type must be solved by scientific instruments on the basis of the clarification
of the underlying biological mechanisms and that this in tum requires 'high
technology' .
Concerning the biology of cancer, we know too little in the largest sense of this word to
be able to work out a rational chemotherapy for cancer, i.e., the pharmaceutic products
we are working with today mean that we are using crude sledge-hammer methods which
kill healthy cells as much as diseased cells. We are not yet able to strike at cancer while
protecting the healthy parts of the organism which we need to protect to maintain life
(24).

The other idea held in common by basic scientists and workers in experimental cancer research, is that work is the enterprise of experts who must
have free choice in selecting the problem areas they intend to attack:

152

Rainer Hohlfeld

There are but a few experts in this field. Who should be entitled to tell us what we
should be doing - policy-makers in the public health system or perhaps journalists who
have taken up this or that trend in this business or computer experts, or anyone else for
that matter? (25)

Unlike the basic scientists, experimental cancer researchers are motivated

by the goal that their research must benefit man in the first place:
Personally I believe, that many of those who call themselves cancer researchers are
actually engaged in a study of rat pathology or mouse pathology and are involved in all
kinds of programmes, theories and discussions, forgetting at the same time that cancer
research in principle is applied research intended to help the patient (26).

Once one goes beyond the level of the common features of experimental
cancer research, it is possible to distinguish three different types of research
enterprises, different both with respect to the maturity of theory development and to the 'intrinsic' or 'instrumental' role of the scientists concerned
(27).

Oriented research in mature areas. Within the combination of mature science

and the intrinsic role, a type of research evolves which mostly resembles
basic research: this is research work in such areas as cellular biochemistry or
chemical carcinogenesis, or parts of viral oncology, and is an example of
the "fmalization" of basic research. Although, as indicated, the scientists
concerned are doing goal-oriented research, they are still guided by the strong
academic bias of their mother diSciplines:
Relevance for human systems cannot become the criterion for the biochemist who is
searching for a model he can work with - we have not reached that point yet. The ideal
model would be one with which one could do biochemical research but which at the
same time would have relevance for human oncogenesis (28).

These researchers, then, are working with the high standards and methodology of the academic disciplines. For this reason this work is not regarded as
being 'bread-and-butter research' .
A structure of the research enterprise which is best described as a specialty
has evolved (29), with such features as its own reward system, communication
exchange, workshops and meetings. The forming of specialties of this type is
exemplified by the. case of research on the metabolism of nitrosamines, a
group of carcinogenic agents synthesized in the acid environment of the

The Pattern of Cancer Research in Germany

153

stomach, which cause cancers of the liver, throat and oesophagus. In the
words of one of the workers in this field:
There are perhaps some 120 scientists working on the nitrosamines. We all know each
other and what each of us is doing. Once a year, or perhaps every two years, we arrange
meetings to debate the relevance of our results and to coordinate our work in a very
informal way. So we do not need any external research programming (30).

As to their thinking about medical scientists and physicians in general,

their recommendation is that they should come into their laboratories to
inform themselves what properties a clinical preparation must have so that
the worker engaged in basic research can do something with it.

Applied research behind the fronts of theory building. When intrinsic scientific motivation combined with the 'external attitude' of goal orientation
occurs in research fields in which the theoretical key questions are still
unsolved - as is the case of molecular cell biology, developmental biology
and immunology - a type of research evolves which most closely corresponds
to 'applied' research in the usual sense of the term. Scientists doing this type
of research try to transfer the progress made on pure systems at the research
fronts to clinically relevant experimental systems, by trial and error methods;
they attempt to elaborate systems which are more complex but immediately
relevant for cancer treatment. Such systems are often deSignated as 'dirty'
systems. This designation refers to systems which are too complex and too
difficult to reproduce to use the highly sophisticated methods of the research
fronts and which therefore offer no scientific breakthrough. This type of
'transfer research' is research behind the fronts of scientific progress (31).
The scientists concerned with this kind of cancer research, and who
are from their training biologists, have split loyalties. The external goal is
convincing and it is easy to identify with it. But cancer research is not frontline research, it has no great status. No clear-cut identification with disciplines
and research fronts is possible here.
'Cancer researcher' is nearly an invective. The cancer researcher is placed between two
worlds. These are the clinic on the one side and prestigeful basic research, for instance
within the Max Planck Society, on the other. The people engaged in basic research will
claim that cancer research is not a solid basis, that such preoccupation with cancer as the
main theme is a restriction of large-scale basic research (32).

154

Rainer Hohlfeld

Many view cancer research as a blind alley, career-wise; it offers few

chances for advancement as there are too few centers where one could
continue one's approaches.
If you have signed up with a cancer research center, you have sacrificed your career. In
that case you have to be enough of an idealist to really want to do something about the
problems (33).

Nonetheless, the scientists concerned will defend their work, or rather its
ranking, by stressing its importance for the patient:
As I see the matter - and this may be a subjective view - decades of biochemical and
virological research involving great expense has in no way helped the patient in the
clinic. As far as I am concerned, therefore, chemotherapeutic research has greater value
than does biochemistry or virology (34).

This interviewee's comments on the problem of the mutual ignorance

prevalent between molecular biologists and medical scientists run clearly
counter to those of the biochemist's:
The researcher working with cells, a research activity which is essential, at some point
must also spend half a year's work in a clinic because otherwise he will never come into
contact with the patient. I am sure that there is many a cancer researcher who has never
in his life seen a cancer, and that is a bad thing (35).

The daily business: strategic planning of biomedical research in the pharmaceutical industry. For researchers in the pharmaceutical industry it is
not regarded as improper to be involved in practical matters. The strategic
organization of experimenting and coordination of activities of departments
are such everyday practical matters. All kinds of scientific models, methods
and systems, such as 'finalized' type of research in biochemistry, 'transfer
research' to develop new inhibitors of tumor growth, and simple trial-anderror empiricism in drug screening are undertaken and funded. The one
exception is basic research concerned with the fundamental theoretical
problems. Here one does not know in advance where and how to set the
priorities and this is considered to be too risky for the industry.
Instrumental reason defines the identity of the industrial research worker.
He Or she will not hesitate to cooperate with public institutions, as for instance
with the U.S. National Cancer Institutes, in programmes for cancer drug
development. Proceeding in line with the economic objectives set by the

The Pattern of Cancer Research in Germany

155

pharmaceutical company the orientation of research here is growth- and

product-oriented. Differently from academic work where the product is a
'good' paper, in industrial research
the goal is a pragmatic contribution to therapeutical action - in the form of a specific
substance. The organizational structure of this type of research therefore is goal-<>riented.
The result should be a good and useful product (36).

The scientist working in industrial research and development has to keep

to different rules than does the researcher in academic science and these rules
define his role as an organizational, instrumental one. These rules are defined
as follows:
1. The scientist may not publish the results of his work whenever he wants to. 2. If his
work evolves into an independent research problem which deviates from the given
research goal, he must abandon it. 3. The work he does requires considerable flexibility
and does not allow him to become a specialist in one particular field. 4. There is a lot of
routine work. Forsaking scientific reputation is compensated by outward recognition for
his work and by remuneration (37).

3.3. The Difficulties of Clinical Cancer Research

Medical scientists have defined clinical cancer research as a type of research
which lays the scientific foundation for new therapeutic schemes to be
evaluated in clinical studies. This involves the transfer of results gained in
work with animal and cellular systems to the patient's tumor cells.
But here a dilemma presents itself:
These results were obtained by work on animal models which per se are highly artificial.
Subsequently, when one takes the step from the model into the clinic it is immediately
apparent that the relevance of all this preliminary research is nil. If these models are to
become relevant, they will have to be brought much closer to the patient (38).

To demonstrate the potential and significance of model approaches for

humans it is said to be necessary to abandon the isolated model approach and
to develop new, institutionalized forms of communication serving to bridge
the growing gap between molecular biologists and medical scientists. What
needs to be done is for workers in experimental cancer research and workers
in clinical research to elaborate the problems in a joint venture (39) and from
a clinical point of view. Yet any such joint projects see themselves confronted

156

Rainer Hohlfeld

with major problems arising from the career pattern of the medical scientists.
So far the physicians' and clinicians' career-pattern do not encompass research. The clinical researcher has been trained as a physician and must then
specialize so that he is never able to free himself from the constraints of his
'original' training and career. This actually means that the clinical researcher
cannot simply pursue his own research interests and develop competences
comparable to those of workers in basic research. In medical training, research
is regarded as a dead-end street. In addition, the physician who is not engaged
in research but who 'treats' patients, i.e., manipulates them directly, in the
eyes of people who are doing research enjoys high prestige and is sure of a
high income. Research in hospitals for all these reasons is not considered
attractive for a German physician.
3.4. The Belief System of the Medical Profession: The Medical Model
In their self-perception, clinicians feel obliged to adopt a bed-side orientation
in their work. Not the originality of the scientific background but the pressure
of immediate and responsible action gains priority in their everyday work.
So far, in their eyes there has been no real breakthrough nor even a real
contribution to medical progress in fighting cancer from molecular biology
(40).
The strategy of treatment is still steel, beam and drug. But it is acknowledged by leading scientists that so far there has been no real standardization
of methods, tumor classification, documentation, and consequently no
possibility to control the success of therapy. Randomized clinical trials are.
recommended as examples for necessary and desired planning:
. . . this includes national and international studies ... on therapy control and therapy
comparisons. Comparisons of the effects of one specific therapy as against the successful
outcome of another while at the same time taking account of the various side-effects, the
remission rates, duration of remission and time of survival require very careful planning
and the performance of cooperative studies (41).

Yet this assessment which focus on the treatment of cancer only is countered by the scientists who have the entire course of the disease in mind, that
is the epidemiologists. The latter draw attention to the restricted vision of the
physicians:

The Pattern of Cancer Research in Germany

157

All the physicians want to do is to treat diseases once they have made their appearance.
Mostly the physician is not motivated to prevent them. All he sees is the event as such
and all his efforts will be put into treating it; all his thinking and action is directed to
curing (42).

The focus on only one or two aspects of the natural history of cancer
diagnosis and treatment reflects an established tradition of medical thinking
which still governs the medical profession: the 'medical model' (43).
Theoretically, the medical model takes its orientation from the objectifying
sciences, i.e., from physics, chemistry and biology and has absorbed their
empirical-hypothetic procedures. Phenomena are worked on, processed to
be stripped out of their practical context and idealized into laboratory
phenomena, to make them reproducible. The scientist becomes the observer
who fmds himself confronted by an object which he himself has stylized.
Thus, the patient is isolated from his life context and is objectively examined
and observed by the physiCian, guided by the measurable and localizable
phenomena of the disease. The medical perception is somatic, organ-centered,
and aimed strictly at the physiological and pathophysiological functions
(44).
This theoretical model accords with a specific practice of treatment of
disease which takes a purely reactive line: the sick organ, the sick cell must be
cured. Accordingly curing can only set in for the localizable diseased entity,
the cell, the organ, the individual who is ill, but never can be applied to the
environment and to the living conditions of humans.
This scientific concept of disease implies a fragmentation of therapy, as is
exemplified from the canon of the specialities established on the basis of
organs and systems of organs in the medical disciplines and in their special
domains. At the highest level of technical power of control available to the
specialist, treatment of disease is split up into a multitude of disconnected,
singular practices (45). This disciplinary separatism can clearly be identified
in the attitudes and practice of physicians working in the specialities involved
in cancer diagnosis and treatment.
In Germany one-track specialism takes the form of the patient winding up in the hands
of the internist although he needs radiological treatment or needs to be transferred to the
surgeon. But the internist hesitates to send the patient on to radiology because of the
pecuniary losses that this entails (46).

Internists resist the establishment of clinical oncology as a new speciality

158

Rainer Hohlfeld

for the same motives: they are afraid of losing their patients to the new
oncologists.
The ignorance prevalent in one clinical domain about the work and orientation of the others and the tendency of the heads of clinics to set up 'fiefdoms'
were further demonstrated by the controversies about which discipline is to
have primacy in clinical oncology teams. For these reasons so far it has not
proved possible to establish combined tumor treatment in Germany of the
type current in the Swiss oncological services.
The separation of medical chairs from each other even affects the chances
of cooperation evolving between clinical and empirical research.
Nowhere have I seen so much animosity and at the same time so much cronyism as
is now current in medical practice and in the medical clinics. This is a state of things
which leaves its mark even on the attitudes and work of the assistant physicians, which
affects research, which gives rise to schools of enemies and schools of friends, and which
ultimately has an extremely inhibiting effect (47).

The fight for dominance in cancer research and treatment not only characterizes the disastrous state of the communication between the clinical
domains, but gives rise to sharp controversies between science and medicine
regarding their fields of competence.
Quarrels and squabbles prevail in all the areas in which science impinges on the clinical
domains, except for those in which medicine plays a primary role ... The physician's
social position is a unique one and thereby a priori creates difficulties for cooperation
with the sciences (48).

On the other hand, when scientists argue for tumor centers with integrated

research units and particular 'research patients' physicians reply "We cannot
allow the patients to be left in the hands of the scientists ... We will not be
reduced to the role of agents of the scientists" (49).

3.5. The Stepchild ofHuman-Related Cancer Research: Cancer Epidemiology

The actual task of cancer research is not to stay stationary in its traditional fixation on
animal models but to relate to man (50).

The object of cancer epidemiology, is man, not rat nor mouse. Nor is
the object of cancer epidemiology the individual cancer case but groups of
individuals who, compared to other groups within the population, share a

The Pattern of Cancer Research in Germany

159

high risk of getting cancer. The thinking of the cancer epidemiologist, therefore, focuses not only on the categories of illness but also on the categories of
non-illness. In other words, the epidemiologist's question is what makes
the healthy ill. But his way of thinking creates difficulties for him. "The
clinicians are not able to think in this dichotomy, their thought is always
centered around those who are already ill" (51).
Epidemiology as a science dealing with the human population must start
with risk factors and go on through to the social causes of disease. Its results
can encompass the actual causes of illness and lead to cancer etiology. Etiological research on the basis of population studies may have consequences
which politically are 'uncomfortable'. Stopping production of polyvinylchloride-based plastics in Norway, because it was found that this process
caused occupational cancer, can be seen as a case in point.
The status of cancer epidemiology has been improved by the cases of
occupational cancers and cumulating evidence that many substances which
are annually put on to the market or are by-products of industrial society
may act as carcinogens.
After all, this is the population we are really interested in, namely the human population
... Because this is the case, cancer epidemiology has in the past few years been receiving
increasing attention and this in turn has led to increased activities (52).

Despite this increasing attention, the state of data collection as a base for
epidemiological work in Germany is viewed as being disastrous. There are
only two cancer registers, which is wholly inadequate to supply representative
data for a population of some 60 million. Various reasons are put forward to
explain this wholly inadequate situation for starting any proper work. On the
one hand, there is the question of data protection:
Epidemiological analysis must relate to persons, must encompass the identity of the
patients. This is particularly difficult in our country, one can only say: A burnt child
dreads the fire. The experiences of the Third Reich in this respect after all are not a
pleasant memory. This has to some extent brought epidemiologists into disrepute.
And so we keep on running, trying to catch up with the international standard. Still,
these matters are extremely difficult in all the countries of the West. Here we enter the
domains of privacy and confidentiality, the legal requirement concerning confidential
medical communication, the right of the person to anonymity regarding the private
personal data (53).

On the other hand, this state of affairs is attributed to the strong resistance

160

Rainer Hohlfeld

of the medical profession inspired by their fear of being controlled in terms

of the mortality rate in the respective hospitals.
As a consequence of this combination of resistance and ignorance, epidemiology in Germany is not what it should be. There are only very few
experts, not to speak of those competent to teach. It is field of work deprived
of status and recognition. As to those of the younger people who work in
cancer epidemiology: "Most of them do not work exclusively in the field but
will in the majority try to get a safe spot somewhere as medical specialists"
(54).
4. Two Power Centers which Shape the German Cancer Research Scene
In carrying through the plans for research coordination and for establishment of needed new research areas within the frame of a research policy
programme, the German Federal Government must deal with the existing
cancer research scene. The scene, and consequently also the existing research
panorama, is governed by two centers of intellectual and institutional power:
by basic science, defined primarily by the paradigm of molecular biology, and
the medical profession, defmed by the medical model. In this sense, control
is exercised over the concepts of the disease, the organizational status of
the national research efforts, the research priorities, as well as the levels of
technology intended.
Common to both molecular biology and medicine is a particular view
of disease which, in turn, implies a particular mode of reductionism: the
reduction of illness to monocausal physiological events. The reduction scheme
applied by modern science and medicine can be characterized by three steps
(55).
(1) The connection between awareness and experience of the illness and
the natural history of the disease, and with it the mind-body complementarity, are reduced purely to the body aspect. How the patient deals with his
illness and his life problems is screened out as is the doctor-patient dialogue.
That the dialogue with the doctor is needed to understand the patient, and
the fact that this is as relevant to the diagnosis as are the physiological data,
are ignored. The social and inter-subjective phenomenon of 'illness' is converted into the research object 'disease'.
(2) Disease itself is reduced by stripping it of the psychic and social

The Pattern of Cancer Research in Germany

161

dimension comprising such factors as problems of living, life events, living

conditions, behaviour diet, and the like. Disease is interpreted as a purely
physiological event according to the medical model. The dimensions of health
problems which have to do with maintaining health and not combating disease
are simply left out of consideration.
(3) A holistic, multicausal phenomena is reduced to a monocausal event,
to one factor, or only a set of aspects. Thus, the natural history of cancer is
mainly reduced to diagnosis and treatment in medical practice and at the
theoretical level the etiology of cancer is reduced to its physical origin by
irradiation, or to its chemical origin by the action of carcinogenic agents, or
to its viral origin, or to the loss of immune control of tumor growth. These
three steps are common to both the medical model and to biomedicine.
(4) With the emergence of biomedical research a further step is incorporated into the reduction process. This is the reduction step from a macro- to a
microconcept of disease, made possible by molecular biology and providing
the high technology in fighting chronic diseases. Thus, for instance, the
concept that a specific chemical substance induces cancer in animals by skin
application is replaced by the concept that a metabolic intermediate of the
compound may act as a mutagen causing loss of cellular growth control.
By reducing the complex phenomena of the chronic disease of cancer to
its physiological aspects and further to key events in molecular cell biology,
the health problem is perceived only in terms of the 'medical' and 'technological' fix and is thereby Simplified. The result is an inadequate representation
of the illness in terms of both the medical model and theories and concepts of
biomedical research.
This overall reduction scheme defmes an intellectual hierarchy according
to which the different subject areas of cancer research can be ordered and
the basic research 'force field' can be described more precisely. The structure
underlying the system of scientific recognition and rewards is the converse of
the reduction scheme. The most rewarded defmition of the cancer problem is
the defmition conceived in terms of the most advanced research fronts of
molecular biology and elaborated by the purest test-tube system. The defmed
research programme becomes the focus of intellectual competition among
the scientific elites. Human understanding of the last secrets of nature, the
"challenge of life", gives this type of research high status within the scientific
community and among the pUblic.

162

Rainer Hohlfeld

On the other hand, for researchers doing this advanced work, research
in mature areas like cell biochemistry and molecular genetics of bacteria is
regarded as 'homework', done to solve the respective problems down to the
last chemical detail. Once the main problems of a subject area are solved, the
leading scientists start migrating with the perspective that a new research
front can now be tackled (56). Less attractive, both from the point of view
of rewards and career opportunities, is work with more complex and 'dirty'
systems.
Even more complex is bed-side oriented clinical research, which is most
relevant for the patient. This type of research is recommended as highly
important, but nobody wants to do it. The methodological standards are
regarded as low and theoretical breakthroughs cannot be expected. So far,
this type of research has not been undertaken on any large scale.
The subject area focussed on the environmental-body relationship and the
etiology of the disease (and which therefore represents the prevention phase
of the disease process) is degraded to a system of medical statistics by the
curatively-oriented medical and basic scientists and enjoys little scientific
recognition.
The attractiveness of the different approaches within cancer research for
'good' scientists and the resultant options for recruitment of new manpower
for the non-academic type of approaches depends largely on this hierarchical
pattern, which is determined by the scientific recognition gradient from 'hard'
to 'soft' science (57). The most rewarded scientific activity is the one which
reduces complex phenomena to immobile stable micro-units - the molecules.
The least rewarded activity is that which deals with the non.reproducible,
changing and mobile units - the psychological and social conditions in the
etiology of cancer (58).
The field of force of basic science and the overall reduction scheme in
conceiving illness is supplemented by an independent center of enormous
institutional and political power: the medical profession and its institutions.
Medical practice and health care, in West Germany monopolized by the
medical profession (59), are not directly linked to the research programme
of the biosciences. The intellectual identity of the medical profession is still
determined by the traditional medical model. The high status of the profession and of the physician is due to his expertise in the exercise of a public
function which has high priority. The reward system within the profession

The Pattern of Cancer Research in Germany

163

works mainly via recognition of the physician's successful therapeutic action,

rather than via recognition of research work. Therefore, bedside research
offers no career chances and is not highly regarded among the clinicians, who
feel that those involved in this line of work no longer satisfy the demands of
direct medical care.
In addition to the lesser scientific reputation of epidemiology and medical
sociology, institutional constraints are put upon this type of research, because
it would displace the war on the causes of the disease into the political and
social area outside the frame of medical health care. The medical profession
cannot admit institutionalization of these research fields as this would mean
the loss of clients, and the loss of the health care system's monopoly (60).
Unlike the situation in the United States of America (61), in West Germany
not much indicates converging tendencies between the concepts of biomedical
research on the part of molecular biologists and the traditional medical model
upheld by the medical profession which might bring into existence a 'biomedical model'. Notwithstanding their common reductionism and curative
fix, present conditions are still determined by the contest between and within
the two establishments for dominance in national cancer policy. There are
deep barriers between the test-tube orientation of the scientists on the one
hand and the clinical orientation of medical professionals on the other 'worlds that do not understand each other' (62). When the low standards in
experimentation, the lack of any theoretical background, and the special
status of the physician are criticized by the scientists, the physicians respond
by criticizing the esoterics of the 'model systems', which are totally irrelevant
to the clinical problems because 'man is not a tissue culture', and by attacking
the touch of arrogance and the aesthetics of the test-tube workers.
Since the end of the Second World War, this controversy between science
and medicine about the right way of fighting cancer has prevented national
coordination of cancer research in Germany. So far there have prevailed at
least two different conceptions of cancer research policy. First, that of the
German Cancer Society, representing the medical perception of the problem
and the clinical point of view. Second, the policy promoted by science
represented by the 'Deutsche Forschungsgemeinschaft' (DFG), the 'Max
Planck Gesellschaft' (MPG) , and the 'Deutsches Krebsforschungszentrum'
(DKFZ).
That this fight for the leading position is still being waged is clearly

164

Rainer Hohlfeld

documented by a recommendation of the DFG-Committee concerned, which

stated the following relative to the establishment of a Board for national
coordination of cancer research:
... The Coordination Board should be attached to the DFG so that from the very fIrst
the primary of the scientific objectivity of the Board will be beyond any doubt (63).

Also beyond doubt is the fact that in spite of all the efforts of cancer
organizations, research funding associations, and of the Federal Government,
the present state of German cancer research and cancer policy can still be
characterized by one basic feature: pluralism without organization (64).
Notes and References

*
1.

2.

3.

4.

5.
6.

I would like to thank Suzanne Libich, Max-Planck-Institut Starnberg, for the

translation of the manuscript.
U.S. Department of Health, Education and Welfare (DHEW), The Strategic Plan.
National Cancer Programm, DHEW Publication No. (HIH) 74-569, 1973. For the
history and the state of the art of U.S. cancer research policy, cf. S. P.
Politics, Science and the Dread Disease, Cambridge, Mass.: Harvard University Press,
1972; R. A. Rettig, Cancer Crusade, Princeton: Princeton University Press, 1977;
K. E. Studer and D. E. Chubin, The Cancer Mission, London: Sage 1980.
Deutscher Bundestag, 7. Wahlperiode, 1975: Grosse Anfrage der Abgeordneten Dr.
Ing. Laermann, Kern, Egert, Dr. Bardens, Spitzmiiller und der Fraktionen der SPD,
FDP vom 19.2.1975, Sachgebiet 12, Drucksache 7/3236; Deutscher Bundestag, 7.
Wahlperiode, 1976: Antwort der Bundesregierung. Betr. Krebsforschung, Bundesdrucksache 7/4711, Sachgebiet 212, 1976.
Der Bundesminister fUr Forschung und Technologie,Programm der Bundesregierung
zur F6rderung von Forschung und Entwicklung im Dienste der Gesundheit 19781981, Bonn: Der Bundesminister f1ir Forschung und Technologie, 1978.
Deutsches Krebsforschungszentrum, 1976: Memorandum zur Errichtung einer
integrierten klinisch-onkologischen Einrichtung (IKO) in Heidelberg, Heidelberg:
DKFZ 1976, Schmidt C. G., 1976: 'Die Entstehung der Essener Tumorklinik',
Mitteilungsdienst der Gesellschaft zur Bekiimpfung der Krebskrankheiten NordrheinWestfalen e. V. 4 (1976) 3-7.
Stifterverband ftir Deutsche Wissenschaft, 1976: 'Pluralismus und Organisation in
der Wissenschaft: das Beispiel Krebsforschung', Wirtschaft und Wissenschaft 24
(1976) 22-4.
By this line of thinking I am referring to the concept of 'resistance' and 'receptivity'
of science to political direction, cf. W. van den Daele, W. Krohn and P. Weingart,
'The ?olitical Direction of ScientifIc Development', in E. Mendelsohn, P. Weingart
and R. Whitley (eds.), The Social Production of Scientific Knowledge, Sociology of
the Sciences Yearbook, I, Dordrecht and Boston: Reidel, 1977, pp. 219-22.

The Pattern of Cancer Research in Germany

165

7. I have interviewed 29 scientists both in leading and non-leading position in the areas
of molecular biology (basic research); experimental cancer research, clinical cancer
research, cancer medicine, and epidemiology in the years 1975 and 1976. For a
more systematic foundation of the classification see Section 3. (The author has the
tape recordings and transcribed records of the interviews.) I refer to these records
as, e.g., 'Expert Interview Experimental Cancer Research'.
8. For the development of classic genetics theory, see A. H. Sturtevant,A History of
Genetics, New York and Tokyo: Harper & Sons, 1966.
9. The term molecular biology was introduced by Astbury, 'Adventures in Molecular
Biology', Harvey Lectures 46 (1950) 3-44, to denote the very complex molecules
playing the key role in cellular processes like proteins, and has since then often
been used synonymously with the term 'molecular genetics', e.g. the molecular
conceptions of the phenomena of heredity. The biochemists have used the term
synonymously with biochemistry, that is the approach that traditionally investigates the metabolism of cellular and organismic compounds. From a systematic
point of view, it is by far more conclusive to use the term molecular biology for
the entire reductionist programme and to differentiate between the particular
theoretical approaches such as 'biochemistry', 'molecular genetics', 'molecular
embryology' to exemplify.
10. For my use of the term 'reduced' and 'reductionism', I am referring to K. Schaffner,
'The Peripherality of Reductionism in the Development of Molecular Biology',
Journal of the History of Biology 7 (1974) 11-119. For historical reconstruction I
would prefer to emphasize, the incommensurability of classical and molecular
genetics and in the light of the fate of the old theory to use the term 'replacement'
(cf. D. Hull, cited by Schaffner, ibid., p. 119). The replacement of a theory by
another is equivalent to a paradigm revolution in the sense of T. S. Kuhn; The
Structure of Scientific Revolutions, Chicago: The University of Chicago Press,
1970.
11. This listing makes no claim to completeness but only attempts to grasp the hierarchies inasfar as they are relevant for cancer biology. The principles of 'assembly
of molecules' which characterizes the higher degrees of order of biological processes
and structures defined as 'boundary conditions' (L. L. Gatlin, Information Theory
and the Living System, New York: Columbia University Press, 1972, 14-17) or as
levels of emergence (M. Polanyi, 'Life's Irreducible Structure', Science 160 (1968)
1308-12.) were used by Polanyi to argue against reductionism by pointing out
that precisely these structural principles cannot be reduced to chemical theories.
This position was rejected by K. Schaffner, ibid., 113, who stated that even the
structural relations in biological systems can be conceived in chemical terms.
12. For the developments of theories in biochemistry, see R. E. Kohler, 'The History
of Biochemistry: A Survey', Journal of the History of Biology 8 (1975) 275-318;
G. Allen, 'The Chemical Foundation of Life', in G. Allen, Life Science in the
Twentieth Century, Cambridge: Cambridge University Press, 1978, pp. 147-85.
13. Theoretical maturity of a theory shall be preliminary deImed as the state of a
theory, where the basic principles of explanation and constitution of a definite
range of subjects are coherent with the empirical data., cf. G. Bohme, W. van den
Daele, and W. Krohn, 'Finalization of Science', in Social Science Information 15
(1976) 306-30.

166

Rainer Hohlfeld

14. For the establishment of the molecular genetics approach and its science policy
preconditions, see E. Yoxen, 'Giving Life a New Meaning: the Rise of the Molecular
Biology Establishment', this volume. For the history and periodization of molecular
genetics, see R. Olby, The Path to Double Helix, London: Macmillan, 1974; G.
Stent, The Coming of the Golden Age, Garden City, N.Y.: The Natural History
Press, 1969. For a more systematic interpretation in the sense of a phase model of
scientific development (cf. Bohme, van den Daele, Krohn, ibid.) see R. Hohlfeld,
'Theory Development in Molecular Biology', in: W. Callebaut, M. de Mey, R.
Pinxten, F. Vandamme (eds.), Theory of Knowledge & Science Policy, Ghent:
Communication and Cognition, 1979, pp. 346-57.
15. See E. E. Luria, 36 Lectures in Biology, Cambridge, Mass.: The MIT-Press, 1975,
pp.213-73.
16. The term was proposed for a new level of technology in medicine, achieved by
applying theories of molecular biology to medical subjects (cf. note 19) by L.
Thomas, Aspects of Biomedical Science Policy. Washington: National Academy of
Sciences, 1972. Systematically development of high technologies means that on the
basis of a mature fundamental theory of the subject area a specialized 'daughtertheory' is constructed to deal with the particular technical constraints. This case of
goal oriented theory construction has been termed 'f"malization' in Science; cf.
Bohme, van den Daele, Krohn, op. cit., 1976, Note 13; Hohlfeld, op. cit., 1979,
Note 14.
17. cr. B. Hartley, 'The Bandwagon Begins to Roll',Nature 283 (1980), 122.
18. Luria,op. cit., 1975, (Note 15), p. 248.
19. In this sense, the term is used by one of the leading protagonists of the philosophy
of combining molecular biology and medicine to achieve a high technology in
fighting disease, C. L. Thomas, op. cit., 1972, Note 16.
20. By this classification I refer to P. Weingart, 'On a Sociological Theory of Scientific
Change', in: R. Whitley (ed.), Social Processes of Scientific Development, London
and Boston: Routledge & Kegan Paul, 1974, pp. 45-68.
21. Breuer, H., zur Hausen, H. Oettgen, H. F. Schmidt, C. G. (Hrsg.), Bericht uber ein
Expertentreffen zur Frage der lftiologie und der therapeutischen Beeinjlussung
maligner Tumoren. Schriftenreihe der Deutschen Stiftung fliI Krebsforschung, Bd.
1, Bonn, 1977, p. 41.
22. Deutsche Forschungsgemeinschaft, Bestandsau/nahme Krebsforschung in der
Bundesrepublik Deutschland 1979, Boppard: Harald Boldt Verlag, 1980, p. 200.
23. Ibid., p. 199.
24. Expert Interview Experimental Cancer Research I.
25. Ibid.
26. Ibid.
27. To discriminate between the social roles of scientists with respect to their motives
for doing research, I would hold to the classification of S. Box and S. Cotgrove,
'Scientific Identity, Occupational Selection and Role Strain', The British Journal
of Sociology 27 (1966) 20-8, which draws the line of distinction between the
'intrinsic' role of scientists mainly operating within specific scientific communities
and the 'instrumental' role of scientists working mainly in industrial research and
development.
28. Expert Interview Experimental Cancer Research II.

The Pattern of Cancer Research in Germany

167

29. R. Whitley, "Cognitive and Social Institutionalization of Scientific Specialties and

Research Areas", in: R. Whitley (ed.), op. cit., 1974 (Note 20) pp. 69-95.
30. Expert Interview Experimental Cancer Research III.
31. Cf. R. Hohlfeld, "PraxisbezUge wissenschaftlicher Disziplinen. Das Beispiel Krebsforschung". In: G. Bohme et al., Die gesellschaftliche Orientierung des wissenschaftlichen Fortschritts. Frankfurt: Suhrkamp, 1978, pp. 178-80.
32. Expert Interview Experimental Cancer Research IV. The scientists concerned have
put the perception of the prestige gradient between basic research and transfer
research into the slogan: the purer the system, the higher the IQ.
33. Ibid.
34. Op. cit., Note 25.
35. Ibid.
36. Expert Interview Experimental Cancer Research IV.
37. Ibid.
38. Expert Interview Clinical Cancer Research I.
39. Cf. Deutsche Forschungsgemeinschaft, op. cit., 1980 (Note 23) p. 200.
40. This statement was accepted by both molecular biologists and clinicians (Expert
Interviews Molecular Biology I and Expert Interview Cancer Medicine I).
41. The president of the German Cancer Society in his statement on the inquiry of the
SPD/FDP-parliamentary group on the state of cancer research in Germany (cf.,
note 2) C. G. Schmidt, Essen, 18.7.1975.
42. Expert Interview Epidemiology I.
43. The line of thinking of the medical profession and the term 'medical model' has
been worked out by the critics of established traditions in medicine and has been
first demonstrated by the critics of psychopathology; cf. T. S. Szasz, The Myth of
Mental RInes" New York: Harper & Row, 1961. Among the literature in this
context I want only to refer to the most instructive analyses: G. L. Engel, 'The
Need for a New Medical Model: A Challenge for Biomedicine', Science 196 (1977)
129-36; E. Freidson, Profession of Medicine, New York: Dodd, Mead & Company
1970; G. L. Klerman, 'Mental Illness, the Medical Model, and Psychiatry', Journal
of Medicine and Philosophy' (1977) 220-43.
44. Cf. W. Siegenthalter, Klinische Patophysiologie, Stuttgart: Georg Tieme, 1976, p. v.
45. To overcome this separatism in the clinic of cancer, the German Cancer Society has
proposed to institutionalize a new medical specialty: clinical oncology to integrate
the scattered knowledge about cancer diagnosis and therapy (C. G. Schmidt and E.
Scherer, "Zur Situation der Onkologie in Deutschland", Deutsches A'rzteblatt 72
(1975) 2009-21.) For the emergence of the medical specialization and separatism
in England, cf. J. Sadler, 'Ideologies of "Art" and "Science" in Medicine' in:
W. Krohn, E. L. Layton and P. Weingart (eds.), The Dynamics of Science and
Technology, Sociology of the Sciences Yearbook, 2, Dordrecht and Boston: Reidel,
1978.
46. Expert Interview Experimental Cancer Research I.
47. Expert Interview Experimental Cancer Research V.
48. Expert Interview Experimental Cancer Research IV.
49. Expert Interview Cancer Medicine I.
50. Expert Interview Epidemiology I.
51. Ibid.

168
52.
53.
54.
55.
56.

57.

58.
59.
60.
61.
62.
63.
64.

Rainer Hohlfeld

Ibid.
Expert Interview Epidemiology II.
Op. cit., Note 52.
The basic idea of this reduction scheme is very clearly outlined by Engel, op. cit.,
1977,Note44.
Historically, as reported by an eye-witness, the consensus reached among scientists
at the Cold Spring Harbor Symposium in 1963, at which the clarification of the
genetic code was demonstrated, made molecular geneticists attack the eucaryotic
cell as the subject following area (Expert Interview Molecular Biology II).
In the reduction scheme outlined above I have focussed on the reduction of the
complexity of phenomena. With the connotation of 'dirty' and 'pure' systems, the
focus is on the experimental reproducibility of research objects. The hierarchy
defined from this more operational point of view is discussed by A. Rip under the
terms of 'restrictedness' and 'unrestrictedness' ('The Development of Restrictedness
in the Sciences', this volume). The clarification of the relation of both hierarchies
needs further investigation.
This scheme might serve as a model for the more general value scheme outlined by
N. Elias, 'Scientific Establishments', this volume.
For an analysis and criticism of the German medical establishment and the health
care system see P. LUth, Kritische Medizin, Reinbeck: Rowohlt 1972; H. U. Deppe
(ed.), Vemachliissigste Gesundheit, Koln: Kiepenheuer & Witsch, 1980.
Cf. M. Janicke, Wie das Industriesystem von seinen Miflstiinden projinert, Opladen:
Westdeutscher Verlag, 1979, pp. 82-7.
Cf. A. Weinberg, 'The Coming Age of Biomedical Science', Minerva 4 (1965) 3-14.
Expert Interview Molecular Biology III.
Deutsche Forschungsgemeinschaft, op. cit., 1980 (Note 23), p. 203.
Cf. also the present German parliamentary debate on cancer policy, 'Programm zur
Krebsbekiimpfung', Das Parlament, 31.1.1981, pp. 1-4.

DEVELOPMENT AND ESTABLISHMENT IN ARTIFICIAL

INTELLIGENCE

JAMES FLECK
University ofAston in Birmingham

1. Introduction
In this paper, I discuss the role played by scientific establishments in the
development of a particular scientific specialty (1), Artificial Intelligence
(AI), a computer-related area which takes as its broad aim, the construction
of computer programs that model aspects of intelligent behaviour. As with
any discussion of a scientific specialty, the identification of what is involved
is not unproblematic, and the above serves as an indication rather than a
defmition. While the term 'Artificial Intelligence' is used in a variety of
ways (2), there is a discernable group (perhaps approaching the degree of
commonality to be called a community) of researchers who recognize the
term as descriptive of a certain sort of work, and who, if they themselves are
not willing to be directly labelled by the term, can locate themselves with
respect to it.
Unfortunately, there is little or no cOminonly available literature that
systematically charts the scope of this area. It is worthwhile, therefore,
to consider the distinctive socio-cognitive characteristics of research in AI
as a prelude to a fairly specific discussion of the social and institutional
processes involved in the development of the area (3), thus providing a basis
for exploring the usefulness and applicability of the concept of establishment.
2. Socio-Cognitive Characteristics of Artificial Intelligence
The patterns of research in AI exhibit distinctive characteristics, forming a
paradigmatic structure which includes such elements in the scientific activity
as research tools, practices, techniques, methods, models, and theories, as well
169
Norbert Elias, Herminio Martins and Richard Whitley
Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982.169-217.
Copyright 1982 by D. Reidel Publishing Company.

170

James Fleck

as the normative and evaluative aspects for selecting among them (4). They
serve as guidelines and a basis for future research, but are complexly interrelated, often encompassing contradictory facets in tension.
The elements in the paradigmatic structure of AI are as follows:
(1)

(2)

(3)

(4)

(5)

The general-purpose digital computer provides an instrumental base

and a disciplinary context - computer science - for research in the
area. Adequate computing facilities are essential for AI work, and
hardware limitations have had a constraining effect. Consequently,
the availability of funding has been of crucial importance for the
development of the area.
List processing languages, a subset of the high level programming
languages available for exploiting the power of the computer, have
been developed as tools for research in AI. The community of people
using list processing languages, such as LISP, in the United States, or
POP-2, in Britain, and their variants, can serve as a first approximation for the AI community.
These list processing languages are oriented towards non-numerical
uses, and, hence, contrast with more conventional programming
languages such as FORTRAN or ALGOL which are numerically
oriented. This non-numerical emphaSiS, with a focus on logic and
structure rather than Immber, distinguishes AI from areas such as
Pattern Recognition, for example, which depends heavily on the use
of statistics.
Associated with list processing languages, there has developed a
distinctive body of craft knowledge. A high level of skill, gained
through first-hand use and practice, is required for the effective use
of any programming language and there are many 'tricks of the
trade' as well as distinct programming styles, which can only be
absorbed through an extended period of apprenticeship (5).
Embedded in this craft knowledge are numerous elements such as
techniques for problem solving, for representing knowledge, for
achieving learning ability, etc. Particularly important and well
developed among these are procedures for carrying out searches,
often employing rules of various kinds - heuristics - to guide the
search and cut down the possibilities to be explored.

Development and EstabUshment in Arti/iciallntelligence

(6)

(7)

171

The craft knowledge of AI is deployed in the construction of computer program models - computational models - of some aspect
of intelligent activity. These models are generally pitched at the
symbolic level of meanings rather than at the physiological level of
the underlying mechanisms. This distinguishes AI from many other
cybernetic approaches, and from much computer simulation work.
The focus on intelligent behaviour provides a disciplinary context psychology - but, due to the great variety of social interpretations
and applications of the term 'intelligence', specific goals for research
are not thereby dictated. This lends AI a similarity with what can be
termed instrument or technique-based specialties, such as Xray
crystallography (6),
are free to be applied to various goals.
Associated with the wide variety of specific examples of intelligent
activity that have been modelled, a clear research area differentiation
has emerged since the early 1960s in which subareas have developed
their own particular specialist guidelines and techniques, focussed on
their own more circumscribed concerns. The research areas that
could be identified in the early 1960s were game playing, theorem
proving, cognitive modelling (an emphasis on models with psychological verisimilitude). natural language, machine vision, and a range
of specific applications (7) some of which have themselves subsequently differentiated out into well defined research areas. These
research areas (or strands of research (8 constitute a primary
setting for scientific activity, and consequently have been one of the
basic arenas for competition among practitioners, as will become
evident.

These cognitive characteristics, or elements of the AI paradigmatic structure are, of course, at a very general level. They open up a huge cognitive
space which offers wide opportunities for exploration, and were elaborated
at a fairly early stage in essentially their complete form, while subsequent
work has largely exploited the possibilities opened up. This overview of the
development of AI invites comparison with Edge and Mulkays' account of the
development of radio astronomy: the initial discovery of radio waves from
space opened up the possibility of a new source of astronomical information
- a new cognitive space - which was subsequently exploited by ever more

172

James Fleck

sophisticated methods of detection, leading ultimately to a revolution in

the conception of astronomy (9). However, while radio astronomy was
apparently allowed to develop without much external conflict (10), the same
cannot be said of AI.
3. Competition and Establishment in Artificia1lntelligence
Conflict in AI has been bound up with the focus on intelligence. Intelligence
is not a socially or cognitively well-defmed goal and every distinctive social
group tends to have its own implicit definition, couched in terms of its own
interests. Consequently research in AI has been oriented towards a variety of
goals. This multigoal character leads to a range of struggles between various
groups and establishments within and around AI, and is institutionally manifested in the high degree of research area differentiation, with interdisciplinary
and multidisciplinary affiliations, and associated multiple funding sources.
This leads to competition on the one hand, between research areas for resources, and on the other hand over the defmition of what AI is. This has had
quite clear effects on development, as for example, with the debate in the
Science Research Council in the early 1970s, which led to separate funding
mechanisms being set up for cognitive science (linguistics, philosophy, and
psychology) applications of AI, and for research within Computer Science.
This multigoal characteristic, involving competing groups with different
aims has exerted centrifugal pressures on research in the area and has resulted
in the non-emergence of a specialty wide general theoretical dynamic. Attempts at the elaboration of theories of intelligence have informed work in
the area - for example, the early programme of research (evident in work
in the 1960s in systems such as GPS (General Problem Solver, towards
forming general mechanisms of inference that would embody the essence of
intelligence - but these attempts foundered upon the diversity of concepts
and applications involved. Later developments such as the attack upon the
problem of knowledge representation (a major theme of resea,rch in the
1970s) effectively accepted the contingent diversity of intelligence and turned
it into a virtue. What theoretical developments there have been, however,
have been very specific and localized, often pertaining to the status of the
methods and languages employed.
Nevertheless, the absence of a uniform goal, or a general theoretical

Development and Establishment in Artificial Intelligence

173

dynamic raises the question of the source of cohesion and coordination

for research in the area. The answer seems to lie in the craft nature of the
paradigmatic structure. While there are many divisions over short- and longterm goals, and between different research areas, there is a shared body of
technique and practice based on the use of list processing programming
languages and transmitted by apprenticeship and personnel migration, thus
constraining the historical development of the area. Access to this body of
knowledge and skill is restricted by the need for rust-hand contact and for
adequate computing facilities, consequently leading to tight intercentre and
intergenerationallinkages in the area. The group of people who control access
to these resources clearly constitute the establishment in AI, and it is at this
level that much of the internal research area competition takes place. It would
seem, therefore, that this case demonstrates that scientific establishments
need not be characterized by a high degree of solidarity. It would also seem
to be the case that a common basis in technique is adequate to hold an area
together in the face of strong centrifugal tendencies, especially where it is
associated with an instrumental basis for research in the area. The: need for
adequate computing facilities has restricted access to the field and encouraged
the development of a strong communication infrastructure, particularly in the
United States where the ARPA (Advanced Research Projects Agency) computer network enables researchers at geographically distant sites to communicate as easily as if they were at the same place. Such expensive instrumental
needs have, of course, opened the area up to influence from funding agencies.
Research in AI has, without doubt, depended on substantial support from
various agencies, such as the United States Department of Defense, and the
United Kingdom Science Research Council, and has consequently been
shaped by the concentration policies of these agencies which have had the
effect of consolidating the position of the establishment in the area. But it is
difficult to fmd evidence for any positive direction of research by the funding
agencies during the 1960s, although in the stringent fmancial climate of the
1970s this changed, and tighter demands for the attainment of particular
goals were made, resulting in a restructuring of funding patterns among AI
centres, both in America and in Britain. And during the early 1980s, there
appears an initial emergence of "AI-Technology", where specific lines of
research, considered to have commercial potential (not necessarily those
of prime scientific importance), are being picked up out of the university

174

James Fleck

context, along with supporting personnel, and transferred for industrial

development. At this level of a broad overview of AI, there appears to be
some similarities to Yoxen 's description of molecular biology in terms such as
"directed autonomy" (11). But at the more detailed level of the following
discussion, it is very hard indeed to identify elements of long-term direction
that might fit in which concomitant long-term strategies on the part of the
funding agencies. Perhaps what is at issue here is the appropriate size and
nature of the envelope within which autonomy is exercised by the practitioners, while yet remaining suitably circumscribed in accordance with the
externally imposed direction.
The availability of funding and institutional resources for the area as
a whole is, of course, controlled by a wider establishment - the funding
agencies, and the universities - and in the following discussion the processes
of negotiation between the specialty and wider establishments stand out
-clearly. It will become clear that the response of the wider establishment to
AI is by no means uniform, thus illustrating that too monolithic a character
should not be imputed to the establishment at this level either, but that an
adequate understanding needs to take into account the particular features of
the area, that is the specificity of AI.
The allocation of resources and the negotiation between the establishments in and around AI have been clearly affected by what Elias termed a
struggle for monopolization of the means of orientation (12). Research in
the area is often seen as constituting a further thrust of mechanical materialist
science into an area - the nature of mind - hitherto under the exclusive sway
of traditional cultural values. By and large, mind or intelligence is regarded
as the most characteristic and unique of human attributes - extremely rich
and complex, and undoubtedly beyond the reach of scientific analyses
more appropriate for the understanding of inert matter. Thus, the focus on
intelligence and mind brings AI into an arena of conflict at a deep-seated
emotional level which touches immediately upon everyone's image of themselves, and induces strong for and against alignments. It is an issue of general
public rather than narrowly scientific interest, as is evidenced by the high
relative exposure AI receives in the, press, on television, etc. It is doubtful
whether many other scientific areas could have the same effect - with the
exception of some areas such as genetic manipulation which undoubtedly
bear comparison. The nature of mind is an area where the religious and

Development and Establishment in Artificial Intelligence

175

philosophical establishments still claim authority, and AI has to fight for

legitimacy. Even where explicitly religious commitments do not seem to be
involved, those brought up under the western humanist culture often feel
threatened by what they see as the reductionist nature of AI, and work in
the area has been denounced as bad science, non-science, gross reductionism,
and even immoral science. At this level then, there is negotiation and conflict
with establishments outside science, as well as between establishments within
science.
As far as competition between establishments within science is concerned,
the case of AI illustrates an important, characteristically 20th-century development in scientific thinking - the software sciences - with a focus on
pattern and organization rather than on the properties of substance or matter.
Yoxen points out in his paper the importance of such metaphors as code,
information, read-out, program, etc., in the reconstitution of biology: with
AI such ideas are at the very core of the subject. Moreover, with the increasing
penetration of the computer into all areas of science and scholarship, the
features of AI related to the software science nature of research in the area,
may well become typical of many fields of science. In particular, the diffuse,
method-based character of AI, with its contingent adaptation to diverse
substantive issues, poses a contrast and challenge to the coherent, theoretically centred nature of the current scientific ideal, deriving from the example
of the dominant physical sciences tradition. The former would not seem to
so readily support a monolithic unified establishment as does the latter and,
consequently, may have implications for the future development of the
sciences.
Thus, AI appears to be an interesting case in the context of a discussion
of scientific establishments, for a number of reasons. The more diversified
nature of establishments in the software sciences may have wider implications
for the sciences as a whole; negotiation and conflict between establishments
at a variety of levels is clearly illustrated - including a struggle for the monopolization of the means of orientation; the power bases of these various
establishments (control over cognitive, instrumental or financial resources)
are clearly evident; and finally AI provides many examples of the problems
arising from struggles between the various establishments involved in a multidisciplinary and interdisciplinary area of research - illustrating many of the
points commented on by Elias.

176

James Fleck

4. Early Development in the United States

The Second World War acted as a melting pot for various quite different lines
of research and disciplines. In the intense concentration on the common goal
of winning the war, traditional disciplinary boundaries were broached and
new areas of research emerged, such as information theory, operations
research, cybernetics and, of course, the development of the digital computer
itself. These areas of research can be broadly characterized as the software
sciences, in that they focussed on pattern and organization rather than on
substance or matter - the concern of the natural sciences such as physics and
chemistry.
Cybernetics, a rather general field given a name and identity by Norbert
Wiener's classic book: Cybernetics - Control and Communication in the
Animal and the Machine, was concerned with the essential similarities between machines and biological processes (13). Work in the area developed
during the 1940s, and involved such approaches as the comparison of biological and neurophysiological processes with electrical circuits and networks of
artificial neurons, or the investigation of the general principles of adaptation
in self-organizing systems - systems which were rich in feedback connections,
and would settle into stable configurations after being disturbed (14).
The advent of the digital computer in the early 1950s heralded a new
approach which sought to build models of intelligent processes at the symbolic level (15). Concepts were represented and operated on directly in the
computer using high level programming languages. These 'symbolic' models
represented intelligent activity at the level of thought itself, rather than at the
level of the physiological mechanisms underlying thought, thus contrasting
sharply with other cybernetic approaches.
In 1952, a conference was held under the rubric 'Automata Studies' (16).
This conference, organized largely by John McCarthy, was intended by him
to attract proponents of the symbolic modelling approach. It failed in this
aim, and attracted contributions more clearly in the other cybernetic traditions. This determined McCarthy to "nail the flag to the mast the next
time", which he did by explicitly using the term 'artificial intelligence' in a
subsequent summer school held at Dartmouth College in the United States
in 1956, to discuss 'the possibility of constructing genuinely intelligent
machines' (17). The official title was 'The Dartmouth Summer Research

Development and Establishment in Artificial Intelligence

177

Project on Artificial Intelligence' and did succeed in isolating the symbolic

modelling theme. Among those present were J. McCarthy, M. L. Minsky,
H. A. Simon and A. Newell (18). After the meeting, Simon and Newell
were to start a group at the Carnegie Institute of Technology (now the
Carnegie-Mellon University) with the aim of developing models of human
behaviour, while McCarthy and Minsky built up a group at Massachussets
Institute of Technology (MIT), with the goal of making machines intelligent
without particular reference to human behaviour. Later in 1962, McCarthy
was to move to Stanford University, where he initiated another AI project.
These three centres, along with Stanford Research Institute, dominated AI
research in the United States in the 1960s and 1970s. Also present were C. E.
Shannon (known in AI for his outline of the chess playing paradigm which is
essentially the same as the one underlying the microelectronic machines
that can now be brought off the shelf) and A. L. Samuel (who developed an
impressive checkers playing program which incorporated an elementary
learning mechanism).
It was at that meeting that the broad outlines of a distinctively AI approach, indeed what might be called a proto-paradigmatic structure, emerged.
This involved the use of high -level programming languages to provide symbolic
models of various aspects of intelligent activity. The first areas attacked,
chosen partly because they seemed to epitomize intelligence, and partly
because they were sufficiently well-dermed to be readily programmable, were
theorem proving in mathematical logic, and games such as chess and checkers.
While chess and other games employed numerically-based techniques for
choosing board moves, the 'Logic Theorist' of Simon and Newell (which was
presented at the Dartmouth conference, the first working, characteristically
AI program developed) employed non-numerical techniques (19). During the
late 1950s, programming languages designed specifically for non-numerical
symbolic information processing, were developed by those two researchers,
along with J. C. Shaw (20), and in 1960, McCarthy formulated LISP (llit
processing language) which became, and still is, the most widely used AI
language (21). During the late 1950s, it also became clear that organizational
techniques of search were of paramount importance in attaining the desired
ends, and that the numerical aspects were of secondary importance. The
principle of looking for and using certain heuristics, that is, rules of thumb
which might help in finding a solution but which would not guarantee a

178

James Fleck

solution, became established (22). By the early 1960s, various successful

programs had been written, resulting in a general air of optimism, and indeed
by this time the paradigmatic structure of AI had been elaborated in essentially its complete form, as already described.
5. The Establishment in the United States
At first sight it might seem remarkable that McCarthy, Minsky, Simon, and
Newell, without doubt the four 'great men' of the AI establishment, all
should have been present at the Dartmouth meeting. But I would argue that
the emergence of the American establishment in AI was part and parcel of the
process of defining the paradigmatic structure of research in the area and the
organizational structure of the field.
In the first place these four were actively involved in the organization of
the field. McCarthy, as already noted, arranged meetings to bring together
those interested in the very loosely-defined goal of constructing genuinely
intelligent machines. Subsequently, he went on to found a group at MIT,
along with Minsky (a fellow student with him at Princeton), and later the
group at Stanford. Simon and Newell developed the group at Carnegie.
In the second place, these four were centrally involved in defining the
substantive cognitive elements of the AI paradigmatic structure already
outlined. Simon and Newell, as well as presenting the rust working AI program, the Logic Theorist, had also developed a series of Ust processing
languages, the IPL series, the forerunners of the basic element of the AI
research activity. McCarthy had produced the definitive AI programming
language LISP on the basis of these forerunners, and had also incorporated
certain features which embodied the emphasis on logic rather than numerical
mathematics. Minsky had written an influential systematizing paper which
explicitly outlined the importance of heuristic search. As well as producing
the basic tools for subsequent research, these four also defined in broad terms
higher-level guidelines for future research. Simon and Newell pioneered the
focus on investigating human cognitive processes as a source of inspiration for
computational models, while McCarthy and Minsky went in more for the idea
of investigating mechanisms for achieving intelligent activity in the abstract,
without prejudice towards specifically-human forms.
The fact that these four were involved in the forming of the AI research

Development and Establishment in Artificial Intelligence

179

activity from the beginning has contributed twofold to their success in

becoming members of the establishment. In the first place, the founders of
any field, simply because at that stage they are competing with fewer people,
gain a visibility which later contributors are unlikely to attain, unless they in
turn can produce work that will lead to subsequent distinctive and fruitful
development. In the second place, these four have been around for a long
time, and consequently the process of normal scientific career progression has
ensured their continued visibility and position within the establishment (23).
Moreover, once recognized as members of the establishment, they have, in
fact, continued to be influential within the field. McCarthy's suggestions for
a mathematical theory of computation (24), and his emphasis on the use
of the predicate calculus have been themes which have been taken up and
developed. Minsky has continued to produce influential synthesizing research
programmes, as for example his presentation of the theme of semantic
information processing (25), or more recently with his explication of 'Frames'
- high-level data structures for organizing and mobilizing the vast knowledge
bases with which effective AI programs have had to work (26). He has also
used his establishment position to argue effectively against the formal theorem
proving strand of research (27). Simon and Newell have continued to pioneer
new approaches at the psychological interface of AI and the 'production
systems' formulation developed out of their earlier work on general problemsolving, and promoted by Newell, has been widely taken up (28).
However, the members of the AI establishment did not arrive from nowhere. Their success, without doubt, owed much to their having attended
prestigous institutions as students, and to their being sponsored by people
who were already members of the wider scientific establishment, not necessarily in the cybernetic area. McCarthy, Minsky, and Newell all attended
Princeton as graduate students, for instance, and McCarthy worked for
Shannon on the organization of the 1952 Automata Studies conference,
while Minsky was associated with W. McCulloch whose 1943 paper with Pitts
is recognized as another of the texts marking the emergence of cybernetics
(29). Simon had already established his reputation in the fields of political
science and economics (30), and he himself acted as sponsor for Newell.
Moreover, this intergenerational establishment reproduction process
continued, and students of McCarthy, Minsky, Simon, and Newell have
dominated the field by and large. Figure 1 gives a graphic illustration of the

180

James Fleck

m
IIWiener -Rosenbleuth

Princeton
von Neumann

I
I

Morgetstern

7 L

\Sha'fmon -McCu1.1och

Newell
SRI

cCart

CMU

.)L

{FeigenbaUlll.}-- ---<
Hart
Duda
(Manna

Nilsson

(Feldman

(Evans

Lederberg
Buchanan

<Ems

f--(Fikes)

Walker

<- !---<Waldinger)

Sacerdoti

<- !---<ColeS)

Fischler

(Quinl

(Colby

Green)

Pople

Binford
<-

Shortliffe
<---

(Quillian

Slagle)

Banerji

Hewitt

Kaplan
Eastman

Robinson
Abelson

(Greenblatt}---

lTenenbaum)

Luckham

Wang

Raphael)

Earnest
Floyd

Elsewhere

Bobrow)
(Guzman

(Falk
(Montanaro

<Roberts

Samuel
<-

Berliner
Barrow

Minsky

Papert

Fredkin
Winograd)

Pohl
Schank)

Amarel

Charniak
McDermott
Bledsoe)

Sussman

Boyer

HWlt

Goguen_

Goldstein

Rieger

Marr

Sinunons

(Waltz
Winston

Rosenfeld

Nevins

Fig. 1. The Establishment in the United States: 1960-mid-1970s. The members of the
establishment were derived from a consideration of the editorial board of the Arti/icial
Intelligence Journal, conference organizing committees, invited conference s,Peakers and
panel members, supplemented by well known researchers as judged on the basis of a
reading of the literature. They include 73 out of a total of upwards of 500 contributors
to the area. Available data was limited, but indicated that only some 11 out of the 73
had not worked at some time or done a PhD at one of the big four Artificial Intelligence
centres: Massachussetts Institute of Technology (MIT); Carnegie Mellon University
(CMU); Stanford University (SU); and Stanford Research Institute (SRI). At least 24
of the 73 received their doctorates from one of MIT, CMU, or SUo There is no particular
significance in the ordering, nor is the record of movements complete. Intergenerational
and inter centre linkages are probably underestimated due to lack of data.
movement of personnel.
student of! worked for.
colleague relation.
Boundary between the wider establishment, and the Artificial Intelligence establishment.

Development and Establishment in Artificial Intelligence

181

prevalence of such links - links which have led to charges of nepotism being
levelled at the AI establishment (31). This structure of very strong intergenerationallinkages turns out to be characteristic of the development in Britain as
well, and in the section on the establishment in the United Kingdom, some
underlying reasons for the strong linkages are discussed.
The emergence of this group as the establishment in AI was undoubtedly
consolidated by their success in getting the backing of the United States
Department of Defense, mainly through the Advanced Research Projects
Agency (ARPA), which provided some 75% of United States AI funding for
the ten years from 1964, and through the Air Force (32). Furthermore, the
preference on the part of ARPA for concentrating resources in a few selected
centres guaranteed the position of the establishment, especially in view of the
great expense of adequate computing facilities, which effectively barred other
groups from competing.
Another aspect of development in AI that has been characteristic and of
importance for the field, and that has served to further reinforce the position
of the establishment should be noted. This is that the general aim of research
in the area to product intelligent machines haS excited extreme reactions and
has tended to lead to very strong for and against (33) alignments. Such a
reaction is not at all surprising given the sensitivity of such a goal to peoples'
images of themselves. Here Elias' comments about the competition for the
monopoly over the means of orientation are relevant (34). The AI approach
is seeking to establish and legitimate a view of intelligence and the nature
of mind which challenges the received commonsense view of mind and
intelligence as something rather special and certainly well beyond the reach of
scientific analysis. Moreover, this received view is very much under the sway
of the religious establishments, or where religious authority does not hold,
under the sway of a liberal humanist tradition. Strong reactions are commonplace in AI and, on the sociological level, have probably had the effect of
heightening the difference between those on the inside and those on the
outside, and consequently have reinforced and concentrated the position of
the establishment.
Thus it can be observed that the emergence of the American establishment
was very much bound up with the development of AI as a distinctive area
of research, and their position was consolidated by their success in gaining
backing from the Department of Defense. The American establishment was

182

James Fleck

not only involved in providing an organizational basis for research in the area,
but was also very closely concerned with the elaboration of a distinctive
cognitive basis for research in the area, the AI paradigmatic structure. In the
following discussion of the development of AI in the United Kingdom, some
of the themes already introduced will be reiterated, while other issues will
become evident.
6. Development in the United Kingdom
In Britain during the 1940s, there was a similar flourishing of interest in
general cybernetic concerns as occurred in the United States, and discussions
of the possibility of machine thought were common (35). A. M. Turing was
an enthusiast for the possibility of intelligent machines, and his 1947 and
1950 papers still stand in many respects as definitive surveys of the arguments
for and against AI (36). R. J. W. Craik, whose 1943 book The Nature of
Explanation is recognized as one of the texts marking the emergen.ce of
cybernetics, wrote passages that bore a remarkable foreshadowing of the
actual AI paradigmatic structure, as for example in the following passage:
... thought models, or parallels, reality - that its essential feature is not 'the mind', 'the
self', 'sense data', nor propositions but symbolism and that this symbolism is largely of
the same kind as that which is familiar to us in mechanical devices which aid thought and
calculation (37).

Others were interested in more specific cybernetic approaches: W. R. Ashby,

whose name is perhaps second only to N. Wiener's in association with cybernetics, contributed many ideas and books on the subject (38); W. G. Walter,
who achieved a degree of fame with his electronic tortoises which exhibited
elementary reflexive behaviour (39); F. George who wrote on a cybernetic
approach to the brain (40); and A. M. Uttley, D. Mackay, and others who
worked on neural net models of cognition and perception. Furthermore,
through the informal RATIO club which existed during the early 1950s (to
which Turing, Mackay, Ashby, Walter, Uttley and others belonged) there was
frequent interaction and discussion on these issues - discussion which often
involved researchers from the United States as well: McCulloch for example,
attended the first meeting ofthe club in 1949 (41).
These discussions continued through the 1950s. There were several British

Development and Establishment in Artificial Intelligence

183

contributors to the 1952 Shannon and McCarthy Automata Studies meeting,

and in 1958 a conference on the 'Mechanisation of Thought Processes' was
held at the National Physical Laboratory mEngland, to which people such
as McCarthy and Minsky contributed as well as the proponents of other
cybernetic approaches and practioners of the art of using the digital computer
(42). Indeed, it was there that McCarthy presented his 'Advice Taker', a
suggestion for using the predicate calculus for modelling common sense
reasoning, an approach which proved influential within the more specific AI
context (43).
Nevertheless, despite all the activity and discussion in the area, the specifically AI approach was not developed in Britain until the mid 1960s, and then
the prime mover was someone quite external to the cybernetic network. This
was Donald Michie, a geneticist at Edinburgh University. During the war he
had worked with Turing at metchley and became fascinated in the possibility
of constructing machines that could think (44). There were no opportunities
for him to follow up these interests after the war, except on the hobbyist
level, so he took medical sciences at Oxford and subsequently specialized in
genetics (45). In 1962, however, during a visit to the United States (arising
out of his hobbyist work on tri!l1 and error learning in the game of noughts
and crosses) he became aware of the developments in AI there and was
impressed by the computer facilities available. On his return from the United
States, and depressed by the lack of such facilities in Britain, he became
active in agitating for something to be done; lobbying, writing newspaper
articles and so on (46). At about the same time, he had grown dissatisfied
with his position in the Department of Surgical Science at Edinburgh, where
he held the post of reader, and in 1963, very much on his own initiative, he
moved out of the department and with his secretary and part-time helpers set
up an unofficial unit - the Experimental Programming Unit. Meanwhile he
continued his lobbying for improved computer facilities and had buttonholed
C. Jolliffe, the deputy grants director of the then Department of Scientific
and Industrial Research (DSIR) whom he impressed with his concern over
the state of United Kingdom computer provision, and was consequently
commissioned to carry out a survey of computing interests and views among
British scientists (47). Lord Halsbury had just been charged with setting up a
spcialized computer board in the DSIR (subsequently the SRC) with the aim
of reviewing and supporting research in computer science. The results of

184

James Fleck

Michie's report indicated a widespread positive assessment of the potential of

AI among young computer scientists and this undoubtedly formed the basis
for the proportionately generous funding by the computing board of the
SRC for research in the area during the mid to late 1960s (48). This took
place against the background provided by the Flowers Report of 1966 on
computing in universities and colleges, which recommended a large expansion
in computer provision and training, and set the outlines of the presently
existing regional computing centre structure, based in London, Edinburgh,
and Manchester (49). In this context, Michie, with his energy and enthusiasm,
was seen very much as a bright young man, and he succeeded in attracting
several large grants (50), which no doubt encouraged the university authorities
to give official recognition in January 1965 to his irregularly set up unit,
which grew rapidly over the next few years.
It is interesting at this point to consider the substantive lines of research
that emerged at Edinburgh (51), for the research proftie bore a remarkable
similarity to the patterns evident in the United States. Michie's own immediate
concerns were with game-playing and heuristic search, then regarded as central
to the field. Also in the experimental Programming Unit there was a project
involved with human problem-solving studies, similar in flavour to the work
of Simon and Newell at Carnegie, and there were also projects concerned with
developing the instrumental base for AI research. One element was the MiniMac project, an interactive multi-access system (and the second such system
to be developed in Britain) so called after the first project of its kind, carried
out on a grander scale - project MAC at MIT, and also partly motivated by
AI concerns according to McCarthy (52). Another element in the instrumental
base was the development of a list-processing language POP-2 which became
the staple AI language in the United Kingdom (and with refinements is still in
use today) (53), just as LISP was the staple in the United States.
An independent group worked in the Metamathematics Unit with a focus
on automatic theorem proving. This unit had been set up by Bernard Meltzer,
a reader in the Department of Electrical Engineering who had become interested in the use of the computer in the course of his work and combined
this interest with his hobby of mathematical logic. He, like Michie, took the
fairly dramatic step of moving out of his official department and into a new,
unofficial unit after a visit to the United States, where he had visited similar
projects (54). Due to their common use of symbolic rather than numerical

Development and Establishment in Artificial Intelligence

185

information processing, and the intelligent nature of their goal, theorem

proving, these approaches had developed as an autonomous but central strand
in the AI area of research.
Thus, by the mid 1960s, there had emerged in Edinburgh a centre for
research in AI which reproduced the same features as had developed in the
United States. In September 1965, Michie organized the first of a series
of meetings - the Machine Intelligence workshops - which were held in
Edinburgh and attracted leading AI researchers from the United States, as
well as interested people from elsewhere in Britain (55). These workshops
played an important part as a forum for discussioQ and the communication
of the AI approach, and influenced such people as E. W. Elcock and J. M.
Foster of Aberdeen University, where in an SRC-sponsored computer unit
they worked on game-playing programs and high-level programming languages
and systems which incorporated some AI elements (56). M. B. Clowes was
another interested researcher who attended. Encouraged by Michie, he had
set up AISB (the Society for Artificial Intelligence and the Simulation of
Behaviour) in 1964, which, after an informal and hesitant start has grown
into a thriving learned society for the area, holding two yearly conferences
and publishing a regular newsletter (57). Clowes too had visited the United
States and had been impressed by what was happening there. He did not work
on specifically AI projects in the mid 1960s, and depressed by the lack of
suitable computer facilities, he went to Australia for several years. However,
he had met and impressed N. S. Sutherland, who had been interested in
mechanistic models since the 1950s, and who built up a centre in experimental psychology at the newly established University of Sussex at Brighton
in the 1960s. On his return from Australia, Clowes went there to work in AI
on a grant held by Sutherland. This work heralded the emergence of Sussex
as a major centre for AI in the 1970s.
It was during meetings of the AI community following the technical
business of the workshops that suggestions were first put forward for starting
a specialty journal for the area -;- Artificial Intelligence - An International
Journal, which was eventually founded in 1970 and of which Meltzer became
the editor (58). It also appears that suggestions for establishing international
conferences on AI were discussed at Machine Intelligence Workshops (59) the first was held in 1969, and since then these conferences (held every
second year) have grown steadily in size and importance (60).

186

James Fleck

Thus, it is clear that the Machine Intelligence Workshops were of great

importance for the social development of the field at the international level
and consequently firmly established Edinburgh as an AI centre of international repute. This reputation was further enhanced when Michie succeeded
in attracting to Edinburgh from Cambridge the research group of Richard
Gregory, the psychologist who became known for his book on the eye and
brain (61), and H. C. Longuet-Higgins, a theoretical chemist of international
standing. The basis for the merger was the goal of building an intelligent
robot - a goal that was also being pursued at the other major research centres
in AI in the United States; MIT, Stanford University, and Stanford Research
Institute.
The robot project was seen to pose a challenge for AI in that it required
the integration of many of the strands of work within the area - machine
vision, problem solving (often based on theorem proving methods), manipulation of a hand in three-dimensional space, and even natural language
for communication. Gregory's group was to provide the perception and
engineering aspects, while Michie and Longuet-Higgins worked on the problem
solving and cognitive aspects. For this project the SRC awarded a major grant
and provided a new computer. In addition the Nuffield Foundation provided
funding for eqUipping an engineering laboratory to build the robot and
associated hardware and other devices (62).
Part of the deal involved in attracting these research groups was that the
University of Edinburgh, largely as a result of the good offices of Michael
Swann (now Lord Swann), then the Vice-Chancellor, would set up a new
department - the Department of Machine Intelligence and Perception and provide chairs for the new senior people involved (63). This was the
first specifically AI focussed department anywhere in the world, and was
seen by some as an exciting venture, though others were less welcoming. In
particular, as an independent institutional entity in the university, it was in
direct competition with the new computer science department, resulting in
rather distant and, at times, antagonistic relations between the two departments (64). This contrasted with the situation in America, where AI was
usually carried on within the departments of computer science and electrical
engineering.
This institutional innovation would not have been possible had it not been
for two favourable factors. The first was the general context of university

Development and Establishment in Artificial Intelligence

187

expansion of the early and mid 1960s. This expansion, in fact, started being
curtailed just after the establishment of the new department, and the resulting
squeeze contributed to the problems that beset the department, as the level
of University Grants Committee support could not keep pace with the large
Research Council funding that the new, rapidly growing area attracted. Had
the curtailment of expansion come a few years earlier, it is highly unlikely
that a separate department would have been approved. The second factor
was the support afforded by Swann: as Dean of Science he had backed
Michie's previous initiatives, and newly-elected in January 1966, as Principal
of Edinburgh University, he was very receptive towards new departures and
constantly promoted the status of Edinburgh as second only to Cambridge
in research (65). Without such sponsorship, it is again doubtful whether a
new department would have been instituted, or whether Michie would have
succeeded in attracting Gregory and Longuet-Higgins. However, the department was established in October 1966, and while in the event the institutional
attractions were evident, it is interesting to consider the scientific motivations
for these people with a non-AI background to change their area of research.
Gregory had engineering interests which led him to seek a new methodology involving a closer study of the physical basis in the brain for perception
and cognition than was usual in psychology at that time, and the AI approach
seemed to promise developments along these lines (66). He also brought with
him other members of his group, notably -S. H. Salter who had extensive
engineering competence and built the robot hardware (and was later to
become known for his wave power system - the Salter duck) and J. A. M.
Howe, a psychologist who was to explore the applications of AI in educational
research, and who became head of the AI department at Edinburgh in the late
1970s.
Longuet-Higgins was very much a scientific high flyer, achieving international distinction in his work in theoretical chemistry with C. A. Coulson
at Oxford, and gaining a professorship at the early age of thirty. For his
eminence in Chemistry, in 1958 he was elected a Fellow of the Royal Society,
and in 1968 a Foreign Associate of the United States National Academy of
Science, the highest American honour available to someone not a United
States citizen. Despite his great success in chemistry, or perhaps because of it
in that he was motivated to seek similar success in a new and potentially
exciting but unexplored field, he had joined with Gregory in planning a Brain

188

James Fleck

Research Institute. Negotiations were in hand for funding from the Nuffield
Foundation, and for accommodation at Sussex University, when Michie, at a
meeting with Gregory early in 1966, suggested that Edinburgh would be an
ideal centre in view of its already established AI work. With the institution of
the new department at Edinburgh, Gregory's group and Longuet-Higgins
moved to Scotland, and great hopes were entertained for the future of
coordinated research in the area.
The anticipated cooperation failed to materialize. Problems over accommodation, personal, political, administrative, and scientific factors were involved
in what became a very complex and confused situation during the late 1960s
and early 70s. Gregory never really settled in at Edinburgh nor became
involved with the computational approach though he remained favourably
inclined towards it, and in 1970 he finally left to go to Bristol University. The
engineering workshop in the Bionics Laboratory had proceeded, however,
with the building of the robot hardware, and a prototype was connected to
the computer for the first time in May 1969. Longuet-Higgins did absorb the
computational approach, but irreconcilable differences between him and
Michie over the installation of the new computer and the robot project, as
well as over their approaches to work in the area, soon emerged and resulted
in Longuet-Higgins moving into separate accommodation and thenceforth
running his unit (then called the Theoretical Section) quite independently
apart from access to the common facilities.
Michie favoured a rather swashbuckling style of directing large team
projects oriented to goals which could be linked with industrial applications
and, in fact, was involved in launching a university based company to market
compiler systems and other software for the POP-2 language, which was
developed in the department (67). In addition, he was extremely energetic
and persuasive and very successful in obtaining funding from many different
sources (68). Longuet-Higgins in contrast, favoured a more restrained, academic style, preferred an individual basis of working with a few colleagues on
research chosen purely for its intrinsic scientific interest, and was dubious
about the advisability of mixing commerce and industry with research. These
differences in style, aggravated by contrasts in personality, were associated
with conflicting views on AI: LonguetHiggins thought that 'artificial intelligence' was not a science or technology in its own right, but was a new way
of tackling problems in those existing sciences which were relevant to the

Development and Establishment in Artificial Intelligence

189

phenomena of intelligence. It set new standards of precision and detail in the

formulation of models of cognitive processes, those models being open to
direct and immediate test (69). Michie's position, on the one hand, was closer
to the view:
.. , that success in achieving the long-term aims of Machine Intelligence should be
regarded as the major goal of Computing Science. Furthermore, progress in Machine
Intelligence is continually generating pressures for solutions to fundamental problems
of Computing Science in an environment where they will be used; an environment which
by its very nature, demands quality and generality. As a result a decision to invest
apparently disproportionate sums into Machine Intelligence could only have beneficial
effects to the whole of Computing Science (70).

This lack of consensus among the practitioners within AI was undoubtedly

a complicating factor when the SRC Computing Science Committee came to
review its funding policy in the early 1970s. The 'cognitive science' view of
AI put forward by Longuet-Higgins was not seen by the reviewing panel as
falling within its scope, while there was strong opposition to the view that
machine intelligence should be regarded as a major goal of computing science.
These differences in views were noted in the SRC Computing Science Review
of 1972, and were given by Sir Brian Flowers, then chairman of the SRC,
among the reasons for commissioning Sir James Lighthill to review the field
(71), a review that was to have a large impact on the area as will be discussed
in due course.
The differences between Michie and Longuet-Higgins also caused great
internal problems at Edinburgh resulting in frequent appeals being made
to the Principal and Secretary of the University, and to the SRC, and a
bewildering sequence of organisational forms were instituted by the university
authorities in attempts to alleviate the embarassing situation; but to no avail.
Michie, however, started losing the support of his other colleagues, and fmally,
in 1974, following a lengthy and, to Michie, unsatisfactory, review of the
situation (72), a new Department of AI was set up with Meltzer at its head
and comprising most of the AI research groups in Edinburgh. Michie was
given his own, independent, Machine Intelligence Research Unit. The bulk of
the very considerable resources which had been built up over the decade,
including the robotic equipment, was settled with the Department, and,
furthermore, limits were placed upon the scope of future research efforts
by Michie.

190

James Fleck

But despite the tension between the senior people at Edinburgh, there
was a thriving research environment in the late 1960s and early 1970s, with
frequent visits by people from elsewhere in the world, including the United
States and the up and coming Japanese AI-oriented groups (73). Young
researchers were very successful, sometimes gaining an international reputation before obtaining their doctorate: P. J. Hayes for example, became well
known after publishing a joint paper with J. McCarthy in 1968 (74), only
receiving his PhD. in 1972.G. D. Plotkin's work on inductive inference was
considered outstanding (75); R. Kowalski made a name for himself with his
vigorous promotion of the predicate calculus as a programming language in its
own right, an approach which became an independent strand of research
termed 'Logic Programming' (76); while R. M. Burstall, one of the original
members of the Experimental Programming Unit, and Michie's second in
command, became established as an outstanding computer scientist with an
international reputation in his specialist area - the Theory of Computation,
in which he built up his own group in the 1970s. In 1978, he was appointed
to a chair with the title 'professor of AI' despite the fact that computation
theory was by that time a general computer science research area, rather than
a specialist AI one. The robot project attracted considerable pUblicity with
some five television and ftlm crews visiting (77): indeed demonstrations
became so frequent as to interfere with the everyday research work and had
to be restricted (78).
However, the concentration of talent, the surfeit of publicity, and perhaps
more than anything else, the predominance of research over teaching in AI,
attracted hostility from other departments weighed down with heavy teaching
responsibilities (79), and strong pressures grew for the area to normalize its
activities. In addition the lack of a career structure for researchers on shortterm contracts, coupled with the increasing uncertainty over the future of the
centre due to the leadership tensions led people to start moving elsewhere:
Hayes went to Essex; Kowalski to Imperial College; several other researchers
to the United States; and Longuet-Higgins himself, along with members of
his group, moved to Sussex University. In the course of a couple of years,
therefore, many of the most highly respected researchers left Edinburgh, and
in some quarters Edinburgh was viewed as being in decline (80).
The problems facing AI were not restricted to Edinburgh alone, nor was
the division of the department the outcome of purely local politics. Rather,

Development and Establishment in Artificial Intelligence

191

these events were tied up with national attitudes especially on the part of the
SRC. The SRC had never been happy with the breakdown of cooperation
over the major robotics grant and had become dissatisfied with the progress
made in work on the project. This dissatisfaction stemmed to some extent
from a basic lack of sympathy with the goals of those involved in the project.
Michie's very ambitious plans for a seven-year industrially oriented programme of research in robotics failed to win favour with the SRC and was
never formally submitted (81). More modest proposals were put forward to
maintain the level of effort on robotics, but, despite a years very intensive
work on the project, in which programmable assembly using visual recognition
of parts was attained (82) (at that time one of the foremost achievements in
robotics in the world, comparable to leading work in America and Japan),
these proposals were turned down. At that stage, the SRC had become
very impatient with Michie, as his entrepreneurial talents did not fit in with
their expectations, and the previously-mentioned survey of AI by Sir James
Ughthill, FRS, Lucasian Professor of Applied Mathematics at Cambridge, and
an eminent hydrodynamicist, undoubtedly influenced their decision.
Lighthill's report created a major controversy and was published' in April
1973 along with other assenting and dissenting views. Lighthill was highly
critical of AI in general and suggested there were three basic categories of
research in the area: work aimed at advanced automation on the one hand,
and at computer based Central Nervous System research on the other, with
in addition a bridge category with the basic component of building robots,
which he saw as the essential underpinning for AI to have any claims to unity
and coherence. Progress in this cateogry Lighthill suggested was virtually nonexistent and the building of robots a mistaken enterprise possibly motivated
by a desire on the part of those concerned to 'minister to the public's general
penchant for robots by building the best they can', and possibly also by
'psuedomaternal' drives to compensate for male researchers' inability to
give birth (83). (It is not hard to detect a reference to Michie's polemical
enthusiasm in these comments.) Furthermore, what success there had been,
he suggested, was evident only in particular applications and derived from
knowledge contributed from the substantive fields modelled, rather than
from any AI component. In time, he saw the bridge category as withering
away, while work directed towards the two extremes would become integrated
with other research in their general areas.

192

James Fleck

Not surprisingly, this caused a major stir in the AI community across the
world (84), and the resulting controversy received much public airing in
the press and even on television (85). Without a doubt, despite Ughthill's
protestation that his report:
would simply describe how AI appears to a lay person after two months spent looking
through the literature of the subject and discussing it orally and by letter with a variety
of workers in the field and in closely related areas of research (86),

it delivered a blow to the prestige of research in the area from which it has
never fully recovered. While UghthiIl's comments on robots were directed
at the specifically AI category, it appeared they also had some effect on
inhibiting robot research and use in Britain in general (87), whereas in other
countries robotics has been a steadily expanding area throughout the 1970s.
In practical terms, the report did affect financial support for research in AI
in Britain, particularly in the case of Michie's proposals for robotics research
and also had some influence on funding in the United States where AI robotic
projects were cut back, and the Advanced Research Projects Agency , ARPA,
the main sponsor of American work in the area, started insisting on missionoriented direct research, rather than basic undirected research (88). These
cutbacks took place in the context of the general reduction in public spending
in the early to mid-1970s, which affected scientific research in all areas,
especially those not seen to be of 'social relevance'. However, the effects
were to some extent mitigated in the case of AI: partly by the variety of
funding sources supporting the area; partly by the SRC's identification of
machine intelligence as an important area of long range research in its 1972
Computing Science Review, which underlined the fact that there was, in
any case, no one predominating view on the value of AI; and partly by the
expanding nature of computing science in general. Consequently, particular
projects were able to get support, especially if their relevance was emphasized
and explicit reference to robotics avoided (89).
Furthermore, the debate over the Ughthill report also led to cognitive
science being recognized by the SRC, and a panel was set up to review applications in this area. Thus there was to some extent a shift in resources to
this area, rather than a straightforward cut back of AI as a whole.
That the reorganization of AI at Edinburgh, with the effective removal of
Michie from a central position, was not a purely local affair, but was rather

Development and Establishment in Artificial Intelligence

193

bound up with changes in attitude in the wider scientific establishment, given

expression by Ughthill, is borne out by the similar pattern of events occurring
at Aberdeen (90). There the computer unit was dissolved in 1972, after the
SRC refused to renew its grant. Elcock, the organizational prime mover
behind the AI interests there (and a colleague of Michie's) had some differences of opinion with the university authorities and the newly-established
computer science department, in which he was not offered a position to
his satisfaction, and left for Canada, where he became involved in building
up another AI research group. J. M. Foster, the other senior figure in the
computer unit, went to Essex University, where R. A. Brooker, at that time
chairman of the Computer Science Department, was in the process of building
up AI research interests in an attempt to dispel the non-publishing lethargy
prevalent there (91). However, the situation at Aberdeen never became quite
so fraught as at Edinburgh.
In Edinburgh, Michie attempted to fight back against the reorganization.
He marshalled support from the Dalle Molle Foundation to keep research
going under his direction (92), and he had previously published an article in
the New Scientist, in response to Ughthill's report, asking why it was that AI,
which required peanuts in financial terms should suffer cuts, while nuclear
physics, absorbing huge amounts of money and providing little proportionate
return should not be cut (93): but to no avail. Essentially, by the time of the
reorganization he had lost the support of his colleagues, and Swann, who as
principal had supported and protected Michie, had left in 1973 to take up the
chairmanship of the BBC (94).
In the Machine Intelligence Research Unit after 1974, Michie's energies
were channelled into promoting, directing and carrying out research on chess
playing programs and organizing further Machine Intelligence conferences,
one in the United States with Elcock (95), one in Russia (96), and a third,
again in the United States. He spent a considerable amount of time on visiting
professorships abroad and took up scientific journalism, where in his regular
column, "Michie's Privateview" in Computer Weekly, he often commented on
the importance of AI research. Latterly, in the late 1970s and early 1980s, he
started vigorously promoting 'expert systems', (AI frameworks for representing and mobilizing highly detailed specialist knowledge, which, given their
essentially simple structure, have achieved remarkable levels of competence
comparable to those of human experts), by organizing conferences and

194

James Fleck

schools to disseminate the approach to industry (97), as well as directing and

sponsoring relevant research in his unit.
Once the new regime of the Department of AI had settled down, activity
did become normalized, with the conventional university emphasis on teaching coming to the fore. This owed much to the mid to late 1970s fmancial
stringencies ensuring that the universities looked to efficiency in their sphere
of production, namely the training of students. Undergraduate course were
experimented with at Edinburgh, and an AI textbook produced (98). The
application of AI in education itself - intelligent computer aided instruction
- clearly an area of direct social relevance - became a major concern at
Edinburgh. This was an area originated by, among others, Seymour Papert,
a colleague of Minsky at MIT, and was taken up by Michie and refined and
developed at Edinburgh in a variety of approaches by J. A. M. Howe and his
colleagues. In general, research activity remained at a high level, producing
some 250 publications during 1975-1980, and with a strong postgraduate
school of about thirty being built up after an initial weakening due to the
lighthill report.
In some quarters, however, there was the impression that Edinburgh had
declined in importance as an AI centre. This impression was partly due to
the departure of highly respected researchers from Edinburgh, as already
mentioned, but there were also other contributing factors. Firstly, Edinburgh
no longer attracted the publicity over the robotics work as it had formerly
done, though research with the robot equipment continued at a modest level;
indeed publicity was shunned as a matter of departmental policy, because of
what was felt to have been overexposure by lighthill. Secondly, an influential
PhD. thesis by T. Winograd in 1972 at MIT (99) had brought the Natural
Language research area to the centre of the AI stage, displacing the Theorem
Proving approach, a shift which owed something to a 'witchhunt' against
theorem provers led by Minsky (100). Edinburgh, however, with a strong
with the research group of Bernard
tradition in this latter area
Meltzer, at that time professor of Computational Logic and head of the
department, had no strong competence in natural language and did not
appoint a specialist on a permanerit basis. Theorem Proving had developed
a very strong internal theoretical dynamic, deriving from mathematical logic
and based on refming the 'resolution' method of machine-oriented inference
devised by J. A. Robinson in 1965, which had brought it into the centre of

Development and Establishment in Artificial Intelligence

195

in AI the first place (101). This internal dynamic, coupled with the Theorem
Provers' assured confidence in their formal mathematics-based status, had
led to their being always rather autonomous; and under easier funding conditions, they would probably have become a completely independent specialty
of computer science, much as Pattern Recognition, itself based on a strong
internal dynamic, had done in the early 1960s (l02). However, the Theorem
Provers made somewhat of a comeback in the late 1970s with the logic programming approach embodied in the language PROLOG, thus re-establishing
themselves to some extent as a source of techniques of utility to AI in general
(103).
A third contributing factor to the perceived decline of Edinburgh as an AI
centre was that computation theory, another of the major research themes in
the department there, under the leadership of R. M. Burstall, professor of AI,
had become more central to computer science in general during the 1970s
and less of a specifically AI approach (104). This situation was rationalized in
the late 1970s with the transfer of the Computation Theory group from the
department of AI into the department of Computing Science. A fourth and
fmal contributory factor to the perceived decline of Edinburgh, was that
other major concentrations of AI interest had developed in Britain: at Sussex,
Essex, and later in the Open University. At the same time, numerous oneperson AI projects were pursued elsewhere, often based on 'colonization' or
'infection' from the established centres.
Professor R. A. Brooker, always an enthusiast for AI (he was one of the
panel members in favour of AI in the 1972 SRC Computing ReView), had
wanted to invigorate the research atmosphere in the computer science department at Essex, and to this end had recruited J. M. Foster from the Aberdeen
AI research group to a chair in the department (105). However, this did
not work out and Foster left after about a year, having done little on the
AI research side, but having developed the elements of a course in the area.
There was considerable interest in the AI approach on the part of young
researchers in the department such as J. M. Brady and R. Bornat, and when
P. J. Hayes arrived in the late 1972 from Edinburgh, bringing with him his
extensive familiarity with the AI literature and research front, he catalyzed
the development of research projects in the field at Essex. One such project,
supported by the SRC, was the development of a system to read handwritten FORTRAN coding sheets using high level knowledge to guide the

196

James Fleck

interpretation - a project of clear practical utility (although it never paid off)

that nevertheless incorporated the AI approach (106).
Once established as a centre for AI other people were attracted there. J. E.
Doran, who had been an early member of the Experimental Programming
Unit and the Department of Machine
and Perception in Edinburgh, joined in 1973, bringing with him his research interests in using AI
techniques for the reconstruction of cultural evolution in prehistoric settlements from data arising out of the archaeological excavation of graves - yet
another illustration of the divergent nature of the AI activity (107). Bruce
Anderson, also from the Edinburgh AI centre,joined for a while before going
to the department of Electrical Engineering, while Yorick Wilks, who made a
name for himself with his AI work on natural language and his numerous
publications, was appointed to a chair in the Department of Linguistics marking a new stage in the penetration of other subject areas by AI (108).
This blossoming of AI interests at Essex was aided by the energetic and
aggressive activity of Brady. He rapidly established himself as a competent
researcher in the field and became involved at the organizational level, being
elected to the chairmanship of the AISB society in the late 1970s, and
organizing a summer school, which attracted leading researchers from the
United States. He soon became known in the international AI community
and made frequent visits to America, eventually leaving Essex to go to MIT in
1979. The MSc course and research environment at Essex proved an effective
medium for the training of AI practioners, some of whom subsequently went
on to start AI groups elsewhere - notably H. J. Siekmann who built up a
group in Germany, B. Wielinga, in the Netherlands, and C. Bearden who
founded an organization similar to AISB in New South Wales in Australia.
The level of interaction with other centres was high, particularly those in
Britain. There were links with Edinburgh through Hayes' connections there,
and increasingly with Sussex through exchanges of students. In particular S.
Hardy, after doing the MSc course at Essex went on to do doctoral work in
AI and then moved to Sussex, where he was largely responsible for the design
of an AI computing environment for the cognitive studies programme (109).
The emergence of Sussex as a major AI centre in the mid 1970s had a
basis that went back to the mid 1960s. N. S. Sutherland, an experimental
psychologist who had been interested in the cybernetic and information theory
developments of the 1950s, was appointed to a chair at the newly-established

Development and Establishment in Artificial Intelligence

197

University of Sussex. He was favourably disposed towards the AI approach,

about which he had heard while he was at MIT in the early 1960s, and was
much impressed by the ideas of M. B. Clowes, whom he had met at Oxford
and who, as already mentioned, had founded the AISB society. Moreover,
AI had been taught by Sutherland as an ingredient of the experimental
psychology course from the start at Sussex, and he had arranged for two
people from Edinburgh - Burstall and Doran - to visit Sussex on a regular
basis to give lectures on technical aspects in the area. As part of the conditions for Sutherland coming to Sussex, he had insisted on the founding of a
brain research institute, and had negotiated with A. M. Uttley, Gregory, and
Longuet-Higgins, for them to join him (110). The latter two decided to go
to Edinburgh instead, but Uttley came to Sussex and started work on the
simulation of networks of an artificial neuron (the informon) (111), and on
the application of these to perception. This was clearly a cybernetic rather
than an AI project, but the computer provided for the project by the SRC
made possible a subsequent characteristically AI attack on machine vision,
which was carried out by Clowes when he returned from Australia in 1969
- work which rapidly established him as a leading figure in machine vision
research (112).
At that time in Sussex there was a broad base of interest in AI, favoured
by the explicit focus on interdisciplinarity of the distinctive Sussex 'school'
organization which contrasted with conventional departmental divisions and
their associated impervious boundaries (113), while Asa Briggs, then vice
chancellor of the university, was supportive of new ventures (114).
In 1970 Sutherland put forward a radical proposal for a new School of
Cognitive Studies with an intellectual focus on knowledge and understanding
to include teaching and research in a range of subjects: computing science
and AI; experimental psychology; linguistics; logic and philosophy; and
mathematics (115). But in the restrictions on growth of the universities in
the early 1970s, this proposal was turned down. However, a more modest
development, which came to be called the 'Cognitive Studies Programme' was
eventually launched within the existing School of Social Sciences, and in
1973 Clowes moved from experimental psychology to take up a chair in AI
instituted for the programme. Several other members of the university were
associated with the programme: among them M. A. Boden who had become
familiar with the computational approach during her doctoral research in the

198

J ames Fleck

United States on purposive behaviour in psychology (116), and whose book

Artificial Intelligence and Natural Man (117) is one of the most accessible
introductions to work in the area; and Aaron Sloman, a philosopher with a
mathematics and physics background who was influenced by Clowes and
Boden to consider work in AI and who spent a year in the Edinburgh centre
before returning to start research in vision at Sussex (118). His book, The
Computer Revolution in Philosophy (119) argues enthusiastically for the
great potential of the AI approach in matters philosophical. Meanwhile,
in 1974 Longuet-Higgins and some members of his research group from
Edinburgh had joined the Centre for Research in Perception and Cognition,
a research unit associated with the Laboratory of Experimental Psychology.
Following Longuet-Higgins' move, other members of experimental psychology, Professor P. N. Johnson-Laird and C. Darwin, also developed research
interests in AI (120), which was taught as a compulsory part of the course
presented there.
These two AI oriented groupings ensured Sussex emerged as a major
centre for AI, especially in the areas of language studies and vision, in the
1970s. Moreover, the emphasis on the 'soft' applications of AI - the study
of cognition - rather than on the 'hard' areas - the computer science and
engineering applications - marked the rise to importance of cognitive science
- the computational approach to linguistics, psychology, and philosophy,
based primarily on the methods of AI. This rise was encouraged by the
formation of a Cognitive Science panel in the S.R.C. Similar cognitive science
concentrations emerged elsewhere, usually as a supra-departmental federation
for research, as in the Edinburgh School of Episternics (founded by LonguetHiggins and others in 1969, but only having really taken off in the late 1970s)
and the Cognitive Science Institute at Essex, also implemented at the end of
the 1970s. At the Open University similar developments had occurred. The
Cognitive Psychology course included a substantial AI component, while the
presence of other researchers,interested in the application of AI to education,
meant that the Open University itself comprised an emerging centre for AI
in the late 1970s.
As well as being an essential element in the emergence of cognitive science,
AI has became a recognized specialty within computer science (121). While
these developments in the 1970s bear some resemblance to Lighthill's predicted fission of AI research between the categories of computer-based central

Development and Establishment in Artificial Intelligence

199

nervous system research on the one hand, and advanced automation on the
other, it is difficult to align his prediction with the continued coherence of AI
- that is, the continued identifiable existence of the AI paradigmatic structure. Rather than the established areas of linguistics, psychology, and philosophyabsorbing AI, it would seem that cognitive science is an emerging synthesis
based on the unifying computational modelling approach of AI. Indeed, one
could equally well argue, against UghthiU, the AI practitioner's extreme view
that what is happening is merely the process of colonization of other areas by
the AI approach:
I see the future of AI as a very long haul towards computational theories of physics,
chemistry, linguistics, sociology, visual perception, locomotion and every other aspect
of what it means to be human (122).

Clearly, therefore, views still differ over the assessment of the place and
future of AI.
7. The Establishment in the United Kingdom
One of the main features of the development of AI in Britain was the initial
and continuing strong American influence. Nearly every one of the leaders of
AI research in Britain had visited the United States and been impressed by
developments there, before moving into the area, or promoting it themselves
in Britain: Michie, Meltzer, Sutherland, and Clowes had all visited AI projects
in the United States in the early 1960s. These links with the American AI
community were maintained and strengthened in the ensuing years, and
continue today with a high international exchange of personnel between the
various centres.
These links also underlie the substantive similarity of research pursued in
Britain and America, at the general level of the paradigmatic structure, as
reflected for instance in the research profIle at Edinburgh which matched the
patterns evident in the United States. This has remained true since the mid
1960s when AI took off in Britain and, by and large, it has been the case that
the initiative in the development of AI in terms of the broad content of
research has remained in the hands of the establishment in the United States.
Consequently, the emergence of the establishment in Britain has been bound
up with organizational aspects of development to a greater extent than in the

200

James Fleck

United States, where organizational and general substantive innovation were

both important. Nevertheless, just as in America, the emergence of the establishment in Britain has been inextricably bound up with the development of
the field. The organizational aspects of development have not only predominated in Britain but have also had an appreciable impact on the shaping of the
wider international AI community: for instance, the Machine Intelligence
Workshops played an important role leading to the establishing of a journal
and an international conference structure for the area; the first dedicated
AI department was established at Edinburgh; and the Cognitive Science
concentration has received its firmest institutional expression in Britain, with
the Cognitive Studies Programme at Sussex, and the School of Epistemics
at Edinburgh.
Moreover, in the development in Britain, a clear division between what
have been called organizational and intellectual leadership roles (123) is
identifiable, especially with respect to the first generation establishment.
For instance, at Sussex N. S. Sutherland was energetic in supporting.the AI
approach there, getting grants for researchers' to work in the area without
himself being actively involved. Similarly, at Essex, R. A. Brooker deliberately
encouraged the development of an AI research group within the Department
of Computer Science, but did not himself actively contribute to research
to any great extent. It was also the case at Edinburgh with Michie, whose
contributions in substantive terms were overshadowed by his role as an
organizational leader: indeed he ranks as a scientific entrepreneur of the
first order. He alone, almost single-handedly, was responsible for getting AI
launched in the United Kingdom, and his influence appears to have lurked
behind nearly every event of major importance concerned with AI in Britain
in the 1960s (124). While his enthusiastic promotion of robotics was to
eventually backfire with Ughthill's condemnation, nevertheless his activity
was instrumental in putting Edinburgh on the map with respect to AI, and for
providing an environment in which young researchers were able to establish
international reputations. Despite his fall from favour after the Ughthill
report, Michie remained active - organizing a further three Machine Intelligence Workshops; promoting and directing research on chess playing programs; and, latterly, working to bring the 'Expert Systems' applications area
of AI to the notice of a wider audience, including indpstry. Also at Edinburgh,
Meltzer established himself as an 'elder statesman' for AI, channelling his

Development and Establishment in Artificial Intelligence

201

energies to maintaining high critical standards in the specialty Journal, and to

providing an environment in which major deveiopments in theorem proving
were made, while his own substantive contributions have been in the form of
synthesizing reviews, directing attention to certain problems and suggesting
fruitful possibilities. Even Longuet-Higgins, with his dislike for organizational
involvement, preferring purely scientific activity, has played a role of organizational importance, in arguing for the cognitive science applications of AI,
and in helping to found the School of Episternics at Edinburgh.
An important part played by the organizational role has been the provision
of facilities and opportunities for young researchers and students to rise into
the establishment, in many cases developing an international reputation for
themselves with their PhD work. The high number of important contributions
to the field at PhD level seems to have been characteristic of AI, and has
combined with the very fluid and informal nature of the organization of work
in the area, to maintain a shallow internal hierarchy with no elaborate vertical
division of labour. Terry Shinn's description of the organizational structure in
laboratories concerned with computer research in vector analysis could apply
equally well to the case of AI (l2S). While this organizational fluidity could
stem to some extent from the holistic and diffuse nature of the research goal
of modelling human intelligence, it seems likely that the situation also arises
out of the youth of the field. The preponderance of young researchers at
about the same stage in their lll"ofessional careers, and hence of more or
less equal status, together with the general shortage of personnel with the
appropriate computer background, has rendered it difficult to support a
steeply hierarchical structure - even where such a structure would be possible,
as, for example, in some well routinized commercial computing where there
has already been a drive towards extreme stratification and division of labour
along scientific management lines (126). However, there has been a horizontal
division of labour, with specialization in different research areas and, of
course, organizational roles have tended to be filled by professionally senior
practitioners. It is likely that the taking up of such roles by the frrst generation
establishment was the outcome of two factors: on the one hand, the clear
opportunities offered by these roles and their availability in the early days
of the emergence of AI in Britain; and on the other hand, the difficulty of
getting to the research front without extensive specialist training, particularly
in the activity of programming, at that time very much an esoteric art. And of

202

James Fleck

course, such training required some sort of organizational structure for

efficient transmission. Consequently, the organizational roles were more
important and probably demanded more effort during the initial emergence
of A! in Britain, before they became institutionalized and suitable recruits
were readily available from among the ranks of the trained A! specialists.
With the development of the institutional structure during the 1960s and
19708, there has emerged a similar tight pattern of intergenerational and
intercentre linkages in Britain as that evident in the United States: indeed,
the American and British structures were connected, as is illustrated by
Figure 2.
This tight intergenerational pattern has been reinforced by two related
characteristics of the AI paradigmatic structure: the constructive and the
craft nature of work in the area. The writing of a computer program to carry
out some task clearly involves making or constructing something, rather than
investigating something that is naturally given (although such investigations
can be and are involved). Not only are there many ways of constructing a
program on AI principles, but there are also many other non-A! ways; for
instance the construction of a stochastic model. Furthermore, it is possible to
construct models not based on the use of the digital computer, for instance,
in the building of electronic analogues of neuron networks. Hence it is clear
that the A! paradigmatic structure delineates only one broad way of making
models among many possible; this arises from the constructive nature of the
activity in the area. Central to the paradigmatic structure is the activity of
programming, based on the use of list processing languages, associated with
which are many characteristic techniques and "tricks of the trade". Effective
programming also involves a high degree of skill, and all these features together
lend the AI approach a distinctly 'craft' nature, which requires for its transmission a lengthy apprenticeship and some degree of contact and interaction
with experienced practitioners. This craft and constructive nature of work in
the area ensures that it is extremely unlikely that the specifically A! approach,
even in broad terms, would be developed spontaneously and independently
outside the community of people already using it. Consequently, these
aspects of the paradigmatic structure place constraints on the development
and transmission of the AI approach, and thus reinforce the tight intergenerational pattern, which is also encouraged by the purely social aspects of
communication and the favouring of those already known in the network.

203

Development and Establishment in Artificial Intelligence

ABERDEEN
EDINBURGH

AMERICA

SUSSEX

ELSEWHERE

ESSEX

64

!!L..
!!L..

Z!L

=
l-

llIi-

ZL.

I-

Tit

=
:55

1L.
BO

Fig. 2. Movement of research personnel between Artificial Intelligence Centres in the

United Kingdom: 1964-1980. Each line shows the trajectory of an
represents the start of a career in Artificial Intelligence. The density of horizontal
lines gives some indication of the frequency of movement between centres. The figure
represents trajectories for SO out of a total of about ISO personnel involved in Artificial
Intelligence research in Britain during 1964-1980; a further 43 did not move, and 15
left the field altogether. Information was not available in the remaining 40-50 cases,
and is less complete after 1978, while short term visits of six months or less (which
were quite frequent) have not been included. Thus the above is probably an understatement of movement between centres. The following features are apparent: strong
links with the United States; movement away from Edinburgh in the mid 1970s;
increased movement in the early to mid 1970s involving the newly emerged centres at
Sussex and Essex; increasing links with centres elseWhere in the 1970s; and finally,
the importance of Edinburgh, and latterly Essex, as sources of Artificial Intelligence
personnel.

204

James Fleck

A major feature of development in Britain was that the initial AI establishment did not emerge from the strong pre-existing cybernetic or computer
science network, despite the clear prefigurement of AI research in the work
of Turing and Craik. Rather, it came from people external to such work:
Elcock at Aberdeen; Michie, Meltzer, and Longuet-Higgins at Edinburgh; and
later, Boden and Sloman at Sussex. Even Sutherland and Clowes at Sussex,
although they had links with the cybernetic tradition, certainly were not
centrally involved with it. Such an entry into the AI establishment from
outside the cybernetics tradition had also been evident in the United States,
with Simon and Newell. But, just as was the case in the United States, the
first generation members of the British AI establishment, although they were
marginal to cybernetics and computer science, certainly did not arrive from
nowhere. Michie, for instance, had some sixty publications to his credit in
his previous specialist areas of genetics, immunology, and reproduction, and
held the post of reader in the Department of Surgical Science at Edinburgh.
He was recommended by W. H. Waddington and M. Swann on his appointment to Edinburgh in 1958, and Swann continued to support him during
his AI activities. Meltzer, too, as a reader in the Department of Electrical
Engineering, had a solid reputation for his work on electron beam dynamics
(used by NASA for the design of ion propulsion for space vehicles) and
solid-state electronics; and Longuet-Higgins, with his international standing
in theoretical chemistry was clearly aheady a member of the wider British
scientific establishment.
The migration of outsiders with some standing, and hence the freedom to
move fields (127), seems to have had, therefore, as its major consequence
the construction of an organizational structure within which the pursuit
of AI research subsequently developed. In some cases (especially that of
Sutherland) the organizational leadership role merged with another, probably
necessary, role in the emergence of a new interdisciplinary area - namely a
sponsorship role. The institutional developments at Edinburgh would not
have been possible without some strong support from sponsors placed in
fairly influential positions with the university government, especially in view
of the tendency for the publicity attracting, research intensive, AI activities
to arouse suspicion and resentment: both Sir Edward Appleton and Sir
Michael Swann, who succeeded Appleton as vice chancellor of Edinburgh
University, were active in encouraging and supporting those developments.

Development and Establishment in Artificial Intelligence

205

And at Sussex there was also fairly widespread support for AI oriented
developments among many of those in positions of influence.
As well as this positive sponsorship within the wider establishment, which
was accompanied by a positive evaluation of the status of AI, there was a
negative sponsorship as well, as demonstrated by Sir James Ughthill's report.
Ughthill's opposition could clearly be seen to support the status quo: He
affirmed the value of the currently existing areas included in Advanced
Automation (e.g., control engineering) and in Central Nervous System research (e.g., neurophysiology), and moreover attributed any success in AI to
contributions arising from these areas. The validity of the emerging interdisciplinary area of AI was thus challenged and denied, explicitly in terms of the
aheady-established disciplines surrounding AI. In particular, Ughthill's focus
on the established Central Nervous Systems areas of research - neurophysiology and neurochemistry - and his use of the term 'central nervous system'
happened to align with the dominance in Europe of the neurosciences which
study the 'hardware' of the brain, over the cognitive sciences - linguistics
and psychology - which might be said to study the 'software' of the brain
- a dominance which is not as clear-cut in the United States (I28). In this
dominance relation we see a prestige hierarchy, with those sciences closest to
the physical sciences accorded most prestige. In this context, the study of the
brain at the reductive level of biochemical or_neurophysiological mechanisms
is considered more prestigious than the AI approach at the information
processing levels.
Moreover, it is difficult to account for the impact of the Ughthill report
(by his own admission a two months layman's view of the area) (129) except
in terms of the authority carried by Ughthill's eminence. It is interesting to
note that reference is still offered to observations that Ughthill made on the
difficulties for search arising out of the combinatorial explosion (I 30), as if
he were their originator, whereas, in fact, the combinatorial problems had
been regarded as the raison d'etre for AI - the huge size of the space of
possible moves in chess, for example, estimated by Shannon in 1950 as in
excess of 10 120 , dictated the need for heuristic strategies to restrict the
search space to manageable proportions (131).
But what is also of interest about the Ughthill report, apart from its
importance as an authoritative pronouncement on the status of AI research,
is that it is one among a multitude of attacks on the field (132). Such attacks

206

James Fleck

on AI have been commonplace, and while they purport to deal with the
particularities of the subject matter of research in the area, it is quite clear on
closer inspection, that they are more concerned with the general goal of
constructing an intelligent machine. It would be too lengthy to argue this
fully here, but it is perhaps pointed by the continued relevance of Turing's
comments on the arguments for and against the possibilities of constructing
intelligent machines (133), despite the fact that the distinctive AI approach
had not emerged when Turing was still alive. It is also pointed by the fact that
these attacks on AI are not the prerogative of any particular group: criticism
has come from all shades of political opinion, and from all areas of research;
scholarly as well as technical.
Ironically, this variety in attacks can only be matched by the diversity
in the sources of support for AI, or the range of, often conflicting, views
within the field itself. Something of this has already become evident in the
differences between LonguetHiggins and Michie, one favouring the cognitive
science defmition of AI, and the other the Machine Intelligence approach.
There are other divisions: those supporting a theoretical formal approach,
such as McCarthy, for instance, and those supporting the exploitation of
practical applications, such as Feigenbaum; those who see no problems with
accepting military funding (McCarthy and Feigenbaum) and those implacably
opposed (Meltzer, while Michie was opposed to classified work).
These divisions, which are legion in the area, are coupled with a rather
amazing state of substantive partisanship, or scientific ethnocentricity, in
which proponents of the various different research areas each tend to see
their own approach as the real AI approach - to the theorem provers, theorem
proving is central, for the natural language proponents, language is the basis
for reason, and so on (134).
Many of these differences can be related to the background competences
of the practitioners, and can be interpreted as competition between groups on
the research area level for resources and authority within AI as a specialty,
constituting perhaps the primary locus for competition over cognitive com
mitments, quite distinct from the individualistic level of competition for
recognition, long identified in the sociology of science as a basic motor of
scientific development. An important point to note is that such differences
are not precluded by the paradigmatic structure of AI, outlined earlier
as providing guidelines for the common computational approach and its

Development and Establishment in ArtijiciDl Intelligence

207

programming basis, but which does not dictate a dogmatic monolithic atti
tude, nor inculcate a unifying solidarity. Moreover, this variety of views in
and around AI can be related to its position as an interdisciplinary area, with
particular research areas associated with particular neighbouring disciplines for example, the natural language research area is associated with linguistics,
while theorem proving has links with metamathematics. And as an interdisciplinary area, the status of AI research is still very much in process of
negotiation. The cognitive science developments appear to have led to an
acceptance, on the part of those involved, of the validity of AI: indeed the
impression in that context is that AI is the 'hard' formal core, and therefore
of high status. But in a computer science or general scientific context, AI
is still seen very much as a 'freaky', rather dubious fringe activity, and con
sequently of rather inferior status (135). Moreover, there are two broad
categories of attacks which can be related to these contexts. On the one hand
there are attacks, often by philosophers and others, on AI for being reductionistic and impossible (136) - in a sense it is 'harder' than is appropriate
for the study of intelligent activity . On the other hand there are criticisms,
often by computer scientists, of AI for being morally wrong (137), bad
science (138), or undisciplined and sloppy (139).
Finally, perhaps the variety and depth of feeling of the many attacks on
AI derive, not so much from what is, in fact, done in AI research, but rather
from the fact that the very broad aim of research in the area, namely the
construction of intelligent machines, bears uncomfortably on our conception
of ourselves. In Elias' terms, AI research is seen to be involved with a very
sensitive area of the means of orientation: the area which is concerned
with the nature of mind. Furthermore, during several hundred years of
development and struggle with other competing establishments, a scientific
establishment has yet to succeed in gaining a monopoly over the means of
orientation in this area. In making its challenge in this area, therefore, AI is
inviting violent attacks, and its practitioners should hardly be surprised when
they suffer them.

8. Conclusion: The Process of Development and Establishment in Artificial

Intelligence
The process of development and establishment in AI in broad terms, therefore,

208

James Fleck

appears to have been as follows. Around the period of the Second World War,
catalyzed by the war-time weakening of traditional disciplinary boundaries,
and brought into being by exigencies deriving from the unprecedented problems of organization and communication posed by the increasingly complex
social structures and conditions, there emerged the software sciences. These
had their focus on pattern rather than substance, and included operations
research, computer science, and cybernetics. In particular, the cybernetics
area, with its focus on the processes common to animals and machines,
promised a realization of the age-old desire to make an artificial man, a
machine that could think. Within the general area of cybernetics various
approaches were made to the construction of intelligent machines, some
based on electronic analogue of neuron networks, others on the simulation
of processes by means of the newly developed digital computer. In this
context, the paradigmatic structure of AI was articulated in the United States
during the late 1950s.
The American establishment in AI consisted essentially of those who had
contributed to this articulation, and who had provided the institutional
structure within which subsequent research based on the distinctive AI
paradigmatic structure, could be undertaken. The source of authority and
reputation of the establishment was derived from the effective demand which
developed for this distinctive approach, on the part of those who wanted to
follow the approach themselves and those who thought the approach was
worthwhile and promising. In addition, the emerging establishment secured
the backing of the funding agencies, aided by their good connections with the
wider establishment, in competition with other approaches within computer
science. The preference of the funding agencies to concentrate resources in
a few centres, coupled with the expense of the instrumental base required
(the digital computer), ensured that an effective monopoly over material
resources as well as the cognitive ones was maintained.
As the paradigmatic structure was already elaborated before it was exported
to the United Kingdom in the early to mid-1960s, it left less opportunity
for substantive contributions on this general level while there was ample
scope for extending the institutional facilities for carrying on AI research
in Britain: consequently the first-generation British establishment in AI
consisted of those who were able to set up organizational forms to exploit
paradigmatic structure, thus giving them a monopoly
the already

Development and Establishment in Artificial Intelligence

209

over the cognitive resources in the area. Because of the constructive nature
of AI, which was manifested in it being only one among several competing
cybernetic approaches, and since people already committed to a particular
approach tend to stay with that approach, members of the first-generation
establishment included people from outside the cybernetic and even computer science traditions. This was the case in the United States, but was
more marked in Britain. Moreover, the members of the first-generation
establishment, in fact, were drawn from those who already had some standing
or prestigious backing in another field, and being thus well connected, they
were able to secure the backing of the Science Research Council in competition with other approaches within computer science, ensuring their effective
monopoly over material as well as cognitive resources in the area.
This monopoly over resources was reinforced by the constructive and craft
nature of the AI paradigmatic structure, and the patterns of development
in both the United States and the United Kingdom followed the lines of
personnel mobility and contact, thus leading to a tight intergenerational and
intercentre structure of linkages. In particular, students and 'descendents'
of the members of the first-generation American establishment have tended
to dominate the field. The paradigmatic structure of AI, however, while
providing general guidelines for the methodological approach employed in
AI research, does not dictate the direction of research. Due to the very wide
ranging nature of the focus on intelligent activity, research in AI has become
involved with the subject matter of many other disciplines, and has therefore
developed as an interdisciplinary area. This interdisciplinary character of
AI has induced many mutually competing divisions within the area, as well
as leading to many external views on the status of the field, which has consequently been very much a matter for negotiation: from the point of view of
the 'soft' sciences, such as linguistics and psychology, AI has appeared 'hard'
and therefore of superior status; from the point of view of the 'hard' sciences,
such as computer science and physics, AI has appeared somewhat 'freaky'
and therefore of inferior status. Furthermore, as a newly emerging specialty,
AI has been in competition with already established disciplines: the Ughthill
report, critical of AI, can be interpreted in this context as an affirmation of
the status quo. Finally, because of the general aim of constructing intelligent
mechanisms, AI has been seen as challenging the monopoly on the means
of orientation with respect to the nature of mind. As this bears directly on

210

James Fleck

peoples' conceptions of themselves, deep feelings have been aroused, as is

evident in the many and varied attacks on AI.
It is therefore evident that there has been competition on a variety of
levels over AI, with differing consequences for the development of the field,
and, moreover, engaging distinct groups or establishments. At the most
circumscribed level, within the field, there has been competition over the
choice of techniques to be used in a particular research area, and competition
between research areas themselves. This is of immediate consequence for
research at the practical level, and involves groups negotiating for ascendency
within the AI establishment. As a less circumscribed level, the establishment
of AI as a whole has been involved in competition with neighbouring scientific
and scholarly establishments. This concerns the general scientific validity of
the field rather than being of immediate practical import, and has consequences in broad terms for the status of the field, the availability of funding)
and the continued demand for the AI approach on the part of potential
entrants. Finally at the most general level, and most clearly underlining Elias'
remarks about the means of orientation, the validity of the AI approach is
discussed in terms involving political, religiOUS, moral, philosophical, and
cultural issues. The question being asked here is not whether AI is valid as a
scientific approach, but rather 'whether any such approach to the mind is
viable. This engages a far wider group, and the prevalence of the debate at this
level is perhaps pointed by the fact that, at most, a couple of dozen full-time
occupational opportunities are available in AI - very few compared to many
other areas of endeavour - while nearly everyone, it seems, has something to
say on the issue of whether machines can think.

Acknowledgements
I should like to acknowledge the help given by the interviewees to whom a
draft of this paper was circulated. The comments of J. A. M. Howe, H. C.
Longuet-Higgins, D. Michie and N. S. Sutherland were particularly valuable.
I should also like to thank R. D. Whitley for his suggestions and the SRC for
their support.

Development and Establishment in Artificial Intelligence

211

Notes and References

1. The term specialty is used here in the sense of a community of practitioners
differentially located with respect to a common paradigmatic structure which
defines a general focus of attention. Compare R. Whitley, 'Components of Scientific Activities, their Characteristics and Institutionalisation in Specialities and
Research Areas', in K. Knorr, H. Strasser, and H. G. Zilian (eds.), Determinants
and Controls of Scientific Development, Dordrecht: Reidel, 1975, pp. 37-73.
2. See, for example, I. Aleksander, 'Artificial Intelligence', Electronics and Power
22 (1976) 242-44.
3. This paper is based on a detailed MSc/PhD study of the area: J. Fleck, The
Structure and Development of Artificial Intelligence: A Case Study in the Sociol
ogy of Science, unpub. MSc diss. University of Manchester, 1978. Also PhD
diss. in preparation.
4. Compare R. Whitley,op. cit., 1975, Note 1.
5. Sandewall has discussed these aspects, for example, E. Sandewall, 'Programming
in an Interactive Environment: The "LISP" Experience', ACM Computing Surveys
10 (1978) 35-71.
6. Compare J. Law, 'The Development of Specialties in Science: The Case of X-ray
Protein Crystallography', Science Studies 3 (1973) 275-303.
7. This is clear from a consideration of the contents of the book, a classic in the
area, E. A. Feigenbaum and 1. Feldman (eds.), Computers and Thought, New
York: McGraw Hill, 1963.
8. Compare D. E. Chubin and T. Connolly, 'Research Trails and Science Policies:
The Shaping of Scientific Work by Hierarchies and Elites', in this volume.
9. D. O. Edge and M. 1. Mulkay,Astronomy Transformed: The Emergence of Radio
Astronomy in Britain, New York: Wiley-Interscience, 1976.
10. Though for a dissenting view, see B. R. Martin, 'Radio Astronomy Revisited: A
Reassessment of the Role of Competition and Conflict in the Development of
Radio Astronomy', Sociological Review 26 (1978) 27-55.
11. See E. Yoxen, 'Giving life a new meaning: the rise of the molecular biology
establishment' in this volume.
12. N. Elias, 'Scientific Establishments', theme paper for Conference on Scientific
Establishments and Hierarchies, Oxford, July 1980.
13. N. Wiener, Cybernetics - Control and Communication in the Animal and Machine,
New York: Wiley, 1948.
14. For instance, W. Pitts and W. S. McCulloch, 'How We Know Universals', BUll.
Maths. Biophysics 9 (1947) 127-47, or O. G. Selfridge, 'Pandemonium: A
Paradigm for Learning', in Mechanisation of Thought Processes, London: HMSO,
1960, pp. 513-26.
15. Machine Translation was an early example of this approach. Based on syntactical
analysis and dictionary lookup, it failed at that time in its aim of providing high
quality translation - see Y. Bar-Hillel, 'The Present Status of Automatic Translation of Languages', in F. L. Alt (ed.),Advances in Computers, Vol. 1, New York:
Academic Press, 1969, pp. 92-163.
16. C. Shannon and J. McCarthy (eds.), 'Automata Studies', Annals of Mathematics
Studii; No. 34, Princeton N. J.: Princeton University Press, 1956.

212

James Fleck

17. Comments by J. McCarthy, at an 'AISB Summer School on Expert Systems', held

at Edinburgh University, July 1979.
18. The Dartmouth Conference is mentioned in several places, for example, M. 1.
Minsky, 'Artificial Intelligence', in Information, (a Scientific American book), San
Francisco: Freeman, 1966, p. 194; and in P. McCorduck, Machines Who Think,
San Francisco: Freeman, 1979, pp. 93-114.
19. A. Newell and H. A. Simon, 'The Logic Theory Machine', IRE Trans. on Info.
Theory IT-2 (1956), 61-79.
20. G. W. Ernst and A. Newell, GPS: A Case Study in Generality and Problem Solving,
New York: Academic Press, 1969.
21. J. McCarthy, 'Recursive Functions of Symbolic Expressions and Their Computation by Machine', Part 1, Comm. ACM 3 (1960) 184-95.
22. M. 1. Minsky, 'Some Methods of Artificial Intelligence and Heuristic Programming',Mechanisation of Thought Processes, London: HMSO, 1969, pp. 3-28.
23. In addition, the 'Matthew Effect' would be operating: R. K. Merton, 'The
Matthew Effect in Science', Science 159 (Jan. 1968).
24. J. McCarthy, 'Toward a Mathematical Theory of Computation', Proc. IFIP
Congress 1962, Amsterdam: North Holland, 1963.
25. M. 1. Minsky (ed.), Semantic Information Processing, Cambridge Mass.: MIT
Press, 1968.
26. M. 1. Minsky, 'A Framework for Representing Knowledge', in P. Winston (ed.),
The Psychology of Computer Vision, New York: McGraw Hill, 1975, pp. 21177.
27. See, for example, M. 1. Minsky and S. Papert (and staff), 'Proposal to ARPA for
Research on Artificial Intelligence at MIT 1971-1972', MIT AI Lab. Memo.
No. 245. Oct. 1971.
28. A. Newell and H. A. Simon, Human Problem Solving, Englewood Cliffs, N. J.:
Prentice Hall, 1972.
29. W. S. McCulloch and W. Pitts, 'A Logical Calculus of the Ideas Immanent in
Nervous Activity', Bull. Maths. Biophysics 5 (1943) 115-33.
30. Indeed, Simon received the 1978 Nobel Prize in Economics, and his 1947 book:
Administrative Behaviour, New York: Macmillan, was explicitly mentioned in the
prize announcement.
31. As noted in McCorduck, op. cit., 1979 (Note 18), 109 ff.
32. J. M. Brady, 'Report on a Visit to the U.S.', Essex University, 1975 (edited
version of a report submitted to the Science Research Council, Sept. 1975),
outlines the funding situation for Artificial Intelligence in the United States.
McCorduck,op. cit., 1979 (Note. 18), 117 ff., discusses the 'no strings attached'
role of the Air Force funding in allowing the work of Simon and Newell to get
started.
33. For attacks on the field see: M. Taube, Computers and Common Sense: The
Myth of Thinking Machines, New York: Columbia University Press, 1961; H. 1.
Dreyfus, What Computers Can't Do: The Limits of Artificial Intelligence, New
York: Harper and Row, 1972; and J. Weizenbaum, Computer Power and Human
Reason, San Francisco: Freeman, 1976.
34. N. Elias, op. cit., 1980, Note 12.
35. P. Armer, 'Attitudes Toward Intelligence Machines', in Feigenbaum and Feldman,

Development and Establishment in Artificial Intelligence

36.

37.
38.
39.
40.
41.
42.
43.
44.

45.
46.
47.
48.
49.

so.
51.

52.
53.
54.
55.

213

op. cit. (note 7), pp. 389-405. This was also commented upon in interviews: R. A.
Brooker, Essex University, 28/3/79; and A. M. Uttley, Sussex University, 21/3/79.
A. M. Turing, 'Intelligent Machinery', (1947), in B. Meltzer and D, Michie (eds.),
Machine Intelligence S, Edinburgh: Edinburgh University Press, 1969, pp. 3-23
and A. M. Turing, 'Computing Machinery and Intelligence', Mind 59 (1950)
433-60.
R. J. W. Craik, The Nature of Explanation, Cambridge: Cambridge University
Press, 1952, p. 57.
W. R. Ashby, Design for a Brain, New York: Wiley, 1952; and An Introduction
to Cybernetics, London: Methuen, 1965.
W. G. Walter, 'An Imitation of Life', Scientific American 182, No.5 (1950)
42-5; 'A Machine that Learns', Scientific American 185, No.2 (1951) 60-3.
F. H. George, The Brain as a Computer, Oxford: Pergamon Press, 1961.
The Ratio club was discussed in an interview: A. M. Uttley, Sussex University,
21/3/79; and is also discussed by McCorduck, op. cit. (Note 18), p. 59.
The proceedings of this conference are published in: Mechanisation of Thought
Processes, London: HMSO, 1960.
J. McCarthy, 'Programs with Common Sense', ibid., 75-84.
D. Michie, On Machine Intelligence, Edinburgh: Edinburgh University Press,
1974, pp. 37,51, and 66-7; where he states: "It was from my personal association with Turing during the war and early post-war years that I acquired my
interest in the possibilities of using digital computers to simulate some of the
higher mental functions that we call 'thinking'."
This, and much of the following information pertaining to Michie's activities is
derived from interviews with Michie (Edinburgh University, 29 and 31/8/78),
backed up by other available sources.
For example, D. Michie, 'The Effect of Computers on the Character of Science',
University of Edinburgh Gazette 34 (Oct. 1962) 23-8; and 'The Computer
Revolution: Where Britain Lags Behind', The Scotsman, 12/7/63.
Computing Science in 1964, A Pilot Study of the State of University Based
Research in the U.K., Prepared for the Research Grants Committee by Dr Donald
Michie, London: The Science Research Council, 1965.
Interview: Lord Halsbury, London, 18/12/79.
Council for Scientific Policy/University Grants Committee, A Report of a Joint
Working Group on Computers for Research, London: HMSO, Cmnd. 2883, 1966.
Interview: Lord Halsbury, London, 18/12/79.
Reconstructed from information in the archives of the Department of Artificial
Intelligence, and the Machine Intelligence Research Unit, Edinburgh University,
and from interviews.
J. McCarthy, 'Review of the Lighthill Report', Artificial Intelligence S (1974)
317-22.
R. M. Burstall and R. J. Popplestone, 'POP-2 Reference Manual', in E. Dale and
D. Michie (eds.) , Machine Intelligence, Vol. 2, Edinburgh: Oliver and Boyd, 1968,
207-46.
Interviews: B. Meltzer, Edinburgh University, 1/8/77 and 23/1/79.
N. L. Collins and D. Michie (eds.), Machine Intelligence, Vol. 1, Edinburgh: Oliver
and Boyd, 1967. There have been nine workshops in the series to date.

214

James Fleck

56. E. W. Elcock, 'Report of the SRC Computer Research Group', Aberdeen University, May 1970; and interviews: A. M. Murray, Aberdeen University, 9/11/78;
P. M. D. Gray, Aberdeen University, 23/7/79; and J. M. Foster, R. R. E. Malvern,
19/6/79.
57. Information on the activities of Clowes derives from a letter: M. B. Clowes,
23/11/78; and interview: M. B. Clowes, Sussex University, 22/3/79.
58. Interview: B. Meltzer, Edinburgh University, 23/1/79.
59. Interview: P. J. Hayes, Essex University, 27/3/79.
60. These conferences, The International Joint Conferences on Artificial Intelligence,
have been one of the main organs of communication in the area, and a major
outlet for publications in the area.
61. R. L. Gregory, Eye and Brain, London: Weidenfeld and Nicholson, 1966.
62. Information on the robot is derived from numerous documentary sources in the
archives of the Department of Artificial Intelligence, and the Machine Intelligence
Research Unit, and from various interviews.
63. Interview: J. A. M. Howe, Edinburgh University, 24/1/79.
64. As was evident from various interviews, including one with S. Michaelson, the
director of the Computer Science Department, at Edinburgh University, on
18/7/79.
65. Lord Swann commented on the new departures in an interview: London, 10/10/
79. At a press conference in connection with an open day for industrialists, he
is reported as commenting that Edinburgh was among the top two centres of
research, despite having no costly 'big science' projects. University of Edinburgh
Bulletin 8, No.2 (Oct. 1971).
66. Interview: R. L. Gregory, Bristol University, 25/7 /79.
67. Conversational Software Ltd., launched in 1970.
68. Interview: D. Michie, Edinburgh University,.31/8/78. He commented that at one
stage some twenty sources were involved.
69. Computing Science Review, London: Science Research Council, 1972, p. 17.
70. Ibid., p. 19. Compare also D. Michie, 'Schools of Thought about AI', University
of Edinburgh, School of Artificial Intelligence, Experimental Programming
Report No. 32,1973.
71. Artificial Intelligence: A Paper Symposium, London: Science Research Council,
1973,p.i.
72. This review was carried out by a Special Committee, chaired by Prof. N. Feather,
and set up by the University Court. It reported in 1973, and extracts were made
available to me by C. H. Stewart, at that time secretary to the University, in a
letter,18/7/78.
B. As was evident from interviews and also from the Departmental Newsletter.
74. J. McCarthy and P. J. Hayes, 'Some Philosophical Problems from the Standpoint
of Artificial Intelligence', in B. Meltzer and D. Michie (eds.), Machine Intelligence,
Vol. 4, Edinburgh: Edinburgh University Pre8s,1969.
75. G. D. Plotkin, Automatic Methods of Inductive Inference, unpub. PhD diss.,
University of Edinburgh, 1971.
76. Interview: R. A. Kowalski, Imperial College, London, 15/6/79.
77 . From the Departmental Newsletter.
78. From the Minutes of the Round Table, a sort of departmental board.

Development and Establishment in Artificial Intelligence

215

79. Interview: Lord Swann, London, 10/10/79.

80. Various interviews.
81. D. Michie, A Six Year Project To Develop an Intelligent Problem-Solving System
(modified version of a seven-year project proposal to the SRC), Edinburgh University, 1972.
82. P. A. Ambler et al., 'A Versatile Computer Controlled Assembly System', Proc.
Third International Joint Conference on Artificial Intelligence, Stanford: 1973,
pp. 298-307; and Artificial Intelligence 6 (1975) 129-56.
83. Sir James Lighthill, 'Artificial Intelligence: A General Survey', in SRC, op. cit.
(Note 71), p. 7.
84. As is evident in theAISB Newsletter, SIGART (The American Equivalent), and in
numerous other places.
85. In the public press there were articles in The Times Higher Education Supplement,
1/6/73 and 8/6/73,in Science 180 (June 1973) 1352-3,and in the New Scientist,
22/2/73 and 1/3/13. There was a public debate on July 4th., 1973, at the Royal
Institution, Albemarle Street, London, under the chairmanship of Sir George
Porter, between Sir James Lighthill and Professors R. L. Gregory, J. McCarthy,
and D. Michie on Sir James' theme: 'The General Purpose Robot is a Mirage'.
This was televised by the BBC as one of its 'Controversy' series and broadcast
on August 30th., 1973. Later videorecordings of the debate were taken to the
United States as the 'Lighthill Tapes', where they were shown to packed audiences
around the centres of Artificial Intelligence.
86. Lighthill,op. cit. (Note 83), p. 1.
87. In Proposed New Initiatives in Computing and Computer Applications, Swindon:
Science Research Council, March 1979, p. 8, there is the comment: ''The Panel
has no doubt that the reluctance of the present community to take up the challenge (of industrial robots research) is due at least in part to the general discouragement of Artificial Intelligence which took place in this country several years
ago and that it is now up to SRC to take steps to remedy the situation."
88. Reported in AISB 20 (July 1975), and in The Times Higher Education Supplement, 14/3/75, p. 9.
89. It is impossible to give defmite figuxes in brief for Artificial Intelligence funding
because of the multiplicity of sources, and because of the difficulty in deciding
what money went to specifically Artificial Intelligence research of the type
discussed in this paper; the SRC, for instance have sometimes included under
the category of !llachine intelligence work that is clearly along cybernetic
lines.
90. Sources detailed in Note 56.
91. Interview: R. A. Brooker, Essex University, 27/3/19.
92. Interview: D. Michie, Edinburgh University, 31/8/78.
93. D. Michie, 'Machine Intelligence in the Cycle Shed',New Scientist 57 (Feb. 22nd.,
1973) 422-3.
94. Interview: Lord Swann, London, 10/10/19.
95. E. W. Elcock and D. Michie (eds.), Machine Intelligence, Vol. 8, New York: Wiley,
1977.
96. J. E. Hayes, D. Michie, and L. l. Mikulich (eds.), Machine Intelligence, Vol. 9,
Ellis Horwood, Chichester, 1979.

216

James Fleck

97. Michie organized the 'AISB Summer School on Expert Systems', held at Edinburgh University, July 1979.
98. A. Bundy et al., Artificial Intelligence: An Introductory Course, Edinburgh:
Edinburgh University Press, 1978.
99. T. Winograd, Understanding Natural Language, Edinburgh: Edinburgh University
Press, 1972.
100. Various interviews, and Note 27.
101. J. A. Robinson, 'A Machine Oriented Logic Based on the Resolution Principle',
Journal ACM 12 (196S) 23-41.
102. J. Fleck, 1978, op. cit. (Note 3), pp. 44-8.
103. D. Warren, 'PROLOG on the DEC system-l 0', paper presented at the AISB
Summer School on Expert Systems, Edinburgh University, July 1979, notes
some applications. The position of theorem proving was also discussed in several
interviews.
104. Interview: R. M. Burstall, Edinburgh University, 10/8/77.
lOS. Interview: R. A. Brooker, Essex University, 27/3/79.
106. Interviews: J. M. Brady, ES'sex University, 26 & 28/3/79; P. J. Hayes, Essex
University, 27/3/79.
107. J. E. Doran, 'Knowledge Representation for Archaeological Inference', in Elcock
and Michie, op. cit. (Note 9S), pp. 433-S4.
108. Interviews: B. Anderson, Essex University, 30/7/79; Y. Wilks, Essex University,
8/8/79.
109. Information on movements through Essex was derived from interviews with
Brady,op. cit., Note 106, backed up by other sources.
110. Information on developments at Sussex University was derived from various
interviews and documentary sources, in particular the interview: N. S. Sutherland,
Sussex University, 21/9/79.
111. For example, A. M. Uttley, 'Simulation Studies of Learning in an Informon
Network', Brain Research 102 (1976) 37-S3.
112. M. B. Clowes, 'On Seeing Things', Artificial Intelligence 2 (1971) 79-116.
113. This organization is frequently commented on in the Annual Reports, Sussex
University, and is discussed by the first Vice Chancellor, Asa Briggs, in his 'Drawing
a New Map of Learning', in D. Daiches (ed.), The Idea of a New University: An
Experiment in Sussex, London: Deutsch, 1964, pp. 60-80.
114. Letter: Lord Briggs, 28/11/1979.
l1S. 'Working Party on School of Cognitive Studies', University of Sussex, June 1970.
116. Interview: M. A. Boden, Sussex University, 19/9/79.
117. M. A. Boden, Artificial Intelligence and Natural Man, Hassocks: Harvester Press,
1977.
118. Interview: A. Sloman, Sussex University, 20/3/79.
119. A. Sloman, The Computer Revolution in Philosophy, Hassocks: Harvester Press,
1978.
120. Interview: P. N. Johnson-Laird, Sussex University, 20/9/79.
121. For instance, approximately one quarter of the MSc computer science courses
outlined in Graduate Studies 1974-75, Cambridge: CRAC. 1974, mentioned
Artificial Intelligence, and it is often mentioned as an acceptable interest in job
adverts.

Development and Establishment in Artificial Intelligence

217

122. J. M. Brady, 'A Glimpse of the Future of AI', text ofa lecture delivered at 'Computing 79', Syndey, Australia, August 1979, p. 2.
123. B. C. Griffith and N. C. Mullins, 'Coherent Social Groups in Scientific Change',
Science 177 (1972) 959-64.
124. I do not want to argue a 'great man' view of history here. The situation can be
interpreted in terms of the emergence of socially defined possibilities which in the
event were exploited by Michie. If Michie had not been, development in Artificial
Intelligence in Britain would still have taken place with other people stepping in
to a greater extent. Perhaps what is required is a 'great opportunities' view of
history, with the emphasis on the socially given possibilities rather than on the
people who exploit them.
125. T. Shinn, 'Scientific Disciplines and Organizational Specificity: The Social and
Cognitive Configuration of Laboratory Activities', in this volume.
126. See for instance, P. Kraft, Programmers and Managers: The Routinization of
Computer Programming in the United States, New York: Springer Verlag, 1977.
127. M. J. Mulkay, in The Social Process of Innovation, London: MacMillan, 1972,
concludes: " ... intellectual migration, whether into established networks or
into virgin areas, will normally be led by mature researchers of known repute.
These men use their eminence to attract funds and graduate students; and in
various ways they try to use their existing knowledge and techniques as a point of
departure for the construction of the new intellectual framework" (p. 54.). The
pattern of development of Artificial Intelligence in Britain certainly conforms
with the first part of what Mulkay writes, but it is difficult to square it with the
second part. It would appear that the extent of articulation of the paradigmatic
structure in the United States preempted attempts by migrants in Britain to use
their existing knowledge in the construction of new intellectual frameworks, and
left them scope only for negotiation over the organizational aspects.
128. See N. S. Sutherland, 'Neuroscience Versus Cognitive Science', Trends in Neurosciences 2, No. 8 (1979) i-ii, which discusses some of the particulars of this
dominance.
129. Lighthill,op. cit. (Note 83), p. 1.
130. For example, Sir Geoffrey Allen (Chairman of the Science Research Council), in a
talk in the Department of Liberal Studies in Science, the University of Manchester,
17/5/79.
131. C. E. Shannon, 'Automatic Chess Player', Scientific American 182, No.2 (1950)
48-51.
132. See Note 33.
133. Turing,op. cit., Note 36.
134. Fleck,op. cit. (Note 3), pp. 138-9.
135. Such an impression was given by those in Artificial Intelligence (for example,
interview: A. Bundy, Edinburgh University, 17/7/79) and those outside, in the
course of interviews.
136. Dreyfus,op. cit., Note 33, is the classic example here.
137. Weizenbaum,op. cit., Note 33.
138. Taube, op. cit., Note 33.
139. E. W. Dijkstra, 'Programming: From Craft to Scientific Discipline', Proc. International Computing Symposium, 1977, Liege, Belgium, April 1977 , pp. 23-30.

THE DEVELOPMENT OF RESTRICTEDNESS IN THE

SCIENCES

ARIE RIP
University of Leiden

1. Introduction
The technical side of science and its contribution to the dynamics of scientific
developments are often neglected. As Derek de Solla Price complains:
It is unfortunate that so many historians of science and virtually all of the philosophers
of science are born again theoreticians instead of bench scientists (1).

Theory is considered to be the distinguishing characteristic of science, and the

source as well as the legitimation of the force of scientific knowledge. Again,
some voices are raised agaiitst this received view:
.
Much of the potency of science must be related to the wide range of applicability of its
standard procedures, and not to the possible deductions which can be made from its
general laws. The role of deduction in science is grossly over-emphasized; natural applications of techniques are insistently perceived as logical or deductive moves (2).

In a more balanced view, theory as well as techniques should get their due,
as in Bernal's aphorism: "Science, in one aspect, is ordered technique; in
another, it is rationalized mythology" (3). The prevailing neglect of the
technical side of science then becomes something to be explained, one of the
reasons obviously being the higher status of mental labour compared with
manual labour in our culture.
Philosophers of science, apart from their professional interest in theory
and speculation, may be excused for their neglect of technical aspects because
the discourse of science, especially in its publications, is much more accessible
to outsiders than its empirical practices (4). The same excuse applies to
sociologists of science, although with less force. Those philosophers of science
who do give a lot of weight to the technical side, the actual control of nature,

219
Norbert Elias, Herminio Martins and Richard Whitley (ed,.), Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982.219-238.
Copyright 1982 by D. Reidel Publishing Company.

220

Arie Rip

are often more interested in the consequences for the foundation of scientific
knowledge (e.g., Janich, Mittelstrass (5)) or for ontological issues (e.g.,
Bhaskar (6)) than in elaborating a balanced view. B6hme and van den Daele,
lately of the Max-Planck-Institute in Starnberg, are the exceptions, but their
work has been discontinued for the moment (7).
For the sociologist of science, the important issues are how the technical
side of science is involved in the dynamics of scientific developments, and
what the impact of technical innovation and its diffusion is on scientific
practices. These are large issues, and in this article I shall limit myself to a
discussion of the concept of 'restrictedness' and how it may contribute to
these issues.
The characterization of sciences as 'restricted' or 'unrestricted' has been
introduced by Pantin (8) and elaborated by Whitley (9). Especially from
Whitley's work, it is clear that 'restrictedness' is a bridging concept, combining features of the objects of the sciences and of their social organization.
Neither Pantin, nor Whitley emphasize technical aspects in their discussion of
'restrictedness'. I shall argue that the restrictedness of a science depends on its
control over the relevant part of reality, its ability to transform circumstances
into parameters and variables: Such an interpretation of the concept of
'restrictedness' makes it possible to discuss technical dynamics in the sciences
while including their social aspects. The approach of this article is thus the
same as when Shinn argues that "the logic contained in experimental operations and the manipulation of instruments impinges significantly on the cast
of social relations" (10). The difference is that Shinn looks at the relation in
contemporary laboratories, while I focus on its development over time.
2. The Concept of 'Restrictedness'
Whitley has shown how Pantin's rather unarticulated distinction between
restricted and unrestricted sciences (for example, between physics and
geology or parts of biology) can be made more precise and sociologically
interesting by looking at science as productive labour (11). Scientific work is,
as is work in general, concerned with transforming objects with tools for
some goal. Differences between sciences will therefore at least to some extent
be related to the specific features of what is being transformed and how, and
such features will give rise to differences in work organization.

The Development of Restrictedness in the Sciences

221

For example, when unrestrictedness of a science is seen as the consequence

of the "richness" of the constitutive phenomena, or of the complexity of the
systems analyzed (as Pantin does, and Whitley sometimes), it seems plausible
to associate it with field work and observation as the predominant styles of
work. Restricted sciences, on the other hand, will rely on laboratory work.
Geology or biology, and physics are paradigmatic examples of these relationships (12). Whitley elaborates this view further and considers also precision
of work tasks (which will be high in restricted sciences and formally assignable), differentiation of work and enforcement of standards (again, high
in restricted sciences) and nature of PhD training (routinized in restricted
sciences).
Whitley's and Pantin's discussion of cognitive and social differences
between physics and chemistry (considered to be highly restricted sciences)
and biology and geology (partly or mostly unrestricted sciences) shows that
there are clear differences. What is not shown, however, is that the root of
these differences is a difference in knowledge object. The characterization of
the nature and organization of work in restricted sciences, as given by Whitley
and summarized above, may be seen as reflecting a high level of articulation
and institutionalization, and not, or not immediately, a more restricted
knowledge object. Given time and social support, unrestricted sciences may
become institutionalized as well and show the same characteristics. The
difficulty of distinguishing between the observed differences and their causes
is compounded by the fact that differences between sciences, in style and
organization of work, tend to diminish with the rise of professionalized
science with a comprehensive "scientific method" (even if this is in fact the
method of physics as perceived by outsiders).
One way of getting around the difficulty of analyzing differences between
sciences in this way is to focus on the criterion of simplicity or complexity of
the objects of the sciences. However, this approach faces methodological
problems. The degree of complexity. of the material object of a science is
known only after the fact, when a consensus about its structure is reached.
What has to be explained, however, is the relation between work style and
organization of work, and features of the object at every stage of the historical
practice of the science.
In Whitley's later presentation of the concept of restrictedness, there are
still two aspects mixed up (13). He emphasizes ideals of explanation, e.g.,

222

Arie Rip

the arithmomorphic ideal of working with objects constituted by simple

elements that can be aggregated and dis-aggregated as one wishes. But a
necessary component of restrictedness is also the ability to actually fmd or
construct such elements in a satisfactory way. My point can be illustrated by
quoting Ziman's discussion of physics as the art of the soluble:
It is not simply good fortune that physics proves amenable to mathematical interpretation; it follows from careful choice of subject matter, phenomena and circumstances
(14).

Such choices are guided by explanatory ideals, e.g., of arithmomorphic

explanation. But the success of the choices is based on actual transformation
of his object:
In his experiments, the physicist contrives to create conditions that are simple to observe
and analyse; ( . . ) He constructed artificial systems with nearly ideal properties; ( ... )
(e.g., purity, homogeneity) (1S).

The success of such transformations will depend on the (unknown) nature of

the material object, as well as on subcultural norms as to what constitutes
success. Technical achievements are never just technical, there is always also
a social component included.
Starting from the social aspects of scientific work, the decision whether
a result is significant, or just noise or an artefact of the method, appears
to be a matter of negotiation between actors, as has been shown by Collins
in a number of cases (16). The goal of scientific work may be taken to be
the reduction of ambiguity, and ambiguity of meaning is reduced by sociocognitive processes of argument, negotiation and institutionalization. An
important input into this process is the reduction of ambiguity of outcomes
with the help of technical means. Qn a few cases, the outcomes are relatively
unambiguous by themselves.) Although such inputs are still dependent
for their significance on a negotiated consensus, they cannot be dissolved
completely into social processes.
The arguments presented here show that there is room for special attention
to the technical aspect of scientific work, and that it may be fruitful to take
the ability to transform the object so as to have relatively unambiguous
outcomes as the feature that distinguishes restricted from unrestricted sciences
at anyone moment in history. Restrictedness will then not be connected with
the material or knowledge object of the science, but with the amount of

The Development of Restrictedness in the Sciences

223

control over the knowledge object. In restricted sciences, the behaviour of the
knowledge object can be restricted, often very narrowly, to those aspects the
researcher wants to study. In unrestricted sciences, this is not, or not very
well possible, so that the behaviour of the object is unrestricted.
Since the definition proposed here is still directed toward the reduction
of ambiguity, the implications for the nature of scientific work and its
organization derived by Whitley on the basis of his definitions will still be
applicable. In addition, increases in the degree of restrictedness of a science
over time can now be linked directly to an increase in control over reality.
This point will be developed in the next section.
In principle, all cases intermediate between a completely unrestricted
science (if that may still be called a science) and a completely restricted
science (if that still shows dynamic development) are possible. In contemporary SCience, however, unrestricted sciences are not free to choose whether
to develop or not towards a higher degree of restrictedness. The high status
of physics as the paradigm of a restricted science forces the practitioners
of unrestricted sciences (e.g., the life sciences, but to some extent also the
social sciences) to aim for more restrictedness, even if they have to import or
imitate it. The alternative is to react against the status of physics as the ideal
science and emphasize other approaches: holism in biology, configurations in
sociology (17).
3. The Production of Restrictedness
Control over the knowledge object can be seen to increase through the
successful production of pure samples (in chemistry and medicine), "pure"
effects (in the physical sciences) and through tightly formulated procedures
(18). The historical reconstruction of the dynamic leading to increased control
is made difficult by the way modem science takes its power to control for
granted. For instance, in chemistry pure compounds are considered to be
non-problematic technically (they can be made) and conceptually (their
defmition is straightforward), while in fact, many chemists are not able to
give an acceptable definition, and depend on commercial suppliers for their
pure compounds (19). In this section, I shall use a few secondary sources
to sketch an account of the development of pure samples and pure ef(ects,
without attempting a full reconstruction.

224

Arie Rip

(a) The Production of Reproducible Samples

The notion of a pure compound, or even of a sample with reproducible
properties is absent in the lore and speculation of antiquity and medieval
times. Chemical recipes for jewellers and colour-makers (for cosmetics and
painting) became standardized to a certain extent, but did not guarantee
reproducible results because the starting materials were not standardized as is still the case in cooking recipes (20). The craftsmen had to produce
desired effects regularly, teach their pupils how to obtain these effects and
assure themselves of a regular supply of reliable starting materials. Tests were
used systematically only to detect frauds, especially in mint metal (21).
For the alchemists in antiquity, in the Arab world and in medieval Western
Europe, purification of materials by fire was important, but only as a method
or a therapeutical treatment to effect a transformation into gold. They did
not bother to derme their procedures and products, and the practical alchemical tradition was repeatedly turned into esoteric alchemy (22). Neither
the alchemical theories of material change, nor the Aristotelian theory of
generation and corruption provided incentives to develop the skills and
procedures necessary to obtain pure samples.
Kuhn has emphasized the priority of a theory, as a promise of potential
order, in the motivation of men to develop skills to convert potential into
actual order, and has given an example of such a bootstrap operation in 19thcentury analytical chemistry (23). The development of chemical techniques
from the 14th century through the 17th century, however, shows the theory
(as commonly understood) is not necessary and other processes are at work
(24). The search for quintessences and the technical innovation of distillation
to produce more efficacious drugs was one of these processes, the other being
the change in Moslem Spain from Galenic pharmacy, with its compounded
vegetable drugs, to "simple" chemical remedies and the controversies in
Western Europe over Paracelsian medical chemistry (including the attempts
of the Paracelsians to remove the toxic properties of their medicines by
reducing them to their essence) (25). The conception of purity as something
to be separated from impurities evolved in this way, since people wanted to:
learne the manner to separate by Arte the pure and true substance as well manifest as
hidden, the which in Phisicke is a great helpe to the taking away of diseases, harde or
rebellious to be cured (26).

The Development of Restrictedness in the Sciences

225

The demand for drugs made it possible to gain a livelihood from the
preparation of substances in the 16th and especially in the 17th century,
while at the same time, a demand for chemical recipes and instruction grew
(27). In the growth of analytical techniques (mineral waters and assaying
being the most important), there is a continuity extending from the 16th into
the 19th century (28). Natural philosophy did play an independent role,
especially in the introduction of corpuscular notions and the Cartesian
explanation of the behaviour of acids and alkalis, but it was one component
of a matrix in which medicine, alchemy and technology played their parts
(29).
During the 18th century, especially in France, chemists who had been
trained as pharmacists or physicians found a new and steadily expanding use
for their skills in the solution of industrial problems (30). In Germany, Stahl
and his followers formulated the phlogiston theory, partly in attempt to
dissociate themselves from lowly pharmacists and disreputable goldmakers
(31). The rationalization of chemistry by Lavoisier and his followers could
build upon the results of these earlier developments. The notion of a pure
sample was then related to the new theoretical structure, and this guided
attempts to improve techniques and procedures.
The technical dynamic operated within a framework of theory, but after a
time it became institutionally distinct from the ongoing research efforts in
chemistry. When teaching laboratories were established in the course of the
19th century, and later, when chemical analysis for monitoring purposes (e.g.,
water quality, industrial tests, food additives) became important around the
tum of the century, the regular production of pure compounds became a
necessary condition for obtaining reproducible results, and a commercially
interesting proposition. At present, a number of big companies offer a large
catalogue of chemical compounds and reagents ("chemicals"), graded according to purity.
(b) Changing Know/edge Objects
This historical sketch of the production of pure samples can be interpreted
as an account of a changing knowledge object. Slowly gathering momentum,
the changes become more rapid from the 16th century onward. The end
result is that the outcome of experiments depends less on the skill of the

226

Arie Rip

experimenter in handling diverse materials, than on preparing his materials to

specification. Variability is restricted in this way and control over outcomes
increased. A further step towards universal reproducibility was the institutionalization of procedures and tests and their incorporation in commercial
production. Only at the research front do scientists still prepare their own
samples.
The interpretation of the production of pure samples in terms of changing
knowledge objects is supported by the occurrence of similar developments
after the establishment of the concept of pure compounds and the techniques
of preparation. When radioactivity was discovered by Becquerel in 1896 as
mysterious "uranic rays", others tried to discover and isolate other radioactive
substances by using the emitted radiation as the identifying indicator, the
work of the Curies being the best known of these attempts (32). A lot of
effort went into the study of the nature of the rays, but another line of
research was initiated by Rutherford in 1899 through studying the changing
radiation intensity of thorium oxide and discovering an "emanation" that
itself showed some properties of radioactive substances. In this way, the
study of radioactive degradation started, but it turned out to be very difficult
to identify the substances produced. Experimental results were unsystematic
because nobody knew the system behind it, it depended on the history of
the sample, and different researchers produced mutually diverging results.
Gradually, and mainly through Rutherford's efforts, standardized procedures
to control the experimental set-up and identify substances by radiation
patterns emerged. The resulting samples were identified by code names like
Radium Emanation, Radium A, Radium B, Radium C, and, after further
efforts, given a place in the periodic system of elements. The theory of
radioactive transmutation, in the meantime, had to wait till the knowledge
object had become manipulable at will; only then could it get down to
systematic explanation.
This episode of confusion about radioactive substances was closed by
1913. A second example of changing knowledge objects is a still unfinished
story. Thanks to techniques of isolation and fractionation, present-day
biomedical research is able to produce all kinds of samples with biological
effects. In a number of cases, standardized procedures are available to obtain
samples with - hopefully - standardized effects, which are then adorned
with code names like CRF, corticotropin releasing factor (33). In exceptional

The Development of Restrictedness in the Sciences

227

cases (which are then hailed as breakthroughs), chemically identifiable

samples are produced, and the research then proceeds in terms of the chemical
compounds. On many occasions chemical identification is not, or only
tentatively, possible. Often, the identification methods that are used (e.g.,
immunological methods in the case of proteins) have their own, sometimes
unexpected uncertainties (34). Latour and Woolgar provide fascinating inside
views of scientific activity in these biomedical fields (35). In spite of the
sophisticated instrumentation, the difficulties in standardizing and identifying
samples are similar to 16th.18th-century chemistry and radio degradation
studies around the tum of our century. The sophistication, combined with
the large scale of biomedical research, has made it possible for commercial
firms to produce standardized samples ("biologicals"). It is an open question,
however, if the knowledge object can be controlled sufficiently to initiate a
theoretical dynamic typical of a restricted science.
(c) The Production ofRestricted Effects
The aim of modem experimental science has been described as the production
of effects:
The facts given in pre-scientific, everyday life become intersubjectively identifiable and
reproducible: the phenomenon becomes an effect, the multiplicity of "influences" is
differentiated and isolated (36),

In fact, as this quotation already suggests, for an effect to be scientifically

relevant it should be a "pure" effect, unadulterated by "noise", in a defmite
situation and free from disturbances (37). One should, more neutrally, speak
of restricted effects here; but purity (and order) are powerful concepts in our
culture and their use dignifies even technical work.
Pure samples - that is, well-defined samples - are one way of increasing
the probability of obtaining pure effects. Certain types of apparatus, for
instance air pumps and Leiden jars, to name two that were popular in the
18th century and important for the development of science, can be characterized as aimed at producing pure effects. If successful, they incorporate
the control that has been won over some aspect of reality: they produce
restrictedness where it was not present before.
According to Bohme and Van den Daele, instruments and apparatus (and

228

Arie Rip

also procedures) like air pumps and Leiden jars that produce or demonstrate
certain effects are synthetic instruments, and should be distinguished from
analytic instruments that extend and calibrate observation (38). They emphasize the role of both kinds of instruments in the development of science
in the 17th and 18th century. Research into the way the instruments worked
and attempts to refine and extend them provided the basis for theories of the
phenomena produced or measured by them. Only after successful theory
development (and the consequent rationalization of the instruments) is it
possible to view instruments as embodied theory.
The first use of analytic instruments was to explore new domains (e.g.,
telescopes, microscopes); scales were used to indicate the occurrence of a
loss or increase of weight, and only later in measuring the extent of the
change. New instruments like barometers and thermometers were themselves
objects of research, and gave rise to important scientific developments.
Bohme and Van den Dae1e trace the design and performance norms regUlating
the technical-scientific dynamic, and show that universality, that is, independence of surroundings and of the materials used in the instrument, is more
important than accuracy as such. Synthetic instruments also, and obviously,
provide opportunities to study the effects they produce. This, as well as their
possible entertaining and technological usages, determines the dynamic of
their development.
From the 17th century until now, effects continue to be important: new
domains are opened for exploration and their discoverers are honoured by
eponymy. After a time, the production of the effect is brought under control
and embodied in standardized apparatus (which sometimes takes off to a
separate career in other sectors, for instance Rontgen apparatus in medical
diagnosis and technological monitoring). The dynamic of the development at
the research fronts, where new effects have to be domesticated, is still the
same as in the 18th century. But the instrument makers of that time, with
their close relations to the evolving scientific community, have been replaced
by a flourishing branch of industry. With the production of monitoring
apparatus for applied science, industry and government as its origin and
backbone, the scientific instrument industry has profited from the expanding
market provided by modern scientific research, and now produces sophisticated instruments like NMR-spectrometers and special gaschromatographs.
Industrial decisions and innovations influence scientific developments, as well

The Development of Restrictedness in the Sciences

229

as vice versa; the present trend towards automation of analysis and measurement is a prime example (39).
4. Some Implications of Increasing Restrictedness

The increase in restrictedness in different sciences seems to be a rather

general feature. Reasons could be several, from the ingenuity of homo faber
and the scientific goal of universalism (requiring standardization) to the social
and cultural goal of Westem civilization to increase its control over nature
and man. In any case, success in the sciences appears to be connected with
restrictedness and increased control.
Within science, successful control is institutionalized through the reproduction of procedures and experimental set-ups, at first involving much skill
and tacit knowledge (40), but after a time in a more standardized way and
supported by regular training. The availability of ready-made restrictedness
(in the form of instruments, directions for use and written instructions)
allows further delegation of tasks. A division of labour results where technical
assistants (and students) are responsible for the production of research outcomes. To ensure reliable results, a researcher now has to exert social control
over his assistants, which replaces the direct technical control of the individual
researcher of a previous age. Besides cognitive resources, now also social
power becomes important, which is reflected in the internal struggles in the
laboratory, as Shinn shows in his discussion of a solid state physics lab
(41).
In the previous section, I have shown how the manufacture of pure samples and apparatus becomes independent of the scientific practices that gave
rise to them. Wide use of such pre-packaged restrictedness implies then that
scientific practices (in the same, but even more pronouncedly in other fields)
will depend on the routinized supply of samples and apparatus. The effect of
importation of restrictedness will depend on the nature of the object of the
science and the existing organization of work. In many unrestricted sciences,
instruments and standardized procedures are used primarily for gathering
data, and do not have a great impact on its development (except building up
"data pressure"). The imported restrictedness makes routinization of work
possible, for instance, in the mineral chemistry laboratories studied by Shinn
(42).

230

Arie Rip

There are obvious advantages in importing restrictedness. For an unrestricted science to really profit, however, it has to reorganize its practices.
Organic chemistry during the decades before 1940 is an example of a semirestricted science with accepted, "chemical", criteria for proofs of molecular
structures (43). After the Second World War, physical instruments were
increasingly used to provide information about molecular structure and
interaction, at first hesitantly and in the face of resistance from the old
elite. Later on, it became commonplace, and the interpretation of the data
provided was considered unproblematic, although chemical confirmation
was still sought. At present, the use of sophisticated apparatus has become
an indicator of respectable chemical work, and sometimes a criterion for
the acceptance of articles (which cannot be met any more by the poorer
laboratories).
The routinization of work and division of labour that is characteristic for
restricted sciences according to Whitley, becomes possible in unrestricted
sciences through importation of restrictedness, as well as indigenous growth.
The impact of routinization is felt at the lower strata of the scientific work
force, including research students. The career perspective of most scientists,
however, is to leave the drudgery of actually performing research tasks after
a time. This makes them accept, and later forget their earlier experiences,
and insufficiently aware of the influence of instrumentation and automation
on selection and outcome of research problems (44). Quality control thus
becomes difficult, but as an issue it normally remains part of scientific folklore. When restrictedness is imported, even the recognition of quality control
problems recedes. Only when controversies erupt, as happens every now and
then in food testing and environmental monitoring, will such issues become
manifest (45).
Restrictedness in the sciences also has effects outside science. Human
capacities to produce restricted situations outside the laboratory, transforming a local reality to be similar to the original experimental situation,
are the essential precondition for applying science. A lot of effort, technical
creativity and often social power is required to effect such a transformation.
The necessity of transforming reality to a restricted situation, before
science can be applied, helps to explain the "external", often unexpectedly
negative effects of the application of science. Agricultural science, for instance, although only partially restricted, needs monocultures to be applicable

The Developmentol Relltrictednellll in the Sciences

231

- as it were, the experimental plots enlarged n times. The required intervention in the ecosystem has its repercussions, which can and should be
distinguished from the side-effects of the applied science (for instance, with
pesticides). Another example is that the scale and general infrastructure of
power plants is often just as important as the specific technology (nuclear,
oll- or coal-fired) being used - as is now slowly being realized in technology
assessments (46).
S. Concluding Remarks
Viewing scientific developments as being partially determined by technical
dynamics, and describing such developments in terms of restrictedness may
also shed light on processes of theory formation. Increasing restrictedness
leads to empirical generalizations and conceptual distinctions, i.e., "bottomup" theory formation (47). Other dynamics playa role, for instance, the
speculative dynamic of world views like the mechanical philosophy of the
17th and 18th century, or atomism untll the 20th century. Such "top-down"
theory formation may induce an explanatory dynamic or process of paradigm
articulation, when the problems in the original explanation of the knowledge
object of a research area are taken as the starting point for theoretical and
experimental studies.
The existence of different dynamics in science emphasizes its heterogeneity, in spite of unified science ideals. In this respect, Kuhn's seminal
article on the mathematical and experimental traditions in science is particularly interesting (48). According to Kuhn, the mathematical tradition, which
goes back to antiquity and covers the "rational ordering sciences" (49), was
transformed during the Scientific Revolution through the innovation of
puzzle-solving groups, while at the same time a new tradition was born, the
Baconian approach of "twisting #le lion's tall", that is, subjecting nature
to artificial experiments. The latter tradition carries the technical dynamic
supporting increasing restrictedness, but ideals of rational ordering provide
stimuli to the experimental tradition.
Kuhn emphasizes the separate evolution of the two traditions, sometimes
far into the 19th century, because he is arguing against a unitary view of
science (50). He also draws attention, however, to a reshuffling of the traditions from circa 1800 onwards, when (pure) mathematics split off, many

232

Arie Rip

Baconian sciences were mathematized, and mathematical physics took over

the mathematical and mechanical tradition of the preceding centuries (51).
In that period, the basis was laid for the combination of bottom-up and
top-down theory formation, which in spite of its uncertain integration, is
sometimes glorified as the scientific method (52).
Kuhn's analysis has limitations (53), but it does provide a historical basis
for an analysis of the cognitive and social relations between disciplines, and,
perhaps, a re-evaluation of the position of physics. One point is that, given
the status of physics as the provider of the most basic explanations (54),
theory formation in other sciences can proceed in two directions: further
articulation according to its own explanatory dynamic, and connection with
physical theory (by deduction, approximation, analogy or even more indirect
means) (55). For instance, in organic chemistry bottom-up theory formation
has been combined, first with the electron theories of the first decades of this
century, and later with quantum theory.
By now, a separate specialty of quantum chemistry has evolved, which
tries to specify the general equations of quantum mechanics to the level
of molecular structure, developing specific approximations to reduce
the complexity of chemically still rather simple molecules. Such an approach is an example of "top-down" theory formation: theoretical explanations are constructed speculatively in terms of axioms, general laws
or approaches. Experiments may be designed to test and/or articulate
such theories further. Especially if the status of the starting theory is high,
however, the resulting theory may be used primarily as a ritual, as is the
case in large parts of quantum chemistry (56). Practitioners in a science
sometimes condemn the more baroque outgrowths of top-down theory
elaboration. There are some cases, for instance, the quantum theory of
the solid state in the 1930s, where an independent experimental dynamic
comes to the rescue by providing opportunities for systematic articulation
(57).
Another point is the question why the mathematical tradition, as carried
on in contemporary mathematical physics, is more powerful than other
traditions. Part of the answer lies in the extent of control over its knowledge
object that physics has achieved and now exports to other disciplines. Other
factors are the ideals of explanation that are favoured at some time (and
which now may be slowly shifting in the direction of more holistic ideas, with

The Development of Restrictedness in the Sciences

233

a corresponding rise in the status of biology); and the social status of the
practitioners of the different sciences.
Although data on social backgrounds and social status of practitioners of
different disciplines are practically non-existent, some information about
physicists and chemists provides a tantalizing glimpse of historical processes
at work to keep a distance, both socially and intellectually, between the two
disciplines. Kuhn notes that during the 17th and 18th centuries, practitioners
in the mathematical tradition were more or less established, and held posts in
universities and the newly-founded academies. Amateurs were the mainstay
of the Baconian sciences, and even when practitioners of these sciences were
members of an academy, they were most often second-class citizens (58). The
role of different social groups and intellectual traditions in the origin and first
phase of the Scientific Revolution is difficult to disentangle. In any case, the
establishment of the royal academies in Britain and France may be seen as the
first step towards pacification between the elite of the new science and social
and political powers (59). Other such moves occurred at later stages; for
instance, Stahl's efforts in Germany to put theoretical chemistry on the map
(60).
When the mathematical and the experimental traditions become associated
with different social statuses, this difference _will be maintained by social
processes like co-optation, differential recruitment and social and intellectual
distancing. Medically trained men for a time provided a middle ground, socially
belonging to the higher classes, but intellectually often inclined toward the
experimental tradition (or to holistic approaches as in the Naturphilosophie).
Mter the complete mathematization of physics, the distinctions may become
quite sharp, as shown for example by Pyenson and Skopp for physics and
chemistry around 1900 (61).
Such thoughts are speculations, however. They have been set down
because they derive from looking at science from the point of view of
restrictedness, and combine social and cognitive aspects in a bridging
concept. The fruitfulness of the concept of restrictedness itself lies in its
bridging the gap between social and cognitive aspects of scientific developments, which is a necessary condition of achieving a real understanding of
science.

234

Arie Rip

Notes and References

1. Derek de Solla Price, 'Philosophical Mechanism and Mechanical Philosophy: Some
Notes Towards a Philosophy of Scientific Instruments', manuscript paper, Clark
Library Seminar, UCLA, March 10, 1979.
2. Barry Barnes, Scientific Knowledge and Scientific Theory, London: Routledge and
Kegan Paul, 1974, pp. 84-5.
3. J. D. Bernal, Science in History, Harmondsworth: Penguin, 1969, p. 3.
4. The bipartition of empirical approaches to the subject matter of science and the
discourse about it is taken from Gernot Bohme, 'Cognitive Norms, Knowledge
Interests and the Constitution of the Scientific Object: A Case Study in the Functioning of Rules for Experimentation', in Everett Mendelsohn, Peter Weingart and
Richard Whitley (eds.), -The Social Production of Scientific Knowledge, Dordrecht/
Boston: D. Reidel, 1977, pp. 129-41.
5. See for example Peter Janich, 'Physics - Natural Science or Technology?', in
Wolfgang Krohn, Edwin T. Layton, Jr. and Peter Weingart (eds.), The Dynamics of
Science and Technology, Dordrecht/Boston: D. Reidel, 1978, pp. 3-27; and JUrgen
Mittelstrass, Die M6glichkeit von Wissenschaft, Frankfurt a/Main: Suhrkamp, 1974.
6. Roy Bhaskar, A Realist Theory of Science, Leeds: Leeds Books Ltd., 1975.
7. See, for instance, Gernot Biihme, Wolfgang van den Daele, 'Erfahrung als Programm
- tiber Strukturen vorparadigmatischer Wissenschaft', in Gernot Bohme, Wolfgang
van den Doole, Wolfgang Krohn, Experimentelle Philosophie, Ursprlinge autonomer
Wissenschaftsentwicklung, Frankfurt a/Main: Sukrkamp, 1977, pp. 183-236; and
Gemot Bohme, Wolfgang van den Daele, and Wolfgang Krohn, 'The "Scientification" of Technology', in Krohn, Layton, JI. and Weingart, op. cit., 1978 (Note 5),
pp.219-50.
8. C. F. A. Pantin, The Relations Between the Sciences, London: Cambridge University Press, 1968. Definitions are given on pp. 18 and 24.
9. R. D. Whitley, 'The Sociology of Scientific Work and the History of Scientific
Developments', in Stuart S. Blume (ed.), Perspectives in the Sociology of Science,
Chichester: John Wiley. 1977, pp. 21-50. Some further refmements appear in
Richard Whitley, 'Changes in the Social and Intellectual Organization of the Sciences: Professionalization and the Arithmetic Ideal', in Mendelsohn, Weingart and
Whitley,op. cit., 1977 (Note 4), pp.143-69.
10. Terry Shinn, 'Scientific Disciplines and Organizational Specificity: The Social and
Cognitive Configuration of Laboratory Activities', in this volume.
11. See Whitley, op. cit., Note 9, and for the general view Richard D. Whitley, "Components of Scientific Activities, Their Characteristics and Institutionalization in
Specialties and Research Areas: A Framework for the Comparative Analysis of
Scientific Developments', in Karin D. Knorr, Hermann Strasser and Hans Georg
Zilian (eds.), Determinants and Controls of Scientific Development, Dordrecht/
Boston: Reidel, 1975, pp. 37-73.
12. Pantin,op. cit., 1968 (Note 8), seems to imply that there is a one-to-one relationship between restrictedness, artificial experiments and laboratory work on the one
hand, and unrestrictedness, natural experiments (and observation) and field work
on the other hand. His own designation of astronomy as a restricted science shows
that other combinations are possible: astronomy of the planetary system relies on

The Development of Restrictedness in the Sciences

13.

14.
15.
16.
17.
18.

19.
20.

21.
22.
23.

235

natural experiments. (It should be noted that astronomy is a difficult case to

classify: as soon as it is concerned with origins and cosmology, disciplined speculation becomes important.) Examples of a combination of unrestrictedness and
artificial experiments are provided by animal psychology and some parts of human
psychology.
Whitley, 'Changes in the Social and Intellectual Organization of the Sciences',
op. cit., 1977 (Note 9), p.147. In the earlier article Whitley says "In some sciences,
events and phenomena are embedded in a highly esoteric theoretical structure
which requires elaborate technical facilities for their production. ( ... ) As I understand him, this view corresponds to objects in Pantin's 'restricted' sciences. In other
fields, scientific objects are constructed in a less theoretically specific manner. They
are not so narrowly conceived and incorporate a range of attributes and properties".
Whitley, "The Sociology of Scientific Work and the History of Scientific Developments",op. cit., 1977 (Note 9), p. 26.
John Ziman, Reliable Knowledge. An Exploration of the Grounds for Belief in
Science, London: Cambridge University Press, 1978, p. 28.
Ibid., pp. 29-30. Ziman does not elaborate this point because he is interested in
epistemological issues. The moral he draws is: "Physics is the harvest of this Kantian
fisherman ( ... )" (p. 30).
For example, H. M. Collins, 'The Seven Sexes: A Study in the Sociology of a
Phenomenon, or the Replication of Experiments in Physics', Sociology 9 (1975)
205-24.
Compare Norbert Elias, 'Scientific Establishments', in this volume, also Whitley,
'Changes in the Social and Intellectual Organization of the Sciences', op. cit., 1977,
Note 9.
Procedures in the sense of conceptual techniques, for instance Linnaeus' "artificial"
classification of plants by selecting the reproductive organs as determinants of
species and conceiving all other characteristics to be irrelevant; see Bohme and Van
den Daele, op. cit., 1977 (Note 7), pp. 207-9.
See Barnes,op. cit., 1974 (Note 2), pp. 65-6 and P. Bulthaup, Zur gesellschaftlichen
Funktion der Naturwissenschaften, Frankfurt a/Main: Suhrkamp, 1973, p. 70.
R. P. Multhauf, The Origins of Chemistry, London: Oldbourne, 1966, provides
many examples for early chemistry. In some cases, the necessity of skillful handling
can be traced until much later times, for example in the use of the blow pipe
(Lotrohr) for analytical purposes (Ferenc Szabadvirry, Geschichte der Analytische
Chemie, Braunschweig: Friedr. Vieweg and Sohn,1966, p. 70). Bulthaupt, op. cit.,
1973 (Note 19), p. 71, quotes a textbook from 1913 saying that the uncertainties
of situations calling for such skills contribute to the joy of discovery.
Multhauf, op. cit., 1966 (Note 20), p. 37; SZabadvary, op. cit., 1966, pp. 16-9.
An assay of ore based on cupellation was specified in writing by royal decree in
Hungary in 1342 (ibid., pp. 28-30).
Multhauf,op. cit., 1966, Note 20.
Thomas S, Kuhn, 'The Function of Measurement in Modern Physical Science', in
The Essential Tension. Selected Studies in Scientific Tradition and Change, Chicago
and London: University of Chicago Press, 1977, pp. 178-224. On p. 196 he
discusses the work on multiple proportions that reached its final results only after
Dalton's corresponding law was taken as a prior criterion.

236

Arie Rip

24. If Janich, op. cit., 1978 (Note 5), pp. 19-20, is followed, theory is considered as a
means for making knowledge communicable, including the knowledge obtained by
an experimenter in terms of what is and what is not successful. On such a view of
theory, it does play an important role.
25. Multhauf, op. cit., 1966 (Note 20), pp. 216, 222, also 281; also Allen G. Debus,
Man and Nature in the Renaissance, London: Cambridge University Press, 1978,
pp. 29,45-7,53. There is a curious parallel in Debus's terminology for humanist
methods, when he speaks about the humanist adherence to ancient philosophy
"provided that they were assured that their texts were pure and unadulterated"
(ibid., 7).
26. Debus,op. cit., 1978 (Note 25), p. 46.
27. Multhauf,op. cit., 1966 (Note 20), p. 353; also Marie Boas, The Scientific Renaissance 1450-1630, London: Fontana Books, 1970, p. 154. Other developments are
the wide circulation of recipe books and the start of industrialized drug production
(Multhauf,op. cit., 1966 (Note 20), pp. 258, 262).
28. Szabadv3.ry,op. cit., 1966 (Note 20), pp. 41-6 and 58-9.
29. Thus Multhauf, op. cit., 1966 (Note 20), p. 349.
30. Henry Guerlac, 'Some French Antecedents of the Chemical Revolution', Ambix 5
(1959) 73-112.
31. Karl Hufbauer, 'Social Support for Chemistry in Germany during the 18th Century:
How and Why Did It Change?' Historical Studies Phys. Science 3 (1971) 205-31.
The strategy of intellectual distancing noted here is treated also by R. G. A. Dolby,
'On the Autonomy of Pure Science: The Construction and Maintenance of Barriers
between Scientific Establishments and Popular Culture', in this volume.
32. The discussion in this paragraph is based on Alfred Romer (ed.), Radiochemistry
and the Discovery of Isotopes. New York: Dover, 1970. See also T. J. Trenn, The
Self-Splitting Atom. A' History of the RutherfordSoddy Collaboration, London:
Taylor and Francis, 1977.
33. Nicholas Wade, 'Guillemin and Schally: The Years in the Wilderness', Science 200
(1978) 279-82; also Bruno Latour and Steve Woolgar, Laboratory Life. The Social
Construction of Scientific Facts, Beverly Hills and London: Sage, 1979, p. 60.
34. Lois Wingerson, 'Mistaken Identity of Proteins Threatens Biochemists' Results',
New Scientist 86 (24 April 1980) 192-3.
35. Latour and Woolgar, op. cit., 1979, Note 33.
36. Bohme and Van den Daele, op. cit., 1977 (Note 7), p. 223 (my translation).
37. The experimenter in fact attempts to control chaos (compare Ziman, op. cit., 1978
(Note 14), p. 59) and in spite of the successful exclusion of ever more "disturbing
factors", ambignity remains: is the effect real, or a result of a "dirty system"?
38. Bohme and Van den Daele,op. cit., 1977, Note 7. The same distinction is made by
Janich,op. cit., 1978, Note 5.
39. Studies of the development of the instrument industry have mostly concentrated
on the economic aspects. The inter-relationships with scientific developments and
the problem of quality control are well worth further study. In chemistry, for
instance, the classical elemental mass analysis is now fully automated, and often
performed by special laboratories. Folklore has it that one in ten analyses goes
wrong, and that mistakes are not always spotted. Clinical analyses are troubled by
the same problem, aggravated by the greater variability of the samples and the risk .

The Development of Restrictedness in the Sciences

40.
41.
42.
43.

44.
45.

_46.

47.

48.
49.
50.
51.
52.

237

of human suffering caused by errors. Indicative is a conclusion of Ferreira and

Bold, when they state: "An accurate analysis is not as simple as is implied in
advertisements. Reliable results can only be obtained by a skilled operator who has
a thorough knowledge of the possible pitfalls involved" (P. Ferreira and A. M. Bold,
'Clinical Laboratory Evaluation of the Orion SS-20 Ionized Calcium Analyser',
J. Automatic Chem. 1 (1979) 94-102, p. 101).
See, for example, H. M. Collins, 'The TEA Set: Tacit Knowledge and Scientific
Networks', Science Studies 4 (1974) 165-86.
Shinn,op. cit., Note 10.
Ibid.
Organic chemistry was a restricted science in the sense that it could dispose of
well-delmed samples and standardized techniques to prepare them. It was unrestricted in the sense that properties and reactions of the samples and compounds
could not be controlled, in spite of a huge amount of accumulated experience and
some empirical generalizations. (See also Emil Broesterhuizen, Arie Rip, Organic
Chemistry, A Techno-5cience. Speculations about Theory Development in Organic
Chemistry, unpubl. report (m. 7801), Leiden: Chemistry and Society Programme,
1978).
This view was suggested to the author by P. Groenewegen.
A number of historical and contemporary examples are analyzed by Ross Hume
Hall, Food for Nought, The Decline in Nutrition, Hagerstown (Mid.): Medical
Dept., Harper & Row, 1974. In -environmental controversies about levels of lelld,
DDT and PCBs, the suspicion that concern is caused by increased sensitivity of the
instruments is often voiced by industry.
A (dramatic) example of some repercussions of application, undercutting the
universal validity of the original scientific Imdings, is the case of DDT: it killed
mosquitoes in the 1950s, but fails to do so in the 1970s - at least, in some parts of
the world. G. Bohme, Wolfgang van den Dae1e, Wolfgang Krohn in 'Finalization in
Science', Soc. Sci. Information 15 (1976) 307-30, p. 323, comment as follows:
"The perspective of nature as an inlmite reservoir for intervention cannot be
abandoned as long as the theories of natural science do not make it possible to
integrate the historical impact of such intervention into their concepts. Therefore,
scientific knowledge will continue to refer to reproducible possibilities, even if the
presuppositions of these possibilities are undermined precisely by the process of
their reproduction." They continue to plead for ecology as a normative, strategic
science.
Bohme and Van den Daele, op. cit., 1977 (Note 7), p. 214. One of their examples
that supports my point about restrictedness giving rise to bottom-up theory formation is experimental electricity in the 18th century (p. 195). This appears to be also
the conclusion of J. L. Heilbron, Electricity in the 17th and 18th Centuries. A
Study of Early Modem Physics, Los Angeles and London: University of California
Press, 1979.
Kuhn,op. cit., 1977, (Note 23), pp. 31-65.
Bohme and Van den Daele,op. cit., 1977 (Note 7), p. 206.
Kuhn,op. cit., 1977 (Note 23), pp. 31-3,46.
Ibid., pp. 60-5. Compare also the quotation from Heilbron, Note 47.
This discussion does not do justice to the complexity of the social, cultural and

238

53.

54.
55.

56.
57.
58.
59.

60.
61.

Arie Rip

technical transformations of this period, of which the transformation of science is

but one part. An important aspect of this transformation is, for instance, how
"natural history" is taken over by the mathematical and mechanical tradition;
compare David Knight, The Nature of Science. The History of Science in Western
Culture Since 1600, London: Andre Deutsch, 1976, and Wolf Lepenies, DasEnde
der Naturgeschichte, Miinchen, Wien: Carl Hanser Verlag, 1976.
Kuhn,op. cit., 1977, Note 23, notes his neglect of the life sciences and medical
sciences (footnote 9, p. 40-41), and makes a passing remark about elaborate data
being important for scientific development, but available only when their collection
fulfills some perceived social function (footnote 6, p. 38).
See Arie Rip, 'Science Policy and the Scientific Picture of the World', in W. Callebaut, M. de Mey, R. Pinxten, F. Vandamme (eds.), Theory of Knowledge and
Science Policy, Ghent: Communication and Cognition, 1979, pp. 358-77.
There is an interesting analogy with the way "rules of the thumb", the empirical
rules important in technology and industry, are articulated: they can be refined
in an empirical way, or an attempt is made to relate them to the body of scientific
knowledge. The connection between more empirical or application oriented knowledge and general, fundamental theory is in both cases in the opposite direction
from the one predicted by the finalization thesis (Bohme, Van den Daele, Krohn,
op. cit., 1976, Note 46).
Broesterhuizen and Rip, op. cit., 1978, Note 43.
E. Homburg, personal communication, based on the case-study of solid state
physics and solid state physicists in.Loet Leydesdorff e.a.,Philips en de wetenachap,
Amsterdam: SUA. 1980.
Kuhn,op. cit., 1977 (Note 23), pp. 51-2, aiso Multhauf,op. cit., 1966 (Note 20),
p. 269, and Debus, op. cit., 1978 (Note 25), p. 140.
The English case has been studied by Wolfgang van den Daele, 'The Social Construction of Science: Institutionalization and Defmition of Positive Science in the Latter
Half of the 17th Century', in Mendelsohn, Weingart and Whitley, op. cit., 1977
(Note 4), pp. 27-54.
See Hufbauer,op. cit., 1971, Note 31.
Lewis Pyenson and Douglas Skopp. "Educating Physicists in Germany circa 1900",
Social Stud. Sci. 7 (1977) 329-66. They note that chemical training is an avenue
of social mobility, while physics is closely connected with the elite.

SCIENTIFIC DISCIPLINES AND ORGANIZATIONAL

SPECIFICITY: THE SOCIAL AND COGNITIVE
CONFIGURATION OF LABORATORY ACTIVITIES

TERRY SHINN
C N R S,Paris

In some respects this essay complements the analysis formulated in three of

the previous studies. James Fleck and Arie Rip both allude either directly or
obliquely to the specificity of the social and cognitive processes present in the
domains of artificial intelligence and chemistry (1). This paper takes that
emphasis one step further as it compares the intellectual and social structures
of a range of disciplines (2); this with the aim of drawing fuller attention to
the unique character of research practices in different fields of scientific
investigation. In addition, Edward Yoxen, together with Fleck and Rip, all
stress the importance of instrumentation to the activities of their respective
sciences. Again, in this contribution the use of different scientific instruments
in various sub-fields is examined in detail as a key determinant of laboratory
morphology. Thus, while Fleck, Rip and Yoxen concentrate on the macroscopic issues of disciplinary development and evolution, here the emphasis
will be placed on the microscopic facets of laboratory organization and
operations.
In the following pages two dimensions oflaboratory activity are explored:
first, the social organization of inquiry: that is, the configurations of authority, hierarchy and communication present in the research setting. Secondly,
cognitive organization: by this is meant the pattern of the intellectual division of labour occurring within research. These twin parameters are scrutinized in three areas of scientific endeavor (mineral chemistry, solid-state
physics and computerized vector analysis) in order to demonstrate the
degree to which disciplinary specificity conditions laboratory morphology
(3).
239
Norbert Elias, Herminio Martins and Richard Whitley
Establishments
and Hierarchies. Sociology o/the Sciences, Volume IV, 1982. 239-264.
Copyright 1982 by D. Reidel Publishing Company.

240

Terry Shinn

1. The Social and Cognitive Morphology of Inquiry

Observations conducted in three laboratories of mineral chemistry point to

the predominance of highly formalized and elaborate social relations, and to
cognitive patterning predicated on differentiated operations and marked by
strictly delineated connective mechanisms (4). Authority here is concentrated
in the hands of a single person. The hierarchical lattice is rigid and complex.
Communications tend to be channelled and codified. Intellectual activity is
equally prescribed, as experimentation is largely dissassociated from the
gathering, collation and organizing of fmdings. The manipulation and interpretation of results, and their relevance to hypotheses and eventually to
theory, is also removed from other mental procedures.
In these chemistry research units all facets of laboratory life are strongly
centralized. Authority is monopolized by a research director who sets forth
general policy and, at the same time, regulates the minute details of routine
activity. Research directors are, in a sense, onmipotent and onmipresent.
Decisions, both scientific and administrative, are taken in complete isolation,
as even senior scientists are seldom consulted. The influence of research
directors is further enhanced since, in most instances, it is they alone who
establish and maintain contact with extra-mural groups and institutions for
the purpose of publication, fmance and formulation of research strategy.
Control is thus personified in the form of the research director who consistently and effectively rejects the tenets of pluralistic leadership.
These laboratories are typified by a five-tier hierarchy (see Figure 1). The
research director occupies the position at the apex of the pyramid, and below
are situated, in descending order, two categories of scientists (senior and
junior) and two strata of technical personnel (technicians and laboratory
assistants). The numerical ratio of the technical personnel to the two groups
of scientists is somewhat in excess of 2.2 : 1. Also, this hierarchy is not
simply elaborate, it is also relatively static. The probability of achieving
upward mobility is remote: only one scientist in two succeeds in rising, and
less than 10 per cent of the technical staff manage to better their condition.
Although these statistics apply specifically to intra-laboratory movement,
supplementary professional progress cannot even be attained by taking
work in other research units. Finally, hierarchic position is not merely a
matter of status and remuneration, since, as will later be demonstrated,

Scientific Disciplines and Organizational Specificity

----1

Fr:equcncy
of contact

--

241

infrequont

mod"rat"y frequent
Ir,qucnf

Fig. 1. Chemistry laboratories.

a sharply defined intellectual function corresponds to each level in the

pyramid.
The combination of this hierarchic arrangement and the unmuted authority
exercised by research directors tends to stultify and circumscribe communications within these research centres. Communication follows a predominantly
vertical axis where the bulk of interactions are both initiated and received by
the research director. Laboratory directors are thus in direct and frequent
contact with all categories of personnel.
But while vertical communication is abundant, horizontal interactions are
precarious. Contact between individuals holding posts in the same hierarchic
strata is restrained. This is, in part, due to the fact that written exchanges are

242

Terry Shinn

regarded as imperative to the smooth operation of the laboratory, while oral

discourse is seen as less central. Most communication, then, takes the form of
written memoranda, personal administrative and research notes and data
reports. Some verbal exchanges do nevertheless occur during the periodic
laboratory meetings, but such gatherings are generally formal in tone, and
thereby constitute an inhospitable terrain for
conversati0I?-'
As in the case of their social organization, the cognitive patterning of
chemistry laboratories is simultaneously fragmented and focussed. Each
group within the research unit is responsible for a body of precisely defined,
nonoverlapping, tasks; the result of the ensuing endeavour is subsequently
transmitted to the research director. The selection of a scientific theme,
designation of a narrower field of investigation within the chosen field, and
specification of research strategies and methods are among the principal
activities of research directors. It is also they who assess experimental data,
who weigh it against the requirements of the original working hypothesis
and integrate it into existing chemical theory. Research directors thus reign
supreme over the initial and terminal stages of research projects. In effect, in
these chemistry research units, most of the cognitive operations connected
with scientific conceptualization and the processes of associating fmdings
with a theoretical framework tend to be the quasi-exclusive preserve of a sole
individual.
The upkeep and preparation of apparatus and materials, experimental
manipulations, and in many instances observations themselves are carried
out by the two strata of technical personnel. Laboratory assistants and
technicians hence undertake all menial work and most routine tasks. Yet
since they have immediate contact with the desiderata of experimental
procedures and are best acquainted with the vicissitudes of particular experiments, the two groups also comprise an essential element in the observation
process. Thus, while largely cut off from theoretical matters, the technicians
and their aides, more than any other segment of the staff, record the characteristics and dynamics of the phenomena under observation and, by so doing,
playa decisive and immensely delicate role in determining the laboratory's
scientific output, a role totally incommensurate with their subordinate status
within the formal organization.
The activity of senior and junior scientists involves linking the theoretical
programme and exigencies elaborated by the research director, on the one

Scientific Disciplines and Organizational Specificity

243

hand, to their materialization in the form of bench work, on the other. This
countenances two intellectual operations. First, individual scientists transform
specific aspects of the original theoretical options and hypotheses into cognitive terms suitable for experimental treatment. They thus translate sometimes
vague or inoperable theoretical phrasing into experimental language. Secondly,
the scientists corroborate and validate observation. To some extent they
supervise experimental procedures, and in those cases where findings do not
coincide with expectations, it is they who attempt to identify the cause of
discrepancy. (For a representation of cognitive patterning in the chemistry
research units, see Figure 2). The cognitive patterning of these research units
conceptual

theory

data I information

collation

Fig. 2. Chemistry research.

thus features a marked degree of specialization as particular operations theorization, experimentation, validation of observations, etc. - accrue to a
predetermined segment of the personnel, and as activity in more than one
domain of research is uncommon.
In many respects the social and cognitive characteristics of laboratories
engaged in computerized vector analysis differ dramatically from those
involved in mineral chemistry. The source and tenor of authority is diffuse

244

Terry Shinn

and ill defined. The hierarchic apparatus is unsubstantial and attenuates group
distinctions within the laboratory. Exchanges between the personnel are
largely oral, multidirectional and profuse. In like fashion, cognitive specialization is minimal. Individuals take on one task or another in conformity
with the needs of a given research project. Work is neither allocated nor
permanently fixed.
While, in principle, the organization and operation of computer laboratories falls under the aegis of a research director, in reality, decisions are taken
collectively by the ensemble of the personnel and are arrived at informally.
It cannot be said that research directors "wield authority". In de facto termS
their prerogatives are largely nominal, as in most situations their voice counts
for no more than that of their colleagues. Here, authority is not an affair
of orders and coercion, but rather a matter of persuasion and communal
responsibility. Hence, as a matter of routine, scientists and technicians alike
are caught up in the process of establishing research strategies and selecting
appropriate methods. In almost all instances, these portentous decisions, as
well as lesser ones, are reached through casual and lengthy discussion which is
suspended only after some sort of consensus is reached. Decision-making is
therefore a relatively cooperative and flexible venture in which peers come
together to share their individual reflections, and agree to a common course
of action.
The hierarchic configuration of computer laboratories is both unimposing
and uncomplicated. They feature only three categories of personnel: research
directors, a single stratum of scientists and one contingent of technicians
(see Figure 3). Scientists constitute the largest group, as they account for
roughly 60 percent of the staff, outnumbering technicians by 2 : I. The
geometric representation of this hierarchic disposition could be thought of as
an elongated parallelogram, where the negligeable distance separating the
peak and the base evokes the tightly-knit nature of the organizational setting.
The setting is fluid, semi-amorphous and free of encumbering barriers and
divisive intermediary strata. This organizational malleability even extends
to professional mobility, as almost one scientist in two becomes a research
director before retirement, and over 80 percent of technicians are promoted
to the position of scientist. It is not incidental that the technical personnel of
these laboratories enjoy a much more advanced education than the analogous
categories in mineral chemistry labs. While the two contingents of chemistry

Scientific Disciplines and Organizational Specificity

245

Inlrequent

mod,r.'l/y Ir., ulnt

Fig. 3. Computer laboratories.

technicians have, at best, one or two years' training, and far more often do
not even possess a, secondary school diploma, those in computer centers
usually have one if not two university degrees.
The integrating dynamic of these research units is also manifest in their
propensity for unprogrammed and unfettered communication. All three
categories of personnel stand equal as initiators and recipients of administrative and scientific information. The place occupied in the communication web
by scientists and technicians is no less salient than that of research directors.
Data and instructions move with the same regularity along the vertical and
horizontal axis. The quantity of information exchanged among the research
staff is prolific and contact tends to occur in the most unsystematized manner. As a rule, exchanges take the form of spontaneous conversation involving
from two to four individuals. Scheduled meetings are rare, and written
documents are practically regarded as anathema. This suppleness of contact
applies not merely to intra-laboratory exchanges, but also to inter-research
unit communications. Albeit research directors seem somewhat privileged
in their dealings with other research centres, scientists, and for that matter,
technicians as well, also maintain working relations with outside scientists and
occasionally with government bodies and industry.
In computer laboratories, intellectual activity is not subjected to the
severely constricting effects of the division of labour. The full range of tasks

246

Terry Shinn

engendered by research is undertaken at different times by different individuals in conformity with their momentary tastes, temperament and current
state of intellectual preparedness. People are not assigned work; they gravitate
toward particular operations for the duration of a specific project and then
shift to different types of work as new projects emerge. Within the confmes
of the computer research centre there exists, therefore, a bastardized variety
of the "butterfly society" where intellectual migration comprises an overriding feature of cognitive inquiry.
Research themes are sometimes proposed by a scientist and at others by a
research director or member of the technical staff. In the three laboratories
studied for this project, about half of the themes were suggested by scientists,
a fifth by research directors, and a third by technicians. The task of translating
the research topic into a coherent and mathematically manipulable set of
preliminary hypotheses is similarly shared by the entirety of the research
team. This extremely difficult, abstract, and often abstruse work, does not
All
belong to the research director or to a select group of senior
of the personnel are actively included in the process. The same is true for the
assessment and interpretation of the numerical values provided by the computer. It is virtually certain that over the span of two research programmes,
each person within the research unit will participate in the fmal step of
research, attempting to make sense of the quantitative data, to fit it into the
outline provided by the working hypothesis and to relate it to questions
of theory either in the domain of physics or mathematics. Hence, in these
laboratories the domains of conceptualization and analysis are not restricted
to the few, since it is seen as fundamental that each individual contributes
fully in the creative dimension of mathematics-based investigation (see
Figure 4).
In the same vein, all levels of personnel are involved in the more tedious
and less imaginative work of the laboratory. Technicians are not alone in
carrying out such tasks as secondary programming, transferring arithmetic
functions on to tape, and later collecting and collating computer data sheets
and claSSifying the information they contain. Scientists and research directors
are also customarily occupied with such activity. Granted, the amount of
time they invest in routine operations is less than that of technicians, yet,
they nevertheless shoulder over half of the more banal work; and this,regardless of their seniority or formal position in the research unit. Inside computer

Scientific Disciplines and Organizational Specificity

247

conclptual
theory

data I information
coliaiion

Fig. 4. Computerized research.

laboratories, it seems then, that cognitive operations are not, or are but slightly, bound to a set of rigid conventions. It is the established norm for each
individual to take up the challenge of the complete research programme, which
implicitly necessitates contact with the more arduous aspects of investigation
as well as the opportunity to take part in the creative conceptual tasks.
While the social and cognitive patterning of physics research units evince
some of the traits common to chemistry and computer laboratories, they also
manifest a host of unique aspects (5). Leadership is semi-participatory, as
various factions continually contest the power of decision. The hierarchy is
segmented, resulting in internal cleavage, and while communications are
definitely channelled, they nevertheless frequently follow a circuitous path.
Cognitive operations are also torn between the effects of centripetal and
centrifugal influences. Subordinate personnel are formally relegated the task
of carrying out and overseeing some of the less creative work, yet they also
initiate opportunities to take part in conceptual operations. Conversely,
senior staff seek to maintain a grasp over matters touching on the formulation
of hypotheses and the assessment of fmal results, but they are often compelled
to take on less purely cerebral tasks.

248

Terry Shinn

In the three physics research units here discussed, authority is neither

predetermined nor diffuse. Instead it is a continuing process in which two
parties perpetually negotiate for domination over laboratory operations. A
research director, on the one hand, and a tiny circle of scientists, on the
other, each seek to impose their respective priorities and judgements on
the unit as a whole. Research directors generally found their claim to preeminence on seniority, scientific experience, and most particularly, on their
fruitful connections with the institutions on which research so heavily depends
- journals, government funding agencies, industry, etc. Senior scientists, in
contrast, insist that their direct involvement with the very stuff of research
places them in an unassailable position to assess laboratory requirements.
A sharply polarized context thus emerges where the prerogative of control
is legitimized in terms of social criteria on the one hand, and in terms of
cognitive imperatives on the other. Relationships between the adversaries
is conflictual, taking the form of impassioned debate. The conflict occurs
sometimes in formal circumstances such as planned meetings, and at other
times in a more random and haphazard setting, for example, chance encounters in corridors, visits to colleagues in their offices, and contacts stemming
from the use of the same laboratory instruments. But in any case, debate is
extraordinarily laborious and protracted, as in many instances up to two
years are needed to decide on particularly crucial or sensitive issues. The final
outcome of this tortuous process is almost always a compromise which,
although denying full control to either research director or scientist, is,
nevertheless, sufficiently amenable to both parties to allow the perpetuation
of a collegial style of leadership.
The hierarchy of these laboratories displays a dual character as the organizational framework vacillates between unity and relative disaggregation.
The hierarchy contains four strata: research director, senior scientist, junior
scientist and technician - the technical staff standing in a ratio of 1 : 1 with
the scientists (see Figure 5). But the most telling aspect of this hierarchy is
the acuteness of its internal stratification. Although the research director and
senior scientists struggle unceasingly and bitterly with one another over the
question of authority, they are quick to join forces when dealing with junior
scientists and technical personnel. The upper echelons of the organization
thus ally in order to guarantee their distinct identity and hegemony, but also
to ensure the inviolability of the research strategies and methodologies so

Scientific Disciplines and Organizational Specificity

I
I
I

249

,
I

,
,,
I

I
I

I
I

,,

I
I

,
\

Fig. 5. Physics laboratories.

painfully formulated. For their part, junior scientists and technicians similarly
coalesce as they strive to affect or penetrate the
of decision-making. Indeed,
attacks on the bastion of power rarely prove felicitous.
The lower echelons seldom succeed in reorienting laboratory planning or
operations. This apparent impasse is partially mitigated, however, as a fairly
sizeable group of junior scientists and technicians climb into the upper echelons and thereby become full-fledged participants in policy determination.
Almost one junior scientist in three rises to the post of senior scientist or
research director as does one technician in six.
The communication network mirrors the compartmentalized dimension of
this hierarchic arrangement. The network consists of two semi-hermetic
conjoint communication circuits; the one composed of senior scientists and
a research director, the other of junior scientists and technicians. Within each
of the circuits, contact is irlcessant and personal, and interactions take both a

250

Terry Shinn

horizontal and vertical path. But as already reported, while in the subordinate
echelons, the spirit of solidarity prevails and colours discourse between
individuals, the bias of exchanges in the upper echelons is far more erratic and
problematic. Despite this dual organizational configuration, a substantial
body of information nevertheless does move between the two circuits. This
generally transpires during the process of daily work, and the agents responsible for the flow are primarily senior and junior scientists whose research tasks
inevitably bring them together on a regular basis. Of no less importance in
these physics laboratories, a balance is struck between written and oral
exchange. Interactions inside each circuit are largely oral, as are most of the
contact between the two categories of scientists. But perhaps in an effort to
neutralize informal exchange, research directors make certain that written
memoranda represent an alternative communications channel, and one of no
little importance at that.
In these physics research units, intellectual tasks tend to be both narrowly
defined and predesigned. Each category of personnel is responsible for a
specific mental operation. But in addition to this, there exists a parallel
covert structure where certain areas of research activity escape the strictures
of official rules. Here, different members of the staff address given elements
of the research project in terms of the programme's inherent logic, rather
than as a function of their formal position in the laboratory hierarchy.
Theory-related questions are without exception coopted by research
directors and senior scientists, as are the bulk of matters associated with
scientific strategy and methodology (see Figure 6). Together, these two
groups constitute a clique which excludes other laboratory agents from
abstracting activities. This interdiction is as unwaveringly upheld at the close
of research projects as at their genesis. Thus, a restricted group takes full
charge of the entirety of conceptual synthesis and analysis. At the opposite
end of the spectrum, junior scientists and technicians are called on to design
and set up experiments, to conduct them, and to present the resulting observations in an orderly and comprehensible manner. Put in another way,
they are responsible for the experimental apparatus, its effective operation,
and for the quality and codification of experimental results.
These intractable cognitive matrices do not, however, seem to provide a
wholly adequate intellectual framework for this kind of phYSics research. The
distance separating conceptualization on the one hand, and experimentation

Scientific Disciplines and Organizational Specificity

251

conceprual
th.ory

Fig. 6. Physics research.

on the other, is seemingly too vast. The mechanism required to link these
two elements frequently proves inadequate. In fact, this linking mechanism
introduces two associated processes: (a) the implementation of extraordinarily precise experimentation. This necessitates scrupulous and unflagging
procedural care as well as more than a modicum of insight into the phenomena
under investigation; (b) the conversion of frequently esoteric and disjointed
observational commentary into significant data. These twin sensitive intellectual operations are dependent on advanced familiarity with both technique
and theory, and the required mix of the two components is shaped by the
exact parameters contained in the research programme. This aspect of mental
activity thus cannot be preprogrammed and in these research units informal
and makeshift arrangements are consequently elaborated so as to neutralize
the shortcomings of the officially advocated cognitive organization.
To summarize the main argument up to this juncture: scientific research
activity gives rise to a range of social and cognitive organizational configurations. Moreover, cognitive patterning and social dispositions are symmetric.
In each one of the three cases described above, the internal structure of the

252

Terry Shinn

cognitive and social arrangements match. Now, borrowing the vocabulary of

organizational sociologists, it becomes possible to represent at least some
scientific operations in terms of classical organizational components:
(1) The setting of the mineral chemistry laboratories presented in this
examination, may be fitted into the organizational system which Burns and
Stalker label the "mechanistic model" (6): that is, organizations characterized
by a single powerful source of command, complex and petrified hierarchy,
circumscribed communication flow and a stringent division and precodification of all work activity.
(2) Again employing the classificatory schema of Burns and Stalker, solidstate physics laboratories exhibit many of the traits associated with the
"organic model" of organization. This includes features like limited pluralistic
leadership based on negotiation, a moderately constraining hierarchy made
more supple by the introduction of some unplanned dynamic elements, the
polarization of communications around several distinct yet juxtaposed nuclei,
and a dichotomy in the work process whereby certain tasks are allocated
while others are undertaken in accordance with the exigencies of the specific
project.
(3) Interestingly, sociologists of organizations have not thus far forwarded
a model which subsumes the characteristics common to the type of research
centers engaged in vector analysis; this perhaps because they have tended to
concentrate on the study of enterprise, and particularly on production
departments. Be this as it may, such a model could well prove useful for
future investigations of both scientific and non-scientific systems. In this
event the paradigms found in computer laboratories are highly instructive as
they are palpably dissimilar to those of other organizational types: a fully
participatory decision-making apparatus or, stated alternatively, an absence of
assertive authority; an uncomplicated and minimal hierarchy which attenuates
internal distinctions; multi-directional and plentiful communication channels;
almost total non-differentiation of tasks as work is freely chosen and changes
in time. Henceforth, this organizational configuration will be referred to as
the "permeable model".
The motive for presenting this organizational terminology, however, is
to advance still another procrustean formulation for the unfolding of
scientific research. Indeed not. It is merely intended to furnish an admittedly
primitive conceptual tool which, it is hoped, will facilitate the following

Scientific Disciplines and Organizational Specificity

253

discussion of some factors that infuse different domains of scientific enterprise with their specificity of practice and structure.
2. Thoughts on Causality

Inevitably the sum of factors contributing to the social and cognitive morphology of scientific activity is incalculable. Certainly, however, elements like
public policy, the availability of funding, and the formulation of priorities
by highly sophisticated research institutes have, in recent years, played an
important role in shaping scientific processes; and the dreams, ambitions and
eccentricities of individual investigators have, always, been central to the
mould of research operations. Yet while determinants such as these undeniably merit much additional attention, this analysis of organizational causality
focuses on two alternative agents. First, social disposition will be assessed in
the light of the imperatives and inhibitions engendered by specific experimental procedures and by the use of certain kinds of experimental apparatus.
In short, this hypothesis posits that the logic contained in experimental
operations, and the manipulation of instruments, impinges significantly on
the cast of social relations. Secondly, cognitive patterning will be examined
in terms of the various components which constitute the epistemic structures
of different sub-disciplines. The key idea is that distinct areas of research are
built around a specific mix of thought categories which delimit the contours
of cognitive organization.
To take the case of the chemistry laboratories described above: the correspondence between mineral chemistry research and the mechanistic organizational model stems from both distinctive experimental practices and mental
processes which, in their turn, are contingent on the category of matter and
the particularity of phenomena under investigation.
Conventionally, mineral chemistry deals exclusively with a narrowly
defined class of objects and their. most immediately ascertainable manifestations; or, as expressed by the historian and philosopher of chemistry, F. A.
Paneth, it directs attention to "elements" and "properties" (7). The class of
matter examined in this instance is mineral, or mineral-associated, and the
goal of the researcher is to identify, claSSify and qualify such substances
and to understand their metamorphosis when in mixture or solution with
analogous or dissimilar chemical agents. Mineral chemists are often taken up

254

Terry Shinn

with the study of mineral-based compounds and alloys. In addition, this subdiscipline is preoccupied with the physical character of mineral substances,
for example, their colour, transluscence, texture, density, elasticity and so
on. This is achieved either through direct unmediated visual observation or
through the use of a battery of rather traditional laboratory equipment which
ranges from interminable numbers of beakers, test-tubes, balances, etc. to
somewhat more elaborate devices for gauging light defraction or mechanical
distortion. In a word, the area of inquiry referred to as mineral chemistry
tends to concentrate on the most palpable and directly measurable features of
a given group of elemental substances.
The unavoidable reliance of this form of research on extensive versus
intensive laboratory bench work and on an abundance of germane chemical
substances, tends in itself to dictate the bounds of what might be regarded as
the only genuinely feasible social framework for its effective operation. As
inquiry in this domain necessitates the simultaneous deployment of multiple
experimental parameters, and the innumerable repetition of each type of
experiment, a relative abundance of subordinate personnel is required to
prepare and conduct experiments, and to record observations (8). Furthermore, since experimental procedures are normally straight-forward, and
the nature of observation is generally unproblematic, the educational and
conceptual qualification of the technicians and laboratory assistants, who
carry out the major part of this bench work, is elementary. This experimentation and observation necessitates a large-scale fIltering and arrangement of
fmdings which frequently prompts supplementary bench work in order to
control and validate recalcitrant items. Such verification is always carried out
directly by the better trained and more learned intermediate-level scientific
personnel. The work of the scientist is thus not as mechanistic as that of the
technical staff, although it is often painfully time-consuming and arduous (9).
It is also essential to point out that in this type of laboratory, logistics-related
questions also have their importance. The need to assure an uninterrupted
and massive flow of appropriate chemical agents to the laboratory, and to
guarantee an adequate stock of experimental instruments, becomes a monumental administrative task when demand is urgent and requirements diverse.
In the same manner, because of the simultaneity of a host of experimental
paradigms, and the presence of several strata of personnel within these
research centres, the task of liaison sometimes takes on looming proportions.

Scientific Disciplines and Organizational Specificity

255

From this, it should be clear that the sheer volume of experimental practices,
the tedium of translating observation material into sound data and the need
for an assertive brand of authority, go far to establish the type of centralized,
restrictive and formal social relations found in chemistry research units (10).
like the social configuration, the epistemic. structure, and hence the
cognitive organization of mineral chemistry laboratories, is also powerfully
influenced by the importance and specificity of observation data. The fundamental cognitive aim of this research is to identify elementary substances and
to establish categories of permanent relationships. Several mental operations
enter into this process: (a) the perception of bits of information which are
held to represent defmite relationships between clear-cut parameters; (b)
the instantaneous selection of pertinent elements of information from the
total observational universe; (c) the transcription of observations into an
accurate and communicable set of signs and the assembly of these signs into a
manageable format, (d) the analysis of observation data and the extraction
of significance. Of course, the initial suppositions underlying the research
programme are founded on indirect knowledge of previous work, or on
immediately preceding experimentation. The process of research is thus
circular in character; and this is particularly true in mineral chemistry which
is highly dependent on the veracity of anterior conceptual canons.
It seems from this that while theory is just as crucial for achievement in
mineral chemistry as it is in other scientific fields (11), the place which it
occupies is nevertheless specific, and different from that of many other
sub-disciplines (12). In mineral chemistry the prinCiple of "reductionist
development" constitutes an important methodological tenet. This principle
stipulates that a maximum of observations must be encapsulated in a minimum
of theoretical axioms (13). This orientation is further strengthened by the
principle of "theoretical clusters" which submits that a given zone of inquiry
retains validity if the bulk of the field's axioms are revealed as adequa}e even
though a number of its fundamental propositions have been shown to be
invalid (14). In effect, the application of the reductionist and cluster tenets
radically attenuates and sometimes altogether obviates, the impulsion for
fresh theory, for the active development of new conceptual canon is frequently seen as inconsistent with the epistemic thrust of the sub-diScipline.
The dominant mental operation in these research units is instead bound
to analogy construction. This exercise gives focus to the myriad bits of

256

Terry Shinn

information accumulated during experiments, and afterwards converted into

data. Analogy construction uses the body of extant static theory as the basis
for evolving new arrangements and images of matter. In the course of the
process, models are frequently developed. They are generally iconic models
which offer a visual representation of phenomena rather than the abstracting
variety (15). This form of strict analogy construction proves so important to
the sub-discipline because it provides the best approach for the qualification
and ordering of mechanistic-state conditions; that is, of highly concrete,
measurement-prone passive agents. It is, in fact, this heavy reliance on analogy
construction, and concommitantly on iconic models, which is the principal
cause for an acute division oflabour; for, as has already been remarked, a vast
amount of rudimentary work is first required before it is possible for a single
well-informed individual to undertake the necessary correlating operations
requisite to obtain the best fit between the body of data and current theory.
Thus, the cognitive pattern, as the social one, is characterized by centralization, differentiation of tasks and rigid canalization of thought and discourse.
Now turning to computer laboratories, and more particularly to computer
research centres involved in vector analysis: their research processes as well
as cognitive orientation, stand in sharp opposition to those of chemistry
research units, this because the entirety of research activity is centred on pure
formalization. It is wholly abstract; in its essential form a numerical logic,
a symbolic calculus. All operations are thereby connected with numerical
arrangement and juxtaposition which are derived from a series of prestated
formulae of symbolic suppositions. The object of research is to advance
adjunct formulae and values, coherent with the initial presuppositions and
which are extensive in that they can be satisfactorily fitted to a range of
analogous symbolic situations. This sort of investigation generally makes use
of computers to examine the transitory dimensions of mechanistic-state
phenomena in transition, where the aim is to describe mathematically or as a
calculus, either total or constituent vector characteristics. Such an analysis is
sometimes used to examine wholly imaginary conundrums (for example,
point displacement in phase or Hilbert space) or to explore problems like
drop dynamics. In all instances, however, the setting is entirely abstract,
and the sole task is to establish a set of mathematical values or to identify
symbolic relationships (16).
Research practices here entail four operations: (a) definition of problem,

Scientific Disciplines and Organizational Specificity

257

selection of methodology and development of appropriate symbolic calculi;

(b) formulation of program and translation of data into the program; (c)
computerized treatment of information; (d) analysis and interpretation of
computer output. Steps (a), (d), and to a great degree (b), are carried out in
the complete absence of scientific instrumentation. The only implements
necessary to undertake this work are a good supply of pencil and paper, and
a small number of books and journals. Even that segment of phase (b), which
calls for transposition of data into the established program, necessitates little
equipment as it requires a bare minimum of uncomplicated software. Only
phase (c) of research activity, then, is associated with scientific apparatus. Yet
in this case the instrumentation is fully automatic and consequently requires
no intervention from laboratory personnel.
Research practices in these laboratories thus escape the shackling effects
imposed by the pervasive presence of cumbersome instrumentation and
quantities of material substances. Inquiry is instead intimately coupled to
individual and group reflection; The work may be spoken of as cerebralintensive. Moreover, it is above all the cerebral nature of this activity, which
elicits a fluid system of social relations. Since research success does not
require personnel to assume either highly technical or automaton-like manipulatory tasks, and as the very nature of the symbolic problems at hand
invite breadth of mental input and participation, it becomes quite possible,
and even affIrmative, to reduce social constraints to a low level; hence, an
attenuation of authority and hierarchy, and the mushrooming of free and
structureless exchange.
The permeable organizational model similarly embraces the cognitive
pattern of computer research centres due to the preponderance of abstraction
in the sub-discipline. In this field of investigation the spirit is absolutely
divested of immediately pertinent sensory data. Activities are disembodied;
language, as commonly intended, is emptied of meaning and hence the need
to create a substitute or meta-language (17). The mind is thereby set to
elaborate or extend uncustomary and original sense systems which are based
on correspondences such as series, discontinuity, analogues, constants and
transitory constants.
The epistemic structure of computerized analysis largely marginalizes
theory, at least as conceived of in the physical sciences where theory is irrevocably attached to immediately or hypothetically identifiable phenomena.

258

Terry Shinn

The fundamental epistemic components of this sub-discipline consist of

definitions, axioms and deductive theorems. Thus, thought tends to proceed
from general abstract categories to particular idiosyncratic cases. Here, the
role of analogy is limited and, though occasionally brought into play, it
is inclined to introduce overly mechanistic and direct dispositions which
simplify and falsify understanding. Model construction, on the other hand,
offers an indispensable epistemic element. Naturally, iconic models are discarded as image portrayal is incidental, and more often than not, altogether
impossible. Rather, the models are formal, expressing juxtaposition and state
through wholly abstract calculi.
In this sub-discipline the epistemic structure is consequently formed
around quantitative and symbolic relationships and combinations. Such
mental activity is inconceivable in a context of disjunctive units; that is,
in terms of discreet disconnected elements (18). The epistemic structure
requires that issues be regarded and treated as wholes, as ensembles which,
although multi-faceted, must be approached as an indivisible research program. Under these conditions, cognitive differentiation can only provoke
the disaggregation of the problems' essential wholeness; and although such
a disaggregation might result in a series of internally coherent statements,
and perhaps even in interconnected calculi, it would undoubtedly spell the
end of the integrity of the original research problematic. These episternic
components thereby require a highly informal, supple, and non-regulatory
cognitive framework where symbolic combinations can easily be realized.
lastly, the organic organizational model common to the laboratories
engaging in solid-state physics research is consistent with the hybrid (19)
character of the sub-discipline's experimental and epistemic orientation;
experimental practices which involve small quantities of material agents, but
necessitate unrelenting observational precision, and an epistemic thrust which
integrates both theoretical and observational constituents and which requires
'
an intricate complex of intermediary mental operations.
These research units principally explore the electromagnetic qualities of
elements. Substances, like germanium and silicon, are studied in order to
acquire a more thorough-going understanding of the correspondence between
matter on the one hand, and energy discharge and receptivity on the other.
Such studies treat the electron donor and receptivity potential of substances
as well as their capacity to withstand heat, conduct power, and in particular

Scientific Disciplines and Organizational Specificity

259

instances, to ensure the fIxity of characteristics during extreme vacillations of

exogenous conditions. Some of this research has given rise to inventions such
as the transistor, tunnel diode, and most recently, the chip. Yet in addition,
solid-6tate physics also focuses on the fundamental relationship between
matter and its transformation by processes like particle loss and energy
dissipation. In this manner, these laboratories bridge the gap separating
electromagnetic events from low-energy physics. Hence, in this sub-discipline,
research centres are not involved exclusively with material substances, but
also with questions related to force phenomena.
Experimental activity is of two orders: the fIrst consists of testing the
electrochemical effects of combining one class of elements with other substances such as arsenic, potassium, etc. In this process, microscopic amounts
of material are transposed onto other substances through the use of precision
instruments or by photographic methods. While relatively unsophisticated
parameters are controlled, the fInite values involved necessitate the utmost
care in each experimental practice (20). Accuracy is guaranteed through
experimental repetition. Thus, even though this phase of experimentation
is sometimes routine, it cannot reasonably be conducted in a mechanical
fashion by only moderately-trained personnel, since success calls for an
advanced understanding of the phenomena under observation, and a mastery
of laboratory techniques.
In a sense, the second experimental operation relies less on material
substances than the first, but in contrast, depends even more heavily on
precision instrumentation and observational accuracy. In this instance,
investigation concerns molecular and atomic motion within a given category
of matter. Atomic trajectories and interconnections are examined in terms of
fragility or permanence of pattern. Ancillary research consists of identifying
the factors that destabilize or consolidate the arrangements. The equipment
employed for this work is incomparably elaborate and delicate, and therefore
diffIcult to handle. This is highlighted by the fact that results are frequently
so gossamer-like that the event being observed may be influenced by the
intervention of the scientist carrying out the experiment, thereby modifying
the sense contained within the ensuing data.
One feature particularly emerges from these experimental practices:
the unmitigated need for observational and theoretical unity. In this subdiscipline, experimentation cannot be pursued by individuals unfamiliar with

260

Terry Shinn

the exact nature of the experimental problems and with the details of observational expectations. Indeed, although problems like the notion of "non-separability" and the Heisenberg indeterminacy principle are rarely voiced, they
nevertheless do represent very real methodological and practical obstacles
to sound research. For this sort of reason, the social organization of these
research centres must, by all means, institutionalize mechanisms which
oblige senior scientists to collaborate closely both with junior scientists and
technicians, and furthermore, which force them to enter into the experimental
operations themselves. This trait constitutes the essential dimension of the
organic organizational model which ensures a high degree of integration, but
within the confines of functional specificity.
As the terrain of investigation of solid-state physics spans a variety of
physical phenomena and forces, research activity, ipso facto, incorporates
a broad scope of mental processes, and hence episternic structures. Theory is
assuredly central to the growth of knowledge in this sub-discipline, as is the
amassing of refmed and detailed observational particularities; and between
these two operations, there are situated equally crucial thought categories like
retroductive hypothesis formulation and metaphor-based model construction
(21).
Within this field, narrowly declarative descriptive statements, expressing
mechanistic elements of observables, are granted some legitimacy. Acceptance
of this type of statement is, however, sometimes illusory, as its long-term
viability is strictly contingent on integration into a broader context whereby
it either adds to, or provides a basis for, comprehensive explanatory formulations. Thus, while declarative descriptive statements do not in themselves
constitute advancement, they are regarded as satisfactory if they eventually
contribute to a global representation of phenomena (22). Theory development
is similarly precarious, and theory lacks the sacrosanct status which it enjoys
in many other sub-disciplines. On the one hand theory is patently transient; it
is routine to see sets of ideas brutally forced aside by competing notions. On
the other, it is remarkably stable since seemingly faltering theories are often
retained through fine adjustments and reconstrual (23). In solid-state physics,
anomaly tends to figure importantly in the process of theory construction
and in the credibility accorded to a given theory. Depending on specific
circumstances, some anomalies are viewed as sufficiently damning to lead to
the outright suppression of a theory. Under other conditions, the presence of

Scientific Disciplines and Organizational Specificity

261

anomalies is instead simply regarded as an occasion to reinforce theory by

altering it in a manner so as to take into account the salient observational or
conceptual objections.
The problems posed by anomalies are regularly resolved in one of two
fashions: first, it is frequently possible to salvage a threatened theoretical
framework by recasting one or several of the models contained in the theory.
Usually, this process of model transformation involves the use of metaphor
rather than analogy. Metaphor has proven especially fruitful in this subdiscipline because it allows the maintenance of a direct link with the initial
theory while, at the same time, permitting substantive associational diversions
through the inclusion of a body of fresh material, be it iconic or purely
formal. Secondly, in solid-state physics, theories are also tempered, and new
ones created, by means of retroductive reasoning (24). This approach
resembles a feed-back circuit. The intellect follows a relatively complicated
trajectory: it begins with the contemplation of orthodox learning which is
used to construct a probabilistic zone of analysis and causality, then takes
up the matter of methodology, its application and observation, and lastly
returns to the original point of departure equipped with relevant data. While
this mental practice is overtly circular, it nevertheless provides numerous
opportunities to interrupt or redirect the original conceptual orientation.
In brief, the epistemic structures common to this form of physics research
are notably diverse. Well-established mental categories coexist with more
recent thought modes, and their frictionless encounter not infrequently
forges the basis for innovation. The type of organizational matrix required
by such an epistemic structure must, then, both supply a framework for
traditional intellectual procedures and give free rein to untested mental
orientations. This configuration coincides with the thrust of the organic
model which simultaneously channels customary activity and acknowledges
the emergence of optional processes.
This study, on a micro-scale, and on a macro-level, the work of Fleck, Rip
and Yoxen, have gone some way in substantiating the general point made by
Norbert Elias in the theme paper of this volume: that scientific establishments
are constituted around a highly complex and multi-faceted composit of institutions and practices where distinctive sets of characteristics are associated
with particular areas of scientific activity. This point, although transparent, is

262

Terry Shinn

nevertheless fundamental in its implications, and is today unfortunately often

overlooked by the reductionist movement in the sociology of science (25). It
must be conceded, however, that the in-depth appraisals of molecular biology,
artificial intelligence and chemistry do reveal a tendency toward reductionist
disciplinary evolution. Yet in each of these fields, as well as the one examined
in this essay, elements of social and cognitive specificity persist, and are
moreover sufficiently telling to differentiate separate sub-disciplines. Hence,
albeit there is some cause to believe that shifts in the organizational morphology of subdisciplines represent a primary feature of scientific establishments,
that fact does not suppress the pattern of scientific mUltiplicity as outlined
here, nor does it redeem reductionist thought which implicitly posits the
uniformity of the scientific enterprise.
Finally, while this contribution has sought to call attention to social and
cognitive patterns of various scientific fields, and has particularly suggested
epistemic factors and scientific instrumentation as determinants of such
patterning, it is also important to indicate that these two causal agents do not
constitute a complete explanatory schema (26). A more thorough analysis of
intra-laboratory configurations certainly necessitates the integration of a
larger number of variables designed to introduce additional social elements
and to refme epistemic considerations. It is possible that the adoption of a
five-prong approach would, at least in part, satisfy these requirements where
questions of (a) social Conditioning, (b) scientific strategies, (c) cognitive
categories, (d) specificity of problematic and (e) scientific instrumentation
would each be assessed separately and in conjunction with one another (27).
Indeed, the utilization of this approach might well afford a clearer vision
of the social and epistemic dimensions of research processes, and thereby
generate a better understanding of the entire scientific enterprise.
Notes and References
1. This analysis of science intentionally eliminates considerations like ''restrictedness'',
"configuration" and "unrestrictedness" which, although evocative, nevertheless
tend to mask some of the particularistic dimensions of the scientific enterprise.
Among the most interesting and useful studies of restricted and configurational
science are the works of Richard Whitley, 'Changes in the Social and Intellectual
Organization of the Sciences: Professionalization and the Arithmetic Ideal', in
Everett Mendelsohn, Peter Weingart and Richard Whitley (eds.), The Social

Scientific Disciplines and Organizational Specificity

2.
3.

4.

5.

6.
7.
8.

9.

10.
11.
12.

13.
14.

263

Production of Scientific Knowledge, D. Reidel, Dordrecht, 1977, pp. 143-69; see

also Norbert Elias, 'The Sciences: Towards a Theory" in R. D. Whitley (ed.) , Social
Processes of Scientific Development, Routledge and Kegan Paul, London, 1974;
and C. F. A. Pantin, The Relations Between the Sciences, Cambridge University
Press, Cambridge, 1968, Chapter 1.
In this text the terms "discipline", "subdiscipline", "field", and "area of inquiry"
are used interchangeably.
In addition to laboratories in mineral chemistry, computerized vector analysis and
physics, the research project on which this study is based also gathered
information on a small number of laboratories working in the fields of organic
chemistry and mechanics.
The observations presented in Part I of this paper were made in 1977. During a
ten-month period, thirteen research laboratories (ten in industry and three in the
university setting) were examined in some depth. Biographical, education and
professional data was collected for a population of 517 researchers and technicians:
98 research directors, research scientists and technicians were subsequently interviewed, the interviews being about two hours in length. The total amount of time
spent in each laboratory ranged from ten to fourteen days.
For information on the organization of another category of physics research - in
this instance high energy physics - see Gerald M. Swatez, 'The Social Organization
of a University Laboratory', Minerva 8 (1970) 36-58; see also A. Bitz, Andrew
McAlpine and R. D. Whitley, The Production and Flow and Use of Information in
Different Sciences, The British Library Report Series, London, 1975.
See Tom Burns and George Macpherson Stalker, The Management of Innovation,
Tavistock, London, 1966 (3rd impr. 1971).
See F. A. Paneth, 'The Epistomological Status of the Chemical Concept of Element',
The British Journal for the Philosophy of Science 8 (1962), Part I, pp. 1-14;
Part II, pp. 144-60.
Ravetz's analysis of "craft and skills" - in particular their importance in descriptive
sciences - is applicable to the situation in these laboratories of mineral chemistry.
See J. R. Ravetz, Scientific Knowledge and its Social Problems, Clarendon Press,
Oxford, 1971, pp. 76-94, particularly pp. 91 and 92.
See E. F. Caldin, The Structure of Chemistry in Relation to the Philosophy of
Science, Sheed and Ward, London and New York, 1961.
Ibid.
See D. K. C. MacDonald, 'Physics and Chemistry: Comments on Caldin's View
of Chemistry', The British Journal for the Philosophy of Science 2 (1960/62)
222-3.
Toulmin underlines the importance of the continuity of traditional scientific
conceptualization. This continuity is particularly present in the case of the mineral
chemistry laboratories. See Stephen E. Toulmin, Human Understanding, Vol. I,
Clarendon Press, Oxford, 1972.
See F. Suppe, The Structure of Scientific Theories, University of Illinois Press,
Chicago, London, 1974, p. 58.
See Hilary Putnam, 'What Theories are Not', in Nagel, Suppes and Tarski (eds.),
International Congress for Logic, Methodology and Philosophy of Science, Vol. 1,
Stanford, Stanford University Press, 1962.

264

Terry Shinn

15. See Max Black, Models and Metaphors: Studies in Languages and Philosophy,
Cornell University Press, Ithaca, New York, 1962, chapters 3 and 13, particularly
pages 220-6; also see Mary Hesse, Models and Analogies in Science, Sheed and
Ward, London and New York, 1963, pp. 30-54; 65-75; 84-98; see also Hesse,
The Structure of Scientific Inference, Macmillall, London, 1974; see also Harre,
The Principles of Scientific Thinking, Macmillan, London, 1970, chapter 2.
16. See A. Sloman, The Computer Revolution in Philosophy, Harvester Press, Hassocks,
1978.
17. See Benson Mates, Elementary Logic, Oxford University Press, New York, 1965;
see also Ian Mueller, 'Euclid's Elements and Axiomatic Method', The British Journal
for the Philosophy of Science 20 (1969) 229-308.
18. It is unclear as to whether Sloman agrees with this analysis; at certain points he
seems to suggest the highly integrated nature of mathematical operations, while at
others he apparently posits a more disaggregated approach. See Sloman, op. cit.
19. Here the term "hybrid" is used as a metaphor and does not allude to Ben-David's
concept of "hybridization". See J. Ben-David and R. Collins 'Social Factors in the
Origins of a New Science: the Case of Psychology', American Sociological Review
31 (1966) 451-65.
20. For a discussion of "tool experts" see Ravetz, op. cit., pp. 90-1.
21. See Black, op. cit.; see also N. R. Hanson, 'The Logic of Discovery', Journal of
Philosophy S5 (1958) 1073-89; 'More on the Logic of Discovery', Journal of
Philosophy S7 (1960) 182-8; Patterns of Discovery, Cambridge University Press,
Cambridge, 1958.
22. For an example of this process, see Nicholas Georgescu-Roegen, The Entropy Law
and the Economic Process, Harvard University Press, Cambridge, Mass. 1971,
chapters Sand 6.
23. For an analysis of several theory types'see A. Rappaport, 'Various Meanings of
"Theory" ',American Political Science Review S2 (1958) 927-88.
24. Hanson, 'Logic and Discovery', 'More on the Logic of Discovery', Patterns . ..
op. cit.
25. For an evocative, albeit fumbling reductionist approach in the analysis of science,
see B. Latour and S. Woolgar, Laboratory Life. the Social Construction of Scientific
Facts, Sage Publications, London, 1979.
26. In another article when treats the organizational specificity of research laboratorjes,
Shinn utilizes the social variables of education as an explanatory device. See T.
Shinn, 'Division du Savoir et Specificite organisationelle des laboratoires de recherche industrielle en France', Revue Franfaise de Sociologie 21 (1980) 3-35.
27. A new study, just under way, employs these five analytical variables. See T. Shinn,
'Transparence et opacire scientifique', project outline for a research project for the
DGRST-CNRS programme science technologie societe.

PART III

ESTABLISHING BOUNDARiES AND HIERARCHIES IN THE

SCIENCES

ON THE AUTONOMY OF PURE SCIENCE: THE

CONSTRUCTION AND MAINTENANCE OF BARRIERS
BETWEEN SCIENTIFIC ESTABLISHMENTS AND POPULAR
CULTURE

R.G.A.DOLBY
University of Kent at Canterbury

1. Introduction
Modem science has grown into a large-scale and complex activity. In an age
of increasing bureaucracy and specialization, the knowledge construction
industry is also bureaucratized and segmented. The intellectual divisions
within science and between science and comparable activities, now hallowed
by tradition, have become locked into the institutions of science, to be
exploited by the relevant interest groups. Cognitive boundaries have thus
been turned into social barriers. When these barriers were being built up, they
could be justified in terms of a convergence of the cognitive requirements of
efficient knowledge production and the social advantage they conferred upon
insiders. The arguments can conveniently be encapsulated in terms of the
notion of expertise. In pure science, the more expert an individual is on
a topic, the more he can be trusted as an authority. Expertise is acquired
through specialist training, association with other experts and by making
recognized contributions to knowledge relevant to the topic in question.
The institutionalization of knowledge advancement particularly involves
organizing the activity of individuals high in hierarchies of expertise. Any
contribution to knowledge is ideally directed to those who can best appreciate
it, that is, to those who are most expert. And contributions from those who
are most expert are the most readily appreciated. Informal communication
networks naturally emerge among the elite of science; more formal institutions
try to limit themselves to the higher part of hierarchies of expertise by setting
267
Norbert Elias, Herminio Martins and Richard Whitley (ed,.j,Scientijic Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 267-292.
Copyright 1982 by D. Reidel Publishing Company.

268

R. G. A. Dolby

minimum standards for participants in their activities. Any individual who

fails to reach the minimum standards set by some scientific organization finds
considerable barriers obstructing his participation in the activities of the
science. His exclusion allows accredited scientists to interact more effectively.
Similar processes operate at the divisions between sciences, as claims of
expertise tend to be limited to specific subject areas and skills. Thus, the
institutionalization of separate sciences turns the separation between the
hierarchies of expertise of each science into social barriers. As a consequence,
the possibilities for overlap in the hierarchies of expertise are limited still
further, and the segmentation of science is increased in a spiral of positive
reinforcement.
The tendency of intellectual barriers within and around science to cumulate and to become harder to cross has inevitably come to be recognized as
a problem. It can be difficult to establish new and promising forms of inquiry
which would cross previous disciplinary boundaries. We cannot, therefore,
rest content with the time-honoured modes of justification of the occupants
of secure scientific establishments, for they obscure objective examination
of the new problems. Since the 1960s, there has been increased suspicion of
the rationale of expertise, particularly when the authority of professional
pure science is used as a platform for offering advice in public forums. The
role of the expert who claims to offer technically compelling but value-neutral
advice has come into disrepute over such politicized issues as nuclear reactor
safety (1). In such a context, if the scientific expert claims professional
standing, he is likely to be regarded merely as attempting to increase the
authority of his pronouncements. Furthermore, a critic may be unimpressed
by the claim of professional status, and regard professions as no more than
monopolies which control the practice of their members in order to protect
their collective privileges.
Our discussion so far has assumed that intellectual barriers are constructed
by those in a position of power in order to protect themselves and to maintain
control of their power. Those inside such a closed barrier form a scientific
establishment, and the justificatory rhetoric of the professional expert
functions as insider ideology. In the theme paper for this volume, Nobert
Elias offers us a contrary ideology - that of the outsider. In his identification
with a low status discipline, sociology, Elias feels keenly the power exercised
by the more prestigious academic establishments of philosophy and physics.

On the Autonomy of Pure Science

269

As custodians of knowledge, they control a means of orientation for individuals in our society and, in practice, put pressure on sociologists to conform
to their methods and standards of production of knowledge. Elias's paper
illustrates the perceptions of the articulate outsider, and offers a revealing
repertoire of the strate gems of the heterodox figure in attacking the social
institutions of the orthodoxy. However, it is not obvious that we should
simple replace the ideology of the insider by that of the outsider. And we
should certainly not take Elias's historical claims at face value (2).
Any exploration of the barriers around science must cross territory scarred
by a grand issue: that of the autonomy of pure science. Although now
increasingly under attack, the doctrine of the autonomy of pure science
is still well entrenched in science and in the disciplines studying science.
Apologetics for science of the mid-century put the traditional view in a sharp
and now familiar form: pure science has its own methods and standards of
knowledge advancement, and any external social interference in science will
either be resisted or will lead to a degenerate form. Nazi rejection of Jewish
science and Stalinist Russian rejection of bourgeois capitalist science are
therefore to be seen as unacceptable state interference; the success of Lysenko
in promoting his eccentric agricultural ideas by relating them to official state
ideology is to be seen as producing mere pseudoscience. That is how the
autonomy of pure science was defended in the time of totalitarian regimes
in Germany and Russia; echoes of the same rhetoric can still be heard (3).
The issue of the autonomy of science and in particular of the production
of scientific knowledge is complex and multifaceted. Empirical or theoretical
study of the processes of boundary construction around science will inevitably
be coloured by prior commitments to its disciplinary manifestations. Where
some see the rational institutionalization of objective features of science,
others see the ideological workings of professional self-interest.
My purpose in reviewing these difficulties of coming to a satisfactory
understanding of scientific establishments and the barriers that surround
them is to suggest that the issues are so blurred that those who claim to
provide clear and practical answers are likely merely to be attempting to
legitimate their own interests. This paper is limited to seeking to enrich our
background knowledge of the construction of intellectual barriers and to
finding ways in which methods of learning about such issues can become
more decisive. The intellectual barrier I will examine in most detail is that

270

R. G. A. Dolby

between established forms of pure science and quasi-scientific activities in the

popular culture, and I will discuss examples which fall on a major border
within science, that between the biological and human sciences. Successive
efforts to increase or decrease the permeability of this internal boundary will
therefore recur in the examples discussed.
It will be helpful at this stage to illustrate the kind of explicit argument for
barrier construction which will repeatedly recur in the historical examples. In
its most general terms, consider two competing forms of intellectual activity
which are not sharply distinguished by an audience whose favours both sides
desire. An actor on one side may seek to establish 'intellectual distance'
between the two activities. That is, he will try to show that the two activities
are not at all alike by sensitizing the audience to the distinguishing features.
Such a strategy is appropriate if an actor can exploit or establish a favourable
hierarchical order of authority between the two activities. A familiar example
(to which we will return) is provided by the famous exchange between T. H.
Huxley and Bishop WJ1berforce over the Darwinian theory of evolution at the
1860 Oxford meeting of the British Association for the Advancement of
Science (4). Huxley sought to distinguish expert discussion of evolution as
a biological theory from popular discussion of evolution as a challenge to
accepted religiOUS views. He sought to show that although Darwin's theory
might have religious implications, it must first be evaluated as a contribution
to biology, and in this decision, those, like Bishop WJ1berforce, who could
not show the relevant expertise, revealed themselves as incompetent judges.
The process of creating 'intellectual distance' is an important one in intellectual controversy. For conflicts in general tend to escalate, to draw upon
ever wider divisive issues as each party seeks to gain advantage by attracting
bystanders to its side. Intellectual conflicts show the same tendency. But
within science, there is an institutionalized tradition that scientific issues
can most readily be settled if they are confined to the facts and to what
can legitimately be inferred from the facts. Within this methodological
perspective, it is only those (if any) who know the facts to very demanding
standards who are competent to judge. Conflicts which escalate beyond
what can be settled by the facts are no longer purely scientific. Although the
history of science (supported by recent sociological studies) shows that the
very question of what counts as a fact can be at issue in a controversy, the
accepted methodology remains valuable; as it provides a set of resources for

On the Autonomy of Pure Science

271

avoiding the escalation of controversy and for giving priority to limited and
more manageable forms of disagreement. T. H. Huxley, for example, could
counter Bishop Wilberforce's appeals to general sentiments in the audience
by proclaiming the status of Darwin theory to be something which only
the expert biologist was competent to judge. Huxley was able to use this
strategem persuasively for that audience, because Wilberforce's descent to
ad hominem rhetoric against Huxley (he asked whether it was on his grandfather's or his grandmother's side that he was descended from a monkey)
licensed Huxley to make an ad hominem reply. Huxley could expose the
errors in the Bishop's speech, and thus suggest that here was one non-biologist
who was not competent to judge such a technical biological matter. Thus
Huxley was able to sharpen up a relevant distinction between expert and
non-expert and give persuasive reasons why the more confined setting should
be given epistemological priority.
The strategy of constructing intellectual distance between the forms of
discourse prevailing on two sides in a dispute can, then, be very effective
when linked to acceptable hierarchical notions of relevant expertise. It can be
used (not always successfully) in conilicts between rival disciplines within
science, and as we shall see, it has had great success in confrontations between
science of the establishment and popular science.
I now wish to introduce one of the battlegrounds in which the rhetoric of
true and false science has been employed in the construction of intellectual
barriers. If we accept that the scientific establishment is contained within
the institutions of the elite of hierarchies of scientific expertise, we must
accept that a proportion of the interest in science lies outside such exclusive
institutions, in the numerous less expert individuals whose primary access to
science is through popular culture. Much of the science of the popular culture
popularization of orthodox science, but in addition, intellectual enterprises
are mounted which fail to penetrate the scientific establishment. These are
one of the targets to which the rhetoric of the philosophical images of science
are regularly applied. Barriers are often invoked or constructed to separate
such enterprises from true science, and the notion of intellectual distance is
employed in showing just how different in kind and quality they are from
true science.
In the modern age of professional science, the barriers between the qualified practitioner of a scientific discipline and the unqualified amateur are

272

R. G. A. Dolby

so clearly present that problems of exclusion are usually minimal. But the
case we will look at in detail, that of evolution just before Darwin, took place
when there was no relevant professional science. There were certainly exclusive
institutions for the elites of the sciences, but in amateur science, many members of each new generation must earn their own admission to the elite. The
battle was thus over who deserved to be admitted. The different institutional
settings for orthodox and popular science encourage some differences in the
prevailing styles of reasoning and argument. Orthodox science typically
strives for consensus through rigorous argument from agreed observational
data with modest theoretical postulates. It is specialized, and each region of
inquiry is deliberately limited to that which is recognized as achievable by
the accepted methods used. In contrast, popular science aspires only to
plausibility, for there is insufficient expertise for consensual judgements
to rigorous standards to be made. Rather than limit itself to what can be
achieved by established methods, popular science seeks systematic generality
and maximum significance on questions that trouble its supporters, and it
flourishes by maintaining popular interest. New ideas are often transient in
popular science, appearing only while they are newsworthy, but those ideas
which are successfully related to the lasting interests and commitments of a
sector of society may persist in that sector even though they have dropped
from wider attention.
Very often the forms of popular science against which the scientific
establishment finds it desirable to construct barriers has a style more suited
to the popular forum than to that of scientific orthodoxy. It is ,wide ranging
in coverage, offering insight into and explanations of many popular issues.
It is visionary and programmatic rather than rigorous and testable. It gains
its support without the mediation of the scientific expert. This deSCription
applies especially well to the case we will go on to consider, of evolution just
before Darwin, but we are encouraged to take the case seriously because,
after Darwin, very sirnilar ideas became the scientific orthodoxy.

2. Vestiges of the Natural History of Creation: Evolution Just Before Darwin

The Enlightenment conception of social progress was occaSionally applied to
nature in the late 18th century. If animal and plant species are variable, rather
than fixed, they might progress up the chain of being by becoming more

On the Autonomy of Pure Science

273

complex and highly differentiated. One of the fuller and better known of
such presentations was that of J.-B. Lamarck. But by the 1830s, considered
scientific opinion, especially as represented by such influential men of science
as Georges Cuvier and Charles Lyell had rejected Lamarck's idea of the mutability of species. Lamarck was thought to have produced mere speculation
which lacked a satisfactory causal mechanism for species change, and which
was against the general run of available evidence. In spite of this scientific
rejection, evolutionary speculations continued throughout Europe. In Britain,
discussion focussed for a time on an anonymous work, Vestiges of the Natural
History of Creation (1844). The essential claim of this book was that the
world is regulated by law and that God works in the world by means oflaw.
The author sought to show that there is a universal law of progressive development which applies to the formation of the solar system from a primordial
nebula, to geological change Ipld to the creation and transformation of life on
earth. life is spontaneously generated under favourable physical conditions,
and under suitable conditions each species is gradually modified, generation
by generation, from simpler to more complex forms. Man is the highest
product so far of this development process. The author did not attempt to
provide a mechanism for the evolutionary process, but had assembled a much
richer collection of supporting arguments than Lamarck had been able to
do some decades earlier. Vestiges generated considerable popular interest.
In Britain it had sold over 20,000 copies in eleven editions by 1860 (5).
Reviewers noted that it was particularly influential among those who had no
specialized knowledge of its subject matter. Mllny of its enthusiasts were
women or artisans seeking to educate themselves, but it was also much
discussed in the general culture of the influential classes of society. The
ideas it expressed were rich in their human implications and stimulating in
their controversial status. And discussion was frequently rekindled by new
rumours as to its authorship. It was not until some years after his death in
1871 that it was fmally revealed that the author had been Robert Chambers,
an Edinburgh publisher.
In contrast to its popular interest, Vestiges was vehemently condemned by
men of science. We shall examine the reasons the reviewers offered for their
rejection and the social and cultural interests they defended. We will then
look more closely at the distinction between popular and scientific forums
of discussion, as it was made by the critics of Vestiges and accepted by its

274

R. G. A. Dolby

author in his reply (6). After looking at the ways in which Vestiges differed
from orthodox science by being in the popular domain, we will examine its
influence - its lasting positive influence, and its negative influence on the
strategy forced upon Darwin and his supporters in the subsequent presentation of Origin of Species (1859).

The Rejection by Scientists ofYestiges (7)

The argument of Vestiges is clearly and simply written, and if one judges by
internal standards alone, the case it makes is plausible and restrained. The
wide range of fascinating implications are clearly indicated, but there is no
attempt at sensationalism. Those of its readers who had Htd" prior knowledge
of its subject matter were favourably impressed. But the leading men of
science who read it regarded it quite differently. The work spanned many
fields of science, and although most well informed on geology, seemed to
them to have been written by someone lacking any relevant expertise. Although the author had clearly read something of the subjects he covered, he
confused speculation, opinion and incompetent factual claims with authentic
science. All his evidence was dogmatically presented as if it was uncontested.
He displayed extreme credulousness towards anything that might support his
case and was simultaneously sceptical about apparent counter-evidence (8).
Many critical reviews of Vestiges followed the pattern of first challenging its
factual claims, then the processes of inference from them, and fmally saying
that this example of pseudo-science was also socially dangerous. The scientific
critics used the cultural resources of the prevailing philosophy of science to
give a rational form to their opposition. In the 1840s, in the wake of the
great treatises of Herschel, Whewell and Mill, it was easy to argue that any
reasoning that does not follow the inductive method is not science. The
critics also pointed out that the development hypothesis had already been
considered and rejected by science (9). The true man of science knows to
avoid matters which lie outside what can be established by the inductive
method, and that is where "the dabbler in science, the lover of the marvellous
in nature, begins" (10). To its scientific critics, the development hypothesis
in Vestiges was mere speculation. What supporting argument the author did
employ relied heavily on analogy and, in particular, on the idea that objects
with some external resemblance were causally related in the development

On the Autonomy of Pure Science

275

process of Nature. For example, mammalian embryos at one stage look a

little like fish, therefore mammals have developed from fish. This kind of
argument, said the critics, was most unreliable. There are so many resemblances in Nature that" ... let an ingenious person select from the unlimited
store and arrange his analogies to suit himself, and he can prove just what he
likes, so far as analogy affords any proof" (11).
Men of science objected in general to such 'speculation, breaking loose
from induction' (12), because it undermined the authority of science. "There
would be no objection to these things if they did not assume the form of true
science" (13). This particular delusion was worse than many others however,
because it undermined religion and threatened the harmonious relationship
between true science and true religion. Since all true accounts must be
consistent with one another, the books of Nature and of Revelation can only
conflict in imperfect interpretations. But Vestiges blantantly contradicts the
story of creation in Genesis (14). One especially full attack on Vestiges was
by the Cambridge geologist Adam Sedgwick. Sedgwick identified Vestiges
with earlier materialistic proponents of spontaneous generation and the
transmutation of species, which, he claimed, ''formally denied all proofs of
design in nature and all indications of an overruling Providence; and thus
struck at the foundation of Natural Religion" (I 5). Although the author of
Vestiges gave expression to worthy religious comments, Sedgwick insisted
that the philosophy he had been caught up in was 'rank materialism' (16). As
he wrote to Charles Lyell, Sedgwick thought that
If the book be true, the labours of sober induction are in vain; religion is a lie; human
law is a mass of folly, and a base injustice; morality is moonshine; and our labours for
the black people of Mrica were works of madness; and man and woman are only better
beasts (17).

Thus Sedgwick was not merely prepared to describe the conclusions of the
book as mischievous nonsense, but as antisocial nonsense (I 8).
Sedgwick made it clear that in attacking Vestiges, he was defending social
and cultural interests - those of the orthodox man of science, committed to
the existing methods and results of science, and to their harmonious relationship with true religion. He interpreted Vestiges as insidiously undermining
those interests, as being the more dangerous because it simulated real science.
He did not hide his social outrage, because he was appealing to others who

276

R. G. A. Dolby

shared his interests, though they might not have had sufficient scientific
knowledge to be aware of the corruption the bland packaging of Vestiges
concealed.
Vestiges as Popular Science
The scientific critics of Vestiges insisted that its author could not be a man of
science and that he was writing for a popular audience. David Brewster, for
example, contrasted the single-minded search for truth of the astronomer or
naturalist with
those revellers in speculation who practice their orgies in the temple of science, ransacking its storehouse for the materials of hypothesis, and not infrequently adulterating
them for popular taste, or fashioning them for vulgar apprehensions ... (19).

Adam Sedgwick confessed to having thought that the book might have been
written by a woman, explaining in a mixed metaphor,
We were led to this delusion by certain charms of writing - by the popularity of the
work - by its ready boundings over the fences of the tree of knowledge, and its utter
neglect of the narrow and thorny entrance by which we may lawfully approach it above all by the sincerity of faith and love with which the author devotes himself to any
system he has taken into his bosom (20).

At a time when a number of women had written very successful works of

popular science, but none had contributed Significantly to science itself, the
attribution of feminine authorship clearly identified the work as popular.
Furthermore, Vestiges was soon appearing in inexpensive editions designed
for sale to those of modest means, and this was noted by reviewers. When the
author of Vestiges came to reply to his critics in Explanations, he accepted
that the men of science had been uniformly hostile and made the counter
charge that they might not be the best people to evaluate a bold and general
thesis such as that of Vestiges. Men of science were mostly
engaged each in his own little department of science, and able to give little or no attention to other parts of that vast field. Experiments in however narrow a walk, facts of
whatever minuteness, make reputations in scientific societies; all beyond is regarded
with suspicion and distrust (21).

Men of science encouraged each other to be excessively cautious. The author

asked the reader whether he agreed that among men of science, in spite of

On the Autonomy of Pure Science

277

"laudable industry and zeal, there was also an intellectual timidity rendering
all the results philosophically barren?" (22). And the conclusion was finally
drawn,
Do men of science have their minds suitably prepared to receive with candour, or treat
with justice, a plan of nature like that presented in the Vestiges of Creation? No, it must
be before another tribunal, that this new philosophy is to be truly and righteously
judged.

Chambers' biographer suggests that this 'other tribunal' was presumably the
educated layman who was buying out every edition of Vestiges (23).
We see, therefore, that Chambers and his critics agreed that Vestiges was
a work of popular science rather than a contribution to the community of
specialized scientific expertise. Even though this occurred at a time before
science was professionalized, the distinction between the institutionalized
single-minded search for truth and the unspecialized contribution to popular
culture, was clear to the participants. Once we recognize clearly the popular
nature of Vestiges, we can understand better some of its characteristics and
some features of the reaction against it.
Vestiges sought, not to establish modest conclusions with disciplined
rigour and inductive certainty, but to stimulate interest in topics which were
regarded as important in the wider society. Its technique, therefore, was to
be suggestive and thought provoking rather than to prove its points. Knowledge claims in orthodox science build only on work which is widely accepted
or which the author hopes to persuade his audience to accept readily. For in
that way they can become authoritative. Scientists avoid drawing points into
their argument which have generated inconclusive controversy. An individual
mid-19th-<:entury scientist might have been inclined to accept the nebular
hypothesis, or spontaneous generation, or phrenology. But to invoke any of
these areas of controversial interest in a separate different scientific argument
would only help bring the latter, too, into disrepute. In contrast ,Chambers
appealed to all these areas in his argument, and maintained his commitments
to each of them as long as he thought them defensible in successive editions.
1his was a very bad strategy in orthodox science, but it was exactly what a
popular audience wanted. The development hypothesis gained its interest by
being shown to be relevant to a wide range of existing issues, and by drawing
upon examples which caught the popular imagination. Furthermore, as

278

R. G. A. Dolby

popular science, Vestiges did not limit itself to the scientific ideas it covered.
Unlike Darwin's presentation of evolution fifteen years later, Chambers
spelled out clearly his conception of the implications of the theory for man
and for man's relation to God. A popular audience was less interested in
science as such, than in the implications of a scientific idea for matters of
greatest human importance.
Some modem commentators on Vestiges have asked themselves why it
was that scientists reacted so strongly to the work when they could have
condemned it simply by ignoring it (24). And part of the answer seems to be
because it was a form of popular science that threatened the main thread of
social justification of science which had built up in popular science literature.
For several decades all the respectable leaders of scientific opinion, of whatever school,
had been actively engaged upon the project of popularizing natural philosophy among
mechanics and disseminating science to untutored multitudes. The result, it was thought,
could only elevate the moral sentiments of the working class and stabilise its situation by
demonstrating the providential and material necessity and the comprehensive beneficence
of industriallU"rangements. But now, it suddenly appeared, it made a great deal of
difference what sort of scienCe was popularized. Hitherto scientists themselves, misunderstood to be sure, had been unfairly charged with encouraging infidelity and moral
anarchy. And just as they were successfully solidifying the religious cement of industrial
society, a real heretic, cleverly and falsely got up in the guise of science, came to undo
their labours and to. demolish the framework (25).

Vestiges was not a threat as a contender for scientific knowledge - most men
of science rejected it out of hand. But its general readers, who had not the
advantage of expert knowledge could easily be seduced and corrupted by it.
The criticism came because it was so plausible and so lucid and was generating
popular interest, and although attack gave it still further publicity, perhaps
it would eventually discredit it in the public domain.
Vestiges and Later Evolutionary Theory
One of the problems arising in the vast body of historical discussion of the
Darwinian revolution has been to clarify the relation between Darwin and
his evolutionary predecessors. On the one hand, it is possible to read such
early evolutionists as Lamarck as providing a well thought out theory which
differed from Darwin's only in that it did not employ the particular mechanism of natural selection (26). On the other hand, it is possible to regard

On the Autonomy of Pure Science

279

Darwin's as the first truly scientific theory of evolution and his predecessors
as merely vague speculators. Modern historians of science have developed a
dislike of looking for precursors of the recognized proponents of scientific
ideas. Such chains of precursors can be forced back indefinitely into the past,
provided the documents are reinterpreted in the perspective of a modern
framework, rather than in terms of the interests of the time in which they
were written. The search for precursors has been ridiculed as the disease of
'precursoritis'. But in the case of Darwin, to refuse to recognize the significance of precursors is to take sides in a late-19th-century-debate - a debate
in which Darwin's supporters stressed his originality and his critics (such as
Samuel Butler) accused him of plagiarism. I think that the present study can
show why it was. that Darwin and his supporters tended to exaggerate the
difference between The Origin of Species and its precursors; in brief, it was
because those precursors, and especially Vestiges of the Natural History of
Creation, had come to be regarded as popu1ar pseudo-science. The strategy of
gaining acceptance for Darwin's theory required him to stress the intellectual
distance between his own work of scientific natural history, which was
'proved' by the standards of the science in question, and the imperfect
popular speculations which preceded it. The Origin of Species kept to a
minimum the references to man and to God, on which controversy was most
heated. It was not open to Darwin to appeal to the authority of an earlier
tradition of evolutionary ideas, for that would have encouraged his audience
to regard Darwin as having merely produced yet another unproven speculation. T. H. Huxley, for example, in an earlier commentary, described the
author of Vestiges as a dreamer, "by whose well-intentioned efforts the
Lamarckian theory received its final condenmation in the minds of all sound
thinkers" (27), and in contrast [mis] described Darwin as one who "abhors
mere speculation as nature abhors a vacuum ... The path he bids us follow
professes to be, not a mere airy track, fabricated of ideal cobwebs, but a solid
and broad bridge of facts" (28).
One of the interests of more recent history of biological science has been
to explore the extent to which biological theories were influenced by contemporary social ideas. Has the history of biology followed its own internal logic
of problem generation and problem solution, or has it merely reflected the
state of the external culture? This has been a special interest because of the
frequency with which biological theory has been applied back to arguments

280

R. G. A. Dolby

about the psychological and social nature of man. Is biology functioning as an

ideology which provides empty rationalization of social prejudices? In the
case of evolutionary theory, one especially strong form of this argument
is the evidence that social and biological theory in the early 19th century
interacted within a common context (29). According to R. M. Young, the
common intellectual context lasted until the 1870s and 1880s, when it came
to pieces in a period of increasing professionalization and specialization (30).
What survived of the common context was at a much lower intellectual level.
This issue can be clarified by examination of the contrast of and relationship
between popular and establishment Qater 'professional') science. For, as
we have already seen, the popular context was one in which every effort
was made to interlink the topics of interest. In contrast, the positivist and
inductivist elaborations of the empiricist outlook of orthodox science tended
to encourage the treatment of each scientific issue in isolation from its wider
implications. Therefore, the common context argument is easy to establish
if popular literature is employed but more implausible if the historian concentrates on the severest and most rigorous parts of science. In the case of
evolutionary ideas in the mid-19th century, it is not difficult to find many
lines of popular evolutionary discussion which are closely linked at all stages
to social' ideas, but it would be an error to reduce the more disciplined
developments within biology to that popular context. However, popular
science was in many ways an unavoidable influence on the men of science in
their efforts to conduct dispassionate and objective inquiry. In this case social
and cultural interests have influenced the science through the mediation of
the associated popular science. The new evolutionary biology could not
completely distance itself from the popular interest in evolutionary theory
which preceded and accompanied its emergence.
Popular and semi-popular British writings on evolution in the mid-19th
century came in particular from the social circle of adventurous thinkers
discussed by Young (31). Their enthusiasms included phrenology, mesmerism
and their combination in phrenomesmerism, evolutionary theory, positivism
and new sciences of human and social phenomena. Some also involved themselves in spiritualism and its investigation in psychic research. The interests
of Robert Chambers, Herbert Spencer and A. R. Wallace, who all wrote on
evolution before parwin had published, cover this range. Chambers revealed
his commitment to phrenology in Vestiges, drawing supporting argument

On the Autonomy of Pure Science

281

from it and so provoking further wrath from his scientific critics. When
spiritualism became the rage later in his life, he showed enthusiastic interest
in that. Spencer wrote early articles on phrenological topics. He had a critical
interest in positivism, and wrote major treatises on psychology and sociology.
The causes championed by A. R. Wallace, whose paper on evolution by
natural selection stimulated Darwin into publication, are comparable. In
his youth, Wallace had dabbled in and become committed to phrenology,
mesmerism, and phrenomesmerism. He had become convinced of the fact of
evolution by reading Vestiges. This interest was a spur to taking up a career
as a naturalist. Later in life, he took up other unorthodox interests, including
spiritualism and psychic research. Wallace's evolutionary work illustrates one
of the routes for popular ideas from unorthodoxy to scientific respectability.
Although Darwin did not belong to this social circle of intellectual adventurers, he was undoubtedly influenced by it in the presentation of his ideas,
particularly in his later writings. But not all the influence was positive. When,
in 1859, Darwin fmally published the theory he had been gestating for more
than twenty years, he went to some trouble to separate it from the popular
discussions. He made only minimal reference to man and to religion, although
he had always been personally interested in the implications for man. Instead
he concentrated on an inductively arranged argument and a response to the
scientific objections he had had ample time to anticipate. Having won over a
small group of supporters in his own area of natural history, they helped him
convert more of the scientific orthodoxy. Darwin was not actively involved in
the confrontations in the controversy which was generated, always pleading
ill health. But T. H. Huxley, in particular, was an effective advocate. Huxley,
who had earlier written a savage review of Vestiges (32), maintained his
intellectual consistency by stressing the differences between Darwin's theory
and its popular precursors and by insisting on the methodological superiority
of the present theory. Although Darwin's supporters did not avoid the wider
implications of an evolutionary biology, they insisted that the status of
the theory should be settled as a purely scientific matter. I have already
mentioned Huxley's opportunistic exploitation of Bishop Wtlberforce's
mistakes at the Oxford meeting of the British Association for the Advancement of Science in 1860. It appears that Huxley, like Wtlberforce, had not
desired or planned for this meeting (33). The day before the exchange with
the Bishop, Huxley had sought to avoid discussion of the theory because,

282

R. G. A. Dolby

" ... a general audience, in which sentiment would unduly interfere with
intellect, was not the public before which such a discussion should be carried
on" (34).
Interest in evolutionary theory rapidly became widespread in British
society. Among the many currents of evolutionary discussion, writers such
as Spencer were prominent. Spencer's evolutionary ideas now had a new
scientific status. Spencer became a philosopher of the age amplifying newly
fashionable ideas into a world view. Many late-19th century intellectuals
passed through a phase of interest in (and often later reaction against) his
systematic expositions. When the leading natural historians eventually turned
to the detailed examination of the evolution of man, they found it impossible
to limit their discussion to 'pure biology'. Inevitably, the themes, preoccupations and prejudices of the popular literature found some expression in their
work. The inheritance of acquired characteristics, particularly habits returning
as instincts, a Lamarckiat). idea given full support by Spencer, was increasingly
accepted by Darwin in his writings on man. Ideas of inequality between the
sexes, social classes and races, which had been much discussed in the popular
literature were often given unthinking recognition in the work of the most
serious scientists. The idea that there is a single line of evolutionary development (part of the chain of being), which had been a common feature of
popular discussions since Lamarck, and which made it natural to suppose
that one could study other cultures and races as corresponding to earlier
evolutionary stages of European man, inspired mid- and late-19th-century
anthropology. The idea that the developing child, like the unborn embryo,
recapitulates the evolution of the species, was carried via the popular literature from its pre-Darwinian origins into late-19th-century psychology (35).
According to this doctrine, children instinctively climb trees and hide in caves
as our ancestors habitually did. The idea that nature is (like modern capitalist
society) a place of open competition, of struggle for survival in which only
the fittest perpetuate their kind was applied by social Darwinists to justify
laissez-faire capitalist society. All these features of evolution came, not
from the pure biology of Darwin and his friends, but from the popular and
semi-popular evolutionary philosophy which preceded and accompanied
their work, and from which they could not completely insulate their own
reasoning.
The common context argument, then, is effective for those aspects of

On the Autonomy of Pure Science

283

evolutionary theory which were primarily developed as popular and semipopular literature. But scientists of the late-19th century became increasingly
concerned to separate their own activity from popular science and to ensure
that all of the new science followed a rigorous methodology. Let us return
for illustration to the writings of T. H. Huxley.
In Huxley's generation, British science was for the first time dominated
by those gaining a living from science. There was a new self-consciousness
among this group, and Huxley's campaigns and enterprises on occasion
defended the interest of the new occupation (he used this word rather than
'profession'). Huxley presupposed a sharp distinction between the established
science to which he contributed and science in the popular culture. He
frequently addressed both audiences and was occasionally obliged to defend
his willingness to devote effort to popularization of science. But he also
played his part in strengthening and maintaining the boundary between the
two. For example, he spoke against the difficulties produced by those "who
think that what they have picked up from popular exposition qualifies them
for discussing the great problems of science" (36). In his writings on pseudoscience, he deplored
the waste of time and energy bestowed on the endeavour to deal with the most difficult
problems of science, by those who have neither undergone the discipline, nor possess the
information, which are indispensable to the successful issue of such an enterprise (37).

The author of Vestiges was among those he was alluding to. What he wished
to attack was
a process of mystification, based upon the use of scientific language by writers who
exhibit no signs of scientific training, of accurate scientific knowledge, or of clear ideas
respecting the philosophy of science, which is doing very serious harm to the public (38).

In addition to his attacks on popular writers, Huxley made a target of

those who sought to extrapolate science beyond what could be justified
methodologically. He resented being thought of as a positivist and admonished
associates who turned Darwinism into a new dogma. For example, one of the
interests of semi-popular scientific philosophers in the mid-century had been
to provide a scientific basis for morality. Spencer sought to construct an
evolutionary theory with implications for human values and behaviour which
could serve in this way. But in the 1890s, Spencer's laissez-faire political
values were no longer seen as progressive by the intellectually adventurous.

284

R. G. A. Dolby

Huxley, though he was a friend of Spencer's, was among those who objected.
The need to separate scientific knowledge from popular speculation could be
accomplished in this case by pointing to the fallacy of arguing from 'is' to
'ought'. Spencer had only succeeded in constructing his laissez-Iaire values
from evolution by assuming that society should take a form in which competition rather than cooperation is crucial for survival (39).

3. Sociobiology
No attempt is made in the discussion which follows to provide a comprehensive introduction to or analysis of the recent controversy over sociobiology.
However,it is of interest to apply the understanding developed in the previous
section to a contemporary issue, one in which the heat of battle has yet to
die away, and with which many readers will already be familiar. The present
controversy has focussed around the 'new synthesis' of E. O. Wilson, who
sought to redefme the disciplinary boundaries between studies of animal
behaviour and population genetics by drawing the material into a single
quantitative theory. However, in his large textbook (40), Wilson did not limit
himself to a synthesis of existing scientific learning; he added a fmal chapter
in which sociobiology was applied to man. It was mainly this extrapolation
which triggered the controversy. The central claim was that Wilson was
resurrecting biological determinism, a favourite ideological device for rightwing viewpoints. Wilson's presentation, it was claimed, failed to separate
out the effects of "the personal and class prejudices of the researcher" (41).
Although Wilson claimed to feel intimidated by the hostility (42), which he
insisted was based on misconstrual or misrepresentation (43), he responded to
the challenge. Indeed, his readiness to maintain his side of the controversy led
him to produce a whole book on the sociobiology of man (44).
One problem Wilson had to face was that, although he claimed to offer a
new synthesis, he was by no means operating in an intellectual vacuum. The
efforts to biologize human social behaviour were increasingly discussed within
popular culture. There, publications could attract appreciative attention by
being internally plausible, comprehensive and giving a new sense to sociallysignificant issues. Many of them appealed to similar sectional social interests
to their predecessors earlier in the century. But some new ones were added.
There was, for example, Elaine Morgan's The Descent 01 Woman (45) which

On the Autonomy of Pure Science

285

replaced male chauvinist accounts of the evolutionary origin of man by a

feminist one. However, because such quasi-scientific popular speculations as
those of Morgan, or of Robert Ardrey (46) or Desmond Morris (47) could
not all be true and, because there was no objective way of deciding between
them, there was no real need for the scientific establishment to take any of
them seriously.
Wilson devoted a small early section of his textbook to a discussion of his
immediate popular predecessors. The move he made to create intellectual
distance between them and his own work was to expose the methodological
weaknesses of such theories, which are advocated rather than inferred and
which are not formulated so as to be falsifiable (48). In the main body of his
'
text, Wilson ignored such popular work.
Early sociobiology could 'not simply exploit existing boundaries and
barriers in setting itself up as a new disciplinary synthesis. One of the popular
works Wilson dismissed in the section just referred to was The Imperial
Animal by L. Tiger and R. Fox. Tiger and Fox had explained in their preface
that they had at first wished to write a textbook "but it was hopeless to
write a textbook for a discipline that did not exist" (49). Tiger and Fox had
discovered that those who wish to be pioneers must venture beyond the
conventional boundaries of the scientific establishment. Although writing for
a non-specialist audience, they had tried to distance themselves from their
more disreputable predecessors, alluding to
... some biologists who have recently with undue innocence leaped from discussion of
simple animals to the most complex of all. The leap can be made, but not in ignorance
of the details and uniqueness of human social institutions (50).

Wilson, at least, failed to appreciate any significant difference between Tiger

and Fox and those innocent biologists. Wilson's problems were especially
acute when he, too, came to write his book on human sociobiology. In his
preface, he conceded that a book in this area could not be a textbook. He
described On Human Nature as a work about science rather than a work of
science, having at its core a speculative essay (51). Thus, Wilson himself was
now entering the territory of the kind of non-scientific speculation which he
had barred from his sociobiology textbook.
Let us turn, then, to the intense critical reaction which the Wilson synthesis
originally stimulated. The primary strategem of the critics was to force the

286

R. G. A. Dolby

conflict into a general forum. This was not a matter to be settled by some
narrow group of experts. The devices Wilson had used to distance himself
from the immediately preceding popular writers were turned against him. He
had, for example, accused them of advocating unfalsifiable theories (52). His
most outspoken critics, the Sociobiology Study Group of Science for the
People argued in one criticism:
When we examine clIIefully the manner in which sociobiology pretends to explain all
behaviours as adaptive, it becomes obvious that the theory is so constructed that no
tests are possible: There exists no imaginable situation which carmot be explained; it is
necessarily confirmed by every observation (53).

The Wilson material was construed by the critics as comparable to earlier

socially biased biological treatments of man. It was reactionary ideology for
Western capitalism and male sexism. Of course, this critical stance provoked
argument as to whether a deliberate misreading of Wilson was involved, and
also whether ifit was a misreading, it was one to which Wilsonian sociobiology
was constitutionally susceptible. But however these secondary matters were
to be decided, the strategy of preventing localization of the conflict has
been successful so far. The matter has become one not just for those expert
in the general area of sociobiology, but for a much wider community of
intellectuals. It has become appropriate for people to write entire books on
the subject while openly admitting that they have not had a relevant scientific
training (54). It has been difficult for Wilson to localize the conflict. He
cannot seek to limit it to some community of experts on human sociobiology,
for that is the very community he is trying to create. While the status of
animal sociobiology as a discipline within biology might conceivably have
been delegated to the experts, even now the quarrel over human sociobiology
can be construed as a quarrel over whether we can accept a new kind of
expert.
My discussion does not lead to predictions as to how the sociobiology
debate will tum out. But historical analogues suggest that as long as the
sociobiologists allow the conflict to be conducted along the present very
broad front, they cannot win. They would have to find a way to appeal to the
prejudices of all competing social classes and factions in a mutually acceptable
formula. One attempt at dispassionate evaluation ends with the suggestion
that

On the Autonomy of Pure Science

287

Human sociobiology should be given a chance to prove its worth. If it cannot deliver on
its promises, it will collapse soon enough .. ; but if it does prove viable, then its success
could pay scientific dividends of the highest order (55).

fu terms of my account, this amounts to proposing that the human sociobiologists should be allowed to set about creating a new scientific community,
the leaders of whom we might come to accept as the experts in such matters.
That is what 'proving viable' amounts to in social terms. And the way in
which this would normally be done is by sociobiologists making and negotiating specific knowledge claims in such a way that critics can no longer
find a SOCially acceptable position from which rational objections can be
made. The cumulation of relevant uncontested knowledge by a community
is normally sufficient for us to concede to it an area of expertise. I think,
however, that it will be extremely difficult to establish human sociobiology in
this way. The conflict has already been widened to a popular forum in which
new claims of expertise are regarded with suspicion. In the absence of some
widely appreciated achievements upon which subsequent research could
build, the strategems available for de-escalating the conflict will take time to
be effective.

4. Concluding Remarks
Professionalism is undoubtedly a factor in the enforcement of the barriers
between and around the established sciences in the present age of professional science. But we have seen that such barriers were first created by the
elite of amateur science. Furthermore, professional self-interest tends to
conserve existing boundaries, but as the range of science has extended and its
internal linkages have been restructed, the boundaries have often had to be
renegotiated. Even within professional science, the boundaries that otherwise
function so readily as barriers do not force conformity upon elite innovators.
Wilson, for example, was quite prepared to cross existing disciplinary boundaries in creating sociobiology. Our studies of the boundary between popular
and established science show that although it is exploited as a barrier on
many occasions, it must be crossed by those who wish to create new outposts
of scientific orthodoxy. The pioneers of social science had to face this problem
in an acute form. Although they could invoke the methods of professionalized

288

R. G. A. Dolby

natural science, the lack of widely accepted achievements of social science

made it especially difficult for them to establish themselves as more authoritative than the rejected alternatives in the popular culture.
The problems posed by existing disciplinary barriers for new forms of
interdisciplinary research are not simply those of establishment ideology.
Such barriers also have a cultural significance and an explicit rationale for
their construction. Furthermore, interdisciplinary programmes which break
down existing barriers and hierarchies of expertise also temporarily break
down the existing mechanisms for the localization of conflicts. Order is
restored only when a new hierarchy of professional expertise is established:
If, as in the case of human sociobiology, the programme itself generates
controversy, the initial hostilities produced cannot be limited to a manageable size, and can be very disruptive.
The problems are less severe in two other types of situation. An interdisciplinary programme which responds to some external issue about which
there is a high degree of social consensus (such as the recognition by the
1970s that new measures were required to deal with pollution) can be forced
upon existing disciplinary groups. They need not resist if they also hold the
consensus view of the external issue. Although it will be harder to localize
conflicts in such cases, there is reasonable hope that conflicts which escalate
by appealing to wider audiences can still be resolved in terms of the wider
social consensus. Eventually, a new hierarchy of professional expertise is
created, and the period of greatest difficulty is over. The other kind of interdisciplinary programme which is subject to less severe difficulties than human
sociobiology is that which seeks to build upon some existing intellectual
achievement which is already accepted by the parties involved. Those who
seek to extend and elaborate such an achievement in an interdisciplinary
programme can immediately defend the intellectual posture that they are
a new group of experts. Provided that they behave cautiously and avoid
provocative claims to large areas of disputed intellectual territory, they can
quickly become like any other growing branch of science. Perhaps that is
what Wilson's animal sociobiology should have been like.
Unfortunately, not every new interdisciplinary enterprise can be fitted
into one of these readily assimilated types. However, the problems to be
faced can then usefully be treated by the framework of categories that I
have been developing in this paper.

On the Autonomy of Pure Science

289

Notes and References

1. The implications of the professionalization of science for the elaboration of barriers
around science have been widely discussed in the sociology of science. See, for
example, R. Whitley, 'Changes in the Social and Intellectual Organization of the
Sciences: Professionalization and the Arithmetic Ideal', in E. Mendelsohn, P.
Weingart and R. Whitley (eds.), The Social Production of Scientific Knowledge,
Sociology of the Sciences Yearbook 1, Dordrecht and Boston: Reidel, 1977, pp.
145-6, who writes of the professionalization of the sciences excluding lay competition; H. Nowotny, 'Science and its Critics: Reflections on Anti-Science', in H.
Nowotny and H. Rose (eds.), Counter-Movements in the Sciences, Sociology of the
Sciences Yearbook 3, Dordrecht and Boston: Reidel, 1979, p. 16, who writes of
science holding a monopoly on claims as to what constitutes a natural phenomenon,
and as to how it should be investigated. A relevant discussion of the inadequacy of
appeals to professional status in the context of the maintenance of intellectual
barriers is given in H. Kuklick, 'Boundary Maintenance in American Sociology:
Limitations to Academic "Professionalization"', Journal of the History of the
Behavioral Sciences 16 (1980) 201-19. See, in particular, page 202, which refers
to "the recent wave of antiprofessional scholarship".
2. Elias is preoccupied with science conducted within universities and with the influence of philosophers in one line of descent from Kant. These concerns may
well be natural for one whose origins lie in German intellectual culture of the early
20th-century, but are inadequate for a comprehensive analysis. Universities undoubtedly became the dominant institutional basis for science during the 19thcentury, but were less important before then. Even in the 19th-century, there were
important national differences in the time at which universities became central to
science. English science, for example, was dominated by amateurs in the fust half
of the century. For a brief discussion which illustrates how the social character of
19th-century English science is now understood, see, for example, M. Berman,
'''Hegemony'' and the Amateur Tradition in British Science', Journal of Social
History 8 (1974-5) 30-43. The main historical study in my paper here describes
an episode in which an amateur scientific establishment sought to contrast itself
from a more radical popular alternative. Universities undoubtedly dominated
science at the beginning of the 20th-century, but since then other major institutional bases for science have emerged. Throughout the history of science, particular
scientific establishments have been very limited in the power they can exercise
beyond local or national or regional boundaries.
The account that Elias gives of an establishment of philosophy which has
reinforced the influence of physics as the primary model for science is also a local
mythology designed for outsiders resentful of philosophy. Philosophers are always
in disagreement, and the kind of consensus on which a philosophical establishment
could build has always been limited in time and space. The version Elias describes in
caricature applies best to certain philosophers of an earlier generation.
3. Among the classics of such defences see, for example, J. Huxley, Soviet Genetics
and World Science: Lysenko and the Meaning of Heredity, London: Chatto and
Windus, 1949; R. K. Merton, 'Science and the Social Order', Philosophy of Science
5 (1938) 321-37, reprinted in R. K. Merton, The Sociology of Science: Theoretical

290

4.

5.
6.
7.

R. G. A. Dolby

and Empirical Investigations, Chicago: Chicago University Press,1973, pp. 254-66;

M. Polanyi, The Contempt of Freedom: The Russian Experiment and After, London: Watts,1940.
Accounts of the meeting are collected in F. Darwin (ed.), The Life and Letters of
Charles Darwin, London: John Murray, 1888, Vol. 2, pp. 320-3, and L. Huxley
(ed.), Life and Letters of Thomas Huxley, 2nd edn., London: Macmillan, 1913,
Vol. 1, pp. 259-274. For a recent survey of the problems of establishing the
historical facts behind the legend, see J. L. Altholz, 'The Huxley - Wilberforce
Debate Revisited', Journal of the History of Medicine and Allied Sciences 35
(1980) 313-6.
M. Millhauser, Just Before Darwin: Robert Chambers and Vestiges, Middletown,
Conn: Wesleyan University Press, 1959, p. 33.
(R. Chambers,] Explanations: A Sequel to 'Vestiges of the Natural History of
Creation' by the Author of that Work, London: John Churchill, 1845.
For a more general account of the criticisms, see Millhauser, op. cit., 1959, Note

5.

8. For example, one reviewer said that Vestiges claimed its hypothesis to be true
because it is "drawn from the response of nature, and the response of nature is
only true when it gives back the watchwords of the hypothesis" (A. Sedgwick,
'Vestiges of the Natural History of Creation', Edinburgh Review 82 (July 1845)
41).
9. See, for example, (F. A. Bowen,] 'Vestiges of the Natural History of Creation',
North American Review 60 (1845) 465.
10. Anonymous reviewer of Vestiges-in The Athenaeum, No. 897, Jan.4.1845, p. 11.
11. (A. Gray,] 'Explanations of the Vestiges', North American Review 62 (1846)
503. Detailed criticisms of the use of resemblances in Vestiges are also made by
Sedgwick,op. cit., 1845 (Note 8), pp. 5-9, and by Bowen, op. cit., 1845 (Note 9),
p.451.
12. Anonymous reviewer, op. cit., 1845 (Note 10), p. 11.
13. Idem.
14. (D. Brewster,] 'Vestiges of the Natural History of Creation', North British Review
3 (1845) 470.
15. A. Sedgwick, A Discourse on the Studies of the University afCambridge, 5th edn.,
London and Cambridge: J. W. Parker, 1850, p. xix.
16. A. Sedgwick,op. cit., 1845 (Note 8), p. 63.
17. Letter from Sedgwick to Lyell. J. W. aark and T. M. Hughes, The Life and Letters
of The Reverend Adam Sedgwick, Cambridge: Cambridge University Press, 1890,
Vol. 2, p. 84.
18. A. Sedgwick, op. cit., 1845 (Note 8), p. 3.
19. D. Brewster, op. cit., 1845 (Note 14), p.4 73.
20. A. Sedgwick, op. cit., 1845 (Note 8), pp. 3 -4.
21. R. Chambers, op. cit., 1845 (Note 6), p. 175.
22. Ibid., p. 176.
23. Millhauser,op. cit., 1959 (Note 5), p. 144.
24. See, for example, C.C. Gillispie, Genesis and Geology (1951) New York: Harper,
1959, p. 162.
25. Ibid., pp. 150-1.

291

On the Autonomy of Pure Science

26. As an example of the search for precursors, see B. Glass et al. (eds.), Forerunners of
Darwin: 1745-1859, Baltimore: Johns Hopkins Press, 1959. The classic of the
genre of precursor'ileeking histories of science is P. Duhem, Le Systeme du Monde;
Histoire des Doctrines Cosmologiques de Platon du Copemic, 10 Vols, Paris:
Hermann, 1913-59.
27. T. H. Huxley, 'The Darwinian Hypothesis' (1859), reprinted in his Collected Essays.
Vol. II, London: Macmillan, 1894, p. 13.
28. Ibid., pp. 20-1.
29. See, for example, R. M. Young, 'Malthus and the Evolutionists: The Common
Context of Biological and Social Theory', Past and Present, No. 43 (1969) 10945; R. M. Young, 'The Historiographic and Ideological Contexts of the NineteenthCentury Debate on Man's Place in Nature', in M. Teich and R. M. Young (eds.),
Changing Perspectives in the History of Science: Essays in Honour of Joseph
Needham, London: Heinemann, 1973, pp. 344-438.
30. R. M. Young, 'Natural Theology, Victorian Periodicals and the Fragmentation
of the Common Context', in C. Chant and J. Fauvel (eds.), Darwin to Einstein:
Historical Studies on Science and Belief, Harlow, Essex: Longman, 1980, pp. 69107.
31. See, for example, the works of Young just cited in Note 29 and 30.
32. T. H. Huxley, 'Vestiges of the Natural History of Creation', British and Foreign
Medico-Chirurgical Review, 13 April 1854.
33. Altholz,op. cit., 1980, Note 4.
34. Quoted by F. Darwin from The Athenaeum, July 14,1860. Darwin, op. cit., 1888
(Note 4), p. 320.
35. See, for example, the writings of J. M. Baldwin and G. S. Hall.
36. T. H. Huxley, Discourses: Biological and Geological. Volume 8 of his collected
essays. London: Macmillan, 1894, p. viii.
37. T. H. Huxley, 'Science and Pseudo-Science' (1887), in Science and Christian Tradition. Volume 5 of his Collected Essays. London: Macmillan, 1894, p. 116.
38. Ibid., p. 177.
39. T. H. Huxley, 'Evolution and Ethics' (1893), in Evolution and Ethics and other
Essays. Volume 9 of his Collected Essays, London: Macmillan, 1894, pp. 46-116.
40. E. O. Wilson, Sociobiology: The New Synthesis, Cambridge, Mass: Harvard University Press, 1975.
41. E. Allen, et al., 'Against "Sociobiology" " New York Review of Books, November
13, 1975, pp. 182, 184-6. Reprinted in A. L. Caplan (ed.), The Sociobiology
Debate: Readings on Ethical and Scientific Issues, New York: Harper and Row,
1978,p.264.
42. N. Wade, 'Sociobiology: Troubled Birth for New Discipline', Science 191 (1976)
1151-5. Reprinted in Caplan, op. cit., 1978 (Note 41), p. 332.
43. E. o. Wilson, 'Academic Vigilantism and the Political Significance of Biology',
BioScience 26 (1976) 183, 187-90. Reprinted in Caplan,op. cit., 1978 (Note 41),
p.291.
44. E. O. Wilson, On Human Nature, Cambridge, Mass: Harvard University Press, 1978.
45. E. Morgan, The Descent of Woman, London: Souvenir Press, 1972.
46. See, for example, R. Ardrey, African Genesis: A Personal Investigation into the
Animal Origins and Nature of Man, London: Readers Union, 1963; The Territorial

292

47.
48.
49.

50.
51.
52.
53.

54.

55.

R. G. A. Dolby

Imperative: A Personal Inquiry into the Animal Origins of Property and Nations,
London: Collins, 1967.
See, for example, D. Morris, The Naked Ape: A Zoologist's Study of the Human
Animal, London: Cape, 1967; The Human Zoo, London: Corgi, 1971.
WIlson,op. cit., 1975 (Note 40), pp. 27-31.
L. Tiger and R. Fox, The Imperial Animal, London: Secker and Warburg, 1971,
p. x.
Ibid., p. xi.
Wilson,op. cit., 1978 (Note 44), p. x.
Wilson,op. cit., 1975 (Note 40), p. 28.
E. Allen, et al., 'Sociobiology: A New Biological Determinism', in Sociobiology
Study Group of Boston (eds.), Biology as a Social Weapon, Minneapolis: Burgess,
1977. From the quotation in M. Ruse, Sociobiology: Sense or Nonsense?, Dordrecht: Reidel, 1979, p. 111.
I have in mind in particular the very helpful book by the philosopher of science M.
Ruse,op. cit., 1979, Note 53.
Ibid., p. 214.

RESEARCH TRAILS AND SCIENCE POLICIES: LOCAL AND

EXTRA-LOCAL NEGOTIATION OF SCIENTIFIC WORK

DARYL E. CHUBIN and TERENCE CONNOLLY

Georgia Institute of Technology, Atlanta, Georgia

1. Introduction
At any given point in the development of a scientific specialty, there exists
some finite set of research topics which are seen as legitimate, interesting,
and feasible by the members of the specialty. How do the members of the
specialty achieve an allocation of their efforts across these topics? What does
the aggregation of these individual decisions imply about the development of
the specialty? To what extent are these consequences shaped by the activities
of scientific 'hierarchies' (Le., formal and informal organizational structures,
policies, and leaders)? And might these consequences be shaped differently
(and perhaps more effectively)?
These questions have been posed before, in various guises (1). Typically,
however, each is addressed separately with little regard for its companions.
For example, the prevailing view on the question of individual topic choice
appears to favor a 'competitive grazing' model. The specialty is pictured as
a bounded meadow into which individual scientists, competitors for recognition and priority, allocate themselves. Areas of heavy over-grazing become
less attractive, virgin pasture is sought out, and the overall territory represented by the available research topics is Uniformly exploited. Thus,according
to Hagstrom (2), those who discover important problems upon which few
others are engaged are less likely to be anticipated and more likely to be
rewarded with recognition. As a result, scientists tend to disperse themselves
over a range of possible problems.
While Hagstrom suggests purposive choice in the matter, Price (3) in
examining the phenomenon of multiple discovery, suggests a simple random
model, in which the ripe apples of discovery are seized by the reaching hands
293
Norbert Elias, Herminio Martins and Richard Whitley
Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 293-311.
Copyright 1982 by D. Reidel Publishing Company.

294

Daryl E. Chubin and Terence Connolly

of the blind scientific harvesters. (It should be noted that Price's demonstration of reasonable fit of his model to Merton's (4) multiple-discovery data
should not be over-interpreted; a similar fit would be found if the harvesters
picked from only a single tree, neglecting the rest of a huge and fruitful
orchard.)
2. Research Trails and Scientific
The simple agricultural images cited above accord poorly with our impression
of how scientific problems are actually chosen - with the passion and concern
shown by individual scientists. These scientists select their projects under the
influence of a complex of organizational, cultural, political, and intellectual
conditions. Our intent in this essay is to sketch the outlines of a richer
and more complex model, in which individual choice of 'research trails' is
embedded in such issues as the distribution of research effort across problems
in a specialty, the development of a specialty over time, and the role of
hierarchies and elites in shaping these processes.
Our central concern is with the implications for an aggregate of scientific
work, or specialty, of the choices of individual researchers - what Schelling
(5) refers to as the relationship between 'micromotives' and 'macrobehavior'.
Our micro motivational assumption will be that individual researchers attempt
to choose their research topics 'sensibly' - neither randomly nor omnisciently
- in pursuit of a bounded self-interest. This assumption, we will suggest,
leads directly to an important and unfortunate set of implications for the
development of specialties: that there exist important pressures which lead to
undue persistence of individuals in some research trails rather than others;
that the social processes associated with the development of these trails tend
toward conservative pressures for intellectual continuity on new entrants; and
that the aggregate result of these processes as is that, far from a wide dispersion
of research effort around the boundary problems of a specialty, there will be
unproductive over-concentration on some few problems, while high-potential
areas go underdeveloped (6).
The present relevance of specialties is their thoroughly relativized definition (7). Once relativized, they are easily reified. To posit their existence and
classify researchers relative to them does not make them real. Yet as Kuhn (8)
intimated, and several others (9) have demonstrated, innovations in a science

Research Trails and Science Policies

295

often come from outside its perceived boundaries. Edge and Mulkay (10) call
this 'marginal innovation'. But what is crucial is the perception that a novel
idea and its bearer are marginal to a particular specialty. If we remember that
novel ideas are routinely transmitted via the published literature, preprints,
and the spoken word, then the analytically decisive process may be the
transi tion of a researcher from one specialty to another.
Despite the fluidity of specialty defmition (11), the migration of researchers, we would argue, is a discernible process which results in the exploitation
of problems. The attraction of migrants may indeed identify a specialty as
'hot', though policy -makers with discretionary resources to allocate are likely
to foster this attraction, and once fostered, to reinforce it (12). Relativism
notwithstanding, then, the construct 'scientific specialty' implies extra-local
intellectual linkages amongst an aggregate of researchers at some point in
time.
Alternatively, a specialty can be conceptualized as the confluence of
several research trails, each representing a sequence of work by an individual
or a small team of researchers. These research trails may be distinguished by
some continuity of focus - be it methodological-theoretical and/or problemoriented - in published research (13). A research trail thus directs attention
to the coherence and development over time of a series or program of research
projects undertaken by an individual or group (the latter often consisting of
local colleagues), while the notion of a specialty directs attention to the
coherence amongst several such trails at a particular point in time. As in the
cable bundles found in complex electrical devices, a particular research trail
may be part of one bundle or specialty for a period of its history, then branch
off to become part of another bundle, and so on. A slice through the bundle
at some point reveals some semblance of the membership and interrelationships between the current members; following each wire from its source to
its destination shows the linkages over time within one single trail. The
'res.earch trail' notion, then, emphasizes the development over time of the
activities of an individual or team of researchers.
If a research trail corresponds to a body of work at the level of a 'problem'
or 'problem area', then a specialty is an aggregation of related problem areas
(14) defined retrospectively. It is the reciprocal and relative nature of trails
and specialties which render them empirically 'slippery' but analytically
fruitful (see below). That is, we observe a distribution of research effort

296

Daryl E. Chubin and Terence Connolly

cross-sectionally, but reconstruct the concentration of such effort - by

individual, organizational site, and problem area - longitudinally. Having
suggested the heuristic value of these constructs, we now seek to apply them
in demythologizing the processes of creative work, problem choice, and
successful claims-making in science.
3. Local Tinkering and Research Trail Development
'Tinkering' has been discussed by Knorr as the 'opportunism' of a process
which produces distinctive goals: successful research products. Local opportunities form the basis for investment, and "are subject to modification
through interpretation, negotiation, or straightforward 'misuse'" (15).
How an individual decides to initiate, persist in, or terminate work on a
certain research trail is a question of 'local tinkering'. The investments made
by an agent in a local site, i.e., the research organization, set the conditions
of production; opportunities for success, operating as a driving and orienting
force, organize those investments around specific topics (16). Following the
economic imagery of Knorr, and Bourdieu (l7) before her, we postulate
a coarse incremental cost-benefit model: the researcher will undertake the
first (or next) study in a trail if she sees the benefits of doing so to exceed
the cost. Without pretending to any special insight into how these costs
and benefits might be assessed, at least the following factors need to be
considered.
(1) The scientific yield of each study is unlikely to be constant across all
studies in a trail. At least three patterns of yield may be distinguished: those
in which the early studies show a high yield, followed by a decline later in
the trail; those in which the early studies are relatively unproductive, but
with relatively greater yield later in the trail; and those in which there is little
yield at any point in the trail (Le., bad ideas or low probability of success).
Obviously, the form of these yield curves is not known with certainty a priori.
The researcher must decide to pursue a trail with considerable uncertainty as
to its fmal yield; and the uncertainty is greater for studies further down the
trail than for the next one or two. It is this rising uncertainty that leads us
to postulate incremental assessment: the yield of the next study may be
reasonably clear, while that of a distant study is quite uncertain, so that the
researcher's planning horizon is necessarily quite restricted.

Research Trails and Science Policies

297

(2) The costs to the researcher will generally tend to peak early in the
research trail, declining to a much lower, and roughly constant, level thereafter. A number of familiar items are included in this non-recurring initial
set-up cost: time and effort involved in reviewing the relevant literature;
learning the techniques of research in the area chosen; acquiring suitable
equipment, instruments, colleagues and facilities; securing research support
as an outsider with no prior track-record of performance in the area; and so
on. Once incurred, these costs generally do not continue, but drop quickly
to a much lower level.
(3) Payoffs to the researcher are not limited to intellectual yield alone. We
assume that publication is positively valued by most researchers, even if
intellectual yield is modest. Further, an additional publication within an
established research trail has the effect of maintaining interest and visibility
- a veneer of success - for the body of work as a whole, so that connected
publications may be valued more highly than are isolated ones. Finally,
we suggest that research trails rarely meet an unambiguous dead end; they
typically peter out in studies of small impact or marginal variation.
Any formal consideration of .the combined effect of these factors would
suggest that investments in research topics, i.e., problem choice, involve
elaborate calculations. We postulate no such elaborate choice process. Rather,
as Knorr observes, "opportunities for success emerge routinely and naturally
from the flow of research" (18). This keeps the risks, or at least the team's
calculation and perception thereof, low. In Knorr's words, "once made,
investments tend to stabilize the effort" or to generate experiments that are
"on the safe side" (19).

4. Evolutionary Theory, Tinkering, and Trails

The 'tinkering' imagery can be pressed further in the present context of an
evolutionary view of scientific development. As Jacob (20) asserts, the
common analogy between natural selection and engineering design is a
misleading one, in that while a competent engineer works to a preconceived
plan, uses specialized materials and tools, and (at least ideally) achieves results
close to the limits of available technology, none of these features should be
assumed for an evolutionary process. Rather, the results of the latter are more

298

Daryl E. Chubin and Terence Connolly

akin to those of 'a tinkerer who uses everything at his disposal to produce
some kind of workable object' (21).
What Jacob's imagery connotes may well appear more descriptive of
the basement inventor than of the gleaming equipment of the professional
researcher. But if we consider the imagery as describing the intellectual,
rather than the physical, equipment of the researcher, the picture is more
vivid. At any point in time, a researcher has available a stock of intellectual
'odds and ends' - techniques, conceptual notions, insights, and intuitions generated by the training and research trails in which she has participated to
date. Further, these 'odds and ends' are not merely free-floating entities,
open to recombination into whatever intellectual objects the researcher may
choose. They possess a strong measure of internal coherence. Just as complex
eyes evolved through a series of simpler, but still functional, primitive photoreceptors (22), so scientific novelty is 'tinkered' onto the end of previous
functional constructs. The history of the research trail, in short, places
powerful constraints on where it may next develop.
While admitting their power, we cannot allow these constraints to be
absolute. Novel tinkering is always possible; and the form of the new intellectual objects is shaped by a second set of constraints, arising outside the
intellect of the individual researcher. It is these external environmental forces
that shape the social production of scientific work, that select from the many
potential novelties those few that will become 'real science'. Of course, the
results of this interaction at a given time will affect the materials available for
'tinkering' at some later time. In this sense, we embrace both the 'internalist'
and the 'externalist' views of scientific development. The internalist view,
here represented by the scientist's perception of coherence within a research
trail, sets limits on the range within which tinkering is possible; the externalist
view stresses the selection, by forces external to the individual scientist, of
the novelties that will be retained and those that will be discarded. Thus,
with evolutionary theory we embrace a contextual view of scientific work
(23).
We hasten to note, however, that the internalist-externalist issue is closely
mirrored in the recent re-emergence within evolutionary theory itself of the
debate between the 'adaptationist' and 'structural integration' traditions.
In Gould's (24) sketch of the extreme positions, one sees the direction of
evolution as channeled "by adaptive requirements of local environments,"

Research Trails and Science Policies

299

while the other stresses the role of "the nature of variation and the morphology of the system in which it arises" (25).
Gould's comments on the differences in emphasis between the two programs are sharply relevant to our concerns here. Natural selection (externalist
pressures) are, of course, the fundamental engine of change. However,
the possible routes of selection are channeled by inherited morphology, building material,
and the amount and nature of variation itself. Though selection moves organisms down
the channels, the channels themselves - rather than the paucity of weU-designed outcomes - impose primary constraints on the direction of change .... Selection on one
part of a structure may impose a set of correlated and nonadaptive modifications on
other parts of an integrated body. Many features, even fundamental ones, may be
nonadaptive (though not, to be sure, strongly inadaptive) either as developmental
correlates of primary adaptation or as 'unanticipated' structural consequences of primary
adaptations themselves (26).

The direction of development, in short, is shaped by both constraints and

consequences, by both the internal logic of the research trail and by the
external forces of opportunities and rewards flowing from the larger environment.
Our contextual argument, therefore, stresses that in the scientific ecology
there may be positive feedback relationships between movement down a
channel (research trail development) and subsequent availability of support
from the environment. Success breeds success, selection becomes 'unnatural'.
In terms of our incremental cost-benefit model, once launched on a research
trail, a researcher or team is likely to persist in it well beyond the point at
which additional studies can be justified by their scientific yield. If research
costs decline rapidly after the first few studies, publication is valued for its
own sake as well as for the size of the contribution reported, and scientific
yield invariably reaches a point in the trail at which incremental yield becomes small, then undue persistence in worked-out, low-yielding research
trails is the likely consequence. Undue persistence becomes a corollary of
modest local tinkering (27). The recognition of such tinkering as novel (Le.,
'successes' within a specialty), and the subsequent investments which accrue
to local sites, nevertheless remain in doubt. For as Whitley rightly asserts:
Once a particular view ... has been institutionalized and legitimate topics and approaches
specified, reformulations and novel developments will be strenuously opposed since they
imply a reordering of priorities and threaten the existing distribution of expertise and
property rights (28).

300

Daryl E. Chubin and Terence Connolly

It is this reordering of research priorities which engages local sites in a com-

petitive struggle for resources and acclaim with various hierarchies - local and
extra-local alike. What we now consider is how these hierarchies, especially
through policy intervention, impinge on problem choice, thereby locking
researchers into and out of research trails.
5. Trails and Hierarchies: Research Persistence and Policy Pressures
The investment model sketched above proposes that two contexts - local
and extra-local (e.g., national policies, international specialties) - structure
opportunities for the pursuit and abandonment of research trails. A retrospective assessment of opportunities and successful local tinkering with them
may be equated with the aggregate growth of specialties and the knowledge
claims therein. However, the amenability of problems to policy-manipulation,
goal-direction and eventual translation of research into utility can be questioned (29). Is certain knowledge resistant to manipulation? Must theories be
'finalized' before they can become policy tools? Does political action - an
attempt to use scientific knowledge - vindicate its claim to truth? Such
questions can be construed as investment rationales. They are part of the
rhetoric that policy-makers and researchers use to justify the funding support
they, respectively, give and receive. Such 'vocabularies of justification' (30)
vary with the context, the audience, and the research problem in question.
Presently, the two contexts of analytical concern are the local research site
and the wider social and intellectual environment in which knowledge claims
are negotiated with political, professional, and lay audiences. The viability
of research trails can be seen as an outcome. of such negotiation. Prior to
it, however, we may ask what minimal conditions must be satisfied for a
researcher or team to pursue a trail.
5.1. Conditions Favoring Pursuit ofa Trail

Four conditions, each with local and extra-local repercussions, can be distinguished as predisposing a researcher to pursue a research trail: legitimacy,
funding, access to local resources, and training capacity. To be faithful to our
relativistic argument, we qualify these conditions as being peculiar to postWorld War II U.S. science.

Research Trails and Science Policies

301

(i) Legitimacy. While legitimacy may well be considered a 'resource' in the

broad sense of that word, it is especially vital for commitment to a research
trail. As has been noted for 'orthodox' as well as 'deviant' sciences, gaining
approval for a given line of inquiry is likely to be problematic for any novel
research effort (31). 'Paradigms' and 'research traditions' (32) circumscribe
definitions of acceptable novelty rather narrowly. These definitions become
institutionalized so as to encompass established trails. Not only are papers
published, but thematic journals may exist. Nor only are research proposals
funded, but funding programs may be created within a federal agency expressly for the purpose. Specialized training programs are likewise founded
at local sites to produce 'specialists' (see below).
Thus, both the individual researcher and those powerful gatekeepers of
the relevant hierarchies who control funding, publication, and acclaim acquire
a reputational interest in the continuance of the trail. Indeed, an accumulation
of such interests lends legitimacy to the trail and advantages to prominent
researchers within it (33). While some compound this reputational capital in
a trail or related set, and simultaneously enhance the identification of the
local site with it, others try to expand their sphere of influence by transferring
their considerable reputational capital to a new problem, thereby imputing
legitimacy to it (34). Of course, even an eminent scientist incurs the risk
of transfer, failure, and loss of legitimacy, if claim to novelty falls beyond
acceptable bounds (35). Finally, for either a non-elite or an elite researcher
to terminate or withdraw from a research trail, might be seen as a confession
of prior errors of judgment, weakening career prospects in the one case, and
future control in the other (36).
(li) Funding. Intimately associated with the legitimacy of a trail is its funding.
Without funding, novel research cannot be exploited; indeed it may become
(un)known as a stillborn idea (37). A research proposal is most readily
prepared, and most readily accepted, in an area with which one is already
most familiar - which is likely to be the next study in an established research
trail. Reviewers are directed to look favorably on established track records,
as on established local facilities and equipment (38). Similarly, conference
organizers, symposium editors, and other conferers of visibility are likely to
seek out researchers known to be established and competent in given fields,
providing both social and intellectual support for those in these fields. For

302

Daryl E. Chubin and Terence Connolly

established participants, flourishing secondary occupations emerge in the

form of internecine disputes on minor variants, reviews of existing research,
refereeing of one another's papers, and so on. Indeed, as many of us have
found, a considerable effort is required to get out of any particular research
trail, once one has done the first two or three studies in it. Again, a pragmatic,
two-6ided incentive structure is at work.: the conference organizer needs
presenters of established competence and credibility as much as the researcher
needs the rewards of participation; the intellectual merit of the materials
presented may, in the limit, become almost irrelevant to the satisfaction of
these needs.
(iii) Access to resources. Our investment model supposes that 'a researcher
embarked on a given trail will have access to the equipment,research facilities,
and collegial climate necessary to capitalize on locally-perceived opportunities.
If such opportunities are either perceived as blocked locally, or more readily
available elsewhere, a change in organizational site may be necessary. If such
a change is infeasible (Le., the inchoative trail does not match existing local
conditions and investments), the research will be deferred and/or pursued on
a reduced scale with non-local colleagues. We might call such deferred or
non.,local research a 'latent trail'. Among active researchers who collaborate
in most or all of their work, such trail-6pecific cooperative arrangements with
fellow specialists may be common (39). For most local tinkering, however,
the incremental costs of the next study in the trail are likely to be small.
With most of the intellectual and material necessities in place, only small
increments of effort and money are required for the next study. Thus, risks
are small and modest success is all but assured. The productivity and visibility
of the team and its members are preserved (40).

(iv) Training capacity. One resource that warrants separate discussion is the
training capacity of the organizational site. Academic labs typically serve as
training and certification sites for neophyte scientists. As several authors have
observed, the presence of such pre- and post-doctoral personnel enhances
research continuities (41). Put another way, the practices of training and
entry to research careers will tend to exacerbate the tendencies to undue
persistence in worked-out or low-yielding research trails. Graduate students,
for example, are likely to seek dissertation advisors who can provide well-

Research Trails and Science Policies

303

specified projects, with a high probability of successful completion and

publication. Upon receipt of the PhD, the new entrant is faced with an urgent
need to establish a reputation and publication record within local promotion
and tenure time-frames. Again, both participants gain from the arrangement:
the professor maintains his reputation by further work on a problem which
he pioneered earlier, keeping alive a research trail; the student benefits by
the safety and legtimacy of competent extension and publication, carrying
forward the banner of the professor's seminal work. Of course, such symbiosis
is unlikely to persist indefinitely, as the neophyte breaks mentor ties (42),
perhaps even repudiating the trail(s} blazed by the former teacher-collaborator
(43). Nevertheless, the modal phenomenon will be undue persistence in a trail
or cluster of related trails.
(v) Residual conditions. Oearly, the four conditions on which we have
dwelled do not exhaust the factors which favor persistence in, or abandonment of, a research trail. Old researchers and obsolete research ideas do, eventually, fade away. More common, however, is the perpetuation of a trail, or
at least a research tradition, by new generations long after the disappearance
of its founder(s}. Within such traditions lurk 'dormant' or 'cold' trails which
may be renewed even in the absence of funding, legitimacy, etc. A residue
of intellectual resources might, at some local site, be converted into an
opportunity for exploitation. Given the age-stratification within science, a
rediscovery of premature ideas or reopened interest in an old problem might
prompt return to a cold trail - though the ahistorical bent of most scientists
and the momentum of trail development imply that such revivals will be
rare.
In sum, we would hypothesize that increasing tedium will drive researchers
out of their trails at some point (though whether they then take up other
trails or leave research altogether is unclear). As relevant hierarchies and their
elites (e.g., funders and editors) decide that a trail has been worked out oris
dead-ended, they will gradually withdraw their support. Young researchers
with reputations to make or skills to market in a shrinking marketplace may
take the dangerous plunge into a new research trail. Persistence, in short, is not
perpetual; rather, it is undue. At any given time, then, a specialty will tend to
a state of unbalanced development with old trails excessively developed, and
new trails inadequately explored. Where Price sees problem choice as random,

304

Daryl E. Chubin and Terence Connolly

and Hagstrom sees it as naturally dispersing, we see it as excessively concentrated. We also see it - problem choice that begets trails which, in turn,
converge to form specialties - as extraordinarily sensitive to local exigencies
and extra-local pressures that must be socially negotiated to ensure intellectual viability.
5.2. Political Pressures and the Evolution of Specialties

In retrospect and at an aggregate level, research trails sustain scientific specialties. Specialties envelop a set of research problems and a cadre of researchers
whose local tinkerings attract national investments. Science policies and
funding programs are formulated to rationalize those investments. Investments are made on a competitive basis and mediated through mechanisms
such as peer review. Scientists, as appointed government advisors, erstwhile
consultants, and proposal reviewers, playa collusive role in defining 'scientific
merit', encouraging certain research trails and creating 'hot' specialties. This,
too, is 'interpretive' science practice. For policies make a difference - both
in scientific careers and in the evolution of those knowledge-producing
communities we call specialties. And scientists - not all, but the advisingconsulting-reviewing elite - mediate those policies and the funding decisions
they warrant (44).
The elite (which is demographically older, more productive, and located at
more prestigious sites than the community at large (45)), is in the advantageous position of helping both to dispense and to receive extra-local resources.
The elite legitimates and funds. like all gatekeepers these elite scientists wield
power by allocating and withholding both the approval and the information
that is construed as 'specialty growth and decline'. The point is this: with
such a concentration of power among a relative few, the scientific elite plays
a functional role in th::: amelioration of political pressures and the evolution
of research. This elite represents, indeed is employed by, an interlocking set
of institutions which, at least in the U.S., garner the lion's share of available
resources, students, and faculty talent (46). Through reputation and iocation
alone, this elite can direct or redirect resources so as to affect the activity
perceived to be within a specialty and the subsequent viability (through
local project funding and pUblication) accorded it. The elite can, in short,
determine what problems are chosen and whose trails are sustained.

Research Trails and Science Policies

305

Such (re)direction imbues research goals and science policies with a fluid,
i.e., 'negotiated', character. Scientists may neither originate nor ostensibly
satisfy these goals and policies, but they intervene in significant ways. Just
as local teams generate idiosyncratic interpretations of literature and ongoing
research, extra.J.ocal hierarchies and elites operate so as to shape policyformulation, -implementation and -evaluation. Their role, while advisory, is
manipulative and selective; it is not purposefully or systematically conspiratorial. But in the evolution of a specialty, elites wield a disproportionately
large influence relative to their number or the richness of their research
tradition(s) (47).
Consider this in the perspective of evolutionary theory, as discussed earlier.
In this theory there exists a clear tension between processes that lead to
variation and those that lead to retention. A reproductive system that allows
no variation is doomed, because there is no chance of increasing adaptation
to the species' environment. Conversely, inadequate retention processes imply
that the selective filter is insufficiently sharp, and that unsuccessful as well as
successful variations are retained, with no overall gain in adaptation. A species
that produces only mutations is as doomed as a species that produces none
(48).
In the 'evolutionary-investment' model we have outlined, a scientific
specialty that retains only its currently-active research trails is doomed to
extinction; as the trails become exhausted, the practitioners move on, and
new entrants grow scarce. Conversely, a specialty which is excessively open
to novelty fails because of a lack of consensus as to what is known, what is
worth knowing, and what knowledge-generating practices are acceptable;
coherence is lost, faddism reigns, and the intellectual development of the
specialty is foreclosed. Therefore, elites must exercise some selectivity. But
as our case for undue persistence attests, the balance between variation and
retention favors the latter. Each study is likely to be all too similar to its
predecessors in the research trail; promising alternative trails are li1,{ely to be
ignored or inadequately explored; and the specialty drifts into pre-revolutionary rigidity, from which crises - the resistance of elites defending particular
orthodoxies and their status prerequisites (advisory and gatekeeping) notwithstanding -lead to radical restructuring of the lines of inquiry.

306

Daryl E. Chubin and Terence Connolly

6. Concluding Thoughts on Policy Implications

As our analysis has shown, scientific hierarchies and elites are eventually
implicated in the mechanisms leading to the unproductive over persistence
in research trails. The challenge to science policy-makers is to find ways
to offset, rather than exacerbate, the destructive feedback loops we have
outlined. Examination of the role of elites in mediating between science
policy, at the national level, and individual problem choice, at the local level,
underlies this challenge.
In an era of problematic support for science, meeting the challenge becomes even more crucial to the overall health of the scientific enterprise. For
example, there is (at least in the U.S.) considerable variation between funding
agencies in the distribution of resources across novel and established research
areas, between promising neophytes and established veterans, and in readiness
to terminate old research trails in favor of seeding speculative novelties. Are
these variations mirrored in the rates of turnover of the relevant elites? In the
productivity of the work funded? In the epistemological allegiances of the
of the specialties supported?
members? Or in the social
The crucial implication of our model - the phenomenon of undue persistence - represents a huge potential waste of scientific resources. The
amelioration of such waste, and the revivifying of stagnant areas of inquiry,
seems to us a high aspiration for the student of science practice. Thus, in one
sense, we are proposing a developmental route allowing enough flexibility
for specialties to evolve; in another sense, we find ourselves advocating
mechanisms for the disestablishment of elites, for the loosening of the bonds
that tie researchers to waning lines of inquiry.

Notes and References

1. For the present, we ignore the questions of how one delimits the boundaries and
membership of a specialty, although these relate centrally to the role of establishments in ordering and validating areas of discourse, defining permissible treatment
of topics within them, and more generally, favoring particular forms of organizing
intellectual work. For characterizations of how establishments and elite figures
influence the definition and development of scientific specialties, see Daryl E.
Chubin, The Conceptualization of Scientific Specialties', The Sociological Quarterly
17 (1976) 448-76; D. O. Edge and M. J. Mulkay, Astronomy Transformed: The
Emergence of Radio Astronomy in Britain, New York: Wiley, 1976, Chapter 10;

Research Trials and Science Policies

2.
3.
4.

5.
6.

7.
8.
9.

10.
11.

12.

307

David O. Edge, 'Quantitative Measures of Communication in Science: A Critical

Review', History of Science 17 (1979) 102-34; and Norbert Elias, this volume.
W. O. Hagstrom, The Scientific Community, New York: Basic Books, 1965, p. 222.
Derek de Solla Price, Little Science, Big Science, New York: Columbia University
Press, 1963, pp. 65 f.
R. K. Merton, 'Priorities in Scientific Discovery: A Chapter in the Sociology of
Science', American Sociological Review 22 (1957) 635-59; R. K. Merton, 'Singletons and Multiples in Scientific Discovery: A Chapter in the Sociology of Science',
Proceedings of the American Philosophical Society 105 (1961) 470-86.
Thomas C. Schelling, Micromotives and Macrobehavior, New York: Norton, 1978.
We are indebted to Stephen Kerr, of the University of Southern California, for his
lighthearted conjecture (made in a private communication to Terry Connolly in
1978) that there exists a 'Law of Perpetuallnquiry' in research trails, a law driven
by mechanisms roughly parallel to those proposed below.
S. W. Woolgar, 'The Identification and Defmition of Scientific Collectivities', in
Gerard Lemaine et al. (eds.),Perspectives on the Emergence of Scientific Disciplines,
Chicago: Aldine, 1976, pp. 238--45.
Thomas S. Kuhn, The Structures of Scientific Revolutions, Chicago: University of
Chicago Press, 1962.
For example, D. A. Schon, Displacement of Concepts, London: Tavistock, 1963;
M. J. Mulkay, 'Conceptual Displacement and Migration in Science: A Prefatory
Paper', Science Studies 4 (1974) 205-34; Robert Olby, The Path to the Double
Helix, Seattle: University of Washington Press, 1974; K. E. Studer and D. E. Chubin,
The Cancer Mission: Social Contexts of Biomedical Research, Beverly Hills: Sage,
1980, chapters 2, 5 and 6.
Edge and Mulkay, op. cit., note 1, p. 382.
Opting for more content"",pecific, non-institutionalized defmitions than 'specialty',
Gernot B6hme ('The Social Function of Cognitive Structures: A Concept of the
Scientific Community within a Theory of Action', in K. D. Knorr et af. (eds.),
Determinants and Controls of Scientific Development, Dordrecht, D. Reidel, 1975)
speaks of 'cognitive regions', while D. E. Chubin and K. E. Studer ('Knowledge and
Structures of Scientific Growth: Measurement of a Cancer Problem Domain',
Scientometrics 1 (1979) 171-93) speak of 'problem domains'. Note that as definitions change, the number of 'migrants' fluctuates. Note, too, how content-centered
defmitions relate to 'restrictedness', as discussed in Richard Whitley, The Sociology
of Scientific Work and the History of Scientific Developments', in the S. S. Blume
(ed.), Perspectives in the Sociology of Science, Chichester: Wiley, 1977, pp. 21-50,
and by Arie Rip in this volume, and to the concentration of intellectual control
among establishment researchers who can impose the greatest precision on objects
for manipUlation and explanation.
Daryl E. Chubin, Social Trappings of Knowledge, unpublished book manuscript,
1981, chapter 2, subsumes this under 'mission-making'. W. van den Daele and
P. Weingart, 'Resistance and Receptivity of Science to External Direction: The
Emergence of New Disciplines under the Impact of Science Policy', in Lemaine
et al. (eds.), Perspectives on the Emergence of Scientific Disciplines, Chicago:
Aldine, 1976, pp. 247-75, call this a 'relabeling' phenomenon that triggers a
bandwagon effect. For evidence that scientists' research 'follows the money', see

308

13.

14.

15.

16.

17.
18.
19.
20.
21.
22.
23.

24.
25.

Daryl E. Chubin and Terence Connolly

Studer and Chubin, op. cit., 1980 (Note 9), chapter 3; M. Useem, 'Government
Influence on the Social Paradigm', The Sociological Quarterly 17 (1976) 146-61;
and Edward Yoxen, this volume.
For a similar usage of a pUblication series as a unit of analysis, see Diana Crane,
Invisible Colleges, Chicago: University of Chicago Press, 1972.
M. J. Mulkay, G. N. Gilbert and S. Woolgar, 'Problem Areas and Research Networks
in Science', Sociology 9 (1975) 187-203; also see R. Whitley, 'Cognitive and Social
Institutionalization of Scientific Specialties and Research Areas', in Richard D.
Whitley (ed.), Social Processes of Scientific Development, London: Routledge and
Kegan Paul, 1974, pp. 69-95.
Karin D. Knorr, 'Producing and Reproducing Knowledge: Descriptive or Constructive?' Social Science Information 16 (1977) 677. Barry Barnes' (Scientific Knowledge and Sociological Theory, London: Routledge and Kegan Paul, 1974, p. 58)
discussion of the scientist as 'bricoleur' is relevant to the notion of tinkering as
creating opportunism.
Bruno Latour and Steve Woolgar, Laboratory Life: The Social Construction of
Scientific Facts, Beverly Hills: Sage, 1979; K. D. Knorr and D. W. Knorr, 'From
Scenes to Scripts: On the Relationship between Laboratory Research and the
Published Paper in Science', Social Studies of Science 11 (1981), forthcoming; also
Shinn, this volume.
P. Bourdieu, 'The Specificity of the Scientific Field and the Social Conditions of
the Progress of Reason', Social Science Information 14 (1975) 19-47.
Knorr, op. cit., 1977 (Note 15), p. 680.
Ibid., pp. 680-81.
Francois Jacob, 'Evolution and Tinkering', Science 196 (1977) 1161-6.
Ibid.
T. Dobzhansky, Evolution, Genetics and Man, New York: Wiley, 1963.
The hazards of applying evolutionary theory to scientific work and intellectual
change are outlined in L. J. Cohen's essay review of Toulmin's Human Understand
ing, 'Is the Progress of Science Evolutionary', British Journal for the Philosophy of
Science 24 (1973) 41-61. In what follows, we have tried to heed Cohen's critique
of Toulmin's 'evolutionist philosophy of intellectual history' without succumbing
either to the pitfalls of the Darwinian metaphor as it describes a population of ideas
and the scientists who advance them or to Cohen's niggardliness in demanding more
than a parallel between the development of a scientific discipline and the evolution
of a biological species. Also see Ron Johnston, 'Contextual Knowledge: A Model
for the Overthrow of the Internal/External Dichotomy in Science', Australian and
New Zealand Journal of Sociology 12 (1976) 192-203.
Stephen Jay Gould, 'The Evolutionary Biology of Constraints', Daedalus 109
(1980) 39-52.
In an intriguing echo of our current concerns, Gould (Ibid., p. 40) notes in passing
that while "no one would, of course, embrace either extreme position in its entirety" the choice of one or the other as a starting point for inquiry critically shapes
the course of study. The exfoliation of the two research trails has been independent
to the point at which structural integrationist notions are almost entirely European,
adaptationist models almost entirely American.
Additional details from the evolutionary debate might clarify the parallelism we

Research Trials and Science Policies

26.
27.

28.
29.

30.

31.

309

see here. The notion of a 'morphological space' is frequently used in evolutionary

analysis to indicate a bounded region of alternative organismic forms. In some
simple cases, this space can be reasonably well specified. Raup (cited in Ibid.), for
example, shows that a three parameter space suffices to specify the alternative
morphologies of coiled shells (snails, clams, brachipods, and so on). A plot of
extant shell species into this space reveals extreme variations in density: some
regions are extremely well populated, others virtually empty. This variation is
explained quite differently by the two alternative evolutionary views, adaptationists
stressing the external logic of environmental fit, the structural integrationists
emphasizing the internal logic of architecture and history. Predictably, the resulting
debate tends more towards an 'arguing past' than an 'arguing with' format.
Ibid., p. 44.
Seizing upon certain opportunities and passing on others attributes to the researcher a calculation and success that may be entirely bogus. Such 'bogus wisdom'
is another myth attached to individuals, teams, and whole organizations via 'rational
reconstructions'. Again we emphasize the relativism of such reconstructions of local
tinkering, research trail development, and the ultimate recognition of successes. The
term 'bogus wisdom' derives from Karl W. Deutsch and William G. Madow, 'A Note
on the Appearance of Wisdom in Large Bureaucratic Organization', Behavioral
Science 6 (1961) 72-8; its application to scientists' careers is explored in S. J.
Turner and D. E. Chubin, 'Change and Eminence in Science: Ecclesiastes II', Social
Science Information 18 (1979) 437-49.
Whitley,op. cit., 1977 (Note 11), p. 32; also see Chubin, op. cit., 1976 (Note 1),
pp. 459-65. For case evidence of such opposition, see Fleck, this volume.
See, for example, Johnston, op. cit., 1976 (Note 23); G. Bohme, W. van den Daele
and W. Krohn, 'Finalization of Science', Social Science Information 15 (1976)
306-30; Michael J. Mulkay, 'Knowledge and Utility: Implications for the Sociology of Knowledge', Social Studies of Science 9 (1979) 63-80; Gerard Lemaine,
Roy MacLeod, Michael Mulkay and Peter Weingart, 'Problems in the Emergence
of New Disciplines', in Lemaine et al. (eds.), Perspectives on the Emergence of
Scientific Disciplines, The Hague and Paris: Mouton, 1976, pp. 1-23; Studer and
Chubin,op. cit., 1980 (Note 2), epilogue; Hohlfeld, this volume.
The rhetorical dimension of scientific practice is addressed in Michael J. Mulkay,
'Norms and Ideology in Science', Social Science Information 15 (1976) 637-56;
Michael A. Overington, 'The Scientific Community as Audience: Toward a Theoretical Analysis of Science', Philosophy and Rhetoric 10 (1977) 143-64; D. E.
Chubin and K. E. Studer, 'The Politics of Cancer', Theory and Society 6 (1978)
55-74, illustrate the use of rhetoric involving congressional testimony over the
U.S. war on cancer.
For so-called marginal sciences, e.g., astrology, mesmerism, and parapsychology,
approval is withheld by the orthodox sciences. See R. Wallis (ed.), On the Margins
of Science: The Social Construction of Rejected Knowledge, Staffordshire: University of Keele, 1979. The Laetrile movement is another example of a social movement that has been denied scientific and medical legitimacy. The political and
medical establishment views it as 'counter-cultural' and 'anti-scientific'; see James
C. Petersen and Gerald E. Markle, 'Politics and Science in the Laetrile Controversy',
Social Studies of Science 9 (1979) 139-66; James c. Petersen and Gerald E. Markle,

310

32.
33.
34.
35.

36.

37.

38.

39.
40.
41.
42.
43.

Daryl E. Chubin and Terence Connolly

"The Laetrile Controversy', in D. Nelkin (ed.), Controversy: Politics of Technical

Decisions, Beverly Hills: Sage, 1979, pp. 159-79.
Kuhn, op. cit., 1962, Note 8; L. Laudan, Progress and Its Problems, Berkeley:
University of California Press, 1977.
See R. K. Merton, "The Matthew Effect in Science',Science 159 (1968) 56-63.
See Chubin, op. cit., 1976 (Note 1), p. 471. Elsewhere in this volume, Yoxen and
Fleck, respectively, present evidence of two such attempts to gain legitimacy.
For example, Nobel laureate Albert Szent-Gyorgyi ('Albert Szent-Gyorgyi, Electrons and Cancer', Science 203 (1979), 522-3) suffered the indignity of repeated
rejection of his research proposals on the 'bioelectronic' theory of cancer by the
U.S. National Cancer Institute. His advanced age was also cited as a liability since,
in our terms, death might doom work in a promising trail if no successors are
willing or capable of pursuing it. As Daniel S. Greenberg ('Compared with Scientists,
Artists and Babes in Grantland', Science and Govemment Report 6, 1980) puts it,
" ... the peer review system is an inherently bland responsibility-diluting method
that generally tends to caution and conservatism in giving away money.... The less
money that's available for giving the less the spirit of adventure .....
An intriguing line of research in social psychology has recently examined the
mechanisms of such self-justifying continuation in low-yielding courses of action.
See B. M. Staw and J. Ross, 'Commitment to a Policy Decision: A Multi-theoretical
Perspective', Administrative Science Quarterly 23 (1978) 40-64.
A recent report on 'twin research' (Constance Holden, 'Identical Twins Reared
Apart', Science 207 (1980) 3123-8) indicated that no institute or study section of
the U.S. National Institutes of Health was considered the appropriate funding home
of such a mUlti-<iisciplinary project. The National Science Foundation finally provided support.
S. Cole, L. Rubin and J. Cole, Peer Review in the National Science Foundation:
Phase One of a Study, Washington, D. C.: National Academy of Sciences, 1978.
For a contrasting view, see I. I. Mitroff and D. E. Chubin, 'Peer Review at NSF:
A Dialectical Policy Analysis', Social Studies of Science 9 (1979) 199-232; D.
E. Chubin, 'Competence Is Not Enough: Essay Review of Cole et al. 's Peer Review
in the National Science Foundation', Contemporary Sociology 9 (1980) 2047.
See Knorr op. cit., 1977 (Note 15), p. 680; also see Chubin, op. cit., 1976 (Note 1),
pp.466-8.
.
K. D. Knorr and R. Mittermeir, 'Publication Productivity and Professional Position:
Cross-National Evidence on the Role of Organizations', Scientometrics 2 (1980)
95-120.
See especially, J. Ravetz, Scientific Knowledge and Its Social Problems, Oxford:
Oxford University Press, 1971;Harriet Zuckerman,Scientific Elite: Nobel Laureates
in the United States, New York: Free Press, 1977; Whitley, op. cit., 1977, Note 11.
T. S. Kuhn, "The Essential Tension: Tradition and Innovation in Scientific Research', in C. W. Taylor and F. Barron (eds.), Scientific Creativity: Its Recognition
and Development, New York: Wiley, 1963.
D. Chubin, op. cit., 1981 (Note 12), chapter 3, asserts that such repudiation must
have occurred in the numerous cases of Nobel laureates who themselves were
trained by laureates.

Research Trials and Science Policies

311

44. See M. J. Mulkay, 'The Mediating Role of the Scientific Elite', Social Studies of
Science 6 (1976) 445-70; J. J. Salomon, 'Science Policy and Its Myths: The
Allocation of Resources', Public Policy 20 (1972) 1-33.
45. L. Groeneveld, N. Koller and N. Mullins, The Advisors of the U.S. National Science
Foundation', Social Studies of Science 5 (1975) 343-54.
46. H. Zuckerman, 'Stratification in American Science', in E. O. Lauman (ed.), Social
Stratification: Theory and Research for the 19708, Indianapolis: Bobbs-Merrill,
1970, pp. 235-57.
47. According to A. W. Gouldner, 'Prologue to a Theory of Revolutionary Intellectuals',
Telos 26 (1976) 3 -36; also see Elias, this volume.
48. Yet, in W. Schafer's ('Finalization in Perspective: Toward a Revolution in the Social
Paradigm of Science', Social Science Information, 1979, 915-43) words, .. 'Finalist'
alternatives in science invert the sequence postulated by classical evolutionary
theory. Variation no longer precedes the selection of new mutants but results from
prior selection" (p. 918).

THE ESTABLISHMENT AND STRUCTURE OF THE

SCIENCES AS REPUTATIONAL ORGANIZATIONS

RICHARD WHITLEY
Manchester Business School, University of Manchester

1. Sciences as Reputational Organizations

Much work in the sociology of science has focussed on the identification and
development of scientific communities in the contemporary sciences in an
attempt to identify the social processes leading to the emergence of new
specialisms or disciplines (1). These units of scientific organization are often
seen as crucial to the development of scientific knowledge because they
combine the production and the evaluation of knowledge claims and so
control the direction of research. Insofar as. they are the exclusive units of
knowledge production and evaluation, scientific communities constitute the
sciences; what they, and only they, produce and certify is scientific knowledge. In this view, the sciences are a set of loosely connected and largely
autonomous groups of producers and consumers of knowledge which form
distinct "communities" around particular knowledge goals, control research
facilities which are widely available to certified practitioners, decide their own
priorities and procedures in relative isolation and are more or less equivalent
in terms of the prestige of their aims and achievements. Once established,
such groups follow similar patterns of knowledge development, as in the
Starn berg group's model for example (2), so that scientific change tends to
be located at the stage when new scientific communities emerge and form
distinct identities. Given these assumptions, it seems reasonable to concentrate on the social processes by which new communities develop if we are to
understand changes in scientific knowledge.
This rather undifferentiated view of the sciences, and the work of Kuhn
(3) from which it derived some justification, have been criticized by observers
who stressed the plurality of the sciences and the inadequacy of relying on

313
Norbert Elias, Herminio Martins and Richard Whitley
Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 313-357.
Copyright 1982 by D. Reidel Publishing Company.

314

Richard Whitley

part of the history of physics for a model of knowledge development in all

fields (4). Recent studies of scientists at work and the social processes by
which controversies are resolved or set aside have further emphasized the
diversity of social processes in the sciences and the important influences of
employment organizations and local circumstances upon scientists' goals
and decisions (5). The diversity and plurality of patterns of change and
development in the sciences necessitate taking the domination of a particular
pattern of social organization and knowledge development, such as that of
physics, as historically contingent and sociologically problematic rather
than the inevitable outcome of epistemological assertions. Furthermore, the
employment of scientists as producers of original knowledge by organizations
such as universities and academies of science resulted in scientific work
becoming subject to the everday contingencies and exigencies of employment
organizations. The later employment of scientists in industrial and government laboratories and the production of scientific knowledge for a variety of
goals again demonstrated the importance of employment structures in the
organization and direction of scientific work (6).
The development of employment status for scientists oriented to the
production and validation of original knowledge reified the sciences as
reputational communities around particular goals and conceptual approaches.
By identifying intellectual goals and research procedures with academic
employment units through the dependence of university jobs and careers
on scientific reputations, the German university reforms in the early 19th
century institutionalized particular purposes, concepts and evaluation criteria
in pedagogic departments and fixed intellectual boundaries (7). The subsequent domination of the sciences by academics led to the identification of
scientific establishments with academic establishments as discussed by Elias
in his theme paper for this volume. However, this identification of science
with the universities was neither a necessary nor a universal feature of the
modern sciences, and the diversity of employment organizations conducting
scientific research today, coupled with the manifest differences between the
structure and organization of scientific fields, suggest that it would be useful
to conceive of the sciences as a particular kind of work organization which
can be, and has been, variously incorporated into employment organizations.
Such an approach would enable us to study different relations between the
sciences and employment organizations and their connections to the degree

Sciences as Reputational Organizations

315

of professionalization and autonomy of various fields. Changes in the degree

of centralization of authority, of standardization of technical procedures and
of specialization of topics and skills in different fields can also be the focus of
attention in this conception of the sciences.
The sciences form a subset of professional organizations, what might be
termed reputational organizations. That is, they are systems of work in which
practitioners control the way in which work is carried out and the goals for
which it is carried out in the light of the particular beliefs and purposes of
the reputational community of which they are members. Tasks are selected,
carried out and coordinated by scientists seeking reputations on the basis of
their contributions to the intellectual goals of the field. These goals differ
in their susceptibility to manipulation and determination by individuals
or groups of practitioners across the sciences, as does the extent to which
different groups are able to control the dominant definition and interpretation of the organizational goals. The extent to which a single scientific
establishment is able to dominate a given reputational organization thus also
varies between fields. This variation is, in tum, related to the centralization
of funding sources, autonomy from other organizations, standardization
of techniques and the existe!lce of a formal communication system for
comparing and coordinating task outcomes.
As a subset of professional organizations, sciences differ in the extent to
which they control jobs and other resources and the goals of employing
organizations. While control over work processes by workers is a general
feature of "craft" systems of administration (8), control over work goals
is the outcome of negotiations and bargains between professional groups
and employers. Where work flow and income flow is relatively unstable,
Stinchcombe suggests that fully "bureaucratic" systems of administration are
not as efficient as craft ones. In these latter systems, the day-to-day control
over how work is to be carried out is decentralized to practitioners whose
competence is certified by some professional agency, such as trade unions
in the construction industry. Authority over task definition, allocation
and performance is here shared between the employer and such agencies
because control over the production and certification of marketable skills is
monopolized by the profession which thereby reduces employers' powers (9).
In these craft systems of work administration, status in the labour market
as a certified possessor of definite skills is more important than status in

316

Richard Whitley

employment organizations; indeed, the latter is largely dependent on the

former in the most professionalized work organizations.
The degree of professionalization of work organizations is here understood
as the extent to which control over work goals, in' addition to control over
work processes, is exercized by some professional agency through its monopoly of the production and certification of particular skills. The more work
goals and assessment of task performance are organized in terms of practitioners' reputations in the professional community rather than by their employers, the more they are professionalized. So, reputational organizations in
which job statuses and careers are determined by practitioners' contributions
to organizational goals, rather than to employers' goals, are highly professionalized. In viewing the sciences as reputational organizations, I am therefore
assuming that the knowledge goals of scientific communities dominate work
goals, and scientists careers in reputational terms dominate careers in terms of
employment statuses.
Just as professions in general vary in the extent to which they control
labour markets and work goals, so sciences control access to jobs and research
facilities to differing degrees. In much industrial research work, goals are
largely determined by employers and reputational communities have little or
no impact on them or how performance is assessed. Although employment is
full-time and continuous, routinization of work flow is difficult to achieve
because task uncertainty is relatively high and so control over work processes
is decentralized to scientists in a similar manner to skilled workers in the
construction industry. Hence, work organization is here administered along
"craft" lines but the degree of professionalization is low.
In contrast, sciences where work goals and evaluation of task outcomes are
dominated by the scientific community, in the manner mythologized by
Kuhn and others as ''normal science", are highly professionalized (10). Here,
scientists are oriented primarily to the knowledge goals and procedures of
their reputational field and employment statuses are largely dependent on
practitioners' professional reputations. Local contingencies and exigencies
have little impact on work goals relative to specialty or discipline opinion.
Scientific establishments in these fields control work through their control of
the reputational system. As controllers of professionalized reputational organizations, establishments in the sciences acquire control and allocate resources
for the production, validation and extension of particular knowledge by

Sciences as Reputational Organizations

317

awarding reputations to individuals, groups and employment organization on

the basis of their contributions to the intellectual goals of the organization
as interpreted by the establishment. However, not all sciences are as professionalized, nor are they all as consensual as Kuhn suggests. In looking at the
sciences as reputational organizations, these differences and other aspects of
their organizational structure can be examined without assuming that anyone
form is inherently "natural", "normal" or "rational". Different patterns of
intellectual and social organization occur in different circumstances and
viewing the sciences as reputational organizations, variously connected to
different sorts of employment units, seems a useful way of understanding
these patterns and changes in them.
2. The Establishment of the Sciences as Reputational Organizations in
Universities
Considering the sciences as reputational organizations results in their connections with employment units being taken as socially contingent and variable.
fu particular, their identity with university departments and the general
identification of science with academic sciences becomes historically contingent rather than intrinsically necessary. The integration of advanced
teaching and research in the German university system in the 19th century is
but one model of linking reputational systems with training structures and
employment organizations, just as, according to van den Daele, the particular
concept of science which became entrenched in the 17th century is but one
of a number which could have been institutionalized then (II). There is
nothing in the nature of modern science which inevitably leads to this form
of organization and other ways of connecting recruitment and training
systems with employment organizations are increasingly apparent (12).
Nonetheless, the dominant role of university employment organizations in
the professionalization of the sciences has had a considerable impact on how
they are organized and the development of their ideals. In particular, the
location of employment positions in teaching posts, which controlled the
production of secondary school teachers, enabled scientists to extend their
control of recruits- and training programmes into the school system and,
therefore, the public image of science as experienced by school children.
Equally, intellectual boundaries and priorities became identical to pedagogic

318

Richard Whitley

ones and reified in curricula and textbooks so that reputational communities

were reproduced through educational structures. Scientific ideals also became
similar to the dominant academic values - such as "neo-humanism" in
Germany (13) - so that the emphasis on "pure" and "high" science was
reinforced and applied science and technology were devalued (14).
The employment of scientists in universities, as distinct from, for example,
academies of science probably resulted in reputational communities and their
objectives and procedures being more strongly bounded and fixed than if
they had been institutionally separated from the educational system. As it
was, disciplines became seen as both the locus of training scientists, employment of them and of intellectually based communities. While the development
of research schools in universities - such as that of Uebig (15) - enabled
scientists to control a plentiful supply of cheap and semi-trained labour, it
also restricted research goals to those institutionalized in existing teaching
posts and curricula. Major diversions of orientations and purposes away from
existing disciplines became very difficult without strong external - usually
state - support for new chairs and departments. In other words, the resources
required to develop new directions became very large once communities were
institutionalized in universities, because that meant creating new training
programmes, curricula and posts. New intellectual fields had to become new
disciplines and departments if they were to develop as sustained research
traditions and programmes; the dominance of academic departments as the
primary way of organizing science resulted in new sciences being forced to
become departments if they were to reproduce themselves as reputational
communi ties.
Of course, all forms of professionalization involve exclusion and closure
(16). Once reputations in scientific fields became tied to jobs and access to
research facilities, investments in the skills and goals specific to those fields
were more considerable and had greater consequences. Control over labour
markets is an essential feature of professional organizations and strategies of
social closure become more central and important to reputational communities when they are professionalized. Nonetheless, the specific identification
of intellectual goals and procedures with units of education and training
probably increased the degree of intellectual and social closure to a greater
extent than if professionalization had taken a different form. While this
is partly because any change in direction involved changes in educational

Sciences as Reputationai Organizations

319

curricula and a reordering of relations between disciplines in universities, and

thus had ramifications beyond the immediate reputational system, it is also
because the academic model of science brought together so many of the
major components of professionalized science into a single employment unit.
With the development of paid positions in universities dependent upon
reputations awarded for contributions to knowledge, as assessed by the international research community (17), science as a reputational organization
became entrenched in the educational system and, through the increasing
importance of that system to the state, linked to dominant groups in the
wider society. In combining the traditional functions of universities - the
inculcation of received wisdom, knowledge and correct beliefs and modes
of behaviour in the privileged young - with the pursuit of new knowledge
and training new practitioners in research, the reformers of the Prussian and
other universities assimilated the sciences to the dominant ideals and elite
groups at the same time as transforming the universities and reasserting state
control over them (18). The universities continued to develop the "whole
man" in certain elite individuals and provide members of the traditional
professions and higher civil service but they now did this through advanced
research and the search for truth rather than through Latin disputations (19).
Rather than providing avenues for upward social mobility, as perhaps was the
case for a time in France (20), or useful inventions, science in the German
universities became part of a general ideology which served to reinforce the
position of existing elites and the "mandarin" stratum in the civil service.
While this usage of the universities was scarcely novel in Europe, its connection to professionalized science was.
The reorientation of the German universities in the 19th century to the
pursuit of original scholarship provided a strong organizational base for
systematic and sustained research. By legitimating, and providing resources
on a considerable scale for, large-scale scientific research, the reforms lead
to the growth of the professional scientific community and the domination of
that community by German research schools. Even though the gentlemanly
"amateur" tradition continued to produce major figures and contributions in
England (21), the sheer number of practitioners in the German universities
and their use of students on a large scale for research, which was in part
subsidized by state support for laboratories (22), led to European science
becoming dominated by academic professionalized science (23).

320

Richard Whitley

One result of this growth of professionalism, in the sense of jobs depending

upon reputations, was that science became "normalized". Since research was
increasingly done with students, reliable means of obtaining results which
could be compared and integrated became very, important. Inexperienced
neophytes could not be expected to contribute to knowledge where task
uncertainty was very high because techniques produced variable results which
required great experimental and interpretative skill for their meanings to be
elucidated. Research schools relying extensively on student labour, then,
required fairly simple and reliable technical procedures if they were to make
contributions to knowledge. As Morrell points out, Uebig's apparatus for
the combustion analysis of organic compounds was an essential prerequisite
for his successful domination of organic chemistry (24), it reduced task
uncertainty sufficiently for his students to analyse compounds on a large
scale and thereby transformed the field. The example of the Saxon scientists
in the early 19th century is similarly instructive. Jungnickel (25) emphasizes
the extensive collaboration between senior scientists required to develop
precise and reliable methods of experimental investigation in the natural
sciences. This collaboration was largely accomplished through the Saxon
Society of Sciences rather than in the Universities of Leipzig and Jena. After
precise and reliable methods and instruments had been developed, the pattern
of research collaboration changed: as Jungnickel says
methods and instruments that made possible the participation of relatively inexperienced
helpers such as advanced science students ... tended to make superfluous the cooperation between highly skilled professionals that had been essential to research earlier (26).

The professionalization of science in universities, then, tended to emphasize

technical precision and reliability and the formulation of problems which
could be dealt with by these techniques. Task uncertainty was reduced to a
level where neophytes could produce useful results.
By organizing research aroupd the training process, and developing formulations and procedures which permitted trainees to produce valid knowledge
claims, the academic professionalization of science developed both a hierarchical structure for conducting scientific research and a style of research
which enabled original knowledge to be produced by such hierarchies. In
contrast, the dominant mode of knowledge production in pre-professional
science - or "amateur" science - seems to have been more akin to literary

Sciences as Reputational Organizations

321

and philosophical production than to modern professionalized normal science.

Individual craftsmen, or "geniuses", worked in splendid isolation on relatively
diffuse and general problems which varied in conception and preferred approach across practitioners. Occasionally supported by assistants or disciples,
these gentlemen amateurs (27) frequently abhorred notions of science as a
profession and preferred to work on their own rather than organize research
groups with a systematic programme of research and "rational" division of
labour. While they may have completed for reputations amongst each other in
terms of solving the major problems of the field, and this involved being able
to specify those problems as being major, they did so largely as equals and
as individuals rather than as schools of knowledge producers entrenched in
employment structures. Because training was not systematically organized
around employment posts, which monopolized access to research facilities
and communication media, it occurred in a largely haphazard and ad hoc
manner and thus did not impinge greatly upon the form or direction of
research. Neophytes may have acquired the basic skills necessary to make
valid contributions by working with an "acknowledged master" but they did
so as individuals, usually limited in number, and not as part of the organization of knowledge production. Collaboration between scientists seemed to be
relatively uncommon and when it did occur was on the basis of equality
rather than as members of a hierarchically structured team, as the above
quotation from Jungnickel indicates.
The crucial innovation of Uebig and other producers of knowledge producers was, of course, to change this system radically and to incorporate the
training of recruits systematically into the practice of doing research. Henceforth, the attraction of a steady flow of research students became a major
factor in the development of dominant research programmes in some sciences,
and the nature of the problems considered, and procedures appropriate for
their solution, changed accordingly. Given the norm of originality, a relatively
formal system of communication and the importance of reputations for jobs,
the professionalization of the sciences would probably have resulted in some
of these changes anyway but the linking of the training of recruits to the
production of new, valid knowledge in the universities accelerated and
reinforced the move to organized division of tasks and coordination of
outcomes by workplace supervisors. Specialization of topic areas and skills
amongst practitioners became echoed and reinforced by specialization and

322

Richard Whitley

differentiation within research laboratories which were integrated and controlled by an administrative hierarchy organized on the joint basis of technical expertise and knowledge and academic rank. Problems which required
collaboration between highly skilled equals or which were relatively diffuse
and general so that they could not easily be divided up into separate parts and
tasks were less likely to be taken up and worked on in these systems of work
orgimization and control than in more "amateurish" systems (28).
The systematic organization of scientific research in employing institutions
with clear hierarchies of authority based on differences in expertise and
knowledge demonstrated how the arcane and esoteric activity of discovering
the laws of the universe, undertaken by a few exceptionally gifted individual
"geniuses", could be rationally administered and controlled. It also showed
how discoverers could be produced on a relatively large scale so that when
science became seen as useful knowledge, and as a method for producing
useful knowledge rather than as a system of fixed knowledge (29),
did not need to wait for "geniuses" to appear but could simply expand the
existing system of scientist production. In a sense, the academic domination
of science through the production of knowledge producers, and their systematic use as research workers solving differentiated tasks which were coordinated by the laboratory head, provided the model of research organization
for employers in other spheres. By showing how scientists could be trained
and how neophytes could be organized to produce knowledge, the development of research schools in universities demonstrated not only that science
could be useful as systematic, fixed knowledge but also that the actual
process of knowledge production could be planned and organized. Rather
than waiting for the fruits of a few individual geniuses to fall off the tree of
knowledge and then be applied to extra scientific problems, the organization
of research schools demonstrated how scientific work could be administered
as a process oriented to a variety of goals. It was thus capable of being rationally organized and, by extension, planned for goals other than purely
academic ones. Science as a system of knowledge about the world which was
stable, true and coherent became transformed into a process or method of
knowledge production which could be organized and planned. The academic
professionalization of science thus paved the way for the large scale use
of science in non-academic environments and for non-academic purposes
(30).

Sciences as Reputational Organizations

323

3. The Domination of Academic Establishments in the Sciences

Another consequence of the professionalization of the sciences in universities

was the emergence of powerful academic establishments, as Norbert Elias
emphasizes in his theme paper in this volume. By combining the control of:
training of recruits, curricula in the schools, the organization of research
workers in pursuit of research programmes, jobs dependent on reputations
based on contributions to knowledge goals and the increaSingly expensive and
complex facilities needed to make those contributions, university departments and their controlling establishments dominated the sciences. While
university employees did not monopolize the production and validation of
scientific knowledge in all countries in the 19th century, academics did come
to hold dominant positions in scientific "communities", even in countries
where the "amateur" tradition was still a powerful ideological force (31), and
the academic model of professionalization became the "obvious" means by
which reputational communities could obtain control over jobs and research
facilities and exclude outsiders. The increasing domination by academics of
the sciences, both established and emerging, meant that scientific establishments became academic ones. It also resulted in distinctions between the
sciences, and between them and non-sciences, becoming matters of academic
concern and institutionalized in academic boundaries.
Academic establishments were, and are, powerful in the sciences not
only because of their control over the production and training of the labour
force, a feature they share with other professional establishments, but also
because of their domination of the knowledge production process through
the extensive use of advanced students as research workers. By monopolizing
the inculcation and certification of skills required to carry out scientific
research, so that it became impossible for people who had not been through
an academic training programme to make "competent" contributions, academics control entry to the labour market and the meaning of the skills
which are "scientific". Insofar as universities are the major research sites,
academics also control access to jobs in the dominant employment organization so that only those with the skills and interests which follow academic
defmitions of the major knowledge goals obtain posts. Furthermore, by using
students to undertake research tasks which are coordinated into a research
programme, academics become the major knowledge producers in the sciences

324

Richard Whitley

because they control more resources. While not all research programmes are
equally "successful", and not all sciences are equally programmable, academic
establishments are more able to define what is "successful" because they
control the writings of the "official history" of the sciences in textbooks
and their use in training new recruits. Also, because they control the bulk of
the scientific labour force, both staff and students, they are more able to
dominate interpretations of the major goals of a science than are individual
"amateurs". As well as controlling the labour market they dominate the
definition of tasks and coordination of task outcomes through their control
of the major part of the knowledge production system.
Control over a substantial part of the scientific labour force is not the
only means of control over knowledge production exerted by academic
establishments. As part of the process of increasing precision, purity and
standardization of tasks and technical procedures, which partly developed
from the use of students in research, the technical apparatus required for
undertaking scientific research which resulted in competent contributions
became more complex and expensive. The raw materials similarly became
more refined and restricted in availability so that the individual practitioner
found it more difficult to conduct scientific research without collective
resources. The technical means of producing scientific knowledge increasingly
outgrew the capacity of individual provision and control so that they had to
be collectively organized and controlled. The more concentrated they were in
universities and under academic control, the more academic establishments
dominated reputational communities. The reduction of task uncertainty
through increased precision and standardization of measurements and observations rendered science more esoteric and professional, and more under the
control of academics.
The standardization of materials and procedures in many sciences in the
19th century was part of the general process of professionalization of scientific work. It aided the academic domination of the sciences through enabling
trainees to make contributions to knowledge while distancing the objects
and procedures of science from everyday, lay concepts and substances. The
simultaneous transformation of scientific research into an activity involving
systematic use of esoteric standard techniques conducted by teams andgroups
in a relatively extensive division of labour and into an activity necessitating
collective, and collectively organized, resources which were controlled by a

Sciences as Reputational Organizations

325

single major institution, the university, was a crucial part ofthe development
of professionalized academic science. The standardization and formalization
of much scientific research was also important in the development of professional reputational communities. As Kuhn points out (32), the mathematization of a science enables, ceteris paribus, controversies to be speedily resolved
and "normal" science to progress smoothly. By standardizing technical
procedures and symbol structures, communication across social and spatial
boundaries is facilitated and the reputational community more able to control
work practices and outcomes. The reduction of what might be termed technical task uncertainty enabled practitioners with a common set of relatively
standardized techniques to communicate task outcomes quickly and easily
and so coordinate them. Competence in such fields then becomes defined
as competence in the use of these techniques for dealing with appropriate
problems. Relatively clear social and cognitive boundaries can be demarcated
and reified and strong collective social control over the details of work
processes exerted. While sciences existed as reputational organizations beforehand, it was their identification with training and employment units built
around research programmes using standard procedures which strengthened
their identities and power. Equally, the domination of academics over employment and training units which inculcated and applied these procedures
facilitated their control over reputational organizations to the extent that
scientific fields have become synonymous with academic "disciplines" or
"specialties" .
4. Variations in Academic Domination of the Sciences

This rather monolithic characterization of academic establishments and

academically professionalized science emphasizes the powerful consequences
of combining training, employment and reputational organizations in a single
structure. However, not all scientific fields became dominated by academics
to such an extent, nor did all sciences develop such a strong professional
identity. Furthermore, the proliferation since the Second World War of
employment units outside universities where scientific research is carried out
for non-academic purposes has severely mitigated the academic monopoly of
jobs and resources.
Indeed, recent changes in the organization of scientific work,or the

326

Richard Whitley

extension and intensification of existing tendencies (33), have emphasized

the contingent nature of scientific disciplines' identity with academic employment units. Reputational communities are no longer completely identical
to university departments - if they ever were - and contributions from
particular employment organizations may be relevant to a number of distinct
reputa tional organizations' goals. In many of the con temporary sciences there
are numerous work organizations directed to a number of distinct goals which
may, or may not, coincide with traditional academic disciplinary goals and
existing reputational communities. Where such "goal directed" research (34)
does not fit in to existing reputational structures - as is arguably the case
in cancer research and fusion research - new reputational communities
are sometimes formed which may later become the basis of new academic
departments, such as those of oncology. Equally, they may mount takeover
bids for existing departments and impose their own theoretical goals on
academic disciplines, as molecular biology did in many biological fields.
These sorts of changes indicate how the development and organization of
reputational communities are linked to changes in the organization and
direction of knowledge production and how, in professionalized academic
science, resources can affect the organization of science. Once the production
of knowledge producers is organized and directly linked to the production of
standardized, formal knowledge, the systematic administration of research
directed towards a number of divergent goals becomes feasible. Where this
sort of research is organized around reputational communities, as opposed to
being oriented to employers' goals, it affects the structure of reputational
organizations directly and, hence, the organization of knowledge and knowledge goals.
The control that reputational organizations have over goal definition, jobs
and resources in fields where employment is concentrated in organizations
established for social, medical or political purposes is usually less than that
evident in predominantly academic fields. Howevermuch scientists may
appropriate the "external" goals of the organization and re-interpret them
into "scientific" terms, which enables them to set up a reputational community and impose their own goals on the work organization, social constraints
are likely to occur sooner and more forcibly here than in universities. While
local considerations, -and the pedagogic function of universities, meant that
reputational control was never total in academic fields, and in any case

Sciences as Reputational Organizations

327

varied considerably between them, the goals of reputational communities,

and their application, will be more clearly the outcome of negotiation and
conflict between employing organizations and "scientific" ideals in less
academically entrenched fields.
With the development of employing organizations dominating the bulk of
scientific work, then, reputational organizations no longer exist in a "pure"
form which can be studied in isolation from administrative structures and
work control procedures in employment organizations. Just as the establishments controlling scientific reputations became incorporated into academic
organizations and transformed into academic establishments controlling jobs,
credentials and facilities as well as reputations, so scientific-academic establishments have now come to terms with the medical, business and political
establishments and transformed themselves into general problems solvers and
administrators. In the process of these transitions scientific ideals and goals
have become altered to accommodate themselves to these new relations, as
new reputational communities have become established around non-academic
- or quasi-academic goals. Furthermore, as new sources of research support
have become available, academics have altered their research programmes to
attract resources and attempt to dominate their reputational communities.
This has sometimes been tied to the redirection of scientific goals in favour of
greater precision etc., which implied the use of particular skills and expensive
apparatus thereby raising the "barriers to entry" in that field and effectively
monopolizing contributions (35). Thus, not only has the expansion and direction of the sciences produced new fields and communities, it has weakened
old definitions and boundaries so that scientific goals have become more fluid
and open to negotiation.
Indeed, the development of full time research laboratories on a relatively
large scale, which are at least partly oriented to reputational communities
and public goals, decreases the advantages academics have by virtue of their
control over trainee labour. As long as university teachers have to undertake
other duties, and not all their teaching involves the employment of research
workers, they remain part time contributors to the reputational community
(36). The more research goals require highly expensive and complex machinery which has to be centralized in full time research organizations, and/or
extensive teams of. technicians , the more workers at full time research centres
will come to dominate - or at least threaten the status of - academics. The

328

Richard Whitley

more reputational communities become dominated by goals reflecting the

interests and resources of full-time work organizations, the less control will
academics have over the direction of the sciences. Although this process is
scarcely complete, and varies in its importance across the sciences, it has
already resulted in numerous attacks on "big science" and struggles, both
within and without universities, over research goals and strategies involving
disputes about scientific ideals and conceptions of knowledge. The autonomy
of academic reputational organizations, and hence, the power of academic
establishments, has been reduced by these developments. Insofar as "big
science" is taken to be the ideal of scientificity, science can be viewed as
having outgrown the university which now becomes just another site where
research is conducted.
These developments, coupled with the obvious differences between
scientific fields in terms of their links to, and control over, employment
organizations, highlight the contingent nature of the academic monopoly of
science and the plurality of relations between reputational communities and
employment organizations. Different sciences may control jobs in a variety
of different organizations and scientists in any particular organization may
contribute to the goals of a number of different reputational communities.
This flexibility of skills and research facilities, vis-a-vis, specific scientific
fields and their goals is higher in some sciences, such as bio-medical research,
than in others, such as physics. The entrepreneurial "tinkering" of some
biological scientists demonstrates their relative autonomy from the central
goals of the dominant reputational community and the corresponding greater
influence of local considerations (37), in strong contrast to the highly goal
specific nature of skills and techniques in physics and chemistry. In these
latter fields, reputational organizations control the use and combinations of
skills for particular purposes to a much greater extent. Here, opportunities for
local initiatives and redefinitions of goals are limited.
S. Dimensions of Organizational Structure in the Professionalized Sciences
These sorts of differences between scientific fields can be partly summarized
in terms of their degree of professionalization, or the extent to which a central
scientific establishment controls work goals and procedures through its
control of the dominant reputational system. However, this would suggest

Sciences as Reputational Organizations

329

that the sciences only vary in their similarity to, say, sociology or philosophy
on the one hand and physics and chemistry on the other hand. In other words
they are either "polyparadigmatic", and apparently highly disputatious, or
highly restricted, centralized and bureaucratic. The problem with relying
entirely on this one dimension is that many other differences between fields
are glossed over or ignored and that some areas simply do not easily fit; e.g.,
economics, molecular biology, mathematics. Furthermore, there is a strong
implication in adopting this dimension on its own that the sciences inevitably
tend to develop in the direction of physics which is exactly the sort of
unidirectional assumption under attack in much recent sociology of science.
The plurality of the sciences, and of their paths of development, are better
understood by viewing them as reputational organizations which differ on a
number of dimensions.
Treating the sciences as one form of work organization in which control
of work practices, and sometimes goals, is dominated by the professional
reputational community, suggests two major dimensions for comparing and
analyzing changes in the structure of sciences: task interdependence and
task uncertainty (38). The extent to which tasks need to be coordinated
for significant contributions to be made to knowledge goals, and the predictability, visibility and repeatability of tasks and task outcomes are key factors
in organizing and controlling work, in science as elsewhere. In this section
I will sketch the use of these dimensions' for analyzing the sciences and
subsequently outline how they are interrelated and change.
(a) The Degree ofMutual Interdependence
Task interdependence among scientists as members of reputational organizations is derivative from their need to rely upon each other to produce
knowledge claims which are significant for organizational goals, and receive
validation from their professional colleagues. The more dependent practitioners are upon each other, as opposed to other groups, for reputations, the
more they have to coordinate their work and interrelate task outcomes. High
dependence on colleagues for validating contributions as significant and
competent implies dependence on the work of others and their goals and so
a high degree of ''functional integration" (39). It also leads to the objectification of knowledge objects and their detachment from the immediate interests

330

Richard Whitley

of knowledge producers by collectively organizing knowledge production and

validation. Interdependence of tasks diminishes personal, diffuse and tacit
evaluation criteria and encourages the development of standard formal
procedures and languages for comparing and coordinating task outcomes,
hence enabling general reputations to be awarded. Object adequacy and
increasing object centredness, as discussed by Elias, are thus related to increasing mutual dependence among practitioners in professionalized sciences.
This need to rely upon a particular group of colleagues in a scientific field,
which is a part of the general need to rely on scientific opinion as a whole to
produce certified knowledge and obtain jobs as a result of the professionalization of the sciences, varies between scientific fields and within the "same"
field in different periods. It increases when the availability of resources
declines relative to the number of practitioners seeking reputations in a
particular field. Given the norm of originality which dominates work in
modern science, increased numbers tend to result in increased specialization
and differentiation of tasks and skills as practitioners seek to demonstrate the
novelty of their contributions and so obtain recognition. Novelty, however,
cannot be too great since contributions in a reputational community are
assessed in terms of their relevance to the dominant goals of that field.
Knowledge claims which do not fit in with established procedures and ways
of interpreting these goals will not be seen as relevant and so will not lead to
improved reputations in that field. Consequently, scientists are ambivalent
about the norm of originality, and in the professionalized sciences are likely
to differentiate their contributions in terms of the system studied or technical
procedures used rather than by indulging in theoretical deviance or major goal
conflicts. Increased competition for reputations and resources in fields with
common technical procedures and formalized communication systems results
in increasingly narrow and specific research problems being undertaken as
practitioners attempt to show their originality simultaneously with their
fidelity to the dominant goals of the reputational system.
The more specialized research topics become, in relatively established
and "closed" fields, the more scientists come to depend on each other as
the relevance of much work to the general goals of the field becomes more
tenuous. Highly specialized and specific contributions have to be coordinated
and integrated if they are to be linked to theoretical goals and so acknowledged as useful contributions. This coordination requires standardized and

Sciences as Reputational Organizations

331

formalized procedures and languages, such as mathematics, and so specialization encourages the development and extended application of standard
techniques and highly formal media of communication. Interdependence and
coordination needs thus increase as the number of practitioners rises and
specialization intensifies, to the extent in physics of creating an entire class of
theoreticians which is itself subdivided (40). These theorists are "needed" to
make sense of the mass of task outcomes produced by highly specialized
experimentalists working on very specific problems and sub-problems, in
terms of the dominant goals of the field which are interpreted and "guarded"
by the establishment of theoretical physicists.
This greater dependence on professional colleagues arising from increased
specialization is further intensified by the concentration and centralization
of research facilities, journal space, funding agencies and other necessities
for conducting research and obtaining reputations. The more limited these
facilities are, and/or the more they are concentrated in a few centres or under
the control of a few people, the more crucial it becomes for scientists to
demonstrate that their contributions are important for the accomplishment
of major organizational goals. Their dependence on established colleagues for
validation of their work is therefore much greater under these circumstances
than when facilities are cheap and widely available as in, perhaps, mathematics. The hierarchies of specialties and sub-fields in physics - as contrasted
with chemistry - are partly a result of the high degree of concentration of
resources and funding sources. By successfully institutionalizing particle
physics as the primary field of physics and thus ensuring that major reputations in physics depended on access to very complex and expensive apparatus,
contemporary phYSicists so concentrated and centralized resources that a
relatively small number of scientists controlled the careers of large numbers
of physicists. Only by convincing established leaders and administrators can
practitioners obtain access to facilities that are essential to make major
contributions to the dominant goals and so acquire major reputations. Furthermore, the sheer size of the facilities necessitates a high degree of division
of labour and administrative coordination in the workplace so that the
independent scholar has to depend on a host of technicians and colleagues
simply to generate some output and organize it into a form where some
sense can be made of it. Not only, then, are scientists here highly dependent
on a small group of colleagues for the opportunity to do research but they

332

Richard Whitley

also depend on other experimentalists to produce and analyze task outcomes,

and on theoreticians to demonstrate the meaning of the results. The old ideal
of the single virtuoso producing a distinct commodity termed original knowledge seems a long way from the fragmented, specialized and hierarchized
world of contemporary experimental physics (41).
While high energy physics may be an extreme case of such centralization,
it cannot be dismissed as a freak example. This is partly because of its prestige
and dominance of scientific identities and ideals which results in other fields
being organized in similar ways as "big sciences" but also because centralization can occur without highly complex and expensive apparatus being used.
Whenever jobs are relatively scarce and dependent upon reputations based on
public contributions to knowledge, control over the communication system,
Le., access to journal space in the "top" periodicals, leads to control over the
reputation system and thus over the validation of knowledge and individuals'
careers. The more jobs are tied to reputations in a particular field which is
dominated by a few journals organized in a clear hierarchy of prestige, the
more centralized is that field. Where jobs can be obtained on the basis of
reputations in a number of areas and/or where the dominant journals are
relatively eclectic in terms of knowledge goals and techniques, obviously the
field is not so centralized. Centralization of control over communication
media is reinforced by the knowledge goals necessitating research which
requires additional funds which are monopolized by a single agency. The
ability of an establishment to control a scientific field through the allocation
of journal space is greatly enhanced by its adoption of knowledge goals which
require practitioners to seek resources outside their employment organization
from a central source which they also control. The definition of sociological
research as survey research which is funded by only a few agencies is an
example of how a particular concept of a field assists central control of
reputations and, when jobs are scarce in the dominant employment organization, i.e., universities, jobs. Dependence on colleagues, then, is high when
contributions require external funds from a limited number of agencies under
the control of. professionals. In these circumstances, with a relatively formal
language, increased numbers of practitioners will lead to a greater standardization of procedures and a strong emphasis on the attainment of a particular set
of theoretical goals. Theoretical pluralism is not likely to be encouraged.
However, a high degree of dependence on one's professional colleagues is

Sciences as Reputational Organizations

333

not always associated with strong centralization of control and the domination
of a single theory. While centralization of control over key resources in the
reputation system encourages increased dependence upon one's colleagues
in the sciences, the reverse is not necessarily the case. Strong reputational
organizations with control over jobs and careers may exist without being
organized around a central goal and approach to problems. A common educational background which develops similar skills and ensures that practitioners
share technical procedures here leads to a common professional identity and
enables the community to exclude outsiders from jobs without centralizing
control over facilities. Work in these fields, such as mathematics (42), is
highly specialized and involves formal communication systems for the award
of reputations which affect career prospects and yet no single view of the
field is dominant to the extent that particular problems or topics are regarded
as central. Resources for conducting research are cheap and widely available,
as are jobs and journals, so different goals may be pursued with little need to
order them into an overall hierarchy. Increased numbers of practitioners in
these fields simply results in greater differentiation and specialization of skills
and tasks which are coordinated through the formal language and common
educational background. As long as the various topic areas can obtain access
to jobs, pressure to organize them into some sort of hierarchy of importance
will remain slight. If jobs become scarce,
and relevance to disciplinary
on the other hand, some means of ordering contributions and topic areas in
terms of their importance will probably develop and more efforts will be
devoted to general syntheses of the field as a whole.
The development of strong reputational organizations in the sciences
with a high degree of interdependence among practitioners is associated with
increasing autonomy and distance from lay and non-scientific concerns and
goals. By definition, the relevance of exoteric audiences decreases as the
importance of esoteric goals and criteria for the award of reputations rises.
Equally, the development of a strong reputational community controlling
access to jobs and research facilities implies increasing autonomy from other
sciences and their goals and technical procedures. For a distinct reputational
community to form around separate goals and practices some autonomy must
be granted; once organized, such strong professions (43) seek to increase
and extend their autonomy by controlling access to jobs through insisting
on particular technical skills which can only be acquired through training

334

Richard Whitley

programmes controlled by that community. This in itself increases practitioners' dependence on each other as appeals to external audiences and
funding sources become channelled through the professional establishment.
Direct assess to the lay public and/or powerful non-scientific establishments
is seen as non-professional behaviour and sanctioned accordingly, as in the
Velikovsky case. Autonomy is thus both a condition and a result of increasing
mutual dependence among scientific workers.
Strategies for social and intellectual closure are not always successful
though. The social sciences show how difficult it is for some fields to establish
autonomy from lay concerns and approaches, as Elias and Katouzian discuss
in their papers for this volume. Attempts to professionalize around a common
body of techniques and methods allegedly derived from, or consonant with,
more established fields have been only partly successful. The relative success
of economics and psychology in distancing themselves from lay control and
problem formulations has been accompanied by attacks from philosophers
and other establishments as well as internal dissension during the expansion
of university posts. The extent to which the goals and dominant approaches
of these fields is indeed separate from lay goals and purposes is, in any case,
open to dispute (44). Furthermore, even when autonomy from non-scientific
groups has been achieved, to the extent that particular technical competencies
become necessary for access to jobs and reputational communities control
careers, autonomy from other scientific groups may still be limited. As
Fleck's study of the development of Artificial Intelligence in this volume
shows, existing scientific establishments were able to exert considerable
influence on funding agencies and the direction of much work in this area
in Britain.
The degree of autonomy over goals, techniques and jobs possessed by any
reputational community, and thus the extent of mutual dependence among
its members, is obviOusly dependent upon its ability to monopolize access
to posts and research facilities, as well as its control over prestigious journal
space. As I have already discussed, the academic model of reputational
organizations exhibited strong tendencies towards such monopoly, at least for
some fields, but the dominance of this model has declined with the growth
of non-academic research institutes. Where the goals of work organizations
do not easily fit in with those established in academic training programmes,
and other established groups have direct interests in the goals pursued by

Sciences as Reputationai Organizations

335

employment organizations, then the degree of dependence on a single, particular reputational community will decline. Plurality of work organizations
and goals represented in them is likely to lead to a plurality of reputational
communities available to practitioners such that their dependence on anyone
group of colleagues is reduced and the boundaries of such organizations
become more fluid. Intellectual and social closure here become difficult to
enforce and scientific identities are often multiple and overlapping. While a
common body of technical approaches may produce a degree of "normative
integration" (45) among, say, biochemists, the goals to which these techniques may be put are so various in many of the bio-medical sciences that
task interdependence and functional integration among this group of practitioners is considerably reduced. The increasingly attested fluidity of research
boundaries, goals and identities among many biological scientists is an exemplication of this phenomenon, as is the "tinkering" described by Knorr and
others (46). The general development of "directed" research in industrial
nations is likely to accelerate these sorts of changes in the organization of the
sciences and promote the formation of new reputational communities and the
dissolution of old boundaries and knowledge structures.

(b) The Degree of Task Uncertainty

The sciences are systems of work organization and control which, in general,
exhibit a relatively high level of task uncertainty. However, they do differ
among themselves in their degree of task uncertainty. As I have already
suggested, the degree to which scientific fields are professionalized and have
standardized technical procedures varies. Furthermore, the meaning, relevance
and significance of research results for theoretical goals vary in clarity and
straightforwardness in different fields. Even where techniques are standardized, the overall significance and importance of results may remain vague and
subject to disputes. Reputations fQr major contributions to organizational
goals are fluid and changeable in such fields. Generally speaking, the more
control scientists can exert over their cognitive environment so that task
outcomes are predictable and visible - Le., reliably replicable (47) - the
lower is task uncertainty in that field. What might be termed theoretical task
uncertainty is low, in addition, when the Significance of results for dominant
goals is clear and widely accepted. In fields with relatively low theoretical

336

Richard Whitley

task uncertainty connections between theories and techniques are easy to

draw and implications of task outcomes rarely problematic or the focus of
major conflicts. Errors and anomalies are not too difficult to distinguish
and identify and there is a large amount of background knowledge and
assumptions which are widely accepted. Theoretical goals are readily translated into research programmes with the necessary sWls and resources easily
identified.
Control over cognitive objects is partly a consequence of cognitive closure
in that the fewer aspects of objects which are taken into consideration, and
the simpler they are, the more likely task outcomes can be successfully
predicted and disputes resolved. By producing phenomena under highly
controlled and artificial conditions and focussing on a restricted number of
qualities expressed as quantities, scientists attempt to increase their control
over objects, as Rip points out in his paper in this volume (48). The search
for purity in 19th-century chemistry can be seen as an attempt to increase
closure and control as well as standardizing materials sufficiently for results
to be coordinated and integrated throughout the reputational organization.
This led to the development of professionalized chemistry with extensive
division oflabour and systematic employment of assistants and neophytes.
There are limits to the extent to which uncertainty can be reduced, however, without threatening the professional status of the sciences. Routinization
of scientific work may lead to its becoming so programmable that scientists
are reduced to being technicians in the service of non-scientific goals and
employers. Chemical analysis, for example, was sufficiently uncertain an
activity in the early 19th century for Davy, Dalton and Faraday, among
others, to make a living as consultants from this activity (49) while today
it is increasingly conducted by low grade technicians and computer directed
machines. In this case not only are work goals being determined by anyone
who can pay for the skills but the work processes themselves have become
so routinized that professionals no longer control them and employment
organizations increasingly plan them directly.
The norm of originality ensures that this degree of routinization is rarely
encountered in the sciences; or, rather, that whenever techniques are produced
that effectively routinize scientific work they become used as instruments
in the pursuit of other goals which have higher uncertainty. By constantly
searching for novelty - of a particular and limited kind - the modern sciences

Sciences as Reputational Organizations

337

ensure that they remain relatively uncertain activities in comparison to the

bulk of other work. Indeed, they claim to control uncertainty itself through
developing increased understanding of the world so that other professional
groups can operate more effectively. By monopolizing the right to generate
truth claims, and the right to assess and validate them, they legitimate others'
claims to competence and uncertainty reduction. The sciences thus both
generate and control uncertainty. Without being able to demonstrate their
ability to produce predictable results in particular circumstances, and therefore routinize uncertainty, they would have much less claim upon resources
from other establishments, yet the scientific method which purportedly
produces this control must be sufficiently esoteric and complex for specialized
training to be required which is under the control of scientific establishments.
Increased specialization of skills and techniques which results in ever extended
training programmes serves to maintain scientists' control over practitioners
as well as ensuring continued originality and task uncertainty (50).
The organizational correlates of reduced task uncertainty in scientific
fields are similar to those found in other work organizations. The more
predictable are task outcomes, the more work can be systematically planned
outside the work processes, work roles allocated on a full time basis, tasks
highly differentiated and results coordinated and controlled through a formal
hierarchy with an elaborate communication system. Terry Shinn's deSCription
of work organization in chemistry laboratories illustrates these points. While
not employment bureaucracies, these scientific fields exhibit many similar
characteristics and the hierarchy of employment statuses tends to be identical
with that of professional statuses so that authority is organized around a
single dimension. Because the hierarchy of goals is often fairly clear and
uncontentious, and the relevance of contributions to these goals are not hard
to establish, reputations and rewards seem to follow smoothly from task
outcomes and the successful accomplishment of research programmes. In a
sense, the whole science operates like a smooth machine.
Scientific fields where task uncertainty is higher are less likely to formulate
and carry out research programmes in a systematic way which directs work
across employment organizations. Theoretical goals are more fluid and subject
to a number of interpretations which have divergent consequences for research
tasks and techniques. The relevance of any particular task outcome for a
single theoretical viewpoint is not easy to establish here, and contradictory

338

Richard Whitley

interpretations of the significance of results are not uncommon. Multiple

conceptions of goals and interpretations of dominant theories result in
competition over the ordering of priorities and variations between work
organizations' orientations. Local considerations and exigencies will have
more impact on the nature of the work carried out, and how it is done, than
in fields with lower task uncertainty. Similarly, employment organizations
will differ in the particular skills they contain and how these are combined
for a variety of purposes. Because goals are fluid and uncertain or conflicting
in their consequences for task outcomes, reputations do not follow on so
smoothly as in other fields and are more susceptible to short-term changes
and influences. Criteria of success and significance are less institutionalized in
these fields so that work is evaluated differently at different sites and reputations not as widely established throughout the field as in more integrated
sciences. Sub-groups may form around different goals - while using similar
methods and skills - which allocate reputations amongst themselves without
unifying them into a science-wide reputational system. Particular skills and
techniques may be commonly held throughout the field, and these may
be flexible enough to be applied to a variety of tasks and subgoals, but
practitioners often develop more esoteric and specialized ways of working
around different concerns in order to distinguish themselves from colleagues
and so continue to be original. Where coordination needs are low there may
be little need to order these subfields into a hierarchy of importance and so
conflicts may be minimal. In other circumstances, though, these subgoals may
be the basis of disputes and competition over their significance and, hence,
over the reputations of collectivities and individuals.
A further point to be noted about fields with high task uncertainty is
that, because goals are not so clearly linked to the use of particular skills and
techniques, and problems cannot be so easily be divided up into distinct tasks
to be undertaken by people with different skills which can subsequently be
coordinated and integrated by others, skills tend to be more general and
broadly applicable to a range of problems. Also, the particular combination
of skills adopted in anyone instance is more likely to lead to novel results
than in fields where outcomes are more predictable. The more specific are
skills to particular problems and goals, so that results can be easily coordinated with each other for the solution of a highly structured problem, the
less likely are outcomes to vary and be difficult to interpret. Consequently,

Sciences as Reputational Organizations

339

the less need is there for intensive collaboration between scientists during the
data production phase and the more routinized can this become - as shown
by Shinn's paper here. The more skills can be related to a variety of goals, on
the other hand, and the more the particular way they are combined in likely
to produce novel and uncertain outcomes, the more important it is to monitor
the generation of information very carefully and
less can it be left to
neophytes and technicians. Different patterns of collaboration and control
thus arise at different stages of the research process in different fields.
6. The Organization of Scientific Fields and their Establishments
Differences between scientific fields, and changes in their organization, can
be summarized and analyzed in terms of these two dimensions of structure
- degree of mutual dependence among practitioners and task uncertainty.
Although they are conceptually independent, changes in one dimension
are quite likely to lead to changes in the other in modern professionalized
sciences and some examples of these changes will be discussed later. First,
though, I will briefly summarize some of the major characteristics of fields
with particular combinations of mutual dependence and task uncertainty as
sketched in Figure 1. In all these cases, technical task uncertainty is assumed
to be sufficiently reduced to enable some comparisons of research results to
be drawn and some sort of communication system established so that the
reputational system does function to a minimal extent.
1. In the first case, scientists do not depend greatly on each other for their
reputations which in turn, are not critical for obtaining jobs and facilities.
Standardization of technical procedures is not high, and the meanings of
research outcomes relatively unclear and only tenuously linked to theoretical
goals. General and diffuse interpretative schemes are produced which rely
extensively on tacit skills for their understanding. These skills are likely
to be produced in a limited number of research sites by highly personal
means and will be associated with particular research goals and conceptual
approaches. The formation and reproduction of research schools which
effectively ''talk past" each other is a notable feature of these scientific
fields. Intellectual conflicts are wide ranging and frequently involve personal
disputes. Because the need to rely on all practitioners "in" the field is low,
scientists can focus on local reputations and compete for jobs and facilities

340

Richard Whitley

LEVEL OF TASK UNCERTAINTY

LEVEL OF
MUTUAL
DEPENDENCE

Low

High

Low

4. Smoothly functioning
formal bureaucracy,
e.g., contemporary
Chemistry

1. I nformal crafts
organization with
unspecialized
intellectuals producing
diffuse interpretative
schemes, e.g., traditional
humanities, European
sociology

High

3. Confl ictual and

2. Collegial-professional

complex pureaucracy
with demarcation
between central areas
and peripheral ones,
e.g., post 1945
physics

form with technical

standardization but
conflicting theories, e.g.,
psychology, parts of
biology, mathematics,
Artificial Intelligence,
economics

Fig. 1. Levels of task uncertainty and mutual dependence in the sciences and their
organizational correlates

within one school of thought and practice rather than developing more
general and widespread skills and procedures which would enable more object
centered and formal assessments of contributions to be made, and reputations
allocated across schools and employment organizations. Here, facilities are
relatively widely available and simple to use so that extensive technical
staffs are not required. Resources can largely be supplied on a local basis
without substantial recourse to external funding agencies. Also, the degree of
autonomy from lay audiences and their goals and concepts is likely to be
relatively low in these fields and many practitioners will seek direct access to
these audiences. Some examples of such areas are the traditional humanities
and many of the social sciences where the professional community as a whole
does not dominate the allocation of jobs.
Establishments in these sciences will be relatively decentralized and control
knowledge production and validation through largely local and personal
means. Because the reputational organization as a whole has only limited
control over careers, its establishment's ability to set goals and systematically

Sciencea aa Reputational Organizationa

341

evaluate contributions to those goals, so that reputations can be awarded on

a field wide basis, is relatively low. Direct control over jobs and resources at
the local level is often as important as controlling the overall reputational
system in these fields. While a particular approach and its dominant group
may come to function as the major establishment in the field, as long as
practitioners do not need to coordinate their research to a high degree in
order to contribute to intellectual goals the power of such an establishment
will be limited. Reputations within conflicting approaches are likely to be as
important to obtaining jobs and developing careers as discipline-wide recognition, indeed the latter may be dependent on the former. Within these schools
of thought control will be largely personal and direct so that they cannot
expand too much without fragmenting into sub-groups; patrimonial authority
is, thus, the dominant characteristic of these fields.
2. In the second case, task uncertainty is still fairly high, especially at the
theoretical level. However, scientists need to rely much more on each other to
coordinate their results and obtain reputations. The reputational system as a
whole is more important to jobs and careers here and so task outcomes need
to be easily comparable through a relatively formal communication system.
Technical standardization is higher in these fields so that the ambiguity of
task outcomes is reduced but the theoretical significance of results is subject
to disagreement and dispute. Conflicting theoretical goals mean that reputations cannot be ordered unequivocally in terms of their importance to the
field as a whole but a common core of techniques and concepts ensures that
they can be awarded in a systematic way which does not rely so much on
personal authority as in the previous case. Because of the greater importance
of the reputational system and mutual dependence of practitioners for
reputations and resources, these fields have a higher degree of autonomy
from lay concerns and concepts. Their distinctive expertise is more clearly
demarcated from non-scientific skills, and scientists are able to establish
greater distance from external audiences than in fields where the degree of
mutual dependence is less (51).
This emphasis on technical competence and standardization, combined
with multiple theoretical goals, enables specialization to occur through developing a common body of technical procedures for diverse theoretical goals.
This gives rise to techniques based specialties, such as Artificial Intelligence,

342

Richard Whitley

which share a common core of technical expertise and "craft" skills but
manifest a plurality of goals. Depending on the extent and diversity of
funding sources, and the cost of facilities, these goals will be more or less
closely related and mutually ordered. In mathematics, for example, where
most work is carried out in university departments which also train new
generations, a greater coherence of goals would be expected because employment is concentrated in a single institution and hence dependence on
colleagues is quite high with a strong degree of "normative integration"
provided by a common curriculum. As Fleck shows, the AI community is
more diverse in its employment sites and funding sources. Consequently,
dependence on a single relatively coherent, community is less - as is its
autonomy from lay concerns and other scientific establishments. Fragmentation - both social and cognitive - is therefore higher.
Establishments in these fields validate contributions in terms of the
standardized technical procedures which constitute the core identity of the
field. While competing groups conflict over the importance of particular
theoretical goals, and the overall significance of task outcomes for the field,
they all concur on the common body of skills and expertise which distinguish
practitioners and provide a means for the allocation of reputations. Insofar as
research facilities and/or funding sources are concentrated, one establishment
may dominate and impose its goals on other groups through its control of
resources. Where, however, facilities are widely available and journal space
not controlled by a single elite which also control jobs, groups adhering to
different goals will compete for domination of the field without any single
one becoming established as central. Different subgroups in different employment organizations are able to set their own goals and objectives within the
broad constraints of the disciplinary identity. Also, although the standardization of techniques and development of an esoteric and formal communication
system enable control of work to be exercised at a distance by rules, personal
contact and patronage remain important means of inculcating skills and
controlling task performance as Fleck's discussion of AI shows. Control is
thus less immediate and personal than in the first case, but personal authority
is still a crucial element in setting priorities and realizing objectives.
3. In the third case, theoretical task uncertainty is lower but the degree of
mutual dependence is still high, often because of the limited availability of

Sciences as Reputational Organizations

343

complex and expensive research facilities. Centralization and concentration of

resources here make individual practitioners highly dependent upon their
colleagues for the reputations which are necessary to continue to gain access
to research facilities. Because of the relatively high degree of theoretical
integration, which links dominant goals with conceptual approaches and
techniques in a fairly coherent manner, scientists in this sort offield compete
over the importance of their subgoals for the central disCiplinary concern.
The more a given group or specialism can demonstrate its significance for the
dominant goal, the more resources it can claim and the more its criteria for
cognitive success dominate the field. Theoretical priorities thus are closely
linked to reputational statuses and control over funding, facilities and posts.
Where these are relatively scarce - and concentrated - and theoretical task
uncertainty is fairly low, conflict over the ordering of theoretical subgoals
will be high as sub-groups struggle to demonstrate their centrality. The relatively low extent of task uncertainty ensures that most research can continue
in a comparatively routine manner with a high degree of specialization, as
practitioners seek to demonstrate their originality through differentiating
topics and techniques within an overall common framework, but the theoretical relevance and meaning of such work remains a matter of considerable
concern, because it affects reputations in the discipline as a whole which, in
tum, dominate the likelihood of continued participation in the science. In
other words, because these fields are both intellectually and socially centralized, the importance of task outcomes must be demonstrated throughout the
science if major reputations are to be established and renewed. Theoretical
integration of research results thus becomes a major activity as the large
mass of task outcomes has to be made sense of in terms of central goals.
The emergence of a distinct group of theoreticians in physics seems to be
partly a consequence of this intersection oftask uncertainty and coordination
needs.
The scientific establishment in these fields has a much greater degree of
control of all aspects of knowledge production and validation than in the two
previous cases. Both intellectual goals and technical procedures are controlled
through the reputational system which determines how jobs and research
resources are allocated. Local and organizational pressures have little impact
on research priorities compared to the influence of the international reputational organization. Additionally, control is exercised through formal rules

344

Richard Whitley

and procedures rather than personal authority and direct example. Assessment of task performance and the significance of task outcomes is a relatively
impersonal and formalized activity carried out by scientists dispersed over a
wide geographical area with few, if any, direct personal ties to the originators
of knowledge claims. Although tacit skills and judgements are still important
here, as they are in all the sciences, and personal patronage and contacts are
equally crucial to scientific careers, as they are in business, control processes
in these highly centralized and standardized sciences are less directly personal
than in those previously considered. Whether an experiment has been currently carried out, and the theoretical meaning and significance of the results,
are more easily and readily decided by relatively impersonal and formal
procedures in these sciences than in others (52).
4. Finally, in fields where task uncertainty is low, and the degree of mutual
dependence among practitioners is also fairly limited, planning of work with
a high degree of division of labour is possible and overt conflict over goals
unlikely. If funding is fairly easy to obtain and is available from a number of
different sources - e.g., industry and government as well as from national
research agencies - individuals' dependence on their professional colleagues
will be less than when it is impossible to carry ouf research without obtaining
funds from a single source. Central reputations in the science as a whole,
will therefore be less important in these fields than where resources are
monopolized by a single agency. Coordination of results into a coherent
theoretical scheme, which orders sub-fields into a dominant hierarchy, is less
important here, as in chemistry perhaps, than in physics. Goals, concepts and
techniques are fairly well integrated in these sorts of fields but their ordering
into a strong hierarchy is not of great concern. Theoretical work is, therefore,
not as crucial in these sciences and the involvement of theoretical issues and
goals in the day to day conduct of research not as high as it seems to be
in physics (53). Patterns of collaboration and communication within and
between research organizations tend to resemble other formal structures of
work allocation and control with strong distinctions drawn between levels of
expertise and competence and fairly sharp differences noticeable between
skills. Shinn's
of work organization in solid state physics in contrast
with that in mineral chemistry highlights these points.
Establishments in these fields control both work goals and procedures to

Sciences as Reputationai Organizations

345

a high extent through largely impersonal and formal means, but they are not
as centralized and hierarchically structured as in the previous case. Instead,
segmentation of sub-fields with their own establishments and separate goals
seems to occur with little overt ordering of these into a single scale of prestige
and power dominated by the establishment. Within an elaborate and formalized body of knowledge and skills which defmes and orders tasks, and
evaluates the significance of results, in a relatively uncontested manner,
different groups develop research programmes in a variety of directions which
are integrated through the commonly accepted rules.
7. Changes in the Organization of Scientific Fields and the Emergence of the
Scientific Establishment
Changes in the degree of mutual dependence among practitioners in a scientific field have consequences for the degree of task uncertainty in that area
and for its intellectual and social organization. Equally, the development
of standard techniques and/or theoretical consensus and coherence change
relations between scientists and the organization of the reputational system
and its relations to employment organizations. Both of these sorts of changes
are related to the intellectual and social autonomy of the sciences and their
funding structures.
The degree of mutual dependence among scientists is related to: (a) their
dependence upon external audiences and sponsors as opposed to esoteric
and professional audiences, (b) the number of
whose work is
oriented to the goals of a particular reputational community, (c) the scarcity
of resources for accomplishing relevant tasks and, (d) the scope of the problems which can be adequately dealt with by single individuals. These factors
are interrelated as we have seen and an increase in one is likely to lead to
changes in others and in the overall structure of the field. If, for instance,
the number of practitioners seeking reputations in a particular field, which
already has a fairly formalized communication system, increases, the scope
of problems tackled is likely to decrease as specialization of tasks and skill
becomes higher. Also, technical precision and standardization are likely to be
emphasized, thus reducing technical task uncertainty, as means of ensuring
comparability of results and coordination of task outcomes. Relative scarcity
of resources - which is likely to be reinforced by the move to greater precision

346

Richard Whitley

and complex techniques - will force scientists to orient their work more
towards each other and claim reputations on the basis of contributing to
community goals. Common and standard technical procedures facilitate such
claims and the award of mutual recognition. Provided the environment
is fairly benign, then, overall autonomy is likely to increase and the field
become more professionalized. To some extent, this seems to have happened
in some of the humanities since the 1950s; work has become more technical
and specialized, remote from lay concerns and language and specific to
scholarly goals. Diffuse interpretative schemes seem less manifest, at least in
Anglo-Saxon countries.
The impact of increased numbers of practitioners seeking reputations from
a single community varies in effect according to the extent of technical and
theoretical autonomy of the field, the relative scarcity of resources, especially
jobs, and the existence of a technical, standardized communication system.
This variability can be seen by a brief comparison of Anglo-Saxon sociology
and economics. The expansion 0 f university posts in both of these disciplines
in the 1960s and early 1970s has resulted in a fragmentation of the former
field and a stronger emphasis on mathematization and technical competence
in the latter.
The lower technical autonomy of sociology, its greater reliance on tacit
skills and personal relationships and the lack of a strongly institutionalized
symbol system for communicating and comparing results meant that the
general availability of jobs and other research resources decreased mutual
dependence and encouraged goal conflict and differentiation. Because posts
were much more widely obtainable, the need to acquire a reputation across
the whole field through conducting research according to the central, established tradition decreased, and sub-groups based on divergent goals,
methological beliefs and distinct technical procedures were able to control
access to jobs and so reputational organizations were established around
sub-goals. Personal contacts and allegiances among the establishment were
insufficient to control the expanded organization and the development of a
more formal and standardized style of research which could be inculcated
throughout the training system as the basis of "normative integration" was
inhibited by the expansion of jobs. Central authority declined, specialization
and differentiation occurred through general methodological orientations and
conceptual frameworks - often linked to ideological views and political

Sciences as Reputationai Organizations

347

movements - rather than by specialized techniques and narrowing research

foci, and established cognitive and social boundaries declined in strength.
General, diffuse interpretative schemes became more respectable, as did
evaluative criteria linked to non-scientific spheres; and new intellectual
movements were established which were opposed to the dominant goals,
often at local institutions (54).
The subsequent decline in availability of posts in sociology may well result
in practitioners becoming more dependent upon each other for reputations
and so lead to greater standardization of techniques and attempts to institutionalize a common training system in all departments. However, the closeness of many sociological concerns and procedures to everyday issues and
concepts, and its apparent lack of success in controlling its own environment
through the development of common skills, render external audiences and
establishments highly relevant to the conduct and evaluation of sociological
work. Dependence of the professional establishment is therefore mitigated by
the low degree of intellectual and social autonomy of the field. As Norbert
Elias points out, sociology competes with other powerful establishments
to provide means of orientation and this competition is likely to inhibit
professionalization processes as long as practitioners can claim reputations
from external audiences as legitimate resources in their bids for jobs and
promotion. The "polyparadigmatic" nature of sociology seems likely to
remain a major feature of the field even if pressures for intellectual conformity do grow with greater competition for jobs (55).
The domination of the training system in economics by a few text books
which are similar in orientation and content provides that field with a far
greater degree of normative integration than sociology manifests (56). While
not so formalized and coherent as mathematics, economics has a common
core of skills and beliefs which are taught to almost all students in university
departments and which defme and distinguish a distinct intellectual and social
identity. Adherence to, and elaboration of, this "paradigm" are essential
features of economists, even when their overt goals diverge. The strength of
this system of beliefs and practices, which is reinforced by the hierarchy of
journals and employment institutions, ensures that tasks and the evaluation
of task outcomes are more remote from lay attitudes and criteria than in
sociology and, hence, technical autonomy is higher. Given the strong emphasis on mathematics which is a major component of the dominant paradigm

348

Richard Whitley

in economics, the communication system is highly fonnalized and results

produced across a variety of settings and countries can be compared and
coordinated. In contrast to sociology, then, the expansion of posts for practitioners in economics did not result in fragmentation but rather increased
mutual dependence and a greater emphasis on technical standardization and
formalization. While some goal differentiation and opposition has occurred,
such as the rise of radical political economists, the technical vocabulary is
still shared among practitioners and actual research practices remain broadly
similar. In economics, then, the common educational experience and its
reinforcement by the power structure, coupled with the strongly institutionalized technical procedures and vocabulary, seems to have enabled the
reputational organization to expand considerably without losing its coherence
and disciplinary identity.
In other fields with higher autonomy and low task uncertainty and where
research facilities and funds are centralized and concentrated in a few organizations, the expansion of the labour force is likely to lead to increased
contlict over goals and the centrality of specialisms. Because resources are
relatively concentrated and becoming increasingly scarce with the growth in
the number of practitioners, reputations are harder to obtain in the central
subfields and at the same time the distinction between centre and periphery
is accentuated so that, for example, most of the space in the top journals is
dominated by results from central areas. Increasingly, reputations in, say,
physics will come to mean reputations in the dominant subfields and the
more subsidiary areas will be dismissed by establishment scientists as not
being "real" physics. These tendencies to strengthen hierarchical relations and
exclude peripheral goals from access to reputations resources will occasion
disputes over funding and the operation ofthe peer review system as intellectual priorities are seen to determine the allocation of scarce resources.
The growth of numbers of theoretical practitioners in such fields, and the
relative cheapness of this fonn of work and so widespread availability of
resources, encourages the development of a separate reputational community
with its own goals and positions. While this group's raison d'etre revolves
around the need to order experimental results and integrate highly specialized
task outcomes into some sort of coherent whole, once established as a separate organization the interests of its members increasingly diverge from
those of other groups. Reputations become sought amongst themselves and

Sciences as Reputationai Organizations

349

specialization of skills and topics is heightened. While claiming the right to

legislate for the whole of physics, theoreticians develop as a distinct subgroup
around goals which are remote from those of many physicists. Because of the
number of theoreticians and the general availability of research resources,
pressures to coordinate and centralize their tasks are less than for experimental physics and so research may well be more divergent among theoretical
physicists. Indeed, if the number of jobs controlled by theoreticians becomes
very large, the intellectual and social structure of this subfield may develop
into a similar form to that of mathematics with a corresponding increase in
task uncertainty.
Just as sciences vary in their degree of mutual dependence among practitioners, and these variations have consequences for the organization of
knowledge production and validation, so too the sciences as a whole have
changed in their mutual dependence with consequences for their interrelations. The more scientific fields depend upon each other for validation as
producing scientific knowledge in order to obtain access to jobs and facilities
the more important do definitions and classifications of science and sciences
become. The development of the philosophy of science as a separate activity,
which has subsequently been institutionalized in university departments in
Anglo-Saxon philosophy, was probably related to the increasing importance
of being "scientific" for academic respectability and obtaining resources
in the late 19th century. The professionalization of the sciences in the universities resulted in scientific fields becoming dependent upon academic
concepts and criteria for their status and legitimacy. To gain continued access
to resources for knowledge production, reputational organizations had
increasingly to demonstrate their adherence to the dominant academic ideal
of scientific work. Even where substantial numbers of practitioners were not
in university posts, academics tended to control the journals and substantial
numbers of trainees so they could enforce their conception of the field on
"amateurs" (57). Dependence on establishments in other sciences grew
throughout the 19th century as science became an honorific title for a new
professional group controlling the production and validation of knowledge
which was both true and useful.
The subsequent decline of the academic monopoly of scientific posts and
facilities, and the separation of reputational identities from academic ones,
has meant that academic conceptions of science and scientific ideals are no

350

Richard Whitley

longer as crucial as they were. Through their control of school curricula and
textbooks, academics still play a major part in the public image of science,
and still decide how resources within universities - especially jobs - are
allocated to different fields, but science as a major social institution producing
and validating the "means of orientation" in Elias' phrase, seems to have
transcended the university. Scientific establishments controlling reputations,
setting knowledge goals and controlling task allocation and performance are
no longer identical to academic establishments and academic science not
identical to the whole of the "best" science. Universities still produce the
knowledge producers but employing organizations also train them in specialist
skills, often for novel purposes, and coordinate their work for specific tasks
which often have no counterparts in academic science. Insofar as work goals
in these organizations are partly controlled by reputational organizations,
new scientific fields become established around non-academic goals, which
are regarded as being equally "scientific" as to those which correspond with
academic boundaries. Indeed, some of these fields, such as cancer research,
later become established in universities.
The development of new reputational organizations controlling resources
outside the universities, and on fields which cut across established academic
boundaries and priorities, is part of the general extension of science to new
areas and concerns and its redefinition as a generalized set of procedures for
producing knowledge rather than a fIXed body of knowledge about a given
reality. As science attempts to monopolize knowledge production and validation and impose its criteria of rationality and understanding upon societies,
scientific ideas increaSingly come to dominate the production of all forms of
knowledge, especially once major social establishments such as the military
and business have become convinced of the utility of the dominant form of
scientific knowledge. Forms of understanding which are antithetical to these
ideals are not scientific and so do not lead to knowledge. Consequently
nascent groups of knowledge producers attempt to prove their scientific
status, and thus gain access to jobs and research facilities, by imitating the
current image of science and organizing their work in a "scientific" manner
with the apparatus of profeSSionalized science. As science grows in influence
and prestige, so too do the ideals and methods of science become important
to all knowledge producers and the reputation of being scientific a key resource in the struggle for legitimacy. The mutual interdependence of scientific

Sciences as Reputational Organizations

351

fields thus increases as science monopolizes the means of orientation and

successfully distances itself from competing producers of cognitive structures.
Sciences as reputational organizations themselves, thus, form an increasingly strong reputational organization as the number of fields claiming scientific
status grows and resources become less widespread. The more groups depend
on their scientific status for legitimacy and resources the more likely some
sort of ranking system will become established around distinct scientific
ideals, and the more scientific establishments will seek to influence the
formulation and interpretation of those ideals. In other words, science as a
whole will come to constitute a reputational organization in which different
fields attempt to both make their guiding principles commensurate with the
dominant ideals and to impose their interpretation of scientificity on other
reputational organizations. Obviously, establishments which control central
funding agencies and have accommodated themselves to important groups
outside the sciences are more likely to be successful in this struggle than those
which have not.
The more centralized is research funding, and the more crucial is scientific
status to groups of knowledge producers, the more a central scientific establishment emerges which allocates reputations in terms of adherence to central
ideals. Given that the development of science as a major social institution
involved negotiating with other establishments, these ideals will reflect their
priorities and goals as well as those of the dominant fields. The sort of science
that such a central scientific establishment promulgates, that is, will be the
type of knowledge which "works'. and is useful in the view of dominant
groups. To gain support from other establishments in generalizing science as
the system of understanding and knowledge validation, and resources to
support this expansion, science had to demonstrate its utility and consonance
with dominant structures (58). The control over nature for military and
business purposes shown by physicists and chemists, and the model of rational administration of research which such "restricted" fields exemplified,
reinforced their domination of the image of science and their primary location
in the hierarchy of concepts and methods. To the extent that the scientific
establishment does control the award of differential reputations of being
scientific to groups of knowledge producers, and resources are largely controlled through a central system by such an establishment, these ideals
dominate the production and validation of knowledges.

352

Richaxd Whitley

8. Concluding Remarks
In suggesting that the sciences should be considered as one form of work
organization - reputational organizations - which are a subset of professional
organizations, I have tried to show how intellectual change and differences
between scientific fields can be understood in terms of variations in their
organizational structure. The professionalization of the sciences in the 19th
century tied jobs and statuses to intellectual reputations for the first time in
a systematic way. This meant that the organization of work, ofreputational
groups and of employing institutions became inextricably entwined so that
changes in the one involved changes in the others. Subsequent developments
in the employment of scientists and generalization of science to novel areas
both decreased the institutional specificity of science in universities and
revised the status of science as a fixed body of truths. As the "scientific
method", science became an abstract set of norms for producing knowledge
oriented towards a variety of goals and a variety of employing organizations
arose to control and direct such knowledge production. This variety of relations between employers, scientists and universities means that the academic
model of scientific work and "col'nmunities" is inadequate for understanding
the professionalized sciences. Viewing them as reputational organizations
which vary in their structure and relations to other organizations seems to
me to be a more useful approach than assuming either their autonomy as
self-governing groups producing certified knowledge or their inevitable
reduction to market interests.
Notes and References
1. See, for example, G. Lemaine et al. (eds.), Perspectives on the Emergence ofscientific disciplines, Paris and the Hague: Mouton, 1976; D. Edge and M. Muikay,
'Fallstudien zu wissenschaftlichen Spezialgebieten'in N. Stehr and R. Konig (eds.),
Wissenschaftssoziologie, Kohn: Westdeutscher, 1975.
2. G. Bohme, W. van den Daele and W. Krohn, 'Finalization in Science', Social Science
Information IS (1976) 301-30; R. Hohlfeld, 'Theory Development in Moleculax
Biology', in W. Callebaut et al. (eds.), Theory of Knowledge and Science Policy,
Ghen t: Communication and Cognition, 1979.
3. Especially his Structure of Scientific Revolutions, University of Chicago Press, 2nd
ed., 1970; 'The Function of Measurement in Modem Physical Science' in the
collected essays The Essential Tradition, University of Chicago Press, 1977 and

Sciences as Reputational Organizations

4.

5.

6.

7.
8.
9.
10.
11.

12.

13.
14.

353

'Second Thoughts on Paradigms' in F. Suppe (ed.), The Structure of Scientific

Theories, University of Dlinois Press, 1974.
See, for example, N. Elias, 'The Sciences: Towards a Theory" in R. Whitley (ed.),
Social Processes of Scientific Development, London: Routledge and Kegan Paul,
1974; H. Martins, 'The Kuhnian "Revolution" and its implications for Sociology' in
T. Nossiter et al. (eds.), Imagination and Precision in the Social Sciences, London:
Faber and Faber, 1972.
E.g., B. Latour and S. Woolgar, Laboratory Life, London: Sage, 1979; the papers
by Knorr, Latour, Callon, Pinch, Harvey, Pickering and Travis in K. Knorr et al.
(eds.), The Social Process of Scientific Investigation, Sociology of the Sciences
Yearbook, Vol. 4, Dordrecht and Boston: Reidel, 1980.
For a discussion of some effects of changes in funding and control of the 19thcentury French Science Faculties, see T. Shinn, 'The French Science Faculty
System, 1808-1914: Institutional Change and Research Potential in Mathematics
and the Physical Sciences', Historical Studies in the Physical Sciences, Johns Hopkins University Press, 1979; the importance of industrial employment for the
expansion of academic physics in the 1920s in the U.S.A., and the efforts of elite
physicists to maintain the link between academic and industrial physicists, are
emphasized by S. Weart, 'The Physics Business in America, 1919-1940' in N.
Reingold (ed.), The Sciences in the American Context, Washington, D. C.: Smithsonian Institute, 1979.
These reforms have been discussed by a number of writers in English recently.
See: C. E. McOelland, State, Society and University in Germany 1700-1914,
Cambridge University Press, 1980, for a useful analysis.
As discussed by Stinchcombe, see: A. Stinchcombe, 'Bureaucratic and Craft Administration of Production', Administrative Science Quarterly 2 (1959) 168-87.
The importance of standardizing and monopolizing the production of producers for
the establishment of the service professions is emphasized by M. S. Larson, The
Rise of Professionalism, University of California Press, 1977, Ch. 2.
Kuhn,op. cit., Note 3; W. O. Hagstrom, The Scientific Community, New York:
Basic Books, 1965; D. Crane, Invisible Colleges, University of Chicago Press, 1972.
W. van den Daele, 'The Social Construction of Science: Institutionalization and
Defmition of Positive Science in the Latter Half of the Seventeenth Century',
in E. Mendelsohn et al. (eds.), The Social Production of Scientific Knowledge,
Sociology of the Sciences Yearbook, Vol. 1, Dordrecht and Boston: Reidel, 1977.
The examples of France and Eastern Europe, and pre-19th-century science, indicate
the contingent nature of the university institutionalization of sciences. Kevles
has argued for the positive effect of the U.S. Geological Survey in imposing high
standards on work in Geology and the negative results for physics of being restricted
to colleges and universities, see D. J. Kevles, 'On the Flaws of American Physics',
in G. Daniels (ed.), 19th Century American Science, Evanston: Northwestern
University Press, 1972.
See: F. K. Ringer, The Decline of the German Mandarins, Harvard University Press,
1969 and McOelland, op. cit. (Note 7), 1980, Chs. 3 and 4.
P. G. Wersky briefly discusses the domination of "high science" ideals at Cambridge
University in his The Visible College, London: Allen Lane, 1978; see also: K.-H.
Manegold, 'Technology Academised: Education and Training of the Engineer in the

354

15.
16.

17.

18.

19.
20.
21.

22.
23.

24.
25.

26.
27.

28.
29.

Richard Whitley

19th Century', W. Krohn et al. (eds.), The Dynamics of Science and Technology,
Sociology of the Sciences Yearbook Vol. 2, Dordrecht: Reidel, 1978.
J. B. Morrell, 'The Chemist Breeders: The research schools of Liebig and Thomas
Thomson', Ambix 19 (1972) 1-46.
Partly because of the need to monopolize the production of producers and to
justify that monopoly through institutionalizing knowledge and skill barriers, and
also to claim separate, esoteric expertise which can only be obtained through the
professional certification agency.
This was especially the case in 19th-century Germany of course; see the books by
McClelland and Ringer already cited in Notes 7 and 13 respectively and the two
papers by R. S. Turner, 'The Growth of Professorial Research in Prussia, 18181846, Causes and Context', Historical Studies in the Physical Sciences 3 (1971);
'University Reformers and Professorial Scholarship in Germany', in L. Stone (ed.),
The University in Society, Princeton University Press, 1974.
McClelland clearly links the particular form the institutionalization of scholarship
took in Germany, and its subsequent high social prestige, to 'an attempt to stabilize
and legitimate the rule of the more flexible part of the aristocracy with the aid of a
small elite recruited from the middle class', op. cit., Note 7, p. 98.
McClelland,op. cit., Note 7; Morrell, op. cit., Note 15.
As suggested by Terry Shinn, op. cit., Note 6.
See, among others, M. Berman, , "Hegemony" and the Amateur Tradition in British
Science', Journal of Social History 8 (1975) 30-50; R. Porter, 'Gentlemen and
Geology: the Emergence of a Scientific Career, 1660-1920', The Historical Journal
21 (1978) 809-36.
Morrell,op. cit., Note 15.
The academic professionalization of the field sciences developed later than in
physics and chemistry and, in general, occurred later in England; see, for example,
J. G. O'Connor and A. J. Meadows, 'Specialization and Professionalization in
British Geology', Social Studies of Science 6 (1976) 77-89; Porter, op. cit., Note
21.
Morrell,op. cit., Note 15.
C. Jungnickel, 'Teaching and Research in the Physical Sciences and Mathematics in
Saxony, 1820-1850', Historical Studies in the Physical Sciences 10 (1979).
Ibid., pp. 41-2.
See the papers cited in Note 21 and D. Edge, 'Some Sociological Reflections' in
M. Hoskin (ed.), General History ofAstronomy Vol. 3, Cambridge University Press
(forthcoming). S. F. Cannon has a detailed discussion of preprofessional science in
England in her Science in Culture, New York: Neale Watson Publications, 1978,
.
Chs. 5, 8 and 9.
Cf. Jungnickel, op. cit., Note 25; Berman, op. cit., Note 21 and his Social Change
and Scientific Organization, London: Heinemann, 1978; Porter, op. cit., Note 21.
As discussed by Berman in his historical account of the Royal Institution in Social
Change and Scientific Organization, op. cit., Note 28; see also R. H. Wiebe, The
Search for Order 1877-1920, New York: Hill and Wang, 1968, especially Chapter

6.
30. The standardization and simplification of experimental techniques by Liebig and
others was a key condition of the "finalization" of science. As W. van den Daele

Sciences as Reputational Organizations

31.

32.
33.

34.

35.

36.
37.
38.

355

and W. Schafer point out, an effective division of labour and organization of

research for a variety of goals require such standardization, and the successful
development of agricultural chemistry depended on the ability to conduct long
series of standard experiments. See their 'The Origins and Structure of Agricultural
Chemistry'in G. Lemaine et al. (eds.), op. cit., Note 1. The successful development
of restrictedness in the sense of technical control of cognitive objects, as discussed
by Rip in this volume, also seems a crucial factor for similar "goal oriented" sciences
to become institutionalized. Davy's work at the Royal Institution failed to establish
agricultural chemistry as an effective scien tific field partly because organic chemistry
had not developed effective techniques for systematically analyzing the composition
and decomposition of organic compounds; see W. van den Daele and W. Schafer,
pp. 32-4, and Berman, op. cit., Note 28, 1978, Ch. 2.
Even in England academic geologists and physicists and chemists began to dominate
their fields by the late 19th century, see O'Connor and Meadows op. cit., Note 23;
and the papers cited in Notes 21 and 27. See also, R. Moseley, 'Tadpoles and Frogs:
Some Aspects of the Professionalization of British Physics 1870-1939', Social
Studies of Science 7 (1977) 423-45; C. A. Russell et al., Chemists by Profession,
the Origins and Rise of the Royal Institute of Chemistry, Milton Keynes: Open
University Press, 1978; R. Kargon, Science in Victorian Manchester, Manchester
University Press, 1977, Chs. 5 and 6.
Kuhn,op. cit., Note 3, 1977.
State laboratories have been established in most European countries for a variety of
purposes since the end of the 19th century, but the expansion of direct government
funding of research for combined political and scientific goals since the 1940s has
been so great as to produce a new system of scientific research which transcends
academic science.
See W. van den Daele et al., 'The Political Direction of Scientific Development', in
E. Mendelsohn et al. (eds.), The Social Production of Scientific Knowledge, Sociology of the Sciences Yearbook Vol. 1, Dordrecht: Reidel, 1977 and R. Johnston and
T. Jagtenberg, 'Goal Direction of Scientific Research', in W. Krohn et al. (eds.),
The Dynamics of Science and Technology, Sociology the Sciences Yearbook Vol. 2,
Dordrecht: Reidel, 1978.
The importation of restrictedness into some sciences, as discussed by Rip, can thus
be seen as part of general professionalization strategies which not only increase
control and success in reducing uncertainty but also exclude competitors and lay
audiences.
Connections between employment status, scientific identities and beliefs, and their
changing interrelations are discussed by R. Krohn, The Social Shaping of Science,
Westport, Connecticut: Greenwood, 1971.
Local considerations and contingencies are stressed by B. Latour and S. Woolgar,
op. cit., 1979; see also K. Knorr, 'Tinkering Towards Success, Prelude to a Theory
of Scientific Practice', Theory and Society 8 (1979) 347-76.
These are similar to those developed by Randall Collins, Conflict Sociology, New
York; Academic Press 1975, Chapter 9. Collins does not, however, emphasize the
importance of controlling jobs as much as I have done and tends to focus more
on theoretical task uncertainty at the expense of technical means of reducing
uncertain ty.

356

Richard Whitley

39. Hargens distinguished between functional and normative integration of scientific

fields in his comparison of mathematics, chemistry and political science. See, L. L.
Hargens, Patterns of Scientific Research, Washington D. C., American Sociological
Association, 1975, pp. 7-9.
40. As discussed by Gaston, see J. Gaston, Originality and Competition in Science,
Chicago University Press, 1973.
41. See, for example, I. Grabner and W. Reiter, 'Guardians at the Frontier of Science',
in H. Nowotny and H. Rose (eds.), Counter-Movements in the Sciences, Sociology
of the Sciences Yearbook, Vol. 3, Dordrecht: Reidel, 1979; J.-M. Levy-Leblond,
'Ideology of/in Contemporary Physics' in H. Rose and S. Rose (eds.), The Radicalization of Science, London: Macmillan, 1976; G. Swatez, 'The Social Organization
of a University Laboratory', Minerva 8 (1970) 36-58.
42. As characterized by Hargens, op. cit., Note 39, 1975;W. O. Hagstrom, The Scientific
Community, New York, Basic Books, 1965; 'Competition in Science', American
Sociological Review 39 (1974) 1-18; C. S. Fisher, 'Some Social Characteristics of
Mathematicians and their Work', American Journal of Sociology 78 (1973) 1094118.
43. Collins,op. cit., Note 38, pp. 340-5.
44. As in, for example, H. Katouzian, Ideology and Method in Economics, London,
Macmillan, 1980; M. Hollis and E. Nell, Rational Economic Man, Cambridge
University Press, 1975; N. Armistead (ed.), Reconstructing Social Psychology,
Penguin Book, 1974. See also many articles in the Journal of Economic Issues
which is the main vehicle for attacks on orthodox economics by the "institutionalist" schoo lin the U.S.A.
45. As achieved through a common, integrated undergraduate and postgraduate curriculum, for example, see Fisher, op. cit., Note 42.
46. See, for example, B. Latour and S. WOQlgar, op. cit., Note 5; the papers by Knorr,
Latour and Travis in the Sociology of the Sciences Yearbook, Vol. 4, Dordrecht:
Reidel, 1980 and Knorr, op. cit., Note 37.
47. Replicability is therefore seen as varying across sciences as a consequence of the
degree of technical control achieved by a field and the amount of background
knowledge which is taken for granted.
48. Restrictedness in the sense of technical control is thus a means of reducing task
uncertainty and ambiguities. This enables a systematic division of labour and formal
means of organizational control to become institutionalized but I am not sure that
this always follows, or that restrictedness is the only path to bureaucratization of
research.
49. Berman,op. cit., Note 28,1978; See also Kargon, op. cit., Note 31.
50. Overtraining practitioners to maintain professional status is discussed by Larson,
op. cit., Note 9, pp. 230-1.
51. The examples discussed by Dolby in his paper in this volume illustrate differences
in autonomy and distance strategies between this case and the first one.
52. Work by Collins, Pinch and others has shown the socially contingent nature of such
decisions in physics but this need not imply that experimental results there are as
ambiguous and open to opposed descriptions as in other fields. See H. M. Collins,
'The Seven Sexes: a Study in the Sociology of a Phenomenon or the Replication
of Experiments in Physics', Sociology 9 (1975) 205-324 and the papers by Pinch,

Sciences as Reputational Organizations

53.

54.

55.
56.
57.
58.

357

Pickering and Harvey in the Sociology of the Sciences Yearbook. Vol. 4. Dordrecht:
Reidel. 1980.
As reported by Shinn in this volume. Others comment on the growing separation of
theoreticians and experimentalists in physics. which may lead to theoretical physics
becoming organized like mathematics as a distinct reputational organization: see
Grabner and Reitner. op. cit . and Levy-Leblond. op. cit . Note 41.
Ethnomethodology is the most obvious example of this. cf. N. Mullins. 'The Development of Specialties in Social Science: the Case of Ethnomethodology. Science
Studies 3 (1973) 245-73. See also K. Knorr. 'The Nature of Scientific Consensus
and the case of the Social Sciences'. in Knorr et al. (eds.). Determinants and
Controls of Scientific Development. Dordrecht and Boston: Reidel. 1975. In his
Theories and Theory Groups in Contemporary American Sociology. New York:
Harper and Row. 1973. Nicholas Mullins suggests that causal modelling will become
the new integrating force in North American Sociology which fits in with my view
that increased numbers can increase mutual dependency and lead to technical
standardization. He also claims that (p. 305) the new theories are method centred
rather than being differentiated by topics. If so. this suggests that largely common
sense foci are being replaced by more esoteric. "professional". ways of differentiating tasks. While I think this is partly what has happened. the relative lack of
autonomy from common sense concerns and issues in sociology mitigates such
professionalization tendencies as Mullins shows in his discussion of the "radicalcritical". "social forecaster" and other new theory groups.
See. C. J. Lammers. 'Mono- and Poly-paradigmatic Developments in Natural and
Social Sciences'. in R. Whitley (ed.). Social Processes of Scientific Development,
London: Routledge and Kegan Paul, 1974; Martins, op. cit., Note 4.
Katouzian,op. cit . Note 44.
As in Geology and Physics, see Porter, op. cit., Note 21. O'Connor and Meadows,
op. cit . Note 23; Moseley, op. cit., Note 31.
This is scarcely a recent phenomenon, of course, but the accommodations made by
natural philosophy in the 17th century .were not those of professionalized science.
The development of organized, directed knowledge production linked to reputational organizations, and the monopolization of knowledge production and validation by professional scientists, have resulted in the entrenchment of particular
scientific ideals - which are supported by other establishments - throughout
the research system. The more dependent scientific fields are upon each other for
legitimacy and reputations. the more the image of science which is congenial to
dominant groups will dominate all knowledge production as access to jobs and
facilities is denied to "amateurs", "pseudo-scientists" and "ideologists".

INDEX

Abir-Am, P. 141
academic establishments 62
institutions, dogmatism 104
repressive nature 103
success, disillusion 102
Academy of Sciences of U.S.A. ix
advice from 111-119
advisory role 114
dietary advice, controversy 116
expansion of 112
origins of 111
supreme court of science 11"4
Academy of Sciences of U.S.S.R.
development of 113
links with state 45
Academy of Sciences, Prussian 112
West German, development 113
Agassiz, L. 118
alchemy 224
Aleksander, I. 211
Allen, E. 291,292
Allen, G. 142,165,217
al-Mulk, Nizam 94
Alt, F. L. 211
Altholz, J. L. 290,291
Ambler, P. A. 215
Anderson, B. 196,216
Appleton, E. 204
'AqnY 107
Aquinas, St. Thomas 93
Ardrey, R. 285,291
Armer, P. 212
Armistead, N. 356
artificial intelligence ix, 169-217
attacks on 205
"big four" in 178
conflict in 172
department of, breakdown of 189
first 186
development
conferences and 186

early, in U.S.A. 176

in U.K. 182
emotive nature of 174
establishment of 172
funding and 181
in U.K. 199,208
workshops and 200
in U.S.A. 178,208
expert systems framework 193
funding, in U.K. 190
policy and 173
goal-oriented approach 188
interdisciplinary nature 207
nepotistic hierarchy 181
organizational structure 201
paradigmatic structure 170, 208
professor of 190
protoparadigmatic structilre 177
research programme, in U.K. 184
robot project 186
sociocognitive characteristics 169
survey of, in U.K. 190
technology of 173
Ashby,W.R.182,213
Astbury, W. T. 132,133, 165
Augustine, St. 93
autonomy of academic science ix
of groups and their theories 58-60
of scientific establishments viii
Averroes (ibn al-Rushd) 108
Avery, R. W. 86
Avicenna 95,96,106
Ayres, C. 107
Baldwin, J. M. 291
Bar-Hillel, Y. 211
Barnes, B. 86, 234, 235, 308
Bateson, C. 130
Battelle Institute (sic) 75
Baxter, J. B. 142
Baxter, R. 139

359

360
Beadle, J. 133
Bearden, C. 196
Becquerel, H. de 226
Bell, D. 85, 86
Ben-David, J. 118,264
Bentham, J. 106
Berg, P. 123
Berlin, I. 107,108
Berman, M. 289, 354
Bernal, J. D. 137, 142, 143, 219, 234
Bhaskar, R. 220, 234
biochemistry, service-like 132
biology, differing notions 129
programme management 134
research strategy 131
biomedical research 226
growth, from political in-fighting
135
strategic planning of 154
biotechnology as trend in molecular
biology (q.v.) 124
Bitz, A. 263
Black, M. 264
Blume, S. S. 86, 234, 307
Boas, M. 236
Boden, M. A. 197,198,204,216
Boffey,P. 117,118
Bohme, G. 165, 220, 227, 228, 234,
235,236,307,309,352
Bold, A. M. 237
Bornat, R. 195
Bourdieu, P. 308
Bowen, F. A. 290
Box, S. 166
Brady, J. M. 195,196,212,217
Breuer, H. 166
Brewster, D. 276, 290
Brickman, R. 118
Briggs, A. 197,216
Brody, J. E. 119
Broesterhuizen, E. 237
Brooker, R. A. 193,195,200,213,215,
216
Brooks, H. 73, 117
Brown, E. R. 136,141,143
Brown, S. C. 118
Browne, E. G. 95, 107
Bulthaup, P. 235

Index
Bundy,A.216,217
bureaucratization 4
Burns, T. 263
Bush, V. 134
Burstall, R. M. 190,195,197,213,216
Cairns,1. 142
Caldin, E. F. 263
Callebaut, W. 166,238,352
cancer, occupational 159
research ix
in Germany 145-168
basic vs. targeted 150
clinical 155
epidemiology and 156, 158
experimental 151
medical opinion and
156
molecular biology and 146
patient-oriented 154
political debate 145
political planning 149
power structure 160
Cannan, E. 108
Cannon, S. F. 354
Cannon, W. B. 127
Carlson, E. A. 131,142
Carnegie, A. 125
cause, Kant on 9
cell biology, cancer research 151
Chambers, R. 273, 277, 278, 280, 290
. Chargaff, E. 143
chemists, social status 233
Chubin, D. E. x, 164, 211, 293-311
civilization, acceleration of 11
time perception and 18
Clark, J. W. 290
Clowes, M. B. 185,197,199,204,214,
216
cognitive organization, limitations 253,
257
cognitive psychology 198
cognitive studies 197
Cohen, L. J. 308
Cole, J. 310
Cole, S. 310
Collins, H. M. 222,235,237
Collins, N. L. 213
Collins, R. 264,355

361

Index

computer facilities, in U.K. 183

funding 184
Comte, A. 106
Condorcet 106
conceptualization 9
Connolly, T. x, 211, 293-311
Cotgrove, S. 166
Coulson, C. A. 187
Council for Science & Technology,
Japan 113
Cox, G. 140
Craik, R. J. W. 182,204,213
Crane, D. 308
Crick, F. 132
Culliton, B. J. 87
culture 267-292
Curie,P. andM. 226
Cuvier, G. 29, 273
cybernetics, in post-War U.K. 182
precursor of artificial intelligence
176
Dae1e, W. van den 86, 164, 165, 220,
227, 228, 234, 235, 236,
307,309,317,352
Daiches, D. 216
Dalton, E. 336
Dalton, J. 235
Daniels, G. 118
Darwin, C. (artificial intelligence) 198
Darwin, C. (evolution) 103, 274, 278,
279,281,290
Darwin, F. 290,291
Darwinian evolution 130
Darwinism, debate over 270
sociobiology and xi
Davy, H. 336,355
Debus, A. G. 236, 238
Delbriick, M. 132,133
Deppe, H. U. 168
Descartes, R. 7,10, 17
Deutsch, K. W. 309
Dobzhansky, T. 308
dogma 97
hallmark of scholasticism 102
dogmatism of academic institutions 104
Dolby, R. G. A. x, 107, 236, 267-292
Doran, J. E. 196,197,216

Dreyfus, H. 1. 212,217
Drosophila 130
Dubos, R. 142
Dupree, A. H. 118
Dijkstra, E. W. 217
Edge, D. O. 171,211,295,306,352,
354
Edison, T. A. 47
Edsall, D. 127
education, computers in 194
effects, restricted 227
Einstein, A. 14, 32, 34
Elcock, E. W. 185, 193, 204, 214
Elias, N. D. viii, xi, 3-69, 85, 86, 111,
114,117,118,124,141,168,
174,175,181,207,211,212,
235,261,263,268,269,289,
307,311,353
elites, admission to 272
Emerson 103
Engel, G. 1. 167
environmental protection 74
epidemiology 156
of cancer 158
Ernst, G. W. 212
establishments, political and scientific
45-48
scientific 3-69
development of 48-52
historical 3-4
orientation control by 37-44
theory of 25-37
evolutionary theory 33, 297
debate, 19th century 282
Vestiges (q.v.) and 278
evolution (pre-Darwin) 272
experimental operations, social relations
and 253,257,259
expertise, institutional basis 74
policy making and 78-80
Faraday, M. 103,336
Feather, N. 214
Feigenbaum, E. A. 206,211
Feldman, J. 211
Ferreira, P. 237
Fisher, C. S. 356

Index

362
Fleck, J. ix, 169-217, 239, 261, 310,
334,342
Flexner, S. 126,127
Flowers, B. 189
fork bending, scientific belief challenged
by 97
Fosdick, R. B. 126,141
Foster, J. M. 185,193,195,214
Fox, R. 285, 292
funding
influence on intellectual development
x

strategy, research orientation governed by 127

Galbraith, J. K. 86
Galileo 7
Gaston, J. 356
Gatlin, L. L. 165
gene as information molecule 134
genetic(s)
heredity and 130
inheritance, Darwinian evolution and
130
manipulation, Nobel
award for
123
mutation, as central issue in biology
131
geocentric world-view 11
George, F. H. 182,213
Georgescu-Roegen, N. 264
German Association of Engineers (VDI)
76
Atomic Energy Commission 86
Ghazzali, Imam Muhammad 94
Gilbert, G. N. 308
Gilbert, W. 123
Gillispie, C. C. 290
Gilpin, R. 85, 117
Glass, B. 291
Goodfield, J. 143
Gould, S. J. 299,308
Gouldner, A. W. 71,85,311
Grabner, I. 356
Gray, A. 290
Gray, P. M. D. 214
Greenberg, D. S. 85,86, 118,310

Gregory, R. L. 186,187,188,197,214,
215
Griffith, B. C. 217
Groeneveld, L. 311
Groenewegen, P. 237
Guerlac, H. 236
Gummett, P. 143
Hagstrom, W. O. 293, 307, 353, 356
Hall, G. S. 291
Hall, T. S. 141
Halsbury, Lord 182,213
Hanson, N. R. 264
Haraway, D. J. 142
Hardy, S. 196
Hargens, L. L. 356
Harre 264
Hartley, B. 166
Harwood, J. 141
Hausen, H. zur 166
Hay,C. ix,I11-U9
Hayes,P.J.190,195,214,215
Heirich, M. 142
Heisenberg, W. 64,65,69
heliocentric world-view 11
Henry, J. 112,118
heredity, genetics and 130
Herschel, W. 274
Hesse, M. 264
hierarchies
in scientific establishments vii
research trails and 300
Hill, A. V. 134
Hill, K. 118
Hodgkin, L. 141
Hohlfeld, R. ix, 145-168, 309, 352
Holden, C. 310
Hollis, M. 356
Homburg, E. 238
Hoskin, M. 354
Howe, J. A. M. 187,194,210,214
Hubble, C. 14
Hufbauer, K. 236
Hughes, T. M. 290
Hull, D. 165
Hume Hall, R. 237
Hutchins, E. 86

Index
Hussed, E. 10
Huxley, A. 123,141
Huxley, J. 289
Huxley, L. 290
Huxley, T. H. 270, 271, 279, 281, 283
hybrid community 78,290,291
ibn Sina, Abu Ali (Avicenna, q.v.) 95
ideas, innateness of, Kant's opinion 8ff
inquiry, morphology of 240
institutionalism, hallmark of scholasticism 102
instruments, analytic 228
synthetic 228
intellectual barriers, construction of
270
public/scientist 267-292
intellectual elitism 269
Jackson, D. 143
Jacob, F. 129,137,142,143,297,308
Jannich, P. 220,234,236
Janicke, M. 168
Johnson-Laird, P. M. 198,216
Johnston, R. 143,308
Jolliffe, C. 183
Judson, H. F. 143
Jungnickel, C. 320, 354
Kaiser-Wilhelm-Gesellschaft 125
Kant,I. 7,8,13,17,67,68,289
Kargon, R. 355
Katouzian, H. ix, 89-109, 356
Katz,J.87
Kendrew, J. 139
Kerr, S. 307
Kevles, D. J. 118,141,353
Kilminster, R. 67
Kitschelt, H. 86
Klerman, G. L. 167
Knight, D. 238
Knorr, K. D. 211,234,296,297,307,
308,310
knowledge, scientific
acquisition of 15
institutionalization of viii
theories of 14, 62-65

363
Kohler, R. E. 126,127,132,141,142,
165
Koller, N. 311
Konig, R. 352
Kowalski, R. A. 190,214
Kraft, P. 217
Krohn, R. C. 86
Krohn, W. 86,164,165,234,309,352
Kuhn, T. S. 36,50,100,101,102,108,
165,224,231,232,233,235,
237,238,294,307,310,313,
316
Kuklick, H. 289
Kunze, P. 67
Kiippers, G. 86,87
laboratory
cognitive organization in 254, 258,
260
cognitive pattern 242, 243, 246, 250
data and materials flows in 254
hierarchy 240, 244, 248
communication ladder in 241,
245,249
lower echelons, importance 242,
246,250
liaison in 254
morphology 239,244,247
organization 240,244,247
Lakatos, I. 100
Lakoff, S. A. 85
Lamarck, J. B. 273,278
Lane, R. E. 72,85,86
Laplace 13,106
Lapp, R. E. 85
Larson, M. S. 353
Latour, B. 227, 236, 264, 308, 353
Laudan, L. 310
Lauman, E. O. 311
Lavoisier, A. 225
Law, J. 211
Layton, E. T., Jr. 234
Leibniz 8, 17, 28
Lemaine, G. 307,309,352,355
Lepenies, W. 238
Levidow, L. 141
Leydesdorff, L. 238

364
Libich, S. 164
Liebig, J. W. 318,320,321
life
as information process 137
differing notions of 129
genetic reductionist concept of 132
synthesis of - front-page news 139
Lighthill, J. 190, 193, 200, 205, 215,
217
Lillie, F. R. 127
Linnaeus 235
Locke, J. 103
Loeb, J. 129,130
Long, T. D. 118
Longuet-Higgins, H. C. 186, 187, 188,
197,201,204,206,210
Lorenz, C. 62
Lundgreen, P. 86
Luria, E. E. 166
Luria, S. 133
Li.ith, P. 168
Lwoff, A. 142
Lyell, C. 273,275,290
Lysenko 269,289
MacDonald, D. K. C. 263
MacKay, D. 182
MacLeod, R. 309
Macpherson Stalker, G. 263
Madow, W. G. 309
Malvern, R. R. E. 214
Manegold, K. H. 353
Markle, G. E. 309
Martin, B. R. 211
Martins, H. 107,353
Marx, K. 104
mass communication, molecular biology
and 138
Mates, B. 264
mathematization 65
McAlpine, A. 263
McCarthy, J. 176,177,178,179,182,
190, 206, 211, 212, 214, 215
McClelland, C. E. 353
McCorduck, P. 212
McCulloch, W. S. 179, 182,211
Meadows, A. J. 354

Index
medical opinion of cancer research 155
medicine and cancer research in Germany
145-168
Medvedev, Z. A. 118
Meltzer, B. 184, 185, 189, 194, 199,
200,204,206,213,214
Mendel, G. 130
Mendelsohn, E. 118,164,234,262,289
Merton, R. K. 212,289,294,307,310
Meselson, M. S. 118
metaphysics
ascendancy of 10
beliefs of 13
philosophers and 6-17
Mey, M. de 166,238
Michaelson, S. 214
Michie, D. 182, 183, 185, 187, 188,
193,199,200,204,206,210,
213,214,215,216,217
microbiology 133
Mikulich, L. 1. 215
Mill, J. S. 103,274
Millhauser, M. 290
Minsky, M. L. 177,178, 179, 182, 194,
212
Mitroff, I. I. 310
Mittelstrass, J. 220,234
Mittermeir, R. 310
molecular biology ix
as goal-oriented research 136
cancer research and, in Germany
146
establishment of 123-143
as discipline 124
fundamental nature of 137
historical development of 125, 147
mass communication and 138
new trends in 123
research frontiers of 148
T.V. shows on 139
molecular structure of biological molecules 133
Montesquieu 106
Moore, W. E. 86
Morgan, E. 132,284,285,291
Morgan, T. 130,131
Morrell, J. B. 320, 354

365

Index
Morris, D. 285, 292
Moseley, R. 355
Muller, H. 1. 131,142
Mueller, I. 264
Mulkay, M. 1. 171,211,217,295,306,
307,308,309,311,352
Muller 132
Mullins, N. C. 142, 217, 311, 357
Multhauf, R. P. 235, 236
Murray, A. M. 214
mysticism, Islamic 94.
Nagel, E. 263
National Science Foundation of U.S.A.
135
Needham, 1. 69
Nelking, D. 87
Nell, E. 356
Newell, A. 177, 178, 179, 184,204,
212
Newton, I. 17, 32, 33
Nossiter, T. 353
Nowotny, H. 87,107,289
O'Connor, 1. G. 354
Oettgen, H. F. 166
Olby, R. C. 142,166,307
Oleson, A. 118
opportunism in research 296
orientation, control of scientific 37-44
Origin of Species, Vestiges (q.v.) and
279
Overington, M. A. 309
Paneth, F. A. 253,263
Pantin, C. F. A. 220, 221, 234, 235,
263
Papert, S. 194,212
Pasteur, L. 103
Pauling,L. 132,133
Petersen, C.. 309
Perutz, M. 133
pharmaceutical industry, biomedical research 154
philanthropists, American, research strategyand 125
philosophers, metaphysics and established 6-17

philosophy
relation with physics and sociology
56-67
time as problem for 17-25
transcendental, ascendancy of 1017
errors of 8ff
mind-set of 12
physician, mind-set of 157
physicists, social status 233
physics
mathematization of 222
relation with philosophy and sociology 56-67
physiology 132
Pickstone,1. 141
Pickvance, S. 136,143
Pinxten, R. 166, 238
Pitts, W. 179,211
Plotkin, G. D. 190,214
Polanyi, M. 165,290
political ambition 79
politicization of science 73
consequences of 80-84
Popper, K. 10,68
Popplestone, R. 1. 213
Porter, G. 215
Price, D. K. 72,85,119,294
professionalization ix
professors 4,5
as ruling group vii
progressive ideology, American 125126
purity
as restricting concept 223, 225
of effects 227
Pursell, C. 142
Putnam, H. 263
Pyenson, L. 233, 238
radioactivity research, early 226
Rae,1. 108
Rappaport, A. 264
Ravetz, 1. R. 86,263,264,310
Reason, triumph of, as dogma
Reingold, N. 118,141
Reiter, W. 356
reproducibility 224

105

Index

366
reputational organization, science as
313
reputational systems, control of resources ix
research
costs 297
councils, British 125
director, computer laboratory 244
mineral chemistry laboratory 240
physics laboratory 248
elites, control of resources 136
goals, choice of 294
organization of x
payoffs 297
policy 306
investment model 300
practices 239
dormant 303
trail, funding 301
hierarchies and 300
legitimacy of object 301
problem area 295
resource access 302
training capacity 302
yield of 296
restrictedness
importation of 230
in science 219-238
institutionalization in 229
production of 223
Rettig, R. A. 143, 164
Ricardo 103
Richards, A. N. 134
Ringer, F. K. 354
Rip, A. x, 118,168,219-238,239,261,
307,336,355
Robinson, J. A. 194,216
robot research, in U.K. 190
Rockefeller Foundation 126
influence on research ix
Rockefeller, John D. 125
Roll-Hansen, N. 131,142
Romer, A. 236
Rosenberg, C. E. 142
Ross, J. 310
Royal Society 112
research strategy and 134

Rubin, L. 310
Rumi, Mawlavi 95
Ruse,M.292
Russell, B. 107
Russell, C. A. 355
Rutherford, E. 226
Sadler, J. 167
Saint-Hilaire, G. 29
Salomon, J. J. 115,118,311
Salter, S. H. 187
Samuel, A. L. 177
Sandewall, E. 211
Sanger, F. 123
Schiifer, W. 311, 355
Schaffner, K. 165
Schelling, T. C. 294,307
Schelsky,H. 86
Scherer, E. 167
Schmidt, C. G. 164,166
Schoenheimer 133
scholasticism
hallmarks of 102
Islamic 93
rise of 94
science compared to ix, 89-109
Schon, D. A. 307
Schrodinger, E. 132
science
academic domination 323
variations 325
ambiguity reduction in 222
application of, negative effects 230
as anti-dogmatism 98
as knowledge construction industry
267
autonomy of pure 267-292
basic, and cancer research in Germany 145-168
Council of Japan 113
de-institutionalization of 84
employment structures in 314
heterogeneity of 231
Islamic 93
as product of religious doctrine
94
comparison with Renaissance 95

Index
science (continued)
organization 339
policy planning, in Germany 150
politicization of 117
consequences of 80-84
productive labour 220
professionalized 328
reputational organization 313-357
research trails and policy 293-311
restrictedness in 219-238
scholasticism compared with 89-109
self-centred view of 89
subjectivity of 99
technical aspects of 219
university, establishment of 317
scientific
achievement, dependence on original
thinkers 100
community, 'divergence of opinions
in 83
establishments
balance of power in 4-5
bureaucratiZation of 4
emergence of 345
organization 339
popular culture and 267-292
shift of power base within 73
judgement, non-neutrality of 82
knowledge, as power source 71
structuring of, by Universities
38-39
transmission of 40-43
practice, classic, entrepreneurial and
pragmatic 124
pronouncements, public suspicion of
268
work, legitimate or otherwise 64
scientification of politics 80-81
scientists as class 71
task interdependence 329
task uncertainty 335
Sedgwick, A. 275,276,290
Seidel, R. S. 141
Selfridge, O. G. 211
Shannon, C. E. 177, 179, 182, 205,
211,217
Shaw, J. C. 177
Sheldon, E. J. 86

367
Shinn, T. x, 201, 217, 220,229,234,
237,239-264,337,353
Shirazi, Sa'di 95
Siegenthalter, W. 167
Siekmann, H. J. 196
Simon, H. A. 177,178,179,184,204,
212
Skopp,D.233,238
Sloman, A. 198, 204, 216, 264
Smith, A. K. 87, 92, 104, 107, 108
social relations and experimental operations 253,256
sociobiology 284
critical reaction to 286
Darwinism and xi
sociology, political blocs and 55-56
relation with philosophy and physics
53-67
scientific nature of 53, 65
status of 54
Solla Price, D. de 219, 234, 307
specialization, confluence of research
trails 295
political pressure and 304
Spencer, H. 280, 282, 283
Stahl 225, 233
Stamberg 313
status in labour market 315
Staw, B. M. 310
Stehr, N. 352
Stent, G. S. 142
Stephenson, R. L. 103
Stewart, C. H. 214
Stich, S. 143
Stinchcombe, A. 315, 353
Strasser, H. 211, 234
Strickland, S. P. 135,142,164
St. Simon, A. 106
Studer, K. E. 164, 307,309
Sturtevant, A. H. 165
Suppe, F. 263
Suppes, P. 263
Sutherland, N. S. 185, 196, 199, 200,
204,210,216,217
Svedberg 133
Swann, Lord 139, 186, 187, 193, 204,
214,215
Swatez, G. M. 263, 356

368

Index

Szabadvary, F. 235,236
Szasz, T. S. 167
Szent-Gyorgii, A. 310
Tarski, W. 263
Taube, M. 212,217
Taylor, G. R. 143
Telford, T. 103
theories, law-like and process-like 8
formation 232
thinking, spectrum of forms of 60-62
Thomas, C. L. 166
Tiger, L. 285,292
time, perception of, civilization and 18
philosophy and 17-25
'tinkering' in research 298
Toulmin, S. E. 263
1,'racy, H. de 106
Trenn, T. J. 236
Turing, A. M. 182,204,206,213,217
Turner, R. S. 354
Turner, S. J. 309
Ullmann, A. 142
universality of law-like generalizations

7-8
universities, scientific knowledge and

38-39
Useem, M. 308
Uttley, A. M. 182,197,213,216
Vandamme, F. 166, 238

Vestiges of the Natural History of

Creation 273
acceptance of, by unlearned public

273
criticism of, by learned public 273
expository technique 277
Origin of Species and 279
popular science 276
rejection of, by scientists 274
Vieo, A. M. 106
Viner, J. 108
'vital process' research 127
Vogel, M. J. 142
Voltaire 95,103, 105, 106, 109
Waddington, C. H. 142

Waddington, W. H. 204
Wade, N. 236,291
Wallace, A. R. 280, 281
Wallis, R. 309
Walter, W. G. 182,213
Warren, D. 216
Watson, J. D. 142
Watt, J. 103
Weaver, W. 126, 127, 130, 132, 134,

141

Weinberg, A. M. 85,115,118,168
Weingart, P. ix, 71-87, 111, 114, 115,

117, 164, 166, 234,262, 289,

309
Weinstein, J. 141
Weizenbaum, J. 212,217
Wersky, P. G. 353
Whewell, T. 274
Whitley, P. 114
Whitley, R. D. ix, xi, 67, 107, 111, 117,
124, 141, 164, 166,210,211,
220,221,223,230,234,235,
262,263,289, 299,307,308,
309,313-357
Wielinga, B. 196
Wiener, N. 176,182,211
Wilberforce, Bishop 270,271,281
Wilkinson, B. 141
Wilks, Y. 196,216
Wilson, E. O. 284, 285, 291, 292
Wingerson, L. 236
Winner, L. 86
Winograd, T. 194,216
Wolff 8
Wood, R. C. 85
Woolgar, S. W. 227,236,264,307,308,

353
Wright, C. 117, 118
Wright, R. 85
Young, R. M. 141,280.291
Yoxen, E. J. ix, 123-143, 166, 174,

175,211,239,261,310
Zilian, H. G. 211, 234
Ziman, J. 222,235,236
Zuckerman, H. 118,310,311

SOCIOLOGY OF THE SCIENCES

A YEARBOOK
Editorial Board:

G. Bohme, Technische Hochschule, Darmstadt

N. Elias, Universities of Leicester and Bielefeld
Y. Elkana, The Van Leer Jerusalem Foundation, Jerusalem
1. Graham, Massachusetts Institu te of Technology
R. Krohn, McGill University, Montreal
W. Lepenies, Free University of Berlin
H. Martins, University of Oxford
E. Mendelsohn, Harvard University
H. Nowotny, European Centre for Social Welfare Training
and Research, Vienna
H. Rose, University of Bradford
Claire Salomon-Bayet, University of Paris
P. Weingart, University of Bielefeld
R. D. Whitley, Manchester Business School, University of Manchester
Managing Editor: R. D. Whitley

1- 1977. E. Mendelsohn, P. Weingart, and R. Whitley (eds.), The Social Production

of Scientific Knowledge.
II - 1978. W. Krohn, E. T. Layton, Jr., and P. Weingart (eds.), The Dynamics of
Science and Technology.
III - 1979. H. Nowotny and H. Rose (eds.), Counter-Movements in the Sciences_ The
Sociology of the Alternatives to Big Science.
IV - 1980. K. D. Knorr, R. Krohn, and R. Whitley (eds.), The Social Process of Scientific Investigation.
V - 1981. E. Mendelsohn and Y. Elkana (eds.), Sciences and Cultures. Anthropological and Historical Studies of the Sciences.