+ All Categories
Home > Documents > What is a Translating Translator Doing? by Brian Mossop

What is a Translating Translator Doing? by Brian Mossop

Date post: 23-Feb-2018
Category:
Upload: donna-noble
View: 221 times
Download: 0 times
Share this document with a friend

of 36

Transcript
  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    1/36

    What Is a Translating

    Translator Doing?

    Brian MossopGovernment of Canada Translation Bureau and

    York University School of Translation

    Abstract:Translating is here defined as the quoting, in sequential chunks, of thewording of a written, oral or signed text, with an imitative purpose. These

    features distinguish it from other sorts of language activity intralingualparaphrasing, re-expressing of ideas, fictive quoting, speaking from a script,ghostwritingandthusprovide an object for a theory of translation production.The defining feature 'quoting ' is taken to involve demonstrating to someoneselected features of the source text. Thus the translational quoter is engaged in adual activity: quoting OF the source text (rendering work) and quoting TO thereaders or listeners (pragmatic work). The texts commonly called translationsarise from some combination of rendering, pragmatic and non-translationalwork.

    Rsum: On propose de dfinir l'acte de traduire comme le fait de citer, l'unaprs l'autre, et avec une vise imitative, les fragments et phrases qui constituentun texte crit, oral ou sign. L'acte traductif se distingue ainsi d'autres types de

    production linguistique, comme la paraphrase intralinguale ou la r-expressiond'ides, et on parvient ainsi dfinir l'objet d'une thorie de la productiontraductive. Le dfinisseur citer estprendre dans le sens d'une dmonstration

    de certains traits choisis du texte de dpart. Le traducteur-citeur s'engage dansune activitdeux volets : citationDUtexte de dpart (il rend ce texte) et citationadresseAUXlecteurs ou auditeurs (travail pragmatique). Les textes communment appels traductions sont produits par une combinaison de plusieurssortes de travail linguistique: travail rendant le texte de dpart, travail pragmatique, travail non traductif.

    1. Defining Translation: Five Criteria and Four Deadends

    The title question is not to be confused with the psycholinguistic question:what is happening in the mind of the translator? Nor is it to be taken as

    Target 10:2 (1998), 231-266. DOI 10.1075/target.l0.2.03mosISSN 0924-1884 / E-ISSN 1569-9986John Benjamins Publishing Company

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    2/36

    23 2 BRIAN MOSSOP

    inquiring about the diverse economic, cultural or political functions which thetranslator is serving: extending a literary tradition by retranslating a novel;censoring information from foreign sources; enlarging a company's market bytranslating product information; allowing people to avoid learning the language of a neighbouring community.

    The title question is asking for a defintion of the object of translationproduction theory, as suggested by Anthony Pym's (1993) nice phrase "translating translator" the translator at work alternately reading or listening to a

    bit of the source text and producing a bit of text in the second language. A

    translation reception theory answers the very different question: what dotranslations do? It is concerned with the moment(s) when completed translations affect readers and listeners, and through them, the target culture, including its language and system of texts. Reception theory needs its own definitionof translation, perhaps one involving the venerable concept of equivalence: asPym points out (1992:37ff, 64), equivalence defines translation in the sensethat when people are persuaded to receive a text as a translation, they believethey are somehow receiving the original message directly from the ST author;the meaning-producing activity of the translator is overlooked.

    According to Levy (1967), what the translator is doing is deciding whichof many possible wordings in the second language to select. This answer to thetitle question opens the door to a search for the social and psychologicalfactors impinging on the translator's decision. However it assumes we havealready answered another question: when is a person being a translator? In thisarticle I will define translation production in a way that makes it possible tosay that people called translators are sometimes translating, sometimes not.Their language production work may be entirely, mainly, partly or not at alltranslational.

    I suggest that a useful answer to the title question will meet five requirements. The definition should be (i) non-normative, (ii) general, (iii) narrow,(iv) multidimensional, and (v) neither psychological nor sociological:

    (i) Non-normative.The definition will not be a disguised restatement ofsome translational ideal,1such as Nida's famous "translating consistsin reproducing in the receptor language the closest natural equivalentof the SL message, first in terms of meaning and second in terms ofstyle" (1969: 12). Such ideals reflect the habits of a particular group

    of translators translating a certain type of text. The aim rather should

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    3/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 233

    be to find a definition for the process of translating-in-general (henceforth Translating), valid for all times, places, language pairs, texttypes, and approaches to the task including approaches that resultin "bad translations" and "mistranslations", as distinct from non-translations.

    (ii) General.The definition will embrace work from and to oral, writtenand signed texts as well as work in all modes which, pre-theoreti-cally, we think of as translational: interpreting dialogue in court,subtitling films, translating books, and so forth. If it seems that no

    interesting generalizations can be made about just this range oflanguage production work, then Translating per se will not be atheoretical object: the production of subtitles, court interpretationsand translated novels will be subsumed in film theory, the sociologyof law, and literary history respectively. The criterion of generality isalso incompatible with dividing translation into a myriad of "translational practices" and personal translation projects (Poundian translation, feminist translation, belles infidles, etc.), all self-defining andwith merely a family resemblance to each other.

    (iii)Narrow.Translating will not be characterized as a subtype of somebroader object, such as intercultural communication, metatextualproduction or mediation. The main point of a defining exercise is todraw attention to what is peculiar about translation production. If itseems that generalizations can be made only about one of these

    broader objects, then the project of theorizing about Translating willhave proven to be wrongheaded.

    (iv)Multidimensional. The definition will distinguish translation production from as many kinds of non-translational language production

    as possible.2This means the definition will have several parts. Eachof the parts will be shared with several non-translational activities,

    but together they will single out Translating.(v) Neither sociological nor psychological. Perhaps one day we will

    have a psychological characterization that will distinguish the mentalprocesses of translational comprehension and composition from non-translational processes. Or a sociological characterization that willdistinguish translators from non-translators in terms oftheeconomic,

    political and cultural relationships into which translators enter. Or apsycho-social characterization distinguishing translators from other

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    4/36

    234 BRIAN MOSSOP

    language producers in terms of their scope for creation given thesocial constraints impinging on their work. These would be empiricaldistinctions, answering the question: is there something unique abouthow translators process language in the mind, or how translation isembedded in social structure? Before framing hypotheses about suchdistinctions, we first need to identify their object: which part ofsomeone's speaking or writing is the translational part?

    For definitional purposes, Translating is best seen not in sociological orpsychological terms but as an act of text-linguistic communication. Of course,

    the common notion of the translator as communicator is not of much use initself,since most speaking and writing is communicative. The definition mustdistinguish translational from non-translational communication, includingnon-translational interlingual communication.

    If the above requirements are accepted, then the following four approaches will be rejected as deadends:

    (a) Identifying a cut-off point on the literal-free scale (fails to meet criteria iand iv above)

    Separating translations from adaptations is a hopeless task. The cut-off pointwill inevitably be arbitrary, dictated by the definer's normative concept oftranslation, or the concept currently prevalent in a society, or among professional translators of a particular type of text. Typically the cut-off point isidentified (rather than defined) by pointing at a text deemed to be mereadapting or at the other end of the scale mere transcoding. This approachis unidimensional and negative: a single feature is invoked (e.g. correspondence to the sensof the source text, cf. Seleskovitch 1982), and translation iswhatever is left over once the two non-translational ends of the correspondence scale are eliminated.

    Related to the translation/adaptation distinction is the equally hoaryquestion of whether translation is derivative or creative. For definitional

    purposes, this question too is pointless; any answer will simply reflect theoutlook of a particular time and place. From the definition of Translating thatwill be proposed here, it follows that it is always both creative and derivative,

    but then this is true of all types of language production: in some sense, "everyutterance repeats" (Compagnon 1979: 9, my translation) and yet at the same

    time offers something new because the repeater is responding to a new

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    5/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 235

    situation. The question is: what is the particular way in which translation"repeats"?

    (b) Separating literary from technical translation (fails criteria ii and v)

    The classic statement of this approach is Schleiermacher's (1813) distinctionof Ubersetzen from Dolmetschen. The distinction is not one of written versusoral translation, as the German words suggest, but rather scholarly versuscommercial translation, or literary/humanistic versus scientific/technical. TheUbersetzer works with texts that reflect one person's unique thought, while

    theDolmetscher works with texts that reflect the extralinguistic world. Muchwriting about translation implies the need for two theories a psychologicaltheory of Ubersetzen (seen as the translator's mental struggle to shape thetarget language in order to convey what is culture-specific or author-specificin the source text), and a sociological theory of Dolmetschen (seen as thecirculation of international, authorless concepts).

    Once the Ubersetzen/Dolmetschen division is allowed, the resulting fragmentation cannot be halted in any principled way. There will have to be adefinition for every use of language: for poetry, where rhythmic and phonetic

    effects are especially important; for court decisions, where real events areinterpreted by reference to a world created by the wording oflaws.And so on,and on and on.

    One can certainly relate language production to the various larger spheresof human activity in which it is embedded, but the effect of such a move is tofocus on the social functions of translation rather than its text-linguistic nature.Some translation {Dolmetschen) is then seen as an aspect of internationalcommerce or politics, some (Ubersetzen) as an aspect of cultural history.More generally, the focus shifts to regional or global communication circuits

    and away from the individual communicative act of language production.(c) Using a situation-based definition (fails criteria iii and v)

    Could we start from a primal situation in which interlingual communicatingactivity (ICA) occurs? Two human groups come into contact, cannot understand each other's speech but must have or wish to have dealings. Unless alingua franca is available or a pidgin is developed and suffices for their

    purposes, some members of one or both groups must learn the language of theother and serve as intergroup communicators in one or both directions until

    such time as bilingualism becomes generalized or one of the languages ceasesto be used.

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    6/36

    236 BRIAN MOSSOP

    With the advent of writing (and later, sound recording), a secondaryinterlingual communicating situation arises. It becomes possible to selectearlier communications that took place within one language community, evenif its language is no longer in use, and re-present them in another language fora new audience.

    Could we then define Translating as the linguistic activity that goes on inthe primal or secondary ICA situation? Such a sociological definition willembrace as translational all language activities engaged in by interlingualcommunicators in ICA situations. Unfortunately this includes the creation of

    derived and co-produced texts. In co-production, a selection of themes isexpressed in L1 and, separately, in L2, as happens in certain multilingualadvertising campaigns. In derivation, an L2 text is produced that drawsgeneral inspiration from an existing L1 text, but the L2 producer expressesideas drawn from an independent stock of thoughts rather than working underthe control of the detailed wording of the L1 text. An example would be theEnglish version of the French song which is now Canada's national anthem,discussed further in Section 3.

    A theory of language production in the ICA situation would thus include

    Translating, but it would not be a theory of Translating per se. A similarproblem arises if the primal situation is seen in intercultural rather thaninterlingual terms. It has become commonplace in reaction to the notionthat translation is an application of linguistics to define translation asintercultural communication. Unfortunately, if one asks what kind of intercultural communication it is, the answer must surely be, rather uninformatively,that it is the interlingual kind! It is not clear how translational and non-translational intercultural communication could be related in order to bring outthe peculiarities of Translating. On the intercultural communication scene,

    translation is really rather an oddity. Internationally, language mixtures andlingua francas are used more than translation (e.g. English for communication

    between Japan and the US), while in multicultural societies, immigrantscommunicate with hosts mainly by learning the hosts' language. And in bothcases, non-linguistic communication is very important. In short, inter-culturality is far too broad a concept for the purpose of defining Translating.

    (d) Requesting labels (fails criteria i, iii and v)

    Could we define Translating by asking speakers and writers to label their

    activity, or by asking an observer to label it? This would create two difficul-

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    7/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 237

    ties.First, people may present their work as translation even though in fact itwas produced in some other way. Sometimes the texts so presented arederived or co-produced. Sometimes they are TL originals which are presentedas translations because that is thought to confer prestige on them or, underother circumstances, shield the authors from responsibility for their content. Aself-proclaimed translation may also be simply a reworking of older translations. This is the case with the King James Bible of 1611. Comparison of a

    passage with the original Hebrew or Greek might well lead one to receive it asa translation, and indeed this is how the front page presents it ("Newly

    Translated out of the Originall tongues: & with the former Translationsdiligently compared and revised"). Actually, the production process consistedin selecting the best 16th century translation of a passage, only retranslating ifno previous translation was suitable (which was not often), and then puttingthe result through the English style editing which gives this version itsdistinctive rhythms and flavour (Opfell 1982). Thus most of the language

    production work was not Translating at all.The second difficulty with requesting labels is that we will get not one

    term but a lengthy list: not just 'translating' but also 'rewriting', 'interpreting',

    'adapting', and so on. We would have to decide whether someone who claimsto be 'adapting' or 'rewriting' is Translating or doing something else. Wecould ask third parties, but they might disagree: one might insist that theactivity is adapting rather than translating, another that it is translating, another that adapting is a kind of translating. If the Koran is put into English,anglophone Muslims will call that activity adapting or interpreting and denythat it is translating; they will see the English as a guide to the original Arabic,which alone constitutes the word of God. Non-Muslims will call the activitytranslation. And what of a francophone journalist in Quebec who describes her

    activity as rdiger les nouvelles, 'writing the news', even though an outsideobserver would describe it as translating an international press agency itemfrom English into French?

    This brings us to the problem of other languages. Is an activity labelled'traduire' or 'perevodit" or 'ubersetzen' an instance of Translating? There areseveral complications. Some languages refer to ICA with a word whosemeaning is general enough for that purpose but does not actually include ICAas a meaning component (think of English 'convert' or 'transform'). Otherlanguages have no general word like 'translate' but only specific words, like

    'subtitle', which refer to the translation of texts in a particular medium, genre

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    8/36

    238 BRIAN MOSSOP

    or source language. And of course in all languages, words can change theirmeaning or their reference over time: an activity described in one era as'translating' might be denied that label a hundred years later.

    Let us use the expression 't-words' for the full set of words, in alllanguages, which people use, or have in the past used, to describe their ownactivity or the activity of others as ICA. For the observable activities referredto by t-words, let us use the notation 'translating...'. It seems clear thattranslating... cannot be a fruitful object for translation production theory

    because the t-words cover such a disparate group of activities: if we know only

    that someone is translating..., we do not know whether that person is engagedin the same activity as someone else who is also translating..., that is, we donot know whether one, both or neither of them is Translating in a universalsense.

    Thus the object of Translating theory should not be defined by referenceto social labels. In the approach to be taken here, the producer of any text can

    be engaging in several language production activities at once, or switchingfrom one activity to another as the text is created. The point will be to decidewhich activity or combination of activities should be called Translating for

    theoretical purposes. Passages in an observable text may arise from purelynon-translational activity, from Translating plus non-translational activity, orfrom Translating alone. This points to another aspect of the title phrase"translating translator": I am interested in people called translators only insofar as they are Translating.

    In the next three sections, I will build up a definition of Translating thatmeets the five criteria. The defining features are quoting (Section 2), sequential rewording (Section 3) and imitative purpose (Section 4).

    2. Translating Is Quoting

    2.1.TheTranslatorasRapporteur

    In Mossop 1983, I proposed that Translating is reporting, and represented thisgraphically with the diagram3 reproduced in Figure 1 :

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    9/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 239

    x c

    X is reporting to C what A said/wrote/signed to B

    Figure 1. Translating as Reporting (from Mossop 1983: 246)

    Since the first defining feature, 'quoting', is a kind of reporting, I want toexpand at some length on this diagram and its caption, drawing attention not

    just to what it shows but to what it doesnot show.The relevant definition of 'report' in Webster's Tenth New Collegiate

    Dictionary is 'to relate the words or sense of (something said)', but nothinghinges on the specifics of this English word since the activity diagrammed inFigure 1 is a universal one. The dictionary definition does point to one greatadvantage of the concept of reporting, namely that it circumvents the notori

    ous problem of intended meaning. One does not report intentions; one reportswhat someone said. X's report of what A said to B may include both intendedand unintended meanings.

    Reporting is to be distinguished from stating the meaning of a passage ina text. In a too-much-cited article, Jakobson (1959) says that a translation is areported speech, but a close reading reveals that the article is actually aboutstating in L2 the meaning of an L1 expression. Consider a French text on fruitharvesting that contains the phrase blessures mcaniques de pommes 'mechanical injuries of apples'. I can state the in-context meaning of blessuresmcaniques either in French ('lsions la suite d'un accident plutt qu'uneinfection') or in some other language ('defects of mechanical rather thaninfectious origin'). Now suppose I produce this English, talking either tomyself or to a colleague: "He means defects of mechanical rather than infectious origin". Such a statement is not functioning as a report; it is a heuristicdevice whereby I spell out the meaning of the passage in hopes of eliciting auseable translation (such as 'bruising'). As Jakobson puts it:

    . . . the meaning of any linguistic sign is its translation into some further,

    alternative sign, especially a sign "in which it is more fully developed", asPeirce, the deepest inquirer into the essence of signs, insistently stated. The

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    10/36

    240 BRIAN MOSSOP

    term 'bachelor' may be converted into a more explicit designation, 'unmarried man', whenever higher explicitness is required. We distinguish threeways of interpreting a verbal sign: it may be translated into other signs of thesame language, into another language, or into another, nonverbal system ofsymbols. (1959: 232)

    Jakobson immediately contrasts this with reporting:

    Most frequently, however, translation from one language into another substitutes messages in one language not for separate code-units but for entiremessages in some other language. Such a translation is a reported speech.(1959: 233)

    Nothing more is said about reported speech, and it is hard to see any connection between Jakobson's mention of it and the main line of his discussion,where a 'translation' is an elaborated statement of meaning.

    The past tense in the caption of Figure 1 (A said to B) refers to the orderof creation, not the order of presentation. Film subtitles in L2 are presentedsimultaneously with the on-screen dialogue in L1 but were of course createdafter it. No Translating can occur until at least some of A-B has already been

    produced. Thus if a language producer is thinking in one language and then

    speaking in another, no Translating is occurring because no source text in thefirst language has been produced. Similarly, the process by which thought isexpressed as text is not (contrary to Mel'chuk 1978) Translating, since thethought is not a prior text A-B. Thought, while perhaps executed inlanguage,is not text publicly available language. So when Bandia (1993) findsevidence that West African novelists writing in their second (European)language were thinking of expressions in their native languages, he has notfound a case of self-translation.

    The caption of Figure 1 does not imply interlinguality. The diagram

    therefore covers such intralingual activities as paraphrasing medical talk in laylanguage. This is not a defect that could be repaired by amending the captionto read: "reporting to C in L2 what A said to B inL1".Reporting is reporting;it does not come in an intralingual and an interlingual version. Theinterlinguality of Translating is discussed in Section 5.

    The diagram also covers both Translating and deriving. Deriving, asdescribed in Section 1, involves a source text A-B, but X does not process itsequentially, in chunks ranging in size from a phrase to a couple of sentences,and does not necessarily aim to imitate it. These two defining features of

    Translating sequentiality and imitation are discussed in Sections 3 and 4respectively.

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    11/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 241

    The diagram defines Translating in terms of the activity of the translatorX. The author and addressees of the source text, as well as the readers of thetranslation, certainly appear in the diagram, but they are present only asimagined by the translator.The translator's interpretation of what A said to Bis in the diagram, but not A as such. The translator's concept of those who will

    be reading the translation is present, but not the readers as such. C is theimagined reader, not the real one, who is of concern only to reception theory.

    The diagram reflects the translator's communicative position at the moment of production. It does not show the translator's social position the

    actual economic, cultural and political relationships in which he or she isinvolved. Nor does it show the translator's location in the source-textsociety, the target-language society, some third society or an in-betweenzone.4All these factors are vital for descriptive and explanatory purposes, butnot for the purpose of defining translation production. Indeed their inclusionwould be incompatible with the goal of a general definition: what a community interpreter in a hospital is doing, sociologically speaking, is indeed quitedifferent from what a translator of Homer is doing.

    Because the diagram is not a description, but simply a representation of

    one proposed defining feature of Translating, it omits many complications.For example in the written but not the oral mode, X is physically isolated fromA, B and C, and there may be a gap of thousands of years between A-B and X-C. The diagram also abstracts away from cases in which the translator does notdirectly address the intended TL audience:

    - C is another translator who turns X's translation into a third languageL3 when no one is available to translate directly from L1 to L3;

    - C is a dialogue writer, who reworks a rough translation of film

    dialogue for use by a dubbing actor.In these cases, X is still reporting to C what A said to B, but knows that C is notthe final addressee. Cases in which an editor intervenes between X and Cwithout X's knowledge are not covered, since the diagram represents thesituation from the translator's point of view.

    Figure 1 is further restricted to the translator's activity qua reporter. Ofcourse, he or she may also be doing things as a consequence or goal of re

    porting. For example, a book translator's work may have the consequence ofenriching the target language, or it may have the goal of introducing a new

    genre. In one peculiar class of cases, where C is already familiar with the

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    12/36

    242 BRIAN MOSSOP

    source text, the goal is not even to convey the meaning of the source:translations prepared by second-language students (or translation students!)for their teachers; translations of Greek and Latin texts into European vernaculars for readerships already familiar with the source (Bassnett 1980: 69, 44).These cases are marginally translational since one does not normally reportthat which is believed to be already known. When Chapman and Pope translated the Iliad into English in the 17th and 18th centuries respectively, theyknew that many of their readers would not want to know what Homer said;like the language teacher, they would be interested in the translations as

    translations of a source text they already knew.Translators may also be doing things in addition to reporting. For example, a community interpreter may be serving as an advocate or as anindependent source of information. Consider the following passage and itstranslation by an interpreter in a hospital in Winnipeg, Canada. The patient, aCree Indian, is anemic and the doctor thinks she may be losing blood during

    bowel movements because of a benign polyp that has been discovered bycolonoscopy:

    Doctor: I can take the polyp out without an operation, by putting a tube inside

    the bowel, and putting a wire around it and burning the polyp off. That stopsthe bleeding, no need for an operation.

    Interpreter (backtranslation from Cree): He says they can put in a tube andburn off the growth. If it's not removed, you may end up with cancer. Andyou won't need an operation. This procedure will get it in time. Before it

    begins to bleed or starts to grow, (adapted from Kaufert and O'Neil 1990: 49-50)

    The part about cancer is not reporting, in that it does not relate anything said orimplied by the doctor. Instead, the interpreter is acting as an independent

    source of medical information: she knows that polyps can become cancerousand says so. Some practitioners of community interpreting might respond bysaying that the interpreter is not in fact interpreting (the normative cut-off

    point approach discussed in Section 1 above). Some theoreticians, on thecontrary, might respond by saying that if community interpreters do indeedwork like this, then Translating must be defined to include it. In my view, the

    proper theoretical response is to say that the interpreter is engaging in twodifferent activities simultaneously: she is Translating (reporting) and she isalso providing independent information. This approach is elaborated in Sec

    tion 6.

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    13/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 243

    Figure 1accounts for all media and modes of translation: written, oral andsigned texts; book translating, dialogue interpreting, film dubbing and so on.One source of its generality is that the letters A, B, X and C represent not four

    particular individuals but four roles. Individuals can take two roles at once, orswitch roles. Thus in self-translation, A=X. In oral or signed dialogue inter

    preting, B=C and the individuals playing roles A and C switch as do thesource and target languages with every turn in the dialogue.

    Dialogue interpreting is unusual: in typical written as well as much oraland signed translation, the original communication A-B is independent of the

    translational communication X-C; A is not addressing C. But in dialogueinterpreting, the source text is specifically addressed to someone who does notspeak the source language: an English-speaking doctor addresses a Cree-speaking patient. Nevertheless it remains the case that X is telling C what Asaid to B(=C). More seriously, dialogue interpreting may lead us to questionthe theoretical separation of production from reception. In all other modes oftranslation book translation, simultaneous interpretation and so on reception is irrelevant to production: the actual (as opposed to anticipated)verbal reactions of receivers have no effect on the translator's production

    work. In dialogue interpreting, however, receivers produce the next bit ofsource text in response to the translation of the previous turn in the dialogue.To maintain the production/reception distinction, it may be necessary to takeeach turn in the dialogue as a separate text. I leave this matter unresolved.

    2.2.IndirectReporting, Exact-wordsQuoting, FictiveQuotingand

    Free DirectQuoting

    When I say that Translating is reporting, I do not mean that it islikereporting.

    I mean that a translator is actually producing the grammatical/rhetoricalstructure known as reported speech. In two common cases, this is immediatelyapparent because there is a quotational phrase (italicized in the followingexamples):

    Indirect reporting

    (1) A to B(=C): Je me sens mal.X to C:She says she doesn't feel so good.

    This is the form favoured by non-professional dialogue interpreters, such aschildren translating for their immigrant parents.

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    14/36

    244 BRIAN MOSSOP

    Direct reportingIn a television documentary prepared by the Canadian Broadcasting Corporation, a dancer at the Kirov Ballet in St. Petersburg was being interviewed.Viewers could see and hear her speaking in Russian, with voiceover by theCanadian journalist in English:

    (2) X to C: I'm not as well off as under communism, she's saying.

    Quotational phrases occur in oral translations of turns in dialogue as in (1), aswell as in oral or written translations which are embedded in a larger TL text

    the Canadian journalist 's narrative surrounding (2). Both the indirect anddirect forms can also occur as translations without a quotational phrase:

    (la) X to C: She doesn't feel so good.

    (2a) X to C: I'm not as well off as under communism.

    Someone reading or listening to (la) or (2a), if it occurred as part of a largertranslated text, might or might not be able to recognize it as translation. Withlive oral translation, which may use either the indirect or the direct form, thefact of reporting will be evident from the physical presence of the interpreterand of someone speaking in another language. In written translation, whichtypically uses the direct form (2a), the fact of reporting will often not be clearto the reader. But to the translator, producing text of type 2a is always a.matterof reporting a prior text.5

    Now intralingually, direct reporting is usually called quoting. It is customary to think of quoting as reproducing the exact words of the source, and in thissense of the word, translating obviously cannot be quoting, since all or almostall the words are changed. I suggest, however, that this commonplace view

    makes for a very unsatisfactory theory of quoting. It is true that in certain writtencontexts (in law, scholarship and journalism), there is an ethical requirement toreproduce exact wordings, but these are best seen as special cases. Most often,quoting is a kind of dramatization: when narrating events, people put theirsources on a metaphorical stage and make them speak in the first person:

    (3) I ran into Gwendolyn yesterday and she said I hear you're notgoing to the translation conference; I said no I don't think so\ shesaid I really think you should.

    Of course one reason exact words are not used in reports of oral conversationis that the narrator most commonly cannot remember the exact words. But that

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    15/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 245

    does not make oral quoting defective, for the dramatic purpose of quoting is inno way enhanced by the use of exact words. For clarity, I shall refer to suchdramatized (first-person) reporting of others' words as "free direct" quotingwhen it is necessary to distinguish it from the exact-words variety or fromindirect reporting.

    Herbert Clark and Richard Gerrig (1990) propose a theory of quoting thatseems able to account for free direct quoting, including its translational variety.6

    The theory defines quoting in terms of its nature as a language act, not in termsof its functions. To simplify, Clark and Gerrig argue that an act of quoting is one

    in which the speaker purports to demonstrate rather than describe selectedaspects of something that has been said or written. Consider:

    (4) "I'll get it translated by tomorrow", said Gwendolyn slowly.(5) "I'll...get... it... translated...by...tomorrow", said Gwendolyn.(6) Gwendolyn promised to translate it by the next day.

    The speaker of (4) describes the fact that Gwendolyn spoke slowly whereasthe speaker of (5) demonstrates it (not very successfully in the above writtenversion; in the oral equivalent, the quoter would speak slowly). The speakers

    of both (4) and (5) demonstrate the illocutionary force (promising) of whatGwendolyn said, whereas the speaker of (6) describes it.Clark and Gerrig provide a partial list of the things a quoter can decide to

    demonstrate, including voice pitch, speech defects, level of formality, propo-sitional content, illocutionary force, exact words uttered (just one optionamong others) and, of course, the language used by the original speaker:

    (7) "Je le ferai traduire pour demain", said Gwendolyn.(8) Speaking in French, Gwendolyn promised she'd have it translated

    by Friday.(9) "I'll get it translated by tomorrow", said Gwendolyn in French.

    The speaker of (7) demonstrates the fact that Gwendolyn spoke in French. Thespeaker of (8) describes this fact, as well as the illocutionary force and

    propositional content of what she said. Interlingual indirect reports like (8) donot demonstrate anything. Even if, on the model of sentence (la) above, weconsider the utterance She 'll have it translated by Friday, this too is descriptive rather than demonstrative in relationtoJe le ferai traduire pour demain. Itis thus non-quotational and if quoting is a defining feature of Translating

    non-translational. However, since such sentences are used in interlingualcommunicating situations, by people called translators, we might alternatively

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    16/36

    246 BRIAN MOSSOP

    view indirect reports as marginally translational.Turning to sentence (9), here the speaker describes the fact thatGwendolyn spoke in French, but demonstrates the force and content of whatshe said. The portion inside quotation marks is therefore translational. Clarkand Gerrig's interpretation of sentences like (9) is vastly superior to thatoffered by Ann Banfield, according to whom people take them as paradoxes(the quotation is in English, not French) and then resolve the paradox throughan "implicit belief in ... a kind of universal language which can be represented

    by particular languages" (1982: 248-249). I suggest, following Clark and

    Gerrig (1990: 798), that people do not take such sentences as paradoxes at all,but instead as perfectly ordinary instances of quoting.Now most translations, unlike the one in (9), are not presented as quota

    tions. Nevertheless (9) is a model of what every translator is doing. Given aFrench text containing:

    (10) A-B: . . . Je le ferai traduire pour demain . . .

    I can demonstrate various aspects of it, for example its illocutionary force andpropositional content, some of its lexico-syntactic structure, and its level of

    language, by producing:(11) X-C: . . . I'll get it translated by tomorrow . . .

    Conceiving Translating as quoting has the virtue of strengthening one ofthe classic answers to the old question about the possibility of translation.Aside from the common bio-ecology all human groups share (which providesextralinguistic reference points for two-language communication), translationis made possible in part by the structural universals of human language. As asupplement to this latter point, we can now say that translation is possible

    because quoting is possible: all languages provide lexico-syntactic devices fordemonstrating, or dramatically representing, the discourse of others.

    The dramatization metaphor suggests the creative aspect of the translator's work. However it does have a drawback in that it may suggest a fictivetype of language production, which is not intended. In a discussion of the usesof quoting in conversation, Tannen (1989: 98ff) emphasizes the creativeaspect of quoting but declines to pick out those cases in which the quotationrefers back to a specific real prior discourse. Instead, her discussion is verytarget-context-oriented: quoting is seen as having a variety of functions in the

    rapporteur's context. For example, the quotation may simply be a way oflivening up an otherwise dull narrative: the dramatized persona did not actu-

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    17/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 247

    ally say anything at all. Or the rapporteur is speculating about what someonewas thinking, and presents the thought as a quote, as in the colloquial "Peterwas like: why are you doing that?", meaning that Peter reacted in such a waythat one could speculate that he was thinking "why are you doing that?".These kinds of fictive quoting work reporting of imagined thought orspeech are not included in the concept of free direct quoting. In Translatingthere is always a real textual source, and, as we will now see, the translatorfollows its detailed wording.

    3. Translational Rewording Is Sequential

    The defining feature 'free direct quoting' already contains the notion that inTranslating, there is a real source text. But sources can be treated in verydifferent ways.The feature under discussion in this section sequentiality refers to a very peculiar aspect of Translating, namely that the languageproducer is reading or listening to the prior textashe or she produces the newone,andisprocessingthesource chunk by chunk, a phrase or sentenceat atime.

    (In consecutive interpreting, this may happen in twosteps:the translator writesnotes as the oral text is produced, and then speaks a translation based on thenotes.)

    The main point of this second defining feature is that in Translating, thesource text is not treated as merely a source of thoughts to be expressed; it is

    processed as text.This is to be contrasted with writers who work partly orentirely from a stock of thoughts, as shown in Figure 2:

    text N

    text Btext A

    translating textA into L2

    deriving textin L2 or L1 fromtext A plus stockof thoughts

    stock of thoughts(themes, beliefs, etc.)

    re-expressing thoughtsin L1 or in L2(if both: co-production)

    Figure 2. Textvs.Stock ofThoughtsas Sources for Language Production

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    18/36

    248 BRIAN MOSSOP

    Here are three examples.(i) A popular exposition in German about special relativity will not be an

    intralingual rewording, sentence-by-sentence or even paragraph-by-para

    graph, of Einstein's 1905 article, which the popularizer may well never

    have read even though it is the ultimate source of the thoughts from which

    s/he is working (it would be one of the texts B...N in Figure 2 that is

    connected to the stock of thoughts by broken lines). Interlingually, an

    English-language exposition will not be a translation of Einstein. If a few

    sentences do prove to be sequential rewordings, then the writer was both

    Translating and popularizing.

    (ii) An autobiographer may describe her life to her ghostwriter, or an accused

    may tell his story to a lawyer. Ghostwriter and lawyer then write or speak

    on behalf of their clients, as does a dialogue interpreter in a hospital or

    court. The difference is that lawyers and ghostwriters do not in principle

    work from the detailed wordings of source texts; they may occasionally

    reword from notes they have taken or documents they have been given by

    their clients, but this is not inherent in their activity.

    (iii) The English version of Canada's national anthem was inspired by the

    original French, but is not a sequential rewording of it, as can be seen

    from the first few lines:

    [French original by Adolphe Routhier 1882] [derived English text by R. Stanley Weir 1908]

    O Canada! Terre de nos aeux! O Canada! Our home and native land!

    [O Canada! land of our ancestors!]

    Ton front est ceint de fleurons glorieux! True patriot love in all thy sons command.

    [thy brow is circled with glorious garlands]

    Car ton bras sait porter l'pe, With glowing hearts we see thee rise,

    [for thy arm knows how to bear the sword]

    Il sait porter la croix! The True North strong and free!

    [it knows how to bear the cross]Ton histoire est une pope And stand on guard, O Canada,

    [thy history is an epic]

    Des plus brillants exploits. We stand on guard for thee.

    [of the most brilliant exploits]

    An English translation made a few years earlier had not been well received,

    presumably because the themes of the French did not correspond to the

    English-Canadian outlook at the time.

    Now there are certainly cases where it is hard to distinguish the product of

    sequential rewording from the product of thematic re-expression. On a recent

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    19/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 249

    plane trip, the flight attendant announced in English thatwe have received ourlanding plansand then in French thatnous sommes autoriss atterrir 'we areauthorized to land'. Judging from her delivery, I do not think she was readingfrom a script. The question is whether she was re-expressing a stocked thought(i.e.co-producing, once in English and once in French) or using the English asa source text. And if the latter, was she Translating or deriving? It's hard to saywith so brief a text, because we cannot really ask whether she was sequentially

    processing chunks. And what of those West African novelists writing in theirsecond language (cf. Section 2.1)? If the writer was working from a passage in

    a particular story taken from the oral tradition of his native language, then theactivity would qualify as Translating provided the resulting English or Frenchcan be connected not merely to a theme of the SL story but to its detailedwording. Of course the distinction between theme and wording could bedifficult to make here, since each telling of a story will have a somewhatdifferent wording.

    Such difficulties should not however lead us to conflate rewording andthematic re-expression, as is commonly done under the rubric 'rewriting'.These are two very different modes of language production. It is irrelevant

    whether we, as observers of a resulting text, are able to decide whether it wasproduced by rewording, re-expressing, or a combination of the two. Neubert'suseful definition of translation as "text-induced text production" (1985: 18) isthus insufficient because it does not distinguish the wording of an inducingtext from its meaning. Translating starts from words, not ideas.

    Let us now turn to the final defining feature of Translating. Whenlanguage producers use source texts only as thematic inspiration or when theywork from a stock of thoughts, they may or may not aim to be loyal to theirsources. In Translating, however, loyalty is always an aim.

    4. Translating Has an Imitative Purpose

    Editors process a source text in sequential chunks, but their goal is to improveit, not imitate it. Translating is here defined as having an inherently imitative

    purpose. As Folkart points out (1991:399), once a decision is made to discusstranslation in terms of reported discourse, the framework thus adopted isinevitably mimetic. Translating is thus distinguished from activities which areessentially oriented toward the requirements of the target context; that is,

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    20/36

    250 BRIAN MOSSOP

    activities in which the source text is not important initself.7

    A person called atranslator who at a particular moment in the production of the L2 text iswriting, speaking or signing without an essentially imitative intent is not atthat moment Translating. A substantive proposal concerning the nature ofimitation is outlined in Section6.1.As will be seen in 6.2, however, Translating is by no means purely imitative.

    For the purpose of defining Translating, there is no need to quantify thedegree of imitation sought, or to give a positive definition of imitation. It issufficient to describe it negatively, as the absence of an intention to add,

    subtract, parodyetc.The degree of imitation actually achieved is of course notrelevant to a production theory.Rewording within a language does not, generally speaking, aim at imita

    tion. For example, when plain-language versions of legal and bureaucratictexts are created by rewording an original text that uses "legalese" or"bureaucratese", the purpose is to clarify. Elsewhere (Mossop, unpublished), Iargue that, precisely because of their clarifying purpose, paraphrasing activities should not be conceived as "intralingual translation". If a nurse para

    phrases for a patient a doctor's medicalese, and herself uses medicalese, her

    effort would serve no purpose, and that is why such non-clarifying reword-ings,though theoretically possible, do not occur. A nurse who is clarifying fora patient is engaged in derived language production, in the sense previouslydiscussed. Other instances of paraphrasing are parodies of the original, or insome other way fail to aim at meaning preservation. Of course interlingualwriters and speakers may clarify or parody as well, but this is not essential: atranslation ofaFrench medical text into an English medical text, both equallydifficult for the lay reader, serves a very clear purpose.

    I want to emphasize that I am making a theoretical point here, not a

    practical prescription. I am not saying that it is desirable for people calledtranslators to have an imitative goal; indeed this may be highly undesirable insome circumstances. Nor am I making any judgment about which activitiescan rightly have the English t-word 'translate' applied to them. That is a socialissue, of no concern here. In this article, I am concerned solely to identify aviable object ofstudy,and that object, I am specifying, is amongst other thingsa rewording process that has an imitative purpose.

    Having said that intralingual rewording is not generally imitative, I nowcome to the exceptions. Our nurse could of course repeat, in medicalese butwith different wording, what the doctor had said in order to check whether she

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    21/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 251

    had heard the doctor correctly. Or she could repeat her own lay-languagestatement to the patient if she thought she had not been heard, or wanted toemphasize a point. Repeating is certainly imitative, but it is not quoting. Ourrepeating nurse is not purporting to demonstrate anything about the doctor'sdiscourse. Her repeating work is thus not translational; she is simply expressing a thought more than once.

    A second exception is quoting during oral narration, as in example (3)above. Such non-fictive quoting does indeed often aim to imitate the original.However the intralingual quoter cannot process the source in sequential

    chunks, for the simple reason that the work is almost always done from long-term memory of the conversation being quoted. Typically, the gap betweensource and rewording is hours or days, so that capturing the detailed meaningin anything like its original compositional form is not a feasible goal. InTranslating, the gap is a few seconds in simultaneous oral or signed interpretation, somewhat longer with dialogue interpreting, up to a minute or so withconsecutive (and memory is here aided by note-taking). A much higher levelof imitation can thus be targetted than in intralingual quoting,8 though ofcourse never as high as when the source text is recorded.

    5. The Interlinguality of Translating

    Whenever language production conforms to all three features already described (free direct quoting, sequential chunk processing, imitative purpose),it will, I think, prove to be interlingual. No form of intralingual language

    production meets all three criteria. For example, editing is sequential, but notimitative. Quoting during narration is often imitative, but not sequential.The

    work of a conference rapporteur can be imitative and (indirectly, via notes) itcan be sequential, but it is not direct quoting (first-person quotes appear onlyincidentally).

    Somewhat surprisingly, then, interlinguality is not a defining feature ofTranslating; its inclusion would be redundant. Nevertheless there are a num

    ber of matters related to interlinguality that bear on definitional issues.First, one of our defining features, 'imitation', is closely tied to the

    interlinguality of Translating. As I argue elsewhere (Mossop, unpublished),the achievable degree of imitation when rewording is much greater if a second

    language is used. Consider synonym sets: one word will be more technical,

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    22/36

    252 BRIAN MOSSOP

    another more formal, a third more current, a fourth more literary, and so on. Ifan editor replaces a word with a synonym, there will be a small change inmeaning. But in a second language, it will very often be possible to find aword at the same level of technicality, formality, etc.

    Second, interlinguality means that two different linguas are involved, alingua being some person's native language. A standard language is a lingua,

    but so is a geographical, temporal or social-class dialect. If two linguas aresufficiently different that speakers find a need for linguistic assistance, thenthe assistant's activity is interlingual, and if it satisfies the three defining

    criteria translational. Thus creating subtitles in standard French for a filmspoken in the dialect of the Gasp region of Quebec counts as Translating.Perhaps a distinction might be made between full lexico-syntactic Translating

    where the problem lies in the completely different vocabulary and syntaxof SL and TL and the marginal Translating which occurs when thetranslator's work is just a matter of pronouncing the text in a standard dialector substituting a few items of local vocabulary (e.g. British versus AmericanEnglish).

    Rephrasing between registers of a language (e.g. from medical to lay

    language) is not interlingual because medicalese is not a lingua: no one speaksEnglish medicalese as a native language. It is a sublingua a selection fromthe lexico-syntactic resources of English. Clarification of sublingua talk orwriting has mainly to do with ideas, not vocabulary or grammar per se. If youcannot understand a Russian paper in advanced particle physics because younever got past high school science, a translator cannot help you at least notinsofar as his or her activity is limited to Translating. What you need instead isa science popularizer who reads Russian. Science popularizers fall into a classof mediators (dispute mediators, cultural mediators) who are not in the first

    instance dealing with linguistic problems. The need for Translating, on thecontrary, arises from the fact that some people do not know the lexico-syntaxused by other text producers.

    Third, the interlinguality of Translating distinguishes it from transcribing,a further type of language production which we can add to Translating,deriving and re-expressing. Examples of transcribing are fingerspelling, transliterating, reading aloud, turning oral tradition into written literature, and

    preparing a written transcript of recorded oral proceedings. Machine translation, though apparently interlingual, is perhaps better understood as an ex

    tremely elaborate form of transcribing. A transcriber reproduces a text in a

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    23/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 253

    different medium or a different notation, rather than quoting it in anotherlanguage. Transcribing is very often sequential, as when a news announcerreads a script aloud, but it is not free direct quoting: reading aloud may becreative, but the source script is not reworded, and there is no demonstrating.

    Fourth and finally, if Translating is interlingual, then it is lingual.Linguality is a defining feature of Translating when it is regarded semiotically,as sign production. In this paper, I define Translating in relation to other sortsof language production, but it could also be related to higher-level semioticacts. Thus Clark and Gerrig's theory of (linguistic) quoting is that quoting is

    demonstrating, which is not necessarily lingual; I can demonstrate your dancing style as well as the propositional content of your speech. As definingfeatures of Translating, quoting and imitation are to be understood as lingual,

    but one might also want to speak of architectural or musical quoting andimitation. As for intersemiotic sign production, some instances can perhaps beseen as deriving: Marleen Gorris' recent film Mrs Dalloway is derived fromVirginia Woolf s novel of that name. Deriving, unlike Translating, is notnecessarily lingual.

    6. The Duality of Translational Quoting

    Up to this point, I have engaged in a defining exercise, picking out just thosefeatures needed to distinguish Translating from all other types of language

    production. I have said that Translating is SIQ sequential imitative quoting.In this section,9 I want to show what can be done with this definition, bymaking a concrete proposal about SIQ.

    Like all reporting, free direct quoting is an inherently Janus-like concept.

    It is always both source- and target-oriented. One makes a report OF something and one makes itTO someone; both aspects must be present. I therefore

    propose that, if Translating is reporting, then a Translating translator is engaging not in a single sort of language work but in two. First there is the reportingOF the source text;10 let us call this work 'rendering'. Second there is thereporting TO an audience; let us call this 'pragmatic' work, of which there areseveral varieties. The result is set out in Figure 3, an expansion of Figure 1.

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    24/36

    254 BRIAN MOSSOP

    1 semantic rendering2 lexico-syntactic rendering

    3 phonetic rendering

    1EDITING (poorly written source text)2RETARGETING(fordifferent useorreader

    type)3ADAPTING (to ideology/culture/rhetoric

    of receiving society)4ADJUSTING(formultimedia integration)5SUMMARIZING(e.g. dialogueinsubtitles)6EXPLAINING(e.g. unrenderable expres

    sions)7COMMENTING(metatextual & metalinguistic notes)

    8CO-ORDINATING(the participantsindialogue interpreting)

    9COPY+ PASTING(originalTL material)10PICTURING(working fromadiagramor

    actual object, rather than froman SLdescriptionofit)

    ETC.etc. etc...

    Figure 3. Rendering Work and Pragmatic Work

    The list of pragmatic activities on the right-hand side of thediagram isillustrative only; further consideration might lead to combining, splitting,adding, subtractingorrelabelling items. The term 'pragmatic' isperhapsnotentirely satisfactory since some might takeit toinclude only those activitieswhichaim at conformance with target-culture norms, publishers' demands,space limitationsand soon.Nosuch restrictionisintended: pragmatic work

    can be motivated by the translator's own writing project.

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    25/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 255

    Figure 3 is not to be given a temporal interpretation; that is, pragmaticwork is not to be understood as following rendering. Nor is Figure 3 adisguised literal-to-free scale: pragmatic is not a new term for free, andrendering is not a new term for literal translating. It could be said that a literaltranslation is one in which relatively little pragmatic work has been actualized(i.e. little of it is manifest in the product); a free translation is one in whichrelatively more has been actualized; and a non-translation is one in whichrelatively little rendering has been actualized. But the purpose of Figure 3 isnot to classify texts; its purpose is to enumerate kinds of language production

    work. With this approach, it becomes possible to maintain a unitary definitionof Translating, while allowing for an extremely broad range of final products.

    The next two sections give a very brief account of rendering and ofpragmatic work, passing over many problems and issues which I hope to dealwith in future publications.

    6.1.Rendering

    Rendering is the work of representing the source text in another language.

    Rendering is not representing presenting the text again to new readers inanother language but ratherrepresenting, depicting the source text withoutregard to future users or uses.It is through rendering that the imitative goal ofTranslating is realized.

    What is rendered depends on what the translational imitator decides todemonstrate (in the sense of Section 2.2). We can distinguish three broadcategories: imitation of meaning, of lexico-syntactic form and of phonetics. Ifan intralingual rendering were of any theoretical interest, it would be very easyto construct. A complete English rendering (of meaning, form and sound) for

    hickory dickory dock, the mouse ran up the clock would be hickory, dickorydock, the mouse ran up the clock. The interlingual case, on the contrary, isextremely complex.

    The imitated meaning is whatever meaning is attributed to the source textby the translator. Thus the semantic rendering may contain errors; it mayconcentrate on one type of meaning (e.g. propositional) and ignore others(expressive); it may be deep (taking maximal account of co-text and context)or it may be superficial.

    What notation can be used to display semantic rendering? Consider the

    following passage from an advertisement in French and the two lines of

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    26/36

    256 BRIAN MOSSOP

    English that follow (neither of which would be very useful for an actualadvertising campaign, but that is a pragmatic matter):

    (12) C'est parce qu' il rsiste au temps qu' il fait qu'il rsiste au temps quipasse.11

    (R1) It 's durable because it 's weatherproof.(R2) Withstands the ravages of weather and thus the ravages of

    time.

    R1 captures the propositional content of (12) but misses its expressive content;

    R2 captures some of the expressive content but the propositional content is notquite right: the word ravage is too strong, and there are no ravages of time assuch, just ravages of weather extending over time. Translators may sacrificesome aspect of meaning because the commissioner is not interested in it, or

    because it cannot be expressed at the same time as some other aspect. But suchdecisions are pragmatic matters. When rendering, no such selection is made.This is most easily conceived in graphic terms, as (Rl) and (R2) written oneon top of the other,perhaps in different colours and fonts so that both are stillvisible, in the manner of certain kinds of wall art a sort of palimpsest

    without erasure.Semantic rendering is akin to the process of stating in L2 the meaning ofan L1 expression, as discussed in Section 2.1. I have written the renderings(Rl) and (R2) in English, but if the translator had used another language, orlanguage combination, to think about the meaning of (12), the semanticrendering would be in that language perhaps Hungarian if the translator of(12) into English were a native speaker of Hungarian. Lexico-syntactic and

    phonetic renderings, to which I now turn, are of course always written in thetarget language.

    (Rl) and (R2) tell us virtually nothing about the vocabulary and syntacticstructure of the source text. When recorded texts are being translated, thewording choices of the ST author are present to the translator's mind andclearly influence the production process. Every rendering must thereforeinclude a representation of the lexico-syntactic structure of the source text, butit is only in close written translation and "bad literal translation" that muchevidence of this rendering work remains in the final output. If the translator'slexico-syntactic analysis is erroneous, that will be reflected in the rendering.

    A lexico-syntactic rendering could be given in either one or two lines.

    When TL and SL are structurally similar languages, it will often be possible to

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    27/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 257

    represent both vocabulary and syntax together:(R3) It 's because it resists the weather that is occurring that it resists the

    time that is passing.

    But with some sentences and probably with all sentences when SL and TLare structurally dissimilar this will not be feasible. To represent the syntax,TL lexical items will have to be slotted into SL syntactic structures, yieldingrenderings like the mock-German If you will wait until I my clothes put on, Iwill with your luggage help (Sternberg 1981: 229).

    Collocations of lexical items need not be rendered idiomatically; hencethe somewhat odd resist time in (R3). This aspect of lexical rendering together with the representation of SL syntax provides an account of the

    peculiar translational version of the target language which is found in many(all?) translations and perhaps arises from the sequential processing thatcharacterizes translation production.

    Some SL features such as lexico-syntactic manifestations of illiteracywill not be renderable at all. The same will be true of words referring to SLculture items, and puns: in (12) the pun on temps is unrenderable since no

    English word means both 'time' and 'weather'. Translators must use pragmatic techniques to deal with unrenderable words and structures.

    Turning to phonetics: in film dubbing, when the actor on screen isuttering an SL word containing a bilabial (p, b or m), the dubbing actor ideallytries to pronounce a bilabial at the same moment, because viewers may bedisturbed if the closing of the screen actor's lips is not matched by anappropriate sound (Whitman-Linsen 1992: 20ff). Thus, unless the SL or TL isa signed language, the rendering will need to include a phonetic imitation ofthe source text, in the manner of the Zhukovskys' famous version of Catullus,

    or an actor pronouncing her lines with a foreign accent. Of course, in mosttranslation work, little or none of this will be actualized in the final output.

    Using letters as a notation system for sounds, we could phoneticallyrender (12) as:

    (R4) zipper ski rays east tote on keel fey . . .

    An additional phonetics line may be needed to render sentence rhythm. Notethat, just as R1 imitates meaning, so R4 imitates sound; it does not substitutesounds thought to evoke ameaningequivalent to the source. If a translator of

    written poetry substitutes phonetic effects from TL poetics, that is pragmaticwork.

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    28/36

    258 BRIAN MOSSOP

    It should now be clear that the different lines of a rendering are not to beconceived as alternative translations. Rather they are imitative demonstrationsof various aspects of the source text. The lexico-syntactic lines of a renderingcome closest to one sense of "literal" translation: TL items selected with littleregard for TL idiom or rhetoric. But neither these lines, nor all the lines takentogether, constitute a translation; translation results when rendering is com

    bined with pragmatic work, as described in the next section.

    6.2.Pragmatic Work

    It is through their pragmatic work that translators, working on behalf ofinstitutions which have commissioned them, re-appropriate the source text fortheir readers or listeners. They make changes needed to extend its spirit intothe target culture, given the translation's intended uses and users. Peoplecalled translators may also, to use Barbara Folkart's term (1991: 407ff), seekto confiscate the source for unrelated or even contrary purposes. In this case,their work is non-translational because the purpose is not imitative.

    What is the status of pragmatic work? I have distinguished Translating

    from non-translational language production acts like re-expressing and deriving; and I have said that Translating is a dual activity, its twin aspects beingrendering work and pragmatic work. Now while rendering is unique to Translating, pragmatic work is not. EDITING and SUMMARIZING, for example, canobviously occur during non-translational language production. Sometimes

    pragmatic work can be seen as overriding rendering, but sometimes it simplyreplaces it; that is, the language producer stops Translating and does something else.

    There is an interesting question whether some types of overriding prag

    matic work tend to occur quasi-automatically during translation. For example,Sguinot thinks that some of the EDITINGwork done by translators, improvingthe logic of the source text, is an inherent part of translation production:translators, she says, have to make a special effort to avoid making suchimprovements (1989: 39).

    Let's look at some other types of pragmatic work. First, COPY+PASTING.Copying, unlike quoting and repeating, is not a communicative act. Exampleswould be an actor memorizing lines by saying them over and over, or a scholarwriting down passages from a book. A copy, once made, can (to use a

    metaphor from computer word-processing) be "pasted" into a communicative

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    29/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 259

    act: theactor speaks thememorized lines during aperformance; thescholarinserts a copied passage into an article. Consider the title of the Scott-Moncrieff translationofProust's mega-novelAlarecherche du temps perdu.Terence Kilmartin, in his recent revision ofScott-Moncrieff, translatesthetitleasinsearch oflost time, but Scott-Moncrieff himself did not translateatall;instead he pasted in a copy of part of the second line ofShakespeare's 30thsonnet: "[when to the sessionsofsweet silent thought /1 summon up] remembrance ofthings past ".

    The pragmatic workof PICTURINGiscommonplace in technical transla

    tion. The translator comes acrossacomplex or poorly written descriptionofatechnical object, stops Translating, and simply describes the object, workingfromadiagram or photograph,orfrom the object itself.

    In film dubbing, EXPLAININGand ADJUSTING are often necessary.Acharacter with face turned away from thecamera, who is saying nothing in thesource language, maybemade to explain verbally something that was conveyed by a certain hand gesture which TL viewers would notunderstand(Whitman-Linsen 1992: 35).What a character says may also have to beadjusted toachievea fitbetween language and gesture (ibid: 33ff).

    Cecilia Wadensjo has pointed out thatadialogue interpreter is notonlyrelaying content but also COORDINATINGthe people engaged intwo-languagedialogue, forexampleby smoothing over a comment which theother partymay find insulting (1995: 308-309). Susan Berk-Seligson (1990: 70) gives thefollowing courtroom example,inwhich the final line replaces rendering witha COORDINATINGmetalinguistic COMMENT addressed to the prosecutor:

    PROSECUTING ATTORNEY: Had you expected that you would have to pay

    another eight hundred dollars?

    INTERPRETER:Esperaba Ud. tener que pagar ochocientos dlares adicionales,seora?[. .. eight hundred additional dollars...?]

    WITNESS:Adicionales, cmo? [additional, what do you mean?]

    INTERPRETER:. . .she doesn't understand the word 'additional' in Spanish thatI used.

    The effects of pragmatic writing will permeate the wholeofa translation,simply becauseit isbeing writtenfor aTL context. The resultsofrenderingnever appearin pure form, except perhaps with international technical termi

    nology. Consider this passage from the transcriptof acourt hearing:

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    30/36

    260 BRIAN MOSSOP

    Alors compte tenu Votre Seigneurie que, l'individu videmment, je ne parlepas d'un individu qui ne serait pas ses premiers antcdents en semblablematire, et qui ce moment-l, comme mon confrre vous l'a mentionn dansle cas de P..., que vraiment l'individu a tu dlibrment Votre Seigneurie desang-froid, je pense que l il y a une marge, ce n'est vraiment pas le cas.

    Here is a linguistic gloss of this passage; a gloss is a marginal kind oftranslation that manifests superficial semantic rendering and minimal overriding of lexical-syntactic rendering:

    So given Your Lordship that, the individual obviously, I don't speak of anindividual who is not at his first record in such a matter, and who at that point,as my colleague has mentioned to you in the case of P...., that really theindividual killed deliberately Your Lordship in cold blood, I think that herethere's a difference, it's not really the case.

    Here is the translation that was actually produced:

    I am not speaking, Your Lordship, about an individual who has done suchthings before and, as in the P... case mentioned to you by my colleague, killeddeliberately in cold blood. This is really a different situation.

    This translation manifests, massively, the effects of SUMMARIZING, EDITING

    and RETARGETING (the latter because spontaneous oral style was deemedinappropriate for a written translation to be used for information). Yet as thetranslator, I was indeed working from the detailed wording of the source text.Rendering is not immediately observable here, but its imitative effects shinethrough, so to speak, in the translation I produced.

    6.3.The Primacy ofRendering

    Translating is always a combination of rendering and pragmatic work. The

    pragmatic work is ofgreatest practical interest. Without it, translations wouldnot be receivable by L2 readers and listeners. It is the site of translationalcreativity, where the source text is fitted to the L2 environment, and wherenew equivalences are invented when a passage is unrenderable. But it isthe rendering work which is of the greatest theoretical interest, because theimitative goal of rendering is what makes Translating such a distinctivelinguistic act. Almost all language is reader- or listener-oriented; Translatingis peculiar because it includes, centrally, language production work which isnotso oriented. Where else do we find such an attempt to imitate a text in other

    words?12

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    31/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 261

    7. Philosophical Conclusion

    I have proposed that sequential imitative quoting (SIQ) be taken as an objectof study. Anyone engaged in SIQ is, by definition, Translating. Many of theactivities described by everyday t-words like 'translate', 'traduire' and 'adapt'involve SIQ to one degree or another.

    In essence I am saying: let us carve out of the world all those instances oflanguage production which satisfy the three criteria. But why pick out SIQ?Two reasons. The first is subjective: it strikes me that picking out SIQ will be

    conducive to progress in understanding translation, because it combines narrowness with generality. It's hard to make progress if you take too broad achunk of the world as your object. The combination of imitative goal andsequential processing of the detailed wording of the source text does seem tomake Translating stand out from other language activities and thus make it agood potential object for investigation.

    My second reason for taking SIQ as object is philosophical. I think thepurpose of scientific investigation is to discover the underlying mechanismsthat generate the actual and experienced world the mechanisms that consti

    tute Reality (Bhaskar 1978: 12-20; 244-246). I have picked out SIQ because Ibelieve it is a real distinct activity of the mind. When communication is underway and the mind is engaged in SIQ, an observable resulting text will havecertain characteristics. That text may of course have other characteristics aswell, arising from other activities like re-expressing stored thoughts which are under way at the same time.

    Translating is thus an underlying generative mechanism which provides apartial account of the observable output of certain language producers. At theobservational level, we find a wide range of outputs on the literal-to-free scale,

    and translation simply shades off into non-translation, depending on the extentof non-translational work and of pragmatic activity replacing rendering. But

    by looking at the mechanisms at work at the moment of production, a unitarydefinition of translation is achieved a feat that would be impossible if theobject of study were taken to be the whole multifarious range of observableoutputs in interlingual situations.

    Translating has here been defined without reference to social functions orsocial labels. The functions of translation its embedding in larger socialframeworks do not define it; rather they explain particular choices made by

    translators on specific occasions. What linguistic label is placed on a language

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    32/36

    262 BRIAN MOSSOP

    production activity by the producer or by society is also irrelevant. The objectof theory is not a socially defined corpus of observed texts: there is no reasonto think that any such corpus would be the result of some single language

    production activity.The proposed definition succeeds in distinguishing Translating from a

    wide variety of language production activities, both interlingual andintralingual: deriving, co-producing, re-expressing ideas, fictive quoting,quoting the thoughts of others, mental language conversion when using one'ssecond language, copying, repeating, transcribing, paraphrasing, ghostwrit

    ing. In addition, a number of activities have been described as marginallytranslational: interlingual indirect reporting; pedagogical translation duringsecond-language learning; translation for addressees already familiar with thesource text; TL glossing of a text to help people who have little or no SLknowledge, and phonetic translation from non-standard to standard pronunciation. The notion of marginality implies no value judgment, any more thandoes the assertion that a platypus, being a layer of eggs, is a marginal mammal.Whatever definition of Translating we choose, there will be central andmarginal instances. The central/marginal distinction is in my view philosophi

    cally preferable to the widespread notion of a cline, though there is no space toargue this here.

    There is of course much to be said about translation that is not covered bytranslation production theory. Aside from the concerns of reception theory,there are social issues such as the extent to which translation should be usedrather than co-production or language learning for inter-group communication. Furthermore the facts of translation can doubtless shed light on humanmental functioning, on the potential for interpersonal understanding, on relations between dominant and dominated societies, and on the development of

    languages and literatures.This paper has not been concerned with shedding light on such matters, or

    seeing how translation might help or hinder the solution of social problems ina world marked by increasing human displacement and friction among groups.It has rather been concerned with what might be called the Saussurean

    problem of seeing translation en elle-mme et pour elle-mme (in and foritself).

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    33/36

    WHA T IS A TRANSLATIN G TRANSLAT OR DOING? 2 6 3

    Author's address:Brian Mossop 14 Monteith Street TORONTO, Ontario M4Y 1K7 Canada;e-mail: [email protected]

    Notes

    1. Toury (1995: 31) fears that any deductive approach to theory one that starts from adefinition will be normative. He proposes instead to define not translation itself but a

    body of texts to be used as a methodological starting point for descriptive work, andultimately for determining laws of translational behaviour. This body of texts consists ofall those which are received as 'translational' in the target culture. A text is translationalif people in a culture assume there is a source text in another language from which it wasderived and with which it shares some features. In this article, I am not delimiting a classof texts for study, but identifying non-normatively I think the features of a writingor speaking act which make that act translational.

    2. A few theorists have attempted to relate translations to other metatexts (van Gorp 1978:106; Popovic 1976: 232). Such proposals are not directly relevant to this present article,where the concern is not to classify completed metatexts, but to enumerate the definingfeatures of the production process leading to them.

    3. Since the diagram has received some attention in the literature (Folkart 1991), I shouldmention that the idea of using cartoon bubbles (rather than embedded rectangles) is due toKen Popert, who is listed in the acknowledgments of Mossop 1983 but not specificallycredited with the diagram.

    4. Pym (1995: 19ff) says that translators are by definition in an in-between location, but thelocation he describes is more mental than geographical. I have tried to capture an aspectof the translator's mental in-betweenness in Section 6. Geographically, some translatorsare in an in-between zone (a French-to-English translator working in Montreal, a citywith two large language cultures) and some are not (a French-to-English translatorworking in Boston).

    5. Pym (1992a: 54-56) argues that translation should be understood not as reported discourse:X to C: A said to B "t"

    but asX to C: s translates as t

    Pym is looking at things from the reader's point of view. His schema answers the reader'squestion "what is this text t?" (it's a translation of s), whereas mine is answering thetranslator's question "what am I doing as I translate?" (I'm telling C what A said to B).

    Nord too takes a reception point of view when, referring to Mossop 1983, she says thatonly what she calls documentary translations are reports (1988: 12; 284 fn. 6). What shecalls instrumental translations are not reports, in her approach, because readers do notexperience them as being related to their sources.

    mailto:[email protected]:[email protected]
  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    34/36

    BRIAN MOSSOP

    Gutt (1991: 45-65) similarly suggests using a theory about a more general activity,namely the use of language to interpret other language. He claims that an approach tolinguistic pragmatics known as relevance theory can explain all interpretive uses oflanguage, including translation. I too am invoking a more general language activity,namely quoting, but I am not claiming that a theory of quoting in itself accounts fortranslation. Quoting is just one of three proposed defining features, and the actual work ofexplaining what happens during Translating does not begin until the theoretical objecthas been defined.

    The general process of resaying something in a new context was central to the work ofMikhail Bakhtin, from whom I originally derived my idea that Translating is reporting(cf. Voloshinov 1973 an extensive discussion of reported discourse in novels, thoughwith no mention of translation). This was part of Bakhtin's larger, interest in what he

    called the dialogic nature of language: every text is to be seen as a mosaic of quotations inthe sense that it contains echoes of and responses to things previously read or heard. Thequotations, however, are not imitative. Foreshadowing skopos theory, Bakhtin saw themas subject to "re-accentuation" to serve some purpose in the new context. Furthermore, hemade no distinction between the re-expression of thoughts drawn from a stock of priorreadings and the re-wording of one particular text. Ultimately, in this view, writing isrewriting; even a direct description of reality is understood as a reaccentuated collage ofold descriptions. Bakhtin's approach is far too broad for the purposes of translation

    production theory.

    Folkart (1991: 263ff) discusses inter- vs. intralingual quoting at length, and concludesthat translations are asui generistype of reported discourse; they are not direct quotations

    (which she equates with exact-words quotations). Drawing on the approach of Danishlinguist Louis Hjelmslev, Folkart defines translation as the reporting of discursive form(1991: 257ff). She distinguishes technical-administrative texts, in which the discursiveform is mainly the propositional content (impressionistically: the thought behind thewords), from literary texts, in which the discursive form is mainly the meaning connoted

    by the author's idiosyncratic selection of language forms (the thought in the words). Interms of Clark and Gerrig's quotation theory, these are two features of the source textwhich can be demonstrated.

    Section 6 is based on a paper read at the First Congress of the European Society forTranslation Studies, Prague, September 1995.

    In a review of Mossop 1983, Witte (1986) suggests that my concept of Translating asreporting is similar to Vermeer's notion that every translation "bietet einen Infor-mationsangebot iiber einen A-Text" (1982: 99); that is, it offers information about asource text. However Vermeer's development of the idea has been target-oriented: thenature of the information the translator offers is determined not by the source text but bythe purpose (skopos) of the translational activity. Witte rightly notes that I do notemphasize the skopos. As will be seen, I give theoretical pride of place to the translator'srendering work.

    Cruse (1992), from whom I have taken sentence (12), attempts to define prototypicaltranslation, that is, to identify features the possession of which will make a pair of texts a

    better representative of the category 'X is a translation of Y ' than a pair lacking them. For

    this purpose, he uses three concepts he calls equivalence (of meaning), congruence (ofstructure) and engagement (translating pertinent aspects of meaning) to define what he

  • 7/24/2019 What is a Translating Translator Doing? by Brian Mossop

    35/36

    WHAT IS A TRANSLATING TRANSLATOR DOING? 265

    calls maximal translation, minimal translation and plain translation. All of these hecontrasts with optimal translation translation which "maximizes not degree of correspondence with the original, but fitness for a particular purpose". Such fitness is achievedthrough what I have called pragmatic work, whereas correspondence is achieved throughrendering.

    12. Schreiber(1993:125) suggests something similar when he distinguishes translation fromadaptation by saying that the former is invariance-oriented (except for the languagechange) while the latter is variance-oriented (in my terms: it is centrally a matter of

    pragmatic and non-translational work). Unfortunately his approach is needlessly complex because he tries to distinguish texts (translations and adaptations) rather than acts(translating and adapting). A text, whether labe


Recommended