Corps de l’article

1 Introduction

In this paper I would like to consider the meeting of two programmes for translation research, that of Gideon Toury and that of Antoine Berman. Toury’s Descriptive Translation Studies has been immensely influential in translation research, but does not seem to have taken off in France; this may be due to different national traditions in translational discourses or different intellectual cultures, topics which would necessitate a very broad study. I shall make a contribution to the issue by looking at the relation of one French theorist, Berman, with Toury. The study of the relationship between the two theoretical frameworks will be undertaken as a reflection on the translating and translatability of research frameworks. This is timely because the question has recently been raised in a forum in which two representatives of different ‘camps’ in translation research (empirical-descriptive and postmodernist) sought to set out “shared ground” (Chesterman & Arrojo 2000 and ensuing debate in Target 12.2, 13.1, 13.2).

2 The Translatability of Paradigms or Genres of Discourse

In order to better analyse the Berman-Toury encounter, I shall consider the ideas of three theorists who have discussed the nature of the relations between different research frameworks or discourses.

We shall start with Thomas Kuhn who has famously talked about paradigms. A paradigm is defined as a model (including laws, theory, application, and instrumentation) for scientific research from which a coherent tradition springs (Kuhn, 1996, p. 10). In general Kuhn does not think that the humanities acquire stable consensual research frameworks; rather there is a tradition of claims, counter-claims, and debates over fundamentals (Kuhn, 1974, p. 6). But we can still use the concept of paradigm for the humanities as a conceptual network which forms a coherent whole. What is most pertinent for my purposes is to look at Kuhn’s discussion of the relations between two different paradigms or theories. Kuhn says that successive theories are incommensurable. In using the word ‘incommensurability’ Kuhn wants to underscore the immense difficulty of comparison rather than the absolute impossibility of engaging with another theory.

Most interestingly Kuhn uses a linguistic parallel to discuss the issue: the translatability of paradigms is parallel to the translatability of languages. He points out first that the difficulties of learning a second language are far less problematic than the difficulties of translating between two languages. Interlingual translation is possible but it involves compromises which alter communication. In the same way scientists can engage with an incommensurable theory, but the engagement is never entirely satisfactory (Kuhn, 1974, p. 268). Translation between languages is difficult because languages cut up the world in different ways, and we have no access to a neutral sub-linguistic means of reporting. In Quine’s famous example, the native informant utters ‘Gavagai’ because he has seen a rabbit, but it is difficult to discover how ‘Gavagai’ should be translated: ‘rabbit’, ‘rabbit-kind’, ‘rabbit-part’, ‘rabbit-occurrence’ etc. (Quine, 1960, p. 51). The same situation applies to a scientist trying to communicate with a colleague who embraces a different theory. Theories embody deep differences such that the same stimuli are perceived differently; a different theory implies a different world. A rigorous comparison of the two theories is not possible because it would require a third neutral ‘language’ which doesn’t exist. It might be considered that inter-theory translation is easier than interlingual translation because the same language and in particular the same terms are being used in many cases. However, this is a trap since the meanings or conditions of applicability of terms change radically in different theoretical frameworks (Kuhn, 1974, pp. 266-277).

Communication between two people espousing different theories is nevertheless possible. This is because there are shared factors: shared stimuli, similar neural apparatus, shared history, shared everyday language, shared everyday world, and a partly shared scientific world. It is possible to use shared everyday vocabularies to elucidate communication breakdown; it is possible to learn to ‘see’ from the other’s point of view (to become ‘bilingual’), and thus to some extent translate the other’s theory and the world it describes into one’s own language (Kuhn, 1974, pp. 276-277). In fact this kind of translation is a necessary step in the process of conversion to a rival paradigm (Kuhn, 1996, p. 202).

It is on this point that one may question the applicability of Kuhnian history to the Berman-Toury situation. Kuhn’s discussion leads to the notion of scientists converting to a new paradigm which is better at solving puzzles (Kuhn, 1996, p. 206). This is not the case for our two theorists. Toury would never convert to the Berman paradigm and Berman would never have converted to the Tourian paradigm. Nor can we say that one framework is more intellectually powerful and insightful and will supersede the other. The frameworks are simply very different. What is needed, therefore, is a stronger notion of incommensurability which does not allow conversion. For this we shall look to Jean-François Lyotard.

The question of incommensurability between discourses is an essential part of Lyotard’s philosophy. Lyotard uses the terminology ‘genre of discourse’ to talk about different types of discourse which are defined by their distinct stakes or goals, for example to describe, to narrate, to persuade, to entertain. Discourses establish reality. For Lyotard the distinct goals of genres of discourse are incommensurable, and this means that there is an abyss between discourses. Upon the utterance of a phrase, genres of discourse are in competition for linking on to the phrase. There is thus a conflict between genres of discourse, and there is no tribunal, no third party (cf. Kuhn’s “third neutral language”) which can equitably judge the case — Lyotard calls this a ‘differend’. A differend is unresolvable equitably; all that can be done is to bear witness to the differend and seek a new idiom (in vain) to express it. Respecting the difference and heterogeneity of the other by bearing witness to the differend is, for Lyotard, a matter of justice. Conversely the height of injustice is suppressing the differend, which occurs when one genre of discourse is hegemonous. If one genre of discourse attempts to settle the differend by applying its own rules, or if it translates the other into its own discourse, the result will be subsumption. This is unjust since it silences the specificity of the stakes, the right to difference, of the subsumed discourse —  it causes a wrong to that party (Lyotard, 1983).

Looking more closely at Lyotard’s writings, we discover that there are in fact other relations between genres of discourse, and that the situation is not so clear-cut after all. There is a certain, if unwitting, recognition of non-absolute heterogeneity. Firstly, in spite of their heterogeneity, genres of discourse are all subject to the same principle: to win. More important is the idea of passages between genres of discourse. Drawing on Kant’s discussion of the faculties, Lyotard speaks of the ‘war and commerce’ between the genres of discourse. Not only does a genre borrow functional elements from another which it uses analogically, but the very constitution of a genre depends on its relation with other genres: each genre examines what goes on in other genres, and, by going beyond its borders, it finds its borders. Passages between genres do not eliminate the abyss between genres: they occur above the abyss, taking it into account. Within a genre there is conflict over the rules of the genre and even over the stakes of the genre (Lyotard, 1983). There is thus evolution over time. It seems that the definition of each genre is far from fixed. There is also the issue that a genre is not monolithic; there are sub-genres and sub-sub-genres. The lack of clear definition of each genre, and the multiplication of sub-genres create a generic fuzziness. The notion of incommensurable discourses depends on the possibility of clearly defining the said discourses. That possibility is challenged by the factors of inter-relationships, internal conflict, and change — borders and identities are not clear. There are aspects, then, in Lyotard which point to a Derridean philosophy, which challenges the presence of identity. If there is no absolute uncontaminated identity, there can be no absolute difference or incommensurability. Insofar as it involves absolute incommensurability, the notion of differend is challenged. On the other hand, the multiplication of sub-genres could give rise to an infinite multiplication of differends. Here Lyotard’s philosophy is brought closer to the Derridean notion of différance, infinite differing and deferring (Derrida, 1972). Rather than undermining Lyotard’s differend, these ideas could be considered to introduce greater flexibility into Lyotard’s notion.

I shall end this section with a discussion of Boothman’s study of Antonio Gramsci’s ‘translation’ into his own realist-materialist paradigm of certain key concepts used in the philosophically idealist paradigm of Benedetto Croce. Boothman finds that Gramsci can translate Croce insofar as there are corresponding term/concepts in the theories. Here is an example: the Crocean concept ‘dialectic of distincts’ is reinterpreted so as to apply to the levels or ranks of the superstructure rather than the aspects of the human spirit; the function of the concept in the two paradigms is the equivalent one of actively correlating the different sectors of human activity (Boothman, 2002, p. 112). Boothman says that concepts can be translated between paradigms because different conceptual schemes cover the same reality in different ways (Boothman, 2002, pp. 105, 116). This argument makes an appeal to a common third party, contradicting Kuhn’s and Lyotard’s idea that different theories create different realities (or that if there is a common stimulus, it is not accessible to us in its ‘raw’ state (Kuhn)). Boothman’s notion of equivalent functions does not in fact depend on a shared reality.

What is useful in Boothman, is his mention of the issue of hierarchy. Boothman acknowledges the incommensurability factor between paradigms, and attributes this to “a determining position in the conceptual hierarchy” of each paradigm (Boothman, 2002, p. 116). The determining position could be interpreted as the philosophical basis of the theories, for example Marxism versus Idealism. Boothman also mentions hierarchy when he says that two paradigms might have an analogous concept but the significance of the particular concept (place in the conceptual hierarchy) might vary for each theory (Boothman, 2002, p. 116). This may be a matter of borrowed concepts, for example: ‘ethico-political history’ (Croce) is accepted by Gramsci as forming part of his concept of hegemony, while ‘historical materialism’ (Marx/Gramsci) is accepted by Croce as a subordinate part of his brand of historical research. “There is an inversion in hierarchical rank, in which one thinker’s concepts form merely a subordinate part of the other’s overall view” (Boothman, 2002, p. 111). This could be considered to constitute a shift in meaning, recalling Kuhn’s worry that the same term takes on a different meaning in a different theoretical framework (Kuhn, 1974, p. 269). Boothman points out more generally (not in relation to Croce) that in cases where Gramsci has incorporated a concept into his theory from elsewhere (whether he uses the same or a different term), the concept undergoes “a semantic shift on being carried across (‘trans-lated’) from one paradigm to the other” (Boothman, 2002, p. 107).

3 Berman and Toury

It becomes important to consider the difference between a theorist accounting for or using another discourse (the theorist is doing the translation, eg. Gramsci translating Croce) and a researcher examining the translatability of two discourses. It could be considered that the researcher has a more ‘objective’ position. One theorist talking about another (rival) theory generally has a particular stake in defending his or her own theory. A researcher-observer may seem to be more neutral but nonetheless has his or her own point of view and interpretation which condition the account.

As a researcher analyzing the Berman-Toury relationship, my account of the encounter will be filtered through my own interpretation and conceptual tools. My interpretation will need to be wary of Berman’s and Toury’s self-interpretations, the ways they themselves present their theories. Berman, for example, says that his ideas are not prescriptive (Berman, 1995, p. 16), yet his aim is to judge translations and present ideas for new preferable translations. Of course my own interpretations may also be contestable: I am necessarily influenced by my point of view, in particular my exposure to contemporary French philosophy which is sceptical of “grand narratives”. As far as conceptual tools are concerned, I will use ideas from Kuhn, Lyotard, and Boothman. Use will be made of some philosophical concepts such as ‘incommensurability’, ‘subsumption’, and ‘contamination’. We have seen above that parallels have been drawn between inter-theory and interlingual translation. It therefore would seem pertinent to use some concepts from interlingual translation theory; these are ‘equivalence’, ‘shift’, ‘transference’, ‘naturalizing’, ‘mistranslation’, and ‘skopos’.

My discussion of Berman and Toury will be divided into four sections. I shall first discuss the question of incommensurability between the two research frameworks. Certain aspects clearly fall into this category. There are, however, some translatable or partially translatable aspects between the two frameworks: these will be discussed under the heading of ‘equivalence’. The third section will deal with changes over time in Berman’s and Toury’s ideas. Toury has not written about Berman, but Berman has written an account of Toury, or more precisely of the school which Berman considers Toury to have founded (“l’école de Tel Aviv”) (Berman, 1995). The fourth section will concentrate on that account, showing how Berman has ‘translated’ Toury and discussing the effects of and reasons for that translation. I do not aim to give a ‘complete’ overview of Berman’s and Toury’s theories; I shall only mention aspects of their theories and writings which are pertinent for my concern with the issue of the translatability and translating of research frameworks.

3.1 Incommensurability

The primary source of incommensurability is the general theory of ideas which each theorist assumes, the “determining position” (Boothman, 2002, p. 116). Berman’s base is modern hermeneutics (in particular Paul Ricoeur and Hans Robert Jauss), German Romanticism, and Benjamin’s concept of critique, whereas Toury’s base is the system theory of Itamar Even-Zohar, empirical science, and socio-semiotics. Berman’s approach is hermeneutic and evaluative; Toury’s is descriptive and systemic. These quite different approaches may be related to different traditions in translation discourse. Berman speaks of different national traditions: “La manière dont apparaît la problématique de la traduction n’est pas la même dans la tradition française que dans la tradition allemande, anglo-saxonne, russe, espagnole ou ( a fortiori ( extrême-orientale” (Berman, 1989, p. 679). There does seem to be a fairly strong tradition of evaluation and prescriptivism in France, which would be resistant to Toury’s descriptivism. The Berman/Toury ‘differend’ might also be explained as arising from different intellectual cultures and institutional situations. Different kinds of discourse may be deemed acceptable and legitimate in the different university systems involved, the French system and the Anglophone/Netherlands/Israeli system (a vague entity at this stage of hypotheses). Toury’s aim was to promote the setting up of a discipline, so he had to espouse a legitimizable discourse for his system, and therefore adopted the favoured empiricist/sociological approach. Berman, who is also concerned with traductologie, was well established in his system which happily accepts an elitist, hermeneutic, evaluative approach.

Let us elaborate on the differences between the two research frameworks. Berman’s essentialism is evidenced in the notion of ‘truth’ which returns insistently in his discourse. Every thing or activity has a truth ( critique, a foreign text, a translation, the translator’s task, the translation project:

il y a lieu de...poser [la question de la vérité] à propos de la traduction...

Berman, 1995, p. 57

Mais jamais ils [les traducteurs] n’ont perdu de vue la vérité autonome de leur tâche...

Berman, 1995, p. 59

Toury is reluctant to give a definition of what translation is, and in order to choose objects of study, he takes ‘assumed translations’, utterances which are presented or regarded as translations in the target culture (Toury, 1995, p. 32). As for the translator’s task, for Toury the translator’s activity is first and foremost a social activity and as such is governed by translational norms in the particular cultural environment where the translator is working (Toury, 1995, p. 53).

History is seen as unified by Berman; the varied translation modes at different times are manifestations of a single Idea of translation (Berman, 1995, p. 61). Toury asserts a relativistic position: norms governing translation vary across cultures, within a culture, and at different times (Toury, 1998, p. 12), and the various modes of translation are not seen as being regulated by some overall design.

For Berman the revelation of the essence of a foreign work to the target culture is accomplished after a historical process of a series of rewritings including commentaries and translations, and the revelation is accomplished by a true translation. Berman calls this historical process translation (as opposed to the usual French word for translation, traduction). Note, however, that a canonical translation does not put a stop forever to the cycle of retranslations (Berman, 1995, p. 57).[1] Toury does not pay attention to translation. No doubt he doesn’t agree with its premisses, that a foreign work has an essence, and that the historical progression constitutes a process of improvement (Berman, 1995, pp. 86, 57). Toury puts emphasis on norms changing over time (Toury, 1995, p. 62), which may explain differences among retranslations, but he does not entertain a notion of improvement. Berman considers in fact that reigning norms can impede the production of a ‘true translation’ (Berman, 1995, p. 84).

A question of significant difference between the two research frameworks, which was indicated above, concerns the roles of the individual and of the social. Berman places a lot of emphasis on the autonomy of the translator, and on the translator’s individuality. This individuality results in variety of expression if different translators are given the same task. Berman does not agree that reality can be considered to be a set of deterministic laws and systems ruling over the individual; the individual translator always makes choices. Berman gives individual translators the power of being able to change languages, literature and cultures (Berman, 1995, p. 59). This is an idealist notion of subjectivity. Berman does go on to say, however, that translators have to negotiate prevailing norms, acting within a socio-historical ‘horizon’ (see section 3.2) (Berman, 1995, pp. 59, 79). This comes closer to the functionalist model where subjectivity is circumscribed within certain limits (Brisset, 1998, p. 34). Toury’s concern is mainly with social and systemic constraints; Toury believes that translation is governed in the last instance by systemic position and function (Toury, 1995, p. 13). Toury tends to downplay the role of individual creativity, variability, and idiosyncrasies, while not denying these factors completely. Non-normative behaviour, for example, is recognized as a cause of change in the system (Toury, 1995, p. 64). Berman is also more concerned with hermeneutic aspects of individual translations, whereas due no doubt to the influence of the scientific paradigm (numbers supporting evidence, the aim of replicability of studies), Toury gives little room to the discussion of individual interpretations. While not denying the role of the other pole, Berman thus gives emphasis to the individual, and Toury to the social. This issue is therefore not a matter of strict incommensurability. We may recall Boothman’s notion of analogous concepts to which is attributed a different hierarchical place in the different theories (Boothman, 2002, p. 116).

The goal of the analysis constitutes an area where Berman’s and Toury’s ideas are incommensurable. For Berman an individual translator is always responsible for a translation, and a translation always makes a truth claim ( that is why the translation must be judged (Berman, 1995, pp. 41, 60). Berman’s goal is to analyze and compare translation, original and the translator’s project, then, based on the analysis and following evaluation criteria, to present a judgement of the translation and suggestions for alternative translations. Berman believes that the principles of poeticity (the translation is itself a literary work) and ethicity (see discussion in 3.3) allow the possibility of consensual judgement. Presumably subjective judgements will be shared, for example, the choice for study of passages in the translation which are problematic and those which are excellent (Berman, 1995, p. 66). Berman’s notion of translation analysis is inspired by the German Romantics’ and Walter Benjamin’s concept of critique. He doesn’t want to retain only a negative notion of judgement; for him, critique must also be positive and productive. As far as translation is concerned, this means not only criticizing but explaining why certain translations are defective. Most importantly it means showing the excellence of good translations, or, if the judgement of existing translations is unfavourable, indicating the conditions and suggesting principles for new translations — critique thus becomes the condition of possibility for translations to come (Berman, 1995, pp. 38, 92, 96).

This is significantly different from Toury’s goal which is to describe and explain translational phenomena through generalizations such as norms or laws. The development of the descriptive approach was largely a reaction against a long history of prescriptive and speculative writing on translation, so a strong division is upheld between description and prescription. Toury aims to be ‘neutral’, that is, to avoid value judgements (but of course ‘neutrality’ is never possible in an absolute sense). For Toury, judgement is not the job of theoretical and descriptive translation studies; judgement is relegated to another branch of translation studies, the applied branch which comprises translator training, translation aids, and translation criticism (Toury, 1995, p. 19). For Toury there are no universal principles on which a translation could be judged: notions of what constitute a good or bad translation are as changing as the notion of translation itself, and judgements are based on norms in the culture in question (Toury, 1998, p. 26). Toury says that findings of the theoretical and descriptive branches can be used in the applied branch; it is not “scholars” but “practitioners” (translation critics, teachers of translation, and translation planners) whose role it is to draw conclusions with respect to behaviour from those findings (Toury, 1995, p. 17).

The differences between the two theoretical frameworks discussed above which concern incommensurable philosophies and goals are profound; we can say that in these respects we are dealing with the case of a differend.

3.2 Equivalence

I am using here the word ‘equivalence’ to signify the possibility of finding parallel ideas in the two research frameworks, and thus the possibility of translating one into the other. Candidates are ideas concerning the discipline, traductologie (Berman)/Translation Studies (Toury); the concept of horizon (Berman)/ ‘norm’ (Toury); the concept of la loi (Berman)/ ‘law’ (Toury); and common methods.

Although naturally there will be links with other disciplines, both Berman and Toury call for a certain autonomy of the discipline which has translation as its focus. There has been a wealth of studies involving translation done from within other disciplines, but what is needed is a discipline of Translation Studies/traductologie itself:

most descriptive studies have been performed within disciplines other than Translation Studies eg. Contrastive Linguistics, Contrastive Textology, Comparative Literature, stylistique comparée, or ( in more recent days ( Text-Linguistics, Pragmatics, or Psycholinguistics. Thus, while their subject matter could well have been deemed translational, the theoretical and methodological frameworks within which it was handled could not, if only because their interests lacked the wish to fully account for all that translation may, and does involve.

Toury, 1995, p. 3

Ce que l’on trouve le plus souvent, ce sont des analyses comparatives, produites dans les contextes les plus variés. Il y en a beaucoup, qui vont des plus naïves et des plus simples aux plus fouillées et étendues. Mais justement parce qu’elles apparaissent dans des contextes d’écriture à chaque fois différents, elles n’ont pas de forme spécifique. Elles ne nous aident donc pas à constituer un “genre”.

Berman, 1995, p. 14

Both Berman and Toury emphasize the necessity of rigour in the procedures of the discipline. What is needed is “clear assumptions” and “a methodology and research techniques made as explicit as possible and justified within Translation Studies itself” (Toury, 1995, p. 3). The discipline must aim to constitute a “savoir discursif et conceptuel rigoureux de la traduction et des traductions, essayant de conquérir une scientificité propre” (Berman, 1995, p. 63). However unlike Toury, Berman’s scientificity is not modeled on the exact sciences, but proper to a human science. Furthermore, as we saw in section 3.1, the purpose in undertaking the rigorous analysis of texts is quite different for each research framework.

There seems to be quite a similarity between Berman’s horizon and Toury’s norms. Berman defines horizon as “l’ensemble des paramètres langagiers, littéraires, culturels et historiques qui “déterminent” le sentir, l’agir et le penser d’un traducteur” (Berman, 1995, p. 79). “Déterminent” is put between inverted commas because the horizon has both a constraining and permitting effect: it allows the translator to act at the same time as defining and limiting the range of possibilities (Berman, 1995, p. 80). The horizon is a social constraint acting on translators, and so are Toury’s translational norms. Norms are not fully determining because they are more or less strong, are not always followed, there are competing norms, and norms are reinforced or modified by translators’ actions (Toury, 1995, pp. 4-64). One important difference between horizon and norms is that horizon includes a historical dimension, tradition, and a lineage of past works and translations, whereas norms tend to emphasize what is currently acceptable. Berman considers that the ‘translator’s position’ is the result of a compromise between translational norms and the translator’s own ‘drive to translate’ (Berman, 1995, p. 74). As mentioned earlier, at other moments, norms, the reigning doxa, are depicted as rather negative by Berman, because they may be in conflict with the translator’s sense of the ‘truth’ of his or her task (Berman, 1995, p. 59). For both theorists, indeed, a translator can choose not to adopt the norms which are currently the most favoured at a particular time (Berman, 1995, p. 59; Toury, 1995, p. 64).

With Berman’s loi (law) and Toury’s laws we have a common term being used in their theories, and thus might think that there are grounds for equivalence. Let us then look at their definitions and use of the term. For Toury establishing laws is the ultimate goal of the discipline. From a large number of descriptive studies, laws could be established which state the likelihood that a “kind of behaviour...would occur under one set of specifiable conditions or another” (Toury, 1995, p. 16). For Berman there is “une loi de traduction” which is independent of social discourse, a Law which a translator cannot modify but must submit to if what he or she does is to be called translation, a Law which cannot be formulated because only relative doxa can be formulated. This Law or Idea of translation is manifold, manifesting itself variously in the course of History (Berman, 1995, p. 60). Although the two notions of law have in common the sense of something which is strongly binding, Toury’s scientistic and Berman’s idealist notions are poles apart, and hence cannot be considered equivalents. In criticizing the content of Even-Zohar’s ‘laws’ regarding the position of translations (also espoused by Toury), Berman says rather angrily: “en histoire de la traduction, il n’y a pas de loi” (Berman, 1995, p. 55). The case illustrates Kuhn’s worry that the same words can have quite divergent meanings in different theoretical frameworks.

The concepts of norms/horizon and laws/Loi, despite some convergences, display two quite different approaches in the human sciences. Toury’s norms and laws are both adduced from empirical-descriptive studies of translations. Translations are thus approached as data from which generalizations are drawn, as in the exact sciences. Berman’s horizon and Loi are posited as existing constraints and enabling conditions, and their existence helps us to better interpret and understand the particularities of the translator’s project and translations in the hermeneutic tradition of understanding phenomena in their “unique and historical concreteness” (Gadamer, 1975, p. 6).

With regard to details of methodology, there are some similarities between Toury’s and Berman’s proposed methods for analyzing translations, but also differences associated with those similarities. Both theorists propose to start with an examination of the translation before looking at the original text. Berman wants to see if the translation works as a text in itself; Toury wants to undertake various kinds of comparison with other translations (Berman, 1995, p. 65; Toury, 1995, p. 72). Both select representative samples, but there is a significant difference: Berman’s samples are representative of stylistic highlights in the source text, whereas Toury’s are representative of what is typical in the translation(s) (Berman, 1995, p. 70; Toury, 1998, p. 23). Both are wary of using translators’ statements in research. Berman thinks translators might just churn out received ideas; Toury worries that translators’ statements are partial and biased (Berman, 1995, p. 75; Toury, 1995, p. 65). Both put an emphasis on using translations themselves to discover factors shaping the translation. Berman wants to find out about the general ideas of the individual translator and about the project for the particular translation undertaken; Toury wants to find out about norms governing the translation (Berman, 1995, p. 75; Toury, 1995, p. 65).

We can conclude that there is some partial equivalence of concepts between the two research frameworks.

3.3 Evolution

Theorists always evolve in their thinking, thus it is important to take into account changes in their writings over time.

Berman’s notion of ethics appears to have changed. In an earlier text (Berman, 1985), Berman expresses very clearly that for him ethical behaviour in translation entails welcoming the foreign by respecting the letter of the source text: “La visée éthique du traduire, justement parce qu’elle se propose d’accueillir l’Etranger dans sa corporéité charnelle, ne peut que s’attacher à la lettre de l’oeuvre” (Berman, 1985, p. 90). This means taking into account the various patterns of the signifiers in the source text, and reproducing those patterns in the translation.

Given this strong stand, it is most surprising to find the following statement in the later text, Pour une critique des traductions: “Le traducteur a tous les droits dès lors qu’il joue franc jeu” (Berman, 1995, p. 93, Berman’s emphasis). Being ethical means then that the translator must say what he or she is going to do in the translation and stick to that:

Ne pas dire ce qu’on va faire ( par exemple adapter plutôt que traduire ( ou faire autre chose que ce qu’on a dit, voilà ce qui a valu à la corporation l’adage italien traduttore traditore, et ce que le critique doit dénoncer durement.

Berman, 1995, p. 93

Berman says that omitting chunks of the source text, adding passages, or producing renderings which are very distant from the source text, are thus fully acceptable practices as long as the translator has talked about it (Berman, 1995, p. 93). It seems that Berman is now relativistic in his outlook. He has possibly been influenced by other discourses. However, his tolerance seems to have limits. Just after having talked about this new concept of ethics, Berman says:

Armel Guerne, traduisant les Fragments de Novalis de manière explicitante et francisante (et pour nous par ailleurs totalement inadmissible à cause du mépris du traducteur pour la stylistique, la concision et la “terminologie mystique” du poète), s’en est du moins ouvertement expliqué.

Berman, 1995, p. 94

Here we see that Berman retains his notion of respect for the letter of the source text, which in fact plays an essential role since it is his basis for judgement of a translation. Indeed, prior to talking about the truthfulness of the translator, Berman enunciates the other strand of ethicity which is “un certain respect de l’original” (Berman, 1995, p. 92) (He had also said at one point that “la ‘vraie’ traduction est...celle qui est source-oriented” (Berman, 1995, p. 59).) Berman does not advocate blind literalism. We gain an idea of what a “certain respect” of the original is through examining his ideas on the translation of Donne’s poem ‘Going to Bed’ in the second part of the book (Berman, 1995). What Berman advocates is a translation which would maintain the following features of the poem studied: the mixture of poetic and prose aspects, and of the rhetorical and the colloquial; the mosaic of images, concepts, terms and figures; and mood and emotion including the metaphysical and conjugal resonances. Rhyme and the archaic nature of the language of the original are not important for him (Berman, 1995, pp. 194-198, 228). Certain features of the source text which are considered the most important are given priority, features which are “nécessaire” and not “aléatoire” (Berman 1995, p. 72), and translations which do not reproduce those features are judged negatively.

A further change pointed out by Brisset (1998, p. 40) is the introduction into Berman’s critical method of the concept of horizon which does not appear in his earlier work. The concept adds an important socio-historical element to Berman’s theory.

As for Toury, there have been changes in his discourse over the years. Comparing Toury’s 1980 book with his 1995 book, Gambier says:

[Toury] souligne à plusieurs reprises le rôle novateur du traducteur...[Les normes établies] n’oblitèrent pas mécaniquement la responsabilité, la créativité, la liberté [du sujet traduisant]. Par rapport à 1980, il y a évolution qui accompagne d’ailleurs tout un courant (non référé par l’auteur) déconstruisant la transparence supposée, l’invisibilité du traducteur...

Gambier, 1997, p. 585

An example of the innovative role of translators described by Toury (in this case pseudotranslators) is the way in which pseudotranslation was used by Arno Holz and Johannes Schlaf in order to introduce certain norms and models of Scandinavian literature (or more precisely, ‘Scandinavian-like’ forms) into German literature (Toury, 1995, pp. 47-52).

In a further chronological step, comparing Toury’s previous texts with his 1998 article, Pym says rather sarcastically:

The papers by Toury and Hermans both show signs of an aging structuralist empiricism (which they wouldn’t name that way) adjusting to critical theory with a sociological bent (which they might indeed name that way).

In Toury, the signs of the adjustment are the relative absence of terms like ‘system’ or ‘polysystem’, and the robust presence of items like ‘power relations’, ‘creativity’, and ‘social groups’, as well as vague human things like ‘hunches’ and ‘feelings’. I suspect all these newish elements could be aligned around the active verb ‘to negotiate’ (‘norms are negotiated’, etc.), since the term presupposes active human agents who are scarce, to say the least, in previous texts by Toury.

Pym, 1998, p. 108

Toury cites a new reference not mentioned in his earlier work, anthropologist J. Davis, who explores the notion of “social creativity” whereby people use their sociability to negotiate and create agreements concerning actions (Toury, 1998, p. 14). Pym notes, however, that Toury doesn’t go as far as he would like: there is still an insistence on norms as behavioural regularities which are accessible to statistics, and an emphasis on order and predictability (Pym, 1998, p. 108).

What we find then in both Berman and Toury is that there has been an evolution in their theories by which they in fact have come closer to each other: at certain points Berman promotes relativism, and at certain points Toury gives a significant role to human agents. At the same time neither theorist has abandoned early tenets of their respective theories. Both Berman and Toury seem to have been influenced by ambient discourses, and possibly by critiques of their work. This demonstrates not only that discourses are not static, but that they interact with and are influenced by other discourses, creating intertextual links; the notions of absolute identity and of absolute incommensurability are undermined.

3.4 Berman Translating Toury

In this section we shall examine Berman’s account (or ‘translation’) of Toury in Pour une critique des traductions: John Donne (Berman, 1995), using as starting points concepts from interlingual translation theory.

Berman finds several equivalents for his own work in Toury’s theory. He considers (Berman, 1995, p. 52) that Toury’s location of the various Hebrew translations of Max und Moritz amid the complex network of languages and cultures at particular historical times is similar to his own programme proposed in L’épreuve de l’étranger: “Montrer comment, à chaque époque, ou dans chaque espace historique donné, la pratique de la traduction s’articule à celle de la littérature, des langues, des divers échanges interculturels et interlinguistiques” (Berman, 1984, p. 12).

In a second example Berman envisages rewriting his book using Toury’s ideas; in other words, he would be able to translate certain aspects of his theory into Toury’s:

Si pour Toury, d’une manière générale, les normes translationnelles prescrivent l’adaptation des oeuvres étrangères, leur naturalisation, il peut arriver qu’à certaines époques, dans certaines cultures, etc., elles prescrivent l’inverse: ce serait le cas de l’Allemagne romantique, par exemple, si l’on réécrivait L’épreuve de l’étranger à la Toury.

Berman, 1995, p. 53

Berman also uses Toury’s term ‘norm’ which shows that it can fit into his own theory. He speaks of the norms of the target literary system: “[une oeuvre étrangère] peut être publiée sous une forme “adaptée” si elle “heurte” trop les “normes” littéraires autochtones” (Berman, 1995, p. 57). More importantly he speaks of translational norms, giving his own equivalent of Toury’s term first: “la manière dont [le traducteur] a “internalisé” le discours ambiant sur le traduire, les “normes”” (Berman, 1995, p. 74).

The phenomenon of transference occurs with respect to Even-Zohar’s ideas on the position of the translational system in the polysystem, ideas which are espoused by Toury. For Even-Zohar translation normally has a peripheral position in the target polysystem. Berman initially refuses the concept of centre versus periphery:

C’est en fait tout le schéma centre/périphérie qui est à revoir. Le fait que la traduction ait toujours eu un statut problématique au sein de la “cité” ne signifie pas qu’elle soit “périphérique”. La littérature traduite n’est pas périphérique, ni centrale; elle a été, et reste, ce sans quoi aucune littérature autochtone ne peut exister dans cet espace du colinguisme...qu’est l’Occident. ()

Berman, 1995, p. 54

However, Berman then goes on to use the concept of centre/periphery on two occasions:

L’affirmation [de Even-Zohar] selon laquelle, lorsque la littérature traduite occupe une position secondaire, le traducteur se soumet aux normes d’“acceptabilité”, peut être occasionnellement vraie. Mais dans le cas de la France du XVIe siècle, on a une relation inverse: la traduction occupe clairement le centre du polysystème, ce qui n’empêche pas la plupart des traductions de cette époque d’aller dans le sens de l’acceptabilité!

Berman, 1995, p. 55

[…] si les traducteurs, depuis Saint Jérôme, ne s’étaient souciés que d’obéir aux normes, jamais la traduction n’aurait été en Occident cette façonneuse “primaire”, et non “secondaire” ou “périphérique”, de langues, de littératures, de cultures, etc., qu’elle a été, et reste.

Berman, 1995, p. 59

Berman has transferred the Even-Zohar/Tourian notion into his own discourse. Because Berman is in principle against the polysystem schema of centre/periphery, his use of the notion constitutes a kind of ‘contamination’ of his own theory.

In Berman’s account of Toury’s theory there are some shifts (recall Boothman pointing out the semantic shifts in Gramsci’s ‘translations’ (Boothman, 2002, p. 107)). Berman bases his account mainly on Toury’s early work (Toury, 1980), which as we saw in section 3.3, was more mechanistic than his later work. Nevertheless, the following sweeping generalization by Berman can be considered a shift with regard to Toury’s intent: “je veux échapper au fonctionnalisme ou au “structuralisme” qui réduisent le traducteur au rôle d’un “relais” entièrement déterminé socio-idéologiquement...” (Berman, 1995, p. 81). Toury’s norms are not fully determining (see section 3.2). Similarly, the statement that Toury “nie toute autonomie du traduire” (Berman, 1995, p. 58) is an exaggeration. As we saw in section 3.1, for Berman and Toury it is a question of differing emphasis rather than a question of a choice between individual creativity and social constraints.

There are several examples of mistranslations in Berman’s account.[2] An initial mistranslation is where Berman says in a parenthetical comment: “[les] “normes translationnelles”...ne sont pas des normes spécifiques pour la traduction, mais des normes valant pour toutes les pratiques d’écriture” (Berman, 1995, p. 56). Toury says there may be a partial overlap of norms governing translation and norms in other fields, but never full identity (Toury, 1998, p. 23). Another guise of the notion of lack of specificity is Berman’s belief that for Toury translations are fully integrated into the target literary system (Berman, 1995, p. 57). Toury in fact says that translations can be regarded as constituting a special system or genre (Toury, 1995, p. 28). There are occasions, however, where there is a blurring of the borderline between translations and non-translations (Toury, 1998, p. 24).

The most significant mistranslation is where Berman says that Toury concentrates on the analysis of target-oriented translations (“son champ d’étude et d’analyse étant la traduction target-oriented” (Berman, 1995, p. 59)). There has been a confusion on Berman’s part between target-oriented research (Toury considers that the forces motivating and governing translation are situated in the target culture which therefore should be the main focus of research (Toury, 1995, p. 24)), and target-oriented translation (translations which adopt target culture norms of expression). Target-oriented research may well study source-oriented translations if those happen to be the translations making up the particular corpus studied; the corpus will simply be studied with a primary focus on the target culture situation.

Shifts and mistranslations make the other more alien than it actually is. For Lyotard, recognizing the other’s difference is a question of justice; but exaggeration of difference can be unjust.

There are a number of cases which seem to be both shifts or mistranslations and subsumptions of Toury’s discourse by Berman’s. Berman says that for Toury: “la traduction “vraie” est celle qui est “adéquate” à tel moment...à la culture d’arrivée (target-oriented)” (Berman, 1995, p. 58). Toury never talks about ‘true’ translations; he says that target-oriented translation is the more common situation (Toury, 1995, p. 272). Finding encouragement in the title of Toury’s article “The Translator as a Nonconformist-to-be, or: How to train translators so as to violate translational norms” (of which the purpose is actually to sensitize translators to the concept of different norms at different times and places, rather than to exhort students to rebel against current (target-oriented) norms (Toury, 1995, p. 258)), Berman goes on to say that: “[L’école de Tel-Aviv] ne peut, au fond, pas ne pas penser que la “vraie” traduction est la première, celle qui est source-oriented” (Berman, 1995, p. 59). This is an imposition of Berman’s own ideas.

In the following passage Berman criticizes Toury for justifying any existing translation: “[le texte traduit] se trouve dans tous les cas justifié, puisque l’analyse même montre qu’il ne pouvait pas être autrement qu’il n’était” (Berman, 1995, p. 62). The point is that Toury is not concerned with ‘justifying’ translations — he wants to describe existing translations. It is Berman who is concerned with justifying or not: he is thus imposing his own goal of judgement in his comment on Toury. Berman also supposes, without any evidence, that Toury and Brisset realise that their aim of neutrality is untenable: “On suppose que les tenants de cette école réalisent, au moins en leur for intérieur, le caractère proprement intenable, à tous points de vue, d’une telle neutralité” (Berman, 1995, p. 63).

The process of Berman naturalizing Toury’s discourse and project into his own is evident in further cases. In his introduction Berman speaks about two forms of analysis or critique of translations: that of Henri Meschonnic and that of Toury. Berman defines critique in the following terms:

[…] si critique veut dire analyse rigoureuse d’une traduction, de ses traits fondamentaux, du projet qui lui a donné naissance, de l’horizon dans lequel elle a surgi, de la position du traducteur; si critique veut dire, fondamentalement, dégagement de la vérité d’une traduction, alors il faut dire que la critique des traductions commence à peine à exister.

Berman, 1995, p. 13

By defining Toury’s project as a form of critique, he subsumes that project under his own. Berman also subsumes Toury’s norm relativity under his notion of the ‘Idea’ of translation. For Berman, changing norms are only the manifestation of the richness of the unified Idea of translation unfolding in History:

Loin d’apporter la preuve que le traduire est chose changeante, relative, sans identité ni frontières, l’Histoire, d’époque en époque, expose à nos yeux la richesse déroutante de la traduction et de son Idée. Les prétendues variations de la notion même de traduction aux différentes époques peuvent ainsi être lues comme des manifestations préférentielles d’un des contenus de cette Idée...

Berman, 1995, p. 61

These cases recall the Gramsci (Marx)/Croce situation where one theorist gave the other theorist’s notions a subordinate place in his own theory (Boothman, 2002, p. 111). In Lyotardian terms the kind of naturalization which results in subsumption of Toury’s discourse by Berman’s creates a situation of injustice, because Toury’s discourse is silenced and robbed of its specificity, its specific stakes.

In order to understand Berman’s ‘translation’ of Toury we need to consider the reason why he engages with Toury in his book Pour une critique des traductions: John Donne. Berman uses the ideas of the two theorists, Meschonnic and Toury (and his ‘school’), in order to position his own brand of translation analysis or critique. Although Berman praises Meschonnic and Toury for having established strong methodologies for translation analysis, he tends to exaggerate what he sees as negative characteristics of their approaches in order to show how his own approach is preferable. Meschonnic is thus depicted as over-hasty and polemical in his negative judgements, and Toury’s theory is portrayed as deterministic, scientistic, negligent of important historical processes, and denying any autonomy of the translator. Meschonnic is shown to represent an extreme of source-oriented methodology and Toury of target-oriented methodology, so that Berman can be seen as occupying a more flexible middle ground. Berman wants to demonstrate how his approach takes account of Meschonnic’s and Toury’s strengths while avoiding their perceived shortcomings. He wants an approach which is both scientific (like Toury’s) in having a well-defined method and involving rigorous analysis, and critical in producing an evaluation (like Meschonnic’s approach but with a careful study prior to evaluation) (Berman, 1995, p. 63). Through his discussion of both Meschonnic and Toury (which involves some shifts and mistranslations), Berman backs up his own ideas. We can say that translation of one research framework by a researcher espousing another is very much a function of the purpose (‘skopos’ in interlingual translation theory) for which it is undertaken.

Unfortunately Toury has not written about Berman, so I am unable to undertake a twin analysis. However I can remark briefly on an article where Annie Brisset (whom Berman (1995) groups with Toury in his text) comments on Berman’s theory. Brisset argues that in Berman’s theory there is a contradiction between the notion of an idealist non-determined subjectivity and the concept of horizon (Brisset 1998, p. 40). But the relationship between individual and social power depicted by Berman could alternatively be viewed as a struggle, or in hermeneutics a negotiation, and I have considered it a question of emphasis (section 3.1). Brisset points out another ‘contradiction’ related to the previous one: that between an essentialist conception of literary works and of the task of the translator (“faire advenir ‘la vérité de l’oeuvre’”) as opposed to the cultural conditioning of the horizon (Brisset, 1998, pp. 41-42). But from the perspective of philosophical hermeneutics there is no contradiction between essentialism and cultural-historical conditioning since truth speaks to us out of tradition and our cultural embeddedness. Understanding the truth of the subject-matter at issue results from a fusion of our culturally conditioned present horizon and tradition or the unfamiliar. On this view, essence is not eternally the same since the interpreter participates in the truth of the object (Gadamer, 1975, pp. XII, 273). If we do accept that Berman has blocked together contrasting positions, this is not necessarily negative. Berman has said that “la traductologie récuse d’entrée de jeu l’idée d’une théorie globale et unique du traduire” (Berman 1989, p. 676); in his list of eleven tasks for traductologie, we find tasks which are exceedingly different: for example – task 2 concerns the issue of ethical respect of the ‘letter’, and task 3 the historicity of acts of translation (Berman 1989, p. 677). With respect to his conception of critique Berman speaks of “la dualité d’un acte critique” which in a descriptive phase reveals the socio-cultural conditioning of a translation, and in an evaluative phase pronounces a judgement on the translation (Berman 1995, p. 40).

Brisset says that “cette méthode [celle de Berman] ne tient pas compte de ce qui structure culturellement l’identité du texte (original ou traduit)” (Brisset 1998, p. 41). This is incorrect as far as translations are concerned. Furthermore, Brisset says that it is the functionalist paradigm (which she espouses) which must have led Berman to adopt the concept of horizon (Brisset 1998, p. 40), whereas Berman says that the concept comes from modern hermeneutics (Berman 1995, p. 79). The above examples show how Brisset’s ‘translation’ of Berman involves subsumption, a shift and perspectival interpretation, in its aim to illustrate inadequacies in Berman’s theory and support the power of her own favoured paradigm.

4 Conclusion

To conclude, there are some major obstacles to translatability between Berman’s and Toury’s theoretical frameworks. There are fundamental differences between the basic ideas underpinning the frameworks: idealism versus relativism, hermeneutics and German Romanticism versus empiricism and system theory, and prescription versus description. The opposition of individuals’ creativity versus social constraints is more a case of different emphasis in the two frameworks. Despite the differences, there is some common ground constituted by corresponding notions, in particular, the similarity of Berman’s horizon to Toury’s key concept of norm. Furthermore, there has been a change over time, which brings the discourses closer together, Toury granting individuals a greater role, and Berman proposing a relativistic notion of ethics.

In Berman’s ‘translation’ of Toury in the sense of recounting another discourse, the various phenomena which occur in interlingual translation are apparent: equivalence, transference, shifts, naturalizing, and mistranslations.[3] Berman is able to engage with Toury despite some misunderstandings and a certain shaping of the account due to the purpose to which it is put. We have confirmed Kuhn’s idea of communicability if not real translatability. The abyss is not so great between discourses as to forbid communication. The case studied therefore supports the more nuanced version of Lyotard’s theory: incommensurability is not absolute; the differend is worked through by parallels, communication, interaction, and change over time. The abyss nevertheless remains due to the fundamental difference in intellectual foundations, which is likely to have grown out of and is reinforced by the different intellectual contexts in which the methodologies were elaborated.

The findings and conclusions of my study are similar in many ways to those of Rosemary Arrojo whose joint paper with Andrew Chesterman regarding shared ground in Translation Studies was commented on by fourteen scholars from many different schools of thought (Target forum 12.2, 13.1, 13.2). Arrojo asserts the inevitability of many different perspectives and interests growing out of different circumstances, because this is simply human. She says that it is possible to make an effort to see beyond one’s own perspective, to consider and respect other views, and possibly to realize that there is some common ground (as she did with Chesterman). However some scholars will only read and criticize in order to present their own opinions; and a reading through a different perspective will often present a different understanding and possibly misunderstanding from the point of view of the author of the text commented on. This may be frustrating for the author but also potentially enriching, leading to refinement of the theory (Arrojo 2002). Berman’s and Toury’s perspectives, with their incommensurability and their interaction, are thus two voices participating in the lively and valuable polyphony of Translation Studies.[4]