Notes on the concept of «translator's competence»

Abstract The first part of the article touches upon metatheoretical constraints to translator’s qualities research represented by the «methodological infancy» of Translation Studies (TS), the role that professional self-esteem of translators plays in TS research, and the lack of importance attributed to a pure science of translation. The main part is devoted to the concept of «translator’s competence» under which the qualities of translators have been typically addressed. I will discuss some problems with the notion of competence as underlying knowledge. Then, I will introduce a point of view adapted from Keen (1988): if understood as aptitude, competence is the result of performance in the translator’s history, not its cause. This could be appropriate to overcome a speculative tradition of translator’s qualities research. Key words: translation studies, translator’s competence, underlying knowledge. Resum. Notes sobre el concepte de «competència del traductor» La primera part d’aquest article tracta de les limitacions metateòriques en la recerca sobre les qualitats del traductor representades per la «infantesa metodològica» dels Estudis de Traducció (ET), pel paper que l’autoestima professional dels traductors juga en la recerca en ET i pel fet que no s’atribueix prou importància a una ciència pura de traducció. El gruix de l’article és dedicat al concepte de «competència del traductor», dins el qual normalment s’han agrupat les qualitats dels traductors. Abordaré alguns problemes relacionats amb la noció de competència com a coneixement subjacent. Seguidament introduiré un punt de vista adaptat de Keen (1988): si l’entenem com una aptitud, la competència es el resultat del rendiment en la trajectòria del traductor, no la seva causa. Això podria ser adequat per a superar una tradició especulativa en la recerca sobre les qualitats del traductor. Paraules clau: estudis de traducció, competència del traductor, coneixement subjacent. Quaderns. Rev. trad. 14, 2007 125-138 Notes on the concept of «translator’s competence» Rui Rothe-Neves Universidade Federal de Minas Gerais 31270-901 Belo Horizonte, MG. Brasil [email protected] Sumary Introduction In Translation Studies (TS), the qualities of a good translator have been typically addressed based on the concept of «translator competence». Originally, «competence» stands for the quality of possessing a skill, knowledge, or qualification. In that sense, it is a synonym for aptitude. Under the influence of theoretical Linguistics, competence has acquired a new meaning in TS, which derives from Chomsky (1964/1970): «the knowledge a speaker-listener has of his language», a mental faculty or underlying knowledge. As such, «competence» is thought of as a psychological attribute of an individual, whereas aptitude clearly implies, beyond mental faculty, also behavioral performance and results. As such, aptitude is also a criterion that can be applied for the judgment and assessment of one’s work by his or her peers. The aptitude to perform a certain task is determined by the expert opinion of a peer who assesses the translator’s performance, under different given circumstances. Thus, aptitude involves socially shared representations. Since translation involves not only different languages, but also different cultural backgrounds, the performance of a good translator is very dependent on a given time and place. In the absence of a «gold standard», when the question «What makes a good translator?» is asked, it is perhaps more important to inquire into the socially shared representations of a translator’s work, than to inquire only into the intrinsic qualities of translators. This is why competence should take on the meaning of aptitude in translation. Though apparently a simple point, this is not part of the mainstream understanding in TS. As referred above, translator’s competence is normally understood as some type of underlying knowledge, an extraordinary skill beyond purely linguistic skills involved in translation. It is said that the skills of comprehension and/or production in more than one language are not sufficient for someone to be able to translate. Koller (1992), among others, have expressed this opinion in a comprehensive review on the German school known as Übersetzungswissenschaft. The translator’s competence surpasses pure foreign language competence as acquired in foreign language classes. The translator’s competence, as the ability to produce a target language text for a source language text according to certain requirements, the so-called equivalence requirements, is qualitatively different from the mastery of the languages involved, thus different from pure language competence. (Koller, 1992, p. 19-20)1 Introduction TS’s methodological infancy The importance of pure translation research Professional self-esteem Some notes on the concept of «translator’s competence» Competence as a result of performance Concluding remarks References 126 Quaderns. Rev. trad. 14, 2007 Rui Rothe-Neves 1. Original: «Die Kompetenz des Übersetzers geht über die rein fremdsprachliche Kompetenz hinaus, wie man sie sich im Fremdsprachunterricht erwirbt. Übersetzungskompetenz als die Fähigkeit, zu It is important to note that nowhere in this passage it is said what makes the translator’s competence different than second or foreign language competence, or what attributes should define translator’s competence. In 300 pages, Koller returns to this subject only towards the end of the review (p. 224), and refers to the abovequoted passage. No descriptive facts are presented to support this opinion. The author, however, does not seem phased by this. In my opinion, Koller thinks it should be obvious enough for anyone in the field that translator’s competence is indeed different from linguistic competence. There is a similar opinion in Dancette (1993), in a thesis on text comprehension during translation. This author is even more emphatic than Koller: «From the fact that “bilinguals” are not necessarily the best translators, one may assume that there is a particular kind of competence that is required in translating» (Dancette, 1993, p. 27).2Shreve (1997) corroborates this line of thought, with similar emphasis, in his review on translator’s competence: «Not everyone can translate; those that learn how to translate do so by acquiring a history of translation experience» (Shreve, 1997, p. 121). These authors are not isolated examples in TS; their opinions, on the contrary, could be taken as representative of the field. In that sense, Koller offers a comprehensive overview of what has been done in Übersetzungswissenschaft, the German school of translation theory in which the scientific basis of knowledge is highly valuable; Dancette and Shreve present new attempts that deal with translator’s competence, from a cognitive point of view and, as such, following rigorous experimental methods. Not one of the authors, however, corroborates the opinion that translating is not limited to linguistic knowledge with a detailed description of what does distinguish the former from the latter. On the basis of the mainstream literature in TS it is, thus, not easy to tackle the question of what qualities should a translator possess. Also, a question arises in why are researchers’ efforts not leading to clear answers? This has certainly to do with the methodological state-of-the-art in our field, a «methodological infancy»; but it also has to do with several assumptions that may be currently hindering research advances on that matter. These assumptions are manifested, among others, in the role that the professional self-esteem of translators plays in TS research, in the lack of attention to a pure science of translation, and in the lack of precision with which the term «translator’s competence» is used in TS. The first part of this article discusses the first three aspects mentioned above, in order to show that there is a clear need in the TS field to pursue such a metatheoretical enterprise. Next, the concept of «translator’s competence» is discussed, a notion under which the qualities of translators have been typically addressed. This paper also tackles the problem of adopting the concept that competence is some sort of underlying knowledge, which renders this concept imprecise and, thereNotes on the concept of «translator’s competence» Quaderns. Rev. trad. 14, 2007 127 einem AS [Ausgangsprache]-Text einen bestimmten Forderungen, sog. Äquivalenzforderungen, entsprechenden ZS [Zielsprache]-Text herzustellen, ist qualitativ etwas anderes als die Beherrschung der betreffenden Sprachen, die reine Sprachkompetenz also. —all English quotes are mine, except otherwise stated». 2. Original: «Partant du constat que les “bilingues” ne sont pas nécessairement les meilleurs traducteurs, on peut supposer qu’il y a un type particulier de compétence requis pour traduire.» fore, not useful. This understanding of competence is very common in TS, so it seems unnecessary to make direct reference to every version presented (discussed by Orozco, 2000, as cited by Hurtado Albir, 2002). Finally, I will introduce a point of view, as adapted from Keen (1988), in which competence, meaning aptitude, is the result of the history of the performance of the translator, not its cause. This standpoint may contribute to further our knowledge about the abilities involved in translating, since it allows for integrating pieces of information from individual, interpersonal and social levels. At the source of the metatheoretical constraints that are manifested in the four aspects addressed here, there is, in my opinion, the possibility of overcoming a speculative tradition in TS. A focus on the performance of translators would be specially appropriate. Such a focus allows researchers to ask questions that can be operationalized in a straightforward fashion and can be, therefore, more readily answered with descriptive statements, providing intersubjective facts about the qualities of translators. Such facts are necessary steps to be taken towards a descriptive approach of the qualities of the translator, as advocated by, for example, Chesterman (1993) and Toury (1995). TS’s methodological infancy By «methodological infancy» I refer to the fact that in translation research there is a want for empirical methods. Despite some opinions on TS and its nature as an independent discipline (e.g., Bassnett-McGuire, 1980, p. 1-11; Snell-Hornby, 1988, cap. 1), it is well-know that TS are still working towards a characteristic methodological profile (Toury, 1995, caps.1-3; Baker, 1996). Without proper tools to objectively assess what activities translators perform, it is not likely that TS will produce more than simply opinions on translation, which are based ultimately on the authority of those who produce them. Königs (1990) made this same point in relation to the Übersetzungswissenschaft. According to the author, the German school produced mainly second-order theories, because they did not produce their theories from observation data of actual translation performance, but rather from an idea of what translators might do. The raw material for this is usually anecdotal experience of the researcher as a translator and/or a translation teacher-or, as Gile (1994, p. 150) once wrote, a science of practisearchers. In other words, theoretical statements about translation performance (what translators do, how they do it and why) need no further demonstration, but only good arguments. In my opinion, this is true not only in relation to Übersetzungswissenschaft. It should be clear by now that without empirical grounds, the investigation of doubtful matters will lack resolve and procedures. Translator’s competence is a case in point. In one of the first contributions aimed at defining translator’s competence, Bausch (1977) produced a set of characteristics that the professional translator should develop: (a) linguistic competences, in a broad sense, both grammatical and communicative; (b) specific competences, represented by the knowledge of extra-linguistic reality; and (c) translational competences, which he considers «an autonomous ability that surpasses reading and writing in complexity» (Bausch, 128 Quaderns. Rev. trad. 14, 2007 Rui Rothe-Neves 1977, p. 519).3Since Bausch, «translator’s competence» has become a dominant concept in the field, although empirical pieces of evidence to support it are scarce (see reviews in Kiraly, 1995, p. 42-48; Séguinot, 1997), partially due to the almost chronic use of think-aloud protocols (TAPs). In that sense, it is known that expert translators verbalize considerably less than novices about their translation problems and thus provide poorer TAPs. After a careful study with on-line measures of conference interpreting, Dillinger (1989) concluded that the interpreter’s performances presented the same features as those of control subjects, who had simply listened to the text. The author, hence, challenges the existence of some extraordinary ability in interpreting. To date, no study approached text translation as carefully as Dillinger’s, and the question of what constitutes translator’s competence remains unanswered. The importance of pure translation research Another cause for the poor empirical grounding of research on the qualities of translator’s is, in my view, that little importance is attributed to pure research. By «pure» I mean research that has nothing to do with translator training. Though there are some hints of change in the past 10 years, most of the published work in TS started at, and is funded by, translator-training programs. Additionally, most translator-training institutions are primarily, or were incorporated by, institutions devoted to language research and/or teaching (Caminade & Pym, 2000) and there is a long tradition in TS to draw on or to adapt assumptions, beliefs, theories and methods for translation research from those fields (Alves, 1992; 1995, p. 8-10; Baker, 1994). These facts are no novelty, but change seems to be gaining momentum. More recently, there has been some theoretical production on translation training experiences based on professional translation (Hurtado Albir, 1996; Pagano, Magalhães & Alves, 2000; Schäffner & Adab, 2000). Nonetheless, in my opinion, the fact that translation has been approached almost always from the vantage point of second language studies seems to obliterate the question of how different they are from each other. In terms of the qualities of a translator, there are very few reliable works on the question of what do translators do that are not related to language (e.g. computer and research skills); much like a civil engineer who has to develop management skills, beyond his or her knowledge about materials, to run a project. Professional self-esteem Professional translation has, for years, been considered a second-order professional activity. Translators’ choices are submitted to those of authors, editors, and others involved in publishing. In this sense, one of the motivations behind TS, whether explicit (Deslisle & Woodsworth, 1995) or not, was and still is to appreciate the Notes on the concept of «translator’s competence» Quaderns. Rev. trad. 14, 2007 129 3. Original: «als eingeständige, die Fertigkeiten Lesen und Schreiben an Komplexität übertreffende Fertigkeit […]». work of the translator. Though a fair motivation, it will not suffice. In fact, considering certain publications by research on translators’ qualities, one is almost compelled to think that the intention of the research to further translators’ professional self-esteem has been counterproductive. These are examples from authors representing different perspectives: «It [translation theory] cannot make a student intelligent or sensitive—two qualities of a good translator». (Newmark, 1981, p. 36); «S/he [the translator] needs very good memory, outstanding intelligence, elevated linguistic sensibility […] as well as the disposition to take responsibility and the ability to think dynamically and creatively.» (Vannerem & Snell-Hornby, 1986, p. 203).4 Translators and interpreters are voracious and omnivorous readers, people who are typically in the middle of four books at once, in several languages, fiction and non-fiction, technical and humanistic subjects, anything and everything. […] A crowd of translators always seems much bigger than the actual bodies present. (Robinson, 1997, p. 27) As a translator myself, I tend to find the statements above very agreeable. But in fact not one of them helps to sort what abilities need be investigated in a systematic study on the qualities of a translator. Some notes on the concept of «translator’s competence» The three above-discussed aspects are important for TS because they touch upon assumptions and beliefs of TS research on the qualities of a translator. As such, they have to be properly discussed. Likewise, some assumptions clustered around the concept of «translator’s competence» seem to actually hinder research on that subject. The present discussion is certainly not exhaustive. The main point here is to advocate a change in perspective. For that, I will refer to an interpretation of the famous Harris-Toury «opposition» and relate it to a discussion on the concept of translator’s competence. In the history of TS, the discussion on the nature of translator’s competence included a debate on the role of nature, represented by a natural skill for translating tantamount to bilingualism, on the one hand, and the role of experience as a translator, on the other. Brian Harris, in 1976 (Harris, 1998), first put forth the opinion that an innate ability parallel to bilingualism plays the most prominent role in the translator’s development. According to Harris, this ability can be identified very early in bilingual children, in situations of «natural translation». Harris defined natural translation as that carried out in day-to-day circumstances and by bilinguals who have never received formal instruction on translation. This notion has received its share of attention in the literature on bilingualism (see e.g. Malakoff & Hakuta, 1991), and also in TS since Toury (1986) presented his notion of a «native translator». Based on the polysystem theory, Toury stated that a translator could not only be made up of an innate ability to translate. The bilingual has to go through 130 Quaderns. Rev. trad. 14, 2007 Rui Rothe-Neves 4. Original: «Er/sie braucht ein sehr gutes Gedächtnis, außergewöhnliche Intelligenz, eine erhöhte sprachliche Sensibilität sowie die Bereitschaft, Verantwortung zu übernehmen und die Fähigkeit zum kreativen, dynamischen Denken.» a process of assimilation of recurrent patterns of behavior in the specific sociocultural situation of translation (more extensively discussed in Toury, 1995). Those two positions have been discussed in TS (e.g., Kiraly, 1995, p. 15-16; Lörscher, 1997; Séguinot, 1997; Shreve, 1997). Harris (1998) summarized it as follows: The concept of NT [natural translation] is similar to Toury’s concept of native translator (see Toury 1986 below); but it differs from the latter in that it is limited to a seemingly innate translation competence which is co-extensive with all bilingualism at any age. NT therefore contradicts the adage, often heard in translation schools and the profession, that «just because somebody is bilingual it doesn’t mean they can translate.» There is, however, a resolution of this contradiction, and it was put very simply by Ljudskanov. According to Harris, what Ljudskanov wrote is that translation can not be taught; what can be taught is how to translate according to the norms of a society. This is not a strange notion to Toury (1995, p. 241-258), to whom, in the development of the translator’s competence, it is of fundamental importance to learn how to translate according to the norms of a society. It may well be that under natural circumstances, a bilingual does translate without ever having been taught to, but it does not necessarily make her a competent, professional translator according to expert peers. Thus, those opinions seem to complement, and not contradict each other. And since they refer to different matters, there is strictly no opposition between them. This excludes the need for a tertius, or somewhat of a compromise, like that found in Lörscher (1997). It is most revealing that Harris’ and Toury’s opinions have been interpreted as opposite by some TS researchers. In addition to a matter of power, an academic dispute around the discipline statute of translatology, there is another possible reason for this interpretation. Though most TS researchers agree with Toury that the object of TS is how translation arises as a cultural fact in which case a strictly linguistic notion of competence is clearly insufficient—these same researchers never abandoned the notion of competence as a mental faculty, a notion that is at the heart of Harris’s natural translation. Thus, the opposition between Harris and Toury was seen as another instance of the debate that prompted the «cultural turn» in TS (cf. Snell-Hornby, 1988). Instead of a more radical criticism to the concept of competence, which set aside the original notion of a mental ability distinct from performance, the paradigmatic answer has been to extend the concept of a solely linguistic ability into a myriad of sub-competences, ultimately leading to «componential models» (discussed below). Paradoxically, a move towards defining competence as aptitude is borne out of the use of «competence» of the translator. In the TS literature, the meaning of «competence» varies from one author to another, or even for the same author. Sometimes competence means aptitude, and sometimes, a repertoire of abilities. This lack of consensus may be attributed, in great extent, to the fact that competence is a novel concept in the long history of writings on translation. For example, there is no reference to it in Vinay & Darbelnet’s (1958) «Glossary» (Glossaire des terms techniques employés dans l’ouvrage). Notes on the concept of «translator’s competence» Quaderns. Rev. trad. 14, 2007 131 Nida (1964) is possibly the first to make use of Chomsky’s Generative Grammar in the study of translation. The author refers to the subject of competence, but there is no reference to the term in the index. Nida introduces the term «generative device» and makes reference to the Chomskyan definition of competence: «[…] for it takes seriously the capacity of the speaker of a language to generate and to decode an infinite series of sentences.» (Nida, 1964, p. 60, note 1). That passage clearly does not refer to a set of procedures, but rather to a mental capacity. In the same work, however, Nida presents a series of operational tips on how to translate: «[…] some of the principal steps in procedure employed by a competent translator can be outlined as follows:» (Nida, 1964, p. 246). Step #8 consists in «Submitting a translation to the scrutiny of other competent translators. Such persons may be either stylists in the receptor language or experts in the meaning of the source-language document.» (Nida, 1964, p. 247). In that case, Nida seems to refer to aptitude in a «competent translator». Likewise, the famous work by House (1977/1981) first refers to what «seems to be the incompetent use of translation in the classroom.» (House, 1981, p. 3). However, in the same paragraph, House states, «that translation may begin to play a truly useful role in developing student’s communicative competence.» (House, 1981, p. 3). In the latter case, competence clearly does not mean aptitude. Despite this lack of precision, House (1981) introduced, for the study of translation, the concept of «communicative competence» first presented by Hymes (1967). Communicative competence is defined as a system of language use including «all components of communicative events, together with attitudes and beliefs related to them» (Hymes, 1967, p. 639-640). The use of «communicative competence» instead of a purely linguistic one, was responsible for the introduction of pragmatics as a basis for translation theories, which lead to the Skopostheorie (Reiß & Vermeer, 1984) and the German functional theories of translation. It also started the broadening of the notion of competence as a mental capacity into the «componential models» previously referred. But House appropriated Hymes’ concept because of its applications to L2 learning, though it was originally part of a research program intended to establish Linguistics anew. The traditional structural approach to language is based, according to Hymes (1967), on the «replication of uniformity», and Linguistics should be directed towards a sociolinguistic approach into the «organization of diversity», from which structures could be inferred. That is, Hymes proposed to reorganize linguistic investigations on empirical grounds, a return to the American ethnolinguistic tradition that was epistemologically opposed to the hypothetical-deductive Generative Grammar. This is perhaps why Fowler (1994, p. 643) considered the concept of communicative competence as «criticism and enrichment» of Chomsky’s linguistic competence, by which means it could be applied to other fields. When it comes to its application to TS, however, it must be said that Hymes’ approach is perhaps more appropriate than Chomsky’s to understand translation indeed as a cultural fact. However, it is not yet followed by all methodological refinements, in the search for organization in diversity, which took place in sociolinguistics, and could perhaps give way to a structure out of variation in actual translations. 132 Quaderns. Rev. trad. 14, 2007 Rui Rothe-Neves Since House (1981), investigation on «translator’s competence» in TS has typically followed three different approaches. Some start primarily from observing translation training or language teaching courses, aiming at applications that may further performance in translating (Hönig & Kußmaul, 1982; Kiraly, 1995; 1997; Shreve, 1997; Toury, 1982). Another research strategy is the so-called expert-novice comparison, in which the mental processes of experts translators are underscored in the comparison with the performance of novice ones. Despite some methodological pitfalls in expert-novice translator comparisons that need yet be overcome—for example, what defines an expert?—some studies have attempted to describe how expert translator’s performance differs from that of novices (Krings, 1986, Königs, 1987, Alves, 1995; 1996; Hönig, 1993, Tirkkonen-Condit & Jääskeläinen, 2000). In these studies, translator’s performance may be understood as the ability or set of skills that makes expert translating more efficient, i.e. a performance that leads to better results with less effort. In the past few years, research on translator’s competence has been summarized into models that combine different sub-components, each being a specific competence. Typically, those «componential» models, as termed by Schäffner & Adab (2000), do not build on empirical research on translation; they are restricted to ad hoc reflections mainly for didactic purposes. The work by the PACTE group is an exception (Beeby, 2000; Beeby et al., 2001; Hurtado Albir, 1999; 2002; PACTE, 2001). These authors have provided appropriate theoretical and methodological discussions (see specially Neunzig, 1999; Neunzig & Orozco, 2001), which included procedures for assessment of their investigation tools (further reference in Hurtado Albir, 2002). Componential models may be criticized along two different lines. First, from a methodological point of view, these components are conceived of as underlying knowledge, and, hence, only through translator’s performance or subjective judgment is it possible to assess translator’s competence, and thus provide empirical evidence to support such models. The way to infer such underlying knowledge from performance in translation involves judgment by the researcher. For instance, one may use translation errors as an index of underlying knowledge. Counting errors means that translators did not render text as they should have done. In some cases, it is not clear whether translators did not render text properly based on a one-to-one comparison, or in comparison with what is normally accepted in translation—that is, according to translation norms. Consequently, the researcher has to rely on what is known about translation norms, and that requires appropriate empirical investigations prior to error counting. Typically, however, there is no such concern; the researcher’s knowledge and experience as a translator or as a translation teacher will be used to provide the parameters against which to assess translation well-formedness (as e.g., Nord, 1999). Thence the objective of every empirical research is subverted, that is, to generate theory grounded on collected data. The problem here is that data is not based on tenable parameters.5 Notes on the concept of «translator’s competence» Quaderns. Rev. trad. 14, 2007 133 5. Rothe-Neves (2002) presents a means to tackle that problem empirically.