JAIS The Interpretive Theory of Translation and Its Current Applications

Choi Jungwha, Ph.D (Hankuk University of Foreign Studies)

Interpreting and translation are two of the oldest activities in the annals of human history. Records of translation activities date back over 2000 years, and since ancient times, translation has been studied by numerous scholars. Interpreting, on the other hand, had no theory of its own, so to speak, until modern times. Although the activity of interpreting dates back to ancient times, it did not begin to take shape in its modern form until 1917 at the negotiation table of the Versailles Treaty. Consecutive interpreting, in which the interpreter begins only after the speaker has finished, came about after the Versailles Treaty. In contrast, simultaneous interpreting, which has become the preferred mode, had its debut at the Nuremberg Trials after World War II. The Interpretive Theory of Translation (aka, the Theory of Sense) was developed by Danica Seleskovitch and Marianne Lederer (researchers at the Ecole Superieure dInterprtes et de Traducteurs (ESIT) at the University of Paris III the so-called Paris School). The following is a brief outline of this theory. Interpretation Defined Interpretation can be universally defined as understanding speech and rewording that understanding in a different language. Theoretically, interpreting requires a faultless command of both the source and target languages, a deep insight of the subject matter and mastery of the correct methodology needed to carry out the interpreting process. In practice, of course, the situation is different. Many so-called theoretical problems of interpretation are merely practical issues borne of individual interpreters, e.g., inadequacies such as insufficient language or subject knowledge, poor interpreting

Choi Jungw ha , T he I nter pr eti ve T heory o f Tra nsla tio n and I ts Curr e nt Appli ca tio ns. Interpretation Studies, No . 3 , December 20 03, pa ges 1 -15 . (c) 2003 b y the Japa n Asso ciati o n for Interpreta tio n Studies

skills, working into a foreign language or the inability of theoreticians to sort out the lexical items instead of text segments, language learning exercises, etc. The correct parameters such as confusion between a language and a text, translation of Interpretive Theory of Translation, in fact, does not deal with the many practical aspects of translating. The Interpretive Theory is built upon four pillars: 1) command of the native language, 2) command of the source language, 3) command of relevant world and background knowledge, and 4) command of interpreting methodology. The first pillar is the command of ones native language. Everyone would agree that translators and interpreters must be able to use their mother tongue in all its nuances and subtleties. The second pillar, the command of the second language, is more difficult to assess precisely. The closed systems: phonological and grammatical, have to be mastered, while the limitless range of lexical items is subject to a lifelong and open-ended learning process. The same is true of both world and background knowledge, which are not static. Rather, they are the result of a continuous, dynamic process of acquisition. The fourth pillar is methodology. In this regard, the Interpretive Theory differs from most other theories as it posits that methodologically the process of translating requires an understanding of sense (language meaning + cognitive complements) and a formulation of the translation on the basis of the synecdoche principle. The Interpretive Process The text to be translated may seem to be a French, English, Korean text. This common impression is hard to dispel, yet the claim that a paper is French or English is inaccurate. In fact, the graphic signs the paper carries are not meanings , but rather symbols . The meanings of those symbols are in the readers or listeners mind. It could be said that this is a paper covered with signs that belong to the graphological system of the French or the Korean language, but not that this is a French or Korean paper. To claim that reading a text is reading a French text or a Korean text is a misconception because the reading activity is twofold: it requires 1) a text composed of signs by an author, and 2) a reader who assigns meanings to those signs. At this stage there is already an interpretive activity assigning language meanings to language signs. On the other hand, believing that a text carries the sentence meanings of a given language is too wide a concept as it credits the text with a quality that is actually an interpretive activity on the part of the reader or listener. This concept threatens to distort the object to be translated, and it results in abortive

attempts to translate one language into another. At the same time this belief is too narrow a concept as the readers or the listeners interpretive activity extends far beyond assigning language meanings to language signs. A reader or a listener not only calls upon his command of a language to interpret the signs but also on his world and background knowledge. In the International Herald Tribune, the headline of one of its articles reads: Singapore SARS a low risk. These signs can be recognized as English but at the same time must be associated to relevant world knowledge to understand what is meant by the author. Grammatical propositions, to be understood, require an attribution of language meanings, texts require in addition relevant world knowledge (such as: Singaporean researchers illness (first case of SARS since the WHO said in July that highly contagious disease had been contained worldwide) is not likely to spread). The full interpretive process comes into play as soon as the translator or the interpreter adds not only his knowledge of language concepts to signs, but also his knowledge of the world. In that case, translating/interpreting conveys an intended meaning; in other words, the sense. Knowledge of the world is language independent . Translating/interpreting an English text into Japanese calls quite naturally on cognitive complements stored nonverbally in our memory and actualized in the target or source languages. Cognitive Complements Native listeners and readers are usually not aware of cognitive complements. Verbal, situational, cognitive contexts and knowledge of the world come into play quite naturally, while language alone seems to be present. In everyday conversation, when listening to each other, the part played by knowledge of the language is difficult to distinguish from background information. Sometimes, however, when overhearing strangers speaking to each other, we realize we do not understand them, even though we understand all the words they say. Imagine hearing I told him so again and again. We would not know who him is or what so represents. We understand the language, not the speakers utterance. The deictics, him and so , consequently point to information that does not belong to language knowledge, but to background knowledge. Thus, background knowledge, just as language, plays a role in the comprehension of speech. Whatever the command of languages interpreters may have, language problems would remain unresolved if it were not for cognitive complements. Let us try now to discriminate between the various types of cognitive complements that enable good

interpreters to convey the speakers meaning. Verbal Context Speech is uttered in a continuous stream of words, each word contributing to the meaning of the words around it and being made more specific by these surrounding words. The meaningful interaction of words present in the working memory (on average, 6 to 8 words) is the first instance of cognitive complements; it dispels word polysemy. In other words, verbal context specifies the appropriate meaning. For example, the Korean word (bae/( )) can mean stomach ( ), pear ( ), boat ( ). when you hear bae with a verb, for instance ( ( ) ), ( ( ) ), ( ( ) ), all ambiguity disappears. Situational Context Interpreters are part of the event at which they interpret. They not only see the participants, but they also know who the participants are and in what capacity they take the floor. Thus, seeing who is being addressed will enable the interpreter, upon hearing Mr. President, to say either , ( daetongryongnim President of Republic) or ( Sajangnim President of a company) or ( Hoejangnim President of the Board of Directors) depending on what is appropriate in the circumstance. Awareness of situational context represents a further cognitive complement that brings forth relevant meanings, and consequently dispelling polysemy. Being present at the discussions and witnessing the proceedings enables the interpreter to gather sufficient knowledge to translate appropriately. For instance, having heard A make a statement and B saying Mr. As comments, B might say, A ( A ga malhankoseunwhat Mr.A just said). Thus, he is translating what happened, not what was said, and setting up an equivalence that could not be forecast at the language level. The term comments corresponds to a number of Korean words such as (koment comment), ( malseum words). Interpreting is not, however, supposed to align corresponding words, since in no two languages do words ever coincide entirely. Rather, it should convey sense on the basis of both verbal information and cognitive complements. These increase the number of possible translations, as Delisle (1988) observes. Cognitive Context As we listen to someone speaking, we remember roughly what was said But

previously. Since it would hardly be possible even for the best of mnemonists to repeat several minutes of speech verbatim, and since observation shows that trained interpreters make use of their memory of things said previously to understand sentences now being uttered, it must be inferred that to a large extent words said previously have lost their verbal shape, and context, since it stems from things said. It is the cumulative knowledge brought by the speech chain up to the point where the interpreter is translating. In consecutive, as in simultaneous interpretation, interpreters make use of this cognitive context. It is best illustrated in consecutive where it accumulates so that the interpreter gets the full sense of all parts of the speech. When delivering his interpretation, he makes use of it to find the appropriate wording. Professional interpreters never interpret sentence by sentence; they prefer to wait for the speech to proceed, so that it imparts maximum information. They know that the accurate meanings of words arise not only out of their language tenor but also out of cognitive context. Whether engaged in court interpretation or conference interpretation, in consecutive or simultaneous, interpreters all over the world claim that understanding is of paramount importance. The question then arises: Does understanding relate only to language or does understanding speech also require knowledge of the world? Linguists often consider that understanding a verbal utterance and understanding its language is one and the same thing. In conversation, when we are listening to each other, the part played by knowledge of language can hardly be discerned from that played by background information. It is only when we overhear strangers speaking to each other that we realize we do not understand them, although we do understand the language, not the utterance. As in the aforementioned example, the deictics, him and so , point to things that do not belong to language but to background knowledge. Thus background knowledge is as important as command of language in understanding speech. When confronted with themes with which we are not familiar, the domestic politics of a foreign country for instance, we also realize that we lack certain knowledge other than that of language: knowledge that is necessary to fully understand what we are hearing or reading. Such will surely be the readers reaction to the International Herald Tribune headline: Singapore SARS case a low risk. Background knowledge is a blanket expression covering a number of cognitive complements that help us understand speech. These include knowledge of the world, of time and place, of the circumstances out of which a speech arises, memory

of things said previously, knowing who the speaker is and who the listeners are. The broader the cognitive complements, the less ambiguity and polysemy there is in language, and the more thoroughly speech is understood. This can easily be shown in the following sentence, which illustrates that knowledge, or a lack of knowledge about a single name, Dominique de Villepin, when mentioned without any background information, makes a huge difference in understanding. At a conference last week, Dominique de Villepin said that the government was vulnerable for two reasons; first, ministerial reform; second, Middle East issues. Listeners will understand this sentence at different levels, depending on whether they know who Dominique de Villepin is. Those who know, understand the political nature of the information. Those who do not, understand that it refers to the French government, but miss implications that are otherwise obvious. The proper name Dominique de Villepin designates a given person, but also an individual with a specific function (Minister of Foreign Affairs). In this utterance the name clarifies the word vulnerable, which loses the shades of meaning of unprotected or defenseless, listed in the dictionary, keeping only that of open to attack. No one will challenge the need for cognitive complements in intralingual communication. There is, however, far less unanimity when it comes to interpreting. Here, the need to add cognitive complements to source language utterances in order to convey them accurately in another language is not always clearly understood. without asking questions. Why cant the interpreter just translate even when he knows all the words of the source language and all corresponding words in the target language? The first answer is empirical: years of practice and observation of practice show that languages, however closely related, do not match in actual speech. This means that translating languages as heard, faultless though the translating might be, does not make sense. The broader the cognitive complements, the less ambiguity and polysemy there is in language, and the more thoroughly speech is understood. When no cognitive complements are present, we are faced with more words and sentences, in sum, more language. Investigations into machine translation have shown the polysemy of lexical items, that is, the number of signifis that can be attached to individual signifiants (see Winograds example; stay away from the bank might be advice to an investor or to a child too close to a river). It has also shown how readily an alignment of words can carry multiple meanings (The sentence All interpreters have had the experience of delegates who want them just to translate

chickens are ready to eat implies that something is about to eat something, but which are the chickens?). Ambiguity is a feature of language, and it is due to the absence of cognitive complements. Language turned into acts of speech always carries cognitive complements. When going to my bank to have a discussion with the manager, I never even think of river banks, just as when smelling the fowls roasting in the oven I know we will soon be eating and never think of live chickens strutting about pecking on food. Isolated words or sentences are language; they elicit meaning, not sense. The only possible level of interpretation is that of graphic signs into multiple linguistic concepts and ambiguous sentences. Words in a text are actualized by the verbal context. Of all the various possible meanings of a given word looked up in a dictionary, only one appears when used in a text. At the same time relevant knowledge in a text is activated as a cognitive complement that combines with the actualized meaning lending it in an ad hoc sense; that rules out any ambiguity. For those who do not know, Bloody Sunday refers to the shooting 20 years ago in Northern Ireland of civil rights marchers by British troops. In stating that translation needs information additional to language meaning and claiming that translators understand texts because of the cognitive complements they bring to bear on them, the Interpretive Theory of Translation introduces the process of translation into the vast area of cognitive research. Sense and Thing While graphic signs taken in isolation are interpreted into concepts, signs in a text are interpreted into sense. Instead of concepts, translators are faced with things. Concepts collect under one name features common to all individuals, leaving aside such features that are not shared by all. In the concept man, we find a body, a head and limbs; in an actual man, we find many more elements, in addition to the common ones; age, height, race, etc. In a text, the concept man is turned into a real man my brother or a well-known personality. Things are much richer than concepts and the richer they are the easier it will be for the translator to find an appropriate designation in any language. A thing is easier to put across in another language than a concept. Because sense is based on the cognitive complements of individual readers/ listeners or translators, it is to some degree an individual matter; its depth will vary according to the knowledge and the world experience of each individual. Whatever

the specifics of sense , however, there is a vast area of overlap of the sense understood by each of the communication partners, so that communication is usually easily established. The translator, acting as a mediator between an author who wants to communicate and readers who want to understand him/her, operates in this area of overlap. The readers of the translation will bring their own cognitive complements to the translated text. The translators rendering enables them to discover the text superficially or deeply, just in the same way as readers of the original. Understanding a text is universal. The translators understanding is only a specific case of the universal process. Observation of a successful interpretation has proven beyond doubt that in normal communication, contrary to popular belief, sounds are not understood first as language meanings and then connected to actual events; they are always instantly understood as the thing meant by a speaker, which Seleskovitch calls sense. Language meaning as such appears only in isolation and it is a misnomer to speak of understanding in connection with words or sentences in isolation. When we have the necessary language competence, we recognize sentences as being English or French; we recognize grammaticality as well as differences between word arrangements; understanding applies only to sense which appears when we listen to a speech such as defined above. Sense is the awareness of the things meant by a speaker. As such, it is much more durable than the language sounds of a speech. Good interpreters, skilled in consecutive interpretation, remember the slightest nuances of a speech when they start rendering it, though the mass of fleeting sounds that carried it has long vanished. Seleskovitch explained the concept of correspondence and equivalence using the metaphor of raisin bread. If the original text is of various ingredients for making raisin bread, the interpreter or the translator will make the dough into the bread. The raisin is the part that can find a correspondence: the form remains unchanged. However, the rest of the dough, made of flour, sugar, salt and yeast, are all incorporated into the dough, and one can no longer separate the original ingredients. Yet, they are all in the dough. The mixed dough can be understood as the equivalent part. The relationship between an equivalent and a correspondence is explained using the following diagram.

When correspondence is possible, the interpreter/translator can move directly from understanding to re-expression. If not, the interpreter/translator retrieves what the original text is saying, that is the sense, and looks for its equivalence in the target language. The Theory of Sense is based on the fact that different languages use different ways of expressing similar content. The theory explains that interpreting and translation is not merely laying down what is said in the source text using corresponding words in the target text; that will result in a text that the target readers cannot understand. The theory emphasizes that the work of interpreting and translation is an extremely creative undertaking. Looking at the process Many attempts have been made to try and determine the process of translation. The various trials and errors of the translator the hesitations, changes of mind and the corrections, assuming they are available for proper scrutiny, however, do not give access to the actual process that goes on in translating, i.e. interpreting graphic signs into sense. Oral translation lends itself better to a detailed examination of the cognitive process of translation. Oral speech is evanescent; its sounds disappear instantly, but sense remains. Formulation in another language shows clearly that sense is made up of two elements: actualized language meanings and cognitive complements. By recording oral translation/interpreting of speeches, it is possible to trace the development of the comprehension of sense , with the false starts, lagging behind, corrections, and anticipation of the rendering in the other language. The interpreters rendering is a direct reflection of his understanding. His formulation is not the transfer of one language into another.

Units of meaning The second-by-second examination of a good simultaneous interpretation shows how the sounds of small segments of utterances deliver their part of the message and fall into oblivion immediately after. The sense of a speech is built upon fragmented utterance after another, the length of which varies. The Paris School calls them units of meaning. Units of meaning are not units in the sense that they can be measured once and for all. Rather, they are born when a sufficient number of words merge with the relevant knowledge, taking on an ephemeral life. One after the other, these units of meaning blend with what has already been stored cognitively and form a more general sense the sense that remains in the minds of the listeners when the speech has come to an end. Compared with listening and understanding a speech in general, the specificity of the interpreters task is to grasp and render each and every unit of meaning. This enables the researcher to visualize the nature of the understanding process, from the fleeting passage of sounds to the full comprehension of speech. Interpreting proves that there are three stages in the process of any oral or written translation: 1) 2) the understanding of sense, a de-verbalization stage, meaning that the words and sentences that gave birth to sense are forgotten, while sense remains present without any linguistic support, and 3) the reformulation of this sense in the other language.

Deverbalization Understanding the concept of sense requires that it be broken down schematically into a double interpretive activity; the interpreting of graphic signs into concepts, then the addition of cognitive complements to such concepts. Actually these two interpretations merge into a single one. Immersed in context, the translator interprets graphic signs directly into sense. This sense remains present as awareness even while the signs fall into oblivion. That stage is the stage of deverbalization. De-verbalization is a natural phenomenon in oral translation, at least in consecutive interpretation. As the sound chain disappears, the good interpreter conveys non- verbal sense into his native tongue. While it is clearly noticeable in interpreting, deverbalization is more difficult to observe in written translation. It is not as obvious because the original text is there. In other words, unlike the sounds of

10

oral speech, it does not disappear. The graphic signs remain, and they call for the proper linguistic concepts and direct correspondences in the other language, short-circuiting the search for appropriate equivalences. Deverbalization is a natural feature of the interpretative process, requiring an effort on the part of the translator. It is nevertheless present in the form of the translator awareness of what an author means in a given passage. Limiting research to grammar and contrastive linguistics in order to overcome the differences in language structures has taken the science of translating no further than what machine translation can achieve. Languages differ in all respects, not only in sound structures, semantics or syntax, but also in the way speakers refer to ideas, facts and events. If interpreters and translators could proceed directly from one language to the other, there would be no point in their existence: the machine would take over. The machines inability and the translators ability are based on the fact that the machine operates solely on the basis of formal signs, while translators and interpreters work mainly on the basis of an awareness of ideas. Understanding a segment of text means de-verbalizing, which enables the translator to discover modes of expression that are not interfered with by the original language. This is the way successful translation works. Formulation or the synecdoche principle The third stage of the process is the search for idiomatic means of expression that render the sense of the original, the actual wording being divorced from the source language, and complying with the usage and customs of the target language. Words in texts are made up of two parts: the graphic sign or explicit part and the implicit part, itself formed of the relevant knowledge readers apply to the explicit part. The Interpretive Theory of Translation claims that languages are not codes so that words as such are not translatable. The object of translation is the sense born out of the merging of language actualized in a text and the receptors relevant knowledge. The explicit parts of a text are known as synecdoches, and languages not only differ in their lexicon and grammar but also in the way the natives express their thoughts. Speech and text are both synecdoches, but in two different languages the same sense is hardly ever expressed with the same synecdoches. Translators who want to be understood have to avoid the structures of the original language and in many instances, its choice of words. In order to achieve a meaningful result, i.e. to transmit sense in a form acceptable to readers of the translation, the translation has to

11

be formulated in a way that agrees with the target language. The whole thing, or sense , is never formulated exhaustively. Only a part of it designates the whole and the parts used are not the same in two languages even though they designate the same whole. To wrap up this explanation of the Interpretive Theory of Translation, it would be interesting to take note of the way Sigmund Freud translated. In his Life and Work of Sigmund Freud , Ernest Jones describes Freuds method: In the first part of the year (1880), Freud was able to cope with the boredom (of military service) by devoting himself to translating a book by John Stuart Mill, the first of five large books he translated. It was a congenial work, since he was specially gifted as a translator. Instead of laboriously transcribing from the foreign language, idioms and all, he would read a passage, close the book and consider how a German writer would have clothed the same thoughts a method not very common among translators. His translating work was both brilliant and rapid. What took place when Freud closed his book after having read a passage from it? What occurred is the phenomenon which provides the key to the mechanism of human language. According to Seleskovitch, the linguistic forms of the original vanished and left behind only the awareness of the sense . What took place when Freud reclothed the foreign thoughts in German linguistic garb? He treated someone elses thoughts as if they had been his own and expressed them as he would have his own ideas. His expressions, being spontaneous and seeking only to convey the sense adequately, conformed to the genius of the German language and thus provided the German reader with a formulation that was crystal clear. Why did Freud work passage by passage, whereas he could have retold the book if his object was to avoid a painstaking transposition? Because discourse is made up of units of sense, and because a translator faithful to the structuring of the original thought is duty bound to reproduce them all. If he had taken portions larger than passages, Freud would have run the risk of missing out part of the content. On the other hand, if he had taken smaller portions for example, if he had worked sentence by sentence he would have relapsed into the span of formal memory (also called immediate or very short term memory), which contains about seven or eight items. He would then have risked falling under the sway of the items of the source

12

language and transposing their meanings; such a transposition, which would have resulted in a text stamped with foreign turns of phrase and betraying the norms of the target language, could only have obscured the original thought. The distance Freud kept from the text enabled him to free the cognitive mechanisms that are necessary to a non-linguistic translation. In consecutive interpretation, the interpreter reproduces the speech in one language after he has heard the entire text in another language. Rather than retain the original formulation, he/she notes only the ideas; accordingly, the reproduction is more in the nature of a narration than a repetition. Comparison of this narration with the original shows the difference which there may be between transposition of meanings from one language system into another and discourse (including both speech and writing), in discourse ideas are formulated spontaneously, thoughts are in direct control. General applicability of the interpretive theory of translation The Interpretive Theory of Translation is based on principles which apply to any language combination; research done on a number of language pairs including French and Chinese, English, German or Japanese validate the assumption that the interpretive process is applicable to all language pairs since it is languageindependent. The methods advocated by the Interpretive Theory of Translation are not only valid for functional translation, but also for economic, political, technical, scientific or commercial texts. There is no a priori obstacle to applying the Interpretive Theory to literature or poetry, as long as the transfer of linguistic elements from one text to another is considered as being not translation but transcoding, the result of which is to show readers of the translations the peculiarities of the original language. The other condition is to define translation as a text that should produce the same cognitive, affective and esthetic effects on readers as the original text does. The Interpretive Theory of Translation also gives guidelines for the treatment of cultural features. The recognition of the fact that sense is not contained in any given language or text but arises from cues given by the text/language and cognitive complements from the target reader is great help for translators in solving problems related to translating a texts cultural aspects. The part played by these cultural aspects in the text, their importance in understanding the authors meaning, and the reasons

13

readers would read the text are weighed by translators who find ad hoc solutions for each cultural point. To conclude, I would like to stress that translation is a very broad notion. It can be used to point to the act of translating or to the result of this act. It covers: 1) 2) translating exercises done in language classes and aiming at improving knowledge of a second language, and comparisons of language pairs which may lead to the establishment of bilingual dictionaries of comparative grammars. These two activities bear on the linguistic aspects of translation and look for correspondences between the source and the target languages. Yet the word translation also covers the cognitive process we are concerned with as translators and translatologists: what is understanding may be quite different from understanding or knowing a language, and why is the designation of a whole concept through a synecdoche the basis for formulating sense in ones own language? *
Editors Notes: This paper is based on the keynote speech given by the author at the 4th Annual Conference of the Japan Association for Interpretation Studies held in Kobe, Japan, on September 23, 2003. We thank Dr. Choi for agreeing to rewrite the speech for inclusion in this issue of Interpretation Studies .

Sohee Jungwha CHOI is a founding member and the current president of the Korean Society of Conference Interpretation. She holds a Ph.D. from the l'Ecole Superieure d'Interpretation et de Traduction in Paris. In 2001, she received the prestigious Prix Danica Seleskovitch in recognition of her eminent contributions to the profession of conference interpretation. She currently resides in Seoul, where she is a professor at the Graduate Program of Translation and Interpretation of Hankuk University of Foreign Studies. Personal Website: http://www.choijungwha.com/

REFERENCES Choi, J.-W. (2002). Note Taking. Seoul: Hankukmunhwasa. Choi, J.-W. (2001). Interpretation & Translation Know-How. Seoul: Nexus. Choi, J.-W. (1998). Introduction to Interpretation & Translation. Seoul: Shinronsa. Delisle, J. (1998). Translation: an Interpretive Approach . (Translation Studies, No. 8) University of Ottawa Press. Lederer, M. (1981). La Traduction Simultanee. Paris: Lettres Modernes.

14

Lederer, M. (1994). La Traduction Aujourd'hui. Paris: Hachette. Lederer, M. (1998). The Interpretive Theory of Translation: a brief survey. Publicos de la Ciudad de Buenos Aires. Seleskovitch, D. (1975). Langues et Memoire. Paris: Lettres Modernes. Seleskovitch, D. (1978). Interpreting for International Conferences. Washington, D.C.: Pen and Booth. Seleskovitch, D & Lederer, M. (1993). Interpreter pour Traduire (3e edition). Paris: Didier Erudition. Seleskovitch, D & Lederer, M. (1989). A Systematic Approach to Teaching Interpretation. Silver Spring, MD: The Registry of Interpreters for the Deaf.

15