The-Translator - Venuti Et Al

This article was downloaded by: [University of Exeter]
On: 15 July 2015, At: 04:09

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: 5 Howick Place, London, SW1P 1WG
The Translator
Publication details, including instructions for authors
and subscription information:
http://www.tandfonline.com/loi/rtrn20
Translation, Authorship,
Copyright
a
Lawrence Venuti
a
Temple University, USA
Published online: 21 Feb 2014.
To cite this article: Lawrence Venuti (1995) Translation, Authorship, Copyright, The
Translator, 1:1, 1-24, DOI: 10.1080/13556509.1995.10798947
To link to this article: http://dx.doi.org/10.1080/13556509.1995.10798947
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the
information (the “Content”) contained in the publications on our platform.
However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or
suitability for any purpose of the Content. Any opinions and views expressed
in this publication are the opinions and views of the authors, and are not the
views of or endorsed by Taylor & Francis. The accuracy of the Content should
not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions,
claims, proceedings, demands, costs, expenses, damages, and other liabilities
whatsoever or howsoever caused arising directly or indirectly in connection
with, in relation to or arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan, sub-
licensing, systematic supply, or distribution in any form to anyone is expressly
forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Downloaded by [University of Exeter] at 04:09 15 July 2015
The Translator. Volume 1, Number 1 (1995), 1-24
Translation, Authorship, Copyright
LAWRENCE VENUTI
Temple University, USA
Abstract. Current copyright law reserves for the author the exclusive
right to copy and circulate a work, and this strictly limits the transla-
tor's control of the translated text, resulting not merely in an economic
disadvantage to the translator but in the continuing cultural margin-
ality of translation. The history of copyright contains alternative
definitions of translation that are useful in challenging its present

legal status. Such definitions demonstrate that the development of an
exclusive authorial copyright depended on an individualistic con-
cept of original authorship that negates the translator's work. But
they also enable the formulation of a collective concept of author-
ship, one in which the translator is seen as a species of author and
originality is redefined. This paper presents a genealogy of copy-
right that contests the cultural assumptions of the law and aims to
foster legislative reform designed to further both the interests of
translators and the practice of translation.
Copyright, the legal codes and conventions that govern the ownership of
intellectual works, describes a narrow space for translation. The history
of copyright since the eighteenth century reveals a movement toward
reserving for the author the right to copy and circulate his or her work,
including the right to license translations of it into foreign languages
(Kaplan 1967; Rose 1993). In current copyright law, with international
treaties that extend the rights of nationals to foreigners, authors world-
wide enjoy an exclusive right in any translation of their works for a term
of the author's life plus fifty years, unless the translation was made in the
service of an employer or on a work-for-hire basis, in which case the
employer enjoys an exclusive right in the translation. 1 Although the pro-
visions of actual publishing contracts can vary widely, in principle
copyright law places strict limitations on the translator's control of the
translated text.
From the viewpoint of translators and translation, these limitations
carry some troubling consequences, both economic and cultural. Insofar
as the law subordinates the translator's rights to the author's, it shrinks
the translator's share in the profits of the translation. A recent survey
ISSN 1355-6509 © St Jerome Publishing, Manchester

Translation, Authorship, Copyright 2
conducted by the PEN American Center indicates that most translations in

the United States are done on a work-for-hire basis, whereby the transla-
tor receives a flat fee with no percentage of the royalties or subsidiary
rights sales from, for example, a periodical publication, a licence for a
paperback edition, or an option by a film production company; in the
relatively few instances where contracts give translators a portion of this
income, the percentages range from 5 to 1 per cent of the royalties for a
hardback edition and from 50 to 10 per cent of subsidiary rights sales
(Keeley 1990). Translators in the United Kingdom face similar contrac-
tual terms (Glenny 1983), although the unequal distribution of profits is
also indicated by the allotment of loan payments under the Public Lending
Right, with the author receiving 70 and the translator 30 per cent.
Because copyright law decisively contributes to this unfavourable

economic situation, it diminishes the incentive for translators to invest in
translation projects. The many literary magazines published in English
today confirm that translators are in fact willing to make such an investment:
they regularly contribute translations of foreign poetry, fiction, and non-
fiction to these magazines without the promise of a book contract, usually
for little or no payment, mainly on the strength of a deep engagement with
the foreign text and culture. Yet the exclusive translation right given to
authors means that it is customarily they (or publishers as their assignees)
who initiate translations in an effort to sell licences and create foreign-
language markets for their works, and so they directly approach foreign
publishers, who then commission translators. The law prevents translators
from acquiring sufficient bargaining power to change this situation, unless
of course the translator is one of the very few who manage to gain public
recognition because publishers repeatedly commission them. But even in
these cases actual publishing practices reveal the subordination of
translators. William Weaver, the leading English-language translator of
Italian fiction since the 1950s, has published over sixty book-length
translations, all of which originated with a publisher's commission (tele-
phone interview: 24 September 1994).
Current copyright law, then, ensures that translation projects will be
driven by publishers, not translators. As a result, publishers shape cultural
developments at home and abroad. Seeking the maximum returns for their
investments, they are more likely to publish domestic works that are also
publishable in foreign countries, yet are not so culturally specific as to
resist translation, and their publishing decisions may target specific for-
eign markets for the sale of translation licences. Paul Goldstein sketches a
hypothetical case: "knowing that the French and German language mar-
kets belong exclusively to it, a publisher of English language works may
Lawrence Venuti 3
decide to invest in works that, once translated, will appeal to these audi-
ences as well" (Goldstein 1983:227). By the same token, publishers who
purchase translation rights are more likely to focus on foreign works that
are easily assimilable to domestic cultural values, targeting specific mar-
kets so as to avoid the potential loss involved in creating new ones. An
increasing trend since the 1980s, for example, has been to invest in the
translation of foreign works involved in 'tie-ins', film or dramatic adapta-
tions that promise wider reader recognition and greater sales.
Publishers thus determine not merely patterns of exchange with for-
eign cultures, but the range of translation practices devised by translators
in the domestic culture. In diminishing the translator's incentive for in-
vestment, copyright law deviates from its "traditional goals" of encouraging
and rewarding creative efforts (Bently 1993:495); the law now curtails
creativity in translation, the invention of translation projects and meth-
ods, as well as the creativity in literature that is inspired by the availability
of foreign works in inventive translations. This situation is particularly
exacerbated in the major English-speaking countries, the United States
and the United Kingdom, where the volume of translation has remained
relatively low throughout the post-World War II period. 2 The global he-
gemony achieved by these countries has made English the most translated
language, but the least translated into.
The history of copyright shows that earlier translators did not suffer
the same legal limitations as their successors today. On the contrary,
translation was advantaged by the centuries-long, sometimes contradic-
tory development of authorial rights in copyright law. There have been
decisions in which the translator's copyright in the translated text was not
only recognized, but given priority over that of an author or employer.
And, ironically enough, cases that proved decisive in reserving copyright
for the author contained alternative definitions of translation that were
much more favourable to translators.
These alternatives from the past can be useful in challenging the present
legal status of translation. They make clear that the historical develop-
ment of an exclusive authorial copyright coincides with, and indeed
depends on, the emergence of a Romantic concept of original authorship
that negates the translator's work. But they also enable the formulation of
a different concept of authorship, one in which the translator is seen as a
species of author, and originality is revised to embrace diverse writing
practices. What I shall present here is a genealogy of copyright that con-
tests the cultural assumptions of the law and aims to foster legislative
reform designed to further both the interests of translators and the prac-
tice of translation.
1. The current situation
Current copyright law defines translation inconsistently. On the one hand,

the author is distinguished from, and privileged over, the translator. Copy-
right is reserved for the author, the producer who originates the form of
the underlying work, and it covers only that form, the medium of expres-
sion as opposed to the idea or information expressed. The author's copyright
encompasses not only reproductions, printed copies of the work, but also
derivative works or adaptations, a category that explicitly includes trans-
lations, as well as such other derivative forms as dramatizations, film
versions, abridgements, and musical arrangements. On the other hand,
however, copyright in a derivative work can be reserved for its producer,
although without excluding the right of the author who produced the
underlying work. 3 Here, the translator is recognized as an author: accord-
ing to contemporary commentary, a translator can be said to author a
translation because translating originates a new medium of expression, a
form for the foreign text in a different language and literature. 4 Yet this
difference in the linguistic and literary medium is evidently not so sub-
stantive as to constitute a truly authorial originality for the translator,
since it does not in any way limit the right of the foreign author in the
translation. When copyright law treats derivative works, it contradicts its
key principle: that authorship consists of original expression, and hence
that legal protection is given only to forms, not ideas. 5 In current law, the
producer of a derivative work is and is not an author.
This contradiction indicates that copyright law must be protecting some-
thing else to the detriment of derivative works like translations. And that
something else, I want to suggest, includes an individualistic concept of
authorship. According to this fundamentally Romantic concept, the au-
thor freely expresses personal thoughts and feelings in the work, which is
thus viewed as an original and transparent self-representation, unmediated
by transindividual determinants (linguistic, cultural, social) that might
complicate authorial identity and originality.6 A translation, then, can never
be more than a second-order representation: only the foreign text can be
original, authentic, true to the author's psychology or intention, whereas
the translation is forever imitative, potentially contaminating or false.
Copyright law reserves an exclusive right in derivative works for the
author because it assumes that literary form expresses a distinct authorial
personality - despite the decisive formal change wrought by works like
translations.
This is evident in an American case concerning translations, Grove
Press, Inc. v. Greenleaf Publishing Co. (1965), in which the decision
Lawrence Venuti 5
waffled on the definition of originality as the criterion of authorship.7

Grove Press was seeking an injunction against Greenleaf, who published
without authorization The Thief's Journal, Bernard Frechtman's 1954
English version of Jean Genet's Journal du Voleur. The court found that
Greenleaf's publication infringed Genet's copyright in the French text:
It is obvious that Greenleaf copied not only the words of Frechtman,

the translator, but also the content and meaning of those words as
created in Jean Genet's original biographical story. This creation in-
cluded the entire plot, scenes, characters and dialogue of the novel,
i.e., the format and pattern. Greenleaf copied two things, (1) the
words and (2) the story. (524-525)
Although this decision linked Genet's authorship to the specific formal

organization of the French text ("the format and pattern"), the sense of
form was inconsistent and confused. Elements of literary form were cited
("plot, scenes, characters and dialogue"), but copyright was vested in "the
content and meaning of those words as created in Jean Genet's original
biographical story". The medium of expression vanished before the ideas
expressed. The "words" in this instance were English, not French, and
they were "created" or chosen by Frechtman, not by Genet. Yet they
communicated a "story" that was "original" because it originated with the
French author, with his life. The judge was uncertain about the precise
genre of Genet's work, describing it as both an autobiography and a
novel, because the criterion of authorship was ultimately not formal, but
thematic or semantic. The judge's certainty was that Frechtman's transla-
tion reproduced the meaning ofthe French text and therefore the author's
intention.
The Romantic concept of authorship thus elides any distinction be-
tween reproducing a work and preparing a derivative work based on it,
even though copyright law lists these two actions as distinct rights re-
served for the author. An unauthorized translation is an infringement of
the author's copyright because the translator produces an exact copy of
the form and content of the underlying work. A translation is not regarded
as an independent text, interposing linguistic and literary differences which
are specific to the translating culture, which are added to the foreign text
to make it intelligible in that culture, and which the foreign author did not
anticipate or choose. The foreign author's originality is assumed to tran-
scend any such differences, so that the translation can be viewed as
effectively identical to the foreign text. What copyright law protects is a
concept of authorship that is really not inscribed in a material form, but is
rather immaterial, a god-like essence of individuality that lacks cultural
specificity and permeates various forms and media.

The most explicit legal version of this concept is droit moral or rights
of personality, which developed in French, German and Scandinavian
jurisdictions during the nineteenth century and achieved international cur-
rency with the Berne Convention for the Protection of Literary and Artistic
Works (Rome Revision 1928).8 Under the droit moral, the identity be-
tween author and work is phrased in moralistic terms, with the work
considered an embodiment of the author's person. In his 1934 commen-
tary on the Rome Revision, Marcel Plaisant described the legal thinking
behind this concept:
Above and beyond the pecuniary and patrimonial right, we under-

stand that the author exercises a lofty sovereignty over his work, such
that when it is damaged he is injured. Publication is envisaged as a
phenomenon that extends the personality of the author and thus ex-
poses him to further injuries because the surface of his vulnerability
has been enlarged. (Saunders 1992:31)
The droit moral gives the author various personal rights, including the
right to be identified as author, the right to control the first publication,
and the right to object to a distorted treatment of the work which may
damage the author's reputation. Derivative works like translations could
conceivably provoke a legal action under this last right, which has been
included in the droit moral section of the Berne Convention since the
Brussels Revision (1948). In principle, legal protection against distortions
endows authors with enormous power over every aspect of the translating
process, permitting them to develop their own idea of what constitutes the
integrity of their work in a foreign language. Interestingly, British law,
although it recognizes the author's moral rights, is alone in specifically
excluding translations from the right to object to a distorted treatment. 9 Is
translation excluded in this case because it is assumed to communicate the
foreign author's personality without distortion? Or is the assumption that
another authorial personality has intervened, the translator's, which is
communicated in the translation and therefore requires protection in deal-
ings with the domestic publisher and the foreign author? Lionel Bently
suggests that "the legislature effected a broad exclusion of translations in
order to recognise the difficulty and subjectiveness of determinations of
the quality of translations" (Bently 1993:514). Whatever rationale may be
offered for this exclusion, it seems clear that droit moral further restricts
the translator's rights, yet without in any way resolving the inconsisten-
cies in current legal definitions of translation. Copyright law admits that
Lawrence Venuti 7
translation sufficiently alters the form of the foreign text to be copyrightable

by the translator. Yet to allow the foreign author to assert a moral right of
integrity over the translation would be to deny this basis of the transla-
tor's authorship. The economic disadvantage to the translator (and the
publisher of the translation) is clear: as Bently puts it, "to require the
author's approval...would be to give him a second opportunity to bargain
in a situation where the derivative user has made considerable invest-
ment" (ibid 1993:513).
The inconsistencies arise, moreover, not just between copyright codes
at different levels of jurisdiction, national and international, but within
the very international treaties that were designed to foster greater uni-
formity in the protection of intellectual works. The Berne Convention did
not recognize the translator's copyright in the translated text until the
Paris Revision (1971), yet this new awareness of translation produced no
change in the author's exclusive right to license derivative works. The
pertinent article reads: "Translations, adaptations, arrangements of music
and other alterations of a literary or artistic work shall be protected as
original works without prejudice to the copyright in the original work".l0
The repetition of original here calls attention to the shifting concept of
authorship in international copyright law. The autonomy of translation as
original work is enhanced by separating author from translator. But the
originality that entitles translators to legal protection is obviously not the
same as that of foreign authors, who still enjoy "the exclusive right of
making and authorising the translation of their works" (article 8).
2. The contradictory development of original authorship
The Romantic concept of original authorship emerged relatively late in

the history of copyright. Although the first English formulations of this
concept occurred in literary treatises like Edward Young's Conjectures on
Original Composition (1759), it did not prevail in copyright law until the
middle of the nineteenth century. In an 1854 case before the House of
Lords,leffreys v. Boosey, a justice answered the claim that copyright "is a
mental abstraction too evanescent and fleeting to be property" by invok-
ing the distinction between the medium of expression and the idea
expressed, only to collapse it. "The claim is not to ideas", he argued at
first, "but to the order of the words, and ... this order has a marked identity
and a permanent endurance".l1 Yet it quickly became clear that the iden-
tity the justice had in mind was in fact a mental abstraction, since the
work was analogous to the author's physiognomy:
Not only are the words chosen by a superior mind peculiar to itself,
but in ordinary life no two descriptions of the same fact will be in the
same words, and no two answers to your Lordships' questions will be
the same. The order of each man's words is as singular as his counte-
nance. (ibid.)
Although copyright was vested in the medium of expression, the medium

was characterized as a transparent representation of the author's personal-
ity, a "mind" of a "superior" and "peculiar" kind. The importance assigned
to an abstraction like personality inevitably evaporated form, with the
result that the scope of the author's copyright was expanded to include
any alteration in "the order of the words", no matter how substantial.
Accordingly, the period that saw the authorial personality prevail in the
courts also saw the institution of statutes that gave the author the right to
prepare derivative works like translations. Although the Statute of Anne,
the first act to protect authorial rights, was instituted in 1710, British law
did not give the author an exclusive translation right until 1852, American
law not until 1870P
The law was slow to recognize this right partly because another, con-
flicting concept of authorship had prevailed before the mid-nineteenth
century. According to this concept, copyright was reserved for the author,
not because the work represented a personality, but because it was a prod-
uct of labour, not because it expressed thoughts and feelings, but because
it resulted from an investment of time and effort, both mental and physi-
cal. As one justice asserted in Millar v. Taylor (1769), a landmark case in
the establishment of authorial rights, "it is just, that an author should reap
the pecuniary profit of his own ingenuity and labour".13 Copyright was
found to exist in the common law: the author enjoyed a perpetual right in
the work. The decision assumed that this right was natural, following John
Locke's theory of private property. In his Second Treatise of Civil Govern-
ment (1690), Locke argued that
every Man has a Property in his own Person. This no Body has any
Right to but himself. The Labour of his Body, and the Work of his
Hands, we may say, are properly his. Whatsoever then he removes
out of the State that Nature hath provided, and left it in, he hath mixed
his Labour with, and joyned to it something that is his own, and
thereby makes it his Property. (Locke 1960:305-6)
As this passage suggests, the concept of authorship as labour invest-

ment is just as individualistic as the Romantic insistence on personality:
an author is completely autonomous from nature and from other persons;
Lawrence Venuti 9
authoring is a free appropriation of natural materials. And the defining

characteristic of authorship, namely labour, turns out to be just as imma-
terial as personality: the author's labour grants a natural right over a work
that is itself natural, with both right and work transcending any specific
cultural determinations or social constraints. Of course the very fact that
the author's copyright requires legal protection, developed in various cases
and enacted by various statutes, indicates that the relation between an
individual and the product of that individual's labour is not natural, but
legally constructed in response to changing cultural and social conditions.
In Millar v. Taylor, these conditions included Locke's liberal theory of
private property, as well as a book industry that functioned as a market for
copyrights and so devised a concept of authorship by which authors were
entitled to transfer their rights to booksellers. As Susan Stewart has re-

marked, "the right of authors was merely a thread amid the complex
relations between state interest, common-law rights to intellectual prop-
erty, and commercial competition emerging throughout the eighteenth
century" (Stewart 1991:15).14 The material conditions of authors' rights
are denied as much by Lockean possessive individualism as by the Ro-
mantic theory of personal expression.
The concept of authorship as labour is interesting, not for its liberal
assumptions, but for its enlargement of the scope of copyright law to
address what today are classified as derivative works. The cases that
defined authorial rights in the wake of the Statute of Anne acknowledged
a translation to be an independent work which did not infringe the copy-
right of the author who produced the underlying work. A key case is
Burnett v. Chetwood (1720). The executor of Thomas Burnett's estate was
seeking to enjoin the defendant from publishing an unauthorized English
translation of Burnett's Latin work, Archaeologia Philosophica (1692), a
theological treatise which included a dialogue between Eve and the ser-
pent that embarrassed the author when translated. 15 The court granted the
plaintiff's suit, although the decision was neither an application of the
statute to protect Burnett's copyright nor an implicit recognition of his
moral right to protect his reputation. The justice was less interested in
interpreting copyright law than in making a paternalistic gesture of cen-
sorship:
Lord Chancellor said, that though a translation might not be the

same with the reprinting the original, on account that the translator
has bestowed his care and pains upon it, and so not within the
prohibition of the act, yet this being a book which to his knowledge
(having read it in his study), contained strange notions, intended by
the author to be concealed from the vulgar in the Latin language, in
which language it could not do much hurt, the learned being better
able to judge of it, he thought it proper to grant an injunction to the
printing and publishing it in English; that he lookt upon it, that this
Court had a superintendency over all books, and might in a sum-
mary way restrain the printing or publishing any that contained
reflections on religion or morality.16
The decision wound up supporting what was "intended by the author",

but it actually involved a legal definition of translation that put it outside
of the author's copyright. Agreeing with the defendant's counsel that
authorship consisted of labour invested in the production of a work, the
Lord Chancellor distinguished between "reprinting the original" and trans-
lating it and hence assumed that the translator was an author, not a copyist.
In Millar v. Taylor, the justices drew this distinction even more sharply.
Although they found that the author held a perpetual copyright, one be-
lieved that "certainly bona fide imitations, translations, and abridgments
are different; and in respect of the property, may be considered as new
works", whereas another asserted that a purchaser of a book "may im-
prove upon it, imitate it, translate it; oppose its sentiments: but he buys no
right to publish the identical work"Y In the early history of copyright
law, the author was given only the right to reproduce the work, not to
prepare a derivative work based on it. In fact, a translation was seen, not
as derivative, but as original, or new, because it resulted from the transla-
tor's labour. TlYatt v. Barnard (1814) found that "Translations, if original,
... could not be distinguished from other Works", and so a copyright could
be held in a translation by the translator or by the translator's employer,
unless that translation copied another translated text (i.e., it was not origi-
nal).18 Originality was assumed to be a precise selection and arrangement
of words, regardless of whether those words were intended to imitate
another work.
The concept of authorship as labour investment thus led to an empha-
sis on form as the basis of copyright, and this emphasis supported the
translator's right in the translation. In Burnett v. Chetwood, the defend-
ant's counsel observed that the Statute of Anne, insofar as it was intended
to promote creativity and the dissemination of knowledge, protected only
the form of the author's work, not the content (the sense), and therefore
the translator's creation of a different form for that content excluded the
translation from the author's copyright. Translation, counsel concluded,
"should rather seem to be within the encouragement than the prohibition
of the act" (1009). The assumption here was twofold: on the one hand, the
ideas in the underlying work were regarded as public knowledge upon
publication, so that an author could own no more than their initial me-
Lawrence Venuti 11
dium of expression; on the other hand, the translator's form-creating la-

bour - the "skill in language" that resulted in the production of "his own
style and expressions" - made him the owner of the translation that dis-
seminated those ideas (1009). A similar assumption underlay the decision
in Donaldson v. Beckett (1774). This crucial case upheld the Statute of
Anne, but repealed the perpetual right given to the author in Millar v.
Taylor precisely because, in Lord Camden's words, "science and learning
are in their Nature publici Juris, and they ought to be as free and general
asAir or Water" (Parks 1975:53). For Camden, any perpetual right, whether
grounded in the author's ideas or form, would hinder their circulation in
derivative works. If copyright were vested "in the Sentiments, or Lan-
guage", he pointed out, "no one can translate or abridge them", an effect
that was contrary to the aims of the statute (ibid.:52).

This line of thinking received its most extreme articulation in anAmeri-
can case, Stowe v. Thomas (1853). The court found that an unauthorized
German translation of Harriet Beecher Stowe's novel, Uncle Tom's Cabin
(1852), did not infringe her copyright in the English text. 19 Citing such
earlier cases as Burnett v. Chetwood and Millar v. Taylor, the judge recog-
nized the decisive intervention of the translator's labour: "The same
conceptions clothed in another language cannot constitute the same com-
position", since "to make a good translation of a work often requires more
learning, talent and judgement than was required to write the original"
(208). The judge limited Stowe's right to the actual language of her novel
because granting her control over translations would interfere with the
circulation of her ideas, thereby contradicting the constitutional view of
authorial copyright as a legal means "to promote the Progress of Science
and useful Arts". 20 The decision sought to foster the cultural creativity
reflected in derivative works, however uneven in quality they might be,
while strictly defining copyright infringement as unauthorized reproduc-
tion:
By the publication of Mrs. Stowe's book, the creations of the genius

and imagination of the author have become as much public property
as those of Homer or Cervantes. All her conceptions and inventions
may be used and abused by imitators, play-rights and poetasters. All
that now remains is the copyright of her book; the exclusive right to
print, reprint and vend it, and those only can be called infringers of
her rights, or pirates of her property, who are guilty of printing, pub-
lishing, importing or vending without her license, "copies of her book".
A translation may, in loose phraseology, be called a transcript or copy
of her thoughts or conceptions, but in no correct sense can it be called
a copy of her book. (208)
Stowe v. Thomas in effect gave translators an exclusive copyright in their

translations, distinct from the copyright in the underlying work held by its
author. And this meant, in principle, that translators could control every
step in the translation process, from choosing a foreign text to translate, to
developing a translation method, to authorizing the publication of the
translated text.
Yet Stowe v. Thomas never achieved the authority of a precedent; in
the history of copyright, the case has proved to be eccentric. For precisely
during the period when it recognized translators as authors by virtue of
their form-creating labour, the Romantic concept of authorship came to
dominate the law, dooming translation to the ambiguous legal status that
it currently occupies. This development can be glimpsed in Byrne v. Statist
Co. (1914), a British case that is sometimes cited for its recognition ofthe
translator's rights, but that actually circumscribes them within narrow
bounds. The court decided that a newspaper had infringed a translator's
copyright by publishing his translation without his permission. The judge
agreed with the plaintiff's counsel that the translator owned the copyright
in the translation according to the recently instituted act:
This translation was an "original literary work" within s. 1, sub-so 1,

of the Copyright Act, 1911. It is "original" because it is not a mere
copy of the work of another person. Originality of idea is not neces-
sary; it is sufficient if the work is in substance a new thing involving
fresh skill and labour. This translation is "original" work in that sense,
and it is "literary" work .... The plaintiff is the "author" of the work,
and is therefore the owner of the copyright therein. 21
Although the concepts favourable to the translator seem to be in place

here - authorship as labour investment, originality as form - they were
radically qualified by the Act of 1911. This same act defined translation
as a "mere copy" by reserving for the author the exclusive right "to
produce, reproduce, perform, or publish any translation of the work" .22 In
Byrne v. Statist Co., both the translator and the infringing newspaper had
in fact purchased a translation right from the foreign author; the newspa-
per, however, neglected to approach the translator as well for permission
to reprint his translation. This case certainly recognized the translator as
an author, but not one whose copyright in the translation superseded or in
any way limited the foreign author's. The act, therefore, was implicitly
defining authorship as something less tangible than labour, something
that transcended formal changes, an abstraction that negated the transla-
tor's work: the foreign author's ideas, intention, or personality.
Lawrence Venuti 13
3. The formal basis of the translator's authorship
The history of copyright may indeed contain alternative definitions of

translation that favour translators. But the neglect into which these defini-
tions have fallen, their sheer lack of legal authority today, indicates that
they require substantial rethinking to challenge the dominance of the
Romantic concept of authorship and to prove useful in legislative reform.
This rethinking must encompass the basic concepts of copyright law,
beginning with the understanding of form that defines authorship.
The early cases conceive of linguistic and literary form as transparent
communication. Meaning is assumed to be an unchanging essence em-
bedded in language, not an effect of relations between words that is
unstable, varying with different contexts. Hence the clothing metaphors

that recur in the cases: an author is said to clothe meaning in language; a
translator then communicates the meaning of the foreign text by changing
its linguistic clothes. In copyright law, this concept of form first appeared
in Burnett v. Chetwood, where, however, it was simultaneously put into
question. The defendant's counsel argued that a translation "may be called
a different book" because
the translator dresses it up and clothes the sense in his own style and
expressions, and at least puts it into a different form from the original,
and forma datesse rei. (1009)
The Latin axiom was drawn from the Aristotelian metaphysics that pre-
vailed in medieval scholastic philosophy: in a fairly close rendering, "form
brings things into existence". The counsel apparently cited this meta-
physical principle to establish the relative autonomy of the translation
from the foreign text: translating is seen as form-creating, and therefore
the translation can be said to exist as an object independent of the under-
lying work on which it is based. Yet the axiom also suggests that the
translation effectively creates the foreign text in another language, that
the different form created by the translator brings into existence another
text with a different meaning. Ifforma dat esse rei, form cannot easily be
detached from content, nor can formal changes preserve the same content
unchanged. Hence, the translator's new "style and expressions" must pro-
duce a new "sense".
The decision itself supports this understanding of form, because it
documents the fact that the meaning of Burnett's Latin treatise changed
when translated into English. The plaintiff's counsel found the translation
a mixture of error and parody, "the sense and words of the author mis-
taken, and represented in an absurd and ridiculous manner" (1009). The
Lord Chancellor saw the change wrought by the translation in social

terms: the "strange notions" of the Archaeologia Philosophica, he noted,
were "learned" and innocuous in Latin, but "vulgar" and potentially harm-
ful in English. The meanings of the two texts, then, were determined by
the writers' creation of different forms that addressed different audiences.
The reference to these audiences demonstrates that authorship is not indi-
vidualistic, but collective: the form of the work does not originate simply
with the author as "his own style and expressions", but is in effect a
collaboration with a specific social group, wherein the author takes into
account the cultural values characteristic of that group.
This collective concept of authorship applies to both the translation
and the underlying work. The texts at issue in Byrne v. Statist Co. were a
Portuguese speech delivered by a Brazilian governor to the state legisla-

ture and plaintiff's English translation published as an advertisement in
an influential London-based newspaper, the Financial Times (624). The
different social situations for which the texts were written ensured that
they would take different forms and carry different meanings for their
readers. The governor's speech was political, serving as "a message to the
GeneralAssembly of that State dealing with its finances", whereas Byrne's
translation was commercial, designed to provide information for potential
investors (623). The social function of each text was inscribed in its form,
most obviously in each author's use of a specific language for a specific
audience, but also in the different literary and rhetorical structures chosen
by each author to signify in a different social context. The collective
nature of authorship becomes clear in the judge's statement of the facts,
which reports Byrne's detailed description of his own translation:
He cut down the speech by about one third. He edited it by omitting

the less material parts. He divided it into suitable paragraphs, and
supplied head-lines appropriate to those paragraphs. He told me too
that the Financial Times sets a high standard of literary style and that
his translation conformed to that high standard. (624)
The commercial function that Byrne's translation was intended to per-

form required not only that it communicate the same financial information
as the governor's speech, but that this information be assimilated to do-
mestic cultural values, rewritten according to a new stylistic "standard" in
English, edited according to a new, distinctively journalistic format ("para-
graphs" and "head-lines"), and reinterpreted according to an English
investor's sense of pertinence (the omission of "less material parts").
Byrne v. Statist Co. indicates that the form of a work is not only
collaborative, constituted by a relation with an audience, but derivative,
Lawrence Venuti 15
not originating in the author's personality or productive labour on raw

nature, but drawn from pre-existing cultural materials. The Brazilian gov-
ernor's speech was written in the style of a political address, Byrne's
translation in the style of business journalism. The styles preceded the
composition of the texts and determined their meanings, however much
those styles were elaborated and fitted to a specific purpose and occasion.
The copyrightable form in a work, then, is not self-originating, but uniquely
derived: the precise selection, arrangement, and elaboration of materials
that already exist in a culture, not merely the lexicon, syntax, and phonol-
ogy that define a particular language, but the structures and themes that
have accumulated in the various cultural discourses of that language:
literary, rhetorical, political, commercial, and so forth. It is from these
materials, never raw or natural, always culturally coded by previous uses,

that an author produces a form determined by an address to a particular
cultural constituency.
Still, the collective authorship of a translation differs in an important
way from that of the underlying work. Even though every work appropri-
ates other works to some extent, a translation is engaged in two,
simultaneous appropriations, one of the foreign text, the other of domes-
tic cultural materials. The relation between translation and foreign text is
mimetic and interpretive, governed by canons of accuracy and methods
of interpretation that vary culturally and historically, whereas the relation
between translation and domestic culture is mimetic and communicative,
governed by an imitation of cultural materials to address audiences that
are culturally and historically specific. In translating, the interpretation of
the foreign text and the address to an audience are mutually determining,
although in any given translation one of these determinants may outweigh
the other: the projected audience may decisively shape the translator's
interpretation, or the translator's interpretation may decisively define the
audience.
Contemporary translations, unlike such other derivative forms as dra-
matic or film adaptations, are bound to a much closer relation to the
underlying work, partly because of the Romantic concept of authorship.
The dominance of this concept instills in translators and their publishers a
deference to the foreign text that discourages the development of innova-
tive translation methods which might seem distorting or false in their
interpretations. Today, a dramatic or film adaptation of a novel may devi-
ate widely from the plot, characterizations, and dialogue in that novel, but
a translation is expected to imitate these formal elements without revision
or deletion.
Nonetheless, the closeness of the relation between translation and
foreign text should not be taken as implying that the two works are
identical, or that the translation is not an independent work of authorship.
If authorship is collective, if a work both collaborates with and derives
from a cultural context, then the translation and the foreign text are dis-
tinct projects because they involve different contexts. The significance of
a foreign novel in the foreign literature where it was produced will never
be exactly the same as the significance of that novel in a translation
designed for circulation in another language and literature. This goes
some way toward explaining why bestsellers do not always repeat their
success in a foreign country when translated. The variation in signifi-
cance, moreover, cannot be limited or pre-empted by the appearance of
the same author's name on the foreign text and the translation: for readers
of the foreign text, that name will project a different identity, tied to the
foreign language and the cultural traditions of the foreign country, than
the somewhat domesticated identity projected by the translation. To take
an extreme yet illuminating example, ever since Islamic fundamentalists
called for the death of the British writer Salman Rushdie because they
judged his novel The Satanic ~rses to be blasphemous of the Koran, the
name "Salman Rushdie" has differed in meaning, depending not only on
the cultural values that a reader brings to any book attributed to this
writer, but also on the language in which it circulates. The identity linked
to Rushdie's name is likely to vary according to whether a book of his is
published in English or in an Arabic translation.
Copyright law has failed to acknowledge the manifold relations that
determine any translation because it has been dominated by individualis-
tic concepts of authorship, whether Lockean or Romantic, whether
grounded in labour or in personality. These concepts have diminished the
legal status of derivative forms, while concealing the degree to which the
underlying work is itself derivative. A collective concept of authorship
offers a precise definition of form to distinguish between a translation and
the foreign text it translates: the collaborative and derivative dimensions
of form result in linguistic and cultural differences that can serve as the
basis for the translator's claim to copyright, but also for an argument in
favour of restricting the foreign author's right in the translation.
Current copyright law, however, lacks the conceptual tools to formu-
late such a restriction. British and American codes (among others) provide
for a "joint work", for instance, but the concept of authorship assumed
here is not in fact collective, but individualistic. Thus, a joint work is
regarded as seamlessly unified: the "contributions" of "each author" are
"not distinct" or are "merged into inseparable or interdependent parts of a
unitary whole".23 In the case of a derivative form like translation, the
Lawrence Venuti 17
contributions of the translator and the foreign author can be distinguished:

the translation imitates the linguistic and literary values of a foreign text,
but the imitation is cast in a different language with relations to a different
cultural tradition. As a result, the translator contributes a form that partly
replaces and in general qualifies the form contributed by the foreign au-
thor. A foreign novelist may be said to contribute the characters in a novel
to the translation, but the nature of those characters as evidenced in dia-
logue or description will inevitably be altered by the values of the
translating language and culture. The notion of indistinct contributions
still rests on the individualistic assumption that linguistic and literary
form enables transparent communication by a single person, as opposed
to communication determined collectively by cultural materials and so-
cial contexts.
The definition of a joint work is particularly inhospitable to derivative
forms like translation because it stipulates an "intention" to collaborate
shared by the authors "at the time the writing is done".24 The assumption
is that the work is produced by two individuals in concert and over a well-
defined period of time. Yet this does not take into account the reality of
translation projects today. According to current practices, several years
are likely to elapse between the publication of a foreign text and its
translation, unless the foreign text was written by an author of previous
international bestsellers and is therefore of immediate interest to publish-
ers worldwide. The development of a translation project requires numerous
tasks that vary in complexity, but all of which are time-consuming: these
tasks begin with the domestic publisher's selection of a foreign text to
translate and include the negotiation of translation rights with the foreign
author or publisher, the commissioning of a translator, and the editing of
the translation. The publication of a translation can thus be considered a
collective project, involving the collaboration of many agents at different
stages. The foreign author's participation is of course indispensable, but it
may finally be limited to the writing of the foreign text that is the basis of
the project. What argues against viewing a translation as a joint work is
not merely the different times at which foreign author and translator make
their contributions, but the absence of a shared intention. Foreign authors
address a linguistic and cultural constituency that does not include the
readers of their works in translation. Translators address a domestic con-
stituency whose demand for intelligibility in the terms of the translating
language and culture exceeds the foreign author's intention as realized in
the foreign text.
Recent cases and commentary suggest that a translation may be con-
sidered a fair use of a foreign text which is exempt from the foreign
author's exclusive copyright in derivative works. A use of a copyrighted

work is defined as fair when it serves "purposes such as criticism, com-
ment, news reporting, teaching (including multiple copies for classroom
use), scholarship, or research". 25 Many kinds of translations, both literary
and technical, serve such purposes, and in the case of literary works a
translation can always be seen as an interpretation of the foreign text, a
criticism or commentary that determines its meaning for a domestic audi-
ence. A fair-use argument for translation can be developed further on the
basis of Campbell v. Acuff Rose Music, Inc. (1994), in which the United
States Supreme Court held that a rap song, 2 Live Crew's "Pretty Woman",
may constitute a fair use of the rock ballad which it parodied, Roy Orbison's
"Oh, Pretty Woman".26 The court stated that "like less ostensibly humor-
ous forms of criticism", parody "can provide social benefit by shedding

light on an earlier work, and, in the process, creating a new one" (1171).
Parody, like translation, involves an imitative rewriting of an underlying
work, while the mimetic relation between translation and foreign text may
sometimes be parodic (the English translation in Burnett v. Chetwood, for
example, was described as an "absurd and ridiculous" version of Burnett's
Latin treatise). A translation can be viewed, more generally, as one of
those "less ostensibly humorous forms of criticism" to which the justice
referred, a commentary on the foreign text that is subtly enacted through
imitation.
Yet a fair-use argument for translation may falter on the additional
factors that must be considered for any such exemption from the foreign
author's exclusive copyright. Recent cases make clear that the most im-
portant of these factors are "the purpose and character of the use, including
whether such use is of a commercial nature", "the amount and substantial-
ity of the portion used in relation to the copyrighted work as a whole", and
"the effect of the use upon the potential market for or value of the copy-
righted work".27
A translation, insofar as it is written in a different language for a
different culture, does not limit the potential market for the foreign text in
its own language and culture; in fact, the translation of a work into many
languages could increase its literary and commercial value at home by
demonstrating its value abroad. Nor does a translator use too much of the
foreign text to sustain a fair-use defence. Today, a translation is expected
to render the foreign text in its entirety; if a translation alters or omits
substantial portions of that text, it would no longer be considered a trans-
lation, but another kind of derivative form, such as an adaptation or
abridgement. More importantly, the peculiar kind of writing involved in
any translation forces a distinction between copying and imitating the
Lawrence Venuti 19
foreign text. A translation does not copy in the sense of repeating that text
verbatim; rather, the translation enters into a mimetic relation that inevita-
bly deviates from the foreign language by relying on target-language
approximations. Even though a contemporary translation is required to
imitate the entire foreign text, their linguistic and cultural features are
sufficiently distinct to permit them to be considered autonomous works.
The factor that might finally mark an unauthorized translation as an
infringement under the fair-use provision is the purpose and character of
the use to which the translator puts the copyrighted work. Certainly, trans-
lators select and translate foreign texts for purposes that can be described
as cultural or even educational. Translations do not just increase knowl-
edge in diverse humanistic and technical fields; they can also maintain,
revise, or create conceptual paradigms, research methodologies, and clinical

practices in the disciplines and professions of the translating culture. And
translations can be enlisted in the service of democratic political agendas
that promote cultural and social change, for example, by gaining author-
ity for the cultural forms of marginalized social groups, by questioning
exclusionary values like nationalism and racism, and by challenging do-
mestic stereotypes of foreign cultures to improve political relations with
those cultures. 28 At the same time, however, translators are also motivated
by a significant commercial interest, since they aim to profit from their
translations. It is this very interest that copyright law was designed to
protect so as to encourage the creation of cultural and educational works.
But the fair-use provision frustrates this design by assuming, quite contra-
dictorily, that authors of derivative works like translations should not
share the commercial motives of other authors.
Perhaps the most effective way to calibrate the competing interests in
a translation project is the one that takes into account the actual dealings
of translators, publishers, and authors. By far the most important consid-
eration here is time. If an author or publisher does not sell the translation
rights for a work soon after its first publication, any project to translate it
will most likely originate in the translating culture and require several
years to develop. During this period, a work that initially lacked value in
the translating culture comes to be valuable through the efforts of a trans-
lator or publisher, notably through translating and publishing strategies
that address domestic cultural constituencies and locate or establish mar-
kets for the translation. A limitation of the foreign author's right in the
translation to a definite period - say, five years - would encourage transla-
tors and domestic publishers by increasing the incentive for investment in
translations. If the foreign text is not translated within the five-year pe-
riod, the first translator or publisher to publish a translation of it thereafter
would not only be permitted to copyright the translation, as current law

provides, but would also enjoy an exclusive translation right in the for-
eign text for the full term of the copyright. Such a limitation would motivate
translators to apply and enhance their expertise in foreign languages and
cultures by allowing them to invent translation projects that answer to
their own sense of domestic cultural values - without fear of legal repris-
als from foreign authors or of uninformed, cost-conscious rejections from
domestic publishers. A limitation of the foreign author's copyright would
also motivate publishers to develop and issue more translations without
the added burden of paying that author for the right to translate the for-
eign text. 29
Current copyright law, however, does not define a space for the trans-
lator's authorship that is equal to, or in any way restricts, the foreign
author's exclusive right. And yet it acknowledges that there is a material
basis to warrant some such restriction. The collective concept of author-
ship outlined here puts the translator on an equal legal footing with the
author of the underlying work. According to this concept, copyright would
be grounded on precise formal features which show that similar proce-
dures are involved in creating the foreign text and the translation, and
these procedures occur with sufficient autonomy, in different linguistic
and cultural contexts, to allow the works to be viewed as independent.
Without a greater recognition of the collective nature of authorship, trans-
lators will continue to be squeezed by unfavourable, if not simply
exploitative contracts, and publishers around the world will continue to
support the unequal patterns of cross-cultural exchange that have accom-
panied economic and political developments in the post-World War II
period. It is the sheer global reach of translation, its strategic and irre-
placeable value in negotiating cultural differences, that lends urgency to
the need for a clarification and improvement of its legal status.
LAWRENCE VENUTI
Department of English, Temple University, Philadelphia, Pennsylvania
19122, USA
Notes
This essay has benefitted from the helpful comments of several readers whom I
would here like to acknowledge: Lionel Bently, Steven Cole, Ian Mason, Ewald
Osers, and Susan Stewart.
Lawrence Venuti 21
1. For the United Kingdom and the United States, see Copyright, Designs and
Patents Act 1988 (c. 48), sections 2(1), 11(1) and (2), 16(1)(e), 21(3)(a)(i),
and 17 United States Code, sections 101, 106(2), 201(a) & (b) (1976). For a
comprehensive account of the legal status of translation, see Bently (1993).
2. The latest statistics for world translation publications are given in the
UNESCO Statistical Yearbook, 1990. The data indicate the volume of trans-
lations from and into selected languages between 1982 and 1984. English
tops the list as the most translated language, with figures ranging between
22,000 and 24,500 publications; French ranks second with figures ranging
from 4,400 to 6,200. Translations into English for the same period range
from 950 to 1300, whereas French translations range from 1800 to 3800.
See also Tables A and B in Grannis 1991. British statistics are given in the
annual Whitaker's Alamanack.
3. Copyright, Designs and Patents Act 1988, sections l(l)(a), 16(1)(e),

21(3)(a)(i); 17 United States Code, sections 102(a) and (b), 103(a), 106(2)
(1976).
4. See, for example, Skone James et al. 1991:3-34, and Chisum and Jacobs
1992: 4C[1][c]. In Signa Trading International v. Gordon, 535 F. Supp.
362, 214 U.S. P.O. 793 (N.D. Calif. 1981), the court held that a machine-
translated list of English words and phrases into Arabic could not be
copyrighted because the translation did not contain sufficient originality
(the translation included phrases like "how are you"). The court commented
that "Translations of many things, such as literary works, are copyrightable
to the extent that translation involves originality. ... It is not the transla-
tions of individual words that makes these works copyrightable, it is rather
the originality embodied in the translator's contributions, for example, con-
veying nuances and subtleties in the translated work as a whole" (535 F.
Supp. 364; 214 U.S.P.O. 795).
5. This contradiction appears in copyright codes in other national jurisdic-
tions: for Canadian law, see Braithwaite (1982:204); for French law, see
Derrida (1985:196-99).
6. Abrams 1953 offers a literary history of the Romantic concept of author-
ship. For histories that address the economic and legal conditions of this
concept, see Woodmansee (1984), Saunders (1992), and Rose (1993).
7. Grove Press, Inc. v. Greenleaf Publishing Co., 247 F. Supp. 518 (E.D.N.Y.
1965).
8. For a history of the emergence of droit moral, see Saunders (1992: chapter 3).
9. Copyright, Designs and Patents Act 1988, section 80(2)(a)(i).
10. Berne Convention for the Protection of Literary and Artistic Works, Paris
1971, 2(3). The UNESCO Recommendation on the Legal Protection of
Translators and Translations and the Practical Means to Improve the Status
of Translators (adopted by the General Conference at Nairobi, 22 Novem-
ber 1976) adopts the wording of the Berne Convention and thereby
subordinates translators to authors of the underlying work: "Member states
should accord to translators, in respect of their translations, the protection
accorded to authors under the provisions of the international copyright

conventions to which they are party and/or under their national laws, but
without prejudice to the rights of the authors of the original works trans-
lated" (11.3).
11. Jeffreys v. Boosey, 4 H.L.e. 815, 869; 10 Eng. Rep. 681 (1854). The key
American case that espoused the Romantic concept of authorship isBleistein
v. Donaldson Lithographing Co., 188 U.S. 239 (1903), where the court
asserted that "[the work] is the personal reaction of an individual upon
nature" (250). The different concepts of authorship assumed in copyright
law are carefully distinguished by Ginsburg (1990:1873-88).
12. Copyright Act of 1852 (15 & 16 rict., c. 12); United States Act of 8 July
1870, ch. 230, s. 86, 16 Stat. 198. \
13. Millar v. Taylor, 4 Burr. 2303; 98 Eng. Rep. 201 (K.B. 1769).
14. The social conditions of the Statute of Anne are discussed in detail by Rose
(1993: chapter 3) and Saunders (1992: chapter 2).
15. Rose (1993:49-51) describes the circumstances of Burnett v. Chetwood.
16. Burnett v. Chetwood, 2 Mer. 441; 35 Eng. Rep. 1008 (1720).
17. Millar v. Taylor, 98 Eng. Rep. 203, 205.
18. Wyatt v. Barnard, 3 Ves. & B. 77; 35 Eng. Rep. 408 (Ch. 1814).
19. Stowe v. Thomas, 23 Fed. Cas. 201 (No. 13514) (e.e.E.D.Pa. 1853).
20. United States Constitution, article I, section 8, clause 8 (1790) provides that
"The Congress shall have Power ... To promote the Progress of Science and
useful Arts, by securing for limited Times to Authors and Inventors the
exclusive Right to their respective Writings and Discoveries".
21. Byrne v. Statist Co., 1 K.B. 622 (1914).
22. Copyright Act of 1991 (1 & 2 Geo. 5, c. 46), 1(2)(b).
23. Copyright, Designs and Patents Act 1988, section 10(1); 17 United States
Code, section 101, 201(a) (1976). The definition of a "joint work" as a
"unitary whole" shows that British and American copyright law assumes
the individualistic concept of organic unity which has long dominated liter-
ary criticism: see Venuti (1985/86).
24. The phrase "at the time the writing is done" is quoted from the gloss on the
definition of a "joint work" in the U.S. code: H.R. Rep. No. 1476, 94th
Cong., 2nd Sess. 103, 120. See also Jaszi (1994). Although Jaszi does not
consider translation, he includes a pertinent discussion of how the concept
of "joint authorship" fails to recognize "serial collaborations" - "works
resulting from successive elaborations of an idea or text by a series of
creative workers, occurring perhaps over years or decades" (ibid:40; see
especially 50-55).
25. 17 United States Code, section 107. British law contains a comparable
limitation of the author's exclusive right by providing for "fair dealing",
the use of a copyrighted work for the purposes of "research or private
study", "criticism or review", "reporting current events": Copyright, De-
signs and Patents Act 1988, sections 29(1) and 30(1) and (2).
26. Campbell v. Acuff-Rose Music, Inc., 114 S.Ct. 1164 (1994). The decision
is discussed by Greenhouse (1994).
Lawrence Venuti 23
27. 17 United States Code, section 107(1), (3), and (4).

28. For detailed discussions of the cultural and political agendas served by
translation, see Venuti (1995).
29. This proposal resembles, but ultimately goes beyond, previous attempts to
limit the author's translation right. In the United Kingdom, for example,
such a limitation was established for foreign authors by the Copyright Act
of 1852. Foreign authors were given three years in which to translate their
work; if a translation was issued within this period, the author enjoyed the
translation right for five years from the date of publication. By the 1911
Act, however, the translation right was assimilated to the author's exclu-
sive reproduction right. See Bently (1993:501-5) for a discussion of the
legislative changes.
References
Abrams, M. H. (1953) The Mirror and the Lamp: Romantic Theory and the
Critical Tradition, New York and Oxford: Oxford University Press.
Bently, Lionel (1993) 'Copyright and Translations in the English-speaking
World', Translatio 12: 491-559.
Braithwaite, William (1982) 'Derivative Works in Canadian Copyright Law',
Osgoode Hall Law Journal 20: 192-23l.
Chisum, Donald S. and Michael A. Jacobs (1992) Understanding Intellectual
Property Law, New York and Oakland: Matthew Bender.
Derrida, Jacques (1985) 'Des Tours de Babel', in Joseph Graham (ed) Differ-
ence in Translation, Ithaca, New York: Cornell University Press.
Ginsburg, Jane C. (1990) 'Creation and Commercial Value: Copyright Protec-
tion of Works of Information', Columbia Law Review 90: 1865-938.
Glenny, Michael (1983) 'Professional Prospects', Times Literary Supplement, 14
October, p. 1118.
Goldstein, Paul (1983) 'Derivative Rights and Derivative Works in Copyright',
Journal of the Copyright Society of the U.SA. 30: 209-52.
Grannis, Chandler B. (1991) 'Balancing the Books, 1990', Publishers Weekly, 5
July, pp. 21-23.
Greenhouse, Linda (1994) 'Ruling on Rap Song, High Court Frees Parody from
Copyright Law', The New York Times, 8 March, pp. AI, A18.
Jaszi, Peter (1994) 'On the Author Effect: Contemporary Copyright and Collec-
tive Creativity', in Martha Woodmansee and Peter Jaszi (eds) The
Construction of Authorship: Textual Appropriation in Law and Literature,
Durham, North Carolina, and London: Duke University Press.
Kaplan, Benjamin (1967) An Unhurried View of Copyright, New York and
London: Columbia University Press.
Keeley, Edmund (1990) 'The Commerce of Translation', PEN American Center
Newsletter 73: 10-12.
Locke, John (1960) Two Treatises of Government, edited by Peter Laslett,
Cambridge: Cambridge University Press.
Parks, Stephen (ed) (1975) The Literary Property Debate: Six Tracts, 1764-
1774, New York and London: Garland.
Rose, Mark (1993) Authors and Owners: The Invention of Copyright, Cam-
bridge, Massachusetts, and London: Harvard University Press.
Saunders, David (1992) Authorship and Copyright, London and New York:
Routledge.
Skone James, E. P., John F. Mummery, J. E. Rayner James and K. M. Garnett
(1991) Copinger and Skone James on Copyright, 13th ed., London: Sweet
and Maxwell.
Stewart, Susan (1991) Crimes of Writing: Problems in the Containment of Rep-
resentation, New York and Oxford: Oxford University Press.
Venuti, Lawrence (1985/86) 'The Ideology of the Individual in Anglo-American
Criticism: The Example of Coleridge and Eliot', Boundary 2 14: 161-93.
------ (1995) The Translator's Invisibility: A History of Translation, London and

New York: Routledge.
Woodmansee, Martha (1984) 'The Genius and the Copyright: Economic and
Legal Conditions of the Emergence of the Author', Eighteenth-Century
Studies 14: 425-48.
This article was downloaded by: [University Library Utrecht]
On: 15 March 2015, At: 18:07
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK
The Translator
The Moral Dilemmas of Court

Interpreting
a
Ruth Morris
a
Bar-Ilan University, Israel
To cite this article: Ruth Morris (1995) The Moral Dilemmas of Court Interpreting, The
Translator, 1:1, 25-46, DOI: 10.1080/13556509.1995.10798948
Downloaded by [University Library Utrecht] at 18:07 15 March 2015
The Moral Dilemmas of Court Interpreting
RUTH MORRIS
Bar-Ilan University, Israel
Abstract. In court interpreting, the law distinguishes between the

prescribed activity of what it considers translation - defined as an
objective, mechanistic, transparent process in which the interpreter
acts as a mere conduit of words - and the proscribed activity of
interpretation, which involves interpreters decoding and attempting

to convey their understanding of speaker meanings and intentions.
This article discusses the practicability of this cut-and-dried legal
distinction between translation and interpretation and speculates on
the reasons for its existence. An attempt is made to illustrate some of
the moral dilemmas that confront court interpreters, and an argu-
ment is put forward for a more realistic understanding of their role
and a major improvement in their professional status; as recognized
professionals, court interpreters can more readily assume the lati-
tude they need in order to ensure effective communication in the
courtroom.
Among members of the linguistic professions, the terms interpretation

and interpreting are often used interchangeably to refer to the oral trans-
fer of meaning between languages, as opposed to translation, which is
reserved for the written exercise. Interpretation, however, becomes a
potentially charged and ambiguous term in the judicial context, where it
refers to a specific judicial process. This process is performed intralingually,
in the language of the relevant legal system, and effected in accordance
with a number of rules and presumptions for determining the 'true' mean-
ing of a written document. Hence the need to adopt a rigorous distinction
between interpreting as an interlingual process and interpretation as the
act of conveying one's understanding of meanings and intentions within
the same language in order to avoid misunderstanding in the judicial
context.
Morris (1993a) discusses the attitude of members of the legal com-
munity to the activities and status of court interpreters, with particular
reference to English-speaking countries. The discussion is based on an
extensive survey of both historical and modern English-language law
ISSN 1355-6509 © St. Jerome Publishing, Manchester

The Moral Dilemmas of Court Interpreting 26
reports of cases in which issues of interlingual interpreting were ad-

dressed explicitly. The comments in these reports record the beliefs,
attitudes and arguments of legal practitioners, mainly lawyers and judges,
at different periods in history and in various jurisdictions. By and large,
they reflect negative judicial views of the interpreting process and of
those who perform it, in the traduttore traditore tradition, spanning the
gamut from annoyance to venom, with almost no understanding of the
linguistic issues and dilemmas involved. Legal practitioners, whose own
performance, like that of translators and interpreters, relies on the effec-
tive use and manipulation of language, were found to deny interpreters
the same latitude in understanding and expressing concepts that they
themselves enjoy. Thus they firmly state that, when rendering meaning
from one language to another, court interpreters are not to interpret - this
being an activity which only lawyers are to perform, but to translate - a
term which is defined, sometimes expressly and sometimes by implica-
tion, as rendering the speaker's words verbatim.
When it comes to court interpreting, then, the law distinguishes be-
tween the prescribed activity of what it calls translation - defined as an
objective, mechanistic, transparent process in which the interpreter acts
as a mere conduit of words - and the proscribed activity of interpretation,
which involves interpreters decoding and attempting to convey their un-
derstanding of speaker meanings and intentions. In the latter case, the
interpreter is perceived as assuming an active role in the communication
process, something that is anathema to lawyers and judges. The law's
attitude to interpreters is at odds with the findings of current research in
communication which recognizes the importance of context in the effec-
tive exchange of messages: it simply does not allow interpreters to use
their discretion or act as mediators in the judicial process. The activity of
interpretation, as distinct from translation, is held by the law to be desir-
able and acceptable for jurists, but utterly inappropriate and prohibited
for court interpreters.
The law continues to proscribe precisely those aspects of the interpret-
ing process which enable it to be performed with greater accuracy because
they have two undesirable side effects from the legal point of view: one is
to highlight the interpreter's presence and contribution, the other is to
challenge and potentially undermine the performance of the judicial par-
ticipants in forensic activities.
1. Interpreting as a communicative process
The contemporary view of communication, of which interpreting is but

one particularly salient form, sees all linguistic acts of communication as
Ruth Morris 27
involving (or indeed, as being tantamount to) acts of translation, whether

or not they involve different linguistic systems. Similarly, modern trans-
lation theorists see all interlingual translation as being essentially
communicative in nature. Seen from this complementary perspective, the
distinction between communication and language that the law seeks to
make in respect of interpreting becomes untenable. The law's insistence
that the interpreter is not a communicator within the judicial process thus
becomes logically unsound, as does its premise that court interpreters
should not put their own interpretation on speakers' words.
A further issue concerns the unreliable nature of the communication
process in general. As a subset of communicative activity, interpreting is
inevitably subject to constraints on success due to a variety of physical

and psychological factors. It is further influenced by differences in cul-
tural and other conditioning factors which shape our thought patterns and
perceptions. In order to convey these aspects, the interpreter needs to
understand not only linguistic but also many other elements related to
speakers' and listeners' worlds of knowledge. When the law calls for
interpreters to restrict themselves to verbatim translation and prohibits
the use of techniques which go beyond the referential use of language, it
is making it impossible to achieve anything approaching the already unat-
tainable goal of 'true' communication.
A critical examination of written judicial statements on interpreters'
skills and performance shows that some members of the legal profession
have recently become aware of the range of issues involved in interpret-
ing. However, individual pronouncements, whether in case reports or
other types of legal documents, have rarely led to the adoption of the
requisite administrative action by judicial and political authorities. The
ongoing New Jersey Interpreter Project (New Jersey Supreme Court Task
Force Report 1985; Tayler 1990; Tayler et al. 1989) is an outstanding
example of this state of affairs: a vast amount of valuable research led to
the formulation of precise policies, but these ran into difficulties on the
level of political implementation. Another typical instance is the British
Crown Colony of Hong Kong (Norton-Kyshe 1898). Right from the colo-
ny's beginnings in the mid-nineteenth century, the need for properly trained
interpreters was identified and a training policy proposed but never im-
plemented. Frequent press reports decried the resulting negative impact
on the doing of justice, but the authorities took no steps to remedy the
situation.
Judicial opinion as reflected in case reports from a variety of jurisdic-
tions is found to vary widely in respect of attitudes to interpreters. On the
whole, the dominant view of the interpreting process is of something
which is performed in a mechanical fashion by a transparent presence.
The Moral Dilemmas a/Court Interpreting 28
Much recent work, including my own, adopts an opposing view, one

which sees interpreting - in court as elsewhere - as a particular form of
communication in which performance of the activity is grounded in a
judicious sensing of speaker meaning. This is then conveyed in a form in
which latitude to depart from a verbatim standard may, and frequently
must, be taken in order to convey what the interpreter judges to be the
speaker's intention, and not merely the speaker's words. Apart from seek-
ing to identify and understand the intentions of speakers, interpreters may
at times also see fit to act in certain ways precisely in order to come closer
to the vital goal of achieving enhanced accuracy in their performance.
Such behaviour may draw attention to their role in the judicial proceed-
ings and may, potentially, be perceived as critical of the functioning of

judicial participants. It is thus not normally acceptable for an interpreter
to point out to an examining lawyer that, for cultural reasons, a particular
form of questioning is either impossible to render in the target language
or would be understood erroneously by the non-English speaker, or to
explain the cultural implications of the witness's reply. A notable excep-
tion to this rule is to be found in the 1820 proceedings against Queen
Caroline of England, where the interpreters were frequently asked to
clarify connotations in foreign-language material.! Even the possibility of
such active behaviour by the interpreter is ignored or precluded under the
distinction, widely accepted in English-speaking legal circles, between
interlingual interpreting and intralingual interpretation. The verbatim per-
formance expected of court interpreters, an activity which lawyers call
'translation', is treated as a mere technical adjunct to proceedings, which
are not seen to be affected by it in any way. Modern studies (Berk-
Seligson 1990a, 1990b; Morris 1989, 1993a; Shlesinger 1991) demonstrate
the fallacy of such an approach, using numerous real-life examples. Judi-
cial frustrations with the essentially communicative difficulties that arise
in interpreting situations are often extended to the interpreters themselves,
assuming what at times become vitriolic forms. At times, representatives
of the legal systems take their frustrations, which arise from their depend-
ence on the intralingual interpretation process and, occasionally, from
incompetent and/or unethical interpreting performances, and transfer them
not only to the providers of those services, but also to their clients (nor-
mally witnesses and defendants).
Legal attitudes to court interpreting in a number of English-language
jurisdictions may also be explained in terms of juridical rejection of and
antipathy towards the alien element, that is the non-English-speaking
individual. The law's denigratory attitude to foreigners, and its related
distaste at having to deal with problems which arise from their presence
in the host country, exclude its making proper interpreting arrangements
Ruth Morris 29
for its dealings with them. In this way, its dire fears about defective
communication become self-fulfilling.
As I have discussed elsewhere (Morris 1993a, 1993b), whatever the
reasons for the low quality of court interpreting, the law overwhelmingly
ignores the legal implications of relying on what is inevitably a flawed
product when interpreting services are provided by unskilled, untrained
individuals, often deficient even in high-level skills in two languages, let
alone in interpreting skills as such. Despite this prevailing state of affairs,
the product of the interpreting process is almost always treated as a le-
gally valid equivalent of the original utterance. The words of an individual
who bears the misnomer of court interpreter, often modified by the even
more misleading term official, to all intents and purposes literally sup-
plant the words of the foreign-language speaker. Tape recordings of
non-English utterances produced in the courtroom hardly ever exist; writ-
ten transcripts are almost never provided. The alarming implications for
the doing of justice are rarely considered by the law.
1.1. The legal view of interpreting: sources of misconception and rea-

sons for upholding them
In his challenging discussion of justice and translation, White (1990:258)

observes that translation represents
an attempt to be oneself in relation to an always imperfectly known

and imperfectly knowable other who is entitled to a respect equal to
our own. It is ultimately a question of understanding and attitude:
recognizing, while we compose our text, its inadequacy as the repre-
sentation of another, and finding a way to express that recognition in
what we say. To put it differently, it means the perpetual acknowl-
edgement of the limits of our minds and languages, the sense that
they are bounded by the minds and languages of others. It is in these
ways that the activity I call 'translation' - making texts in response to
others while recognizing the impossibility of full comprehension or
reproduction - becomes a set of practices that can serve as an ethical
and political model and, beyond it, as a standard of justice.
The prevailing legal view of interlingual interpreting identified above

embodies the complete antithesis of the approach so eloquently expounded
by White. Its consequences are almost entirely ignored by jurists.
Instructions and guidelines provided to interpreters in a judicial set-
ting are anchored in an overall view of language which Robinson (1991)
identifies as romantic. This view insists on achieving perfect identity in
the translation process between the source and target texts or utterances,
between source and target languages. 2 Robinson (ibid:88) speculates that

the source of this approach may lie in the mystical tradition of Kabbalism, 3
where "absolute cosmic correspondence, translating sense-for-sense, word-
for-word, even letter-for-letter, was essential, or more than essential, crucial
(anything less meant doom and destruction)". An echo reverberates here
between the law's demand for a verbatim performance of interpreting
activities and the verbal ritual that had to be followed meticulously in the
oral pleadings of the early law. 4 The verbatim prescription that contempo-
rary legal practice seeks to impose on the interpreting process would thus
seem to derive its ethos directly from the oral tradition of the early law.
But whereas in bygone days the smallest deviation could nullify proceed-
ings, the law faces no such dangers today as a result of slips in interpreting:
the prescription of verbatim (and hence, presumably, reliable) renderings
is taken for the performance. In other words, the law assumes that no
deviation whatsoever is taking place.
This view of the interpreting process has a major pragmatic advantage
from the law's point of view. It enables the court to function effectively as
a monolingual setting, since the absolute verbatim requirement has been
laid down and will, it is presumed, be met. This may be defined as the
legal fiction that enables the court to hold that what is stated originally in
a foreign tongue can, after ostensibly being switched into the language of
the proceedings, continue to function, with few exceptions,S as an original
text. A two-level supposition is at work here: that, as a matter of general
principle, one language can be switched into another with no loss of
substance or form, and, furthermore, that a standard of absolute accuracy
will be achieved in a particular interpreting performance.
White (1990:253) suggests that the classical, positivist view of trans-
lation - that what is said in one language can be said in another - is the
result of a defective view of language generally, namely that language is a
code into which messages are encoded. In this view, language is con-
ceived of as transparent, which means that the device which performs
switching between one language and another can similarly act as a trans-
parent conduit or decoder through which messages can flow unimpeded
and undistorted from one code into another. The interlingual interpreter is
thus ideally viewed as a mere disembodied or mechanical presence which
can, to all intents and purposes, be ignored. The law may recognize that
the input is in a foreign language and the output is in the language of its
monolingual setting, but in practical terms it relates only to the output.
The process can thus be ignored and its outcome treated as identical with
its origin. 6
I have already explained that this view of the translation process runs
entirely counter to what is now generally accepted: that no rendering can
Ruth Morris 31
exactly replicate an original text or utterance, and that failure to repro-

duce an identical replica across the language barrier is inevitable. This
inevitable failure and treacherousness of the translation process is pre-
cisely what the law cannot allow itself to recognize, and for very good
reasons. The law has to displace and personalize the failure, attaching it to
the individuals whom it engages to pursue the unattainable Holy Grail of
translatory perfection that will enable it to ignore the differences that
exist between speakers of different languages. The principle is sanctified:
the instructions 'to translate truly' are given; any fault lies with the imper-
fections of the human performers, the translators or interpreters. No taint
can be attached to those who gave the instructions, who conduct the
proceedings, or who listen to the performance.

Apart from refusing to acknowledge inevitable failures in the transla-
tion process, the law cannot allow itself to recognize or accept a related
fact, namely that its own activities are also at risk in an inherently flawed
communication process. If the difficulties of interlingual interpreting are
inherent in all communication, then intralingual communication is also
inevitably flawed. But how can the law respond to this internal difficulty
without undermining its own authority? Once again, it has to try to con-
trol the difficulty by projecting it on to the interpreters themselves. It
prescribes the requisite behaviour, namely word-for-word translation, and
proscribes the forbidden activity of interpretation. It assumes that it can
dictate different rules for communication through interlingual interpret-
ing from those that apply in its own sphere, as if contingent
meaning-in-context were suspended in the interlingual exercise, being
reserved exclusively for judicial interpretation. Where there is no context
to an ambiguous phrase, interpreters are nevertheless expected to render
the ambiguity exactly in the target language. Where ambiguity is deliber-
ate and bound to a particular language, the law still expects it to be
reproduced and preserved in the interpreted utterance.
Lawyers pride themselves on their ability to manipulate language and
express themselves with precision; if they are not understood by those
who rely on interpreters to participate in the proceedings or by interpret-
ers themselves, the fault clearly lies with the latter, not with the lawyers.
To admit that an argument has not been made cogently, that a sentence
has not been completed, that a word has been misused, that a grammatical
construction has been flawed, that hesitation has been present, is to admit
to imprecision and imperfection. The mechanical, transparent provider of
interpreting services is not supposed to interrupt or comment on lawyers'
performances, for this can generate a negative impression of judicial func-
tioning. 7 Instead, in an unspoken and unwritten code of good practice,
interpreters, like court reporters, are expected to improve on such defects
in lawyers' and judges' performances. The verbatim prescription is sus-

pended in these cases. 8 In witness statements, by contrast, such
imprecisions are held to be sacred and the interpreter is not expected to
intervene. On the surface, therefore, the prescriptive approach appears
monolithic, but on closer examination the 'translate, don't interpret' ad-
monition turns out to be relativist in philosophical and practical terms.
For interpreters to be able to playa more active role in the courtroom,
they have to be able to adopt an interactional stance which takes account
of both speaker intention and listener understanding, they have to be able
to use their own strategies for identifying misunderstandings, elucidating
context, investigating intention, and clarifying meaning explicitly. These
are all activities which may force the law to admit and confront its own
uncertainties and inadequacies, an admission it can ill afford to make. In
this sense, the whole interpreting process contains within it the potential
to undermine the entire edifice of legal procedure. For an interpreter to
seek to clarify with a cross-examining lawyer a question that was deliber-
ately framed in an ambiguous fashion is to restrict the examiner's freedom.
To seek clarification of a badly phrased question is to draw attention to
sloppiness. To point out that a particular question cannot be rendered into
the witness's language in the particular form in which it was asked is
potentially to make the lawyer aware of his/her own deficiencies, or to
draw attention to the foreignness of the witness. Each strategy adopted in
the interpreting process entails a certain cost. Not to clarify means to
guess, to conjecture, to put the interpreter's own interpretation (or belief)
on what was said. For an interpreter to seek clarification means identify-
ing ambiguities and potentially querying time-honoured legal conventions.
Challenges to language use - to the 'languaging' which is the very es-
sence of the law - can shake its foundations. No wonder then that the
law's reaction to difficulties which are inherent in the interpreting process
frequently becomes personalized in vitriolic diatribes directed against all
court interpreters. No wonder also that interpreters are sometimes ex-
ploited in the tactical manoeuvres employed by lawyers, who take
advantage of current legal views of the interpreter's status and role in
court.
A few examples from the multilingual Demjanjuk war crimes trial,
(The State o/Israel v. Ivan John Demjanjuk, Criminal Case 373/86) might'
help illustrate the complexity and Catch 22 situation in which court in-
terpreters often find themselves. During his cross-examination, the
Ukrainian-speaking defendant, John Ivan Demjanjuk, was questioned in
Hebrew about a description he had given in his examination-in-chief of
the appearance of a Russian general. Back-translated into English, the
prosecutor said:
Ruth Morris 33
You (m.) said he was tall (tamir) - that's how you (f.) translated it,
didn't you?
The Ukrainian-Hebrew interpreter, who (rightly) thought that the second

part of the prosecutor's question was addressed to her in an attempt to
verify her rendering of the defendant's answer, began to comment in
response: "That's what I said ... ". The Presiding Judge, seeing a seem-
ingly personal exchange beginning between counsel for the prosecution
and an interpreter, immediately reprimanded both: the interpreter for (al-
legedly) making a personal contribution, and the lawyer for commenting
on interpreting-related matters, on which he should consult later with the

head of the interpreting team. Thus, a commendable attempt by a lawyer
to verify the precision of both his memory and the interpreter's rendering
was foiled by the presiding judge's determination to reduce the interpret-
ing process to a mechanical cut-and-dried product.
A further example illustrates the interpreter's dilemma in dealing with
unclear material. Rules imposed by the court forbade the court interpreter
to address questions of her own directly to a witness. Consequently, she
attempted on various occasions to draw the attention of the bench to her
uncertainty about the precise meaning of the English-language original
where an item or utterance had been difficult to hear or understand. How-
ever, such requests by the interpreter were largely ignored, except in
cases where the three members of the bench, who followed the testimony
directly, were themselves interested in a clarification. The upshot of this
practice was that the legally authoritative Hebrew language record of the
proceedings was flawed and incomplete, despite the interpreter's repeated
attempts at drawing the court's attention to potential deficiencies in her
renderings. The court interpreter was eventually forced to ignore the rules
and clarify material directly with the speaker, thereby improving the quality
of the record.
A similar dilemma occurred in the case of German-language testi-
mony which was rendered simultaneously into English, consecutively
into Hebrew, and (on relay from the Hebrew consecutive or English
simultaneous versions) into Ukrainian using the 'whispering' technique.
At times during his examination-in-chief, counsel for the prosecution
(who had understood the testimony in the original German) politely indi-
cated to the bench that his understanding of what the witness stated was
different from what was conveyed to the court through the court interpret-
er's Hebrew-language rendering. Such statements by the prosecution,
which embodied an implicit challenge to the official interpreted version,
were themselves interpreted simultaneously into English. The English-
speaking defence lawyer, who had to follow both German-language testi-

mony and Hebrew-language comments through English interpreting,
charged his opposite number with putting words into the witness's mouth.
Mter a number of such incidents, the English-language interpreters were
asked to produce their version on relay from the Hebrew consecutive
rendering, rather than directly from German. This unprofessional sugges-
tion was adamantly rejected by the interpreters. It is worth noting here
that all the corrections to the Hebrew renderings suggested by the pros-
ecutor were justified and were subsequently incorporated into the record
of the proceedings. The English interpreters' rendering always coincided
with the prosecutor's understanding of the German. Hence, the defence

lawyer's charge that the prosecutor was putting words into the witness's
mouth was a classic trial tactic, one which took advantage of the inherent
difficulties in the interpreting process and the readiness with which the
law is willing to dismiss them.
Slips of the tongue, as in the inadvertent confusion of two dates, are
often a source of further dilemmas for court interpreters. In the Demjanjuk
trial, for example, one lawyer referred to a triangle when the referent
(visible in a photograph) was clearly a rectangle. In this instance, the
interpreter corrected the error, which passed without comment. A more
delicate problem arose when an expert witness testifying in German inad-
vertently gave two incorrect dates in a single sentence: 1976 (instead of
1946), and 1978 (for 1948). In both instances, the English interpreter
substituted the correct date, which could readily be inferred from previ-
ous references. By replacing the wrong dates with the right ones, the
interpreter avoided the charge of incorrect interpreting, but she also in-
fringed the rule that interpreters should repeat all original material,
including errors. The second error, as it happens, was also corrected by
the examining lawyer, who understood the original-language testimony.
And since the English version produced by the interpreter did not contain
the second error, those who listened to it were confused by the prosecu-
tor's correction of what appeared to them to be a non-existent error.
In the following example, the interpreter gave both the original, erro-
neous date and the correction by the bench, and because of her delayed
delivery was forced to act more as a reporter than an interpreter:
Prosecutor (in Hebrew): When did the Russo-German War break out?
Witness (in Hebrew): 22.6.1921.
Bench (in Hebrew): '41.
English interpreter (late rendering, after bench): Witness says 1921;
Bench corrects: 1941.
Ruth Morris 35
At times the interpreter deliberately drew attention to the slip, indicating

its origin with the speaker, and also interjecting a comment as to pre-
sumed speaker intention. In the following example, where the reference is
to witnesses examined in Israel for a war crimes trial in the Federal
Republic of Germany, the interpreter again acted as a reporter, and also
added some editorial comment of her own:
Witness (in German): Some said they would not travel to Israel.
Interpreter: ... to Germany; witness says Israel, but it must be Ger-
many.
On one occasion, the court interpreter was reprimanded for indicating

what she presumed to be the intention of the German-speaking witness,
rather than her precise words:
Interpreter (in Hebrew): I imagine that she wanted to say the photo-
graph.
Bench (in Hebrew): Did she say document?
Interpreter (in Hebrew): She said document, yes.
The deliberate use by interpreters of techniques such as those illus-

trated in the above examples reflects an attempt to distance themselves
from what mayor may not be speaker error. In using those techniques,
they inevitably draw attention to themselves as individuals in their own
right and thus flout the legal authorities' implicit 'out of sight, out of
mind' policy towards court interpreters. On occasion, as Harris (1981:198)
reports in relation to another World War II war crimes trial, interpreters
feel obliged to specifically distance themselves from material that they
have uttered, such as a question asked through them. The interpreter
observed by Harris began each of her renderings with a formula identify-
ing the speaker whose particular words she was about to convey; this
allowed her, when a witness asked her why she was asking pointless
questions, to respond by saying that the questions were being asked by
the judge and the lawyers, not by herself. Conference interpreters some-
times use a similar technique by adding a comment such as says the
speaker, which is equivalent to the editorial [sic], to qualify what they
perceive as a statement which makes no sense or which is incorrect in
some way. Such distancing tactics clearly draw attention to the interpreter
as a participant in the communication process in his or her own right.
In the transparent view of language, an interpreter is a non-person. This is
the ideal conception for the law. Examining the issue in a philosophical
context relevant to our present concern, White (1990:259-60) asks whether
that inherently marginal figure, the interpreter, who lives "in the space
between two languages", can ever have a voice or identity of his or her
own. Going even further, he queries why it should not be possible for the
interpreter to actually enter one or another of these worlds, and "speak
with momentary, if qualified, confidence within it". The attitude of the
law, as reflected in law reports, does not seem to welcome the possibility
of interpreters speaking with their own voices, rather than as mere alter
egos.
In determining its attitude to interpreting, the law finds itself facing
uncomfortable implications about its own role. 9 Iflaw, as White suggests,
is a form of translation, then it is marred by the same lack of certainty and
perfection which is characteristic of the process of interpreting. 10 At its

best, according to this argument, the law acknowledges its own shortcom-
ings. Logically, at its worst, it denies them. In denying the inadequacies
of the interpreting process, the law seeks to deny its own inherent defi-
ciencies. A particular interpreted utterance may be acknowledged as
flawed, but it will be considered an aberration. Even if attention is drawn
to an accumulation of interpreting deficiencies, these will be attributed to
the failings of the approach adopted by the individuals responsible for the
product. The inherent fallibility of the interpreting process will be denied,
for to acknowledge it is to question the very nature of the law, itself a
system of translation. An interpreter who goes beyond the literal meaning
or verbatim, word-for-word norm is using his/her own subjective under-
standing, his/her own words, putting words into the witness's mouth - is
interpreting, not translating. And interpretation, as we have seen, is a
proscribed activity for court interpreters.
2. Court interpreting in action: the case for interpreters' latitude
The pitfalls and impracticability of the 'translate, do not interpret' pre-

scription outlined in the previous sections can be further illustrated by
reference to a number of extremely common expressions in the judicial
context: the English character reference, certificate of good conduct and
criminal record, and their counterparts in other languages.
In English-speaking countries, a character reference or certificate of
good conduct (also known as a reference, character, or testimonial) may
be issued by anyone who has known the individual in question in a formal
capacity, particularly as an employer. The French acte de bonne vie et
moeurs or certificat de moralite, and the Dutch verklaring van gedrag are
issued by the police, as is one version of the German Filhrungszeugnis.
The Dutch document is obtained by making an application to the local
town hall, which forwards it to the Ministry of Justice in The Hague.
Ruth Morris 37
The term criminal record is often rendered in French as easier

judieiaire, which is a misleading term insofar as every citizen of a coun-
try influenced by the Napoleonic Code has the equivalent of a easier
judicia ire, whether vierge (clean) or otherwise. Thus the question Does
he have a record? cannot be rendered by Est-ce qu'il a un easier
judicia ire ?, since for all citizens of these countries the answer to the
French rendering must be in the affirmative, giving an entirely false im-
plication in an English-language legal context. When a citizen of such a
country needs to prove that slhe has no criminal record, the Ministry of
Justice issues the equivalent of an extra it de easier judicia ire, which
indicates the absence of any prior convictions. This term is rendered in Le
Docte (1982) as "certificate of non-punishment or of penalties incurred",

showing clearly the absence of the institution in English-speaking coun-
tries and the vital modifying effect of context. The same dictionary renders
easier judicia ire as "(UK) convictions record; (US) record of prior con-
victions". A verbatim rendering of extra it de easier judicia ire would be
"extract of a legal/judicial compartment, locker, drawer or cabinet"; even
taking account of some of the connotations would still give the entirely
misleading "abstract (or excerpt or copy) of convictions (or police) record".
Only the pragmatic approach will work here, and it is therefore reason-
able to argue that the court interpreter should be given a wide degree of
latitude, which may include providing an explanation of the relevant pro-
cedure in order to site the document within its cultural context. In written
texts, the translator can always append a translator's note, but considera-
tion needs to be given by the judicial authorities to the correct approach to
be adopted by the court interpreter in this kind of context.
2.1. The issue of latitude
There would appear to be no way out of the conundrum: the interpreter

has to use his or her own words, and judgement, in rendering speaker
meaning. The issue of latitude may be examined in more detail by draw-
ing a comparison between court interpreters and another group of
professionals who work in a similar context and under similar restric-
tions, namely court reporters. This comparison should help bring into
sharper focus the issues facing the interpreter, who essentially does much
the same thing as the court reporter. Both attempt to transfer meaning
from one domain to another: the court reporter transfers meaning from an
oral to a written mode, and the court interpreter transfers it in an oral
mode only but across a foreign language divide.l1
A major dilemma facing both types of professional concerns the extent to
which the material being transferred may be edited. Strictly speaking,
editing by the court reporter or court interpreter in a judicial setting is

prohibited; yet even court reporters are not supposed, for example, to
include "false starts, stutters, uhms and ahs and other verbal tics" in
transcripts, unless the exclusion of such verbalizations "could change a
statement's meaning" Y The court reporter is therefore supposed to edit
these "verbal tics" out of the transcript, except in cases where they are
significant to the meaning of the statement. In order to follow such a rule,
reporters must clearly use their subjective judgment, and are governed
not by verbatim standards of performance but by the need to evaluate and
convey meaning. In court reporting, Walker (1988) identifies the issue of
editing as involving the tension between verbatimness and readability. 13
In a conference interpreting situation, the tension is less obvious, since

professional ethics on the whole require the interpreter to render the speak-
er's intended meaning (as identified by the interpreter) in as eloquent a
form as the speaker would probably have wished to achieve, rather than
to reproduce the often imperfect form of the original. For the interpreter
to do "somewhat better than the original" (Herbert 1952:62) is accepted
practice in a conference setting. In the courtroom, on the face of it, the
same practice would be taken to constitute highly unprofessional behav-
iour.
Like the court reporter, the court interpreter is confronted by difficul-
ties and dilemmas which are inherent in the activity being performed, as
well as by unattainable or undesirable prescriptions laid down by the
system. Walker's studies of the performance of verbatim court reporting
identify related problems encountered in attempting to comply with the
law's ostensibly objective standards of what to record and how to render
an oral event in a written medium. She ascribes these problems to basic
differences between the characteristics of and expectations about speech
and writing. 14
Most of the obstacles to the court reporter's understanding of speech
in the courtroom setting that Walker (1990:214-17) identifies and analyzes
are equally applicable to the performance of court interpreters: lack of
context, inability to hear, insufficient knowledge of linguistic code and
professional jargon, insufficient background knowledge, garbled or dia-
lectal delivery, overlapping or co-speech, and being discouraged by custom
from interrupting speakers for any reason. Walker's studies (1988, 1990)
show that, in addition to the discrepancies likely to result from the ensu-
ing problems, transcripts differ in innumerable ways from the unrealistic
verbatim standards prescribed by the system and are therefore not a 'true'
representation of the original spoken material. 15 Wolchover (1989:787-8),
for example, explains that
Ruth Morris 39
... even when judges stick to doing 'almost nothing' - 'The pattern
that cannot go wrong' of simply recapitulating the evidence in the
order in which it was given - they are still notoriously capable of
performing the role of prosecuting counsel. Thus, with a studied mix
of intonation, inflexion, timing, and movements of head, eyes and
hands, the determined judge can make his opinion of the facts known
in a way that will never find its way onto the official written tran-
script.
The parallel with the court interpreter's activities, although these are con-
fined to the oral sphere, is obvious. 16
Walker (1990) identifies the existence of significant appellate conse-

quences of variable court reporting customs and conventions, which the
system fails to recognize,17 just as it ignores the implications of the vagar-
ies of interpreting activities. She carried out a nationwide survey of court
reporting practices and literature in the United States and analyzed her
own experience as a court reporter. One of the things she discovered was
that the actual degree of editing performed by a court reporter was af-
fected by the speaker's ranking within seven categories, including sworn
or unsworn, educated or uneducated, expert or lay, and liked or disliked. 18
Unsurprisingly, sworn testimony was edited least; most editing was nor-
mally carried out on ungrammatical and similarly substandard material
from judges or lawyers, unless they fell into the 'unliked' category, in
which case verbatim reporting could be used to emphasize the speaker's
incompetence. She also notes that judges are aware of the potentially
disastrous impact of their words being recorded verbatim in a deliberate
adoption by a critical reporter of what she dubs the GIGO policy: Garbage
In, Garbage Out (ibid:229). My own findings from a detailed case study
of interpreting in the Demjanjuk war crimes trial (Morris 1989:33-34) are
similar to those identified by Walker with respect to the behaviour of
court reporters. I found, for instance, that interpreters at the Demjanjuk
trial refrained from editing the speech of testifying witnesses but tended
to edit lawyers' questions in order to improve intelligibility and effective-
ness of communication. Speakers in the Demjanjuk proceedings, like those
in most other contexts, regularly hesitated, failed to finish their sentences,
used incorrect grammar, suffered from slips of the tongue and so on.
Under particularly stressful situations, some tended to speak rapidly. In
addition, certain speakers suffered from speech defects, or were required
(and/or willing) to express themselves in a language of which they had an
imperfect command. In standard conference interpreting practice, inter-
preters normally compensate for foibles of this type as far as possible. But,
as Berk-Seligson (1990a:171) and Gonzales et al. (1991:17) confirm, the
prescriptive rules of court interpreting do not allow for such behaviour.
Strictly speaking, the rules require any errors in the original to be reflected
in the interpreted version, even at the risk of the interpreters themselves
sounding incompetent. In the Demjanjuk trial, however, the practice adopted
by the court interpreters varied considerably, both on a personal basis and
according to the status of the original speaker.
Shlesinger (1991:149) discusses the issue of latitude explicitly and
reports that interpreters at the Demjanjuk trial frequently discussed whether
or not they had a 'legal mandate' to accommodate their listeners by using
what she calls the explicitation technique to clarify culture-bound refer-
ents. She further demonstrates how the interpreters into Hebrew flouted
the so-called 'accuracy' pledge by routinely omitting recurring formulaic
expressions of deference and courtesy used by the American defence

lawyers, because a faithful reproduction of such formulae would have
had a farcical effect in Hebrew (ibid:151). Roberts (1981) similarly dis-
cusses the tension between the legal requirements and sociolinguistic
conventions which determine the performance of court interpreters.
2.2. The professional status of court interpreters
In most of the world's jurisdictions, court interpreting has not, on the

whole, attained professional status in terms of either recognition or per-
formance. Where, exceptionally, the interpreter holds some form of
certification or registration, this can affect the evaluation of specific be-
haviour as either professional or unacceptable. Such was the case in the
1988 Australian civil case of Gradidge v. Grace Bros. Pty. Ltd., where
the litigant was deaf and interpreting was performed between English and
sign language. The issue considered on appeal was whether the inter-
preter had been wrong in ignoring the judge's instruction to stop signing
at a particular phase in the proceedings. The appeal court discussed the
interpreter's behaviour and the judge's control over the courtroom gener-
ally, together with many other issues affecting the provision of court
interpreting. The comments of one of the appeal judges, Judge Samuels,
show an understanding of the frustrations experienced by judicial partici-
pants as a result of interpreters doing what they consider to be their duty:
I readily understand that there are some cases that [sic] the use of an
interpreter, particularly one as indefatigable as this one, might pro-
duce irritations and frictions which heighten the emotions which are
commonly to be tapped in most forensic procedures; but that is sim-
ply a matter which cannot be helped. A judge must resolve these
conflicts if they occur as well as he or she can. 19
Ruth Morris 41
Because the individual concerned in this case was on the government

panel of interpreters, Judge Samuels noted that she was governed by a
code of practice and could therefore be dealt with appropriately if she
were subsequently found to have been in breach of her professional re-
sponsibilities. 20 In the meantime, it was held that the goal of interpreting,
namely to put language-handicapped individuals in the same position as
their hearing or English-speaking peers, should not be artificially ham-
pered.
The following comments by Judge Samuels in Gradidge v. Grace
Bros. Pty. Ltd. contain important guidelines for the behaviour of all court
interpreters and the rights of all language-handicapped participants in
legal proceedings. Legal authorities everywhere should consider them

and draw appropriate conclusions:
The task of the interpreter ... is to remove any barriers which prevent
understanding or communication .... The task of an interpreter is not
restricted merely to passing on the questions when the party is giving
evidence; it must be extended also to appraising a party of what is
happening in the court and what procedures are being conducted at a
particular time. We are all aware that this is not uncommonly done
and sometimes a judge may have to ask an interpreter to speak a little
more quietly or remonstrate when altercations develop, as they some-
times do, between the interpreter and the party. All of these things,
when they occur, must be determined and dealt with by the trial
judge. I emphasize, however, that it is quite wrong to imagine that all
an interpreter is supposed to do is to translate questions for a person
in the witness-box. 21
The case of Gradidge demonstrates that professionally certified status

(commoner in the more regulated world of sign language interpreters than
in that of their foreign language counterparts) can help improve the legal
view of interpreters and their activities, irrespective of the actual level of
proficiency of particular interpreters. When interpreters are perceived as
professionals who are regulated and governed by a code of conduct, they
acquire a status which enables the law to recognize them as something
closer to officers of the court. The interpreting process then becomes a
less controversial and more technical activity. As soon as this happens,
the law's attacks on interpreters as intrusive, meddling outsiders become
less vitriolic. In Gradidge, Judge Samuels acknowledged that certain things
may take place between the client and the interpreter which affect foren-
sic procedures; yet the overriding consideration, that of fairness of
procedure, must prevail and the interpreting process must be allowed to
proceed. It is worth noting here that the silent nature of the sign-language
interpreting activities which were taking place in Gradidge proves con-

clusively that it is not the acoustic element of interpreting which disturbs
judicial figures but the mere fact that something is occurring in the court-
room which is beyond judicial control and, indeed, is likely to be beyond
the understanding of other participants in the judicial process.
3. Conclusion
Despite growing interest in the process of interlingual and intercultural

mediation which takes place in court and community interpreting, few
legal systems seem to acknowledge the delicacy of the interpreter's task
and the moral dilemmas which are inherent in performing it. A study of
case reports from a variety of English language jurisdictions shows that,
on the whole, members of the legal profession appear to be unaware of
these dilemmas and/or unwilling to consider their implications for the
doing of justice. The professionalization of court interpreting can go some
way towards improving the status of court interpreters, thus allowing
them to exercise the necessary latitude in dealing with the inherent diffi-
culties of their profession. On the other hand, the professionalization of
court interpreting can only be achieved if legal practitioners and judicial
authorities are willing to recognize the same need for latitude in interlingual
interpreting as in intralingual (legal) interpretation and to drop their un-
tenable insistence on verbatim translation in the courtroom.
RUTH MORRIS
30/6 Haportsim Street, 92541 Jerusalem, Israel. F2B104@ILAC.BIU.VM
Notes
1. Bill of Pains and Penalties against Queen Caroline (September 1820), Par-
liamentary Debates, New Series, Volume 3.
2. The achieving may be against all odds, as Robinson (1991:88-9) argues:
"Important things are at stake in this notion, so important as to override (for
the romantics and their heirs, at any rate) all practical, commonsensical
objections regarding its impossibility. So what if it is impossible? It has to
be done! It is not something we would sort of like to try to do; it is a
messianic imperative, a question of life or death for all humanity. The
translator is the romantic savior, charged with the task of undoing the
damage done at Babel". Emphasis in the original.
3. Kabbala (or cabbala) is an ancient Jewish mystical tradition, based on an
esoteric interpretation of the Hebrew Bible.
4. Mellinkoff (1963:41) elaborates on the need to use particular words - "not
Ruth Morris 43
words of inherent precise meaning, but magical words that could stir a
God or wreck a soul ... formula, part of a ritual. Its repetition in this exact
form - and in no other - would produce the desired effect".
5. The major exception is the area of libel, in which it has traditionally been
recognized that the "very words" must be given in the original language.
6. Robinson (1991:68) shows how the dichotomous view of the translator as
the saviour who achieves the impossible, a view which opposes salvation
to oppression in a messianic approach, leads to a success/fail mentality:
"Actually, the romantic ideal is word-for-word and sense-for-sense: the
Augustinian display of determined fortitude in submerging despair over the
impossibility of ever knowing or translating God's (or the SL writer's) total
meaning is here intensified into a powerful (although still always frus-
trated) messianic hope. Translation soon becomes an all-or-nothing affair,

either total meaning, total understanding, total liberation from oppression,
or total failure, total untranslatability".
7. Berk-Seligson (1990b:195) shows that interpreters' interruptions of law-
yers were perceived by mock jurors (particularly Hispanics) as showing the
lawyer to be less competent and intelligent.
8. Walker (1988), like Morris (1989), finds that lawyers, judges and other
similarly placed individuals prefer not to have their slips of the tongue
reproduced.
9. White (1990:81) suggests that "While lawyers can of course make the
mistake of thinking that their language is the only one, the pressures of the
law are against it, for the law is a constant linguistic competition. How to
characterize the facts nd the law, how to conceive of and feel about the
case, and what, therefore, to do about it, are the central questions for the
lawyer, who knows that her categories are those of argument and judgment,
not simple factual description. The terms of her language itself are always
arguable. The legal conversation must therefore proceed, if it is to proceed
well, with a kind of structural tentativeness about itself. The law, at its best,
is a system of translation that acknowledges its own inadequacies". Em-
phasis added.
10. White (1990:261-2) explains that "What is true for the lawyer is true for the
law as well: it is a discourse that mediates among virtually all the dis-
courses of our world, all ways of talking, and it does this not on the premise
that meaning can be translated from one discourse directly into a different
one, but the creation of texts that are new compositions. In this sense the
law (like the lawyer) is both central and marginal at once: it exists at the
edge of our discourses, outside all of them, structurally supplementary; yet
it is also the discourse of power in our official world .... It is crucial to its
democratic power that in the end it make sense not only to those who speak
the language of the law by profession but to the men and women of a jury:
the ultimate translation is into the ordinary language of the citizen".
11. Walker (1988:1-2) compares the activities of court reporters and court
interpreters as follows: "The most interesting similarity is not really that
you both deal with someone else's words, but that all those words must
pass through you, and in passing, become to some extent your words too,
shaped by your background, your knowledge, your experience, your beliefs
- your hearing".
12. Federal Guidelines for Transcribers, quoted in Walker (1988:21).
13. Walker (1988:17): "In the practice of their profession, the tug toward
verbatimness vies with readability, objectivity with interpretation, statutes
with common sense".
14. Walker (1990:206) explains that "The central task performed by court re-
porters ... is to transform an event from its spoken manifestation into a
written one, thus performing what some scholars say flatly is an impossible
operation: providing an equivalence in two different media" (see for in-
stance Catford 1965:53). Elsewhere (Morris 1993a:205-68), I have
discussed some of the dilemmas which face both court reporters and court
interpreters.
15. Walker's approach is clear from the title of one of her articles: 'The Verba-
tim Record: The Myth and the Reality' (Walker 1986).
16. When transcribed, interpreting clearly crosses the talk-to-type divide.
17. Walker (1990:242) asserts that "Without awareness, and without inquiry,
institution-based discrepancies will continue to be irregularly characteristic
of verbatim transcripts, and the customs and conventions of court reporters
will continue to carry unknown consequences for the appellate process and
those who enter into it".
18. The seven categories are: sworn or unsworn, educated or uneducated, ex-
pert or lay witness, ins or outs, employer or non-employer, liked or disliked,
and sees transcript or doesn't see transcript (Walker 1990:233).
19. Gradidge v. Grace Bros. Pty. Ltd. (1988) 93 FLR 414 at 427.
20. Gradidge v. Grace Bros. Pty. Ltd. (1988) 93 FLR 414 at 422: "The inter-
preter was a member of the Government panel. If the interpreter
misconducted herself, that would be a breach of her ethical and profes-
sional duties. It could be dealt with accordingly. It could even amount to a
criminal offence. It would warrant action against her to discipline her or
remove her from the list of Government interpreters".
21. Per Samuels J. A. in Gradidge v. Grace Bros. Pty. Ltd. (1988) 93 FLR 414
at 425-6. Emphasis added.
References
Berk-Seligson, Susan (1990a) 'Bilingual Court Proceedings: The Role of the

Court Interpreter', in J. N. Levi and Anne Graffam Walker (eds), Language
in the Judicial Process, New York and London: Plenum Press, 155-201.
------ (1990b) The Bilingual Courtroom: Court Interpreters in the Judicial Proc-
ess, Chicago: University of Chicago Press.
Catford, J. C. (1965)A Linguistic Theory of Translation, Oxford: Oxford Univer-
sity Press.
Gonzalez, R., V. Vasquez and H. Mikkelson (1991) Fundamentals of Court Inter-
pretation: Theory, Policy, and Practice, Durham: Carolina Academic Press.
Ruth Morris 45
Harris, Brian (1981) 'Observations on a Cause Celebre: Court Interpreting at

the Lischka Trial', in Roda Roberts (ed) L'interpretation aupres des
tribunaux, Ottawa: University of Ottawa Press, 189-202.
Herbert, Jean (1952) The Interpreter's Handbook: How to Become a Conference
Interpreter, Geneva: Librairie de l'Universite.
Le Docte, Edgard (1982) Legal Dictionary in Four Languages, Antwerp and
Apeldoorn: Maarten Kluwer.
Mellinkoff, David (1963) The Language of the Law, Boston & Toronto: Little,
Brown and Company.
Morris, Ruth (1989) The Impact of Court Interpretation on Legal Proceedings.
M.A. Thesis, Hebrew University of Jerusalem: Institute of Communications.
------ (1993a) Images of the Interpreter: A Study of Language-Switching in the
Legal Process. Ph.D. Thesis, Lancaster University: Department of Law.

------ (1993b) 'The Interlingual Interpreter - Cypher or Intelligent Participant?
Or, The Interpreter's Turn .. .', International Journal for the Semiotics of
Law 6(18): 271-9l.
New Jersey Supreme Court Task Force on Interpreter and Translation Services
(1985) Equal Access to the Courts for Linguistic Minorities: Final Report.
Norton-Kyshe, James William (1898; reprinted 1971) The History of the Laws
and Courts of Hong Kong from the Earliest Period to 1898, Hong Kong:
Vetch and Lee Ltd.
Roberts, Roda (ed) (1981) L'interpretation aupres des tribunaux, Ottawa: Uni-
versity of Ottawa Press.
Robinson, Douglas (1991) The Translator's Turn, Baltimore and London: John
Hopkins University Press.
Seleskovitch, Danica (1985) 'Interpretation ou interpretariat?', Meta 30(1): 19-24.
Shlesinger, Miriam (1991) 'Interpreter Latitude vs. Due Process. Simultaneous
and Consecutive Interpretation in Multilingual Trials', in Sonja Tirkkonen-
Condit (ed) Empirical Research in Translation and Intercultural Studies,
Tiibingen: Gunter Narr, 147-55.
Tayler, Marilyn R. (1990) 'Update from the New Jersey Department of Higher
Education Interpreter Education Project: The New Jersey Legal Interpreta-
tion Project', in A. Aguirre (ed), Proceedings of the Northeast Conference on
Legal Interpretation and Translation, New Jersey: The Consortium of Edu-
cators in Legal Interpretation and Translation, 31-35.
------, Roda P. Roberts and Ellie de la Bandera (1989) 'Legal Interpreter Educa-
tion: The New Jersey Legal Interpretation Project', in D. L. Hammond (ed)
ATA Proceedings of the 29th Annual Conference, Medford, N. J.: Learned
Information, 389-97.
Walker, Anne Graffam (1986) 'The Verbatim Record: The Myth and the Real-
ity', in S. Fisher and A. D. Todd (eds) Discourse and Institutional Authority:
Medicine, Education, and Law, Norwood, N. J.: Ablex, 205-22.
------ (1988) Court Reporting: Another Kind of Interpretation. Paper given at
UNITAS International Conference for Translators and Interpreters, Arlington.
------ (1990) 'Language at Work in the Law - The Customs, Conventions, and
Appellate Consequences of Court Reporting', in J. N. Levi and Anne Graffam
Walker (eds) Language in the Judicial Process, New York and London:
Plenum Press, 203-44.
White, James (1990) Justice as Translation: An Essay in Cultural and Legal
Criticism, Chicago and London: University of Chicago Press.
Wolchover, David (1989) 'Should Judges Sum up on the Facts?', Criminal Law
Review, 781.
On: 14 July 2015, At: 15:04
The Translator
Observations on Anomalous
Stress in Interpreting
a
Sarah Williams
a
Centre for Research on Bilingualism, Stockholm
University
To cite this article: Sarah Williams (1995) Observations on Anomalous Stress in

Interpreting, The Translator, 1:1, 47-64, DOI: 10.1080/13556509.1995.10798949
Observations on Anomalous Stress in Interpreting

SARAH WILLIAMS
Centre for Research on Bilingualism, Stockholm University
Abstract. A problematic phenomenon in simultaneous interpreting,

and one which is recognized by interpreters, teachers of interpreting
and conference delegates alike, is that of anomalous stress, i.e. when
the interpreter unexpectedly stresses the 'wrong' word. Since one of
the functions of stress in spoken language is to show coherence rela-
tions in a text, anomalous stress can lead to comprehension problems

for the listener. Examples of anomalous stress produced by a profes-
sional interpreter at a live conference were acoustically analyzed in
relation to stress patterns in the speaker's input. In the examples
studied, it was found that while the anomalous stress produced by
the interpreter did not appear to be directly related to semantic or
pragmatic features in the incoming message, it was preceded by
stressed elements in the input. Two possible mechanisms are tenta-
tively suggested as playing a role in the occurrence of anomalous
stress: firstly, an automatic matching mechanism triggered by sali-
ent stress from the input, which may anticipate forthcoming input
prosody or may trigger the most recently stored prosodic pattern,
and secondly, the independent storage of salient prosodic patterns
from the input. Initial findings of this pilot study indicate that anoma-
lous prosody may at least in some cases be a result of automatic
mechanisms beyond the interpreter's conscious control.
An irritating but widespread phenomenon in simultaneous interpreting,

and a problem recognized by interpreters, teachers of interpreting, and
conference delegates alike, is that of anomalous prosody. This can mani-
fest itself in various forms, such as monotonous intonation contours
(Altman 1989; Gran 1989; Kurz 1989), erratic rhythm (Andronikof 1962;
Barik 1972; Altman 1989), high volume output (Gran 1989; Spiller &
Bosatra 1989) and anomalous stress (Shlesinger 1994). While all aspects
of prosody are important in contributing in various ways to the successful
transmission of a message in spoken language, this paper will confine
itself to a discussion of anomalous stress.
Cruttenden (1986:16) defines stress as prominence due to either pitch,
length or loudness, or a combination of these, whereby stressed syllables

Anomalous Stress in Interpreting 48
are indicated most by pitch and least by loudness. Pitch concerns the vary-
ing height of the pitch of the voice over one syllable or over a number of
successive syllables: in other words, the voice goes up and down; length
concerns the relative durations of a number of successive syllables or the
duration of a given syllable in one environment relative to the duration of
the same syllable in another environment (i.e. when one syllable or word
is longer than others); loudness concerns changes of loudness within one
syllable or the relative loudness of a number of successive syllables, that
is it concerns changes in volume (ibid:2). Very generally speaking, stress
falls into two categories, which I will refer to as word-related stress and
discourse-related stress. Word-related stress can be delimitative, serv-
ing to mark beginnings and endings of words and thus aiding segmentation
of the stream of speech, or distinctive, distinguishing between two other-

wise identical-sounding words (ibid:18), such as, for example, black bird
and blackbird. Discourse-related stress has various functions (Bolinger
1986, 1989; Cruttenden 1986): grammatical functions, such as topicalization
and parenthesis; discourse and coherence functions, largely to do with
how a particular piece of information fits into the discourse as a whole,
such as focus, given versus new information, contrastivity; pragmatic func-
tions, i.e. what is contributed to the communicative situation in general,
in terms of speech acts, and speaker's attitude and emotional state. Anoma-
lous stress is when the stress on an element is unexpected in the context
and creates a misrendering of the original message, even creating stress
patterns that are atypical of the target language.
In simultaneous interpreting, it is possible that the occurrence of anoma-
lous stress creates for the listener problems of coherence which either
momentarily or permanently impair comprehension of the text. In the case
of momentarily impaired comprehension of the text, modification of ini-
tially miscomprehended coherence relations in the light of following
information (i.e. back-tracking) will require extra effort on the part of the
listener; conversely, extra effort will also be required if listener expecta-
tions created by the preceding text are suddenly not realized (Cutler 1987).
In both cases, the flow of comprehension is momentarily interrupted and
listener concentration may suffer. In the case of permanently impaired
comprehension of the text, the miscomprehension is never rectified, with
the result that there then exists a discrepancy between the speaker's in-
tended message and the message received by the listener. In a study by
Shlesinger (1994) showing, among other things, that anomalous stress is
a feature of interpreted text, the level of listener comprehension and recall
of information was found to be lower in subjects who listened to inter-
preted texts than in those who listened to the same texts transcribed and
read aloud by the same interpreters.
Sarah Williams 49
That anomalous stress occurs in interpreting, and that it may well have
a detrimental effect on listener comprehension, then, is clear. What is not
so clear is why and when this phenomenon occurs. When listening to the
recorded performance of a professional interpreter at a conference, I no-
ticed that anomalous stress in the interpreter's production (which I refer
to here as output) sometimes appeared to be immediately preceded by
particularly salient stress in the speaker's production or input. The stress
in the output and the input however did not seem to be semantically or
pragmatically related, such that, roughly speaking, anomalous stress oc-
curring in output Al seemed to be a reflection of speaker stress occurring
in B, as can be seen in Figure 1. In other words, the speaker first produced
Sentence A, and then the interpreter rendered Sentence A in the target
language (AI) while listening to the speaker's next sentence, Sentence B.

In some cases, it seemed as though the interpreter reacted to the stressed
word in Sentence B by inadvertently producing salient stress. The prob-
lem was that when this happened, the interpreter was still producing
Sentence A l , so that the stress produced seemed to be unrelated to the
sentence being uttered by the interpreter.
SPEAKER:
[!]~
----~~----------
--------~~~~~~--~~~
INTERPRETER:
GJ~~
Figure 1: Model of input-related anomalous stress
1. Material and method
In order to see whether these perceptual findings could be corroborated

with objective evidence, an acoustic analysis was carried out using a dou-
ble track recording of interpreted material from a one-day conference at
which the working languages were English and Swedish.
Excerpts containing examples of perceived anomalous stress preceded
by salient stress in the input were taken from the recorded material. These
excerpts were then run through a computer program, SOUNDSWELL
(Ternstrom 1992), and a printout was obtained for the speech signal, sound
pressure level and fundamental frequency (F0) on both tracks, i.e. both the
speaker's and the interpreter's production. Intensity and fundamental
frequency are the acoustic equivalents of loudness and pitch, respectively.

Fundamental frequency (F0) is obtained by filtering out as many other
frequencies as possible, namely those produced by overtones (making up
the sounds characteristic of the particular quality of the speaker's voice)
and background noise. 1 In order to make the presentation of the results
more accessible, only the Fo graphs, which show the intonation contours,
are given here. The complete graphs for speech signal, sound pressure
level and Fo appear separately in the Appendix. In all of these graphs, the
length of the syllables can be seen by looking at the horizontal axis, which
gives time in seconds.
The four excerpts analyzed in section 2 below were taken from the
first fifteen minutes of the interpreter's performance and were selected
because they provide clear examples of the phenomenon in question,

namely input-related anomalous stress. 2 I should point out, however, that
there were also instances of salient stress in the input which were not fol-
lowed by anomalous stress in the output. Other instances were found where
the interpreter's anomalous stress may have been influenced by other fac-
tors such as, for example, relief at finally having found the word sought
after; such instances do not provide clear examples of input-related anoma-
lous stress. In what follows, then, the suggestion that anomalous stress
may be input-related is put forward as one of several possible factors oc-
casioning anomalous stress.
There is a noticeable discrepancy in terms of quality between the
speaker's and the interpreter's recording, which is due to the fact that re-
cording conditions for the interpreter were more favourable (soundproof
booth, little background noise, little physical movement towards and away
from the microphone). This is why the Fo readings for the speaker are not
as clear as those for the interpreter; the background noise produced fuzz
which was difficult to filter out without filtering out some of the speech
signals. A related problem was the actual quality of the speaker's voice,
which contained a lot of overtones that were difficult to filter out. 3 De-
spite all this, it is still possible to identify the higher F0 of the stressed
elements in the speaker's utterances.
Each example is first given in a rough transcription, with various fill-
ers such as er and mm simply rendered as er. All stressed items appear in
italics; those which are relevant to the discussion are in bold italics. In-
stances of anomalous-stress in the interpreter's production appear in
italicized bold capitals. An attempt has been made to indicate roughly
where the stressed elements occur in relation to each other by using the
convention of displaced parallel text. Each example is followed by an
acoustic analysis in which only the relevant stressed elements are given
as text. 4 It is here that the exact timing of the elements can be seen, as the
Sarah Williams 51
time scales for both speaker's track and interpreter's track have been syn-
chronized.
2. Results
In all the examples analyzed here, the interpreter was listening to a new
message at the same time as producing a version of the previous message.
In each case, the anomalous stress produced by the interpreter while inter-
preting the previous message was immediately preceded by salient stress
in the input that the interpreter was listening to. Since the stress in the
input belonged to a new message that the interpreter had not yet rendered
in the target language, it was neither semantically nor pragmatically re-
lated to the word which the interpreter actually stressed. Hence the
occurrence of instances of anomalous stress in the interpreter's output.
Example 1
Speaker: somnar monopolisten forr eller senare in ................................ och sussar gott .. .
Interpreter: doing this ....................... and .... therefore ...... the person who has a monopoly
..... och levererar daIig service till ett hogt pris .... Detta ar ett ....... er ................. ett
will more or less go to sleep ............. and ..... will deliver very bad services to a very
................... empiriskt och ........................... vetenskap-

high PRICE .................... And .......... this is an empir-
Gloss: ... the monopolist will sooner or later go to sleep and sleep well and
deliver bad service at a high price. This is an empirical and scientif- ....
In this example, the anomalous stress is on the word price. The correspond-
ing word in the original,pris, was unstressed, and there is no discourse-related
reason for stressing price. The anomalous stress follows a stressed item in
the input, detta, with a time lag of approximately 2.802 seconds (See
Graphs la and Ib).5
Example 2
Speaker: avknoppat foretag med foredetta ......... kommunala- och lanstingsanstallda

Interpreter: we know .. that .. today .. a company ..... that ... has .. started ............... .
kvinnor som har blivit iildrevard dom ldg tju/em procent battre an stans enheter
... er ...... on .... the .. side ......... for instance WOMEN who've been employed in the ge-
nar processen ....... .

municipal geria- .. .
Gloss: ... companies with women who had previously been employed by
the municipality and local councils which have now become old folks'
homes they were twenty-five per cent better off than the city's institutions
when the process ...
'" 300
,=~250
:;: r- 200 della
[-;.
i5 ~ 150
:;:'"
2 g 100
z:~
~ ~ 50
price
2 4
TI~IE IN SECONDS
Graphs la and lb. Upper Section: Speaker stress on detta

Lower Section: Interpreter stress on PRICE
",300
~250
<:>
tjufem
~~ 200
domJag
i5 ~ 150
:;:'"
2
z:~
g 100
~ ~ 50
TIME IN SECONDS
Graphs 2a and 2b. Upper Section: Speaker stress on dom ldg tjufem
Lower Section: Interpreter stress on WOMEN
Sarah Williams 53
The anomalous stress in Example 2 is on women. The corresponding word

in the original, kvinnor, was unstressed, and there is no discourse-related
reason why women should be stressed. Women appears to occur almost
simultaneously with the stress in the input, dom lIlg tjufem. However, on
closer inspection it can be seen that the beginning of the stress pattern in
the input immediately precedes the anomalous stress pattern in the out-
put. The two syllables in the Swedish input,dom lIlg, not only have higher
pitch but also show other signals that indicate stress, namely in the form
of increasing loudness. The prosodic changes in these two syllables, which,
rhythmically, happen to co-occur exactly with the syllables being pro-
duced by the interpreter, may have triggered the anomalous stress in the
output, such that the next syllable produced by the interpreter reflected
the stress in the input. It seems likely that the interpreter's third syllable
would have occurred simultaneously with the speaker's next syllable had
the speaker not produced a syllable group, tjufem, with an unstressed syl-
lable before the main stressed syllable (see Figure 2).
Swedish input dom Hig tiu fern

Time in seconds 5.358 5.509 5.622 5.848
Time in seconds 5.358 5.509 5.759
English output for instance women
Figure 2: Simultaneous occurrence of stress in speaker input and

interpreter output over several syllables
Example 3
Speaker: ... pa ...... sig .... for att komma fran ............. ett ............. typ .... .
Interpreter: .. one who looks at ....... a ......... five year period and one looks
av system till ett annat .... er ............ sa er .... har man nu i fOrvaltningarna
from .. one ............................ er ... from one system, getting from one system
.... mycket ambitiOst tycker jag och de fiesta har varit mycket bra gatt igenom vad iir det
to another one one has decided IN the administration in a very ...... er ... ambitious way ... .
nu som kan konkurrensutsiittas och .. .

.... one has gone through .................. .
Gloss: ... in order to change over from one type of system to another, they
have now, in the administration, I think very ambitiously, and most of
them have been very good, looked in detail at what can be exposed to
competition and ...
TIME DI SECONDS
Graphs 3a and 3b. Upper Section: Speaker stress on mycket ambitiost

Lower Section: Interpreter stress on IN
Here, the anomalous stress is on the word in. The corresponding word in
the input, i, was unstressed, and there is no discourse-related reason for
stressing in. The anomalous stress follows a stressed item in the input,
mycket ambitiOst, with a time lag of approximately 0.79 seconds.
Example 4
Speaker: ... egen utsaga ligger de inte mer an atta tio procent battre beroende
Interpreter: ... they'll allow .. twenty-five .. per cent higher ........ but now ...... .
pa .. att ... de .................... kommunala enheterna (xxx) har nu b6rjat tillampa en del
they're ....... only ............... about ....... er ... eight .................... ten ...... per cent better
av deras arbetsmetoder och fiiljaktligen sankt sina kostnader och fiiljaktligen ar inte
.............. only ........................ than .. the ...... municipal .. sector .... it's because the
skillnaden lika star ........... langre mellan den privata ........... och den egna
municipal sector have FOLLOWED them and lowered their costs .............. .
egen energiverksamhet .. .
........ er ....................... .
Sarah Williams 55
Gloss: ... their own statement they are no more than eight or ten per cent
better, as the municipal institutions have now begun to apply some of
their methods and consequently lowered their costs and consequently the
difference is no longer as great between the private and their own con-
cerns.
In this example, the intonation of the item carrying anomalous stress in the
output sounded perceptually very similar to that of an item stressed earlier in
the input, rather than to the intonation of the immediately preceding item.
'N' 300
C 250 I
~ 1Z 200
1-;..-
slor
~:;::'"~ 150
~ g 100
z'"
f:: e: 50
'N'
C 400 followed
...l<::
~ ~ 300
Zu
~ ~ 200
<=-
co
!§ ~ 100
......
2 4
TIME 1:\ SECO:-;DS
Graphs 4a and 4b. Upper Section: Speaker stress on kommunala and star
Lower Section: Interpreter stress on FOLLOWED
This may indicate the existence of two mechanisms. The anomalous stress
on followed sounds perceptually very much like the intonation in the in-
put kommunala, which occurs quite a bit earlier than followed. However,
it will be noted that followed is immediately preceded by rising pitch and
increase in breath pressure in the input star (see Graphs 4c and 4d in the
Appendix). This may trigger anomalous stress in much the same way as
was suggested in discussing Example 3, except that the stress pattern pro-
duced in this case does not seem to be based on an anticipation pattern set
up by star, but is perceived to reflect the stress pattern from the earlier
input kommunala. The time lag between kommunala and followed is ap-
proximately 6.832 seconds; the time lag between star and followed is
0.341 seconds.
3. Discussion
First, it must be pointed out that any discussion based on a limited number
of examples, as in this case, must necessarily be somewhat speculative.
However, the examples analyzed above suggest that while the anomalous
stress in the interpreter's output does not directly correspond, semanti-
cally or pragmatically, to stressed elements in the input, it does appear to
reflect or be preceded by salient stress in the input, although exactly how,
why and when this happens is still unclear. Indeed, the analysis indicates
that there may be several different mechanisms involved, which mayor
may not be related. In Example 1, anomalous stress is produced slightly
after salient stress in the input; in Example 2, it is produced immediately
after the onset of salient stress in the input and may in fact anticipate
forthcoming salient stress, whereby increasing loudness may playa role;
in Example 3, the anomalous stress in the output sounds similar to that
which has just occurred in the input, and in Example 4, it appears to be
immediately preceded by salient stress in the input but sounds similar to
salient stress that occurred considerably earlier in the input.
Clearly, much more work needs to be carried out before anything more
definite can be said. However, two areas of research may offer some in-
sight into this phenomenon, namely the tendency, in certain circumstances,
for a person to adapt his/her pitch to that of the person s/he is talking to
(referred to here as Fo mirroring) and the tendency to use the sound of
one's own voice to automatically monitor and adjust subsequent speech
production (referred to here as proprioceptive audial control).
It has been noted in the literature that identification with the speaker is
an important factor in conference interpreting. Cary (1962:5), for exam-
ple, suggests that "the essence of conference interpretation as compared
with other forms of translation is not so much a scrupulous analysis of a
text, but rather an immediate identification with the speaker" (my transla-
tion;6 see also Andronikof 1962 and Garcia-Landa 1985). Ironically, this
identification with the speaker, which is felt to provide a basis for good
interpreting, may be precisely what leads to the kind of anomalous stress
that is detrimental to listener comprehension (possibly via Fo mirroring).
In a study in which pairs of subjects met for the first time, were recorded
in conversation and afterwards independently reported their feelings to-
wards the other person, Buder (1993) was able to show that in cases in
which there was a mutual feeling of affinity, there was also a tendency for
rhythmic and F0 matching to take place; in other words, subjects adapted
their intonation to that of the other person. Indeed, there is reason to be-
lieve that Fomatching may actually be innate; infants' Fo in babbling is
higher in the presence of the female caretaker and lower in the presence of
Sarah Williams 57
the male caretaker (Weir 1962). Fo matching may also constitute part of
the more global matching mechanism seen in reciprocal body language in
dyads that is well documented in research on interpersonal behaviour (for
instance in Morris 1977).
A related problem may be the perceptual mixing of the speaker's
prosody and the interpreter's own prosody. Spiller & Bosatra (1989) dis-
cuss various problems involved in co-ordinating the perception of the
speaker's input and the interpreter's own input, and point out that the
audiophonatory reflex is very important in monitoring output; delayed
audio feedback, for example, can cause stuttering, and the higher volume
input sometimes favoured by students can also interfere with this
audiophonatory reflex. Gran (1989:95) also mentions that when students
at the initial stage of their training turn up the volume for fear of missing
part of the original speech, they then tend to speak "in a very loud voice,
in a flat and monotonous tone, while making syntactic and pronunciation
errors because of auditive interference with the source language" and dis-
play "an almost total lack of control of their output in the TL". If, perhaps
because of energy and attention constraints, the interpreter is (momentar-
ily) unable to keep the two incoming channels separate, i.e. the speaker's
voice and hislher own, then the speaker's input at that point may act as
audial control for the interpreter's output, creating an automatic adjust-
ment. It may, then, be the ability to block this posited automatic matching
mechanism that students have to acquire, and it may be this blocking that
sometimes fails to operate in professional interpreters under certain cir-
cumstances.
It is by now perhaps clear that this study raises more questions than it
answers, such as, for example:
• When and why does input-related anomalous stress occur?

• What is the relationship between pitch, duration and intensity in in-
put and output?
• How many mechanisms are involved in input-related anomalous
stress, e.g. mirroring, anticipation, triggering, memory of original
prosodic pattern?
• Is there a distinctive time-lag for input-related anomalous stress?
• Does this phenomenon occur in all interpreters, or just some?
• Does it occur with equal frequency in women and men?
• How does anomalous stress in the output of professional interpreters
compare with that in the output of student interpreters?
• What is the relationship between anomalous stress, discourse-related
stress and word-related stress? Does anomalous stress take precedence
only over discourse-related stress or might it in some cases be so

strong that it also overrides word-related stress?
• Is the occurrence of anomalous stress the same in the first and sec-
ond languages (L1 and L2)?
• Might further work in this area provide support for the hitherto cred-
ible but subjective view 7 that sign language interpreters are easier to
listen to and understand than other interpreters because of a greater
degree of separateness between input and output modes?
Ultimately, further research must pay close attention to detail, in terms of

exact measurements of pitch, duration and intensity, and must also be
carried out on a far larger corpus of material, including more subjects and
more language combinations. Unfortunately, analyses of the kind presented

here are extremely time-consuming, which consequently sets certain
limitations on the size of a realistic corpus. However, since the results
presented in this paper would support the finding that subjective perception
of pitch is relatively reliable (see Kitzing 1979), another way for researchers
to approach this issue is to carry out detailed subjective analyses of data
by listening to tapes of conference interpreting on double-track tape-
recorders and noting instances of perceived input-related anomalous
prosody. Although this would preclude the more exact kind of measurements
presented here, such work would nevertheless be extremely valuable in
providing an insight into the extent and nature of this phenomenon. Ideally,
this kind of research should be monitored occasionally by means of
acoustic analyses on random samples as a way of checking reliability.
There appears to be a tacit assumption in the literature that prosody is
something over which the interpreter has conscious control; students are
encouraged to adopt a lively rather than monotonous intonation pattern
(Altman 1989; Kurz 1989), and Alexieva (1990:5), for example, refers to
"optimum prosodic variant(s)" and "felicitous choice of prosody". While
it is essential to draw attention to the vital role of prosody in interpreting,
it is also important to bear in mind that prosodic features may be influ-
enced by a variety of factors. The preliminary findings presented in this
paper indicate that anomalous prosody, at least in some cases, may be the
result of automatic mechanisms beyond the interpreter's conscious con-
trol.
SARAH WILLIAMS
Centre for Research on Bilingualism, Stockholm University, S-J069J
Stockholm, Sweden. Sarah. Williams@biling.su.se
Sarah Williams 59
Notes
Grateful thanks go to the staff at the Department of Logopedics and Phoniatrics,

Karolinska Institute, Huddinge University Hospital, and in particular to Alf
Hakansson for his generous help with the technical side of this research.
1. The amplitude of the speech signal is measured in volts (V), the sound
pressure level is measured in decibels (dB), and the fundamental frequency
is measured in Hertz (Hz).
2. It would obviously be extremely helpful to know more about the frequency
and distribution of this phenomenon, induding how many instances of
anomalous stress occurred in the first fifteen minutes. Although such an
extensive investigation is beyond the scope of this particular paper, it is
my intention to carry out more detailed research in this area in the near
future.
3. In the graphs that follow, the upper curve Fo extraction is not perfect; peaks
above 250-300 Hz probably reflect the second harmonic rather than Fo.
4. Because of technical reasons, the text given in the graphs is an approxi-
mation of where the stress occurs; in order for the text to fit exactly over
the graph lines, either it would have had to have been reduced to an illeg-
ible size, or I would have had to enlarge the graph itself to such an extent
that it would not have been possible to present this length of excerpt.
5. The time lag was measured from the beginning of the rise in Fo in the
stressed element in the input to the beginning ofthe rise in Fo in the stressed
element in the output.
6. ''l'originalite essentielle de !'interpretation de conference par rapport aux
divers genres de traduction consiste en ce qu'elle suppose non une ana-
lyse minutieuse d'un texte, mais !'identification instantanee a un homme
qui parle".
7. Personal communication, Anna-Lena Nilsson, sign language interpreter,
Stockholm University.
References
Alexieva, Bistra (1990) 'Creativity in Simultaneous Interpretation', Babel 36(1 ):

1-6.
Altman, Janet (1989) 'The Role of the Tutor Demonstration in Teaching Inter-
preting', in Laura Gran and John Dodds (eds) The Theoretical and Practical
Aspects of Teaching Conference Interpretation, Udine: Campanotto Editore,
237-40.
Andronikof, Constantin (1962) 'Servitudes et grandeur de !'interprete', Babel
8(1): 8-12.
Argyle, Michael (1972, 1977) The Psychology of Interpersonal Behaviour,
Harmondsworth, Middlesex: Penguin Books.
Barik, Henri (1972) 'Interpreters Talk a Lot, Among Other Things', Babe118(1):
3-9.
Bolinger, Dwight (1986) Intonation and Its Parts: Melody in Spoken English,
Stanford: Stanford University Press.
------ (1989) Intonation and Its Uses: Melody in Grammar and Discourse, Lon-
don: Edward Arnold.
Buder, Eugene (1993) Synchrony of Speech Rhythms in Conversations. Paper
presented at the Department of Linguistics, Stockholm University, 17th
March, 1993.
Cary, E. (1962) 'Noblesse de la parole', Babel 8(1): 3-7.
Cruttenden, Alan (1986) Intonation, Cambridge: Cambridge University Press.
Cutler, Anne (1987) 'Speaking for Listening', in A. Allport, D. G. Mackey, W.
Orinsz and E. Scheerer (eds), Language Perception and Production, Lon-
don: Academic Press.

Garcia-Landa, M. (1985) 'L'oralite de la traduction orale', Meta 30(1): 30-36.
Gran, Laura (1989) 'Interdisciplinary Research on Cerebral Asymmetries: Sig-
nificance and Prospects for the Teaching of Interpreting', in Laura Gran and
John Dodds (eds) The Theoretical and Practical Aspects of Teaching Con-
ference Interpretation, Udine: Campanotto Editore, 93-100.
Kitzing, Peter (1979) Glottografisk frekvensindikering. Doctoral dissertation,
Medicinska Fakulteten vid Lunds universitet, Malmo: Litos Reprotryck.
Kurz, Ingrid (1989) 'The Use of Video Tapes in Consecutive and Simultaneous
Interpretation Training', in Laura Gran & John Dodds (eds) The Theoretical
and Practical Aspects of Teaching Conference Interpretation, Udine:
Campanotto Editore, 213-16.
Morris, Desmond (1977) Manwatching, London: Elsevier.
Shlesinger, Miriam (1994) 'Intonation in the Production and Perception of Si-
multaneous Interpretation', in Sylvie Lambert and Barbara Moser-Mercer (eds)
Bridging the Gap: Empirical Research on Simultaneous Interpretation, Am-
sterdam and Philadelphia: John Benjamins, 225-36.
Spiller, Edith and Andrea Bosatra (1989) 'Role of the Auditory Sensory Modal-
ity in Simultaneous Interpretation', in Laura Gran and John Dodds (eds) The
Theoretical and Practical Aspects of Teaching Conference Interpretation,
Udine: Campanotto Editore, 37-38.
Temstrom, Sten (1992) SOUNDSWELL - Signal Workstation Software. Version
3.06. Manual, Stockholm: Soundswell Music Acoustics.
Weir, Ruth (1962) Language in the Crib, The Hague: Mouton.
Sarah Williams 61
Appendix
4
2
~
'"'" 0
;;,
.'"'
t:
:E
«
-2
-4
-10
-20
'N'
!; 250
:if- ~> 200 delta
~~ 150
Ci g 100
~~ 50
TIME IN SECONDS
Graph 1 c. Speaker stress on detta
4
2
~
.,'"'" 0
..
f-
:l
:E
«
-2
-4
TI~IE f.\ SECONDS
Graph ld. Interpreter stress on PRICE

4
:> 2
;:;-
Q
0 ~~~~~~-.--~~~~~~
E -2
""~ -4
.."
--~-------------------------
'" -10
~ -20
If~g ~,~~~~~~~~
iI
I
o 2345678
TIME 1:'0 SECO:'ODS
Graph 2c. Speaker stress on dam ldg tjufem
4
:> 2
i O~~~~~____~*M~"~~.
~0:: -2
~ -4
women
I I I ! I
o 1 2 3 4 567 8
TI~IE 1:'0 SECO:'ODS
Graph 2d. Interpreter stress on WOMEN

Sarah Williams 63
4
2
~
'"Q
0
~ -2
.
-'
..: -4
;;
T1~IE I~ SECONDS
Graph 3e. Speaker stress on mycket ambitiOst
4
2
~
'"
Q
0
~
.
:;
;;
..:
-2
-4
-10
-20
o 4 12 14
T1~IE 1:\ SECO~DS
Graph 3d. Interpreter stress on IN

4
2
~
"'
Q
O~~. .~~~~~~~~~~~~~~~
E
.
...J
:;;
«
-2
-4
--~-----------------------------------
2 4 6 8 10 12 14
TIME L'I SECONDS
Graph 4c. Speaker stress on kommunala and stor
TIME (:\' SECONDS
Graph 4d. Interpreter stress on FOLLOWED

This article was downloaded by: [University of Bristol]
On: 29 January 2015, At: 07:24
UK
The Translator
A Descriptive Framework for

Compensation
a
Keith Harvey
a
London, UK
To cite this article: Keith Harvey (1995) A Descriptive Framework for Compensation,
The Translator, 1:1, 65-86, DOI: 10.1080/13556509.1995.10798950
Downloaded by [University of Bristol] at 07:24 29 January 2015
A Descriptive Framework for Compensation
KEITH HARVEY
London, UK
Abstract. A detailed description of compensation as a translation

strategy can be of considerable help to professional translators by
highlighting some of the options available to them. More importantly,
such a description can also facilitate the pedagogical presentation of
the concept. This paper starts with an overview of the various treat-
ments of the concept in the literature before proceeding to elaborate

a new descriptive framework for compensation along three axes:
typological, linguistic correspondence, and topographical. The dis-
cussion is illustrated throughout by examples taken from authentic
translations, including examples from the famous cartoon series
Asterix.
Compensation is a loosely used and poorly defined concept in much of the

theoretical literature, as well as in the kind of practical workshops which
are frequently organized by professional institutions. Passing references
to the concept are often tagged on to general discussions of non-literal
translation procedures, and authentic examples of it are rarely given in the
literature. And yet, in recent years scholars of translation have increas-
ingly had recourse, however brief, to the concept. What is more, they
attribute to it a powerful range of application. Baker (1992:78), for exam-
ple, invokes compensation as a technique for dealing with "any loss of
meaning, emotional force, or stylistic effect which may not be possible to
reproduce directly at a given point in the target text". And Newmark
(1991:143) claims that "compensation is the procedure which in the last
resort ensures that translation is possible".
If compensation is as important as these statements suggest, then there
is a clear need for a descriptive framework that can account for the pro-
cesses it entails. Among other things, such a framework would serve a
useful pedagogical function: trainee translators stand to gain from a more
explicit statement of the means by which they can make up for effects in
their source text that cannot be directly transferred to a target text. This
paper is an attempt to set out a descriptive framework for compensation,
using examples from the famous strip cartoon series Asterix to elaborate
the concept in action. The text I have chosen is Les Lauriers de Cesar
(Goscinny and Uderzo 1972), translated as Asterix and the Laurel Wreath
ISSN 1355-6509 © St Jerome Publishing, Manchester

A Descriptive Framework for Compensation 66
by Anthea Bell and Derek Hockridge (1974). I have chosen Asterix be-
cause much of the humour of this particular series is created by linguistic
jokes of many kinds, including puns, adapted or misquoted idioms, errors
in performance such as spoonerisms, and so on. And compensation as a
technique is of course at its most active in the translator's attempts to deal
with just such devices.
Let me begin with a working definition of the notion of compensation.
I define compensation as:
a technique for making up for the loss of a source text effect by

recreating a similar effect in the target text through means that are
specific to the target language and/or the target text.
Many of the terms in this definition will come under scrutiny as we

proceed. However, it is worth noting from the outset that the use of the
word technique has an important implication. Although the examples that
I will discuss will necessarily be the result of a retrospective critical
analysis of translated products, compensation is first and foremost a tech-
nique available to translators engaged in the process of transferring
meanings and effects across linguistic boundaries. Hence, its importance
in translation pedagogy is possibly greater than its significance as a de-
scriptive category in translation criticism.
Puns are often cited as the privileged site for compensation. For in-
stance, in a reference to the Asterix series, Hatim and Mason (1990:202)
note that "the translators abandon the attempt to relay the puns as such
and, instead, compensate by inserting English puns of their own which
are not part of the source text". A simple example of a compensated pun
occurs on page 14 of Asterix and the Laurel Wreath. 1 At a busy slave
market in Rome, each merchant is shouting his wares on the public place.
In the source text, a merchant cries:
Suivez mes Thraces! Suivez mes Thraces!

(Literally: Follow my Thracians! Follow my Thracians!, with a pun
on the expression suivez mes traces, or follow in my footsteps.)
The target text translates this by using its own pun:
Heavy-duty nimble Hoplites!
The humorous effect of the source text is lost and then recreated by
different means in the target text (Hoplites and its homophone lights
contrasting paradoxically with Heavy-duty). It is worth noting that in this
Keith Harvey 67
example loss and compensation occur in the same place. Also, the same
linguistic device, a pun, is employed in the target text to create a similar
effect.
1. Compensation: a problem of definition
Explicit references to compensation are scattered throughout the litera-

ture on translation studies. These references often represent piecemeal,
non-formalized uses of the term. Particularly in texts dating from before
the mid-1980s, words such as compensation, compensatory and compen-
sate for are usually employed in a loose, common-sense way. Close
examination of examples reveals that practically anything that did not

involve straightforward formal correspondence was subsumed under this
label.
Nida (1964) and Nida and Taber (1969) contain some representative,
early examples of loose uses of the term. In their discussion of "semantic
adjustments made in transfer", Nida and Taber (ibid:105-12) argue for
the priority of content over form in a hierarchy of translation processes.
Except in cases like poetry, form can be sacrificed in translation. Con-
versely, "in any translation there will be a type of 'loss' of semantic
content, but the process should be so designed as to keep this to a mini-
mum" (ibid: 106). In the subsequent discussion of this minimization of
semantic loss, they suggest types of compensatory strategy. For example,
not only do they approve of the translation of source text idioms by target
text idioms, they also argue for the strategic translation of non-idiomatic
forms in the source text by target language idioms: "Such idiomatic ren-
derings do much to make the translation come alive, for it is by means of
such distinctive expressions that the message can speak meaningfully to
people in terms of their own lives and behaviour" (ibid). In a footnote on
the same page, we find a passing reference to compensation:
What one must give up in order to communicate effectively can,

however, be compensated for, at least in part, by the introduction of
fitting idioms.
If this is not done, they warn,
the end result is a weakening of the figurative force of the translation,

since they [the translators] do not compensate for loss of certain idi-
oms by the introduction of others. (ibid)
Examples of this generalized type of compensation with the use of idi-

omatic expressions are found inAsterix and the Laurel Wreath. On page 42,
for example, the source text plays with the source language idiom y faire
de vieux os (literally: make old bones there, meaning 'grow old there')
when the cut-throat Habeascorpus introduces Asterix and Obelix into the
catacombs (a series of underground passages where bodies used to be
buried in Ancient Rome):
Notre repaire: les catacombes ... c'est un refuge sur; on peut y faire de
vieux os.
This is translated closely as

Here's our hide-out: the catacombs. It's quite safe. You'll make old
bones down here.
The target text continues in this punning vein with a play on both senses
of skeleton and on the homophone skull in skulduggery:
Tomorrow night we'll leave a skeleton staff here, and you can try
your hand at skulduggery ...
The source text passage contains no puns or notable stylistic effects to

account for this. Nor does there appear to be any loss in the vicinity which
would explain or justify the target text's inventiveness. This must there-
fore either be an example of displaced compensation (discussed under 3.3
below) or, more probably, of Nida and Taber's generalized type.
Nida and Taber's discussion of idiomatic expressions situates com-
pensation firmly in the domain of stylistic effects, that is effects that are
deliberately employed in a text for specific purposes of register, tone and
colour. Their comments also suggest that compensation does not neces-
sarily involve systematic, one-to-one correspondence of individual source
text and target text effects. The 'equivalence' of specific effects is sec-
ondary to the achievement of a balance of tone across the entire text. By
implication, the text as a whole is highlighted as the unit with which the
translator ultimately works.
1.1. Compensation and cultural mismatch
Early on in his book The Science of Translation, Wilss (1982:39) asks

"How should he [the translator] compensate micro contextually and
macrocontextually during the transfer process for structural divergences
on the intra- and extralinguistic level?". Behind the abstract terms of this
question there seems to be a familiar preoccupation with the problem of
Keith Harvey 69
the unit of translation. The use of 'macrocontextually' acknowledges that

the unit to be worked with is sometimes situated above the levels of word,
clause or sentence.
In the same passage, Wilss points out that "the range of practicable
compensation strategies in any instance varies from language pair to lan-
guage pair" (ibid). This claim clearly ties compensation to the problem of
a systemic mismatch between source and target languages. In other words,
it suggests that compensation arises as a consequence of the mismatch
between the two language systems under consideration and is conditioned
by the limits of those systems. Wilss therefore seems to downplay the
stylistic, text-specific nature of compensation. But he later extends the
notion of compensation to encompass cultural elements as well (ibid:50):
Cultural untranslatibility occurs when sociocultural factors cover a

different range of experience in the SL and the TL and must be made
to coincide in regard to the intended meaning in each instance ....
Translation procedures generally present possibilities for compensa-
tion, since translation is in principle possible whenever the transfer is
preceded by an understanding of the content of the original text.
Thus, Wilss mentions instances where "a lexical by-pass strategy such as
paraphrasing or explanatory translation" is "the only compensatory way
out open to the translator" (ibid: 104).
So far, then, compensation seems to have been explained as a tech-
nique for introducing idioms and as a cover term for paraphrases and
explanations. The latter might be the necessary consequence either of a
systemic mismatch between languages or of the cultural differences that
render practices in source and target cultures mutually opaque. However,
if we are to succeed in establishing compensation as a useful descriptive
category, we need to beware of using it to cover too many problems in the
process of transfer. While stylistic, text-specific devices seem to fall com-
fortably within its remit, the larger issues of the mismatch between social
and cultural practices go well beyond it and threaten to make the concept
too general to be of any pedagogical use or theoretical value.
This is not to suggest that linguistic devices used to achieve compen-
sation do not sometimes have a culture-specific component. They do in
the following example from Les Lauriers de Cesar, for instance. The joke
in this example (on page 9) is at the expense of Homeopatix's maid, a
short, stout woman with black hair and large earrings. She announces that
dinner is served in the following manner:
La Matrone elle est serbie

The appropriate realization of this conventional expression is Madame est

servie (literally: Madam is served). Although the maid's use of Matrone
(a pejorative word in modem French for a matronly or stout woman)
together with the excessively informal repetition of the pronoun after the
noun (La Matrone eUe ... ) is comic in itself, it is the tell-tale confusion of
the French phonemes Ivl and Ibl that confirms what the maid's appear-
ance might already have suggested to a member of the source culture: the
maid is Spanish. It is because the cultural aspect here is mediated through
a linguistic device that it becomes available for compensation.
The maid's nationality sets off particular cultural resonances in France,
where a whole generation of working-class Spanish immigrants appeared
between the wars. Jokes about this immigrant population's misuse of the
French language, and in particular the phonemes Ivl and Ibl, are still current
in France. Such associations are of course unavailable to a member of the
target culture. The target text's translation
Cena is served
takes the appropriate functional equivalent in the target language (Dinner

is served) and adjusts it to accommodate a linguistic joke. However, the
latter is obscure and depends on a knowledge of modem Spanish or of
Latin (cena means supper or evening meal in both). Such knowledge
cannot be assumed on the part of members of the target culture. How,
then, is this loss compensated for in the target text?
Just before the maid appears, there is an exchange about Homeopatix's
house wine where considerable differences emerge between source and
target texts. In the source text, Galantine, Abraracourcix's wife, makes
the following straightforward comment:
l' espere que vous aimerez ce vin, il est fait avec Ie raisin de notre
vigne, sur la butte ...
(Literally: I hope you will like this wine, it is made with the grape
from our vine, on the hillock ... )
The target text elaborates on this with linguistic material that is culturally
marked for a member of the target culture as a pastiche of the language of
a wine connoisseur (which may itself be linked, more or less consciously,
to the cultural habits of a non-target language speaker):
Try some of the 55 B.C., from our own vineyard. It's a modest,
unpretentious little wine, but I hope you like it.
Keith Harvey 71
Soon after this, the target text transforms another simple source text
remark with a significant piece of intertextuality for a member of the
target culture. Where the source text has Homeopatix commenting:
On ne peut vivre qu'a Lutece, tu sais.

(Literally: One can only live in Lutetia, you know.)
the target text has a pastiche of Dr Johnson:
When a man is tired of Lutetia, he is tired of life.

It is just after this sequence that the maid announces dinner and, at least in
the source text, a piece of cultural stereotyping is put to amusing affect. It
would appear, therefore, that the justification for the various target text
elaborations which draw on cultural and intertextual knowledge is that
they compensate for the loss of the joke over the maid's heavy Spanish
accent. Note, however, that the culture-specific loss to be compensated is
manifested in linguistic terms.
1.2. Compensation: location and equivalence
InA Textbook of Translation, Newmark (1988:90) suggests that compen-

sation "is said to occur when loss of meaning, sound-effect, metaphor or
pragmatic effect in one part of a sentence is compensated in another part,
or in a contiguous sentence". Not unreasonably, therefore, he defines
compensation as a response to translation loss. However, this in itself is
not entirely unproblematic. If loss is an inevitable consequence of any
attempt to transfer sense from one language to another, and compensation
is a response to that loss, then the entire translation process could be
accounted for by the twin mechanisms of loss and compensation. Once
this happens, of course, the floodgates are open and both loss and com-
pensation get washed away as useful descriptive terms.
To counter this danger, we must specify what types of transfer be-
tween source and target texts are susceptible to compensatory strategies. I
have already stated that I believe it is important to retain the term for
essentially stylistic, text-specific features and effects. The weakness in
Newmark's definition is that it does not make this emphasis clear enough
and suggests that compensation might cover systemic, language-specific
features as well. In particular, where his definition refers to 'meaning',
are we to understand that it includes those problems inherent in the trans-
lation of lexical meaning, where the target language fails to find a
one-to-one equivalent for a source language item? The problem in cases

of this type is surely systemic and not stylistic: it is a consequence of the
fact that different languages divide the semantic space in different ways.
Newmark's definition also raises the issue of the location of compen-
sation in relation to the instance of loss. The assertion that the loss "is
compensated for in another part, or in a contiguous sentence" rules out -
unnecessarily, in my view - two possibilities for compensation. First, it
excludes those instances of what I call paraDel compensation, such as we
have already seen in the example of Suivez mes ThraceslHeavy-duty nimble
Hoplites. Here, the target text manifests simultaneous loss (of, say, the
source text pun) and compensation, the latter effectively overwriting the
former. Second, this unnecessary restriction fails to account for compen-

sation over much greater distances than simply contiguous sentences,
what I have called displaced compensation in 3.3 below.
Unlike Newmark, Hatim and Mason (1990) make no mention of the
possible location of compensation in relation to loss. Compensation is
glossed as "the making good of some communicative loss by substituting
equivalent effects" (ibid:239). It would therefore seem that they consider
compensation to be characteristically unbound. Similarly, there is no speci-
fication here of the types of linguistic feature that it operates upon.
The reference to effects in Hatim and Mason's definition reinforces
the general thrust of their communicative approach to translation. How-
ever, this approach begs a number of fundamental questions. The notion
of 'equivalent' in "equivalent effects" is problematic. It is not clear from
the definition whether the equivalence of effect is necessarily bound to an
equivalence of textual device. For example, if humour is the textual effect
that has been lost from the source text as the result of the untranslatibility
of a pun, should this loss be compensated for in the target text by the same
type of linguistic feature? This is left unspecified, although examples
abound where the question of type of linguistic device is subordinated to
that of effect. One such example occurs on page 43 of Asterix and the
Laurel Wreath. The source text achieves a humorous effect through the
neologism couiquer (a verb) from the conventional onomatopoeia couic
(for squeak). The cut-throat Habeascorpus uses couic, together with an
appropriate gesture of the index finger running across the throat, as a
euphemism for the murder of an approaching passer-by:
Et s'il resiste ... Couic!

(Literally: And if he resists .,. Squeak!)
A perturbed Obelix asks Asterix:

Keith Harvey 73
Dis, Asterix, on ne va tout de meme pas Ie couiquer?

(Literally: Say, Asterix, we are not really going to squeak him?)
The target text compensates for the loss of this sequence of an onomato-
poeia followed by a neologism by using a colloquial idiom:
- If he makes a fuss ... the chop!

- We're not really going to give him the chop are we Asterix?
The device employed is different, but the intended effect is arguably

comparable.
But how does one measure equivalent effect? Gutt (1991) puts the
point rather well in his discussion of a target text that fails to reproduce
the effect of flattery of its readers' cultural knowledge that is achieved by
a source text in relation to its own readership. Suggesting at first that the
translator should apply the technique of compensation and strive to obtain
the effect of flattery by other means, Gutt (ibid:48) soon recognizes the
difficulties inherent in this solution:
Does he [the translator] do so by checking whether his translation

flatters the receptor language audience in corresponding parts of the
texts, or by making sure that the number of instances of flattery that
occur is equal between original and translation, or by some compari-
son of the cumulative flattering effect of the whole text?
In other words, Gutt concludes that there is no empirical basis for the
equivalent effect argument other than the translator's own reactions to the
texts slhe is reading (source) and writing (target). Effect turns out to be a
function of the reader's own motivation for reading a text, and even of the
various conventions that determine response in different cultures, rather
than the inherent property of a particular text.
2. A descriptive framework for compensation
In a book that is as much practical handbook for English-French transla-

tion problems and procedures as it is a theoretical overview, Hervey and
Higgins (1992:34-40) set out the longest and most detailed breakdown of
compensation in the literature. They introduce the concept of compensa-
tion as a response to the inevitable compromise that translation involves.
However, the translator is not entirely powerless when confronted with
such losses:
It is when faced with apparently inevitable, yet unacceptable,

compromises that translators may feel the need to resort to tech-
niques referred to as compensation - that is, techniques of making
up for the loss of important ST features through replicating ST
effects approximately in the TT by means other than those used
in the ST. (ibid:35)
Hervey and Higgins distinguish four categories of compensation: com-

pensation in kind, where different linguistic devices are employed in the
target text from those in the source text in order to re-create a similar
effect; compensation in place, where the effect in the target text is achieved
at a different place from that in the source; compensation by merging,
where source text features are condensed in the target text; compensation
by splitting, where meanings expressed in the source text have to be
expanded into a longer stretch of text in the translation. Although they
point out that the four types of compensation can co-occur, the last two
categories, 'by merging' and 'by splitting', would appear to be mutually
exclusive. I will deal with each type briefly, using Hervey and Higgins'
own examples, before I proceed to offer an alternative framework in
section 3 below.
To illustrate compensation in kind, Hervey and Higgins discuss a
French narrative which achieves a strong stylistic effect through the inter-
play of past historic and perfect tenses. The text (ibid:35) recounts the life
of a young fighter in the French Resistance:
Quelques jours apres la Liberation, on retrouva son corps dans un

charnier. Elle a ete fusillee Ie 8 juillet 1944 a l'age de 23 ans.
Elle fut une militante exemplaire.
In this account of the fighter's death, the ordinary use of the past historic
tense for narrative (retrouva, Jut) is interrupted by the appearance of the
perfect (a ete Jusillee) to convey shock and immediacy. The English tense
system cannot reproduce the effects obtained by this interplay of tenses.
Consequently, for the last two sentences Hervey and Higgins suggest the
following translation (ibid:36):
This girl was shot on 8 July 1944, at the age of 23.

She was an exemplary resistante.
Here, the demonstrative This, the noun girl rather than the pronoun Elle in
the source text, the strategic placing of the rhetorical comma after 1944,
Keith Harvey 75
and the cultural borrowing of the term resistante all contribute to com-
pensating for the loss of "the emotional impact of the ST's play on
tenses" (ibid:36).
This example is highly significant for our purposes because it illus-
trates that stylistic potential is present in some systemic features of a
language. Another example in French would be the stylistic implications
of the presence in the language system of the pronominal forms of famili-
arity and politeness, commonly known as the TN distinction (see Hatim
and Mason 1990:28; Baker 1992:96-98; Hervey and Higgins 1992:36). In
other words, these examples show that our understanding of stylistic ef-
fects must include those that are not unique to a particular text. Such
effects are inscribed within a particular linguistic system and can be acti-
vated for stylistic purposes within a specific text.
Another point worth making here is that all the examples of compen-
sation in kind given by Hervey and Higgins are of an essentially parallel
type. Despite their assertion that compensation in kind and compensation
in place can hypothetically co-occur, if an instance of compensation em-
ploying very different linguistic devices were to be located a long way
from a particular loss, it would be extremely difficult to identify and
prove the relation between them.
Under compensation in place, Hervey and Higgins (ibid:37) include
compensation "for an untranslatable pun in the ST by using a pun on
another word at a different place in the TT". Another example (ibid:38) is
located at a level that we have not so far considered, yet one at which
stylistic, text-specific effects are typically found, namely the use of sound
for rhetorical effect:
Voila ce que veulent dire les viriles acclamations de nos villes et de

nos villages, purges enfin de I' ennemi.
The target text (ibid) compensates for the inevitable loss of the sequence
of alliteration and assonance by exploiting a different sequence of sounds
in the corresponding sentence:
This is what the cheering means, resounding through our towns and
villages cleansed at last of the enemy.
The phonetic reinforcement here does indeed make use of sounds which
are different from those used in the source text, but only incidentally does
it take place on different 'equivalent' words. The source text's veulent
dire, for example, which contributes two elements of phonetic reinforce-
ment to the chain, finds its standard translation equivalent, means, also
contributing an instance in the target text. We are thus left with the im-
pression that more long-range examples of compensation in place would
have enabled Hervey and Higgins to bring home the possibilities of this
type more forcefully.
Hervey and Higgins' last two types of compensation, by merging and
by splitting, are presented as complementary procedures. Whereas any
problems that arose with the 'in kind' and 'in place' types were a conse-
quence of the quality of the examples chosen to illustrate them, with these
last two categories I believe that serious doubts emerge as to their validity
as compensation at all.
Hervey and Higgins give two examples for each category. To illus-
trate compensation by merging, they point to a phrase from a technical
source text, "en cas de feux et incendies", for which they propose the
target text "in the event of fire" (ibid:38). They justify their choice by
suggesting that "in a technical text like this one, a translation such as
'little fires or big fires/ones', or 'fires and conflagrations/blazes' would
be comically inappropriate or unidiomatic" (ibid:39). However, the rea-
son for this inappropriateness is not to be found exclusively in the text
genre, as they suggest. Rather, it is a consequence of the difference be-
tween two lexical systems. In other words, the French lexical universe
has inscribed within it the distinction between fires of different sizes and
degrees of gravity in a way that the lexical universe of English has not.
The target language alternatives given, conflagrations and blazes, do, it is
true, suggest larger fires, but their connotations in English are more emo-
tive and, thus, stylistically more marked than the everyday incendies in
French. These qualities are part of the systemic meanings of these items
in English.
Compensation by splitting is considered necessary "in cases where
there is no single TL word that covers the same range of meaning as a
given ST word" (ibid:39). As an example, Hervey and Higgins discuss
the 'splitting' of the French papillons into butterflies and moths in the
English title of an article on lepidoptera. But again, the choice here is
hardly a stylistic one for the translator, whose rendering will be deter-
mined by a knowledge of the distinctions made or not made explicit in the
different lexical systems. Thus, English lexicalizes a distinction here that
is brought out in other ways in French (moths would be papillons de nuit,
hence papillons can function in French as a superordinate for the species
of both night and day). The problem here is essentially a systemic one.
The concept of compensation which we are trying to refine cannot be
called upon to cover it.
Keith Harvey 77
3. A new descriptive framework for compensation
I would now like to bring together the issues I have been discussing and to
set out a more systematic framework for compensation. The aim here is to
refine our understanding of compensation as a theoretical concept and
thereby increase its power as a pedagogical tool.
It should be clear from the discussion so far that one of the central
problems we face is establishing what does or does not count as an in-
stance of compensation. We may be certain, for example, that we do not
wish to include straightforward instances of grammatical transposition.
These belong to systemic transfer between languages and do not have a
stylistic, text-specific function. Similarly, where a word in the source

language does not have a straightforward equivalent in the target lan-
guage, we would not include a resulting paraphrase in the target text
within our category of compensation. On the other hand, I have suggested
that we can confidently include puns and phonoaesthetic effects that are
specific to the source text as areas that could prompt target text compensa-
tion.
What should we do, then, with the fairly standard replacement of con-
ventional rhetorical routines between languages? For example, is the
translation of the French
- ~a va?
- ~a va.
by the English
- How are you?

- Fine.
an instance of compensation? If we accept the criteria that (i) compensa-

tion operates on stylistic, text-specific features, and (ii) sociocultural
practices (of which such a conventional exchange is a verbal example)
require cultural substitution, which is itself another distinct category of
transfer, then the answer is negative. Not all non-literal translation proce-
dures, in other words, are instances of compensation, and culturally-
determined routines are examples of those that fall outside the scope of
compensation as I have defined it in this paper.
Assuming that we can develop a reliable method for identifying in-
stances of compensation, we still need to address the question of the
diversity of relationships that may obtain between the source text effect
that is lost and the target text device that strives to compensate for it. The
two main issues that a descriptive framework should attempt to address

are therefore: (i) developing explicit criteria for recognizing instances of
compensation, and (ii) explaining the relationship that obtains between
the relevant stretches of source and target texts. The framework proposed
here addresses both issues along three axes. The first, typological, tack-
les the question of elaborating criteria for recognizing instances of
compensation. The second, degree of linguistic correspondence, and the
third, topographical, attempt to account for the relationship between loss
and compensation of effects.
3.1. The typological axis: identifying instances of compensation

The first axis I wish to propose assumes that while compensation con-
cerns itself with stylistic devices in text, these depend more or less heavily
on the systemic features of a language. I would therefore suggest distin-
guishing between two types of compensation along the typological axis:
stylistic and stylistic-systemic.
Stylistic compensation occurs where the effects achieved in the source
and target texts are text-specific and contribute uniquely to the colour,
tone and register of that particular text. We have seen many examples of
this type of compensation, including Hervey and Higgins' discussion, under
compensation in place (1992:37-38), of the use of sound for achieving
rhetorical effect.
Stylistic-systemic compensation is where the effects have a stylistic
value where they occur in the text, but these draw upon part of the con-
ventional systemic resources of the language. The exploitation of verb
tense relations for rhetorical effect, discussed by Hervey and Higgins
under compensation in kind (ibid:35-36), belongs to this category. I
would also include here the use of idiomatic expressions which, while
representing a conscious choice to employ a marked form for effect,
draws upon the lexical store of a language in a conventional way.
Another example of stylistic-systemic compensation occurs on page
28 of Asterix and the Laurel Wreath. Obelix regrets that Asterix has
decided that circumstances no longer make it necessary for them to buy
themselves out of slavery. The source text has the personal pronoun no us
('us') functioning as the direct object of the reflexive verb me payer (an
informal structure meaning 'buy'):
J'aurais bien aime me nous payer .... <;a m'aurait fait un souvenir a
rapporter de notre voyage.
(Literally: I would have liked to buy myself us (i.e. buy us for myself)
.... It would have made me a souvenir to take back from our trip.)
Keith Harvey 79
Of course, me (reflexive pronoun, 'myself') followed directly by nous

(direct object, 'us') form a string where the semantic oddity of the idea
expressed is deliberately reinforced by the grammatical strangeness. To-
gether they generate humour. The target language translation of me payer
can only be the less idiomatic buy:
I should have liked to buy us.
This target sentence cannot reproduce the source text's grammatical ef-
fect, which depended on the contiguity of two types of pronoun. The loss
here is of a systemic device in the source language, where reflexive verbs
abound, that was put to stylistic effect.

However, the next target text sentence, instead of reproducing the
source text's <;a m'auraitfait ... ('It would have made me ... ') makes the
subject of the sentence We and thus compensates for the loss of an equiva-
lent effect for me nous in the previous sentence:
We would have made a nice souvenir to take home from our trip.
The proximity of "We ... our" goes some way towards achieving a similar
effect through the use of grammatical words in the target language.
3.2. The correspondence axis: establishing the linguistic relationship
Once we have identified an instance of compensation, the next step is to

describe the degree of linguistic correspondence between the devices used
to achieve the effect in source and target texts. Three types of relationship
can be distinguished here: direct correspondence, analogical correspond-
ence, and non-correspondence.
Direct correspondence occurs where a lost effect that is achieved by
a given linguistic device in the source text is compensated for by the use
of the same type of linguistic device in the target text. We have seen many
examples of this, including the various puns discussed earlier, as well as
the phonoaesthetic devices described by Hervey and Higgins (1992:38).
It is important to note that I am not specifically interested here in whether
the effect compensated for in the target text remains comparable to that
lost in the source text. What interests me at this stage, for the purpose of
developing the current descriptive framework, is the means or devices by
which the effect is achieved in both texts.
Langeveld (1988:83) provides us with a succinct example of direct
correspondence in his brief discussion of compensation. The source text
is a sentence from Thomas Mann's Der Zauberberg:
Aber mit seiner Chaussure, hore mal, da steht es scheusslich.

(Literally: But with his shoe, listen, there it stands hideously.)
He notes the stylistic value of the use of the French word chaussure
('shoe') in the source language, German, and how this is not available for
a translation into Dutch:
The French word 'Chaussure' makes this sentence sound pretentious.

In Dutch the word 'chaussure' for shoes, footwear, is never used, not
even by the most affectedly speaking people. The Dutch translator has
therefore opted for the more neutral 'schoeisel' (footwear) and has
inserted the French element elsewhere in the sentence by translating
'scheusslich' (awful) by 'affreus', a French word (affreux = hideous)
which is used in Dutch and does have the affectedness that is called for
by the SL text: "Maar zijn schoeisel, zeg, dat is affreus." (ibid)
An extended example of direct correspondence occurs in Samuel

Beckett's own translation of his French play Fin de Partie (1956:51). In
the source text, there is a dialogue between the characters Clov and Hamm
that culminates in an obscene pun prompted by a malapropism. A flea is
in Clov's trousers:
Clov: La vachel
Hamm: Tu l'as eue?
Clov: On dirait ... A moins qu'elle ne se tienne colte.
Hamm: Coite! Coite tu veux dire. A moins qu'elle ne se tienne coite.
Clov: Ah! On dit coite? On ne dit pas coite?
Hamm: Mais voyons! Si elle se tenait coite nous serions baises.
(Literally:
Clov: The cow! [conventional French exclamation]
Hamm: Did you get her?
Clov: One would say so ... Unless she's holding herself coitus.
Hamm: Coitus! Quiet you mean. Unless she's holding herself quiet.
Clov: Ah! One says quiet? One doesn't say coitus?
Hamm: But let's see! lf she was holding herself coitus we'd be fucked.)
The use of coite ('quiet') instead of corte ('coitus') allows Hamm to relish
the vulgar ambiguity of baises ('fucked/done for'). Beckett compensates
for the loss of this sequence in translation (Endgame 1957:27) by substi-
tuting a target language-specific malapropism that leads to another vulgar
pun. The correspondence of linguistic devices employed in source and
Keith Harvey 81
target texts is thus direct:
Cloy: The Bastard!

Hamm: Did you get him?
Cloy: Looks like it ... Unless he's laying doggo.
Hamm: Laying! Lying you mean. Unless he's lying doggo.
Cloy: Ah! One says lying? One doesn't say laying?
Hamm: Use your head, can't you. If he was laying we'd be bitched.
Direct correspondence is one of the most typical types of compensation. It

is also one of the most straightforward to identify and account for. Other
types are more complex and potentially problematic.
Analogical correspondence occurs where the device used in the tar-

get text to compensate for a lost source text effect is derived from the
same linguistic repertoire as that used in the source text, without being of
an identical type. Hervey and Higgins (1992:37) cite an example that fits
into this category from another text in the Asterix series. In Asterix en
Corse (translated as Asterix in Corsica), there is a passage where the
source text "hinges on the mutual incomprehension of speakers of differ-
ent dialects" (ibid). The translation compensates for the loss of this effect
by substituting a whole sequence of playful punning between the charac-
ters. Although the linguistic devices employed in source and target texts
are different, they are both metalinguistic in character. I would therefore
characterize the target text compensation as analogical. Note that this is in
contrast with the long sequence from Beckett's Endgame cited above.
There, source text puns are replaced by target-text specific puns; in other
words, by the same metalinguistic device.
The example of stylistic-systemic compensation discussed above, where
the source text exploited the syntax of reflexive verbs (me nous payer) is a
further instance of analogical correspondence. The target text's exploita-
tion of the relation between a possessive pronoun (We) and a possessive
adjective (our) depends on a similar, though not identical, lexico-
grammatical device.
Non-correspondence occurs where the loss of an effect generated by
a source text device is compensated for by a target text device that shares
no linguistic features with it. Hervey and Higgins' example, discussed
under compensation in place, of the use of an array of rhetorical devices
(punctuation, lexical borrowing, etc.) to compensate for a source text's
exploitation of the source language's tense system represents an example
of this type of compensation. Similarly, we saw in the example of Couic/
couiquer (see 2.1.) how a sequence containing onomatopoeia and neolo-
gisms in the source text was compensated for by a conventional target
language idiom in the target text. This too is an instance of compensation

with non-correspondence of linguistic devices.
3.3. The topographical axis: location and distance
The third axis is topographical, that is it concerns the respective location

of the effect that is lost in the source text and compensated for in the
target text. I suggest that four broad types of relationship between the two
can be identified along this axis: parallel, contiguous, displaced and gen-
eralized.
A parallel relationship obtains where the compensation occurs at ex-
actly the same place in the target text as the effect that has been lost in the
source text. The compensation here effectively overwrites the loss. Ex-
amples of parallel compensation can be found in the handling of proper
names betweenLes Lauriers de Cesar and Asterix and the Laurel Wreath.
For example, the Gaulish chief Abraracourcix (a pun on the source lan-
guage idiom tomber sur quelqu 'un a bras raccourcis, literally: to fall
with shortened arms on someone, meaning 'to set upon someone') be-
comes Vitalstatistix in the target text. In the source text, Homeopatix's
wife is called Galantine. The translators seem to have assumed that the
cold pressed meat dish, galantine, was less familiar to members of the
target culture. Consequently, the humour is renamed after a sweet dish,
Tapioca.
A contiguous relationship obtains where the compensation occurs in
the target text within a short distance from the lost effect of the source
text. The translation from German into Dutch of the sentence by Thomas
Mann, discussed in Langeveld (1988) and quoted under direct corre-
spondence above, provides a clear example. There, the effect obtained by
the use of the French chaussure ('shoes') in the German source text is
compensated for later in the corresponding sentence in the Dutch transla-
tion by using another lexical item, affreus ('hideous'), whose French origin
is clearly marked for the target audience.
For another example of contiguous compensation, we can turn to page 4
of Asterix and the Laurel Wreath, which contains this sentence:
But Getafix also has other recipes up his sleeve ...
The idiom up his sleeve is a stylistic device that goes beyond the source
text's
Mais Panoramix a d'autres recettes en reserve ...

(Literally: But Panoramix has other recipes in reserve ... )
Keith Harvey 83
There is a conscious choice here on the part of the translator to 'go

idiomatic': in reserve would have done just as well in the target text as in
the source text, and would have maintained the tone of the latter more
closely. I suggest this is an example of contiguous compensation because
of something that happens on the same page. At the end of the source
text's description of Abraracourcix, we have the popular source language
expression
C'est pas demain la veille!

(Literally: Tomorrow's not the eve!)
The popular, informal character is underlined grammatically by the drop-

ping of the ne element of the source language negative particle before the
verb (i.e. n'est). This is translated as
Tomorrow never comes.
This expression is arguably of a more neutral (maybe even formal) regis-

ter in English than the source expression is in French. The difference in
register represents a certain loss in translation, and this leads me to sug-
gest that the apparently unprompted use of the idiom up his sleeve earlier
on the same page of the target text is in fact an example of contiguous
compensation.
A relationship of displaced compensation obtains when the instance
of compensation in the target text is a long distance from the source text
loss. This type of compensation throws up a number of difficulties. First,
it is obvious that the distinction between contiguous and displaced com-
pensation is relative rather than absolute. The decision to categorize a
target text feature as one or the other may largely depend on the size of
the text under consideration. In a long text, such as Asterix, one might
judge that compensation 'on the same page' is essentially contiguous. In a
sonnet, by contrast, an instance of compensation that occurs a mere few
lines after a source text loss might appear to be displaced. Second, when
dealing with a long text, it becomes increasingly clear that the analyst can
do little more than suggest that a certain target text feature is displaced
compensation. Short of asking the translator to recall the process s/he
went through, such a judgement must remain largely speCUlative. By its
very nature, this type of compensation takes the text as a whole as the unit
of translation. As a result, the analyst would have to undertake a quantita-
tive study of devices and effects in the text as a whole in order to produce
convincing evidence of displaced compensation.
Having recognized these limitations, it might be useful to quote a
possible instance of this type of compensation. In a bilingual edition of
Beckett's Happy Days/Oh les Beaux lours (Beckett 1978), Knowlson

notes that the English source text features the following newspaper adver-
tisement (ibid:26):
Wanted bright boy.
This is translated (ibid:27) as
Coquet deux-pieces calme soleil.

(Literally: Charming two-room flat quiet sun)
Knowlson (ibid: 120) accounts for such a radical, and apparently uncalled-
for, semantic change by pointing out that this "new French text also
compensated to some extent for the omission in French of the Cymbeline
quotation 'Fear no more the heat 0' the sun"'. In other words, the refer-
ence to solei! ('sun') in the target text advertisement (with its bearing on
the plight of the characters in the play, marooned on a beach with only a
parasol for protection) compensates for the loss incurred much later in the
act where the source text quotes from Shakespeare's Cymbeline. This is
replaced in the target text by a quote from Racine's Athalie that does not
mention the sun.
Partly because of the practical difficulties of identifying instances of
displaced compensation, I suggest a fourth type of relationship along the
topographical axis, namely generalized compensation. This occurs where
the target text includes stylistic features that help to naturalize the text for
the target reader and that aim to achieve a comparable number and quality
of effects, without these being tied to any specific instances of source text
loss. This is in fact the way that Nida and Taber conceived of compensa-
tion in general (see section 1 above). An example of this type of compensation
occurs on page 23 of Asterix and the Laurel Wreath, where the target text
has Osseus Humerus' dissolute son Metatarsus exclaiming
Hey Pater, Pater! We don't often see oculus to oculus ...
This is inventive and comic in a way that the source text is not:
Ah pater, pater! Nous ne sommes pas souvent d'accord ...

(Literally: Ah father, father! We don't often agree ... )
There appears to be nothing in the vicinity that accounts for the effect
created in the target text. It is, therefore, either a case of displaced com-
Keith Harvey 85
pensation that I have not been able to identify or more simply, perhaps, an
instance of generalized compensation.
4. Conclusion
The framework sketched out in this paper should in theory allow the
analyst to describe any instance of compensation systematically along all
three axes. Thus, the pun in Suivez mes ThraceslHeavy-duty nimble
Hoplites, discussed earlier, is an instance in the target text of parallel
compensation, of a stylistic type, with direct correspondence of linguistic
devices. On the other hand, the effect hinging on the language of the
Spanish maid (discussed in section 1.1 above), which was lost in transla-
tion, is made up for in the target text by contiguous compensation, of a
stylistic type, with non-correspondence of linguistic devices. Obviously,
some combinations are more likely to be identifiable than others. For
example, displaced compensation with non-correspondence of linguistic
devices would be difficult to spot in an extended text.
The framework presented in this paper needs extensive testing against
authentic and varied instances of translation. Its validity and usefulness
should be explored with reference to non-literary translation, in order to
establish whether certain types of compensation are privileged over oth-
ers in the translation of, say, commercial texts. The framework also needs
to be tested on translations between non-cognate languages in order to see
whether this crucial variable affects its descriptive power.
And finally, introspective protocols could be devised to gauge the
extent to which this description actually accounts for the processes that
translators go through. Similarly, proof of its pedagogical efficacy could
be sought in order to see how far it clarifies the issues involved in com-
pensation for trainee translators, and to what extent it can be integrated
into their range of internalized techniques. Given the importance attached
to the notion of compensation in the literature, it seems reasonable to
suggest that any future descriptive framework should satisfy at least these
requirements.
KEITH HARVEY
45 Bonham Road, Brixton, London SW2 5HN, UK
Notes
1. Page numbers are identical in source and target texts. This applies to all
examples from Asterix and the Laurel Wreath.
References
Baker, Mona (1992) In Other Words: A Coursebook on Translation, London

and New York: Routledge.
Beckett, Samuel (1978) Happy Days/Oh! les Beaux lours, bilingual edition ed-
ited by J. Knowlson, London: Faber.
Goscinny and Uderzo (1972) Les Lauriers de Cesar, Paris: Les Editions Albert
Rene; translated by Anthea Bell and Derek Hockridge as Asterix and the
Laurel Wreath, 1974, London: Hodder Dargaud.
Gutt, Ernst-August (1991) Translation and Relevance: Cognition and Context,
Oxford: Blackwell.
Hatim, Basil and Ian Mason (1990) Discourse and the Translator, London:
Longman.
Hervey, Sandor and Ian Higgins (1992) Thinking Translation: A Course in
Translation Method, French to English, London and New York: Routledge.
Langeveld, A. (1988) 'Compensation', in Paul Nekeman (ed) Translation, Our
Future: Proceedings ofXIth World Congress of FIT, Maastricht: Euroterm.
Newmark, Peter (1988)A Textbook of Translation, London: Prentice Hall.
------ (1991) About Translation, Cleve don, Avon: Multilingual Matters.
Nida, Eugene A. (1964) Toward a Science of Translating: with Special Refer-
ence to Principles and Procedures Involved in Bible Translating, Leiden: E.
J. Brill.
------ and C. R. Taber (1969) The Theory and Practice of Translation, Leiden: E.
J. Brill.
Wilss, Wolfram (1982) The Science of Translation: Problems and Methods,
Tiibingen: Gunter Narr.
On: 15 July 2015, At: 07:37
The Translator
The Dawn of a Modern Theory of

Translation
a
Juan C. Sager
a
University of Manchester Institute of Science &
Technology, UK
To cite this article: Juan C. Sager (1995) The Dawn of a Modern Theory of Translation,
The Translator, 1:1, 87-92, DOI: 10.1080/13556509.1995.10798951
Revisiting the Classics
The Dawn of a Modern Theory of Translation
JUAN C. SAGER
University ofManchester Institute of Science & Technology, UK
Les Problemes theoriques de la traduction (collection Bibliotheque des

idees). Mounin, Georges. Paris: Editions Gallimard, 1963. xii, 297 pp.
Reprinted 1976 (collection TEL), Paris: Gallimard; last reprint 1990. Pb.
ISBN: 2-07-029464-1, FF. 52.
A glance through various English books on translation confirms the im-

pression that most authors rarely cite critical works in another language,
and those who do only give the occasional reference to such works. There-
fore, not surprisingly, the name of Georges Mounin is absent from the
reference lists of several English publications in the field, some recent
and some not so recent, induding Steiner (1975), Bassnett-McGuire (1980),
Larson (1984), Robinson (1991), Baker (1992), Neubert and Shreve (1992),
and Gentzler (1993). This fact alone justifies introductions of the type
provided by this feature of the journal as a modest step towards overcom-
ing such cultural and/or linguistic barriers.
Though reprinted many times in France, most recently in 1990, and
although translated versions exist in a number of languages, Les Problemes
theoriques de la traduction has not as yet been translated into English.
First published in 1963, it represents the outcome of George Mounin's
doctoral research and was followed almost immediately by the publica-
tion of his complementary thesis, La machine a traduire, his to ire des
problemes linguistiques (Mounin 1964), which deals specifically with the
concept of machine translation, a major factor in the renewal of interest in
translation studies at the time. The importance of the book lies in the fact
that it is written at the crossroads of several currents which Mounin was
the first to identify and examine for their relevance to translation: he is
aware of the role of computers in translation - the first important work on
the subject, by Booth & Locke, had appeared in 1955; he is strongly in-
fluenced by the structural school of linguistics but at the same time

Georges Mounin: Les Probfemes theoriques de fa traduction 88
recognizes the importance of Chomsky's Syntactic Structures, published

in 1957; he has come to terms with the concept of linguistic relativity
(Whorf 1956) and examines its implications for translation. It is this par-
ticular juncture of ideas and trends which made the book so important for
its time; its lively discussion of these exciting ideas continues to arouse
interest in the French-speaking world, as we can see from the frequent criti-
cism and re-appraisals of Mounin's work by French writers on translation,
for example Ladmiral (1979), Larose (1989), and Schogt and Tatilon (1993).
Apart from Vinay and Darbelnet, who presented their practical method-
ology of translation on the basis of contrastive linguistics and stylistics in
1958, Mounin is undoubtedly the first European writer of our time to have
attempted to develop a theoretical framework for translation on the basis of
linguistics, and his careful choice oftitle,Les problemes theoriques ... , shows
that he is aware of the enormity of the task. In 1961, also in his doctoral
thesis, Jumpelt had raised the question of the possibility of a linguistically-
based theory and had observed certain regularities in the process of
translation. However, this publication was too close to the date of Les
problemes theoriques for Mounin to have considered it, although he was
familiar with Jumpelt's earlier work. Eugene Nida's equally cautious ti-
tle, Toward a Science of Translating, was only published in 1964, and
Catford's A Linguistic Theory of Translation, though known in lecture
notes from the early 1960s, was not published until 1965. Mounin him-
self returned to the subject of translation and linguistics in his later work,
Linguistique et traduction, which was published in 1976. An overview of
his life and work can be found in two fine tributes by Schogt and Tatilon
(1993) and Tatilon (1993), the former having appeared in a special issue
of La linguistique dedicated to Mounin, which includes a full bibliogra-
phy of his writings.
For the English-speaking reader, Mounin's book has the additional
interest of making us see the important period of linguistic research of the
late 1950s and early 1960s in a new light, through the eyes of an observer
who wants to make sense of the new research and seeks to apply it to
translation. He brings to our attention the contemporary French linguistic
thought of scholars like Benveniste, Buyssens, Martinet and Prieto. And
he comes well prepared for his task, because he combines the academic
outlook of scholarship in linguistics with a distinguished career as a literary
translator from Italian to French. This experience is evident in his earlier
monograph,Les belles infidetes, and qualifies him for the role of arbitrator
in the age-old dispute of whether translation is essentially an art (Cary 1956)
or a science with linguistic foundations. It also saves him from the all too
familiar accusation of theorizing in a manner which is irrelevant to the
practical task of solving translation problems. His conclusion is similar to
Juan Sager: The Dawn of a Modern Theory of Translation 89
that reached by Vinay and Darbelnet (1958): translation remains an art,

but an art based on science.
Les Problemes theoriques de la traduction consists of six parts, which
provide the structure for the discussion of the two problem areas identified
by Mounin. The first two parts examine the uneasy relationship between
translation, which belongs to the realm of parole or performance, and lin-
guistics, which is usually identified only with langue, or language as an
abstract system of relations. The third part, where Mounin looks at the prob-
lems associated with the lexicon in translation, serves at the same time as a
bridge to the next major area of enquiry, dealt with in parts four and five; this
is the general question of the relationship between language and society, which
variously supports the arguments for both the possibility and impossibility
of translation. The last part returns to linguistics by re-examining the role of

syntax in relation to the previous main axes of argument. These six parts are
further divided into fifteen chapters, with titles which clearly trace his lines
of argument. There is a substantial conclusion and a bibliography which il-
lustrates the width of the theories and ideas discussed.
Part One, 'Linguistics and Translation', consists of two chapters: 'Trans-
lation as Contact between Languages' and 'Should the Scientific Study of
Translation Activities be Part of Linguistics?'. Translation is described as
the coming into contact of two languages. In this sense, it is a form of
bilingualism and is therefore of interest to linguists. But, Mounin suggests,
instead of - or in addition to - studying translation as a means of shedding
light on various linguistic issues, it is also possible to use linguistics to iden-
tify recurring problems in translation. This shift in orientation, he argues, is
justified on three grounds. First, the growth of the activity of translation
means that the study of translation can no longer proceed on a purely
heuristic basis but requires a theoretical framework to support it. Second,
the increased interest in machine translation requires a detailed study of
the linguistics of translation. And third, the practice of translation poses a
theoretical challenge to contemporary linguistics, whose basic premises
suggest that translation is impossible, so that the use of linguistics in the
study of translation can provide an opportunity to revise and test the as-
sumptions of both fields. Mounin concludes that neither the input of
structural and functional linguistics nor the actual practice of translators
should be dismissed in the attempt to develop a theoretical framework for
the discipline. He then considers the question of whether the scientific
study of translation should be seen as a branch of linguistics and suggests
that translation is situated at the intersection of several disciplines, including
logic, linguistics, psychology, and general pedagogy. Yet, and in spite of the
fact that several prestigious institutions are involved in training translators
and interpreters, none of these disciplines has deemed it appropriate to
Georges Mounin: Les Probtemes theoriques de la traduction 90
treat translation as a field worthy of theoretical study.

In Part Two, 'The Linguistic Obstacles', which consists of three chap-
ters, he sets out to address this challenge by presenting linguistic theories
of meaning and culture as theoretical obstacles which hamper the scien-
tific study of translation. In Chapter Three, 'Translation in the Light of
Linguistic Theories of Meaning', he re-examines Saussure's idea of sense
relations, namely that the value of a sign is totally determined by the
value(s) assigned to other signs in the same system, and rejects it from the
point of view of translation. He argues that if meaning is no more than the
value assigned to a sign within its own system, then there is no way of
comparing signs across systems and therefore no way of communicating
across linguistic boundaries. He also dismisses Zellig Harris's concept of
distributional structure as irrelevant to translation. In Chapter Four, 'Trans-

lation in the Light of Neo-Humboldtian Theories about Languages as World
Views', Mounin re-examines the question of linguistic relativity. He re-
jects the extreme view propounded by Whorf (1956), according to which
each language structures reality in a radically different way. Such an ap-
proach to meaning and reality negates, at a theoretical level, the very
possibility of translation. He argues that translation theory has to accept
that each language is an instrument of communication by which human
experience is analyzed differently in each linguistic community. But this
does not deny the possibility of translation; on the contrary, translation
has given rise to the concept of 'equivalence' to deal with the fact that
language offers monolingual and bilingual speakers many ways of seeing
and representing observed reality.
The theme of meaning remains in focus in Part Three, 'Lexicon and
Translation', where Mounin continues his search for minimal units of
meaning, Vinay and Darbelnet's units of translation, and methods of
analyzing meaning that can both enable and explain the phenomenon of
translation. In the six chapters of this substantial part, his main attention
is devoted to semantic field analysis as documented in the work of
Hjelmslev and Prieto and, quite outside the limits imposed by linguistics,
Gardin's ideas about distinctive and descriptive features for knowledge
classification, arriving finally at the concept of definitions which he con-
siders an essential part of linguistics. With this observation, Mounin moves
into the area of terminology and de-contextualized word meaning, which
provides him with theoretically safe ground for identifying even language-
independent units of meaning. Such units offer the kind of certainty which
affirms the feasibility of translation. To restore the balance, he gives a
detailed analysis of connotation as opposed to referential meaning. Mounin
concludes that meaning is a complex concept which exceeds the limits of
linguistics. With the final chapter in Part Three, Chapter Eleven (,Trans-
Juan Sager: The Dawn of a Modern Theory of Translation 91
lation, Language and Interpersonal Communication'), he begins to move

away from the discussion of minimal units of meaning as he proceeds to
develop his own position on the question of the extent to which language
can express thought and to which individuals can understand the full in-
tention of a message. This question is crucial for determining the dual
role of the translator as mediator or interpreter of messages in communi-
cation.
Having rehearsed and examined all the linguistic arguments he could
muster against the possibility of translation, Mounin goes on to examine
the relative possibilities of translation in Part Four, 'World View and Trans-
lation', through a thorough investigation oflanguage universals. In Part Five,
he picks up again the question of resolving the conflict between 'Multiple
Civilizations and Translation'. This is done in chapters thirteen and four-

teen, provocatively entitled 'Ethnography is Translation' and 'Philology
is Translation', by which he means that the diachronic and synchronic com-
parative study of societies and languages can provide us with the theoretical
means of establishing the possibility of translating between them. In the
final chapter, 'Syntax and Translation', Mounin again searches for the
kind of universals that might provide the theoretical justification for trans-
lation and finds them in logic, axiomatic linguistics, semantics, and finally
in Chomsky's Syntactic Structures. The book concludes with a discussion of
the relativistic nature of translation and translatability. Mounin suggests
that, in theory, the basic units of any two languages, be they phonemes,
monemes or syntactic units, do not correspond. And yet communication,
and hence translation, remains possible by reference to the shared situa-
tion in which speaker and listener, or author and translator, are located.
On the surface, Mounin seems to do no more than pose some of the
main problems encountered in translation and explore the contribution of
various fields to this subject, but in the course of doing so, he questions
the validity of the arguments which present obstacles to the scientific study
of translation. The chief merit of Mounin's book then and now is that it
widens our horizon of reflection on translation. His vast knowledge and
intellectual curiosity, which bring so many different strands of thought to
bear on the paradox of translation, have created a rich foundation for fur-
ther thought and study. In this sense, he has made a major contribution
towards formulating the theory of translation for which we are still wait-
ing. This fact alone justifies our re-reading of Les Problemes theoriques
de la traduction and would also justify its translation into English.
JUAN C. SAGER
Department ofLanguage & Linguistics, UMIST, POBox 88, Manchester
M601QD, UK. jcs@ccl.umist.ac.uk
Georges Mounin: Les Problemes theoriques de La traduction 92
References
Baker, Mona (1992) In Other Words, London and New York: Routledge.
Bassnett-McGuire, Susan (1980, 1991) Translation Studies, London and New
York: Routledge.
Booth, A. D. and W. N. Locke (eds) (1955) Machine Translation ofLanguages,
New York: Wiley, and London: Chapman and Hall.
Cary, E. (1956) La traduction dans Ie monde moderne, Geneva: Georg.
Catford, J. C. (1965) A Linguistic Theory of Translation, London: Oxford Uni-
versity Press.
Chomsky, Noam (1957) Syntactic Structures, The Hague: Mouton.
Gentzler, Edwin (1993) Contemporary Translation Theories, London and New
York: Routledge.
Jumpelt, R. W. (1961)Die Ubersetzung naturwissenschaftlicher und technischer

Texte, Berlin: Langenscheidt.
Ladmiral, J.-R. (1979) Traduire: theoremes pour la traduction, Paris: Petite
bibliotheque payot.
Larose, R. (1989) Theories contemporaines de la traduction, Quebec: Presses
de l'Universite du Quebec.
Larson, Mildred (1984)Meaning-based Translation, New York: University Press
of America.
Mounin, Georges (1955) Les belles infideles, Paris: Cahiers du Sud.
------ (1964) La machine a traduire: histoire des problemes linguistiques, The
Hague: Mouton.
------ (1976) Linguistique et traduction, Bruxelles: Dessart et Mardaga.
------ (1994) Travaux pratiques de semiologie genera Ie, edited by Alain Baudot
and Claude Tatilon, preface by Claude Tatilon, colI. Theoria, Volume 3,
Toronto: Editions du GREF.
Neubert, Albrecht and Gregory M. Shreve (1992) Translation as Text, Kent, Ohio:
The Kent State University Press.
Nida, Eugene A. (1964) Toward a Science of Translating, Leiden: E. J. Brill.
Robinson, Douglas (1991) The Translator's Turn, Baltimore and London: The
John Hopkins University Press.
Schogt, Henry and Claude Tatilon (1993) 'Georges Mounin: la traduction sans
fard', La linguistique 29: 69-77 (Special Issue: Georges Mounin, linguiste
et semiologue).
Steiner, George (1975) After Babel, Oxford: Oxford University Press.
Tatilon, Claude (1993) 'In memoriam: Georges Mounin (1910-1993)"
Traduction, Terminologie, Redaction (TTR) 6(1): 7.
Vinay, J.-P., and J. Darbelnet (1958) Stylistique comparee du franr;ais et de
l'anglais. Methode de traduction, Paris: Didier.
Whorf, Benjamin Lee (1958) Language, Thought andReality, New York: Wiley,
and London: Chapman and Hall.
On: 14 July 2015, At: 23:06
The Translator
Book Reviews
Marilyn Gaddis Rose, Paul Kussmaul, Michel Ballard &
David Morris
To cite this article: Marilyn Gaddis Rose, Paul Kussmaul, Michel Ballard
& David Morris (1995) Book Reviews, The Translator, 1:1, 93-110, DOI:
10.1080/13556509.1995.10798952
Book Reviews
Translating Slavery: Gender and Race in French Women's Writing, 1783-

1823 (Translation Studies, Vol. 2; Series Editors: Albrecht Neubert, Gert
Jager, and Gregory Shreve). Doris Kadish and Fran<;oise Massardier-
Kenney (eds). Kent, Ohio and London, England: The Kent State University
Press, 1994. xvii, 346 pp. Rh. ISBN 0-87338-498-9, $27.
The subtitle of this excellent volume is more apt in relation to its content
than the title. Far from illustrating yet another demeaning metaphor of
translation, it recovers, via scholarly translation and translation scholar-
ship, three French women who spoke out against slavery during the last
two decades of the eighteenth century and the first two decades of the
nineteenth. The French writers featured are Olympe de Gouges, Germaine
de Stael, and Claire de Duras. Each writer is situated in her relevant con-
text by means of a critical essay and a translation of one or more of her
key works bearing on slavery. All the translators are women. The editors
frame these author-oriented sequences with a discussion of the ideologi-
cal aims of the translators, as expressed by the translators themselves.
Finally, the last word is given to the original French texts, which are re-
produced in the Appendix.
Precisely midway occurs the most illuminating section of the volume:
a dialogue between two of the translators after they had translated De
StaeI. Massardier-Kenney, a French native who has lived in the United
States for many years, notes class prejudice. Sharon Bell, an Mrican
American, notes race prejudice. Their reactions to the material they are
translating come from their own cultural perspectives and influence their
translation strategies. Bell, in fact, comments, " ... the statement [that blacks
were sauvages] offended me so much I could not put down what the sen-
tence actually said" (p. 175). None of this seems to distort the translations
in the sense of falsifying them. On the contrary, the leitmotif of this col-
lection is the illustration of ideological translating, especially gendered
translating. None of the translators here intends to be neutral; none con-
siders neutrality desirable.
What is just as illuminating and probably more remarkable is the pub-
lication of this anthology as Volume 2 of the Translation Studies series,
edited by Albrecht Neubert, Gert Jager, and Gregory M. Shreve, as a con-
tinuation of Ubersetzungswissenschaftliche Beitriige, the German language

BOOK REVIEWS 94
series published in Leipzig since 1978. Translating Slavery does not de-
rive from the systematic/systemic aspirations of European translation
studies as pursued by Shreve and Neubert - as well as Justa Holz-Manttari,
Hans Vermeer and Jean Delisle, to name but a few. One need not be Eu-
ropean, or Canadian, to be associated with the procedures of descriptive
linguistics: both Eugene Nida and Mildred Larson in the States, as well as
Gideon Toury and Hamar Even-Zohar in Israel, l tend to rely on the same
set of procedures. Such aspirations have as their implicit goal translations
that are accurate and, presumably, neutral and objective as well. At least,
translators in the systematic/systemic tradition are expected to observe
only the biases in the texts and leave their own out of the equation. Kadish
and Massardier-Kenney, by contrast, illustrate what Shreve and Neubert
in their fair-minded foreword, 'A House of Many Rooms: The Range of

Translation Studies', call the American "literary and humanistic tradi-
tion" (p. xi). While I would argue that, in the dizzying ricochets of Western
intellectual life since World War II, this latter tradition derives ultimately
from continental philosophy, it is, admittedly, espoused at present by
Americans and Canadians, especially comparatists, wherever they have
their academic berth. Douglas Robinson and Lawrence Venuti lecture in
English departments; Suzanne Jill Levine in Spanish; Richard Sieburth in
French as do Doris Kadish and Fran~oise Massardier-Kenney. This means
that they translate - and study translations - from the inside-looking-out,
as opposed to the outside-looking-in of descriptive studies. Whenever, or
if ever, the translators of Translating Slavery distance themselves from
their texts to analyze features such as the lexicon or syntax, what remains
critical in their relationship to the texts is the fact that they maintain an
alert intuition and sympathy throughout their analysis, in spite of the oc-
casional drudgery of the task. In short, they look for the epistemological
Subject, explicit or covert, and take a position vis-a.-vis that. All the trans-
lators in this volume - Kadish, Massardier-Kenney, Marie-Pierre Le Hir,2
Sylvie Molta, Maryann Dejulio, Sharon Bell, Claire Salardenne - have
put themselves inside the works they have translated. Not necessarily as
the authors, for the authors stopped short of condemning slavery qua slav-
ery. Sharon Bell in particular is sensitive to racism disguised as
enlightenment or sentimentality. Her remarks to Massardier-Kenney in
the dialogue mentioned above have implications for reading and viewing
other works. Caucasian readers, I trust, will take Bell's caveats to heart. 3
Nor have any translators inserted themselves into the texts as twentieth-
century women. They have been both exegetical and hermeneutic. Exegetical
because they have recast themselves as women of the authors' own times.
Massardier-Kenney used Mary Shelley'sFrankenstein to acclimatize herself
to early nineteenth-century diction, which is temporally appropriate for De
Stael's Mirza; similarly, Dejulio used Mary Wollstonecraft's Vindication
BOOK REVIEWS 95
of the Rights of Women to accommodate De Gouges's Declaration des

droits de la femme. And they are hermeneutic because they want to fore-
ground the advanced thinking of each author and to background the
concessions each author made to the power structures of her era. They do
not want to misrepresent the author, but, on the other hand, they do not
want her to be needlessly offensive to contemporary readers. This transla-
tion strategy, exegetical and hermeneutic at the same time, means that the
translators simultaneously aimed for a balance between period flavour
and heightened contemporary relevance. Generally speaking, literary trans-
lators have to sacrifice one goal to the other. But Dejulio's translation of
De Gouges's L 'esclavage des noirs, for instance, is a tour de force,4 giv-
ing Western repertory a tragi-comedy (drame) to balance the comedies of
such contemporaries as Richard Sheridan and Royall Tyler.

Successful also in this respect are Massardier-Kenney's translations of
De SHiel's Mirza and, with Sal ardenne, Claire de Duras's Ourika: they
are so authentic, yet readable, that we feel as if we had discovered English
language works of the same era. These two works themselves are docu-
ments requiring special pleading. They purport to be slave narratives, i.e.,
narratives by slaves, but they are in fact third-party testimonials where a
sympathetic but privileged author presumes to give a subaltern a voice. As
such, they call for being read 'across', i.e. deconstructed. I personally went
through all the translations, like a copy-editor, but found nothing to criticize.
Helpfully included also is Kadish's particularly fine translation of De
Stael's De l'Esprit des traductions. Although De Stael does not discuss
slavery here, not even obliquely, she expresses a view of enriching trans-
lation practice that is akin to that of Kadish, Massardier-Kenney and their
collaborators. De Stael exhorts translators to do what Kadish and
Massardier-Kenney have done: bring to late twentieth-century English-
language readers the texture, substance, and even verbal echoes of three
late eighteenth/early nineteenth century French women writers who were
appalled by slavery and said so. This representation of French writers via
translation belies the notion that translation is inevitably an act of trans-
gression. Here at any rate, consciously gendered, beneficently biased
translation is armed support. Translating slavery, even historically recon-
sidered, is striking against slavery. We could say that, in Neubert and
Shreve's 'house of many rooms', the historical victims have found a bun-
ker and are firing back.
MARILYN GADDIS ROSE

Center for Research in Translation, Binghamton University, P. 0. Box 6000,
Binghamton, New York, USA
BOOK REVIEWS 96
Notes
1. Gideon Toury and Hamar Even-Zohar, as well as Jose Lambert, who is

closely associated with them, are housed in departments of comparative
literature, but the poly system approach which they have developed and
which has made such a valuable contribution to translation studies over
the past two decades proceeds like descriptive discourse analysis. In terms
of Neubert and Shreve's translation studies metaphor of the 'house with
many rooms', polysystem theory does real estate listings.
2. I have one minor quibble with this particular translator's use of represen-
tation (p. 79) and represented (p. 81), in her carefully researched discussion
of the De Gouges play, L 'esclavage des noirs. Le Hir appears to mean
simply performance and presented, respectively. Representation, in the
tradition where Neubert and Shreve place these contributors, has a re-
stricted usage. The volume seems to have been proofread by someone
outside translation studies, because Raymond van den Broeck and Wolf-
ram Wilss appear in the general editors' introduction with Anglicized
surnames (Brook and Wills respectively, p. x). I suspect also that the Ameri-
can scholar Neubert and Shreve meant to pair with Eugene Nida is Mildred
Larson, not Lawson (p. ix) - perhaps a confusion with Veronica Lawson,
who is British?
3. For example, speaking of De Stael's Mirza, Bell says: "She creates these
two characters who prove that black people are just like white people ....
And I found offensive the fact that these two were given special status be-
cause they were superior. In other words, blacks didn't have any value unless
they were superior; nobody considered the mass of blacks at all" (p. 174).
4. A proponent of an enlightened monarchy despite her condemnation of
colonialism, De Gouges was guillotined during the reign of terror which
followed the French Revolution.
Language Engineering and Translation: Consequences of Automation

(Benjamins Translation Library, Vol. 1). Juan Sager. Amsterdam & Phila-
delphia: John Benjamins, 1993. xix, 345 pp. Hb. ISBN 1-55619-476-5,
Hfl. 160; Pb. ISBN 1-55619-477-3, Hfl. 60.
As stated in the preface, this book sets out to describe translation in an

industrial setting, focusing in particular on automated alternatives to hu-
man translation. It starts by exploring "existing theories and models in
order to select those which are still adequate for the changed situation or
to reformulate those which can be adapted" (p. xviii). By far the largest
part of the book is concerned with this particular task. There are six chap-
ters altogether: the first four cover theoretical topics in linguistics and

BOOK REVIEWS 97
translation studies; Chapter Five, 'The Automated Dimension of Transla-

tion', and Chapter Six, 'Industrialisation of Translation' , are more directly
concerned with the main aim of the book, namely describing the indus-
trial setting of translation.
Chapter One, 'The Language Industry and its Raw Material', offers an
overview (with numerous diagrams) of the various types of communica-
tion in an industrialized world. For non-specialists in the field of language
processing, the initial survey of automated tools such as spelling and style
checkers, lexical databases and dictionary-making facilities, multilingual
term banks, text editing facilities, and interactive machine translation sys-
tems will be particularly informative. This survey is followed by a
discussion of some aspects of language and communication which are
relevant to translation. There is a detailed discussion of language func-

tions, of the pragmatic dimension of communication, speech acts, and the
distinction between natural and artificial languages. Students of transla-
tion will find all this highly relevant; those with some knowledge of
linguistics will be able to identify familiar models and notions. What struck
me as somewhat unusual was the fact that the author does not always
mention the protagonists of the relevant models. He discusses speech acts,
a central notion in this book, without mentioning Austin or Searle and
talks about speech situations without any reference to Crystal and Davy's
model of situational analysis (1969) nor to the work of their predecessors,
Halliday, McIntosh and Strevens (1964). He does, however, mention Joos
(1967) and uses his stylistic labels of frozen, formal, consultative, casual,
familiar, and intimate style (pp. 25, 78ft). Perhaps these concepts are typi-
cally taken for granted in an Anglo-Saxon academic community.
Moving from the general to the particular, Chapter Two ('Communi-
cation Theory for Translation') goes into the details of situational and speech
act analysis. It seems to me that Sager here uses 'speech act' in a non-
technical sense. For instance, he does not distinguish between locution,
illocution and periocution. On the other hand, he mentions the directive
purpose of speech acts, one of Searle's types. If one applies Austin's and
Searle's theory, Sager seems to have fallen victim to a common misun-
derstanding of what makes communication successful. He suggests that
directives are successful if readers do what they have been told (p. 70).
But according to Austin and Searle, directives, and indeed all speech acts,
are successful if the reader or hearer has understood what is meant, that is,
if s/he has grasped the illocution of the speech act. For instance, a direc-
tive speech act such as Could you get me a can of beer from the fridge,
Jimmy? is successful even if Jimmy remains seated and replies Why don't
you get it yourself, because by saying this he has made it perfectly clear
that he has understood what was asked of him. These are, however, minor
BOOK REVIEWS 98
quibbles. The strength of this chapter is that it covers just about every
conceivable aspect of a communicative situation. There are interesting
situational dimensions discussed here which, to my knowledge, have not
been tackled in earlier models of situational analysis. Sager discusses, for
instance, the psychological state of the sender or receptor (p. 63) in terms
of whether slhe is depressed, optimistic, cheerful, angry, amused, and so
on. All relevant factors are subsumed under the term 'choices', a word
which is perhaps more amenable than 'situation' to the inclusion of such
elements of communication as topic, purpose, language, and technical
means, in addition to traditional situational features such as degree of
formality. On the other hand, translators do not always have access to
such choices, that is, they cannot always choose. They often have to fol-
low specific instructions from those who commission the translations.

Also, once a translator has made one choice, such as agreeing to translate
a large manual on car repair, slhe is no longer free to opt for certain other
choices, for instance using a fax or telex machine to receive the original
or transmit the translation.
Sager deals quite extensively with one particular and important choice,
namely the choice of text type. The types he covers include schedule,
memo, essay and report. From a practical point of view, other text types
such as manual, instruction leaflet and commercial correspondence should
perhaps have also been included: according to a survey conducted by
Schmitt (1990), these are among the most common text types that practis-
ing translators have to deal with on a daily basis. In general, however, this
chapter offers valuable observations on reader expectations and recogni-
tion of text types. There is also a detailed discussion of processes of
reception and comprehension, clearly based on a study of psycholinguistic
models, though here, again, no explicit references are given. This is fol-
lowed by an overview of Grice's principles and maxims and their
application, via Sperber and Wilson, to translation. Sager concludes this
chapter by introducing the notions of 'text modification' (p. 105ft) and
'original as a draft' (p. 112), which allow us to see translation within the
wider perspective of text production and to place translators on a par with
editors, revisers and abstractors. This is an important perspective which
gives the profession room to grow and allows its members to fulfil tasks
other than direct translation, tasks such as technical writing, documenta-
tion, and public relations, where expertise in text production as such is required.
In Chapter Three, 'Theoretical Aspects of Translation', Sager goes on
to discuss a number of existing models and theories of translation, includ-
ing those expounded by Albrecht Neubert, Eugene Nida, Roman Jakobson,
Hans Vermeer, and William Frawley. There are two sections which I found
particularly relevant. The first (pp. 142-45) offers a detailed and critical dis-
BOOK REVIEWS 99
cussion of the notion of equivalence and concludes that we ultimately

have to distinguish between the levels at which it is possible or desirable
to achieve equivalence (grammatical, pragmatic, and so on). The second
(pp. 145-49) presents various criteria for the evaluation of translation in an
industrial setting, where factors such as clarity and intelligibility are of
key importance. Theoretically-minded readers will find much of what Sager
has to say in this chapter rather novel. Issues such as the amount of time
needed to produce a translation, the cost-effectiveness of the method used
or time spent on a translation (p. 174), the value of the source text as re-
flected in the amount of money a client is prepared to spend on having it
translated (p. 175), these are all considerations which do not normally merit
discussion in the theoretical literature. Again, placing translation within
an industrial setting here broadens the definition of translation, allowing

Sager to describe quite a number of activities which are not traditionally
seen as translational. They include, for instance, the production of autono-
mous documents from source language drafts (p. 179), excerpts (p. 181),
abstracts (p. 181) and gist translations (p. 182). These are all "functional
types of translation" (p. 179) in Sager's view, as indeed they are in the
view of many translation scholars, particularly in Germany, including Hans
Honig, Sigrid Kupsch-Losereit, Christiane Nord, Katharina ReiB, Peter
Schmitt, Hans Vermeer, and myself.
Chapter Four, 'A Model of the Translation Process', picks up quite a
number of topics which have already been dealt with briefly in previous
chapters and discusses them in more detail; this seems typical of Sager's
style of writing. The issue of comprehension, for instance, is taken up
again with reference to Lorscher's psycholinguistic research (pp. 213ft),
and the concept of equivalence is reassessed within the framework of Vinay
and Darbelnet's work (pp. 221ft). There is a very detailed description of
various techniques of finding equivalents. From a functional point of view,
which is otherwise adopted in this book, I would have expected some
structuring of the types of equivalents listed. Sager does mention at one
point that the pragmatic level "takes precedence over matching at the lexical
and syntagmatic levels" (p. 230), and I fully agree with this statement.
The bulk of his presentation, however, is concerned with isolated tech-
niques such as borrowing, loan translation, transposition, modulation, and
pre-established matches. All of these can indeed be useful, but within a
functional framework they would have to be clearly related to pragmatic
considerations, and the latter would have to be shown to override them.
Sager's approach to the evaluation of errors creates similar problems at
times. He suggests that "the gravity of the error is further dependent on
the importance the erroneous element has in the document" (p. 241). This
is a functional statement and, it seems to me, a very sensible one. It means
BOOK REVIEWS 100
that we should not judge errors in isolation but should rather consider
their effect within a particular context. From this point of view, however,
a statement such as "An absolute evaluation assesses an individual trans-
lator's precision and can serve a didactic function because a major omission
is obviously more serious than a spelling error" (p. 241) seems to take us
back to a contrastive view of translation. It is not quite clear to me which
view of translation Sager actually favours overall, the functional or the
contrastive.
Chapter Five, 'The Automated Dimension of Translation', begins with
a comparison of human and machine translation (MT) and convincingly
shows the limitations of the machine, the decisive issues being that it can
only translate regular and recurring patterns (as in the case of the Cana-
dian METEO weather forecast system), that it is limited to the sentence as

the uppermost unit of translation, that it cannot infer meaning, and that it
is unable to compensate for translation loss. Sager suggests, however, that
the success of machine translation should not be assessed by reference to
tasks that we know it cannot perform; instead, it should be measured "by
the success with which particular objectives have been achieved" (p. 262).
Machine translation, we are told, has a limited range of possible func-
tions. Unfortunately, Sager is rather vague about these. Readers who are
not particularly well informed about machine translation would have ben-
efited from a more detailed discussion of its practical applications.
In his relatively brief final chapter, 'Industrialisation of Translation',
Sager discusses the increase in demand for translations in recent years,
which is largely due to an increase in the level of sophistication of prod-
ucts, resulting in a greater volume of documentation being generated,
especially by the translation services of the European Commission in Brus-
sels. He also talks about practical matters such as the problem of pricing
translations (per page, line, word or character). Against this background
of increasing demand, there is, or rather should be, a large market for
machine translation. The important question to be resolved for MT at its
present stage of development, according to Sager, is not so much how it
can be adapted to the complexities of all types of text produced by the
human mind, but rather how texts can be adapted to the possibilities of
machine translation. Sager believes the main uses of machine translation
to be in the following areas: in the translation of "stereotype text types on
largely constant subject matter" (p. 300), the translation of restricted lan-
guages (p. 308), the automatic scanning of full texts followed by selective
human or machine translation (p. 301), creating multilingual databases
(pp. 301-02), and producing term banks and dictionaries for translators
(p. 309). Finally, he mentions some recent developments which involve
incorporating suprasegmental semantic and pragmatic relationships in the
BOOK REVIEWS 101
design of machine translation systems (p. 307). But, again, he does not
offer any detailed information about how these relationships are dealt with
in the new systems.
So, does the book deliver what it promises in the preface? It does, to
some extent. It delivers a useful and clear description of theories and models
of translation and presents quite a number of new and stimulating ideas,
but it is not always particularly clear on which of them are still relevant in
our heavily industrialized and largely automated society. The book, I think,
would have gained from focusing more clearly on those models which are
particularly relevant to our present situation. This could perhaps have been
achieved by condensing some sections and avoiding repetition in others,
leaving also some room for additional examples to illustrate the models in
question. As far as the discussion on machine translation is concerned,
the reader will find Sager's arguments easy to follow, and the relevant
chapters should help make this rather esoteric subject more accessible to
a wider audience. Sager, as he puts it in the preface, "looks" at machine
translation. What we see through his eyes seems fascinating, and some
readers will no doubt want to look more closely. For them, some explana-
tion of the way in which various machine translation systems deal with
specific tasks, using examples in the manner of Wilss (1988), would have
been a valuable addition.
PAUL KUSSMAUL
Johannes Gutenberg-UniversitiitMainz, Fachbereich 23, Institut fUr Anglistik
undAmerikanistik, D-76711 Germerscheim, Germany
References
Austin, John L. (1962) How to Do Things with Words, Oxford: Clarendon Press.
Crystal, David & Derek Davy (1969) Investigating English Style, London:
Longman.
Halliday, Michael A. K., Angus McIntosh and Peter Strevens (1964) The Lin-
guistic Sciences and Language Teaching, London: Longman.
Joos, M. (1967) The Five Clocks, New York: Harcourt, Brace & World Inc.
Schmitt, Peter A. (1990) 'Die Berufspraxis der Ubersetzer. Eine Umfrageanalyse',
BDU Mitteilungsblatt for Dolmetscher und Ubersetzer, February: 1-15.
Searle, John R. (1969) Speech Acts: An Essay in the Philosophy of Language,
Cambridge: Cambridge University Press.
Wilss, Wolfram (1988) Kognition und Ubersetzen: Zur Theorie und Praxis der
mensch lichen und der maschinellen Ubersetzung, Tiibingen: Niemeyer.
BOOK REVIEWS 102
La Traduction raisonnee: Manuel d'initiation ala traduction professionnelle

de I' anglais vers Ie franfais (Collection Pedagogie de la traduction). Jean
Delisle. Ottawa: Presses de l'Universite d'Ottawa, 1993. xii, 484 pp. Pb.
ISBN 2-7603-0372-1, 190 FF/$38 (Canadian).
As its title indicates, this book discusses problems of English-French trans-

lation. Clear in its conception and layout, it starts with an informative
four-page introduction, followed by a table of abbreviations and symbols
and a convenient glossary of 180 clearly defined terms, with a system of
cross-references which encourages the reader to go beyond the entry con-
sulted. The bulk of the manual consists of eight general 'objectives', as
the author calls them, and these break down into fifty-six specific objec-
tives (discussed below in some detail). There are two appendices and a
bibliography of 183 titles. The first appendix consists of a list of texts
proposed for translation; it is not immediately obvious what use the reader
might make of this list. The second appendix lists selected references un-
der headings such as 'Professional Journals', 'History of Translation' , and
'Technical Translation', with various hints for further reading. The bibli-
ography, which follows the appendices, is copious, listing a wide range of
articles and books on the translation of 'pragmatic' texts. 1 Despite this
bias, which is inevitable given the aim of the book, there seems to be a
sufficient number of references to general works on translation studies.
But there also remain certain surprising lacunae. For example, there is
one article by Jean Maillot, the late president of the French Translators
Society, but no mention of his main work, La traduction scientifique et
technique (1981), which is more relevant in this context. References also
appear under 'Further Reading' at the end of each chapter.
Despite the complexity of the subject matter and the density of the
text, the book remains user-friendly. The judicious use of paragraph breaks,
titles and subtitles, the running heads at the top of each page, and the use
of bold type and capitals to highlight key items all help the reader to
identify the important concept~ and to follow the structure of the argu-
ment. Each specific objective, which is equivalent to a chapter, is presented
clearly, with a good balance of text, quotations and examples. The theo-
retical part rarely goes beyond five or six pages and is immediately followed
by practical exercises and questions.
Throughout the book, Delisle draws on two fairly different approaches
to translation. The first derives from contrastive linguistics, as illustrated
most clearly in the work of Vinay and Darbelnet (1958). The second is
the interpretive view advocated by the Paris School and expounded in
particular by Danica Seleskovitch and Marianne Lederer (see for instance
1SSN 1355-6509 © St. Jerome Publishing, Manchester

BOOK REVIEWS 103
Seleskovitch and Lederer 1989). The interpretive view of translation, some-

times referred to as the 'theorie du sens', is based on its founders' extensive
experience in simultaneous interpreting, but it has been gradually extended
to account for the transfer of meaning in written translation as well. Much
attention is paid to the process of constructing meaning and to the role of
context and extralinguistic factors in this process. The school is also known
for its emphasis on documentation, an aspect which receives a great deal
of attention in this book. Delisle is of course well known as a theorist in
his own right, and his previous publications include a manual of transla-
tion (Delisle 1980) based on his doctoral thesis. The present book is
essentially the outcome of his own personal research and experience, but
its originality lies in the successful integration of contrastive linguistics
and the interpretive approach, the latter having so far dismissed the meth-
ods of contrastive linguistics as irrelevant.
The first 'general objective', or rather section, consists of an introduc-
tion to the terminology of translation studies. Delisle stresses that the use
of precise terminology is a prerequisite to a scientific study of translation.
This emphasis on terminological precision is also reflected in the glossary,
where he not only pays a great deal of attention to providing unambiguous
definitions but also includes rather unusual exercises to ensure that the
reader has understood them. For example, the reader is given a number of
headings and subheadings and asked to list the contents of the glossary
under the relevant ones; this exercise is followed by conceptual questions
which can only be answered successfully if the grouping of the terms has
been done accurately. But Delisle seems to borrow Vinay and Darbelnet's
terms without questioning their validity and without revising their defini-
tions to reflect more recent research. In particular, the notion of the unit
of translation, 2 as well as the definitions of various aspects of the transla-
tion process, ought to be seriously reconsidered. There is also some
confusion, acknowledged by Delisle himself, in the use of terms which
describe the differences in length between original and translation. 3
The second general objective discusses the tools available to transla-
tors: general reference works, different types of dictionaries, and other
sources of information such as data banks. A number of guidelines for
assessing the reliability of various sources are provided and some of the
limitations of bilingual dictionaries are examined. The third general ob-
jective tackles the various procedures involved in dealing with a translation
task, what must be done before, during and after the translation process.
Delisle looks at details such as the medium used for work - pen, type-
writer, computer - and examines the various stages of translation, including
the initial reading of the text, conducting background research, the actual
writing of the translation itself (the form it should take, the speed at which
BOOK REVIEWS 104
it should be done), the process of putting the translation aside and return-
ing to it at a later stage, and the final editing of the text for style. There is
emphasis on the need to pinpoint potential translation problems during
the reading process, which fits in with the interpretive approach. But
whereas the requisite skills for the various processes discussed are gener-
ally assumed to evolve gradually with experience, Delisle attempts a
short-cut to efficiency here by making use of contrastive linguistics to
develop the same skills consciously and within a shorter period of time.
He does, however, stress that there are no ready-made solutions which
can be applied directly to specific problems. The aim here is to help the
trainee translator identify recurring problems and consequently develop
the ability to find appropriate solutions in context through a process of
reasoning.
General Objective 4, entitled 'The Cognitive Process of Translation',
explores the importance of extralinguistic information in determining the
meaning of the linguistic components of the text. One of the specific ob-
jectives subsumed under this heading examines the way in which terms
are generated in translation. Delisle considers three possibilities. The first
is that of direct or almost direct transfer from the source to the target text:
this often applies in the case of proper names, figures, and similar items.
The second involves choosing from a list of 'equivalent' terms offered in
one or more dictionaries, with due regard to the context. The third re-
quires the translator to be creative in response to the demands and realities
of the discourse situation, which may not be well served by a literal ren-
dering of the original. Objective 5 discusses questions of usage in spelling,
abbreviation, punctuation, typography, and similar areas. Under General
Objective 6, Delisle focuses on problems of vocabulary, starting with words
which are very close to being false friends. For example, automatiquement
sounds somewhat unnatural as a substitute for automatically in certain
French contexts. The meanings and translations of deceptively easy or
polysemous words such as available, challenge, corporate, and develop-
ment are similarly discussed, with numerous examples and exercises to
illustrate their complexity. The relationship between morphology and
meaning is examined, as is the relationship between literal and figurative
meaning and the way in which such relationships may affect the transla-
tion of certain items. And finally, differences in the level of explicitness
between the lexical units of different languages are also discussed, for
example wall-to-wall carpet in English vs. moquette in French.
General Objective 7 deals specifically with problems of syntax. The
polysemy of a number of connectors such as when, as, while, and with is
explored and some of their possible renderings are discussed. Other syn-
tactic items and constructions examined in this section include
BOOK REVIEWS 105
comparatives, long series of adjectives, and verbs which express progres-

sive action, such as s'est accrue in La production industrielle s' est accrue
de 0,8 % en mars. Syntactic choices are shown to be ultimately dependent
on higher-level discourse patterning, which accounts, for instance, for the
more frequent use of negative interrogative structures in French: a ques-
tion such as How about a dog house for your next creation? (to a child
playing with a lego set) is said to be more naturally rendered into French
asPourquoi ne pas construire une niche achien maintenant? ('Why don't
you build a dog house now?'). Delisle treats the distinction between prepo-
sitions and adverbs in phrasal verbs rather casually in this section,
suggesting, for instance, that out is a preposition in to crawl out (p. 298),
but this is really a minor oversight and does not detract much from the
strengths of this chapter, namely clarity of expression, conciseness, and

abundant and convincing examples.
The final section of this book, General Objective 8, is entitled 'Prob-
lems in Composition'. This is a rather modest title considering that what
is attempted here is a rational and systematic treatment of the various fin-
ishing touches which can finally give the translation a distinctive French
flavour. Of the specific 'objectives' included in this chapter, Objective
55, which tackles issues of coherence and cohesion, and Objective 56,
which examines techniques for idiomatizing a translation, are particularly
worth reading. In his introduction to the chapter, Delisle quotes Peter
Newmark as suggesting that "all translation problems finally resolve them-
selves into problems of how to write well in the target language" (Newmark
1981:17; quoted by Delisle on p. 328). The quote, taken out of context
and therefore open to misinterpretation, is meant to stress the importance
of writing well in the target language. It would have been helpful, in this
reviewer's opinion, to remind the reader at this stage of the critical role of
meaning assessment and interpretation in any kind oftranslation activity.
Another problem concerns an important part of this section, entitled' Cor-
rect Expression in French', which deals with discourse-level preferences
such as the predominance of the passive voice in certain types of English
prose as opposed to the predominance of the active voice in French, or the
preference for nominalization in French where verbal structures may be
more natural in English, as well as the frequency and types of anaphoric
references, repetition, juxtaposition and coordination. The problem here
is not one of content but of organization: this chapter is essentially con-
trastive in orientation and might have therefore been included in the previous
section, General Objective 7. Alternatively, General Objective 8 should
perhaps have been given a different title to indicate that it is concerned
both with the process of polishing the text in French and with additional
contrastive issues ranging from morpho syntactic ones (for example the
BOOK REVIEWS 106
use of nominal vs. verbal formulations) to discourse patterning and

stylistics.
Despite this query concerning the homogeneity of the last section,
Delisle's manual deserves much praise for its clarity of conception and
presentation. It offers a rational and realistic programme for introducing
students to the intricacies of translation. It is rational because it is firmly
grounded in theory and realistic because it applies theoretical insights sys-
tematically to achieve specific pedagogical goals, constantly illustrating
theoretical issues with concrete examples and reinforcing them with trans-
lation exercises. The general process of translation is described efficiently,
the terminology necessary for its analysis and teaching is defined clearly,
and the pedagogical aims the author sets out to achieve are structured and
organized according to a graded scale of difficulty. This book will be of

great interest to anyone involved in teaching translation and wishing to do
so systematically rather than by means of a random selection of texts with-
out any order or rationale behind them. Le Livre du maftre, the Teacher's
handbook, contains answers to the exercises and translations of sixty Eng-
lish texts and is available from Tothemes Diffusion.
MICHEL BALLARD
Universite d'Artois, U.F.R. de Langues, Litteratures et Civilisations
etrangeres, 9, rue du Temple, BP665-62030, Arras Cedex, France
Notes
1. Delisle uses this term to refer to informative texts, mainly scientific and
technical, as distinct from literary texts.
2. For a recent discussion of the notion of 'unit of translation', see Ballard
(1993:223-62).
3. See pages 29-30, for instance the definition of etoffement. Delisle acknowl-
edges the confusion on page 30.
References
Ballard, Michel (1993) 'L'unite de traduction', in Michel Ballard (ed)La traduction

al'Universite, Translation in Higher Education, Lille: Presses Universitaires
de Lille.
Delisle, Jean (1980) L'analyse du discours comme methode de traduction, Ot-
tawa: University of Ottawa Press. Part I translated by Patricia Logan and
Monica Creery (1988) as Translation: An Interpretive Approach, Ottawa:
University of Ottawa Press.
Maillot, Jean (1981) La traduction scientifique et technique, Paris: Technique
et documentation.
Newmark, Peter (1981) Approaches to Translation, Oxford: Pergamon Press.
BOOK REVIEWS 107
Seleskovitch, Danica and Marianne Lederer (1989) Pedagogie raisonnee de

l'interpretation, Bruxelles and Luxembourg: Office des Publications des
Communautes europeennes, and Paris: Didier.
Vinay, I.-P. and I. Darbelnet (1958) Stylistique comparee du franc;ais et de
l'anglais: methode de traduction, Paris: Didier.
Traduction, Adaptation et Editing Multilingue. Jacques Permentiers, Erik

Springael and Franco Troiano. Brussels: Telos Communication Group
Editions, 1994. xiv, 221 pp. Pb. ISBN 2-9600071-0-7, FB 900IFF 148.
The authors of this book came together as founders or directors of

EUROLOGOS and LITIERA GRAPHIS, two successful companies based
in Brussels and offering translation, interpreting and desktop publishing
services. The book is the outcome of over ten years of collaboration and
experience in these fields. It offers a detailed study of the various pro-
cesses involved in producing high-quality, polished texts in different
languages. There seems to be a heavy bias towards advertising material
and instructions for the use of commercial products. However, the book
is eminently useful overall in highlighting that producing a translation is
far from being a simple matter of sitting in front of a typewriter or word
processor with a few dictionaries and a text to translate. The last four or
five years alone have radically transformed the possibilities of publishing
a highly polished, professional (and colourful) document from a relatively
small-sized office. An increased awareness of the exciting possibilities
which the immediate future holds out to the young professional translator
can be valuable in bringing training programmes up-to-date. It is also
useful in reminding the more experienced translator of the continued need
to keep abreast of developments in this area.
Quality is a key word throughout this book, and very rightly so. There
are many stages, from the writing of the source text to the final printing of
the target text(s), in however many languages that might be, at which the
quality can go astray. For those who like to see things expressed in dia-
grams and flow charts, the book provides ample visual illustration of the
different stages involved in producing a translation. The very first stage,
the initial writing of the text in the source language (or the original utter-
ance in the case of interpreting) is of prime importance. The translator
sometimes does not have the opportunity, or occasionally fails to take
advantage of it when it is there, to point out ambiguities or other weak-
nesses in the original text, and this can seriously affect the quality of

BOOK REVIEWS 108
subsequent translations. A similar problem arises with interpreting, where

conference speakers may at times be determined to get twenty minutes'
worth of information into a ten-minute slot. It is not unusual to see con-
ference interpreters gesturing frantically to the speaker to slow down or to
hear them apologize to the audience for not being able to keep up with the
speaker.
It is of course difficult to share one's experience with others without
referring to it explicitly and without implying that it represents expert or
superior practice. I am prepared to believe the authors when they claim to
apply rigorous principles to ensure quality control at all stages, and this
would no doubt explain the success of their companies. What does seem
unnecessary in this context is the repetitive attacks on what the authors
call 'letter-box' translation bureaux, those who accept any translation work
from and into any language, using freelance translators and making little
attempt to check the quality of the work they pass on to their clients. These
are disastrously dishonest and unprofessional operations, it is true, but
once said, enough said. Also, there are market forces of supply and de-
mand which sometimes prevent individuals and companies from achieving
their ideal standards of quality. It would be an ideal world indeed if all
translations were done, as apparently they are done at EUROLOGOS, in
pairs, with a translator and a reviser working on every text, and both work-
ing only into their respective native language and alternating roles for different
texts.
The translator's command of the target language, which should nor-
mally be hislher mother tongue, plays a major role in determining the
quality of the final text. The authors make the point that translation is
based on the written parole, that is on the actual rather than idealized use
of language. This reinforces their point concerning the translator's com-
mand of the target language and implies that the writing of material for
translation as well as the actual act of translating itself should only be
entrusted to people who are experienced in the art of writing. The same is
of course also true of spoken discourse, and hence of interpreting. But the
authors go on to plead for the acceptance of new and borrowed terms and
criticize those 'conservatives' who want to protect the national language
in order to enhance clarity of expression. This seems a little contradic-
tory. And while I agree that it may be difficult to find French equivalents
for above-the-line and below-the-line in advertising, l I have to admit that
I find sentences such as "Cette annonce etait munie d'un headline, d'un
claim au-dessous du pack-shot [?] et a cote d'un mini-copy remplace, Ie
plus souvent, par un base-line souvent considere, ason tour comme toujours
trop long" (p. 98) rather difficult to digest. Without condoning the eccen-
tricities of Mr Toubon's law,2 one cannot but wonder whether the need
BOOK REVIEWS 109
for translation into French will eventually be eroded altogether, given that
so many French publicists use English expressions so frequently that even
a French-speaking writer concerned with the quality of parole does not
find it necessary to translate them. It is true, however, that not all transla-
tion bureaux can afford to employ terminologists and that borrowed
expressions and neologisms abound (p. 57). But perhaps the authors could
have given us some advice on what assistance is available in terms of data
banks and other computer facilities; as it is, they only make a brief refer-
ence to SYSTRAN in an appendix.
The authors devote part of the book to discussing the importance, or
otherwise, of using technical texts in training translators. This has, of
course, been the subject of endless debate in the literature and among
professionals. It could be argued that what is important in training is to

cultivate a specific approach to translation per se, an ability to digest the
content of the original and then stand back from it in order to express this
content in the appropriate target language form. This can be done using
any kind of text: political, economic, social or technical. Generally speak-
ing, highly technical texts leave less room for creativity and therefore
represent an essential but different part of the training process; at least at
the Ecole d'lnterpretes Internationaux, where this reviewer works, they
are usually introduced in the later stages of a four-year degree course. I
imagine the authors would not disagree with this general approach, espe-
cially since creativity is clearly more important than accuracy when it
comes to translating the kind of advertising material they seem to be most
concerned with.
A chapter on interpreting follows those on translation and points out
that much of what has been said in terms of the written word is also valid
for the spoken word. The importance of written documents, both those
which have to be studied before a conference and those which result from
it, is discussed. There is also a useful check list of various aspects of
conference organization, including such things as booking the right room
and hiring interpreters.
In Chapter Seven, the authors spend sometime defending what is es-
sentially a political viewpoint, namely that held by supporters of parties
with a neo-liberal economic policy, such as the Conservative Party in Brit-
ain or the PRL in Belgium. Whether or not this reviewer, or any reader,
happens to share this viewpoint, the relevance of many of their remarks in
a book about translation must be questioned. It is all very well to say that
translators and translation bureaux in the private sector have to work long
hours for minimal reward; pretty well everyone in the public sector has
had to get used to reduced financial rewards in recent years, and lecturers,
who are of course state employees, are also feeling the pressure of increased
BOOK REVIEWS 110
teaching hours and reduced earnings. Having expressed this overtly po-
litical viewpoint themselves, the authors nevertheless go on to tell us
that "a firm must abstain from all political activity in favour of whichever
party".3
This is, however, a useful and well-written book overall. The presen-
tation is attractive and the text easy to read, with the exception of the
unhelpful use of anglicisms. There are short summaries provided in the
margins to help the reader locate specific sections more quickly or, alter-
natively, skip those in which slhe is not particularly interested. But do not
look for theoretical explanations of the mental processes involved in read-
ing, understanding, interpreting or writing; that is not what the book
attempts to provide. It is essentially a practical book. I would recommend
it to anyone who wants to know about the nitty -gritty of professional trans-
lation and interpreting; and that is, after all, what the authors set out to
describe.
DAVID MORRIS
Ecole d'Interpretes Internationaux, Universite de Mons-Hainaut, Avenue
du Champ de Mars, B-7000 Mons, Belgium
Notes
1. However, I found excellent definitions of these expressions in the useful

glossary which opens Part Eight of this book.
2. A French law which bans the use of foreign terms in official texts.
3. " ... l'entreprise doit s'abstenir de toute activite politicienne de soutien a
l'un ou l'autre parti" (p.158).
Publishers of Books Reviewed in this Issue:
Editions Gallimard, 5 rue Sebastien Bottin, 75341 Paris, France

Presses de l'Universite d'Ottawa, c/o Academic & University Publishers'
Group, 603 rue Cumberland, Ottawa, Ontario KIN 6N5, Canada
John Benjamins: P. O. Box 75577, 1070 Amsterdam, The Netherlands
The Kent State University Press, P. O. Box 5190, Kent, Ohio 44242-
0001, USA
Telos Communication Group Editions: 197 Avenue de Tervueren, 1150
Brussels, Belgium
Tothemes Diffusion, B. P. 60, 26 avenue de l'Europe, 78141 Velizy,
Cedex, France
This article was downloaded by: [University Library Utrecht]
On: 16 March 2015, At: 18:05
UK
The Translator
MA/Advanced Diploma in BSL/

English Interpreting
a a
Mary Brennan & David Brien
a
Deaf Studies Research Unit, University of Durham, UK
To cite this article: Mary Brennan & David Brien (1995) MA/Advanced
Diploma in BSL/English Interpreting, The Translator, 1:1, 111-128, DOI:
10.1080/13556509.1995.10798953
Course Profile
MNAdvanced Diploma in BSL/English Interpreting
MARY BRENNAN & DAVID BRIEN

Deaf Studies Research Unit, University of Durham, UK
Abstract. Differences between signed and spoken language and be-

tween Deaf and hearing communities may affect the nature and
practice of sign language/spoken language interpreting and hence
the nature of interpreter education. The University of Durham post-
graduate courses on interpreting between British Sign Language
(BSL) and English are profiled in this article against the background
of debates and changes occurring within the sign interpreting pro-
fession and the Deaf community in Britain.
Profiling any course places it in a particular era and context. The diffi-
culty of profiling the MA and Advanced Diploma courses in BSL/English
Interpreting offered by the Deaf Studies Research Unit (DSRU) at the
University of Durham is that we are currently going through a period of
change and re-examination of central issues. 1 This change is reflected di-
rectly in the fact that the courses profiled here have recently been reviewed
and re-structured: the new courses are outlined with a view to offering
some insights as to why changes have been introduced. The re-thinking
that is taking place extends beyond the university context. Central to the
debate are changes occurring both within the Deaf community and within
the BSL/English interpreting profession itself.
1. British Sign Language and the Deaf community in Britain
BSL is a naturally evolving human language which has probably existed

in some form wherever Deaf people have come together in communities
within the UK. Historical information is scant, but Sutton-Spence (1994)
provides a fascinating example of the reality of some form of signing
described in the record of the marriage of a Deaf person in England in
1575. There is certainly plenty of evidence from accounts relating to other

Durham: MAIAdvanced Diploma in BSL/English Interpreting 112
countries that wherever Deaf people come together some kind of sign
communication is likely to develop (Lane 1984). Of course, as with the
beginnings of English, the lack of contact between different Deaf groups
suggests that it is likely that the precursors of present-day BSL were prob-
ably quite distinct varieties which mayor may not have been mutually
comprehensible. Moreover, unlike English, there is no generally accepted
written form of BSL. The lack of a written form has meant that several of
the processes of standardization which have operated with respect to Eng-
lish have not occurred in the same way for BSL. Nevertheless, some degree
of standardization has developed in more recent times under the influence
of such factors as the increased use of BSL on television and within fur-
ther and higher education, the increased mobility of members of the Deaf
community, the growth in the numbers of Deaf people teaching BSL and
the recent publication of the first BSL/English dictionary (Brien 1992).
In order to fully understand the relevant issues and their impact on the
nature of the language, we need to recognize a number of salient factors
about the Deaf community now and in the past. The term 'Deaf commu-
nity' itself is relatively new. There has been a tendency amongst hearing
people to think of deaf people in individual terms, that is as individuals
with a hearing problem. Many Deaf people reject this 'pathological' view
and instead see themselves as constituting a specific linguistic minority.
The convention of using a capital 'D' in Deaf is now commonly adopted to
distinguish those who regard themselves as belonging to a linguistic and
cultural community from those (lower case) deaf people who have a hear-
ing impairment but do not see themselves as having any linguistic
allegiance with members of the BSL Deaf community.
Perhaps one of the most surprising facts, for those not familiar with
the reality of Deaf life, is that the majority of deaf children are born to
hearing parents. Even though Deaf adults are more likely to have Deaf
partners than hearing partners, they, in turn, are more likely to have hear-
ing children. Thus, although a relatively small proportion of Deaf adults
come from families with a long history of deafness and can trace their
Deaf ancestry through several generations, this is not the case for the ma-
jority of Deaf people. This clearly has repercussions for the nature of the
Deaf community itself. In particular, it means that for most Deaf people,
British Sign Language and Deaf culture are not transmitted through their
immediate genetic family. There may well be an ongoing tension between
commitment to one's biological family and commitment to the Deaf com-
munity. As one of our Deaf colleagues in the DSRU has written of her
childhood, "I was sad when the summer holidays were near because I
dreaded being away from my Deaf family in the school who used the only
language I knew - BSL. I love my genetic family, but Deaf people taught
Mary Brennan and David Brien 113
me far more" (Reed 1994:278). This tension is a microcosm of the wider

one, felt throughout the Deaf individual's life, between the Deaf commu-
nity and the demands and pull of the larger hearing community. In terms
of BSL, the lack of direct linguistic transmission means that age of expo-
sure to and age of acquisition of BSL vary considerably, depending on
circumstances. This, combined with the limited educational use of BSL,
particularly during primary and secondary schooling, means that there is
considerable variation in the nature and use of BSL within the Deaf com-
munity as a whole. Fluency in productive morphology, for example, may
vary, depending on age of acquisition. Similarly, the influence of English
on the signing variety may vary depending on the degree of access and
use of the two languages. Such variation clearly has practical implica-
tions for the practising interpreter.
An interpreter entering into the Deaf world needs to understand the
nature of such tensions and also what some would characterize as the
colonial-like domination of the Deaf community by hearing people (Lane
1992). There is a widespread belief within the Deaf community that the
language of Deaf people, i.e. BSL, has been oppressed and denigrated,
with a consequent oppression and denigration of Deaf people themselves.
The last two decades have seen Deaf people in the UK becoming much
more conscious and proud of their linguistic heritage. This pride in their
language and their community has led to a demand for linguistic rights: a
demand to be treated as a linguistic minority. One element in this has
been a focus on the right to have appropriate interpreting services avail-
able to Deaf people. Similarly, there have been increasing demands for
bilingual (i.e. BSL/English) education for deaf children, a demand which,
if taken seriously, would require Deaf people themselves taking on an
active and equal role within the education process. In fact, while there has
been some progress in the attainment of Deaf rights, many of the central
demands of the Deaf community remain unmet. Although it is now at
least twenty years since a linguistic rationale was proposed as the basis of
the deaf child's right to develop BSL as her/his first language (Brennan 1975),
only a small proportion of young deaf children are given access to BSL
soon after diagnosis of their deafness. BSL is still rarely used as the me-
dium of education within schools, although there are notable exceptions
to this (Smith 1994, Keir: Forthcoming). Deaf people wishing to train as
teachers of deaf children still face major obstacles. However, there has
been an improvement in access through BSL within further and higher
education, which has, in turn, led to an increase in the numbers of, and
need for, educational interpreters.
It could be argued that the disempowerment of Deaf people in the past
requires interpreters not simply to act as neutral professionals but to take
on an empowering role. Within the USA, for example, it has been sug-
gested that interpreters should be allies of Deaf people in their search for
equal and adequate access (Philip: Forthcoming). Scott Gibson (Forth-
coming) has argued that the professionalization of BSL/English interpreting
has not been without cost. In particular, she warns of the potential for a
growing divide between interpreters and members of the Deaf commu-
nity. Similarly, Forsman (Forthcoming) resists the imposition of a single
model of the interpreter, deriving primarily from North American and
northern European cultures, on other communities and cultures. She sug-
gests that the notion of interpreter as 'neutral facilitator' may be inappropriate
within contexts where Deaf people have made limited progress towards
equality of access and opportunity. It may be that as we go through a
transitional stage, interpreters will actually be receiving conflicting mes-
sages about what their role should be from within the Deaf community
itself, as well as from the different groups, both Deaf and hearing, they
are trained to serve. Interpreter education must address the issues which
define this debate and ensure that they are adequately incorporated within
training.
2. The visual nature of BSL and its functional implications
In addition to the realities of the Deaf community, there is also the ques-
tion of whether the nature of BSL itself requires a very different kind of
training from that normally offered to spoken language interpreters. In
particular, can the units and elements of competence proposed for spoken
languages be directly applicable to BSL? At first sight, the answer would
seem to be a resounding Yes. After all, members of the Deaf community
and many people within the fields of Deaf studies, sign linguistics and
sign language interpreting have been stressing for years that BSL should
be considered a language on a par with the other languages currently taught
in our schools and universities. However, some of us find ourselves also
wishing to stress those particular aspects of BSL which may render the
activity of BSL/English interpreting rather different from the activity of
French/English interpreting. Almost inevitably, we are focusing here on
the modality difference.
BSL is a visual-gestural language: it is produced by movements of the
hands, face and body and perceived through the eyes. The extent to which
this particular modality has influenced the nature and structure of BSL
and other sign languages is part of an ongoing debate within sign linguis-
tics about the effects of modality on structure. Many sign linguists are
keen to stress that, despite the modality difference, sign languages have
the same underlying rules of structure as other languages. However, some
of our own recent work (Brien 1992, Brennan 1992, Brennan et al. 1993)
has stressed the importance of recognizing the linguistic and practical
implications of the visual nature of BSL. We have argued, for example,
that BSL encodes visual information as a matter of course. Such informa-
tion is built into verb morphology, spatial/linear syntax and discourse.
Given that we live in a world which is more full of shapes and visual
patterns than it is of sound and sound patterns, it is not surprising that a
visual language incorporates spatial and dimensional information into key
areas of its structure. Whether this should be seen as relatively trivial in
terms of the underlying linguistic structure of BSL remains unclear. How-
ever, we would argue that it is not trivial with respect to interpreter
education. The signer in a court case involving a stabbing would nor-

mally produce a sign for stabbing which indicates the way the stabbing
took place, for example with what type of force and on which part of the
body. Similarly, in discussing a car crash the signer would automatically
indicate the nature of the crash, i.e. which parts of the vehicles collided.
This type of information would be incorporated because the Deaf signer
sees it as appropriate to include visual information and, more importantly,
the language lends itself to it. The implications of this practice in a court-
room situation have to be addressed in interpreter education. The other
side of this is that an English interpretation which includes all the visual
information that naturally occurs within a BSL text may appear cumber-
some and overladen. Specialized training is required to enable interpreters
to recognize when visual encoding can raise particular problems for the
translation or interpreting process.
The fact that BSL makes use of a visual-gestural modality also leads
to a functional difference in the activity of BSL/English interpreting. Be-
cause we are dealing with two separate modalities, spoken and sign, it is
possible to express these two modalities simultaneously. It is therefore
the norm for BSL/English interpreting to be simultaneous rather than se-
quential. For spoken language interpreters, this is often only the case in
formal conference settings, where special facilities such as interpreting
booths and headphones make it possible for simultaneous interpreting to
take place. However, BSL/English interpreters undertake such simultane-
ous interpreting even within liaison settings such as doctors' surgeries,
employment interviews, tutorials, and so on. It is important to understand
how this change in the expected function of an interpreter can affect the
nature of the interaction. It is quite common, for example, for the hearing
persons present in an interaction to completely ignore the Deaf person
signing and to focus their eye gaze solely on the interpreter who is simul-
taneously speaking the message. Inevitably, this can affect the dynamics
of the interaction and can lead to the interpreter feeling pressurized into
Durham: MAIAdvanced Diploma in BSLIEnglish Interpreting 116
acting for the Deaf person. This is less likely to happen in a situation
where one first sees one speaker presenting (in a language one does not
understand) and this is then followed by the translation in one's own lan-
guage. It is perhaps this situation more than any other which has led sign
language interpreters in particular settings to have more responsibility
placed upon them than would normally be expected. Thus, educational
interpreters often find that they are expected to take on a more direct,
often clarifying, role within the interaction rather than simply working as
neutral mediators between two languages. Rather than ignoring the real-
ity of such interpreting experience, we must explore how such situations
have developed in order to allow both Deaf people and interpreters to

arrive at an informed view of the process in which they are taking part.
It is also worth pointing out that the use of simultaneous interpreting
across two modalities places the sign interpreter in a much more vulner-
able position than other interpreters. Those who know the two languages
concerned and have direct access to them (particularly hearing people who
know both BSL and English) can (and often do) assess the interpreter's
performance. Whilst this can be helpful at times, it can also be extremely
nerve-wracking: errors can be obvious and dramatic. Rather than being
corrected within the privacy and anonymity of the interpreting booth, they
are available for all to see. BSL/English interpreters have to be trained to
recognize the demands of this type of situation.
The lack of a written form of BSL also has a number of practical ef-
fects with regard to the interpreting/translation process. There are two
ways of defining the notion of translation within the context of BSL and
English. One is to regard translation as occurring wherever there is a writ-
ten text involved. Thus, translating from a written English text to BSL
would be one kind of translation and going from BSL to a written English
text would be another. In both cases, the written language concerned is
English. However, the other element normally involved in translation, as
opposed to interpreting, is the notion that the translator has time and re-
sources to work through several different versions towards a final 'polished'
text in the target language. But for those working from an English text to
a finely tuned BSL translation, the activity of going through different
versions is fraught with difficulty. Inevitably, the translator uses video
tape versions of the signed text, checking through each one for possible
changes and amendments. This in itself is more time-consuming and on-
erous than checking a written version. Difficulties are exacerbated when
the signed draft has to be produced in a final form. There is a very strong
likelihood, given the nature of the modality, that there will always be
some slight changes to what was presented in an earlier draft. In other
words, there is an inherent difficulty in reproducing a signed text several
times in exactly the same form. The only way we can get an inkling of the
difficulty involved is to imagine reproducing a piece of spoken language
in exactly the same form several times without aid of the written language.
Specialized training is clearly required in the specific techniques and strate-
gies which can be deployed to ensure the success of such a process.
It could be argued that translation itself has not received sufficient
attention either in our own or other courses on sign interpreting. We are
beginning to examine whether we should establish a separate translation
pathway and/or whether we should change the title of our present courses
to BSL/English Interpreting/Translating. Currently, we aim to include the
activities of interpreting, relay interpreting and translating at every level
of the course. Sign interpreters are frequently given written texts from
which to work, although it is not always the case that a prepared transla-
tion, as opposed to an 'on the spot' interpreting job, is required. Thus,
interpreters are often given papers prior to conferences which they use for
preparation, even though they know that the presenters are more likely to
'talk to' their papers than read them directly. However, translation is be-
ginning to assume much more importance within the profession for two
reasons. First, it is an area in which Deaf people themselves can play an
important part. Indeed we are seeing more Deaf people becoming actively
involved in producing translations of English texts. Given that Deaf peo-
ple are working in this case from English into their own preferred language,
BSL, it is likely that they can produce a more effective translation, either
working alone or with a hearing translator. Second, the growth of televi-
sion for Deaf people has meant that there is more opportunity for Deaf
individuals to be involved in the translation of prepared scripts for later
transmission. We see it as an essential part of the empowering process to
recognize that Deaf people themselves have a right to training as translators/
interpreters.
One further kind of activity which is not usually considered within
spoken-language interpreting but is beginning to be used in BSL/English
contexts is relay interpreting. This is a process whereby a hearing inter-
preter takes the English message from an English speaker and relays this
message in a signed form to a Deaf interpreter. The hearing interpreter
usually relays the message using what is often termed 'Sign Supported
English': essentially the message is conveyed in English but through a
visual medium. The Deaf interpreter, who now has access to the English
(which s/he cannot hear), then interprets this signed form of English into
BSL. Again, the main advantage of this form of interpreting is that it
allows the individual to interpret into his/her own language. The skills of
the Deaf signer can be exploited to the full and the message conveyed
more effectively to the Deaf individual or group. The use of relay inter-
preting has so far been fairly limited, but when it has been used the feedback
from Deaf people suggests that they themselves tend to prefer interpreta-
tion into their own language by a Deaf interpreter, though this of course
also depends on the skills of the individual interpreter.
There are particular difficulties in relation to using Deaf relay inter-
preters. The main problem is the additional cost of employing two people
when the employer may well feel that one person can do the job. Again,
there are political issues here relating to empowerment and right of access
to employment opportunities. The Durham group have taken the view
that Deaf people should be encouraged into the profession, but we also
recognize that it would be unethical to raise false hopes and to imply that
a Deaf interpreter would currently have the same job opportunities as a
hearing interpreter. If the focus is widened to include prepared interpret-
ing and translation, and interpreting directly between different sign
languages (such as British Sign Language to American Sign Language),
then we can certainly envisage situations where more and more Deaf peo-
ple can be employed.
3. The Durham courses
Training in BSL/English Interpreting is offered at Durham at two levels

of qualification: Advanced Diploma and MA. Students follow the same
route for the first set of modules, but those registering for the MA must
take additional modules and pass all of their assessed work at a higher
level than the Advanced Diploma students. The courses are modular in
nature, with each module having a specific University Credit Unit (UCU)
value. To date, the courses have been offered on a part-time basis, but it is
anticipated that they will be offered on a full-time basis within the next
two years.
Interpreting courses are offered within the context of a wider programme
of postgraduate degrees, namely 'Teaching of Sign Language', 'Sign Lin-
guistics' and 'Deaf Studies'. Thus, for instance, interpreting students take
the first module, Language, Community and Culture, alongside students
who may be working towards one of the other three degrees. MA students
are required to obtain 120 University Credit Units and Advanced Diploma
students 80 Units. The first 80 Units are obtained through obligatory course
components, which means that Advanced Diploma students have no choice
in what they do. MA students can obtain the further 40 UCUs either by
undertaking research leading to a dissertation or by taking two optional
modules (see 3.2 below). An MA student may also choose to do one taught
module and one independent project. The latter option allows the student
to take an area of hislher choice and negotiate a specific topic for inde-
pendent work. In all of these options, the student is expected to relate the
work directly to the practice of interpreting. Even if a student chooses a
relatively theoretical dissertation topic, it is understood that consideration
will be given to the potential practical and professional implications of
such research.
3.1. Obligatory modules
The obligatory components of the course comprise three taught modules

and one practicum. Each taught module involves forty-eight hours of con-
tact time. The teaching on the part-time course is currently delivered in
three-day blocks. Additionally, students are expected to undertake a se-
ries of tasks, some of which will count towards their final assessment,
and make use of video and audio resources to improve their skills. They
are also expected to familiarize themselves with the relevant interpreting,
linguistic and professional literature. The practicum consists of a moni-
tored and assessed set of interpreting tasks which take place over a specified
period. The taught modules are: (a) Language, Community and Culture,
(b) Language Structure and Skills, and (c) Principles and Practice ofSign
Language Interpreting. A selective listing of texts appears after the de-
scription of each module to give an indication of the type of reading and
support materials used and recommended to students.
3.1.1. Obligatory module: Language, Community and Culture
This module examines the nature and structure of Deaf communities, with
particular focus on the British Deaf Community. Starting with the 'Deaf
experience', the module goes on to examine theories of the social con-
struction of deafness, both internally, from within Deaf communities, and
externally, in relation to how such communities are viewed by members
of hearing communities. The so-called 'pathological' model of deafness
is contrasted with the more recent linguistic/cultural models. Specific
models of Deaf culture are described and critiqued. There is also an intro-
duction to the complex linguistic situation of Deaf people, including
examination of such linguistic models as the BSL-English continuum,
diglossia, and bilingualism. The module makes considerable use of input
from Deaf staff, Deaf people on video and the writings of Deaf people.
Assessment is based on one essay of 3000 words or 30 minutes of con-
tinuous signing, plus two written/signed tasks each equivalent to 1500
words or 15 minutes of continuous signing.
Durham: MAIAdvanced Diploma in BSLIEnglish Interpreting 120
Indicative texts: Gregory & Hartley 1990; Lane 1992; Padden & Humphries
1988; Turner 1994.
3.1.2. Obligatory module: Language Structure and Skills
This module provides a descriptive and comparative account of key ele-

ments of the structure and functions of British Sign Language and English.
It examines some of the major differences between a language produced
in the visual-gestural modality and one expressed primarily within the
oral-aural modality. These differences are explored within the areas of
phonology, morphology, syntax and semantics. Particular attention is paid

to areas which are known to be especially demanding within interpreta-
tion, such as the use of role-shift, topic-comment structures and productive
morphology in BSL, and appropriate rhythm and stress patterning, fea-
tures of cohesion and register choice within English. Students are expected
to familiarize themselves with the details of the structure of both lan-
guages so as to be able to discuss interpreting skills in a knowledgeable
way. The module includes practical exercises with respect to the major
areas of both languages as a basis for moving on to interpretation from
one language to the other. Assessment is based on one essay of 3000
words or 30 minutes of continuous signing, plus four practical linguistic
tasks equivalent in total to 3000 words or 30 minutes of signing.
Indicative texts: Ahlgren et al. 1994; Brien 1992; Halliday & Hasan 1976;
Schiffrin 1994.
3.1.3. Obligatory module: Principles and Practice of Interpreting
This module examines the complex cognitive and linguistic processes in-
volved in moving from the source to the target language during
interpretation. Students are introduced to theoretical accounts of such as-
pects as attention, visual and auditory memory, message analysis,
pragmatics and discourse analysis. There are opportunities to develop ap-
propriate strategies relating to these areas through a range of practical
techniques and exercises, including self-assessment techniques. There is
also an opportunity to discuss and apply practical solutions in the areas of
(potential) cultural conflict and cultural mediation. In addition, the mod-
ule examines the cultural, professional and ethical aspects of interpreting,
including differing analyses of the role of the interpreter, for example as
an impartial but competent mediator and/or as an advocate for Deaf peo-
ple. Assessment is based on one essay of 3000 words or 30 minutes of
continuous signing, plus four practical interpreting/translating tasks equiva-
lent in total to 3000 words or 30 minutes of signing.
Indicative texts: Brien et al.: Forthcoming; Frishberg 1990; McIntire 1986;

Scott Gibson 1991.
3.1.4. Obligatory practicum
All students are required to undertake a set of monitored and assessed

interpreting tasks. The precise programme of tasks is negotiated with the
student in the context of the individual's personal situation. However, the
programme of tasks requires students to have experience in a range of
interpreting settings and situations. These must include (a) liaison inter-
preting, for example, within a medical context, (b) group contexts, such
as a meeting or tutorial, and (c) formal settings or other contexts where an
audience is involved, such as a lecture. Students must demonstrate inter-
preting skills in both linguistic directions, i.e. both from BSL to English
and from English to BSL. In addition, they choose two particular inter-
preting tasks which they feel best suit their skills. The monitored tasks
therefore include:
• BSL to English: Liaison • BSL to English: Formal

• English to BSL: Liaison • English to BSL: Formal
• BSL to English: Group • Optional
• English to BSL: Group • Optional
These tasks may be videotaped for viewing later by course staff. However, at
least four of the tasks must be monitored and assessed in situ. In order to
allow some progression within the practical component, the student and
tutor choose four out of the eight tasks to count towards the final grade.
The programme of practical tasks is carried out over a specific time period.
3.3. Optional modules
There is an extensive list of optional modules available for various tracks

of the MA programme, including modules on such topics as Curriculum
Development, Language Assessment and Sign Language Acquisition, in
addition to four specific modules on interpreting. Students on the BSLI
English Interpreting degree are expected to choose two of the options
described below, but they may, with the agreement of the department,
choose a non-interpreting option provided they can justify their choice in
terms of their individual career needs. For example, a student wishing to
work as an educational interpreter may take a module entitled Sign Lan-
guage in Education. Due to lack of space, only the interpreting options
are discussed here.
Durham: MA/Advanced Diploma in BSL/English Interpreting 122
3.2.1. Optional module: Educational Interpreting
This module examines the theory and practice of educational interpreting.

It explores the way in which the concept of educational interpreting has
been developed in other countries, especially in those countries - such as
(parts of) the USA, Sweden, Finland and Denmark - where educational
interpreters are used extensively in the education of deaf children. This
development is compared with traditions of educational interpreting and
communicative support within educational environments for dlDeaf chil-
dren and students in the UK. The roles of the BSL/English educational
interpreter and communication support worker are explored from the per-
spectives of the teacher/lecturer, student, interpreter and communication
support worker. The particular professional, communicative and linguis-
tic demands of this type of interpreting are examined with reference to
authentic examples. Assessment is based on one essay of 3000 words or
30 minutes of signing, plus two practical educational interpreting tasks.
Indicative texts: Green: Forthcoming; Hayes 1992; Patrie: Forthcoming;

Winston 1992.
3.2.2. Optional module: Legal Interpreting
This module examines the theory and practice of sign language/spoken

language interpreting within legal settings, focusing in particular on BSL/
English interpreting. The different contexts in which legal interpreting
may take place, for example within courtrooms or police stations, and the
various interpreting demands of these settings are explored. Special atten-
tion is given to the nature of legal language and of courtroom interactions.
Students examine the potential cross-linguistic and cross-cultural clashes
which may occur and explore authentic linguistic dilemmas and their possi-
ble solutions. Such dilemmas may include inherent problems in interpreting
between a visual-gestural language, which encodes visual spatial infor-
mation as a matter of course, and an oral-aural language which only
optionally includes such information. Other dilemmas may relate to diffi-
culties in relaying attitudes conveyed by such language-specific elements
as address systems, politeness strategies, and hedging and foregrounding.
The ways in which interpreters can mediate cross-cultural information are
analyzed, and the appropriateness of adding or excluding such informa-
tion is discussed. Assessment is based on one essay of 3000 words or 30
minutes of signing, plus two practical legal interpreting tasks.
Indicative texts: Berk-Seligson 1990; Keg! & Turner: Forthcoming; Nuffield

Interpreter Project 1993; Turner & Brown: Forthcoming.
3.2.3. Optional module: Medical Interpreting
This module examines the theory and practice of sign language/spoken

language interpreting within medical and health settings, with particular
attention to BSL/English interpreting. The various contexts in which medi-
cal interpreting may take place (for example within surgeries, clinics,
hospitals and health groups) and the different interpreting demands of
these settings are explored. Special attention is given to the nature of pro-
fessional/patient exchanges and how these are mediated and affected by
the presence of an interpreter. The particular demands of medical lan-
guage are explored in practical sessions, with students becoming more
familiar with exploiting the productive lexical resources of BSL. Students

are encouraged to appreciate the particular personal and emotional de-
mands of interpreting with different types of client/patient, including
individuals with HIV, cancer and mental illnesses. There is also some
consideration of the special requirements of health education contexts,
such as those relating to women's health, men's health, healthy eating,
and fitness. Students examine the potential cross-linguistic and cross-
cultural clashes which may occur and explore authentic linguistic dilemmas
and their possible solutions. The various ways in which interpreters can
mediate cross-cultural information are analyzed, and the appropriateness
of adding or excluding such information is discussed. Assessment is based
on one essay of 3000 words or 30 minutes of signing, plus two practical
medicallhealth interpreting tasks.
Indicative texts: Beck: Forthcoming; Collins: Forthcoming; Denmark 1994;

Wadensjo 1992.
3.2.4. Optional module: Issues in Interpreting
This module examines a range of controversial issues relating to the theory

and practice of sign language/spoken language interpreting, including: the
role of the interpreter and changing perceptions of that role; the relation-
ship between the interpreter and other professionals such as social workers,
teachers and lawyers; problems of role change with the same person tak-
ing on more than one role, for example interpreter and social worker,
interpreter and teacher; the nature of the interpreter's relationship with the
Deaf individual, for example as ally, advocate or impartial mediator; the
relationship between the interpreter and the Deaf community, for example
as active member, fringe participant or detached outsider. Some considera-
tion is given to the way in which the provision of interpreting services has
been affected by wider policy changes influencing the individual policies
Durham: MA/Advanced Diploma in BSL/English Interpreting 124
of voluntary organizations and general provision of interpreting services

at a local and national level. The development of new designations or
professional career routes for sign interpreters is examined. The potential
for conflict between traditionally accepted ethical principles and required
professional practice is discussed within the context of specific types of
setting and in the light of new types of employment patterns for interpret-
ers (for example freelance and contract interpreting). Assessment is based
on one essay of 6000 words or 60 minutes of signing.
Indicative texts: Goldfinch: Forthcoming; McIntire 1986; Philip: Forth-

coming; Scott Gibson 1991.
4. Issues in the Durham experience: where do we go from here?
In establishing an interpreter training programme, we have had to take into

account the situation and background experience of potential trainees, both
new recruits to the profession and those who have been interpreting for
years. When the Durham University courses were first developed, many
people in BSL/English interpreting had not undergone any approved course
of training, though some had been taught BSL explicitly and had a con-
scious understanding of the differences between English and BSL. Others
did not have any academic background in the study of the language and
therefore no metalanguage for discussing BSL. Similarly, qualifications
in English, even of those already undertaking interpreting work, varied
enormously. While groups such as our own have stressed the importance
of ensuring that BSL/English interpreters obtain a standard of compe-
tence and professionalism comparable to that expected for spoken language
interpreters, there have been very strong pressures to develop short, quick
courses which could ensure a fast supply of 'trained' interpreters. We
have resisted such pressures but, inevitably, other types of training have
been developed, usually outside the higher education system.
At this point, it may be worth outlining some of the difficulties we
have faced in developing our courses. The first persistent problem is lack
of funding and what might be termed a lack of embedding within the
system. The courses are run by the Deaf Studies Research Unit within the
Department of Sociology and Social Policy. The DSRU itself does not
receive any core funding. All staff, including senior and highly experi-
enced staff, are on short term contracts. This means that there is a constant
requirement to seek funding to ensure the continuation of both the DSRU
itself and the teaching and research it sustains. The situation is made worse
by the fact that part-time students are not eligible for the main sources of
postgraduate funding. Moreover, because we are dealing with an emerg-

ing profession, not all employers accept a responsibility to train. Students
who attend a course such as ours have usually already made a commit-
ment to high standards. In a sense, their very commitment to high standards
disadvantages them in that they typically have to pay their own fees and
have to use annual holidays to complete training.
A second difficulty is the paucity of interpreter educators with appro-
priate qualifications, experience and skills. We recognize the need to
include competent interpreters as part of our teaching and assessment team
and currently have two full-time and one part-time member of the team
with relevant background in interpreting and/or translating. 2 Both of the
full-time members (one Deaf and one hearing) have worked within televi-
sion and are familiar with a range of different types of translation and/or
interpreting demands. We have been fortunate in having an Honorary Fel-
low and part-time member of staff who is also President of the European
Forum of Sign Language Interpreters and Chairperson of the Scottish
Association of Sign Language Interpreters. We have also been able to
secure funding to obtain the services of Visiting Fellows from the USA,
Denmark and Greece to contribute to the educational programme. Never-
theless, we are very much aware that it is necessary to involve additional
interpreters in the training and assessment programme. However, very
few people have the combination of interpreting skill and academic back-
ground which is normally required. On a positive note, we see these courses
as actively encouraging new interpreter trainers in that it is already clear
that some participants are keen, eventually, to move into training.
A third difficulty relates to the gap between entry qualifications and
experience and the level of the awards. Most entrants to the BSL/English
interpreting courses enrol as Advanced Diploma students. The University
allows a number of such students to enrol without the traditional entry
requirement of a degree: other professional qualifications and/or experi-
ence may be deemed sufficient. The positive side of this is that a number
of such people do make excellent progress and achieve high professional
and academic standards. The negative side has been that others may strug-
gle with both the level and amount of work required. Most students work
as full-time interpreters at the same time as studying for their degree, with
all the additional pressure that entails. Because the courses have been part-
time and the recruitment national, it has not been feasible to provide the
consistent tutorial and advisory support that might enable these individu-
als to work through the courses more successfully. This difficulty is
compounded by the lack of trainers mentioned above. The newly revised
course structure and pattern of delivery (see section 3) aims to address
this problem: by seeing students on a more regular basis, it is hoped that
we can overcome the problem of distance.

A further difficulty is the additional costs involved as a result of deal-
ing with a visual-gestural language. It is essential that during the teaching
we make use of video tapes and video equipment. Similarly, students
need to have access to video cameras and playback facilities. The assess-
ment of student work is more time-consuming, again because we are
dealing often with video recordings. In order to allow the Deaf members
of the team to contribute adequately to the advice and support we give to
students, it is often necessary to provide written transcripts of English
material. This is all feasible and appropriate, but it inevitably raises the
costs in terms of tutors' time and the additional facilities required.

Sign language interpreting is a relatively new field, with the conse-
quence that there is not a wide range of appropriate teaching and assessment
materials available. Inevitably, we have spent a considerable amount of
time developing such materials ourselves. The difficulties and demands
of developing appropriate assessment schedules would require a further
article: suffice it to say that such work is ongoing. Our students do face
some difficulty in obtaining access to the appropriate literature. Because
the courses are part-time, it is more difficult for them to make use of the
resources of the Durham University library. Additionally, as the DSRU is
not funded directly by the Higher Education Funding Council of Eng-
land, it has no automatic right to a library budget. We do of course try our
best to make resources available to students, but the lack of embedding
and funding does make the work of mounting such courses more onerous
than it should be.
Despite all the difficulties involved, the Durham courses appear to be
providing an opportunity for BSL/English interpreters to be educated to
an appropriate standard. We are seeing students actively contributing to
the academic and professional foundations of their profession. Last year,
we organized an international conference entitled Issues in Interpreting
and were delighted that students could take their place quite appropriately
amongst international and distinguished presenters. We don't have all the
answers, but we are delighted to be part of the process, with our students,
of trying to ask the right kind of questions and to work towards achieving
the right kind of solutions. The fact that the courses exist at all is a dem-
onstration, we hope, of the potential of collaboration and the emergence
of genuine empowerment - both for Deaf people and for interpreters.
MARY BRENNAN AND DAVID BRIEN

Deaf Studies Research Unit, University of Durham, Elvet Riverside 2,
New Elvet, Durham DR1 3JT, UK
Notes
Mary Brennan and David Brien are Co-Directors of the Deaf Studies Research
Unit and the MA and Advanced Diploma programme in Deaf Studies and Sign
Language Studies. The authors would like to acknowledge the contribution of
DSRU colleagues to the development and delivery of these courses.
1. The revised structure is still being processed through the University of Dur-
ham validating system. There may be some amendments to the detail of the
information presented here in the final description of the courses.
2. The Durham team consists of three Deaf and four hearing people who work
as full-time members of staff, as well as one honorary fellow, an emeritus

professor and additional part-time staff from a range of backgrounds.
References
Ahlgren, I., B. Bergman and M. Brennan (eds) (1994) Perspectives on Sign

Language Structure, Durham: International Sign Linguistics Association.
Beck, H. (Forthcoming) 'Interpreting in HIV/Aids Contexts', in Brien et al.
Berk-Seligson, Susan (1990) The Bilingual Courtroom: Interpreters in the Ju-
dicial Process, Chicago: Chicago University Press.
Brennan, Mary (1975) 'Can Deaf Children Acquire Language?', American An-
nals of the Deaf 120(5): 463-79. Reprinted in British Deaf News (1976)
February Supplement, 1-13.
------ (1992) 'The Visual World of BSL', in Brien (ed).
------ (1994) 'Pragmatics and Productivity', in Ahlgren et al.
------, David Brien, J. Collins, F. Elton and Graham Turner (1993) 'The Dic-
tionary of British Sign Language/English: A Bilingual Resource', in
Facilitating Bilingualism, Herts: LASER.
Brien, David (ed) (1992) Dictionary of British Sign Language/English, Lon-
don: Faber and Faber.
------ and Graham Turner (1994) 'Presenting Meanings in the First BSL/English
Dictionary', in Ahlgren et al.
------, Mary Brennan and L. Scott Gibson (eds) (Forthcoming) Issues in Inter-
preting: Proceedings of the Conference Held in Durham in 1994.
Collins, J. C. (Forthcoming) 'Who Needs Medical Interpreters?', in Brien et al.
Denmark, J. C. (1994) Deafness and Mental Health, London: Jessica Kingsley
Publishers.
Forsman, H. (Forthcoming) 'Imposing One-Dimensional Solutions: Profession-
alism and the Greek Situation', in Brien et al.
Frishberg, N. (1986, 1990) Interpreting: An Introduction, Silver Spring, Mary-
land: RID.
Goldfinch, A. (Forthcoming) 'Deaf People as Interpreters', in Brien et al.
Green, C. (Forthcoming) 'Interpreting in Educational Settings', in Brien et al.
Gregory, S. and G. M. Hartley (eds) (1990) Constructing Deafness, London:
Pinter Publishers.
Halliday, M. and R. Hasan (1976) Cohesion in English, London: Longman.

Hayes, P. L. (1992) 'Educational Interpreters for Deaf Students: their Respon-
sibilities, Problems and Concerns', Journal of Interpretation 5(1): 5-24.
Kegl,1. and Graham Turner (eds) (Forthcoming) 'The Bilingual, Bimodal Court-
room', Journal of Interpretation, Special Issue.
Keir, M. (Forthcoming) 'Sign Language Assessment', in R. MacAree and H.
Mitchell (eds) Effective Communication for Deaf Children: Policy and Prac-
tice, Edinburgh: Moray House Institute of Education, Heriot Watt University.
Lane, H. (1984) When the Mind Hears: A History of the Deaf, New York: Ran-
dom House.
------ (1992) The Mask of Benevolence: Disabling the Deaf Community, New
York: Alfred A. Knopf.

McIntire, M. (ed) (1986) Interpreting: The Art of Cross Cultural Mediation,
Silver Spring, Maryland: RID.
Nuffield Interpreter Project (1993) Access to Justice: Non-English Speakers in
the Legal System, London: The Nuffield Foundation.
Padden, c.A. and T. Humphries (1988) Deaf in America: Voices from a Cul-
ture, London: Harvard University Press.
Patrie, C. J. (Forthcoming) 'A Confluence of Diverse Relationships: Interpreter
Education and Educational Interpreting', Proceedings of the RID Conven-
tion, Held in Evansville, Indiana in 1993.
Philip, M. J. (Forthcoming) 'Professionalism: from which Cultural Perspective',
in Brien et al.
Reed, M. (1994) 'Review of Bertling, T. A Child Sacrificed to the Deaf Cul-
ture', Signpost 7(4): 278-9.
Schiffrin, D. (1994) Approaches to Discourse, Oxford: Blackwell.
Scott Gibson, L. (1991) 'Sign Language Interpreting: An Emerging Profession',
in Gregory and Hartley.
------ (Forthcoming) 'Open to Interpretation: the Cult of Professionalism', in
Brien et al.
Smith, S. (1994) 'BSL Assessment within a Bilingual School Programme'. Pa-
per presented at Durham University.
Sutton-Spence, R. (1994) 'BSL in 1575', Signpost 7(4): 256-7.
Turner, Graham (1994) 'How is Deaf Culture: Toward a Revised Notion of a
Fundamental Concept', Sign Language Studies 83 (Summer): 103-06.
------ and R. K. Brown (Forthcoming) 'Interaction and the Role of the Inter-
preter in Court', in Brien et al.
Wadensjo, Cecilia (1992) Interpreting as Interaction: On dialogue Interpreting
in Immigration Hearings and Medical Encounters, Linkoping: Linkoping
University.
Winston, E. A. (1992) 'Mainstream Interpreting: an Analysis of the Task'. Pa-
per presented at the Eighth National Convention of the Conference of
Interpreter Trainers, Pomona CA: The Challenge of the 90s. New Standards
in Interpreter Education.

The-Translator - Venuti Et Al

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The-Translator - Venuti Et Al

Uploaded by

Copyright:

Available Formats

This article was downloaded by: [University of Exeter]

On: 15 July 2015, At: 04:09

To link to this article: http://dx.doi.org/10.1080/13556509.1995.10798947

PLEASE SCROLL DOWN FOR ARTICLE

Translation, Authorship, Copyright

definitions of translation that are useful in challenging its present

ISSN 1355-6509 © St Jerome Publishing, Manchester

conducted by the PEN American Center indicates that most translations in

Because copyright law decisively contributes to this unfavourable

1. The current situation

Current copyright law defines translation inconsistently. On the one hand,

waffled on the definition of originality as the criterion of authorship.7

It is obvious that Greenleaf copied not only the words of Frechtman,

Although this decision linked Genet's authorship to the specific formal

specificity and permeates various forms and media.

Above and beyond the pecuniary and patrimonial right, we under-

translation sufficiently alters the form of the foreign text to be copyrightable

2. The contradictory development of original authorship

The Romantic concept of original authorship emerged relatively late in

Although copyright was vested in the medium of expression, the medium

As this passage suggests, the concept of authorship as labour invest-

authoring is a free appropriation of natural materials. And the defining

entitled to transfer their rights to booksellers. As Susan Stewart has re-

Lord Chancellor said, that though a translation might not be the

The decision wound up supporting what was "intended by the author",

dium of expression; on the other hand, the translator's form-creating la-

that was contrary to the aims of the statute (ibid.:52).

By the publication of Mrs. Stowe's book, the creations of the genius

Stowe v. Thomas in effect gave translators an exclusive copyright in their

This translation was an "original literary work" within s. 1, sub-so 1,

Although the concepts favourable to the translator seem to be in place

3. The formal basis of the translator's authorship

The history of copyright may indeed contain alternative definitions of

unstable, varying with different contexts. Hence the clothing metaphors

Lord Chancellor saw the change wrought by the translation in social

Portuguese speech delivered by a Brazilian governor to the state legisla-

He cut down the speech by about one third. He edited it by omitting

The commercial function that Byrne's translation was intended to per-

not originating in the author's personality or productive labour on raw

materials, never raw or natural, always culturally coded by previous uses,

contributions of the translator and the foreign author can be distinguished:

author's exclusive copyright in derivative works. A use of a copyrighted

ous forms of criticism", parody "can provide social benefit by shedding

revise, or create conceptual paradigms, research methodologies, and clinical

would not only be permitted to copyright the translation, as current law

3. Copyright, Designs and Patents Act 1988, sections l(l)(a), 16(1)(e),

accorded to authors under the provisions of the international copyright

27. 17 United States Code, section 107(1), (3), and (4).

------ (1995) The Translator's Invisibility: A History of Translation, London and

The Moral Dilemmas of Court

To link to this article: http://dx.doi.org/10.1080/13556509.1995.10798948

PLEASE SCROLL DOWN FOR ARTICLE

The Moral Dilemmas of Court Interpreting

Abstract. In court interpreting, the law distinguishes between the

interpretation, which involves interpreters decoding and attempting

Among members of the linguistic professions, the terms interpretation

ISSN 1355-6509 © St. Jerome Publishing, Manchester

reports of cases in which issues of interlingual interpreting were ad-

1. Interpreting as a communicative process

The contemporary view of communication, of which interpreting is but

involving (or indeed, as being tantamount to) acts of translation, whether

inevitably subject to constraints on success due to a variety of physical

Much recent work, including my own, adopts an opposing view, one

ings and may, potentially, be perceived as critical of the functioning of