The Task of the Translator

Nowhere, so far as a work of art or an art form is concerned, does taking account of the person receiving that work or that form prove fruitful as regards gaining knowledge thereof. Leaving aside the fact that any link to a specific audience or equivalent constitutes a distraction, the very concept of an ‘ideal’ recipient is an evil in all discussions of art theory, such debate being required only to presuppose the existence and essence of the human being in general. It follows that art itself, while presupposing the physical and mental essence of humankind, does not, in any of its works, presuppose man’s attention. For no poem is aimed at the reader, no picture at the viewer, no symphony at the people who are going to hear it.

Is a translation aimed at those readers who do not understand the original? That would seem adequately to account for the differing status of the two in terms of art. It would also seem to offer the only possible justification for saying ‘the same thing’ over again. What, in fact, does a piece of fine writing ‘say’? What does it communicate? Very little to the person who understands it. Essentially, it is neither communication nor statement. Yet the translation¹ that seeks to communicate could never communicate anything except its being a message – nothing essential, then. In fact, this is also a distinguishing mark of bad translations. But the thing that, apart from communication, exists in a piece of fine writing (and even the bad translator will concede that this is the essence) is surely, is it not, seen universally as the incomprehensible, mysterious, ‘writerly’ component? The part that the translator can reproduce only by – becoming a writer himself, perhaps? Actually, herein lies a second characteristic of the poor translation, which may be defined accordingly as the imprecise communication of an inessential content. And so it will remain, just as long as a translation sets out to serve the reader. But if the translation were aimed at the reader, so too must the original have been. If the original does not exist for the reader’s sake, how shall the translation be understood on the basis of that relationship?

Translation is a form. To grasp it as such, we must go back to the original. For it is here that the law of translation lies: in the original’s translatability. The question of the translatability of a work is an ambiguous one. It may mean: will there, among the totality of its readers, ever be an adequate translator for the work? Or it may, more authentically, mean: does the work, in essence, admit of translation and hence (in line with the significance of the form) ask to be translated? Basically, the first question requires a purely problematic answer; the second an apodictic one. Only superficial thinking will, by denying the autonomous sense of the second question, give equal importance to both. On the other hand, it should be pointed out that certain relational terms retain their full (possibly their fullest) meaning when they are not, from the outset, made to relate exclusively to people. One might, for instance, hear tell of an unforgettable life or moment even when everyone had forgotten them. The fact is, if their essence demanded not to be forgotten, such a predicate would not be untrue but merely a demand that people fail to meet – at the same time, presumably, as pointing implicitly to a realm in which it was met: the memory of God. Accordingly, linguistic constructions should still be deemed translatable even where to the human mind they are not. And ought they not, taking a strict view of translation, truly to be so to a certain extent? In the context of such uncoupling, we need to ask: should translation of specific linguistic constructions be required? For the principle applies: if translation is a form, translatability must be of the essence of certain works.

Translatability is an essential property of certain works. That is not to say that translation of them is essential as such; what it means is that a certain significance possessed by the originals finds expression in their translatability. Clearly a translation, no matter how good it is, can never mean anything so far as the original is concerned. Yet thanks to the original’s translatability, the translation is very closely connected with it. Indeed, the connection is all the closer for the fact that it [the connection] no longer means anything so far as the original is concerned. It may be called a natural connection – more precisely, in fact, a living connection. Just as the expressions of life are very closely linked to the living creature without being of any significance to that creature, so does the translation proceed from the original. Not, admittedly, from the life [Leben] of the original so much as from something higher, its ‘survival’ [‘Überleben’]. The translation comes after the original, of course, and in the case of important works, which never find their chosen translator at the time of their coming into being, it of course denotes the stage of the original’s continued existence [Fortleben]. The idea of the life and continued existence of works of art is to be understood in a wholly non-metaphorical, objective sense. That organic corporeality was not the only thing to which life could be attributed was thought likely even in times of the most hidebound thinking. Yet it cannot be a question of extending life’s dominion under the feeble sceptre of the mind, as Fechner tried to do;² and certainly not of its being possible to define life on the basis of the even less decisive elements of the animal, nor on the basis of feeling, which can only occasionally characterize it. No, only when life is attributed to everything that possesses a history and is not simply its showplace does the concept receive its due. For it is on the basis of history, not of nature, and certainly not of such fluctuating elements as feeling and mind, that the sphere of life must ultimately be defined. That is why it is the task of the philosopher to understand all natural life on the basis of the wider sphere of history. And surely at least the continued existence of works is incomparably easier to recognize than that of creatures? The history of great works of art knows their descent from the sources, their fashioning in the time of the artist, and the era of their basically everlasting continued existence among subsequent generations. The latter, when it occurs, is called ‘renown’. Translations that are more than communications come about when a work, in continuing to exist, has reached the time of its renown. They therefore not only serve this, as bad translators are in the habit of claiming for their product; they owe their very existence to it. In such translations the life of the original attains its (ever-renewed) latest, most comprehensive development.

That development, belonging as it does to a peculiar, superior life, is determined by a peculiar and superior fitness for purpose or functionality. Life and functionality – their apparently tangible yet almost unknowable connectedness becomes accessible only where the purpose towards which all individual functionalities of life strive is again sought not in its own sphere but in a higher one. Ultimately, all manifestations of life that are fit for purpose, like their fitness for purpose itself, are so not as regards life but as regards expressing its essence, portraying its signification. Translation, for instance, is ultimately fit for purpose as regards giving expression to the innermost relationship of languages to one other. It cannot possibly reveal that hidden relationship itself, cannot possibly produce it; but it is able to portray it – by realizing it, either in embryo or intensively. Moreover, this setting-forth of a thing signified by means of taking a stab at it or through the seed of its bodying-forth is a most peculiar mode of portrayal – such as in the realm of non-linguistic life is only rarely to be met with. For in analogies and signs this kind of life knows other types of suggestion than the intensive type, i.e., anticipatory, allusive realization. However, the intentional, innermost relationship between languages is that of a peculiar convergence. It consists in the fact that languages are not ‘foreign’ to one another but instead are a priori and regardless of any historical links related to one another in what they are trying to say.

With this attempt at an explanation, however, our inquiry seems to have followed various futile diversions only to find its way back to the traditional theory of translation. If in translations the relatedness of languages needs to prove itself, how else could it do so than by conveying the form and meaning of the original as exactly as possible? Regarding the concept of such exactness, that theory would of course have been all at sea – so could not, in the final analysis, have given any account of what is essential in translations. Truth to tell, however, the relatedness of languages becomes evident at a far deeper and more specific level in a translation than in any superficial, indefinable similarity between two pieces of literature. Understanding the real relationship between original and translation means launching an inquiry entirely analogous in intent to the argumentation by which epistemology must prove the impossibility of image theory. If it is shown in the former case that there could be no objectivity (not even a claim to objectivity) in knowledge if such existed in images of the real, in the latter case it can be proven that no translation would be possible were it trying its utmost to achieve similarity to the original. For the original, in continuing to exist – that is to say, in its continued existence (which could not be termed such, but for the way in which anything living alters and is renewed) – undergoes a change. There is a further maturing even of words already set down. What in an author’s day was perhaps the slant of his literary language may later cease to be so; immanent tendencies, new ones, may emerge from what has already received form. What once sounded youthful may later sound tired; what formerly sounded colloquial may subsequently acquire an archaic ring. Seeking the essence of such changes (as well as of those, similarly constant, affecting meaning) in the subjectivity of later generations rather than in the inmost life of language and its works would mean (assuming the very crudest kind of psychologism) mixing up the reason for and the essence of a thing; speaking more strictly, however, it would mean denying one of the mightiest and most fruitful of historical processes out of weak thinking. Nor would even trying to enshrine the author’s last stroke of the pen as the work’s mercy killing redeem that dead theory of translation. The fact is, just as the tone and signification of the great works of literature change completely with the centuries, so too does the translator’s mother tongue. Indeed, while the literary text lives on in its own language, even the greatest translation is doomed to wither as its language grows and to die in the renewed version. So far is it removed from being the infertile equation of two dead languages that, of all forms, it falls to translation to perform the special mission of noting the further maturing of the foreign language at the same time as the throes of its own.

If in translation the relatedness of languages manifests itself, this happens in a different way than through the vague similarity of replica and original. As indeed it is clear in a general sense that looking alike is not an inevitable concomitant of being related. Moreover, the concept of relatedness is in this connection also synonymous with its stricter usage [as actual ‘kinship’] in so far as it cannot adequately be defined by identity of descent in both instances, although of course as regards determining that stricter usage the concept of descent will always be indispensable.

Where, other than in the historical dimension, is the relatedness of two languages to be looked for? Not, certainly, in similarity of literatures any more than in a similarity of words between them. No, all supra-historical relatedness of languages rests on the fact that in each of them, considered as a whole, one and indeed the same thing is meant that on the other hand can be attained by no single one of them but only by the totality of their mutually complementary intentions:³ namely, pure language. The fact is, while all individual elements of foreign languages (words, sentences, connections) are mutually exclusive, those languages complement one another in terms of their intentions. Fully understanding this law, which is among the most basic in linguistic philosophy, involves distinguishing, at the level of intention, what is meant from the manner in which it is meant. In Brot [the German word for ‘bread’] and pain [the French word for ‘bread’], while the same thing may be meant, the manner in which it is meant is not. It is because of the manner in which they are meant that the two words signify something different to the German and the French recipient respectively, that they are not interchangeable so far as the two of them are concerned, that indeed in the final analysis they tend to exclude each other; but that in terms of what is meant they (looked at absolutely) signify the selfsame thing. Whereas the manner in which they are meant pulls in such different directions in these two words, in the two languages from which they are taken it amplifies itself. What happens, in fact, is that in them the manner of meaning expands to become the thing meant. In the individual, non-amplified languages, what is meant in them is never to be found in relative autonomy, as with individual words or sentences, but rather in a constant state of flux until it is able, out of the harmony of all those manners of meaning, to emerge as pure language. Until then, what is meant remains concealed in languages. If, however, languages do grow in this way until the Messianic culmination of their history, then it is translation that, aroused by the eternal continued existence of literary works and by the never-ending renewal of languages, repeatedly puts this sacred growth of languages to the test, asking how far removed is what they conceal from being revealed, how present may it become in the knowledge of that removal?

This is to admit, of course, that all translation is simply a somehow provisional way of dealing with the foreignness of languages. A solution of that foreignness that is not temporal and provisional, one that is instant and final, is not open to humankind or is at least not to be aimed at directly. Indirectly, however, it is the growth of religions that in languages ripens the encased seed of a higher language. Translation, therefore, although it cannot (unlike art) claim permanence for its products, does not hide the fact that it tends towards a final, definitive, decisive stage of all linguistic construction. In it the original grows into a (so to speak) higher, purer air of language in which it is not of course able to live with any permanence, as it does not, by a long chalk, reach that far with all parts of its nature, but towards which it at least points in a wonderfully insistent fashion as towards the preordained, failed atonement and fulfilment sphere of languages. The original does not attain that sphere altogether, but that is the location of what, in a translation, is more than communication. To be precise, this essential nucleus may be described as what, in the translation, is not itself translatable. The fact is, one can take from it as much communication as possible and translate that, but a part will be left untouchable, the part at which the work of the true translator had been aimed. It cannot be ‘carried across’ as the literary text of the original can because the relationship of content to language is quite different in original and translation. In the former these may constitute a certain unity, like fruit and skin, but the language of translation wraps its content in a flowing royal robe. For it signifies a higher language than in fact it is and as a result, so far as its own content is concerned, remains inappropriate, high-flown, and strange. Such fragmentation gets in the way of that ‘carrying-across’ at the same time as it renders it superfluous. For any translation of a work from a specific moment in linguistic history stands, as regards a specific aspect of its content, for those into all other languages. Translation therefore transplants the original into what is at least to this extent (ironically) a more final realm of language: that it cannot, by any kind of further translation, be moved on from that realm but can only, always from scratch and in other respects, be elevated within it. It is no accident that the word ‘ironically’ here may recall certain trains of thought of the Romantics. Before others, the Romantics had an insight into the life of works of which translation is a supreme manifestation. Of course, they were scarcely aware of this as such, instead devoting their whole attention to criticism, which also constitutes a factor in the continued existence of works, if a lesser one. Yet even though their theoretical work can scarcely have been directed at translation, the great body of translation that they undertook went hand in hand with a feeling of the essence and dignity of the form. That feeling (everything suggests this) is not necessarily at its most powerful in the writer; in fact, in the writer, as writer, it may have very little room. Not even history suggests the conventional prejudice according to which the major translators are writers and minor writers make poor translators. Several great names (Luther, Voss, Schlegel) are incomparably more important as translators than as writers, while other great names (one thinks of Hölderlin and George) cannot, in view of the whole range of their creative activity, be seen as ‘writers’ alone. Even less as ‘translators’. The fact is, translation constituting a form of its own, the task of the translator is also unique, sharply distinguishable from that of the writer.

That task consists in finding in the language into which the work is being translated the intention on the basis of which, in the translation, the echo of the original will be struck. This is a feature of translation that differs entirely from fine writing since the intention of the latter never targets language as such, the totality of language, but purely (and immediately) specific linguistic contexts of meaning. However, translation does not, like writing, see itself as existing inside the internal forest of language itself, so to speak; no, it is from outside that forest, over against it, and without setting foot inside it that translation will summon the original to enter it – and to enter it, moreover, at that exact spot where the echo of the language of the translation is able to meet with a response in a work in the respective foreign language. Not only does the intention of translation concern something different from that of fine writing (namely, a language as a whole from an individual work of art in a foreign language); it is itself different: the writer’s intention is naive, initial, concrete, the translator’s derivative, ultimate, abstract. For the great motive of drawing the multitude of languages together into one true language fills the latter’s work. This, however, is the one in which individual sentences, literatures, judgements admittedly never talk to one another (remaining, as we have seen, dependent on translation) but in which languages themselves, filled out and reconciled in their manner of meaning, attain mutual agreement. If on the other hand there is a language of truth in which the ultimate secrets that keep all thinking busy are stored in a state of relaxed silence, then that language is truth – true language. And it is this language, in whose surmising and describing resides the only perfection the philosopher can aspire to, that is thoroughly concealed in translations. There is no muse of philosophy, nor is there a muse of translation. But philistine, as sentimental artists like to think, they are not. For there is a type of philosophical genius whose most distinctive trait is a nostalgia for the language that manifests itself in translation. ‘Les langues imparfaites en cela que plusieurs, manque la suprême: penser étant écrire sans accessoires, ni chuchotement mais tacite encore l’immortelle parole, la diversité, sur terre, des idiomes empêche personne de proférer les mots qui, sinon se trouveraient, par une frappe unique, elle-même matériellement la vérité’ [Benjamin cites this passage in French]. If what Mallarmé has in mind with these words can be strictly measured by the philosopher, the fact that it contains seeds of such language places translation midway between literature and theory. In terms of its stamp its work is inferior to both, yet it makes no less deep an impression on history.

If the task of the translator is shown in such a light, the more impenetrably obscure do the ways of solving that task threaten to become. In fact, that task (bringing the seed of pure language to maturity in translation) does seem utterly impossible – identifiable in no solution. For is not any such solution undermined as soon as rendering the sense ceases to be crucial? And that, surely, is what (considered from the negative point of view) all the foregoing implies? Fidelity and freedom (the freedom to ‘give the gist’, as it were, and, in thrall to it, fidelity to the text) are the traditional concepts used in any discussion of translations. A theory that seeks, in a translation, something other than rendering the sense would seem to be one that such concepts can no longer serve. Granted, traditional usage always sees these concepts in irreconcilable conflict. Take fidelity, for instance: what actually can that achieve so far as rendering sense is concerned? Fidelity in translating the individual word is scarcely ever able fully to reflect the sense possessed by that word in the original. For that sense is never (judged by its literary signification so far as the original is concerned) entirely comprised in what is meant; in fact it derives that precisely from the way in which what is meant is bound up, in the specific word, with how it is meant [the ‘manner of its meaning’]. This is usually expressed in a formula to the effect that words carry an emotional connotation [Gefühlston, literally: ‘feeling tone’]. As for literalness with regard to syntax, that will see off any ‘rendering the sense’ completely, threatening to lead to downright incomprehensibility. The nineteenth century was presented with overpowering examples of the literal approach in the shape of Hölderlin’s Sophocles translations [into German, of course]. Finally, it stands to reason how much more difficult fidelity in rendering the form will make that of the sense. Therefore, the literalness requirement cannot possibly follow from any interest in preserving the sense. Preserving the sense is far better served (not so creative writing and language, of course) by the licentious freedom of bad translators. Inevitably, then, that requirement (clearly just but very hard to justify) needs to be understood on the basis of more convincing links. As the shards of a pot, let us say, if they are to fit together, must correspond in the tiniest detail without needing to be identical, in the same way the translation, rather than make itself resemble the sense of the original, must lovingly and precisely mimic the original’s manner of meaning in its own language in such a way that, as two shards recognizably form part of one vessel, both it [the translation] and the original become recognizable as forming part of a greater language. For that precise reason it must very largely ignore the object of communicating something (the sense, that is to say), and in this respect the original is essential to it only in so far as, through its existing, the translator and the work of translation have been spared the effort and ordering of what is to be communicated. In the field of translation, too, it is true that , in the beginning was the Word. On the other hand its language can – indeed must – let itself go in relation to sense lest it sound the intentio⁴ of that sense as echo, as reflection, rather than have its own kind of intentio ring out as harmony, as complement to the language in which that sense communicates itself. So it is not, particularly in the era of its creation, the highest praise that can be heaped on a translation that it reads like an original in its own language. No, the importance of fidelity, which literalness guarantees, is precisely that the great yearning for linguistic completion speaks from the pages of the work. A proper translation is transparent, it does not cover up the original, does not stand in its light; instead it permits pure language, as it were reinforced through its own medium, to fall the more fully on the original. What facilitates that most is literalness in the translation of syntax; it is literalness that proves the word, not the sentence, to be the translator’s most basic element. For the sentence is the wall blocking off the language of the original, literalness the arcade.

If fidelity and freedom in translation have always been seen as conflicting tendencies, not even this deeper interpretation of one of them seems to reconcile the two; on the contrary, it appears to deny all right to the other. For what is freedom about if not reflecting the sense [of the original], which ought now to cease calling itself primary? Only if the sense of a linguistic construct may be equated with what it communicates is it left with something that, while very close to it, is yet infinitely remote, hidden among it or, more clearly, broken by it or, more forcefully, beyond any kind of communication – something final, something decisive. Residual within all language and its constructs, aside from what can be communicated, is an element that cannot be communicated, a symbolizing element or a symbolized element, depending on the context in which it is encountered. Symbolizing only in the finite constructs of languages; symbolized, however, in the very way in which languages develop. And the thing that seeks to set itself forth (more: to body itself forth) in the way in which languages develop is the nucleus of pure language itself. However, where that nucleus, albeit hidden and fragmentary, is nevertheless present in life as the thing symbolized itself, it inhabits linguistic constructs only as a symbolizing factor. If the ultimate entity that is pure language itself is associated only with the linguistic element in languages and with the changes this undergoes, in linguistic constructs it is saddled with a weighty, alien sense. Freeing it from this, turning the symbolizing element into the thing symbolized itself, recovering pure language (now given form) for linguistic usage – that is the stupendous and sole ability of translation. It is in such pure language, which no longer seeks to say anything, no longer gives expression to anything, but as expressionless, creative Word is what is meant in every language, that all communication, all sense, and all intention finally reach a stratum at which they are destined to cease. And it is on the basis of that stratum that freedom of translation is confirmed as a new and higher right. Not from the sense of communication, to emancipate from which is precisely the task of fidelity, does it have its being. No, for the sake of pure language freedom proves itself in terms of its own. To redeem that pure language that is banished into otherness in one’s own language, to release the language held prisoner in the work by rewriting it – that is the task of the translator. For its sake the translator breaks down crumbling barriers in his own language: Luther, Voss, Hölderlin, George all broadened the bounds of German.

Seen from this angle, what significance for the relationship between translation and original is left to sense may be captured in a comparison. As the tangent touches the circle fleetingly and only at a single point, and as that contact (though not that point) is prescribed for it by the law in accordance with which it continues its straight course into infinity, so the translation touches the original fleetingly and only at that vanishingly small point of sense before (obedient to the law of fidelity) pursuing its unique course in the freedom of linguistic usage. The true significance of that freedom was characterized by Rudolf Pannwitz (albeit without naming it or substantiating it) in remarks in his [1917] book die krisis der europäischen kultur. Together with what Goethe says in the notes to his Divan,⁵ these are probably by far the best things published in Germany on translation theory. Pannwitz writes:⁶

our translations including the finest of them proceed from a false basic principle seeking to germanize indian greek english instead of indianizing greekifying englishing german, they have far more respect for their own linguistic customs than for the spirit of the foreign work […] the fundamental mistake of the person translating is to set the fortuitous state of his own language in stone instead of letting the foreign language shift it by force, particularly when translating from an extremely remote language he must push his way back to the ultimate elements of language itself where word image tone merge into one he must widen and deepen his language with the foreign language people have no idea how far this is possible the extent to which every language is capable of changing language distinguishes itself from language almost like dialect from dialect but not if they are taken too lightly in fact only if they are taken with sufficient weight.

How well a translation can match the essence of this form will be determined objectively by the translatability of the original. The less value and dignity the latter’s language possesses, the more it constitutes communication, the less of it translation is able to recover, until the total predominance of that sense, far from being the lever of a formally perfect translation, makes this impossible. The more formal a work is, the more it will still, even in the most fleeting brush with its sense, be translatable. This applies only as regards originals, of course. Translations, on the other hand, prove untranslatable not because of the seriousness but because of the excessive superficiality with which sense attaches to them. Of this (as in every other essential respect) Hölderlin’s translations, particularly those of the two Sophoclean tragedies, stand as confirmation. In them the harmony of languages is so deep that only as an Aeolian harp is brushed by the wind is sense brushed by language. Hölderlin’s translations are archetypes of their form; they also relate to the most perfect German-language versions of their texts as archetype [Urbild] to model [Vorbild], as a comparison of Hölderlin’s and Borchardt’s translations of the third Pythian ode of Pindar will show. For that very reason, they more than others are infected by the immense, original danger of all translation: that the doors of so expanded and highly structured a language slam shut, imprisoning the translator in silence. The Sophocles translations were Hölderlin’s final work. In them, sense plunges from abyss to abyss until it threatens to disappear altogether into bottomless linguistic depths. There is, however, a handhold. But the only text providing this is holy scripture, where sense has ceased to be the watershed for the twin streams of language and revelation. Where a text in its literalness belongs directly, without the intermediary of sense, to true language, to truth or doctrine, it is quintessentially translatable. No longer for its own sake, of course, but purely for that of languages. As regards that text, such boundless trust is required of the translation that, with as little tension as language and revelation in the text, in the translation literalness and freedom must be united in the form of the interlinear version. For to some extent all great writings (but sacred ones supremely so) carry their own virtual translation between the lines. The interlinear version of holy scripture is the archetype or ideal of all translation.

[1923]