Anthony Pym


_Video material

Open letter on hybrids and translation

25-29 June 1996
Christine Schäffner and Beverly Adab
Department of Languages and European Studies
Aston University
Aston Triangle
Birmingham B4 7ET

Dear Christine and Beverly,  

Many thanks for your kind invitation; it's always a great pleasure to participate in this kind of thing. In this case, though, I must confess to some perplexity as to the exact nature of the terms, particularly the bit about 'hybrids'. Do you mean something like Mendelian genetics (you know, with the eye colours that jump generations, or my mother's once-red hair that appears on me  if I let my beard grow)? No, I can't make much sense of that when applied to translation, though God knows just about every other metaphor has had at least one go at our subject. Or perhaps you mean hybrids in the sense of a good old mix, like blue and yellow giving us green? But no, that'd be too easy; it'd simply stir up all those millennial fears about cultures and languages getting lost in each other, producing a murky neutral grey, some kind of bleak imperialised future. Vive la différance! would reply the non-translating Derrideans. No, surely that's a little too facile as well. But if not Mendelian and not colours, then what?  

How about this?  


Now that's gotta be a hybrid! A face on a fish with something like a feather on the side. To say nothing of the funny writing, which I'll explain in a minute. There you have it, ladies: my point of departure, not that the above doubts have been forgotten. Let me also indicate where I want to go from here: You say translations are hybrids, more or less, perhaps like this little picture. Okay. So what happens if I convince you that non-translations are hybrids or at least promoters of hybridisation? Where would that leave us? If translations are X (hybrids, or whatever) and non-translations are also X, then everything is X and we're back to square one. Which in this case may not be a bad place to be (it's hard to start with hybrids; easier to come back to them). Yet I'll take my professional heresy just one step further: What if I convince you that translations are actually agents of dehybridisation, that they work against the producing of things like our little picture? What a cunning bastard I'd be! Mind you, the argument isn't all that recherché: since any translated text marks a line between at last two languages and cultures, it posits the separation and thus the possible purity of both. And if imagined purity is some kind of opposite of hybrids, whatever the latter may be, translations help it rather than hinder it, QED. Far better, I suggest, to do without translations and to see little hybrids everywhere, although we may need something like translations to see them, which is why we can only come back to them. Things are getting complicated. And I haven't even mentioned the more likely hybrids, the human translators, who are quite frankly of far more interest to me than any abstract hypothesis about text genres. But let's explain the picture first.  

Well, it's from a translation. More exactly, it's from manuscript copies of the first Latin translation of the Qur'an, which explains why the funny writing says 'Mahumeth'. The translation itself was carried out in Hispania in 1142-43; the drawing was added later by an anonymous copyist. Derrida would insist the iconic hybrid is a supplement to the translation, though I hesitate to make such references. The drawing was copied again by Marie-Thérèse d'Alverny in her analysis of the Qur'an translation, and again by Jacques Le Goff who stuck it in Les intellectuels au Moyen-Age without telling anyone what the drawing was actually of. That's one reason why I'd like to explain it here.  

The Qur'an translation was carried out along with a whole series of supplementary documents: texts on the nature of Islam, on the life of the prophet, and lists of Islamic heresies, since the translation was supposed to help Christians debate against Islam, not to believe in it. This was the age of the incipient Crusades; the translation comes from the most immediate frontier with the most immediate other of Christendom. The drawing is actually a supplement to one of the supplementary texts: it accompanies the lines where the boss of the project, Petrus Venerabilis (Peter the Venerable, abbot of Cluny), tries to explain why the Qur'an is heretical even though it contains much material from the Christian Bible, which obviously cannot be heretical: there is a lot of the self in the other, so the translated text cannot be condemned outright. His terms are as follows: ' sic, ut ait ille, undique monstruosus humano capiti ceruicem equinam et pennas auium copulat'. Islam isn't wrong in itself; it just breaks the rules of genre, sticking a monstrous human head on a horse's neck (which I struggle to see in the drawing) and adding a few feathers. The result is a hybrid, a monster, precisely of the kind that postcolonial studies often finds when self meets other in a frontier region, when there is anxiety about crossing boundaries: the great risk of translating the Qur'an is obviously that it might attract a Christian to Islam, creating traitors rather than a triumphant identity. Beware, says the drawing, there are monsters out there, like the sea monsters on navigation charts: go no further. But also: Beware, if you venture beyond, you yourself might become a monstrous hybrid, neither here nor there, lost in heresy. Thus does the self produce the imaginary mix so as to affirm its purity, expressing anxiety about the attraction of the other and the quite real possibility of becoming something in-between.  

A postcolonial citation, just to touch a few bases: James Donald analyses 'the grotesque as a boundary phenomenon of hybridization or mixing, in which self and other become enmeshed in an inclusive, heterogeneous, dangerously unstable zone... The point is that the exclusion necessary to the formation of social identity at [one] level is simultaneously a production at the level of the Imaginary, and a production, what is more, of a complex hybrid fantasy emerging out of the very attempt to demarcate boundaries, to unite and purify the social collectivity.' This is from something called 'How English is it?', in New Formations 8 (1988), pp. 36-37, drawing on Stallybrass and White's The Poetics and Politics of Transgression (1986). Sounds pretty good! Not that I've read much of that stuff. I've pinched the citation from Robert Dixon's Writing the Colonial Adventure (1995), parts of which are actually quite a good read, with all sorts of stories about imaginary hybrid tribes, troppo whites and cross-cultural dressing in the wilds of Australia, all cemented into place by abundant references to Homi Bhabha and the rest. Yet the point that really interests me is that the imaginary hybrid - the awareness of possible hybridisation - is produced in a frontier space, as part of an exclusion of the other, as part of a separation of identities.  

Note that postcolonial studies can describe all that without even the slightest reference to translation. Robert Dixon's book, the one I just pinched the citation from, has about 200 pages retelling and analysing Anglo-Australian popular fiction between 1875 and 1914: yards of ripping yarns, tons of hybrids, a fair swathe of references to languages and their mixing, to language learning and language forgetting, but nothing at all on translation. The phenomena of hybridisation would seem not to require translation. If this is so, what reason could we have for putting the two terms together?  

Well, our picture of a hybrid was at least attached to a translation. Yet there is more to it than that.  

Discrepancies exist between the picture and the text that accompanies it: mainly a fish we can see in the drawing but not in the text. So where did the fish come from? In fact, where did the whole idea come from? Any self-respecting translation theorist should have guessed the answer about two pages ago. For the sake of the rest of us, let me cite the beginning of Horace's Ars poetica, which might have something to do with the question of hybrids and translations: 'Humano capiti ceruicem pictor equinam / iungere si velit, et varias inducere plumas / undique collatis membris, ut turpiter atrum / desinat in piscem mulier formosa superne, / spectatum admissi risum teneatis, amici?' Fairclough's translation is fair enough for my purposes: 'If a painter chose to join a human head to the neck of a horse, and to spread feathers of many a hue over limbs picked up now here now there, so that what at the top is a lovely woman ends below in a black and ugly fish, could you, my friends, if favoured with a private view, refrain from laughing?'  

That's it. The anonymous producer of our picture clearly corrected the reference to Horace, adding the fish that Petrus Venerabilis had somehow left out. Yet what interests me even more than the detail is the way the hybrid relates to translation, since Horace has been of some little importance for classical discussions of our subject.  

The Ars poetica is well worth reading and re-reading, from beginning to end. Horace is writing a letter (can translation theory be done in a letter?) to a father and two sons of the Piso family (no one knows much about them) giving advice on how to write poetry, especially drama. His main message, as we see in the opening reference to a hybrid, is that the established (Roman) genres shouldn't be mixed. To do so, wilfully to violate norms by producing hybrids, would simply merit ridicule. Or so say the norms of the rhetorical question.  

The opening reference to a mad painter corresponds to the final image, at the very end of the letter, of a mad poet: 'He, with head upraised, splutters verses and off he strays; then if, like a fowler with his eyes upon blackbirds, he fall into a well or pit, despite his far-reaching cry "Help, O fellow-citizens!" [succurrite... io cives!] not a soul will care to pull him out' (457-460). If you don't observe the rules, you're laughable, necessarily mad, since no fellow citizen will save you when need be. A hybrid, you see, is by definition not like one's fellow citizens. More exactly, by imagining and ridiculing hybrids we can define the fellowship, the affective and iconic relations, the bonds of belonging that ensure mutual help, cooperation, in this case on the level of the civitas, a civilised society. Hybrids, as one-off creatures, may be stronger for a moment, individually gifted or inspired, but they have no fellow citizens in the civilised world. Alone, they do not belong.  

Horace's reactionary defence of established norms is also, of course, a defence against outside interference with those norms. Hence, perhaps, the only piece translation scholars usually care to cite: 'nec verbo verbum curabis reddere fidus / interpres' (133-34), 'do not seek to render word for word as a slavish translator'. We all know that many generations have misinterpreted the line, from Jerome to the early Renaissance theorists studied by Norton right down to André Lefevere and, in her introduction to a 1993 anthology, Siri Nergaard. That is, centuries of citers have believed, as Mildred Larson puts it, 'Horace stated that a faithful translator will not translate word-for-word' (these and similar misinterpretations can be found cited in García Yebra's Traducción: Historia y Teoría of 1994). No, he didn't state that, no matter how much as the Latin of the isolated lines might justify the reading. In context, Horace is telling the would-be dramatist to avoid precisely the word-for-word strategy he attributes to the fidus interpres: the faithful translator works word-for-word; the good Roman poet should not.  

And why not? Horace gives two parallel reasonings. First, the poet should avoid imitating common expression, since slavish imitation is like leaping into an Aesopic well ('nec desilies imitator in artum') 'out of which neither shame nor the laws of your task will keep you from stirring' (134-45). Nor, we might add, will any fellow citizen bother to pull you out. Second, if the poet were merely to repeat something that has been said, the work would hardly make sense in cost-benefit terms: 'parturient montes, nascetur ridiculus mus', 'Mountains will labour, to birth will come a ridiculous mouse!' (139). A nice classical quote: put it in your pocket for use against translation theorists. But here, note well, it is the faithful translator (interpres - a figure between prices, a commercial translator) who is being aligned not just with word-for-wordness but also with falling into a hole (like the mad poet) and producing something ridiculous (like the painter's hybrid).  

Now, Horace does not actually say that faithful translators are bad because their word-for-word strategies bring foreign interference into the Latin language, upsetting the norms and producing linguistic hybrids. Yet I believe his argument could be extended in this direction. In support of my claim, look at a cognate of interpres used just one page prior to the bit about faithful translators: 'post effert animi motus interprete lingua', 'then, with the tongue for interpreter she [nature] proclaims the emotions of the soul' (111). Nature uses the tongue - language - to translate something prior to language. This would be natural translation, if ever there was such a thing. Clearly, not all translating is bad, nor is it all word-for-word (this one is word-for-motus). But which language is involved here? And what happens if the translating goes wrong? Horace explains: 'If the speaker's words sound discordant with his fortunes, the Romans, in boxes and pit alike, will raise a loud guffaw' (113-14). The language of natural translation is the Roman tongue; it is not word-for-word because this language translates directly from a natural order, expressing ideas on which all social classes agree; and if the natural translation goes wrong, if it should upset the transparency of established structure, all good fellow citizens will merely greet it with laughter: it will be a ridiculous hybrid genre.  

This reference to natural translation, you see, acts as a kind of bookmark for what would come later: Horace doesn't talk about any sense-for-sense transfer as such (he only mentions word-for-word), yet this natural translating of motus is the ideology necessary for the sense-for-sense to be found in Cicero, Jerome, and indeed any idealism of deverbalised meaning, at base a romance of transparent language and unchanging social order. The unbearably long tradition of binary translation theory (one method or the other) is already implicit here, even though the question of hybrids is aligned with only one side, the word-for-word.  

At this point you might be worried that I'm a long way from home. I want to argue that translation works against hybridisation. Yet Horace is implying precisely the opposite; he invites us to see translation is an agent of hybridisation. I insist, though, that this and perhaps all binary translation theory is fundamentally reactionary. Translation is aligned with hybrids only in order to make non-translation, the categories of apparently primary social belonging, seem pure and unadulterated, even when accorded the richness of creative dramatic poetry. This imagined purity can then project sense-for-sense.  

Horace's tropes ricochet down the centuries of theorisation. In Schleiermacher, just in case anyone caught my reading of him in the journal Translation and Literature last year, the figure of the Blendling (the-translator-as-bastard) is clearly the hybrid, social isolation is in the translator-as-child-abandoned-to-troupes-of-acrobats (perhaps drawn from Dryden's translator-as-tight-rope-walker), the stilted opposition to nature is word-for-wordness negating 'natural gymnastics', and ridicule is in the smiles with which masters look down on the bad translator's laborious literalism, all this in the form of rhetorical questions. If you haven't read the passage, don't bother. Believe me, Schleiermacher is involved in much the same game as Horace, associating literalist translation with the risk of hybridisation, precisely in order to preserve the ideal of belonging as non-hybrid. Of course, the German theorist seeks a mode of translation that will keep the risk within acceptable limits, developing rather than adulterating Germanic identity. And herein lies a key: Schleiermacher can only control the risks by excluding commercial translation (Dolmetschen, negotiations, the stuff of the interpres) and by insisting that the translator, like any citizen, must belong to only one culture: 'Wie Einem Lande, so auch Einer Sprache oder der andern, muß der Mensch sich entschließen anzugehören', people must finally belong to one country or another, one language or another. If not, if I were a real hybrid using my in-between situation to do translations, who would pull me out of a pit if need be?  

Strange translation theories: translations themselves, as texts, can run the risk of being hybrids, since that preserves the illusory purity of non-translations as non-hybrids. Yet translators as people cannot; they must be qualified as faithful (fidus) for Horace, as German for Schleiermacher, and as general target-culture members for most of our self-describing Descriptive Translation Studies. Strange theories, since it seems to me that translators, by the very nature of their plural competencies, if not always thanks to mixed birth and extensive mental or physical travel, are quite likely to live and work in the cultural overlaps or intersections marked by hybridisation, in what I would like to call 'intercultural space'. I mean that translators, as people, are more likely to be hybrids than are the texts they produce. As such, they have long been controlled by normalising discourses on their allegiance (fidelity, duty, loyalty and the like) or simply shelved behind discourses on the qualities of their texts. Petrus Venerabilis could describe the Qur'an as a hybrid, but he did all he could to avoid the consequences of translators sharing the same status. His translation team comprised at least one Mozarab, one anonymous 'native informant', two scientific Latinists not particularly interested in religion, but then also Petrus's own private secretary, to make sure the Latin was really Latin and not the product of any half-baked infiltrators from the frontier regions (what languages did they use when discussing the Qur'an?). Norman Daniel has called the result 'a Clunaic paraphrase', hardly the sort of textual hybrid we might expect from the drawing or the border situation. The figure of the hybrid might have something to do with translations, but that does not mean all translations are necessarily hybrids.  

I have two things left to say. The first is easy, since it's no more than a question: What would happen if human translators, despite their hybrid situation, were not quite alone in their intercultural predicaments? What if there were many others there, in the intersections of cultures, able to form some kind of intercultural community, able to pull each other out of pits? Surely then, and perhaps only then, what we are calling hybridisation could start gaining respectable press in translation studies, at the same time avoiding at least some of the unwitting affirmations of non-translational purity. If that were possible, I might even recover enthusiasm for translation studies.  

My second point follows on from this but is more difficult to argue: I would like to suggest that the people sharing the translator's intercultural space are increasingly the authors of source texts, and that the real hybrids, the out-and-out weirdos, are more likely to be precisely those source texts.  

Perhaps I should explain that most of my own professional translating these days involves work on texts that are either overtly multilingual or at least multidiscursive, incorporating fragments from various sources and various degrees of lingual competence. Here, for example, is an EU report the drug problem in Guyana: the main author is a Senegalese who writes in bureaucratic French and tries to do something similar in English; the second author, responsible for parts of the same text, is visibly a native speaker of British English who has studied law, since he produces legalistic prose no matter what the circumstances; and interspersed with this French+bad English+legal English are crumbs from previous reports of all kinds, ranging from UN jargon to CIA notes. The result is indeed a horse-necked feathered girl's head on a fish! And that's without explaining how this work comes to me from a Catalan-Spanish client in Barcelona.... So what do I, as a translator, do with this hybrid? Why, I put it all into good English, as neutral and bland as possible (not a trace of my native Strine!), translating, correcting, revising, homogenising, reordering, adding and subtracting where necessary, since I'm paid to produce a text that will attract the necessary subsidy. From extremely heterogeneous sources I write an English that lives almost nowhere (well, is Brussels a place?). Another case, one of my favourites, is a report on the sociology of domestic appliances, written in a Spanish-English-Spanglish mix that can only be properly domesticated when a French subtext is deciphered: for example, the designers of appliances are called 'innovators'/'innovadores', since some previous author has sought to avoid the Anglicism 'designer' in French, thus coining 'innovateur', rendered into Spanish and Spanglish as such before I happily returned it to the original 'designer'. Through an archeology of the multilingual source I, as translator, could happily plaster over most of the fissures.  

Perhaps you'll object that these are mere exceptions. True, not all sources are quite so multi. Yet even the apparently monolingual texts I receive freely incorporate foreign terms and rarely give a damn about the correctness of their national norms. The reason is simple: these texts are only meant to be read by the research teams involved; within restricted intercultural circles they can freely communicate in several languages or bad language; and it's only when the thing reaches me, the translator, that any exit to monolingual or inexpert readers is assumed and dehybridisation has to take place. Within the kitchen, anything goes; the translator's task, in this context, is to make the dish look coherent and to serve it in the form most suited to the external readership.  

In case you still don't believe sources are becoming more hybrid than their translations, in case doubters want to see and touch, here are excerpts from the minutes of an EU meeting held, apparently, in English:  

DK: Denmark supports the Structural Business Statistics Regulation although it is thought to be of too great detail. The Danish statistics have got the responsibility of a general Company register, from the very beginning all sectors were covered.
FIN: the Finnish register has been existing for a long time and is now being revised to meet the requirement of the Regulation. It is quite comprehensive. The statistical units are however a difficulty.
B: Some work is ongoing with the Social Security Office to find the local units. Not considered will be the associations with no employees and the public sector.
L: the company register was published a fortnight ago in two volumes: per activity class and alphabetical order. It is also available on informatic support.
S: Concerning the improvement of the Register the quality of the classification has to be refined as a priority.
A: In 1996 the Register will be extended to deduce a kind of demography of the enterprises.
D. An important prerequisite is the introduction of a law on business services statistics (Dienstleistungs-statistikgesetz).
N. The delegate invokes the problem of which group would be looking after the implementation of the structural regulation.
F: France has been researching these sectors some years ago already. At national level use is already made of the approach of surveys in turn.
GR: Greece would appreciate having the possibility to update the Register regarding new start-ups in order to cover completely new enterprises demography. It is not clear how long it will take to set up the Register.
E: Small problems and gaps still in the services area due to classification.
UK: It is considered that classifications do not always give answers to the needs as they are sometimes not fine enough. In the future, a product breakdown will be asked in annual enquiries.

Note that the macrotext is itself passably hybrid, all in the present tense, when something Anglo-Saxon within me wants minutes to be in the past. As for the details, the various languages and norms should be clear enough: the infinite patience of the Finnish 'It is quite comprehensive'; the German need to make a law and give us its name; the French 'surveys in turn' (enquêtes tournantes) for 'cyclical surveys'; the nice passives that virtually only come from the United Kingdom, though even that text could do with some polish. Leave the linguistic analysis for that Monday morning translation class (yes, translation) for which you never have anything prepared. What interests me here is the fact that, no matter how ridiculous the text on the level of grammar, all these people obviously understood each other's English well enough to carry out the meeting in that language. No mountains were labouring. Of course, one might hope a few francophone participants will improve their English with time, although it might be more cost-beneficial to teach the others some basic French, so that no one gets thrown by the occasional opaque calque, so that the members of this small intercultural community can effectively pull each other out of linguistic pits.  

Whatever the case, this, ladies, is your hybrid space. It is a space of text production from within cultural overlaps. You might argue, if you're clever, that most of the people participating in this meeting were effectively working as translators, expressing their thoughts in a foreign language. Yes, but not quite. Technically, what these people have produced is as much a non-translation as anything else, since there are no paratextual markers to tell us otherwise and the text producers were paid only to produce texts. These people are in the same space as translators, but they are not working as translators. The person paid to translate and mark the text as a translation was in this case my wife, whose task was not only to put it all into good French but also to clear away the mixes, to pretend there was never any real hybrid there, and of course to waste EU money in the maintenance of that illusion. The real question, though, is this: Since all these people communicated in English, why do they need a translation into French (some two months after the meeting!)? Whatever answer you find, it must surely include an ideological warding off of hybrids.  

So much for my examples. Here, in summary, are a few hypotheses that might be matched against your own:  

· Contemporary professional non-literary translation in Europe (isn't that what you wanted to discuss?) is an agent of dehybridisation for the simple reason that source-text generation processes are increasingly multilingual, whereas translational outputs are normally monolingual.  

· Translation in general (my subject) is an agent of dehybridisation in the sense that it creates and projects the illusion of the non-hybrid text.  

· Translators as an intercultural community are nevertheless a focus of massive hybridisation on the ethnic, cultural and linguistic levels.  

· The increasingly intercultural nature of source-text generation processes means that translators are not alone; they share much of their intercultural space with those who do indeed produce textual hybrids.  


I naïvely hope my arguments will have some kind of consequence for translation studies. We cannot continue to re-cite the classical authorities of the past; we cannot passively describe translations in terms of 'interference'; nor should we be content when deconstructionist clones claim that all translations are by definition transgressions, elements of some kind of futureless liberation. Such arguments simply manipulate the translation-as-hybrid to affirm the illusory homogeneity of non-translations. That, I'm afraid, is a blindly reactionary game, forever caught in the geometry of one-side-or-the-other. Far better, I suggest, to look at the mixes that actually occur in the space of translators as people, and at the even greater mixes ensuing from the intercultural communities that provide translators with professional contexts. Far better, that is, to do rather more than translation studies currently allows.  

If there is to be any kind of liberation from nationalist ideals (and that's what I want), it will not come from the ideologies of hybrids. Nor will it come from the employment of translators to produce pretty texts. What we must accept, I believe, is the existence of substantial intercultural communities. From that basis we can ask a few important questions for our future.  

Best wishes to you both,  




Last update 11 March 1999  

© Anthony Pym 2014
URV. Av. Catalunya, 35
45002 Tarragona, Spain
Fax: + 34 977 299 488