TRANSLATION DEPENDS ON TRANSFER

Transfer and translation work on distance

 If there are any closed cultures, we know nothing about them. This might sound merely pious, but if it can be accepted that we do not live within closed cultures - that our own culture is open and is engaged in exchange with other open cultures - , it is also possible to accept that everything we know about cultures beyond our own has come to us, has been appropriated or assimilated, through processes of transfer and translation.

Similarly, and as a necessary consequence, everything we believe or suspect we do not know about other cultures has been at least prefigured by processes of transfer and translation.

It might then be concluded that transfer and translation operate on the semiotic distance between known and unknown signs. This could be what they do as general activities. But as for what they are or should be as practices conditioned by historical factors, as for the way they relate semiotic and material distances, the matter is a little harder to grasp.

Happily there are a few basic principles concerning the way transfer and translation are related as specific practices. The purpose of this essay is to formulate a few of the more obvious principles.

My initial proposition is that if there were no material transfer, if texts were not moved across time and space, there would be no translation. This suggests that translation can be seen as a response to transfer. However, I also wish to propose that translations represent and often misrepresent the time and space crossed by texts. Transfer and translation thus open up two quite different ways of approaching the distances they work on, the first based on responses, the second on representations.

In this chapter I shall work first from transfer, and then from translation, in order to formalise a general approach to the union of the two.
 
Transfer is a precondition for translation

The English nominalisation "translation" is derived from translatus, past participle of the Latin verb transferre, "to carry over or across". It is from no more than the past participle - by definition coming after the event itself - that we have the nomen actionis "translating" (translatio) and the nomen agentis "translator". English would seem to have lost the association these words once had with the less specific and more material sense of transferre. Our common terms are really only articulating translation as "the translated", as the completed result of translational work. Contemporary terminology thus tends to ignore the wider process that might nevertheless be recovered and nominalised, from transferre, as "transfer", to be understood here not in its psychological sense but simply as the physical moving of something from one place and time to another place and time.

In this light, translation can be seen as a special kind of response to things that have been transferred or are meant to be transferred.

In the next chapter I shall describe how this particular kind of response is defined by equivalence. For the moment I simply want to argue that, on the level of general abstract concepts, translation depends on transfer. Let us investigate a few possible objections to this proposition.

First, it might be complained that since no text need actually be moved in order to be translated, translation can take place independently of transfer. But insistence on one-to-one solidarity - demanding one act of transfer for each act of translation - has little to do with what I mean by translation's general dependence on transfer. Just as no person is an island and no culture is entirely isolated, no translator ever works entirely alone or in a strictly one-off situation. Even when translators are not aware of responding to any particular act of transfer, they will necessarily be using linguistic and cultural knowledge accrued from previous translations, depending on previous transfers, which are themselves responses to previous translations, and so on in a series of links that unavoidably chain the particular to the general.

Sceptical minds might then interpret the connection between transfer and translation as a question of chickens and eggs. Yet the relation in this case has none of the cyclical causality of genetic or generative metaphors. In its epistemic dimension, the dependence of translation on transfer is decidedly one-way. This is because although translation depends on transfer, transfer does not depend on translation. That is, if there were no translation, there could still be transfer; but if no text were ever going to move, there would be no reason even to think about translation as a purposeful activity. Whatever the material circumstances, no matter whether the translator is situated before or after the actual movement of a text, the concept of transfer precedes the concept of translation. This is a general principle. It has several practical consequences.

Let us suppose that, because of this general dependence, translators commonly have ideas concerning the transfer which has taken place, is to take place or, in a teaching situation, could take place with respect to the text to be translated. In other words, translators must and do have ideas about the purposes of transfer. If such ideas are necessarily based on previous transfers and translations - since there is always already contact between the cultures we know - , accrued general ideas about where a translated text should go and how it could be received can adequately inform the translator's work quite independently of the actual movement or reception of any particular text.

This is why it would be quite naïve to suggest that material transfer were immaterial to translation, as if intercultural virginity were the necessary condition of immaculate equivalence. Or again, the idealist notion that there can be translation without transfer is like saying that there can be poetry not written for a reader. The world is no doubt full of miraculous conceptions, unread poems and apparently immobile texts, but these particular cases do not annul the general qualities of sexuality, poetry and translation as modes of communication. Texts are only translated because they are transferred.

Transfer thus has logical priority as a necessary precondition for the general practice of translating. If nothing has moved or is going to move from A to B, then there is no reason to translate from the culture of A into the culture of B. If someone is translating or has translated, then something has moved or is meant to move.

Examples bearing on transfer and translation in Spain:

a) Although airplanes transfer passengers, cargo and pilots, they do not require any translation of the pilots' language, since international aircrews use English as their lingua franca. A major air crash in Tenerife was reportedly caused by a Spanish pilot's English being misunderstood by a Spaniard. The disaster thus resulted from transfer without translation.

b) Since printing costs are lower in Spain than in most other west-European countries, Spanish publishers can use a certain economy of scale to print full-colour books for foreign-language clients, keeping the same illustrations and just replacing the text where necessary. Economic imbalance thus gives rise to a transfer situation requiring translations. Translators are sometimes employed by the client publisher but more usually by the Spanish publisher, with the client then undertaking extensive editing of the result. The second combination is the more common because, thanks to a further economic imbalance, translators' rates are usually lower in Spain and professional translators' associations are mostly ineffective. Transfer in this case not only requires translation but also tends to determine the material location of translators.

c) Spanish academics at a symposium on translation history presented papers in which, in a particularly self-serving way, many of them criticised the poor quality of published translations. For each mistake, they usually proposed the "correct" version or at least an improved translation, and often an invocation of untranslatability. That is, they presented alternative translations that appeared to be free from the constraints of any real transfer situation - no deadlines, no salary problems, no demanding client, no economic imbalance, no financial prohibition of non-translation. Is there then translation without transfer? But in this case the pertinent transfer was the movement of texts from the professional translator to the academic critic, from the open marketplace to a university symposium. If this transfer is mostly hidden, it is no doubt because one of the functions of such symposia is to perpetuate the illusion of translation without transfer, as well as several peculiar notions concerning ideal equivalence, the resulting mythology of untranslatability, and hence the apparent need for academic criticisms of translations.
 
Exactly what is transferred?

It is easy enough to say that a text is what is transferred. But exactly what is a text? How do we recognise one when we see it? And how do we recognise the fact that it has been or is being transferred?

In order to answer these questions properly, we need a clear idea of the kind of transfer pertinent to translation. Apologies might be necessary for basing my explanation on a passage from a deservedly little known novel set in Western Australia. It concerns sheep, aborigines and communication:

"Patrick turned away from the ewe to look about the cave, seeing the handprints left by the aborigines, for a purpose unknown.

'My hand is twice as long as that one,' said Jane, pointing to a small handprint low down on the back wall.

'It's a child's,' he said, and they felt strange and sad at the thought of the dead piccaninny who had perpetuated himself in this way." (Randolph Stow, A Haunted Land, 1956, 126-7).
 
Here there are two acts of transfer. The first is the movement of the receiving subjects - Patrick and Jane - deeper into the cave, from a scene of death in natural reproduction - a ewe dying after lambing - to a scene of intercultural communication as artificial reproduction. The second movement is the transfer of rock-marks across time, from the unknown moment of their production to the moment of their reception by Patrick and Jane. Both movements are important, but they have quite different qualities. Patrick and Jane are subjects; their movement opens up possibilities of reception and thus possibilities for the translation of the things they come into contact with. The rock-marks are objects; their movement through time opens up possibilities of them being received and thus translated by the subjects who cross their trajectory. Subjects can receive and translate; rock-marks cannot. Which is why translation studies should generally consider translators to be subjects - or mechanical extensions of subjects - who work on transferred objects. Few theorists would disagree.

This simple principle underlies the rather more interesting proposition that translation studies should accord more priority to the movements of objects than to those of subjects. That is, although the study of intercultural relations has to pay considerable attention to the subjects transferred through trade, migrations, wars and explorations, the specific study of translation finds its privileged point of departure in the objects transferred, quite possibly by the same trade, migrations, wars and explorations. For intercultural relations, it could be important that Patrick and Jane moved, that western eyes encountered pre-industrial cave-space. But for translation studies, which is only interested in Patrick and Jane insofar as they are or could become translators, it is more important to know about the movement of what they found.

Now, having identified the kind of transfer that interests us, how can we identify whatever it is that moves in the process of transfer? What are these rock-marks?

Although theories of translation rarely talk about transfer as such, they do tend to make assumptions about what can ideally be taken from one culture to another. For some, there is no real movement, since the one mark always approximates the same pre-existing "meaning" or "concept": if the handprint meant "hand" when it was made and it means "hand" when received by Patrick and Jane, how could one say that anything has moved? Universalist semantics wants us to believe that everything was always already there. In this way, blindness to transfer does away with the basic reasons for translation. For other theories, however, there is real movement in the sense that the mark functions as "information", "signification", a "message" or even "enlightenment", bringing new meaning to the particular receivers Patrick and Jane. This approach can at least explain why there should have been an act of transfer and thus the possibility of translation. But does it really matter what the mark might have meant before it reached these new receivers?

Patrick and Jane do not know what meaning or concept the mark had for its producer. Nor do they really care. But they can be fairly sure that the original meaning or concept had little to do with a dying ewe as a symbol of natural reproduction, if only because sheep were introduced to the land at about the same time as Patrick's great-grandfather migrated there. A radically changed context means that meaning in production cannot be equated with meaning in reception. But should we then abandon all talk of meanings and concepts? Should we say that the mark is something entirely new in its situation of reception?

Patrick and Jane recognise the mark as being meaningful. However, this is not the kind of meaning that Leonardo found when looking at the forms of clouds or decaying walls, nor that of a geologist who might find in the cave certain inscriptions of gold mineralisation. Patrick and Jane know this particular piece of rock is meaningful because it has been marked by another subject. It is not a natural piece of rock. It is of the same substance as the surrounding rock but its form indicates that it has been intentionally inscribed, that it bears the trace of purposeful work. Ochre liquid spat from an absent mouth outlined an absent hand placed against this rock; the production of this archetypal mark was both oral and manual. Without knowing why the inscription was made, the potential translators recognise it as an inscription made for some purpose. This is thus not a natural object; it is what Rossi-Landi (1975) describes as an artefact; or more provocatively, it is what we might insist on calling a text, an object endowed with meaningful materiality.

When Patrick and Jane recognise this part of the rock as being meaningful, what happens to the natural rock itself? When Jane focuses on the shape and size of the hand, is the rock material suddenly without consequence?

It has become commonplace in linguistic and literary theory to define a text as an intangible complex of semiotic relations, insisting that its status as an object of knowledge not be confused with its material support - a text is said not to be a book - and sometimes declaring that it only completely exists when concretised in reception (after Ingarden 1931). According to such definitions, the text here would be no more than the hand-shape, a structural relation between certain conventionalised curves and lines, with the rock material acting as a merely transitory support. If the form were in relief, it could be inked and transferred to another support, perhaps a sheet of paper; if Patrick were a photographer, it could be transferred to film. And if such simple reproduction were all that was involved, one could happily talk about structure as that which is transferred from rock to paper or film; one could adequately regard the text as a question of forms, a semiotic complex, indifferent to questions of substance or support.

But can texts - including oral texts - ever exist without the materiality of a support? Does their status as an object of knowledge ever not presuppose a level of substance? The kind of transfer that goes from rock to paper or film requires that the supports come into material contact or proximity with the inscribed form. When, as in the case of the rock-mark, such simple transfer is across time instead of space, the contact between form and support is continuous. But the principle of necessary materiality is the same in both cases. It is impossible to find a text devoid of a support, be it rock, electrons, genes, sound waves, or whatever else is able to go from one point to another. The nature of the support can change - from rock to paper, from paper to voice - , but at no point is the text liberated from the materiality of things that move. To imagine otherwise is to pretend that texts fall from the sky and exist forever.

Patrick and Jane are not just concerned with the hand-shape or with the text as form. They find a text whose materiality indicates it has come from another time. Flaking ochre and weathered rock must say more about the handprint's age than does the simple absence of its producer. Reception is concerned with a text which is both hand-shape and rock, form and material, since both these aspects are necessary if the receivers are to conclude that the absent producer is now long dead.

This textual materiality allows Jane to attach importance to the physical dimensions of the text - the hand-shape is of a certain size - , then use comparison to attribute meaning to that size - the shape represents a hand smaller than her own. Transfer thus enables a process of interpretation, a comparison, a figuration of the absent producer, a potential utterance and a complex contextual meaning as an artificial alternative to natural reproduction. For Patrick, as for most of Stow's heroes, writing will sublimate sexuality as transfer of the self. More importantly, no one need insist that ancient piccaninnies had any such meaning in mind.

Transfer in this case enables a process of interpretation which borders translation. Jane projects the absent producer through comparison with her own hand as text; she relates the imagined piccaninny to herself. Simple transfer might thus be enough for some form of knowledge to be produced through intercultural communication.

But is there any translation here? Is there any strictly translated text? If the second, interpretative hand had not been evaluated as significantly larger than the transferred text, if it had not been conceptually attached to the interpreting subjectivity known as Jane, then it might have been possible to consider it as a translation. Or again, perhaps one could consider Patrick as a potential translator, the translated text then being his phrase "It's a child's". But the deictic "it's" separates the object transferred from the subject translating, in the same way as sheer size separates the textual hand from Jane's interpreting hand. For these reasons, Patrick and Jane cannot be seen as translating the hand; they simply comment on it as an object external to their own time and place. There is a difference between translating a text and just talking about it or producing a similar text.

Translations are quite difficult to achieve; they are very particular kinds of communicative artefacts. As we have seen, not all acts of transfer need give rise to complete acts of translation. And as we shall see in the following chapters, translations moreover require fulfilment of a series of specific conditions which go well beyond transfer, including a certain kind of belief on the part of the person receiving the translated text.

Exactly what is transferred? For the purposes of translation studies, the privileged object of transfer is the text, independently of whatever meaning, information, message or signification might have been attributed to that text prior to transfer. But the text must be recognised as inseparable from material support, since it is only through materiality that its transfer can become significant.

The principle of meaningful materiality involves theoretical consequences well beyond our immediate concerns. It is possible, for example, that coherence and cohesion presuppose a continuity of material support both before and during reception, even when this continuity is not realised because of broken or ruptured transmission. It is conceivable that fanfares of intertextuality should be limited by quite reasonable criteria of historical contiguity: if there is to have been some kind of transfer from one text to another, then the two texts concerned must at some time have shared the same locus. But the important point for our present purposes is that the necessary materiality of texts condemns them to displacement. Indeed, not only are texts always available for transfer, they are by definition unable to avoid being transferred, through time if not always through space.
 
Translation can be intralingual or interlingual

It is often assumed that the kind of transfer most pertinent to translation is that which takes place exclusively between different languages. This restriction of the field assumes a radical division between interlingual and intralingual transfer. Unfortunately there is no such division, simply because there are no natural frontiers between languages. The kinds of translation that can take place between idiolects, sociolects and dialects are essentially no different from those between more radically distanced language systems. Consider, for example, the various transformations necessary to rewrite in the English of Queen Elizabeth II a text from American English, working-class Liverpudlian, Shakespeare's English, Chaucer's English, the French of François Mitterrand and Japanese. Although one would expect to encounter a need for increasing transformations with increasing cultural distance, there is no strict cut-off point at which wholly intralingual rewriting can be said to have become wholly interlingual. Those who travel on foot or have read the diachronic part of Saussure know that there are no natural frontiers between languages.

Since "language A" and "language B" are insufficient descriptions of the two places minimally involved in translation, some alternative vocabulary must be sought. A Chomskyan "ideal speaker-listener in a completely homogeneous language community" (1965, 3) would clearly be inadequate for much the same reasons as "language A": since there are many more languages in the world than countries to house them, the fact of bilingual and polyglot communities must be recognised and incorporated into any global approach to translation. Similarly, since numerous languages are spoken in more than one community, it must be admitted that texts can be transferred from one community to another and yet not require translation because the original language of the text is able to seek out its appropriate receivers.

Neither "language" nor "community" are sufficient criteria for the description of the kinds of places minimally involved in translation. A certain retreat to the bunker is necessary, in this case to the suitably vague term "culture".

That is, the kind of transfer I consider pertinent to translation is that which takes place between different cultures.

But what then is a culture? How might one define the points where one culture stops and another begins? The borders are no easier to draw than those between languages or communities. One could perhaps turn to a geometry of fuzzy sets or maybe even deny the possibility of real contact altogether, but neither mathematics nor ideological relativism are able to elucidate the specific importance of translation as an active relation between cultures.

Although questions like the definition of a culture are commonly thought to lie beyond the scope of translation theory, their solution could become one of translation studies' main contributions to the social sciences. Instead of looking for differentiated or distilled cultural essences, it could be fruitful to look at translations themselves in order to see what they have to say about cultural frontiers. It is enough to define the limits of a culture as the points where transferred texts have had to be (intralingually or interlingually) translated. That is, if a text can adequately be transferred without translation, there is cultural continuity. And if a text has been translated, it represents distance between at least two cultures. In this way, translation studies avoids having to link up all the points of contiguity in the way that political frontiers do. After all, there is no obvious reason why points of contact and exchange between cultures should form continuous lines. Culture is not geo-politics. Transfer and translation concern situations of contact and exchange, not lineal separations.

According to the solidarity of these definitions, specifically intercultural transfer is a precondition for general translation, and translation itself therefore logically indicates both the existence of intercultural transfer and the points separating the cultures concerned.

Instead of using preconceptions about cultures in order to form preconceptions about translations, it is thus possible to use facts about translations in order to locate contacts and differences between cultures. Indeed, to do so could be conceptually elegant.

How many cultures to a community, or communities to a culture?

a) The language of many sciences is now exclusively English, no matter where the actual scientific activity is carried out. This situation is reducing the need for translations, since the scientists speak and write directly in English where necessary. According to the above definitions, these sciences are thus becoming cultures in themselves, increasingly independent of their everyday contexts. Indeed, the frontiers crossed by scientific translations tend to be those separating the specialist from the wider public, such that scientific translation is becoming a synonym for vulgarisation or respect for an outdated nationalistic identification of language with community.

This example suggests that, in general, a unified monocultural stratum can be formed through non-translation. A further example would be the non-translation of the Koran, which, in separating those who understand from those who do not, forms a broad monocultural stratum embracing many different communities.

b) "They wrote out all the Mordecai's orders to the Jews, and to the satraps, governors and nobles of the 127 provinces stretching from India to Cush. These orders were written in the script of each province and the language of each people and also to the Jews in their own script and language" (Esther 8:9). Thus the multicultural communities of Xerxes' empire were held together by translation, so that individual cultures might survive and Purim be celebrated ever since by the Jews.

c) Between these two extremes - extensive monocultures revealed by non-translation; cultural frontiers revealed by translation - there are bicultural communities where it is difficult to decide if translation crosses a cultural frontier or not. When the Spanish Moriscos (from 1492 to the definitive expulsion in 1609) used Arabic characters to write in Romance - Castilian, Aragonese, Portuguese, Catalan or Valencian - , many of their texts were in fact loose translations from the spiritually untranslatable Koran (Vespertino Rodríguez 1990). The Moriscos did not know Arabic, so their translators transferred elements of their cultural past into Romance. But they used Arabic script, the script of the sacred Koranic language, as an outward manifestation of continuity with this cultural past: the words were different but the letters were the same. In this case, translation not only crossed a frontier but also symbolised a bridging of the same frontier.

However, historical analysis suggests that the real frontier here was not the linguistic interface crossed by the translators but the different script by which the Moriscos proclaimed their distance from the Christians surrounding them. When expelled from Spain, their literature remained in Romance but was written in Latin script, to further proclaim their cultural difference, this time from the Arabic-speakers surrounding them.

Translation can be approached from transfer

These few comments on the nature of transfer provide us with two basic ways of approaching its relation with translation. On the one hand, translation is partly knowable through the analysis of texts which have been translated (or, more ambiguously, through the past-participle form "translated text" or TT, which, from the perspective of the translating translator, can also be read as "target text"). On the other, to know why and how any particular translational operation was or should be carried out, we have to look at the factors involved in the transfer from a distanced or even imaginary source text (ST) to the place of a manifest TT. We have to ask what came from where and for what reason; and where, why and to whom the translated text is to go. Two complementary approaches are thus available from the outset: one is textual (translation as representation), the other is extra-textual (translation as response).

Almost everyone interrogates translation from the first of these perspectives, making vast use of semiotic science and diverse cultural convictions but in fact basing their observations on no more than translated texts as representations. However, the second set of questions, deceptively simple, can often subvert the conclusions thus reached.

Here, for instance, is a text visibly translated because a writer on translation has enabled us to compare it with a French source:

TT: For remember this, France does not stand alone, she is not isolated.

ST: Car la France n'est pas seule! Elle n'est pas seule! Elle n'est pas seule!

Peter Newmark (1977, 169) has carefully considered the relationship between these two texts, helpfully pointing out that the translator has paraphrased the source text. Newmark also insists that this kind of translation should not be allowed in the case of citations from "authoritative" texts. There can apparently be no legitimate reason for presenting these paratactic and perhaps hysterical French negatives as if they were stable hypotactic English logic. But let us consider the example a little more closely.

A sympathetic linguistic analysis might have tried to locate speech norms that allow the French an exclamatory voice not so readily available in English. A transformational approach might then have suggested that the repeated negatives were derived from obsessive suppression of the idea that France was in fact isolated. Could it then be that the parataxis had to be transformed because it represented a fear pertinent only to an uncertain future seen from the moment of utterance? Could the translation have some justification after all?

But linguistic analysis alone cannot properly explain this TT until transfer analysis reveals that a considerable jump has been made from a speech given by General de Gaulle in 1940 to a biography published by Major E. L. Spears in 1966. The real fear involved in 1940 could not fully be transferred away from its moment of production - we know who won the war - and could at best achieve a weak representation in translation. Confronted by an inevitable loss of discursive force on this level, the real question should be why Major Spears bothered to translate at all. Why should he have made de Gaulle speak English in 1966? Why was it important to have it known that de Gaulle himself produced this utterance? Why should a biographer remind readers that that non-isolation had been important to the France of 1940 before it was presumably of some importance in the Britain of 1966?

In this way, the analysis of transfer leads to the fundamental question to be asked of all translated texts: why?

It is not difficult to argue that, as the centre of the Commonwealth was becoming a satellite of the European Economic Community, questions of identity and national pride were becoming increasingly vexing and a military biographer would have good reason to render some very careful translations. Just as military France had needed Britain, so economic Britain needed France. Moreover, this Britain of 1966 did not need a vision of France that included discursive violence, historical paranoia or excessive Gaullist pride. Closer to material movement than was his analyst, the translator knew that words said in time of war should not be repeated - nor too literally translated - in time of peace. However, unfortunately for the translator and his country, de Gaulle was also something of an expert in transfer analysis: he had used this same principle to block Britain's eec entry in 1963 and was to do so again in 1967. Translators are not alone in their responses to transfer.

Transfer can be approached through translation

The above example should serve to illustrate why an adequate approach to translation requires something more than linguistics. The idea that translation involves more than mere language is of course by no means new, but my argument here is not quite the same as those who stress the importance of extra-linguistic semiotic systems, cultural knowledge or intuitive competence. What worries me is the fact that linguistic models - like the semiotic and pragmatic schemata that have been added to them - fail to conceptualise transfer as a bridging of material time and space. No movement is visible as long as the analyst places two texts side by side, calls one a source text and the other a target text, and attempts to compare the language used in both. The results of such analyses might be of interest to linguists, but they will not necessarily have anything to do with translation.

The most brutal way to subvert textual analysis is to work from the level of extra-textual coordinates, as has been done above with the introduction of the distance between France in 1940 and Britain in 1966. This is to analyse translation from the perspective of transfer. But is it equally legitimate to approach transfer from the simple relation between two texts, or even on the basis of one specifically translational text? How might such an approach be founded?

The most subtle way of incorporating transfer into what can be said about a translated text is to consider the logic of that text's absent alternatives. If a translator has produced a certain TT1, his or her work can be represented as a choice not to produce the alternatives TT2, TT3,... TTn. I think it would be fair to say that Major Spears produced his TT as a conscious negation of the far more obvious literalism blithely recommended by Newmark: "For France is not alone! She is not alone! She is not alone!". So much for the obvious.

Now, any series of possible TTs is necessarily bordered by two radical alternatives which are themselves always possible and pertinent:

- non-transfer, or the absence of both ST and TT in the place of reception (let us call its negativity X)

- transfer to reception, but without production of a TT (which inspires the interrogative symbol Y, to be read as the basic question "why?").

That is, Major Spears could have ignored the cave of de Gaulle's speech altogether (X) or he could have presented contact and yet decided not to translate it (Y).

From the perspective of translation, transfer can now formally be described as the movement from X to Y. But exactly where does this movement take place?

The symbols X and Y represent two particular positions at which things could have happened but did not happen. X would mean that a certain text could have been transferred to a certain point but was not. And Y would mean that the text was indeed transferred but was not translated. Working from translation, we thus locate transfer through conceptual negation rather than through reference to spatiotemporal coordinates. It is important to appreciate how this negation works.

If a text or reports of a text have not been transferred to a certain point, how can anyone at that point know that the text could have been transferred there? Indeed, how could anyone know the text exists? In all honesty, X can only represent the position located through conceptual or theoretical negation of an act of transfer which has actually taken place. Only once Spears has effective access to de Gaulle's text can it become meaningful to consider what would have happened if such contact had not been made. Only through transfer can the point of non-transfer X be projected as the necessarily re-created position of a text at its source.

Similarly, the position Y can only be meaningful as the negation of a translation which could have been carried out. That is, it is only meaningful as a point locating the place of a potential translator. One could of course trace the trajectory of a hand painted in a cave, saying that the transferred text could have been translated at any point over the past 50 or 500 years. But then Y would be little more than a simple description of transfer as a continuum of potential but unmanifested translations. It is far more fruitful to insist that the position Y is a particular point actively pertinent to translation. It is the specific point reached by the transferred text at the moment when it becomes the initial object for a potential translation by a subject. It is the rock-hand as it appears to the potential translators Patrick and Jane.

When the positions X and Y are thus located through conceptual negation, their theoretical significance far outweighs that of their coordinates in time and space. They become fixed positions only in the sense that they refer to specific stages in a text's capacity to provoke meaning or to be translated. In this light, it is fair to refer to texts with these capacities as occupying these positions. We may follow common usage in describing a text in position X as a "source text" or "ST" (despite the fact that the notion of "source" tends erroneously to suggest direct access to this position, assuming undemonstrable degrees of originality or inspirational production). And since there is no common-usage term for the position Y, we are forced to retain the term "transferred text", symbolised as "Y text", to be understood as a text at the logical moment immediately after transfer and immediately prior to translation.

My pedantry on this point is due to two concerns. First, I am trying to open the way for a theory able to address phenomena like pseudotranslations or translated texts for which the ST and Y texts are entirely imaginary. According to the above definitions, the fact that a text is read as a translation is sufficient basis for projecting both the X and Y positions and thus for analysing the text as a translation independently of the existence or non-existence of a source text. A baser materialism would have to exclude such phenomena. Second, I want to avoid the problem of the exact geo-political location of the position Y, the position at which transfer provokes translation, or translation responds to transfer, depending on the approach adopted. The process of translation is essentially indifferent to the physical location of the translator, who can be surrounded by ST culture, TT culture, or, by facsimile machine or modem, neither of these. Like Thomas Mann declaring (at Pacific Palisades?) "Where I am, there is German culture", or like de Gaulle broadcasting French resistance from London, the translator can carry a culture to any point on the globe. Moreover, since the generality of translation is set up by transfer, the place of the translator incorporates at least two cultures and their contact, independently of the geographical centering of the cultures themselves. From this perspective, it is idle to ask exactly where the position Y is located. It is enough to accept it as the imaginary negation of a translated text, and thus as the point pertinent to the question "Why was this text translated?".

The two radical alternatives X and Y now enable us to give a more formal definition of the relation between transfer and translation, dividing the entire process into three distinct moments:

- Transfer: the movement from X to Y, from absence to presence in the place of the translator, supported by the corresponding movement of text.

- Translating: the transformation of Y into TT, of textual presence into translated presence, available as negation of the alternatives TT1, TT2, TT3,... TTn.

- The translated text: the actually selected TT manifesting a translational relationship between itself and its antecedents.

Approaches to translation differ significantly with respect to which of these possible points of departure they choose to privilege. The users of the verb transferre privileged transfer and tended to take the rest for granted. Twentieth-century pedagogical, psychological, cybernetic and even purpose-oriented approaches appear to focus on the moment of translating and take transfer for granted. Linguistic and literary approaches are traditionally concerned with normative criticism of the translated text, being for the most part uninterested in understanding how, or in response to what, a Major Spears might have worked.

Confronted by this tripartite relativism, my choice of a merely double point of departure no doubt requires some justification.

How these approaches are used in this essay

As much as I am interested in what goes on in the translating brain or machine, I have absolutely nothing of importance to say about the matter. This alone is reason enough for approaching translating from the outside. But I am also sceptical as to how much of real value can be said from the inside.

The basic theory of choice between alternatives is obviously no more than an external reconstruction of something that is presumed to have happened. It should not be confused with everything that can happen when the translator actually gets down to work. The translating brain is very much a black box about which hypotheses can only be based on what goes in and what comes out. But then, if the input is what we have described as Y and the output is TT, this kind of analysis is in fact based on the relation TT:Y, the relation expressed in the ambiguous past participle of "that which has been translated", manifested by the TT, coming after the event itself. That is, analysis is in fact based on the third of the points of departure listed above. Alternatively, if translating is approached as a specific work situation, the various plays of influences and purposes forming that situation find their point of departure not in the translating brain itself, but in the intercultural determinants on Y, the why and wherefore of transfer. In answering the question "why?", the analysis of purposes is based on the position Y, on the extra-textual side of translating, the first of the above points of departure.

Approaches to translation thus at best skirt around the summit of translating itself, tending to seek more accessible slopes on either the textual or extra-textual sides. In recognition of this difficulty, I prefer explicitly to avoid assumptions about what happens in the inner intimacy of translators.

Since my overall purpose is to move from traditional translation studies towards wider social sciences, the order of the following chapters will go from the textual to the extra-textual, from the analysis of translated texts to the analysis of transfer. It would be possible to do otherwise, to begin from transfer and then attempt to generate translation, in the way that simplistic analyses of base economic relations once attempted to generate superstructural institutions. Such an approach would perhaps be conceptually easier and entirely appropriate for the analysis of an individual translation, but decidedly confrontational as a general critique, since it would mean insisting that ultimate truth lies on one side of the mountain and not the other. Although more difficult, an approach going from translation to transfer has the advantage of allowing one critically to adopt and undermine certain traditional assumptions, demonstrating the relative weakness or blindness of theories which exclude the fact of transfer. The order I shall adopt is thus less confrontational, but perhaps more subversive.

In this spirit, the following three chapters undo or rework traditional approaches to translated texts. These chapters correspond to the TT's relational value (equivalence), discursive status (there is no translational first person) and quantity (which hides the position of the translator).

I shall then suggest ways in which translations can be read as responses to transfer. This requires two chapters: one on how transfer itself can change the status of texts (since texts can "belong" to social groups or situations) and another on how transfer can be carried out by social groups on an intercultural level (thanks to intercultural regimes which facilitate and regulate text transfer).

The two approaches thus form a critique followed by an explanation of the critique. They are not to be distinguished as descriptive versus normative theory. Nor do they necessarily converge on the place of the translator. Instead, they come together as two different levels on which the union of transfer and translation can be analysed.

Since transfer and translation both work on distance, they can be brought together under a general formula describing this double analysis, hopefully as suggestive as it is neat:
 

Together, these four terms claim that translation is able to relate two kinds of distance: that represented by a translated text in relation to a transferred text (TT:Y), and that manifested by transfer itself (Y­ST) ("position of transferred text, minus position of source text").

Above the line, in the world of visible signs, Major Spears' translation (TT) represents de Gaulle's speech as it existed in Britain in 1966 (Y). This representation is expressed by the relation TT:Y. Below the line, in the world of moving objects, we know that de Gaulle's text crossed the time-space between France in 1944 (ST) and Britain in 1966 (Y), thus creating the distance expressed as the difference between these two points (Y­ST). The formula says that the first level (TT:Y) - the relation that is a result of translational practice - represents the second level (Y­ST) - the result of transfer.

If we want to know why Spears chose to translate de Gaulle's speech as if it had been written, hypotactic and unfrightened, we should thus carefully consider the situation of the transferred text - why should de Gaulle be translated in 1966? - and relate this situation to the distance between France in 1944 and Britain in 1966. On both these levels, the pertinent aspect of the distance represented is "absence of war-time situation". Spears' translation can be analysed and appreciated as a representation of this distance. It can thus be understood on its own terms, although it should not necessarily be praised as ethically astute.

The above formula relates the objects of knowledge with which we are concerned. It enables us to say what translations do as representations of distance and as responses to transfer. But it is a rather unstable formula. There is no guarantee of any constant ratio between these two levels; there is no superpower to insist that all translated texts must represent or respond to all acts of transfer in the same way. The line here is no more than an illusion of authority, interrogated and transformed - through the double appearance of Y - by translation as a diagonal cutting of both universal comprehensibility and incomprehensible cultural specificity. Of course, in the incertitude created by this double interrogation, in the lack of firm ground in heaven or at home, it is very possible and indeed common for translated texts to suggest that there is no transfer at all, that there is no distance between cultures, or that there is no real intercultural communication. I believe that such approaches open the way for relations which, in creating artificial paradises from the hiding of differences, or real conflict from the hiding of everyday exchange, are often as benighted in conception as they are pernicious in practice.

The above formula is thus not neutral with respect to its object. In insisting on the pertinence of transfer, in basing intercultural relations on the distance created by the movement of objects, this double approach must ultimately argue against the false authority of translated that which suggest they were always already there, and against non-translated texts that suggest they can exist nowhere else.
 

Table of Contents

Next chapter