Chapter 11: Self-reflection
1. It is worth exploring in what ways the descriptive approach could be more self-aware and self-critical, and why it should be so. Hermans starts with the notion that translation is norm-governed and impregnated with values. With special reference to historical and cross-cultural studies of translation, Hermans points out that our descriptions of translations are themselves translations of practices and concepts, since they are filtered through our own concepts and conventions. They are thus subject to all the manipulations that are entailed by translation itself (Hermans 1998a, 67). Hermans considers that the critical practice of the discipline should consist in acknowledging the normativity of its representations of translation (Hermans 1998a, 70), and in theorizing the historical contingency of different modes and uses of translation and of discourses on translation including its own (Hermans 1999, 147). Venuti says that norm formulations (whether a researcher’s or anybody else’s) are always interpretations, and they are made in relation to and possibly against previous formulations in the field, but also in relation to the hierarchy of values that define the culture at large. He thinks that the insistence on ‘value-free’ translation research prevents the discipline from being self-critical, prevents it from engaging with contemporary theoretical discourses which would enable it to draw incisive conclusions from the data, and thus it is prevented from considering the wider cultural impact that translation research might have (Venuti 1998, 28-29).
2. Being self-critical necessarily means recognizing that certain distinctions are not absolute. An idealized empiricist scientific model assumes strict demarcations between meta-level and object level, and between empirical work and non-empirical work. Similarly to Hermans above, Matthijs Bakker and Ton Naaijkens point out that a clear distinction cannot be made between translation description (meta-level) and translations (object level), since both are interpretative (Bakker & Naaijkens 1991, 204-206). Hermans also shows how there is no clear dividing line between scholarly and other statements about translation: apart from the fact that both involve perspectival interpretation (see also Venuti, previous paragraph), both are second-order observations involving the impossibility of self-observation (Hermans 1999, 146). For his part Philip Lewis challenges the distinction between empirical work and non-empirical work (introspection, speculation, a priori prescriptivism) which Toury relies on. Lewis notes the centrality of an axiology in undertaking comparative work of translation and original, as reported in Gearhart:
Il [Lewis] ne voyait pas comment on pourrait éviter la rupture avec “l’empirique”. Plus on travaille au niveau qu’on appelle empirique, plus on est obligé d’apporter cet accent sur la totalité des opérations et de le faire revenir dans le commentaire qu’on écrit. (Gearhart 1981, 269)
[He [Lewis] didn’t see how the break with the ‘empirical’ could be avoided. The more you work at what we call the empirical level, the more you have to follow an axiology in the analysis and bring it into the commentary being written.] (my translation)
Any translation theorist must necessarily position him or herself (Arrojo 1998, 43). Being self-critical would mean acknowledging the role of judgements, interpretation and values in descriptive research.
3. Toury himself cannot escape interpretation and judgement in his own method and practice. Examples of the presence of interpretation and evaluation in Toury’s case studies are: the judgement that the use of binomials in Hebrew translations results in informational intensification, stylistic elevation, or change of semantic structure (Toury 1995, 107); the discussion of semantic shifts in the translation of a short German tale into Hebrew (Toury 1995, 150, 161); and the interpretation of a translator’s motivations in stating that the translator took into account stylistics and acceptability (Toury 1995, 190). Another researcher might not agree with such judgements. Bakker points out the significance of the researcher’s interpretative role in undertaking the description of shifts of meaning between ST and TT, and thus the possibility of a discrepancy between different researchers’ studies of the same texts (Bakker 1994). In these kinds of situation the goals of inter-subjective testability and replicability seem difficult to assure in an absolute fashion, even if studies are undertaken within the same theoretical framework and using the same method. There may be relative consensus, but never absolute consensus because of the issue of interpretation, and issues of categorization.
4. Does the recognition of the pervasiveness of interpretation and values mean therefore that there is no difference between Toury’s approach, and the approaches from which he so much wants to demarcate his own? There still are considerable differences between the approaches. Taking actual translations as the object of study differentiates Toury’s approach from speculative writing, and he does not set out to give an overall assessment of a translation or prescribe behaviour. Hermans, who fully acknowledges the value-ladenness of description, supports the distinction between an a priori prescriptive and a descriptive approach when he says that:
the primary task of the study of translation is not to seek to interfere directly with the practice of translation by laying down rules or norms, but to try to account for what happens on the ground, including the ways translation has been conceptualised. (Hermans 1998a, 66)
5. However, even if the aim is not to seek to interfere in the practice of translation by making normative pronouncements, it must be emphasized that normativity is present in descriptions. Rosemary Arrojo says that statements of norms are not in themselves prescriptive but they are not merely ‘descriptive’ either, since they reflect the viewpoint, interests and perspective of those who elaborate them. Similarly, what is considered as a regularity will reflect the interests of a certain translation specialist or research group, at a certain time, in a certain context (Chesterman & Arrojo 2000, 159).