0191-6599&b $3.00+0.

00
Pergamon Journ.11~ Lid

STRENGTHS AND WEAKNESSES OF THE HISTORY OF

MENTALITIES*

The history of mentalities is not easy to define to everyone’s satisfaction. Even

in France, where the tradition is most self-conscious and most continuous, the
approach has undergone modifications in the fifty years which separate Marc
Bloch from (say) Jacques Le Goff. Elsewhere, at least two leading practitioners
persist in denying that the history of mentalities is what they are doing; Richard
Cobb and Carlo Ginzburg. For the purposes of this paper the approach will be
defined in terms of three distinctive features, In the first place, a stress on
collective attitudes rather than individual ones. Secondly, an emphasis on
unspoken or unconscious assumptions, on perception, on the workings of
‘practical reason’, or ‘everyday thought’ as well as on conscious thoughts or
elaborated theories. And finally, a concern with the structure of beliefs as well as
their content, with categories, with metaphors and symbols, with how people
think as well as what they think. In other words, to assert the existence of a
difference in mentalities between two groups is to make a much stronger
statement than merely asserting a difference in attitudes.’
In all these respects the history of mentalities differs from other approaches to
intellectual history, such as the American ‘History of Ideas’, the traditional
German Geistesgeschichte, and even the newer German Begriffsgeschichte, while
resembling certain styles of social anthropology.2 It might indeed be termed a
historical anthropology of ideas. But the problems it raises are too important for
even two disciplines to deal with, and they have been discussed in many more,
including philosophy, psychology, economics, literature, the history of art and
the history of science (although some individuals and groups involved in these
discussions have been unaware of all these interdisciplinary ramifications).
An interest in these problems goes back at least as far as the eighteenth century,
notably to Montesquieu and Vico(and in Denmark to Jens Kraft and his concern
with the Taenke-Maade ofDe ViIdeFolk). As a systematic historical approach it is
rather more recent. It has of course been associated with the French journal
Annales since the days of its founders Lucien Febvre and Marc Bloch, who were
much concerned with what they called mentalit& collectives, or outillage mental,
or, in the phrase they borrowed from Durkheim, repr&entations collectives.’ It
should be pointed out, however, that the interest in mentalities was not a
monopoly of Annales. Marcel Granet the sinologist and Georges Lefebvre the
historian of the French Revolution were already writing this kind of history in the

*This paper has been given at seminars in Cambridge, Copenhagen, London, Oxfordand
elsewhere, and the author would like to thank all his audiences for their comments and
criticisms.
tErnmanuel College, Cambridge.

439
440 Peter Burke

1930s4 It was not even a French monopoly, for it shades into the cultural history
practised by Johan Huizinga, whose Waning of the MiddleAges, first published in
19 19, was concerned with collective attitudes, with the history of feelings, and,
most important, with what he called ‘forms of thought’(gedachtensvormen), such
as the structuring of medieval thought by means of personification and
symbolism. In this last respect Huizinga had moved away from Jacob
Burckhardt, whose model of cultural history he generally followed.5
This essay attempts an assessment of both the strengths and the weaknesses of
the history of mentalities, as practised in France from the days of Bloch and
Febvre to those of Duby, Le Goff, Vovelle and Le Roy Ladurie. A similar
approach is now practised elsewhere, from the British Isles (Peter Brown, Keith
Thomas) to Eastern Europe (Peter Hanak, Aaron Gurevich), but to concentrate
on France should focus the discussion and make it more precise.‘j
The first point to make about the history of mentalities is that we very much
need something to occupy the conceptual space between the history of ideas,
defined more narrowly, and social history, in order to avoid having to choose
between an intellectual history with the society left out and a social history with
the thought left out. Consider the following recurrent problems. Why is it that
individuals from different cultures often find communication difficult? Why does
one find absurd precisely what the other takes for granted? How is it possible to
be able to translate every word in a text from an alien (or even a half-alien)
culture, yet to have difficulty in understanding the text? Because-one is able to
say if one adopts this approach to the past-there is a difference in mentality, in
other words different assumptions, different perceptions, and a different ‘logic’
(at least in the philosophically loose sense of different criteria for justifying
assertions; reason, authority, experience and so on).’
A classic example of the kind of problem which we need the concept of
mentality-or something roughly like it-in order to solve (or to ‘dissolve’ as
Wittgenstein used to say), an example which has exercised educated Europeans
from Montesquieu (if not Wormius) to Peter Brown, is that of the medieval
ordeal, which is easy enough to condemn in the name of reason, but hard to
explain historically without recourse to what Montesquieu called ‘la man&e de
penser de nos pkres’ (in other words, mentality).*
In the last few years, in fields as far apart as economic history and the history of
science, a number of scholars have found it impossible to solve their problems
without invoking a concept like that of mentality, as opposed to a timeless
rationality (which usually turns out to have been defined ethnocentrically).
Witold Kula, in a brilliant account of the workings of the ‘feudal system’ in
seventeenth- and eighteenth-century Poland, has demonstrated that it cannot be
understood without taking into account the attitudes, values, or mode of thought
of the Polish magnates who gained most from itY Edward Thompson’s famous
article on the ‘moral economy’ of the English crowd suggested that food riots
should be seen not as a simple response to hunger but as an expression of
collective moral assumptions, a suggestion which was soon taken up and utilised
in the study of rebellions in societies as distant as South East Asia.” There are
obvious similarities between this approach to economic history and that of
‘economic anthropology’, which also stresses alternative rationalities. In the
history of science too there is a tendency to turn ‘rationality’ from an assumption
The History of Mentalities 441

into a problem, and to supplement explanations of innovation in terms of

internal necessity or great men by referring to collective ‘styles of thought’.”
However, if I am to defend, as well as describe, the history of mentalities, I
should prefer to take a stand on more familiar ground, and cite four concrete
examples, from the early modern period, where it seems particularly difficult to
do without something like the idea of mentality or mode of thought if we are to
give a plausible account of certain beliefs.
1. The idea of ‘correspondences’ between the seven planets, the seven metals,
the seven days of the week, and so on. Correspondences are neither identities nor
similarities but seem to be somewhere in between, a middle category which seems
to have parallels in other cultures, notably the famous Bororo, who claimed to be
macaws, and the Nuer, for whom ‘a twin is a bird’. The Bororo example was
taken up by the philosopher Lucien Levy-Bruhl, who described the relationship
between men and macaws in terms of ‘participation mystique’, a characteristic,
so he suggested, of ‘pre-logical’ thought or ‘primitive mentality’.‘*
2. The idea of the ‘great chain’, ladder or ‘scale’ of being, according to which it
is better to live than to exist, better to feel than to live, better to think than to feel;
in other words, better to be a tree than a stone, a horse than a tree, a man than a
horse.13 As Michael Walzer has commented, ‘The complex system of analogies
up and down the chain cannot be taken entirely literally, nor can it be understood
as a mere convention, a useful and pleasing metaphor.‘i4 The fact that this
projection of the social hierarchy on to the cosmos has now come to seem quaint
indicates a change in western mentality since the sixteenth century.
3. Hence, to take a third example, the modern reader is likely to be somewhat
puzzled by the following statement of cardinal Btrulle’s: ‘l’etat de l’enfance, &at
le plus vi1 et le plus abject de la nature humaine, apres celui de la mart’.” The
oddness is not-or not only-a function of the content of Berulle’s beliefs, of the
fact that childishness has become more acceptable to adults than it seems to have
been in the seventeenth century. It is the hierarchical arrangement, the structure
of his beliefs, that we are likely to find most alien. It has also come to seem rather
odd to describe death as an Aat. Yet personifications of abstractions were
common enough in the medieval and early modern periods. Johan Huizinga,
whose interest in personification has been mentioned already, once discussed a
statement by St Francis to the effect that he was married to Lady Poverty.
Huizinga suggested-like Walzer on the Great Chain of Being-that this
statement should not be taken either literally or metaphorically. Its logical status
was somewhere in between.
4. His suggestion may be helpful in interpreting my last example, which comes
from the rebellion of the ‘bonnets rouges’ in Brittany in 1675. The rebels expressed
their demands in a document known as the ‘codepuysan’. One of its clauses runs
as follows: ‘II est dgfendu, . . . de donner retraite Lila gabelle et Lises enfants, et de
leur fournir ni ri manger ni aucune commoditk; mais au contraire, il est enjoint de
tirer sur elle comme sur un chien enragP.‘16 This ‘gabelle’ was of course the salt tax.
It should be added that some historians believe that the code is a forgery,
precisely-so I suspect-because they find this clause a stumbling-block to
credibility. It is a scandal for historians who do not believe in differences in
mentality. What could the rebels have meant? Whether or not the Breton
peasants had a more archaic mentality than that of peasants elsewhere in France
442 Pt~to. Burkt>

is a question I do not feel competent to answer. They employed personifications

of death and plague as well as the gabelk; but similar personifications can be
found elsewhere in Europe. i’ Even the notorious clause about the gabelle differs
only in degree of elaboration from the cry of some fifteenth-century Tuscans,
‘muoia il catasto’; and this is a special case of a more common Italian slogan,
‘muoia ilmalgoverno’. Is the injunction to be taken literally or metaphorically’? Or
is it another statement which will not fit our own cultural categories of ‘literal’
and ‘metaphorical’, and so indicates an alien mentality?18
Faced with examples like these, historians need a concept like that of
‘mentality’ in order to avoid two opposite dangers. The first danger is that of
dismissing the Breton peasants or cardinal Berulle as irrational or as unworthy of
serious historical consideration. If a specific seventeenth-century attitude strikes
us as odd, we have to remember that it was part of a belief system in which the
different parts supported one another, making it virtually unfalsifiable. The
second danger is that of sweeping the examples under the carpet, of assuming
that seventeenth-century Frenchmen must ‘really’ have thought in the same way
as we do. This second danger is that of succumbing to what might be called
‘premature empathy’, an intellectual illness which was diagnosed some sixty
years ago by Levy-Bruhl (thoughvico and Rousseau had been aware of it).” The
diagnosis has passed into British social anthropology under the name of the ‘If I
were a horse’ problem. The point is that to understand the behaviour of people in
other cultures it is not sufficient to imagine oneself in their shoes, in their
situation; it is also necessary to imagine their definition of the situation, to see it
through their eyes. A similar point was made by Lucien Febvre in an essay which
discussed the problem of ‘psychological and anachronism’, as he called it.‘”
The great strength of the mentalities approach is to make it possible to steer a
course which avoids the two opposite hazards. If we replace it with some
alternative, it seems reasonable to require that the substitute deal in a satisfactory
way with examples of the kind which have just been discussed. There is, of course,
a case for replacing an approach which suffers from weaknesses and has had
serious criticisms levelled against it. One objection, which is not infrequently
heard in Britain, ought not to be taken too seriously; it is that the French treat
mentalities as impersonal forces. In Britain it is obvious that there are no such
forces, but only men thinking, as Herbert Butterfield put it. Or as Vivian
Galbraith used to say, with deliberate sexism, in his Oxford lectures, ‘History is
just chaps’. To the French, however (if I may attempt an act of empathy with
them), it is equally obvious that the term mentalitk is not being used to describe a
thing or a force, but rather to characterise the relation between beliefs, which is
what makes them into a system, The beliefs are ‘collective’ only in the sense of
being shared by individuals, not in the sense of standing outside them. The
contrast between the British intellectual tradition of methodological individual-
ism and the French tradition of holism is so strong, and goes back such a long
way, that one is tempted to call the difference itself one between two different
mentalities.”
I shall be equally laconic, though not I hope, cavalier, in response to the rather
similar objection that historians of mentalities underestimate what human beings
have in common. There is a sense (or level), at which human nature is always the
same and another level at which it is not, and it is very hard to talk about
The History of Mentalities 443

differences in mentalities (or cultures) without exaggerating their importance. To

discuss differences in views of the world need not imply that different groups see
the world in completely different ways.
There are, however, at least four more serious objections to the mentalities
approach to intellectual history. They are usually presented in exaggerated form,
but each contains a kernel of serious criticism.
1. In the first place, to look for broad differences in mentalities encourages
historians to treat attitudes they find alien as if they were homogeneous, to
overestimate the degree of intellectual consensus in a given society in the past.22
This is of course one of the classic objections to the concept of Zeitgeist, and to
traditional Geistesgeschichte, and the objection is only partially answered by
saying that Hegel (for example) did not treat seventeenth-century thought as
homogeneous, but contrasted Bacon (say) with his contemporary Boehme.
Bloch and Febvre (like Huizinga) are not so far from traditional Geistesge-
schichte as their use of the new term ‘mentality’ might make one think. To write,
as they did on occasion, of ‘the medieval Frenchman’ or ‘the sixteenth-century
Frenchman’ as if there were no important variations in attitude among the
inhabitants of France in this period is seriously misleading.
To be fair, it should be pointed out that there is no reason why a mentality
should not be imputed to a social class or other group rather than to a whole
society, but doing this leads to similar problems on a lesser scale. From outside it
may seem reasonable enough to talk of say the ‘legal mentality’ in seventeenth-
century England, but it cannot be assumed that all lawyers had the same
attitudes. We have to allow for individual variation. Hence Carlo Ginzburg’s
study of Menocchio was designed to undermine the history of mentalities
(though he came to employ his hero as a spokesman, however eccentric, for
traditional peasant culture, so that mentalities, thrown out of the door, came
back in again through the window). 23 Homogenisation of beliefs is not a
necessary part of the approach. It is possible-indeed, I think it is necessary-to
do as Jacques Le Goff suggests and to use the term ‘mentality’ to describe only
the beliefs which a given individual shares with a number of contemporaries, and
limit the approach to the investigation of common assumptions rather than
extending it to the whole of intellectual history.24 In practice, however,
homogenisation remains a danger.
2. Linked to the problem of variation is the problem of change, or variation
over time. It is, in the words of a recent critic (Roger Chartier), ‘the problem on
which all history of mentalities stumbles, that of the reasons for and modalities of
the passage from one system to another’.*’ The sharper the contrast between one
mentality and another (‘traditional’ and ‘modern’, for example), the harder it is
to explain how change ever took place. The crucial idea here would seem to be that
of a ‘system’ of beliefs, a circle of thought in which each part at once supports and
is supported by the rest, making the whole system impervious to falsification (a
closed system in Popper’s sense of the term), as in the famous case of Bloch’s
royal touch or the parallel account given by Evans-Pritchard of the Zande belief
in poison oracles. ‘In this web of belief,’ he wrote, ‘every strand depends upon
every other strand, and a Zande cannot get out of its meshes because it is the only
world he knows.‘26 This notion of ‘system’ lies at the heart of the history of
mentalities and it is at once a help and a hindrance. Without it we cannot explain
444 Peter Burke

how ideas persist over the generations in the face of awkward empirical evidence;
but if we use it we make it difficult for ourselves to understand change.
To put the point another way, it is all too easy to reify mentalities, to perceive
them as ‘prisons’ from which individuals cannot escape.27 Again I would insist
that this danger, acute in practice, is not one to which historians of mentalities
must inevitably succumb. In the case of Italy, with which I am particularly
concerned, it is fascinating to see the gradual rise (between the thirteenth and
seventeenth centuries) of what might be called the ‘numerate mentality’, revealed
in practices like the spread and elaboration of censuses as well as in private
account-books.28 For a change in the mentality of a group of intellectuals over
the relatively short term, one might point to the so-called ‘Scientific Revolution’
of the middle of the seventeenth century (to the ‘mechanisation of the world-
picture’, the shift from justifying statements by authorities to justifying them by
experience, and so on). 2yThe anthropologist RobinHorton, elaborating ideas of
Evans-Pritchard and Popper, has sketched out a general picture of change in
modes of thought, emphasising the importance of awareness of alternatives to a
given intellectual system, or, as he now puts it, the relative importance of
competition between theories in different societies.30 In short, there are answers
to this fundamental criticism. However, as a criticism of the practice of historians
of mentalities in the past, this second objection retains considerable force.
3. Another serious objection to the history of mentalities is that it treats belief
systems as autonomous, that it is concerned with the relationship of beliefs with
one another, to the exclusion of the relationship between beliefs and society. The
objection must not be exaggerated; it does not seem to me that Bloch’s Rois
thaumaturges or Febvre’s Problkme de Pincroyance (which emphasises the
impossibility of thinking all kinds of thought at all times) do treat belief systems
as completely autonomous. Compared with historians of ideologies, however,
there is a great difference in emphasis. Historians of ideologies see thought as
shaped (if not determined) by social forces, and they emphasise the cunning
(conscious or unconscious) by which a particular view of the world is presented
as natural, indeed the only possible one. Historians of mentalities by contrast see
belief systems as innocent and fairly autonomous. The two approaches overlap
but they do not coincide.31
4. A still more serious objection- indeed, the fundamental objection-to the
mentalities approach is that it is built on evolutionism and more specifically on
the ideas of Lucien Levy-Bruhl, and that these foundations have been
undermined. The importance of Levy-Bruhl in the intellectual history of the early
twentieth century still awaits adequate assessment. His critique ofsimple notions
of empathy was a valuable one, and his own notion of ‘prelogical’ thought was
more subtle than some of his critics have realised.32 He was influential on
thinkers as diverse as Ernst Cassirer, Jane Harrison and Jphan Huizinga. It was
Levy-Bruhl who launched the term mentalitt in the 1920s (it had been employed
by Durkheim earlier). It was he, together with Durkheim, who was quoted by
both Bloch and Febvre at crucial points in the arguments of their books on the
royal touch and the religion of Rabelais (just as he was quoted in Huizinga’s
study of the late Middle Ages. However, at the end of his life, Levy-Bruhl
abandoned his famous distinction between the ‘prelogical’ and ‘logical’
mentalities, and in any case, anthropologists as different as Evans-Pritchard and
The History of Mentalities 445

Levi-Strauss have agreed in rejecting it, together with the evolutionist schema of
which it formed a part.33
Yet a rather simple contrast along roughly Levy-Bruhlian lines, between two
mentalities, can be found not only in the work of Febvre (and to a lesser extent in
that of Bloch, always the most cautious of the two), but also in some of their
successors, such as Robert Mandrou, Jean Delumeau and Robert Muchembled.
The last two have recently been criticised by Stuart Clark for attempting to
explain peasant religion in early modern Europe almost entirely in negative
terms, in terms of failure, anxiety and the constraints of a harsh environment. As
Clark says (quoting Wittgenstein on Frazer), it seems implausible to treat an
entire world-view as a mistake.34To describe an alien belief system as a failure is a
gross form of ethnocentrism. Where do the criteria of success and failure come
from if not from the speaker’s own culture? Ethnocentrism is encouraged by the
description of mentalities as ‘traditional’ and ‘modern’ (no less than ‘primitive’
and ‘civilised’, ‘prelogical’ and ‘logical’), since these dichotomies tend to slip into
a contrast between Us and Them. The dichotomies are simplistic. They elide the
differences between (say) Chinese mandarins, Renaissance humanists, and
Breton peasants, all of whose beliefs, since they are not like ours, have to be
described as ‘traditional’.
All four criticisms of the approach come in different forms, more or less cogent,
undermining the work of some historians of mentalities more effectively than
others. However, all of them suggest the need to reformulate the approach to
meet the objections. I have three proposals for reformulation; a greater concern
with interests; with categories; and with metaphors.
In the days of Bloch and Febvre, Marxist historians did not have a credible
alternative to the history of mentalities, because they dismissed ideas as mere
‘superstructure’ (or ‘ideology’ in a reductionist sense), and devoted their
attention to economic history. More recently, however, Marxists have
questioned the notion of ‘superstructure’, refined the notion of ideology, and
introduced new concepts like cultural ‘hegemony’. Looking back on Marc
Bloch’s great book on the royal touch after more than fifty years, it now seems
curious that he did not ask the question, in whose interest it was that this belief in
the supernatural powers of kings should continue to exist. The question seems to
me to be a particularly important one in the discussion of the early modern
period, when the monarchy was under attack in the French and English civil
wars, and more especially in the late seventeenth century, when Charles II and
Louis XIV are unlikely to have believed in the virtues of their touch, but went on
with the custom nonetheless. More generally, historians of mentalities cannot
afford to ignore the problems raised by students of ideologies, even if they do not
deal with them in the same confident way as Habermas (say) or Althusser, or
even Bourdieu.3s It is interesting to see some scholars who describe themselves as
historians of mentalities, such as Georges Duby and Michel Vovelle, beginning
to concern themselves with these questions of interests, legitimation, and so on.36
It will not be easy to combine what might be described as the ‘innocent’ and the
‘cynical’ views of thought, but a synthesis might be possible along the lines of the
study of the unconscious harmonising of ideas with interests; conflicts of interest
make the unconscious conscious and the implicit explicit, and so lead tochange.
The second proposal is a greater concern with mental or perceptual categories,
446 PL’IW Hlirkc

schemata, stereotypes or paradigms (the variety of terminology reveals the extent

to which similar problems and insights have been occurring in different fields and
disciplines). Huizinga and Febvre were stimulated by the psychologists of their
day (notably Wundt and Wallon); more recently, some anthropologists and
historians have turned to Piaget (despite the danger of analogies between the
psychology of children and that of ‘primitive’ or medieval man).” In the area of
categories and schemata, important historical work has been done by Ernst
Gombrich, Thomas Kuhn, Michel Foucault, Norman Cohn; the oddness of the
juxtapositions makes one pause for a moment, but whatever their other
differences, all these scholars are interested in categories and schemata; with
what is associated with what, or contrasted with what, in whose thought.38
Stereotypes are amazingly long-lived on occasion, as Cohn’s examples in
particular demonstrate; and they are subject to change, since they are employed
by individuals and groups to structure or interpret their experience of a changing
world. If the great stumbling-block for the history of mentalities is, as suggested
earlier, ‘the reasons for and the modalities of the passage from one system to
another’, then there is an obvious case for taking upThomas Kuhn’s notion of an
intellectual tradition or ‘paradigm’ which may absorb or resist change for long
periods thanks to relatively minor ‘adjustments’, but will finally crack and allow
a ‘Gestalt switch’ or intellectual ‘revolution’. Kuhn’s model of the process of
intellectual change is an attractive one precisely because it is dynamic, concerned
with a sequence. Of course the notion of paradigm is problematic even in the
history of science, where it began, and is likely to become still more problematic if
we try to utilise it in discussions of other kinds or change in ideas.39 As in the case
of ‘mentality’, a ‘paradigm’ must not be seen as a ‘prison’; it is something
individuals use to make sense of their experience. Yet at the same time it is not
entirely at their service, but shapes their thought. Despite these problems, the
notion of paradigm can be and has been utilised with sensitivity, skill and
caution-in studies of the Renaissance, for example, and of early modern
political thought.40
Similar kinds of difficulty arise when one tries to work with the parallel notion
of ‘episteme’ developed by Michel Foucault, in his case too out of a history-of-
science tradition (that of Bachelard and Canguilhem). In Foucault’s case the way
in which change occurs is even less apparent than in that of (say) Gombrich.4’
The greater the emphasis on system, the more difficult the explanation of change
(or ‘rupture’) becomes. On the other hand, Foucault himself has contributed to
the refinement of the notion of paradigm or episteme by his discussions of the
mental ‘grids’ (grilles) as he calls them, employed by different groups, and his
concern with what they excluded as well as what they contained.42 I take it that
these grids are a kind of microparadigm; or to use the language of Warburg and
his followers, they are ‘schemata’ which assimilate different situations to one
another in order to make them easier to understand. How and why do they
change? One way in which change occurs seems to be by the displacement of the
schema or stereotype from one object to another; Norman Cohn, for example,
has charted the history of the early modern stereotype of the witch (the sex orgies,
the eating of babies etc.), pointing out that in the Middle Ages this stereotype was
applied to heretics and in Roman times to the Christians, while in the eighteenth
century some people rejected it altogether; at this micro-level, the schema is not
The History of Menrnlities 447

invulnerable to counter-evidence. If we look at a so-called ‘belief system’ as a

bundle of schemata, which generally support one another but may sometimes be
in contradiction, change becomes a little easier to explain.
The notions of ‘schema’ and ‘system’ may themselves be clarified if we look
more closely at language, and especially at metaphor and symbol. Historians like
anthropologists and sociologists are coming to recognise the importance of
language in ‘constituting’ the reality they study. 43The relation between modes of
linguistic comunication, oral and literate, and their effect on modes of thought
has also attracted much attention from historians as from anthropologists in
recent years.44 Indeed, there is a danger that after years of neglecting language, it
will now be credited with too much power; the idea of the ‘prison-house of
language’-like Braudel’s notion of mentalities as ‘prisons’-is too simple.4s The
limits to the ‘constitution’ of reality by individuals and groups also need to be
borne in mind; the world is not completely malleable. However, if we are trying
to describe the differences between mentalities, it seems a good idea to look at
recurrent metaphors, especially those which seem to structure thought. Obvious
examples in the history of the west are the metaphor of the world as an organism
(a ‘body’, and ‘animal’), and the metaphor of the world as a machine, and the
shift from one to the other in the course of the so-called ‘scientific revolution’ of
the seventeenth century. The idea of the ‘mechanisation of the world picture’ in
the seventeenth century is a fruitful one which deserves a fuller, richer, wider
treatment than it received in the famous book of that title (cited above). So is that
of the decline of ‘natural symbols’ like the ‘body politic’.46 The stress on
dominant metaphors is especially valuable because it is a means of liberating
historians of mentalities from the perils (if not the ‘prison’), of the binary system,
the great divide into traditional and modern. I am not suggesting that a given
period be characterised in terms of one metaphor alone; seventeenth-century
thinkers, for example, were fascinated not only by the image of the machine but
by the metaphor of law (the laws of nature; the court of heaven, with its judges
and advocates; and so on). It is also necessary to ask whether there has been a
decline of metaphor, a gradual shift from a more concrete to a more abstract
mode of thought (associated with literacy and numeracy). I consider that this is
very likely; but to believe that moderns think without the help of metaphors
would be to succumb to a dangerous illusion. The shift might be better described
as a more acute consciousness of metaphor, or a change in the conception of
metaphor, from fundamental ‘correspondence’ to mere ‘analogy’.47 Metaphor is
coming to be taken seriously not only by literary critics but also by social
anthropologists, who study its place in social life, and also by some philosophers,
notably Hans Blumenberg.48 Historians of mentalities would have something to
gain by following their example.4g
It will be noted that of my three points, the last, on metaphors, is concerned
with an ‘internalist’ history of mentalities, in other words with the relation of
beliefs to one another. The first, on interests, is concerned with an ‘externalist’
history, with the relation of beliefs to society. My central point, about categories,
is concerned with both. Categories and schemata are ways of structuring
thought. But they are not neutral; they may be associated with interests, with
attempts by some groups to control others, as is suggested by the sociologists
who have developed what they call ‘labelling theory’. Give a witch a bad name
448 Peter Burke

and burn her. It is essential to any reformulated history of mentalities that it

combine the insights of the internalist and externalist approaches. We will make
no progress in this field if we treat mentalities either as completely autonomous
or as nothing but the reflection of a more fundamental reality.
The essential argument of this paper is that the history of mentalities can be
renewed (and is being renewed) by a selective appropriation and incorporation of
concepts taken from other traditions; a kind of bricofuge but not an uncritical
eclecticism. Whether this will be a mere extension of the historian’s outilluge
mental, or whether it will be a true paradigm shift, we can only find out by trying.

NOTES

1. Discussions of the history of mentalities are now extremely numerous; special issues of
the TijdJchrtft moor Gesrhiedenis (1973), and Kulrur- & Kla.,.se (1983) have been devoted
to the subject. Among the recent critical surveys, with bibliographies, are M.A.
Gismondi, ‘The gift of theory’, Social History 10 (1985), 21 l-30, and V. Sellin,
‘Mentalitat und Mentalitatsgeschichte’, Historische Zeitschrtft 241 (1985), 555-9X.
2. R. Koselleck, Vergangene Zukunft (1979; English trans., Futures Past, Cambridge,
Mass., 1985) explains and illustrates Regrtffigeschichte; for anthropological
approaches, see C.O. Frake, Language and CulturaJDescription (Stanford, 1980), and
K. Hastrup, Culture and History in Medieval Iceland(Oxford, 1985). In the 192Os, the
so-caBed ‘sociology of knowledge’, more especially the work of Karl Mannheim, who
was interested in ‘styles of thought’, was not far from the history of mentalities
practised by Marc Bloch.
3. A. Burguiere, ‘The fate of the history of mentalities in the Annales’, Comparative
Studies in Society and History 24 (1982), 424-37, suggests that Bloch and Febvre did
not give the term mental&! the same meaning. I am not sure he is right, but there is no
doubt that Bloch was the more interested in collective attitudes, Febvre in individuals.
4. M. Granet, La pensee chinoise (Paris, 1934); G. Lefebvre, La grande peur de 1789
(Paris, 1932).
5. Huizinga’s studies with the psychologist Wundt at the University of Leipzig seem to
have left their traces on his history.
6. P. Hanak, ‘Mentalitt et symbolique des mouvements socialistes agraires’, Archives
des Sciences Sociales des Religions 45 (1978), 65-73; A.J. Gurevich, ‘Medieval culture
and mentality according to the new French historiography’, European Journal of
Sociology 24 (1983), 167-95.
7. On the possibility of using the term in a more precise sense, see D.E. Cooper,
‘Alternative logic in primitive thought’, Man 10 (1975), and M.H. Salmond, ‘Do
Azande and Nuer use a non-standard logic?‘, Man 13 (1978).
8. L’Esprit des Lois (Paris, 1744), Book 28, chap. 17; P. Brown, ‘Society and the
supernatural’, Daedalus (Spring 1975), 133-47; cf. C. Morris, ‘Judicium De?, in
Church, Society and Politics, ed. D. Baker (Oxford, 1975), pp. 95-111, and CM.
Radding, ‘Superstition to science’, American Historical Review 84 (1979), 945-69.
9. W. Kula, Economic Theory of theFeudalSystem(1962; English trans., London, 1976).
10. E.P. Thompson, ‘The moral economy of the English crowd’, Past and Present 50
(1971), 76-136; J.C. Scott, The Moral Economy of the Peasant; RebeNion and
Subsistence in South East Asia (New Haven, 1976).
The History of Mentalities 449

11. Pioneers in this field include L. Fleck, Genesis and Development of a Scientific Fact
(1935; English trans., Chicago-London, 1980), with his stress on Denkgemeinschaft
and Denkstruktur, in the tradition of the sociology of knowledge of Karl Mannheim;
and A. Rey, La science orientale avant les grecs (Paris, 1930). Rey collaborated with
Lucien Febvre and thus contributed to the development of the French ‘mentalities’
tradition. Although they have gone their different ways, Gaston Bachelard, Georges
Canguilhem, Michel Foucault and Thomas Kuhn all have something in common with
this approach. Rather closer is the anthropologist Robin Horton, in his ‘African
traditional thought and Western science’, Africa 37 (1967), 50-71, 155-87, a study
which underlies Occult and Scienttfic Mentalities in the Renaissance, ed. B. Vickers
(Cambridge, 1984). A.C. Crombie is at work on a study of Styles of Scientific Thinking
in the European Tradition.
12. E.M.W. Tillyard, The Elizabethan WorldPicture (Cambridge, 1943) is a well-known
description of Renaissance correspondences; compare and contrast M. Foucault, Les
mats et Zes chases (Paris, 1966), chap. 1. The ‘twin is a bird’ example is analysed by
E.E. Evans-Pritchard, Nuer Religion (Oxford, 1956), pp. 128f. On the Bororo, J.C.
Cracker, ‘My brother the parrot’, in The Social Use of Metaphor, ed. J.D. Sapir and
J.C. Cracker (Philadelphia, 1977), chap. 6.
13. A.O. Lovejoy, The Great Chain of Being (Cambridge, Mass., 1936) is the classic
account.
14. M. Walzer, The Revolution of the Saints (Cambridge, Mass., 1965), p. 156.
15. Quoted in G. Snyders, La pedagogic en France aux 17e et 18e sit?cles(Paris, 1964), p.
194.
16. Les revoltes bretonnes de 1675, ed. Y. Garland and C. Nires (Paris, 1975), p. 102.
17. On Breton personifications of death and plague in this period, A. Croix, La Rretagne
aux 16e et 17e siecles, 2 ~01s. (Paris, 1981), pp. 1067f.; on the gabelle, cf. A. Le Braz, La
Iegende de la mort, 4th edn. (Paris, 1922), pp. 136f.
18. Personifications are still employed today of course, but my point is that they are now
self-conscious metaphors. Why so many personifications are female is a fascinating
question which cannot be discussed here; see M. Warner, Monuments and Maidens
(London, 1985).
19. L. Levy-Bruhl, La mentaliteprimitive (Paris, 1927), p. 15; he offered his analysis ‘au
lieu de nous substituer en imagination aux primittfs que nous Ptudions, et de les faire
penser comme nous penserions si nous Ptions a leur place’.
20. L. Febvre, ‘History and psychology’ (1938), trans. in A New Kind of History, ed. P.
Burke (London, 1973), chap. 1.
21. On holism, Susan James, The Content of Social Explanation (Cambridge, 1984).
22. For a vigorous recent formulation of this objection, see Gismondi, ‘Gift of theory’.
23. C. Ginzburg, Cheese and Worms (1976; English trans., London, 1981).
24. J. Le Goff, ‘Les mentalites: une histoire ambigue’, in Faire de (‘histoire, ed. J. Le Goff
and P. Nora, 3 ~01s. (Paris, 1974), Vol. 3, pp. 76-90.
25. R. Chartier, ‘Intellectual history or sociocultural history? The French trajectories’, in
Modern European Intellectual History, ed. D. LaCapra and S. Kaplan (Ithaca and
London, 1982) pp. 13-46.
26. E. Evans-Pritchard, Witchcraft, Oracles andMagic among theAzande(Oxford, 1937)
p. 194. It is worth adding that before he became a social anthropologist, Evans-
Pritchard was a student of mediaeval history.
27. M. Vovelle, Ideologies et menrahtCs(Paris, 1982), pp. 214f., criticising Braudel’s image
of the ‘prison’, to which he prefers Ernest Labrousse’s idea of ‘resistances’. Cf.
Gismondi, ‘Gift of theory’, pp. 21 If.
28. On the early stages of this process, A. Murray, Reason and Society in the Middle Ages
(Oxford, 1978), pp. 180f.
29. E.J. Dijksterhuis, The Mechanization af the World Pil,ture (1950; English trans.,
Oxford, 1961); for a recent discussion ofthe rejection ofauthority, suggesting that this
rejection was sometimes more a matter of style than substance, see P. Dear, ‘Totius in
verba: rhetoric and authority in the Early Royal Society’, Isis 76 (1985), 145-61.
30. R. Horton, ‘African traditional thought’; R. Horton, ‘Tradition and modernity
revisited’, in Rationality and Relativism, ed. M. Hollis and S. Lukes (Oxford, 1982)
pp. 201-60, which draws on E. Gellner’s critique, ‘The savage mind’, in his
Legitimation of Belief(Cambridge, 1974), chap. 8.
31. M. Vovelle, ‘Ideologies and mentalities’, in Culture. Ideology and Politics, ed. R.
Samuel and G. Stedman Jones (London, 1983), chap. I. Cf. Karl Mannheim’s
distinction between the ‘particular’ and the ‘total’ concepts of ideology in Zrieo/og~
and Utopia (London, 1936) pp. 49-96.
32. A balanced assessment is offered by E. Evans-Pritchard, Theories ofprimitive Religion
(Oxford, 1965), pp. 78-99.
33. L. Levy-Bruhl, Carnets (Paris, 1949).
34. S. Clark, ‘French historians and Early Modern popular culture’, Pasr andpresent 100
( l983), 62-99.
35. J. Habermas, Knoll~letlge and Human Interests (English trans., Boston, 1971); L.
Althusser, ‘Ideology and ideological state apparatuses’( 1970), English trans. in Lenin
and Philosophy (London, 1971), pp. 121-73; P. Bourdieu, ‘Ul’te interpretation de la
theorie de la religion selon Max Weber’, Archives europeennes de sociologic 12 (197 I),
3-21.
36. G. Duby, The Three Orders (1979: English trans., Chicago, 1980).
37. C.R. Hallpike, The Foundarions of Primitive Thought (Oxford, 1979); C.M. Radding,
‘The evolution of mediaeval mentalities’, American Historical Revier+x 83 (1978),
577-97. Recent work by psychologists of relevance for the history of mentalities
includes M. Cole and S. Scribner, Culture and Thought (1974).
38. E.H. Gombrich, Art and Illusion (London, 1960); T. Kuhn, The Structure ofScien/iJic
Revolutions (Chicago, 1962); M. Foucault, Les mots er /es chases (Paris, 1966); N.
Cohn, Europe’s Inner Demons (London, 1975). Gombrich acknowledges a debt to
Gestalt psychology; Kuhn also discusses change in terms of a ‘Gestalt switch’.
Foucault’s concern with psychiatry is well known, while Cohn is interested in
psychoanalysis.
39. Criticism and the gro\+,th of knowledge, ed. I. Lakatos and A. Musgrave (Cambridge,
1970); D.A. Hollinger, ‘T.S. Kuhn’s theory ofscience and its implications for history’,
in Paradigms and Rev~olurions, ed. G. Gutting (Notre Dame, 1980) pp. 195-222 (my
thanks to Simon Schaffer for bringing this piece to my attention).
40. J.E. Seigel, Rhetoric and Philosophy. in Renaissance Humanism (Princeton, 196X);
J.G.A. Pocock, Politics, Language and Time (London, 1972) esp. chap. 8.
41. Gombrich’s Art andIllusion was criticised along these lines by Rudolf Arnheim in Ar/
BuNetin 44 (1962) 75-9.
42. The term conceptual ‘grid’ can be found in K.L. Pike, Language in Relation to a
Unified Theory ofrhe Structure ofHuman Behaviour( 1954; rev. edn., The Hague and
Paris, 1967), and in M. Douglas, Natural Symbols (London, 1970). Compare the use
of the term ‘code’ by sociolinguists and ethnographers of communication such as
Basil Bernstein and Dell Hymes.
43. M. Crick, Explorations in Language and Meaning (London, 1976) is an unusually
explicit account of this approach. It should be added that linguistics-notably the
work of Antoine Meillet-was influential on Febvre when he was turning towards the
history of mentalities; but a newer linguistics has something different to offer.
44. A recent influential formulation is J. Goody, The Domestication of/he Savage Mind
(Cambridge, 1977). Historical explorations of the effects of literacy and print on
The History of Mentalities 4.51

mentalities include M.T. Clanchy, From Memory to Written Record(London, 1979),

and E. Eisenstein, The Printing Press us an Agent of Change (Cambridge, 1979).
45. The phrase is Nietzsche’s, put into wider circulation by F. Jameson, The Prison-House
of Language (Princeton, 1972). A similar view is implied by the work of Benjamain
Whorf, which is of great relevance to the history of mentalities; see B.L. Whorf,
Language, Thought and Reality (Cambridge, Mass., 1956).
46. E.H. Kantorowicz, The King’s Two Bodies (Princeton, 1957) is the classic study; there
is a useful general survey by P. Archambault, ‘The analogy of the body in Renaissance
political literature’, Bibliothkque d’Humanisme et Renaissance (1967), 21-53. More
searching is D. Starkey, ‘Representation through intimacy’, in Symbols and
Sentiments, ed. I. Lewis (London and New York, 1977) chap. 10.
47. This major shift deserves more attention than it has so far received. Aspects of the
problem are studied by J. Hollander, The Untuning qf the Sky (Princeton, 1961),
conclusion; V. Harris, ‘Allegory to analogy in the interpretation of scripture’,
Philological Quarterly 45 (1966); B. Vickers, ‘Analogy v identity’, in Occult and
ScientiJic Mentalities, pp. 95-163.
48. The Social Use of Metaphor, ed. J.D. Sapir and J.C. Cracker (Philadelphia, 1977); H.
Blumenberg, Paradigmen zu einer Metaphorologie (Bonn, 1960). Blumenberg’s
approach is reminiscent of E.R. Curtius, while Cracker and Sapir acknowledge the
inspiration of Kenneth Burke.
49. Cf. Geoffrey Lloyd’s Rivers lecture (1985), ‘Mentalities, Metaphors and the
Foundations of Science’. 1 should like to thank Professor Lloyd for letting me see his
script.