There exists widespread discussion about directionality in interpreting, mainly questioning whether an interpreter has to work into her/his own mother tongue, out of it or both ways.
Traditionally, in Western Europe there has been a tendency which prefers interpreters to work only into their mother tongue (A language). This approach is rooted in Seleskovitch’s theorie du sense (1978:36-38) generated at the Ecole Supérieure d’Interprètes et de Traducteurs based in Paris. According to this theory, interpretation is a process of deverbalization . It means that interpreting is a process of decoding and visualization similar to communication. Basically, communication works as follow: speakers have a non-verbal thought in their mind, which is the idea they want to express. In order for this idea to be expressed, it must be encoded through a code. The receiver of this message, in order to understand it, decodes the message unconsciously and automatically. The same happens in interpreting. It means that the process of interpreting is a decoding/encoding one: the speaker speaks in a language or code (source language), this message produces a “mental image” in the interpreter’s mind; this “mental image” is the meaning, it is independent from the language used and it is what the interpreter has to encode in a different code (the target language) (Seleskovitch, 1978:86).
This approach considers the interpreter’s mother tongue as the best and most appropriate code to convey the message in another language whilst interpreting, because a mother tongue is believed to be the most spontaneous and natural code of expression for everybody, including interpreters (Seleskovitch, 1978:120-121)
Therefore, this theory looks at production as the critical phase during the interpreting process. An adequate comprehension or near-native comprehension is given for granted, because it is achievable, while a near-native production is much more difficult to achieve (Seleskovitch & Lederer,1989:134-135 ; Seleskovitch, 1978:121).
On the contrary Denissenko,(1989:157 ) conceives that the comprehension phase is the crucial one, because he believes that it is undeniably true that comprehension in one’s mother tongue is easier than in a foreign language.
Furthermore, what is considered to be an advantage, that is to say an excellent command and the wider variety of expression that interpreters have in their mother tongue compared to their B language, can become an obstacle when interpreting simultaneously. The interpreter, in fact, looks subconsciously and intuitively for the best, most polished way of expression, slowing down the retrieval process because of this selection. On the contrary, when interpreting into one’s B language, the variety of expression is more limited, therefore the selection is quicker.
This introduces the second main approach to directionality, which is the one that prefers interpretation into the interpreter’s B language. The grounds supporting it is the fact that, by doing so, a proper understanding of the message is assured, considering everybody can have comprehension problems (Denissenko, 1989:157). Interpreting out of the interpreter ‘s mother tongue was mainly used during the beginnings of conference interpreting and in the Soviet Union and Eastern bloc (Stevaux,2003:328). Besides any theoretical reasons for choosing this direction, there were also political reasons (Martin, 2005:85). In fact, politicians did not want to leave their words to the interpretation of foreign interpreters. Baker (1997) reports about Saddam Hussein being interviewed during an American TV-show with the assistance of an Arabic mother tongue interpreter. Usually, in the context of TV-shows, the interpreter is provided by the production and s/he is usually mother tongue in the host language, but in this case probably Saddam Hussein wanted to make sure the interpreter could fully understand what he was saying, in order to convey the message in the more trustful way to the public.
At any rate, today we can observe the following situation: bilingual booths on the private market and pure booths (Stévaux, 2003:333) in international organizations (interpreters working into their mother tongue and not out of it). But, due to the EU enlargement, EU institutions have started to accept interpreters doing a retour[1], given the scarce number of interpreters with certain rare language combinations (e.g Latvian-Finnish, Swedish-Romanian, etc).
International organizations follow the AIIC guide lines for simultaneous interpreting (http://www.aiic.net/ViewPage.cfm/page205.htm). The point of view of AIIC on directionality is to try to avoid the use of retour as much as possible, preferring instead, the use of relay. As stated by Anne Martin (2005:85), it does not seem to be casual that such an approach was originated in France, where generally a certain attention was paid on the preservation and purity of French and in addition, a French audience is alleged to be less tolerant of foreign accents (Martin, 2005:95).
On the contrary, on the private market, bilingual booths predominate. This is mainly because of economic reasons, as in a bilingual meeting it is much more convenient to hire just two interpreters who can work both ways rather than four.
Considering the reality of the private market, supporters of interpreting into A expressed their adversity to the systematic practice of the retour for the previously mentioned theoretical reasons, and especially given that they consider performances into B to be of a lower quality, they fear that conference attendants’ dissatisfaction will lead to a switch to only-English international events.
With regards to this, Déjean Le Feal (2005:189) defines interpreting as a luxury good and as such it should always be a high quality service which only a mother tongue can provide. If such service does not meet the requirements expected and communication is not effective, there will be, as already said, a switch from multilingual conferences to only-English conferences. Therefore, interpreters have to reflect on it and act appropriately, otherwise the amount of interpreting jobs available will decrease (Déjean Le Feal, 1998:45)
Notwithstanding, there is a major question with regards to this: what are the criteria that define whether an interpretation is of a high quality or not? Bühler (1986 cited in Jorg, 2009) made a survey amongst 47 members of AIIC, asking them to evaluate[2] the importance of certain aspects of interpretation[3]. In 1989, Kurz (2002) made the same survey amongst conference participants. The comparison between Bühler survey and Kurz’s one shows that interpreters are harsher judges of their performances. This is probably because they have the appropriate knowledge to judge it. This can be supported by Collado Aìs’s work (2002): she asked some conference attendants to assess three interpretations. One was provided with a lively tone and was fully consistent with the original speech; the second one was fully consistent but delivered with a monotonous tone and the last one was delivered with a lively tone but was not fully consistent. Surprisingly or not, the not fully consistent one was rated as better than the fully consistent but monotonous one and in many instances, it received better rates than the fully consistent and lively tone performance. It shows that conference attendants do not have the adequate tools to evaluate competently an interpretation, they can just assess part of it. Namy (1978:26) justifies such results by affirming that a lively tone and a native accent can actually “lull the interpreting users into a false sense of security”.
A key aspect resulting from these surveys is the rating for a native accent. AIIC Interpreters of Buhler survey rated a native accent (probably one of the most evident distinguishing elements between an interpreter interpreting into A and one doing it into B) almost as an important feature for a successful performance, while interpreters users of Kulz survey rated it as a less important parameter. Therefore, it seems that interpreting users do not mind a mild non native accent.
With reference to this, it is interesting to mention the so called “suspension of disbelief” (Martin, 2003:432). Let’s imagine a congress in Russia with several Russian speakers and Russian mother tongue interpreters working into English for an English audience. According to the “suspension of disbelief”, listening to an interpretation into English with a mild Russian accent would add a sense of reality to the situation, as it would be as if the Russian speaker was speaking in English. But on the contrary, Donovan (2005:163) does not agree with it and believes that a non native interpreter increases the gap between the user and the interpreter. Moreover, he believes that when interpreting into B is not mastered, there is a tendency to a “nonsensical parroting” (Donovan, 2003:374) of the source speech, that leaves out nuances of meaning and breach one of the main tasks an interpreter has to fulfill, that is to say to convey the meaning not the words.
It is indeed clear that directionality cannot be investigated just relying on the discipline of interpreting, but must also take into account other disciplines: the approach must be multidisciplinary (Gile et al, 1994:4-5). In particular it would be interesting to study the interpreting process from a neurolinguistic and a cognitive perspective. The former, in particular, has already provided some hints on how a second language functions. Dejean Le Féal (2005:173) refers to a study by the neurolinguist Michel Paradis. Putting aside, the aim and the data of the experiment itself, the study shows that a native language is an unconscious process, while an acquired language is a conscious system. This would support the theory of a better interpretation into the interpreter’s A language because it is the most spontaneous form of expression for the interpreter. Furthermore, it can be supported by Gran and Fabbro’s study (1988:40) of interpreting students at several stages of their training. The outcomes of this survey revealed that self-correction when interpreting into one’s mother tongue is more efficient and more automatic.
Turning now to training, the School of Paris is traditionally against the teaching of interpreting into a B language (Stévaux, 2003:332), which is not to be taught because it is based on the same techniques of simultaneous interpreting into A and it fosters the practice of retour in the interpreting market. It is however tolerated for rare languages, for example Baltic languages and certain Eastern European languages, where the number of interpreters able to interpret from these languages is quite low.
Nevertheless, in a survey carried out by Donovan, it emerges that interpreters working into a B language would have liked to receive such training (Donovan, 2004:208) and, according to Iglesia Fernandez (2005:107), the training should follow the demand of market demand (which, as previously stated, prefers bilingual booths).However, Donovan points out (2003:375) that interpreting classes into B tend to become language classes. This is because they are not focusing on what she considers to be the main weakness of interpreting into a B language: lack of communication skills and a non-automatic expression. Therefore, she suggests, these lessons should keep in mind the communicative aim of interpreting and focus on enhancing communicative skills in the foreign language.
Another interesting study aiming to demonstrate the validity of interpreting into a B language is by Olaf Immanuel Seel (2005), in which he carries out an analysis of German-Greek non-verbal meaning. The study leads him to conclude that the best way to interpret is from the interpreter’s mother tongue because a proper understanding of non-verbal meaning can be granted only by a native speaker of the source language.
This concern about comprehension can be supported by several studies. For example, Anne Martin (2005) interviewed 10 professional interpreters working in the private market in southern Spain. The results showed that even though most of them considered working into a B language more tiring, working into B is considered as standard practice and their main concern was comprehension. In relation to production, their main concern was how the audience perceives their delivery. It is also interesting to note that the majority of those who declared they felt less comfortable working into their B language were those not residing in the country of their B language and whose exposure to their B language was therefore weaker (p. 93).
Particularly relevant is the outcome of Galli’s study in 1990 (as cited in Iglesias Fernández, 2005:107). He interviewed some interpreters and the majority stated they did not consider working into a B language as a source of concern for their profession. They argued that there were other factors that can make the job more difficult, namely interpreting read speeches or fast speakers. So, again, the outcome of this study suggests that interpreters’ main challenges relate to the comprehension stage.
Another argument confirming that interpreting into B is not considered as a great challenge, is presented by Snelling (1992:14). In Snelling’s experience as a professional conference interpreter when working in the booth he had never been asked by an Italian interpreter to take over interpreting from Italian into English but he was asked to do so when non English speakers were giving speeches in English and had a strong foreign accent. Moreover, according to Feldweg (1989:140) it is undeniable that if something is missed in the comprehension phase, there is almost no way for the interpreter to recover.
An important step forward in the debate about directionality was the Forum on Directionality in Translating and Interpreting held in Granada in 2002. At the Forum a change to the classification of languages was suggested: to move from classifying working languages as A language or mother tongue and B or non-native language to a new classification of active languages and passive languages. (Martin, 2003:430). Under this new classification, A language and B language (the languages interpreters can usually interpret into) would be at the same level, while language C would be the passive language.
The vagueness of this classification can also be explained by Gile’s gravitational model of linguistic availability (1995b: 212-230): the more the interpreter actively uses and is exposed to certain words and expressions (in any language), the quicker the retrieval process is. The less frequently words are used, the slower the retrieval process is.
The need to redefine this classification emerged because sometimes the difference between mother tongue and non-mother tongue languages is not clear. An example is provided by Al-Salman & Al-Khanji (2002), who interviewed Arabic mother tongue interpreters working with English. They explained that they felt more confident working into English. The reason for this was that Arabic changes considerably according to the register, while English is less affected by differences in register.
Another comparative study that advocates the language specificity of interpreting and directionality is Galli’s (1990 as cited in Fernández 2005:107), which shows that Italian interpreters working at medical conferences (where the terminology is specialized and the grammar is simple) from Italian into English, could simplify more quickly the complex Italian structure to shorter, simpler English sentences.
The above mentioned study leads to the consideration that directionality in interpreting, its challanges and limits are also dictated by the language pair involved. This is in contrast to the théorie du sens (Seleskovitch 1978:86) in which interpretation is not language specific. Gile (2005:16-17) believes that the characteristics of language pair involved can influence directionality and the strategies interpreters use to cope with challenging pieces of interpretation.
Bartłomiejczyk (2006) reaches the same conclusion. Observing Polish-English interpreters, Bartłomiejczyk concluded that interpreters use different coping strategies according to the language pair involved.
In conclusion, in spite of efforts to carry out studies and surveys on directionality in interpreting, there is not enough empirical evidence to define one direction as better than the other (Gile, 2005:22). A great deal of the information available is based on personal experience, the quantitative basis or sample used is often too small to be able to draw any general conclusions and a more multidisciplinary approach is required.
[1] The practice of interpreting into the interpreter’s foreign language
[2] The rating definitions were: irrelevant, less important, important, highly important.
[3] Namely: completeness, correct grammar, correct terminology, fluency, logical cohesion, native accent, pleasant voice, sense consistency.