Chapter 1
Two Dogmas of Translation
âHow am I able to obey a rule?â â if this is not a question about causes [nach der Ursachen], then it is about the justification for my following the rule in the way I do. If I have exhausted the justifications I have reached bedrock, and my spade is turned [so bin ich nun auf dem harten Felsen angelangt, und mein Spaten biegt sich zurĂŒck]. Then I am inclined to say: âThis is simply what I doâ.
(Remember that we sometimes demand definitions for the sake not of their content, but of their form. Our requirement is an architectural one; the definition a kind of ornamental coping that supports nothing.) (Wittgenstein, Philosophical Investigations I. 217)
When I was a child in Madrid, my friends and I took our literary pleasures where we could find them, and we found them very often in translation.1 RenĂ© Goscinnyâs and Albert Uderzoâs great âAstĂ©rix et ObĂ©lixâ series came to us in Castilian furnished with occasional footnotes: âJuego de palabras intraducibleâ, it would say at the bottom of the page: âuntranslatable word-gameâ or âuntranslatable punâ, referring to a word spoken by a Gaul, Roman, Belgian and so on. At the time, we understood these footnotes to be deliberate characterizations of the Spanish words we had before us, rather than of a French original â and I suspect, in fact, that it never crossed our mind that there might be a French original, no doubt because the notorious national rivalry between Spain and France made the possibility unthinkable. For us, it was the Spanish that was âintraducibleâ, the language into which the comic could not be translated remaining unspecified; it struck us as particularly hilarious that the Spanish words footnoted for us were commonplace and should thus be utterly obvious in their translation, an indictment of that other, misty language as much as it was proof of Castilianâs nuance and particularity. Years later, when I read the works in French and in English, I realized to my shock that these footnotes were not deliberate parabases but marks of melancholia or exasperation, a sort of throwing-up-of-the hands left diacritically by a nameless translator who was unable if not unwilling to render into Spanish the French jeux de mots.2 âJuego de palabras intraducibleâ, he or she wrote when the going got sticky, closing off to Spanish readers of the books the avenue of the linguistically unfaithful but culturally correlative pun â something on the order of turning the name of ObĂ©lixâs minute dog âIdĂ©fixâ into âDogmatixâ, as the brilliant English rendering does, rather than conveying it, as did the tired Spaniard I imagine, merely as âIdefixâ.3
One does not go unscarred through experiences like this. A host of retrospective reorderings, retranslations and recathectings of primal relations can ensue when the ideas one formed in childhood about this or that important expression become unfixed â with often perilous consequences for oneâs professional or psychic health. And my choice of topic, translationâs âindecisive museâ, must surely count as one of the long psychic after-effects of this early shock, the discovery that my fixed, dogmatic ideas about the translatability or untranslatability of names like âIdĂ©fixâ or of referential expressions like âJuego de palabras intraducibleâ were not merely marks of linguistic ignorance and chauvinism but incorrect to boot. Imagine again my notional, exasperated Spanish translator. The confession of a termâs untranslatability speaks to his or her exhaustion, or to a literalistic turn of mind bound to conflict with the subtle extravagance of Goscinnyâs script (one could reasonably argue, for instance, that âIdefixâ is not a bad, but rather a particularly good translation of âIdĂ©fixâ, which like all proper names is untranslatable as a description of what it names) â unless, of course, the hilarity that these footnotes and bad juegos de palabras provoked among one group of pre-adolescents itself served as a sly correlative for the humour of Uderzoâs and Goscinnyâs jeux de mots.4 On this improbably Straussian reading the marks of the translatorâs seeming failure themselves succeeded in conveying some of the âuntranslatableâ wit of the original, as unfaithfully but just as companionably as the more conventionally successful Englishing does. Improbable â but how would one know? On what grounds would one make such a decision? What might be at stake in deciding one way or another whether the imagined, unimaginative Spanish translator did or did not mask an acute, dialectically inclined linguistic druid?
Something more, surely, than the comparative linguistic innocence of one young group of Spanish readers. Say one wanted to argue that the Spanish translation of Goscinnyâs and Uderzoâs bandes dessinĂ©es was poor (because flatfooted or unimaginative, and hence inadequate to the original) or that it was in a way too good (charactersâ proper names were left rigidly untranslated, as if their descriptive function in the original had been shorn at the border), but that it did a certain ideological work (the work of soliciting adolescent assent that Spanish was indeed âdifferentâ, in being untranslatable) with pernicious consequences (to promote Spanish exceptionalism at a time when Spainâs exceptionality with respect to Europe was largely political, was to bolster that âexceptionalâ political model). A reader self-involved enough to examine the source and possible consequence of his literary pleasures might well find himself or herself making this sort of observation; any translator contemplating different ways of rendering this or that juego de palabras, or indeed contemplating any translation whatsoever, might just as well. In both cases, the mechanisms, habits, discursive contexts or rules that we, readers and translators, employ when we make judgements concerning the effect or the felicity of a translation take into account the historical circumstances of a translationâs production and reception: we need history to tell us whether this or that translation âworksâ, and how, and to what ends, to help us to evaluate it. Clio watches over the translatorâs task and it is to that muse that the translatorâs interpreters turn as well.
And yet this banal observation is beset with problems. Consider. Goscinnyâs and Uderzoâs little Gaulish village, under siege by Roman forces, is devised and first published in the shadow of sieges that France felt much more proximately indeed â the Battle of Algiers, whose double siege â the French population in its quarter, the FLN and its supporters largely confined to the casbah and older parts of the city â so firmly gripped the French imaginary in the years from 1954 to 1960; and the terrible siege of the French expeditionary forces in the valley of Dien Bien Phu, in the course of 1954, which led to Franceâs withdrawal from Indochina. Consider, too, that the besieged but invincible Gauls may well have expressed a compensatory French nostalgia for the movable stockades of âresisteursâ who fought against Vichy and against German, rather than Roman, occupation â with considerably less success than their cartoon counterparts. (RenĂ© Goscinnyâs family were Polish Jews. Born in France, he lived through the war in Buenos Aires and the United States; all his European relatives died in the camps.) Note also that Goscinnyâs and Uderzoâs Spanish translator worked still, in 1970, under the careful eyes of the last of Europeâs fascisms, a regime for which the figure of the siege served quite different purposes: recall most notably the official and widespread assimilation of the Roman siege of the Spanish town of Numantia, in the second century BCE, to various Republican sieges of Nationalist fortifications during the 1936â1939 Spanish Civil War and, from the Republican side, the similar characterization of the sieges of Madrid and Barcelona by Francoâs forces and of the entire exiled population as a âNumancia erranteâ, in the marvellously evocative words of the socialist leader Luis AraquistĂĄin. To succeed in making Francoâs Spanish truly âdifferentâ from republican French (or from any other language) â to make the languages mutually untranslatable â and to make a new generation of Spanish readers enjoy their difference was to raise and level Pyrenees of the mind between Spain and Europe, between the little village or fortified alcĂĄzar of Spanish falangismo and a liberalizing continent intent on invading Spain every August. âSpain is differentâ, ran the campaign that the Spanish Ministerio de InformaciĂłn y Turismo opened in 1963, marketing Spanish tourism in the exceptionalist tropes used since the eighteenth century to marginalize the Peninsula: Spanish, this wily translator seems to have been showing us, is different as well. How might a translation negotiate this peculiar set of overdeterminations? What shape might we expect judgments concerning its effects or felicities to take? What work might we expect the claim of âuntranslatabilityâ to do, at the time and under the circumstances?
Other, less clearly âhistoricalâ determinations are at work as well. For our judgements concerning the effect or the felicity of a translation also depend upon assertoric statements like âIdefix and Dogmatix are translations of IdĂ©fixâ or âIdĂ©fix, Idefix and Dogmatix refer to the same thingâ, statements which are true rigidly and independently of historical circumstance. We would not be able to say that âIdefixâ is a translation of âIdĂ©fixâ, good, bad or indifferent; and with this or that consequence; without knowing that âIdĂ©fixâ and âIdefixâ refer to the same thing. How this circumstance came about, whether it may change, what might be the different descriptive charges of the names âIdĂ©fixâ and âDogmatixâ â these questions are of another order entirely. The notion that a translation should be âfaithful to the originalâ founders just here: âfaithfulnessâ is a predicate with two aspects, a concept with two domains, unfaithful in a way to either one. It is drawn, on the one hand, from a structural or grammatical account of reference, which admits, that indeed requires the possibility of synonymy: assertoric statements like âIdefix and Dogmatix are translations of IdĂ©fixâ can be rephrased as âIdefix is synonymous with IdĂ©fix and Dogmatix is synonymous with IdĂ©fixâ. It is also drawn, on the other hand, from a practical, historical account of reference, one which will always understand synonymy to arise as a matter of happenstance. Here on this side we say: âIdefix and Dogmatix are translations of IdĂ©fix now, under these circumstances, for these reasons, with these consequences: it could have been otherwise, it may not last; what one name describes about its referent coincides now and to some extent, but not necessarily completely, with what the other describes about its referentâ. Each side of this couple is wobbly on its own. For instance, the distinction between a proper name and a description is not always, and perhaps never a priori, a clear one: âIdĂ©fixâ and âDogmatixâ look like proper names that work also as descriptions, but âIdefixâ, the Castilian version of both, only works remotely (if at all) as a description, the Spanish version of idĂ©e fixe being something like âidea fijaâ, two terms that donât readily yield a portmanteau. And the second, practico-historical way of approaching a translationâs âfaithfulnessâ proves as unfaithful on its own to its founding terms and rules as it proves when wed to the structural account of reference that its assertoric shape suggests. For when we judge the felicity and effects of a translation, do we not also need to know, as it were, a history of the translation (and the workâs) effects yet to come? How might these be reckoned? What temporal and imaginative âtranslationâ is required of us in order to inscribe upon the translation that we are evaluating the history of its effects to come?
These scattershot questions, addressed to the form that âdecisionsâ take when we make judgements about translation, flow from the incompatible registers in which the determining notion of a translationâs âfaithfulnessâ operates. Each of these registers furnishes us, too, with answers to the questions that their mutual interference provokes: answers to translationâs indecisions, on the one side, reflecting a roughly âIdealistâ approach to linguistic usage and, on the other, answers reflecting a roughly âmaterialistâ approach to linguistic usage â the two dogmas of my subtitle. My goal in casting Quinean doubts on this tired distinction is not, however, a sceptical one. Rather, I want to argue that grounds for establishing judgements concerning translation today find a compelling weakness in both the surprise of their emergence (they are not determined) and in the dividedness of their conceptual domain (the notion of faithfulness, the notion that the translated word is similar to a proper name â though this name isnât solidly either a designator or a description). The examples I have chosen â taken from the work of Jorge Luis Borges and of Ludwig Wittgenstein â are drawn from terrible years: the period between 1937 and 1940, when both writers observed from abroad the rise, consolidation and aggressions of European fascism, one from Dublin and Cambridge, the other from the basement of the Biblioteca Municipal Miguel CanĂ© in Buenos Aires. Neither could know the outcome of the historical âdarknessâ into which they looked, from different sides of the ocean. For both, an ethical disposition towards that great darkness â der Finsternis dieser Zeit, as Wittgenstein wrote some years later â passed through the problem and the question of translation. Between them â in a dialogue one can only wish had actually taken place, in public, and well attended, somewhere between Buenos Aires and Cambridge â there develops a determining account of the need for articulating the field of ethics â where responses to the crisis of their times could be evaluated â with the problem of translation.5
UNCERTAIN TRANSLATION
A felicitous translation, we might first agree, wearing provisionally an Idealist cap, should be a form of transport, as the characteristic humanist etymology would stress: âdel verbo latino traduco, -is, por llevar de un lugar a otro alguna cosaâ, writes SebastiĂĄn de Covarrubias in a 1611 definition of âtraduzirâ: âfrom the Latin word traduco,- is, meaning to take something from one place to anotherâ.6 The thing remains untouched; topology, rather than interpretation, is at issue; of the translator or the displacement âfrom one place to anotherâ, no trace remains upon the thing or anywhere else. As to what this âthingâ or âalguna cosaâ might be, it plainly need not be a thing at all â at least not in any ordinary sense of the term âthingâ (after all, a wordâs âsenseâ has no material consistency). And of course this âthingâ need not be an âideaâ either, or not at any rate the sort of âideaâ whose numinous form we associate with the various avatars of Platonism, any more than the âplaceâ or âlugarâ to which this âthingâ is moved need be a physical space â though both âthingâ and âplaceâ must minimally be coherent and perspicuously regulated entities. The topology of Idealist translation need not remain quite so plodding â indeed, it cannot. The minimal principle of coherence and perspicuity applies a fortiori to the rules according to which the âtransportâ from linguistic âplaceâ to âplaceâ occurs, rules whose âplaceâ within each language designates the totality of âplacesâ and âthingsâ of another (of every other) language. Because another language (because every other language) also has a place for the rules of its âtransportâ or translation into another language (into every other language), the place occupied by the rules of translation in this or that natural language includes hypothetically both every other language and in and through these others the âprimitiveâ language as well.
As I have described it, the Idealist argument for the minimal coherence and perspicuous regulation of the field of language comes closest to the atomic position most commonly ascribed to Russell â even the paradoxical topology smacks of what will become Russellâs theory of types. Wittgenstein, who comes to be among Russellâs sharpest critics, maintains these principles well through the Tractatus and perhaps up to the Philosophical Investigations. Though never, it should be said, without certain qualms â especially when it comes to translating to natural language principles derived from formal ones. âIf I know the meaning of an English and a synonymous German wordâ, he writes in the âfoursâ of the Tractatus, then âit is impossible for me not to know that they are synonymous, it is impossible for me not to be able to translate them into one anotherâ (4.243) â and this âimpossibilityâ, unmöglichkeit, chimes clearly with the companion stress on necessity and continuity typical of the earliest work.7 But is Wittgenstein right? For me to know (or judge) that two words I know, one English, the other German, are synonymous, I must judge that what I call (in English) âsynony...