7/30/2019 de-man

1/26

PAUL DE MAN

"Conclusions" on Walter Benjamin's"The Task of the Translator"Messenger Lecture,Cornell University, March 4, 1983*

Editorial NoteWhat appears here is an edited transcript of the last of six MessengerLectures delivered at Cornell in February and March of 1983.The textis based on a collation of three sets of tape recordings, supplementedwith eight pages of manuscript notes. Aside from differences in detail,formulation, and emphasis, the notes diverge significantly from thetapes only on the last sheet, where de Man wrote: "Im Anfang war dasWort und $as Wort war bei Gott/Dasselbe war bei Gott/ohne Das-selbe" (the ast two words lined out)-the beginning of Luther's trans-lation of St. John's gospel, which Benjamin quotes in Greek and towhich de Man made reference in the question session following theCornell lecture. This text retains traces of the context in which thelecture was delivered, notably in references to the three preceding let-.tures.Though the task of the transcriber-to give to an unwritten textthe afterlife of canonicity-may be undertaken only by suspendingthe ideal of fidelity that underwrites it, I have tried wherever possibleto resist the necessity of fixing or immobilizing passages which ap-peared to be still underway toward formulation. De Man's sometimesunnaturalized English has been preserved, with the exception of a fewmodifications attempted for the sake of coherence. Some sentences,and a few paragraphs, had to be rearranged. Solecisms and redundan-

'From Yale French Studie s 69 (1985):Th e Lesson o f Paul d e Man.'* Hegel on the Sublime" (inMark Krupuick, ed., Disp lacem ent: Derrida and After,Bloo mingto n: Univ ersity of Indiana Press, 19831, "Ph eno me nali ty and Materiality i nKant" (i n Shapiro and Sica, Hermen eutics: Quest ions and Prospects , Amh erst: Univer-sity of M assa chu setts Press, 19831, afid "Kant and S chiller" (un pu blish ed).YES 9 7 , 5 0 Years o f Yale French Studie s, Part 2, ed. Porter and W aters, O 2000by Yale University.

7/30/2019 de-man

2/26

PAUL DE MAN 11cies have been retained, however, where the possibility of foreground-ing a gap between oral performance and printed text seemed to out-weigh the likelihood of inconvenience; in this way I have tried totransmit some of the burden and risk of reconstruction on to thereader. Omissions and emendations are intended to conform to thisprinciple. I have punctuated less with an eye to correct usage thanwith the aim of remaining faithful to the tentative nature of the act oftranscription. Here it was my intention to reproduce the pace of oraldelivery and to close off as few readings as possible, even when leavingambiguities open may have been less true to de Man's intent than to acertain reluctance to compromise the instability of this artifact. Para-graphing generally follows de Man's oral pauses and the repetitions ofthesis statements with which he seemed to demarcate articulations inhis argument; such breaks are to an extent reflected in the manuscriptoutline. Except for a few passages in which de Man adopts HarryZohn's translation, quotations in this text reproduce de Man's ownimpromptu translations, which sometimes bear litt le resemblance tothe available English translations cited in my notes.The article is printed with the kind permission of the University ofMinnesota Press to whom it had been promised. -William D. Jewett

I at first thought of leaving this last session open for conclusions anddiscussion; I still hope for the discussion, but I have given up on theconclusions. It seemed to me best, rather than trying to conclude(which s always a terrible anticlimax), ust to repeat once more what Ihave been saying since the beginning, using another text in order tohave still another version, another formulation of some of the ques-tions with which we have been concerned throughout this series. Itseemed to me that this text by Benjamin on "The Task of the Transla-tor" is a text that is very well known, both in the sense that i t is verywidely circulated, and in the sense that in the profession you are no-body unless you have said something about this text. Since probablymost of us have tried to say something about it, let me see what I cando, and since some of you may be well ahead of me, I look forward tothe questions or suggestions you may have. So, far from concluding orfrom making very general statements, I want to stay pretty close tothis particular text, and see what comes out. If I say stay close to the

7/30/2019 de-man

3/26

12 Yale French Stud iestext, since it is a text on translation, I will need (and that is why I haveall these books) translations of this text; because if you have a textwhich says it is impossible to translate, it is very nice to see what hap-pens when that text gets translated. And the translations confirm,brilliantly, beyond any expectations which I may have had, that it isimpossible to translate, as you will see in a moment.Nevertheless, I have placed this within a kind of framework, aframework which is historical. Since the problems of history havecome up frequently, I thought it would be good to situate it within ahistorical or pseudohistorical framework, and then to move on fromthere. Therefore I start out with a recurrent problem in history and his-toriography, which is the problem of modernity. I use as an introduc-tion into this a litt le essay by the German philosopher Gadamer, whoin a collection called Asp ek te der Moderni ta t wrote, many years ago,interesting articles called "Die philosophischen Grundlagen deszwanzigsten Jahrhunderts" ["The Philosophical Foundations of theTwentieth Century"]. Gadamer asks the somewhat naive but certainlyrelevant question, whether what is being done in philosophy in thetwentieth century differs essentially from what was being done before,and if it then makes sense to speak of a modernity in philosophicalspeculation in the twentieth century. He finds as the general theme,the general enterprise of contemporary philosophy, a critical concernwith the concept of the subject. Perhaps one wouldn't say this now,which perhaps dates this piece a little, but it is still relevant. His ques-tion then is whether the way in which the critique of the concept of thesubject is being addressed by present-day philosophy, differs essentiallyfrom the way it had been addressed by the predecessors of contempo-rary philosophy, in German Idealist philosophy-in some of the au-thors with whom we have been concerned, such as Kant, Hegel, andothers. He writes the following sentence, which is our starting point:

Is the critique of the concept of the subject which is being attempted inour century something else, something different from a mere repeti-tion of what had been accomplished by German Idealist philosophy-and, must we not admit, with, in our case, incomparably less power ofabstraction, and without the conceptual strength that characterizedthe earlier movement?

1. The German text, which appeared in Aspek t e der Modern i ta t (Gijttingen: Van-derhoeck & Ruprecht, 1965), 77-100, is most readily available in Gadamer's KleineSchri f ten (Tiibingen: J . C. B. Mohr, 19671, v. 1, 131-48. An English translation may befound in the collection Phi losoph ica l Herm eneu t i cs , trans., David E . Linge, (Berkeley:University of California Press, 1976), 107-29. Cf. Kleine Schri f ten , v. 1, 141; Philosoph-i ca l Hermeneu t i cs , 119.

7/30/2019 de-man

4/26

Is what we are doing just a repetition? And he answers, surprise:"There is not the case." W hat w e are doing really is som ethin g new,something different, and we can lay claim to being modern philoso-phers. He finds three rubrics in wh ich we-contemporary philoso-phers-he, Gadam er-is ahe ad of hi s predecessors, and he cha rac ter-izes these th ree progressions in t er m s of a decreased naivetC. To usnow i t seems, if we look back o n Hegel or Kant, tha t there is a certainnaivetk there which w e have no w grown beyond. H e distinguishes be-twee n th ree types of naivetC, w hic h h e calls Naivitat des Setzens(naivetk of positing ),N aivita t der Reflexion (na ive tk of refle ctio n), andN aiv ita t des Begriffs (naivetC of th e c on ce pt).Very briefly, wh at is m ea nt by th e first, by a c erta in naivetC of posi-tion, is a critique wh ich w e have been able to develop of pure percep-tion an d of pu re decla rative discourse, in relation to th e problem of t h esubject. We are now ahead of Hegel in th at w e kno w better th at thesubject does not dom inate its ow n utterances; we are more aware thatit is naive to assu m e tha t th e subject really controls its o w n discourse;we kn ow thi s is not the case. Yet he qualifies th is one bit: neverth e-less, understanding is available to us to so m e extent, by a herm ene uticprocess i n wh ich understanding, by a historical process, can c atch upw ith the presuppositions it had m ade ab out itself. We get a develop-m en t of G adam er, disciple of Heidegger, of t h e no tio n of a he rm en eu -tic circle, where the subject is blind to it s ow n utterance, but whe renev erthele ss th e reader wh o is aw are of t he historic ity of t ha t blind-ness can recover th e meaning, can recover a certain am ou nt of con trolover the tex t by me ans of th is particular her m en eu tic patter n. Th ism ode l of und erstand ing is ahead of t h e Hegelian mo del exac tly to th esam e exte nt th at on e could say tha t th e herm ene utics of Heidegger areahead of th e he rm en eu tics of Hegel, in Gadam er's sense .

H e the n speaks of t h e naivetk of reflec tion, and develops furt he rwh at is already posited in the first; namely, he asserts th e possibilityno w of a hist ori city of und erstand ing, in a wa y th at is not accessibleto individual self-reflection. It is said th at Hegel, i n a sense, was no thistorical enough, tha t in Hegel it is stil l too m uc h the subject i tselfwhich originates its own understanding, whereas now one is moreaw are of t he difficulty of t h e rela tion ship be tw ee n th e self and its dis-course. W here in the first progression h e refers to Heidegger's co ntr i-bution, here he refers very mu ch to his ow n contribution: historiciz-ing the notio n of understanding, by seeing und erstan ding (a s he laterRezept ionsasthet ik , which comes from Gadamer to a large extent,will develop it ) as a process between auth or and reader in w hich the

7/30/2019 de-man

5/26

14 Yale French Stud iesreader acquires an understanding of the text by becoming aware of thehistoricity of the movement that occurs between the text and him-self. Here Gadamer also makes a claim that something new is goingon nowadays, and indeed, the stress on reception, the stress on read-ing, are characteristics of contemporary theory, and can be claimed tobe new.Finally, he speaks of the naivetC of the concept, in which the prob-lem of the relationship between philosophical discourse and rhetori-cal and other devices which pertain more to the realm of ordinary dis-course or common language were not, with Kant and Hegel, beingexamined critically. We alluded to an example of that yesterday whenKant raises the problem of hypotyposis and invites us to become awareof the use of metaphors in our own philosophical discourse. That typeof question, which at least was mentioned by Kant, and was men-tioned by Hegel much less, is now much more developed. Gadamer'sallusion is to Wittgenstein, and also indirectly to Nietzsche. We nolonger think, says Gadamer, that conceptual and ordinary language areseparable; we now have a concept of the problematics of languagewhich is less naive in that it sees to what extent conceptual philo-sophical language is still dependent on ordinary language, and howclose it is to i t. This is the modernity which he suggests, and which hedetails by these three indications.Now although this is Kantian to some extent in its critical outlook,it is still very much a Hegelian model. The scheme or concept ofmodernity, as the overcoming of a certain nonawareness or naivetk bymeans of a critical negation, by means of a critical examination whichimplies the negation of certain positive relationships and the achiev-ing of a new consciousness, allows for the establishment of a new dis-course which claims to overcome or to renew a certain problematic.This pattern is very traditionally Hegelian, in the sense that the devel-opment of consciousness is always shown as a kind of overcoming of acertain naivetC and a rise of consciousness to another level. It is tradi-tionally Hegelian, which does not mean that it is in Hegel, but it is inHegel the way Hegel is being taught in the schools. Indeed, Gadamerends his piece with a reference to Hegel:

Th e concept of spirit, wh ich Hegel borrowed from the C hristian spiri-tu al tradition, is still th e ground of t he c ritique of t he subject and of t hesubjective spirit tha t appears as th e m ain ta sk of t he post-Hegelian,th at i s to say mo dern, period. Th is conce pt of s pirit (Ge is t ) ,which tran-scends the subjectivity of the ego, finds its true abode in th e phen om e-

7/30/2019 de-man

6/26

non of language, which stands more and more as the center of contem.porary philo~ophy.~Contemporary philosophy is a matter of getting beyond Hegel inHegelian terms, by focusing the Hegelian demarche, the Hegelian di-alectic more specifically on the question of language. That is howmodernity is here defined, as a Hegelianism which has concentratedmore on linguistic dimensions.If we compare the critical, dialectical, nonessentialist (becausepragmatic to some extent, since an allowance is made for commonlanguage) concept of modernity which Gadamer here advances, withBenjamin's text on language in "The Task of the Translator," then atfirst sight, Benjamin would appear as highly regressive. He would ap-pear as messianic, prophetic, religiously messianic, in a way that maywell appear to be a relapse into the naivetk denounced by Gadamerj in-deed, he has been criticized for this. Such a relapse would actually re-turn to a much earlier stage even than that of Kant, Hegel, and idealistphilosophy. The first impression you receive of Benjamin's text is thatof a messianic, prophetic pronouncement, which would be very re-mote from the cold critical spirit which, from Hegel to Gadamer, isheld up as the spirit of modernity. Indeed, as you read this text, youwill have been struck by the messianic tone, by a figure of the poet asan almost sacred figure, as a figure which echoes sacred language. Allreferences to particular poets in the text put this much in evidence.The poets who are being mentioned are poets one associates with asacerdotal, an almost priestlike, spiritual function of poetry: this istrue of Holderlin, of George, and of MallarmC, all of whom are verymuch present in the essay.(SinceI mention George, one is aware of the presence of George-aname which has now lost much of its significance, but which at thattime in Germany was still considered the most important, centralpoet, although in 1923 or 1924 when this was written this was alreadygetting toward its end. For example, Benjamin quotes Pannwitz, a dis-ciple of George, at the end of the text. And he refers to George in a rel-evant way; in George there was a claim made for the poet again assome kind of prophet, as a kind of messianic figure-George doesn'tkid around with that, he sees himself at least as Virgil and Dante com-bined into one, with still quite a bit added to i t if necessary-thereforehe has a highly exalted notion of the role of the poet, and incidentally

2. Cf. Kleine Schriften, v. 1,p. 148;Philosophical Hermeneutics, p. 128.

7/30/2019 de-man

7/26

16 Yale French Studiesof himself, and of the benefits that go with it . But this tone hangs overthe German academic discourse, and over a certain concept of poetry,which were then current. There are many echoes of it in the way Ben-jamin approaches the problem, at least seen superficially. The same istrue of references to Holderlin, who at that time was a discovery ofGeorge and of his group, where you find a certain messianic, spiritualconcept of Holderlin. Many echoes of this are still to be found in Hei-degger, who after all dedicated his commentaries on Holderlin to Nor-bert von Hellingrath, who was a disciple of George and a member ofthe George circle, and who was, as you know, the first editor of Holder-lin. I sketch in this little piece of background-it may be familiar toyou, i t may be entirely redundant-to show that the mood, the atmo-sphere in which this essay was written is one in which the notion ofthe poetic as the sacred, as the language of the sacred, the figure of thepoet as somehow a sacred figure, is common, and is frequent.)

It is not just in the form of echoes that this is present in Benjamin,it almost seems to have been part of the statement itself. This notionof poetry as the sacred, ineffable language finds perhaps it s extremeform already from the beginning, in the categorical way in which Ben-jamin dismisses any notion of poetry as being oriented in any sense,toward an audience or a reader. This passage has provoked the ire ofthe defenders of Rezeptionsasthetik, who analyze the problem of po-etic interpretation from the perspective of the reader-Stanley Fish orRiffaterre in this country follow that line to some extent, but i t is ofcourse Jauss and his disciples who do this the most. For them, a sen-tence like the one which begins this essay is absolutely scandalous.Benjamin begins the essay by saying:

In the appreciation of a work of art or an art form, consideration of thereceiver never proves fruitful. Not only is any reference to a certainpublic or its representatives misleading, but even the concept of an"ideal" receiver is detrimental in the theoretical consideration of art,since all it posits is the existence and nature of man as such. Art, in thesame way, posits man's physical and spiritual existence, but in none ofits works is it concerned with his response. No poem is intended forthe reader, no picture for the beholder, no symphony for the li~ tener .~

3. Walter Benjamin, "The Task of th e Translator," in Illumin ation s, trans., Harry Zohn,(NewYork: Schock en Books, 1969),69. Qu otatio ns from the French translation of M auricede Ganclillac are take n from Walter Benjamin, Oeuvres (Paris: Editions Denoel, 1971).Pagenum bers referring to either of these versio/ns are given in parentheses; tran slation s not iden-tified wit h a page numb er are the autho r's. Page num bers supplied wit h qu otations in G er-m an refer to the paperback Illumin ationen (Frankfurt, Suhrkamp, 2d ed., 1980).

7/30/2019 de-man

8/26

17AUL DE MANHe couldn't be more categorical tha n in this assertion a t the beginningof th e essay. You can see how this wou ld have throw n th em in to aslight panic in K onstanz, a panic wi th which they deal by saying tha tthi s is an essentialist theory of a rt, tha t this stress on the au thor a t theexpense of th e reader is pre-Kantian, since already Kant had given t hereader, the receptor, the beholder an im porta nt role, m ore im porta nttha n th e author's. Th is the n is held up as an ex am ple of the regressionto a messianic con ception of p oetry wh ich w ould be religious in th ewrong sense, and it is very m uc h atta cke d for that reason.But on th e oth er hand, B enjamin is also frequently praised as theone wh o has return ed t he dimen sion of the sacred to literary language,and wh o has th us overcome, or at least considerably refined, th e secu-lar historicity of lite ratu re on w hic h the noti on of m od ern ity depends.If one can t h in k of m od ern ity as it is described by G adam er as a loss ofth e sacred , as a loss of a certa in ty pe of poet ic experience, as it s re-placement by a secular historicism wh ich loses contact with wh at wasoriginally essential, th en one can praise Benjamin for having re-estab -lished the contact w ith w hat had the re been forgotten. Even in H aber-ma s there are statem ents in th at direction. But closer to home, an ex-am ple of someb ody wh o reads Benjam in wit h a great deal of subtlety ,w ho is aware of th e com plications, and w ho praises him precisely forthe way in which he combines a complex historical pattern with asense of th e sacred, is Geoffrey Ha rtm an , wh o writes i n one of h is lat -est books as follows:

This chiasmus of hope and catastrophe is what saves hope from beingunmasked as only catastrophe: as an illusion or unsatisfied movementof desire that wrecks everything. The foundation of hope becomes re-membrance; which confirms the function, even the duty of historianand critic. To recall the past is a political act: a "recherche" that in-volves us with images of peculiar power, images that may constrain usto identify with them, that claim the " w e a k Messianic power" in use(Thesis2).These images, split off from their fixed location in history,undo concepts of homogeneous time, flash up into or reconstitute thepresent. "To Robespierre," Benjamin writes, continuing Marx's reflec-tions in Th e Eighteenth Brumaire, "ancient Rome was a past chargedwith the time of now j/etztzeit) which he blasted out of the continuumof history. The French revolution viewed itself as Rome incarnate"(Thesis14).4

4. Geoffrey H. Hartman, Cri t ic i sm in the Wi lderness : Th e S t udy o f L i tera ture Today(N e wHav en: Yale University Press, 1980),78.

7/30/2019 de-man

9/26

18 Yale French Stu die sT he reference here is to historical remembrance, to a historical con-cept wh ich th en dovetails, w hic h injects itself in to an apocalyptic, re-ligious, spiritual concept, thus marrying history with the sacred in away w hich is highly seductive, highly attra ctive. It is certainly highlyattractive to Ha rtm an, and one can understand why, since it gives onebo th t h e language of despair, th e language of nih ilism , wit h th e partic-ular rigor that goes with that; but, a t the sam e time, hope! So you haveit all: you h ave th e critical perception, you h ave th e possibility of car-rying on in apocalyptic tones, you have th e particular eloquence thatcomes w ith that (because one can only really get excited if one writesin a n apocalyptic m ode ); bu t you can still talk i n ter m s of hope, a ndBenjam in wo uld be an exam ple of t hi s com bin atio n of n ihil istic rigorwith sacred revelation. A man who likes a judicious, balanced per-spective on tho se things, l ike Hartma n, has reason to qu ote and to ad-m ire th is possibility in Benjamin. T h e problem of the reception of Ben-jamin centers on this problem of the m essian ic and very frequently itis this text o n " Th e Task of th e Translator" th at is quoted as one of t hemo st characteristic indicators in that direction.We now th en ask the simplest, th e m ost naive, th e m ost li teral ofpossible questions in relation to Benjamin's text, and w e will not getbeyond tha t: wh at does Benjamin say? What does he say, in th e m ostimm ediate sense possible? It seem s absurd to ask a que stion tha t is sosimple, th at seem s to be so unnecessary, because we can certa inly ad-m it tha t am ong li terate people we w ould at leas t have som e m inima lagreement about what is being said here, allowing us then to em broi-der upon this statement, to take positions, discuss, interpret, and soon. But it seem s that, in t he case of thi s text, th is is very difficult to es-tablish. Even the translators, w ho certainly are close to th e text, wh ohad to read it closely to som e extent, don't seem to have th e slightestidea of w ha t Benjamin is saying; so m uc h so that w he n Benjamin sayscerta in things rathe r sim ply in one way-for exam ple he says tha tsom ething is not-the translators, w ho at least know G erm an wellenough to know the difference between som ething is and something isnot, don't see it! and pu t ab solutely and literally the o pposite of w ha tBenjamin has said. Th is is remarkable, because th e tw o translators Ihave-Harry Zo hn, wh o translated th e text in English, and Ma urice deGandillac, w ho translated the text in French-are very good transla -tors, and know Germ an very well. Harry Zohn , you m ay know; M au-rice de Gan dillac is an em ine nt professor of p hilosophy a t th e Univer-sity of Paris, a very learned ma n w ho kn ows Ge rm an very well, an d

7/30/2019 de-man

10/26

PAUL DE MAN 19w ho should be able to tell th e difference betwee n, for example, "Ichgehe nach Paris" and "Ich gehe nich t nac h Paris." It is not m ore diffi-cul t than that , but som ehow he doesn 't get i t .An example wh ich has become fam ous and has an anecdote is thepassage nea r the end of B enjamin's essay, w he re Benjamin say s th efollowing: "Wo der Text unm ittelbar, o hne verm ittelnd en Sinn," andso on, "der Wahrheit oder der Lehre angehort, ist er iibersetzbarschlechthin" (62 ) ."Where the text pertains directly, w itho ut me dia-tion, to the realm of t he tr u th and of dogm a, it is, w ith ou t furthe r ado,translatableu-the text can be translated, schlechthin, so there is noproblem about translating it . Gandillac7-I wo n't co m m en t on this-translates this relatively simple, enunciatory sentence: "La oh letexte, immirdiatement, sans l 'entremise d 'un sens . . . relkve de lavirri te ou de la doctrine, i l est pure m ent et s im plem ent intraduisible"(275)-untranslatable. W hat adds som e comedy to this particular in-stance is tha t Jacques Derrida was doing a sem inar wit h th is particu-lar text in Paris, usin g th e French-Derrida's G erm an is pretty good,bu t he prefers to use the French, and w hen you are a philosopher inFrance you take Gan dillac mo re or less seriously. So Derrida w as bas-ing part of his reading on the "intraduisible," on the untra nsla tability ,un t i l somebody in his seminar (so I'm told) pointed ou t to h im tha tth e correct word was "translatable." I 'm s ure Derrida could explainthat i t was the same-and I me an that in a posit ive sense, i t is thesame, but s ti ll , i t is not the same w ithout som e addit ional explana-t ion. This is an example, and we w il l soon see some other exampleswhich are more germane to the quest ions which we wil l br ing upabout th is text .Why, in th is tex t, to begin w ith, is th e translator t he exemplary fig-ure? Why is the translator held up in relation to the very general ques-tions abou t the natu re of poetic language w hich t he text a sks ?Th e textis a poetics, a theo ry of poetic language, so wh y does Benjam in not goto the poe ts? or to th e reader, possibly; or th e pair poet-reader, as in them ode l of re ceptio n? And since his is so neg ative about t he n otio n of re-ception anyway, wha t m ake s th e essential difference betwe en t he pairauthor-reader and the pair author-translator-since one's first, sim pleimpression wo uld be that the translator is a reader of th e original tex t?There are, to some extent, obvious empirical answers one can give.T he essay was writte n, as you know, as an introd uctio n to Benjamin'sow n transla tion of th e T ableaux paris iens of Baudelaire; it m igh t justbe ou t of m egaloma nia th at h e selects th e figure of t h e translator. But

7/30/2019 de-man

11/26

20 Yale French Studiesthis is not the case. One of the reasons why he takes the translatorrather than the poet is that the translator, per definition, fails. Thetranslator can never do what the original text did. Any translation isalways second in relation to the original, and the translator as such islost from the very beginning. He is per definition underpaid, he is perdefinition overworked, he is per definition the one history will not re-ally retain as an equal, unless he also happens to be a poet, but that isnot always the case. If the text is called "Die Aufgabe des ~ b e r s e t -zers," we have to read this title more or less as a tautology: Aufgabe,task, can also mean the one who has to give up. If you enter the Tour deFrance and you give up, that is the Aufgabe-"er hat aufgegeben," hedoesn't continue in the race anymore. It is in that sense also the defeat,the giving up, of the translator. The translator has to give up in relationto the task of refinding what was there in the original.The question then becomes why this failure with regard to an orig-inal text, to an original poet, is for Benjamin exemplary. The questionalso becomes how the translator differs from the poet; and here Ben-jamin is categorical in asserting that the translator is radically unlike,differs essentially from the poet and from the artist. This is a curiousthing to say, a thing that goes against common sense, because one as-sumes (and obviously it is the case) that some of the qualities neces-sary for a good translator are similar to the qualities necessary for agood poet. This does not mean therefore that they are doing the samething. The assertion is so striking, so shocking in a way, that hereagain the translator (Maurice de Gandillac) does not see it. Benjaminsays (inZohn's translation), "Although translation, unlike art, cannotclaim permanence for its products . . ." (75);Gandillac, the same pas-sage: "Ainsi la traduction, encore qu'elle ne puisse klever une prirten-tion a la durire de ses ouvrages, et en cela elle n'est pas sans ressem-blance avec l'art . . ." (267).The original is absolutely unambiguous"~bersetzungalso, wiewohl sie auf Dauer ihrer Gebilde nichtAnspruch erheben kann und hierin unahnlich der Kunst . . ." (55).Asyou come upon it in a text, the statement is so surprising, goes somuch against common sense, that an intelligent, learned, and carefultranslator cannot see it, cannot see what Benjamin says. It is remark-able. Zohn saw it-don't get the impression that Zohn gets it all rightand Gandillac gets it all wrong-basically Gandillac is a little ahead ofZohn, I think, in the final analysis.At any rate, for Benjamin there is a sharp distinction betweenthem. It is not necessary for good translators to be good poets. Some of

7/30/2019 de-man

12/26

PAULDEMAN 21the best translators-he mentions Voss (translatorof Homer), Luther,and Schlegel-are very poor poets. There are some poets who are alsotranslators: he mentions Holderlin, who translated Sophocles and oth-ers, and George, who translated Baudelaire-Dante also, but primarilyBaudelaire, so Benjamin is close to George. But then, he says, i t is notbecause they are great poets tha t they are great translators, they aregreat poets and they are great translators. They are not purely, as Hei-degger will say of Holderlin, Dichter der Dichter, but they are ~ b e r -setzer der Dichter, they are beyond the poets because they are alsotranslators.

A number of the most eminent ones, such as Luther, Voss, andSchlegel, are incomparably more important as translators than as cre-ative writers; some of the great among them, such as Holderlin andStefan George, cannot be simply subsumed as poets, and quite particu-larly not if we consider them as translators. As translation is a mode ofits own, the task of the translator, too, may be regarded as distinct andclearly differentiated from the task of the poet. (76)Of the differences between the si tuation of the translator and that

of the poet, the first that comes to mind is that the poet has some rela-tionship to meaning, to a statement that is not purely within therealm of language. That is the naivete of the poet, that he has to saysomething, that he has to convey a meaning which does not necessar-ily relate to language. The relationship of the translator to the originalis the relationship between language and language, wherein the prob-lem of meaning or the desire to say something, the need to make astatement, is entirely absent. Translation is a relation from languageto language, not a relation to an extralinguistic meaning that could becopied, paraphrased, or imitated. That is not the case for the poet; po-etry is certainly not paraphrase, clarification, or interpretation, a copyin that sense; and that is already the first difference.

If it is in some fundamental way unlike poetry, what, in Benjamin'stext, does translation resemble? One of the things i t resembles wouldbe philosophy, i n that it is critical, in the same way that philosophy iscritical, of a simple notion of imitation, of philosophical discourse asan Abbi ld (imitation, paraphrase, reproduction) of the real situation.Philosophy is not an imitation of the world as we know it, but it hasanother relationship to that world. Critical philosophy, and the refer-ence would be specifically to Kant again, will be critical in the sameway of the notion of the imitative concept of the world.

7/30/2019 de-man

13/26

22 Yale French StudiesUm das echte Verhaltnis zwischen Original und ~ b e r s e t z u n ~u er-fassen, ist eine Erwagung anzustellen, deren Absicht durchaus denGedankengangen analog ist, in denen die Erkenntniskritik die Un-moglichkeit einer Abbildstheorie zu erweisen hat. (53)In order to seize upon the real relationship between the original and itstranslation, we must start reflection of which the intent is in generalsimilar to the modes of thought by means of which a critical episte-mology-there's Kant, Erkenntniskritik-demonstrates the impossi-bility of a theory or simple imitation.

Kant indee d wo uld be critical of a no tion of a rt as im itat ion ; thi swo uld be true of H egel to som e exten t too, because th ere is precisely acritical element th at intervene s here and which takes this image, thismodel, away, w hic h destroys, undoes th is concept of im ita tio n.Translation is also, says Benjamin, mo re like critic ism or lik e thetheory of literatu re, t ha n li ke poetry itself. It is by defining himself inrelation to Friedrich Schlegel and to Germ an Rom anticism i n generalthat Benjamin establishes this similarity between literary criticism(i n he sense of literary theo ry) and translation; and th is historical ref-erence to the Jena Rom anticism here gives to the notio n of criticismand literary theory a dignity which it does not necessarily normallyhave. Both criticism and translation are caught in th e gesture whichBenjamin calls ironic, a gesture wh ich undoes th e stability of th e orig-inal by giving it a definitive, canonical form in th e tran slation or in thetheorization. In a curious way, transla tion can onizes its ow n versionmore than the original was canonical. T ha t the original was not purelycanonical is clear from th e fact tha t i t demands translation; i t can notbe definitive since it can be translated. But you c ann ot, says Benjamin,translate the translation; once you have a translation you cannottransla te it any more. You can tran slate only an original. T he tran sla-tion canonizes, freezes, an original and show s in the original a mob il-ity, an instability, which at first one did not n otice . Th e act of critical,theoretical reading performed by a critic like Friedrich Schlegel andperformed by literary theo ry in general-by m ea ns of w hic h th e origi-nal work is not imitated or reproduced but is to som e extent put i n mo -tion, de-canonized, questioned in a way which undoes its claim tocanonical authority-is sim ilar to w ha t a translator performs.Finally, translation is like history, a nd th at will be th e m ost diffi-cult th ing to understand. In w hat is the mo st difficult passage in th istext, Benjamin says that i t is like history to t he extent th at history is

7/30/2019 de-man

14/26

not to be understood by analogy with any kind of natural process. Weare not supposed to think of history as ripening, as organic growth, oreven as a dialectic, as anything that resembles a natural process ofgrowth and of movement. We are to think of history rather in the re-verse way; we are to understand natural changes from the perspectiveof history, rather than understand history from the perspective of nat-ural changes. If we want to understand what ripening is, we should un-derstand it from the perspective of historical change. In the same way,the relationship between the translation and the original is not to beunderstood by analogy with natural processes such as resemblance orderivation by formal analogy; rather we are to understand the originalfrom the perspective of the translation. To understand this historicalpattern would be the burden of any reading of this particular text.All these activities that have been mentioned-philosophy as criti-cal epistemology, criticism and literary theory (the way FriedrichSchlegel does it), or history understood as a nonorganic process-arethemselves derived from original activities. Philosophy derives fromperception, but it is unlike perception because it is the critical exami-nation of the truth-claims of perception. Criticism derives from poetrybecause it is inconceivable without the poetry that precedes it. Historyderives from pure action, since it follows necessarily upon acts whichhave already taken place. Because all these activities are derived fromoriginal activities, they are singularly inconclusive, are failed, areaborted in a sense from the start because they are derived and sec-ondary. Yet Benjamin insists that the model of their derivation is notthat of resemblance or of imitation. It is not natural process: the trans-lation does not resemble the original the way the child resembles theparent, nor is it an imitation, a copy, or a paraphrase of the original. Inthat sense, since they are not resemblances, since they are not imita-tions, one would be tempted to say they are not metaphors. The trans-lation is not the metaphor of the original; nevertheless, the Germanword for translation, ubersetzen, means metaphor. ~bersetzenrans-lates exactly the Greek meta-phorein, to move over, ubersetzen, to putacross. ~bersetzen,should say, translates metaphor-which, assertsBenjamin, is not at all the same. They are not metaphors, yet the wordmeans metaphor. The metaphor is not a metaphor, Benjamin is saying.No wonder that translators have difficulty. It is a curious assumptionto say ubersetzen is not metaphorical, ubersetzen is not based on re-semblance, there is no resemblance between the translation and theoriginal. Amazingly paradoxical statement, metaphor is not metaphor.

7/30/2019 de-man

15/26

24 Yale French StudiesAll these activities-critical philosophy, literary theory, history-resemble each other in the fact that they do not resemble that fromwhich they derive. But they are all interlinguistic: they relate to whatin the original belongs to language, and not to meaning as an extralin-

guistic correlate susceptible of paraphrase and imitation. They disar-ticulate, they undo the original, they reveal that the original was al-ways already disarticulated. They reveal that their failure, whichseems to be due to the fact that they are secondary in relation to theoriginal, reveals an essential failure, an essential disarticulationwhich was already there in the original. They kill the original, by dis-covering that the original was already dead. They read the originalfrom the perspective of a pure language (reineSprache),a language thatwould be entirely freed of the illusion of meaning-pure form if youwant; and in doing so they bring to light a dismembrance, a de-canon-ization which was already there in the original from the beginning. Inthe process of translation, as Benjamin understands it-which has lit-tle to do with the empirical act of translating, as all of us practice i t ona daily basis-there is an inherent and particularly threatening danger.The emblem of that danger is Holderlin's translations of Sophocles:Confirmation of this as well as of every other important aspect issupplied by Holderlin's translations, particularly those of the twotragedies of Sophocles. In them the harmony of the languages is so pro-found that sense is touched by language only the way an aeolian harp istouched by the wind. . . .Holderlin's translations in particular are sub-ject to the enormous danger inherent in all translations: the gates of alanguage thus expanded and modified may slam shut and enclose thetranslator with silence. Holderlin's translations from Sophocles werehis last work, in them meaning plunges from abyss to abyss until itthreatens to become lost in the bottomless depths of language. (81-82)

Translation, to the extent that it disarticulates the original, to the ex-tent that it is pure language and is only concerned with language, getsdrawn into what he calls the bottomless depth, something essentiallydestructive, which is in language itself.What translation does, by reference to the fiction or hypothesis of apure language devoid of the burden of meaning, is that it implies, inbringing to light what Benjamin calls "die Wehen des eignenenU-thesuffering of what one thinks of as one's own-the suffering of the orig-inal language. We think we are at ease in our own language, we feel acoziness, a familiarity, a shelter in the language we call our own, inwhich we think that we are not alienated. What the translation reveals

7/30/2019 de-man

16/26

is that this alienation is at its strongest in our relation to our own orig-inal language, that the original language within which we are engagedis disarticulated in a way which imposes upon us a particular alien-ation, a particular suffering. Here too the translators, with consider-able unanimity, cannot see this statement. Benjamin's text is, ". . .dass gerade unter allen Formen ihr als Eigenstes es zufallt, auf jeneNachreife des fremden Wortes, auf die Wehen des eigenen zu merken"(54).The two translators-I guess they didn't correspond with eachother, they did this d'un commun accord-translate Wehen, pains, as"birth pangs," as being particularly the pains of childbirth. Gandillacis very explicit about it, he calls it "les douleurs obstitricales" (266) nthe most literal, clinical way; Zohn says "birth pangs" (73) .Why theydo this is a mystery. Wehen can mean birth pangs, but it does meanany kind of suffering, without necessarily the connotation of birth andrebirth, of resurrection, which would be associated with the notion ofbirth pangs because you suffer in producing something-and this is amagnificent moment, you'd be willing to suffer (especially easy for usto say).Benjamin has just been speaking of the "Nachreife des fremdenWortes," translated by Zohn as "maturing process," which again iswrong. Nachreife is like the German word Spatlese (a particularlygood wine made from the late, rotten grape), it is like Stifter's novelNachsommer ("Indian Summeru)-it has the melancholy, the feelingof slight exhaustion, of life to which you are not entitled, happiness towhich you are not entitled, t ime has passed, and so on. It is associatedwith another word that Benjamin constantly uses, the word iiber-leben, to live beyond your own death in a sense. The translation be-longs not to the life of the original, the original is already dead, but thetranslation belongs to the afterlife of the original, thus assuming andconfirming the death of the original. Nachreife is of the same order, orhas to do with the same; it is by no means a maturing process, it is alooking back on a process of maturity that is finished, and that is nolonger taking place. So if you translate Wehen by "birth pangs," youwould have to translate it by "death pangs" as much as by "birthpangs," and the stress is perhaps more on death than on life.The process of translation, if we can call it a process, is one ofchange and of motion that has the appearance of life, but of life as anafterlife, because translation also reveals the death of the original.Why is this? What are those death pangs, possibly birth pangs, of theoriginal? It is easy to say to some extent what this suffering is not. It iscertainly not subjective pains, some kind of pathos of a self, a kind of

7/30/2019 de-man

17/26

26 Yale French Studiesm anife station of a self-pathos w hic h th e poet w ould have expressed ashis sufferings. Th is is certainly not th e case, because, says Benjamin,th e sufferings th at are here being m entioned are not i n any sense hu -m an . They w ould c ertain ly not be t he sufferings of an individual, or ofa subject. T ha t also is very hard t o see, for th e translators. Zo hn, c on-fronted with that passage (I will stop this game of showing up th etranslators, bu t it is always of som e int ere st), ranslates: "if they are re-ferred exclusively to man" ( 7 0 ) .Benjam in very clearly says, "w en n sienicht . . . auf den Men schen bezogen we rden" (51),if you do not relatethe m to m an. T he stress is precisely t hat th e suffering tha t is me n-tioned, the failure, is not a hu m an failure, it does not refer therefore toany subjective experience. The original is unambiguous in that re-spect. Th is suffering is also not a kin d of h istorical path os, th e pathosthat you heard i n Hartman's reference to Benjamin as the o ne who haddiscovered th e pathos of h istory; it is no t th is pathos of re m em bran ce,or this path etic m ixture of hope and catastrophe an d apocalypse w hic hHa rtm an captures, wh ich is present certainly in Benjamin's tone, bu tnot so m uc h i n what h e says. It is not the p athos of a history, i t is notth e pathos of w ha t in Holderlin is called the "diirftige Ze it" b etw eenth e disappearance of th e gods and th e possible retu rn of th e gods. It isno t th is kind of sac rificial, dialectical, and elegiac gesture, by m ea ns ofwh ich one looks back on th e past as a period tha t is lost, wh ich th engives you th e hope of ano ther futu re th at m ay occur.Th e reasons for this pathos, for thi s Wehen, for this suffering, arespecifically linguistic. T hey are stated by B enjamin wi th considerablelinguistic structura l precision; so m uc h so tha t if you com e to a wordlike "abyss" in the passage about Holderlin, where it is said thatHolderlin tum bles in th e abyss of language, you w ould un dersta nd th eword "abyss" in th e nonpa thetic, technical sense in w hich w e speak ofa mise en abyme structu re, the kind of structu re by m ean s of w hic h itis clear that the text itself b ecom es an exam ple of w ha t i t exemplifies.T he text about translation is itself a translation, and th e untranslata-bility which it m entions about itself inhabits i ts own tex ture and willinhabit anybody wh o in his tu rn will try to translate it , as I a m n o wtrying, and failing, to do. Th e text is untranslatable, it w as untra nslat-able for the translators w ho tried to do it , i t is untranslatable for thecom me ntators who talk about it , i t is an example of w hat i t s tates, i t isa mise en abyme in th e technical sense, a story wit hin the story ofwhat i s i t s own s ta tement .W hat are the linguistic reasons whic h allow Benjamin to speak of a

7/30/2019 de-man

18/26

suffering, of a disarticulation, of a falling apart of any original work, orof any work to the extent that that work is a work of language?On thisBenjamin is very precise, and offers us what amounts in very few linesto an inclusive theory of language. The disjunction is first of all be-tween what he calls "das Gemeinte," what is meant, and the "Art desMeinens," the way in which language means; between logos and lexis,if you want-what a certain statement means, and the way in whichthe statement is meant to mean. Here the difficulties of the translatorsare a little more interesting, because they involve philosophical con-cepts that are of some importance. Gandillac, a philosopher whoknows phenomenology and who writes in a period when phenomenol-ogy is the overriding philosophical pressure in France, translates by"visee intentionelle" (272). The way we would now translate inFrench "das Gemeinte" and "Art des Meinens" would be by the dis-tinctions between vouloir dire and dire: "to mean," "to say." Zohntranslates by "the intended object" and the "mode of intention" (74).There is a phenomenological assumption here, and Gandillac has afootnote which refers to Husserl: both assume that the meaning andthe way in which meaning is produced are intentional acts. But theproblem is precisely that, whereas the meaning-function is certainlyintentional, it is not a priori certain at all that the mode of meaning,the way in which I mean, is intentional in any way. The way in whichI can try to mean is dependent upon linguistic properties that are notonly [not]made by me, because I depend on the language as it exists forthe devices which I will be using, it is as such not made by us as his-torical beings, i t is perhaps not even made by humans at all. Benjaminsays, from the beginning, that i t is not at all certain that language is inany sense human. To equate language with humanity-as Schiller did,as we saw yesterday-is in question. If language is not necessarily hu-man-if we obey the law, if we function within language, and purelyin terms of language-there can be no intent; there may be an intent ofmeaning, but there is no intent in the purely formal way in which wewill use language independently of the sense or the meaning. Thetranslation, which puts intentionality on both sides, both in the act ofmeaning and in the way in which one means, misses a philosophicallyinteresting point-for what is at stake is the possibility of a phenome-nology of language, or of poetic language, the possibility of establish-ing a poetics which would in any sense be a phenomenology of lan-guage.How are we to understand this discrepancy between "das

7/30/2019 de-man

19/26

28 Yale French StudiesGemeinte" and "Art des Meinens," between dire and vouloir dire!Benjamin's example is the German word Brot and the French wordpain. To mean "bread," when I need to name bread, I have the wordBrot, so that the way in which I mean is by using the word Brot. Thetranslation will reveal a fundamental discrepancy between the intentto name Brot and the word Brot itself, in its materiality, as a device ofmeaning. If you hear Brot in this context of Holderlin, who is so oftenmentioned in this text, I hear Brot und Wein necessarily, which is thegreat Holderlin text that is very much present in this-which inFrench becomes Pain et vin. "Pain et vin" is what you get for free in arestaurant, in a cheap restaurant where it is still included, so pain etvin has very different connotations than Brot und Wein. It brings tomind the pain fran~a is, aguette, ficelle, bdtard, all those things-Inow hear in Brot "bastard." This upsets the stability of the quotidian.I was very happy with the word Brot, which I hear as a native becausemy native language is Flemish and we say brood, just as in German,but if I have to think that Brot [brood]and pain are the same thing, Iget very upset. It is all right in English because "bread" is close enoughto Brot [brood],despite the idiom "bread" for money, which has itsproblems. But the stability of my quotidian, of my daily bread, the re-assuring quotidian aspects of the word "bread," daily bread, is upset bythe French word pain. What I mean is upset by the way in which Imean-the way in which it ispain, the phoneme, the term pain,whichhas its set of connotations which take you in a completely different di-rection.

This disjunction is best understood ( to take it to a more familiartheoretical problem) in terms of the difficult relationship between thehermeneutics and the poetics of literature. When you do hermeneu-tics, you are concerned with the meaning of the workj when you do po-etics, you are concerned with the stylistics or with the description ofthe way in which a work means. The question is whether these twoare complementary, whether you can cover the full work by doinghermeneutics and poetics at the same time. The experience of tryingto do this shows that it is not the case. When one tries to achieve thiscomplementarity, the poetics always drops out, and what one alwaysdoes is hermeneutics. One is so attracted by problems of meaning thatit is impossible to do hermeneutics and poetics at the same time. Fromthe moment you start to get involved with problems of meaning, as Iunfortunately tend to do, forget about the poetics. The two are notcomplementary, the two may be mutually exclusive in a certain way,

7/30/2019 de-man

20/26

PAULDE MAN 29and that is part of the problem which Benjamin states, a purely lin-guistic problem.

He states a further version of this when he speaks of a disjunctionbetween the word and the sentence, Wort and Satz. Satz in Germanmeans not just sentence, in the grammatical sense, it means state-ment-Heidegger will speak of Der Satz vom Grund;Satz is the state-ment, the most fundamental statement, meaning-the most mean-ingful word-whereas word is associated by Benjamin with Aussage,the way in which you state, as the apparent agent of the statement.Wort means not only the agent of the statement as a lexical unit, butalso as syntax and as grammar. Id you look at a sentence in terms ofwords, you look at it not just in terms of particular words but also interms of the grammatical relationships between those words. So thequestion of the relationship between word and sentence becomes, forBenjamin, the question of the compatibility between grammar andmeaning. What is being put in question is precisely that compatibility,which we take for granted in a whole series of linguistic investiga-tions. Are grammar (word and syntax) on the one hand, and meaning(as it culminates in the Satz)on the other hand-are they compatiblewith each other? Does the one lead to the other, does the one supportthe other? Benjamin tells us that translation put that conviction inquestion because, he says, from the moment that a translation is reallyliteral, wortlich, word by word, the meaning completely disappears.The example is again Holderlin's translations of Sophocles, which areabsolutely literal, word by word, and which are therefore totally unin-telligiblej what comes out is completely incomprehensible, com-pletely undoes the sentence, the Satz of Sophocles, which is entirelygone. The meaning of the word slips away (aswe saw, a word like Auf-gabe, which means task, also means something completely different,so that the word escapes us), and there is no grammatical way to con-trol this slippage. There is also a complete slippage of the meaningwhen the translator follows the syntax, when he writes literally,wortlich. And to some extent, a translator has to be wortlich, has to beliteral. The problem is best compared to the relationship between theletter and the word; the relationship between the word and sentence islike the relationship between letter and word, namely, the letter iswithout meaning in relation to the word, it is asemos, it is withoutmeaning. When you spell a word you say a certain number of rnean-ingless letters, which then come together in the word, but in each ofthe letters the word is not present. The two are absolutely indepen-

7/30/2019 de-man

21/26

30 Yale French Stu diesden t of each othe r. W hat is being nam ed here as the disjunction be-tween gramm ar and meaning, Wort and Satz, is th e ma teriality of t heletter: the independence, or the way i n which the letter can disrupt theostensibly stable mea ning of a sente nce and introd uce in it a slippageby me ans of w hic h tha t m ean ing disappears, evanesces, and by me ansof wh ich all con trol over th at m eaning is lost.So we have, first , a disjunction in language between the herm eneu-tic and th e poetic, we have a second one between gramm ar and me an-ing, and finally, we w ill have a disjunction, says Benjamin, betwe en thesymbol and what is being symbolized, a disjunction on the level oftropes between th e trope as su ch and mea ning as a totalizing power oftropological subs titution s. The re is a similar and equally radical dis-junction, between w hat tropes (wh ich always imp ly totalization) con-vey in term s of totalization an d wha t the tropes accomp lish take n bythemselves. Tha t seem s to be the m ain difficulty of this particular text,because the text is full of tropes, and it selec ts tropes whic h convey th eillusion of totality. It seem s to relapse i nto the tropological errors thati t denoun ces. Th e text co nsta ntly u ses images of seed, of ripen ing, ofharmony , i t uses the im age of seed and rind ( l'ecorce et le n oy au )-wh ich seem t o be derived from analogies between natu re an d language,whereas the co nstant claim is constantly being m ade tha t there are nosuch analogies. In the sam e way tha t history is not to be understood i nter m s of an analogy with nature, tropes should not be based on resem-blances w ith natu re. But th at is precisely th e difficulty, and the chal-lenge of th is particular tex t. W henever Benjamin uses a trope w hic hseem s to convey a picture of tota l meaning, of com plete adequacy be-tw ee n figure and meaning, a figure of perfect synecdoche in w hic h th epartial trope expresses the to talit y of a meaning, he m anip ulates th e al-lusive context within his work in su ch a way tha t th e traditional sym-bol is displaced in a m ann er th at a cts out th e discrepancy between sym -bol and meaning, rather tha n th e acquiescence between b oth.One strikin g exam ple of t ha t is the im age of t he a mp hora:

Fragments of a vessel which are to be glued together must match oneanother in the smallest details, although they need not be like one an-other. In the same way, a translation, instead of resembling the mean-ing of the original, must lovingly and in detail incorporate the origi-nal's mode of signification, thus making both the original and thetranslation recognizable as fragments of a greater language, just as frag-ments are part of a vessel. For this very reason translation must in largemeasure refrain from wanting to communicate. . . . ( 7 8 )

7/30/2019 de-man

22/26

According to this image, there is an original, pure language, of whichany particular work is only a fragment. That would be fine, providedwe could, through that fragment, find access again to the originalwork. The image is that of a vessel, of which literary work would be apiece, and then the translation is a piece of that. It is admitted that thetranslation is a fragment, but if the translation relates to the original asa fragment relates, if the translation would reconstitute as such theoriginal, then-although it does not resemble it, but matches it per-fectly (as in the word symbolon, which states the matching of twopieces or two fragments)-then we can think of any particular work asbeing a fragment of the pure language, and then indeed Benjamin'sstatement would be a religious statement about the fundamentalunity of language.Benjamin has told us, however, that the symbol and what it sym-bolizes, the trope and what it seems to represent, do not correspond.How is this to be made compatible with a statement like the one madehere? An article by Carol Jacobs called "The Monstrosity of Transla-tion," which appeared in Modern Language Notes, treats this passagein a way which strikes me as exceedingly precise and correct. First, sheis aware of the Kabbalistic meaning of the text, by referring to Ger-shom Scholem, who in writing about this text relates the figure of theangel to the history of the Tikkun of the Lurianic Kabbalah:

Yet at the same time Benjamin has in mind the Kabbalistic concept ofthe Tikkun,the messianic restoration and mending which patches to-gether and restores the original Being of things, shattered and cor-rupted in the "Breaking of Vessels," and also (theoriginal being of)his-tory.

Carol Jacobs comments,Scholem might have turned to "Die Aufgabe des ~bersetzers," herethe image of the broken vessel plays a more direct role. . . . Yetwhereas Zohn suggests that a totality of fragments are brought to-gether, Benjamin insists that the final outcome is still "a brokenpartu5(763,note 9 )

All you have to do, to see that, is translate correctly, instead of trans-lating like Zohn-who made this difficult passage very clear-butwho in the process of making i t clear made it say something com-5. Carol Jacobs, "The Mo nstrosity of T ranslation," Modern Language Notes, v. 90(19751, 763, note 9.

7/30/2019 de-man

23/26

32 Yale French Studiespletely different. Zohn said, "fragments of a vessel which are to beglued together must match one another in the smallest detail." Ben-jamin said, translated by Carol Jacobs word by word: "fragments of avessel, in order to be articulated togetheru-which is much betterthan glued together, which has a totally irrelevant concreteness-"must follow one another in the smallest detailu-which is not at allthe same as match one another. What is already present in this differ-ence is that we have folgen, not gleichen, not to match. We have ametonymic, a successive pattern, in which things follow, rather than ametaphorical unifying pattern in which things become one by resem-blance. They do not match each other, they follow each other; they arealready metonyms and not metaphors; as such they are certainly lessworking toward a convincing tropological eotalization than if we usethe term "match."

But things get even more involved, or more distorted, in what fol-lows.

So, instead of making itself similar to the meaning, to the Sinn of theoriginal, the translation must rather, lovingly and in detail, in its ow~llanguage, form itself according to the manner of meaning (Art desMe inens )of the original, to make both recognizable as the broken partsof the greater language, just as fragments are the broken parts of a ves-sel.

That is entirely different from saying, as Zohn says:in the same way a translation, instead of resembling the meaning ofthe original, must lovingly and in detail incorporate the original'smode of signification, thus making both the original and the transla-tion recognizable as fragments of a greater language, just as fragmentsare part of a vessel.

"Just as fragments are part of a vessel" is a synecdoche; "just as frag-ments," says Benjamin, "are the broken parts of a vessel": as such heis not saying that the fragments constitute a totality, he says the frag-ments are fragments, and that they remain essentially fragmentary.They follow each other up, metonymically, and they will never con-stitute a totality. I'm reminded of an example I heard given by theFrench philosopher Michel Serres-that you find out about fragmentsby doing the dishes: if you break a dish it breaks into fragments, butyou can't break the fragments any more. That's an optimistic, a posi-tive synecdochal view of the problem of fragments, because there thefragments can make up a whole, and you cannot break up the frag-

7/30/2019 de-man

24/26

PAUL DE MAN 33ments. What we have here is an initial fragmentation; any work is to-tally fragmented in relation to this reine Sprache, with which it hasnothing in common, and every translation is a fragment, is breakingthe fragment-so the vessel keeps breaking, constantly-and neverreconstitutes i tj there was no vessel in the first place, or we have noknowledge of the vessel, or no awareness, no access to it, so for all in-tents and purposes there has never been one.Therefore the distinction between symbol and symbolized, thenonadequation of symbol to a shattered symbolized, the nonsymboliccharacter of this adequation, is a version of the others, and indicatesthe unreliability of rhetoric as a system of tropes which would be pro-ductive of a meaning. Meaning is always displaced with regard to themeaning it ideally intended-that meaning is never reached. Ben-jamin approaches the question in terms of the aporia between freedomand faithfulness, the question which haunts the problem of transla-tion. Does translation have to be faithful, or does it have to be free? Forthe sake of the idiomatic relevance of the target language, it has to befree; on the other hand, i t has to be faithful to some extent to the orig-inal. The faithful translation, which is always literal, how can it alsobe free? It can only be free if it reveals the instability of the original,and if i t reveals that instability as the linguistic tension between tropeand meaning. Pure language is perhaps more present in the translationthan in the original, but in the mode of trope. Benjamin, who is talkingabout the inability of trope to be adequate to meaning, constantly usesthe very tropes which seem to postulate the adequation betweenmeaning and trope; but he prevents them in a way, displaces them insuch a way as to put the original in motion, to de-canonize the origi-nal, giving it a movement which is a movement of disintegration, offragmentation. This movement of the original is a wandering, an er-rance, a kind of permanent exile if you wish, but it is not really an ex-ile, for there is no homeland, nothing from which one has been exiled.Least of all is there something like a reine Sprache, a pure language,which does not exist except as a permanent disjunction which inhab-its all languages as such, including and especially the language onecalls one's own. What is to be one's own language is most displaced,the most alienated of all.Now i t is this motion, this errancy of language which neverreaches the mark, which is always displaced in relation to what itmeant to reach, it is this errancy of language, this illusion of a life thatis only an afterlife, that Benjamin calls history. As such, history is not

7/30/2019 de-man

25/26

34 Yale French Stud ieshuman, because it pertains strictly to the order of language; it is notnatural, for the same reason; it is not phenomenal, in the sense that nocognition, no knowledge about man, can be deprived from a historywhich as such is purely a linguistic complication; and it is not reallytemporal either, because the structure that animates it is not a tempo-ral structure. Those disjunctions in language do get expressed by tem-poral metaphors, but they are only metaphors. The dimension of futu-rity, for example, which is present in it, is not temporal, but is thecorrelative of the figural pattern and the disjunctive power which Ben-jamin locates in the structure of language. History, as Benjamin con-ceives it, is certainly not messianic, since it consists in the rigorousseparation and the acting out of the separation of the sacred from thepoetic, the separation of the reine Sprache from poetic language. ReineSprache, the sacred language, has nothing in common with poetic lan-guagej poetic language does not resemble it, poetic language does notdepend on it, poetic language has nothing to do with it. It is within thisnegative knowledge of it s relation to the language of the sacred thatpoetic language initiates. It is, if you want, a necessarily nihilistic mo-ment that is necessary in any understanding of history.Benjamin said this in the clearest of terms, not in this essay but i nanother text called "Theological and Political FragmentIu6 fromwhich I will quote a short passage in conclusion. He said it with allpossible clarity, it seemed to me, until I tried to translate that particu-lar passage, and found that English happens to have a property whichmakes i t impossible to translate. Here is the passage:

Only the messiah himself puts an end to history, in the sense that i tfrees, completely fulfills the relationship of history to the messianic.Therefore, nothing that is truly historical can want to relate by its ownvolition to the messianic. Therefore the kingdom of God is not the te-10s of the dynamics of history, it cannot be posited as it s aim; seen his-torically i t is not its aim but its end.

That is where I have a great deal of trouble with English, because theEnglish word for "aim" can also be "end." You say, "the end and themeans," the aim and the means by which you achieve it. And the En-glish word "end" can mean just as well Ziel as it can mean Ende. Myend, my intention. So that if we want to use that idiom, the translation6. Cf. Illuminationen op. cit., 262. An English translation of the "Theologico-Polit-ical Fragment" may be found in Reflections, Edmund Jephcott, trans., Peter Demetz, ed.

(NewYork: Harcourt Brace Jovanovich, 1978),312-13.

7/30/2019 de-man

26/26

PAUL DE MAN 35the n becomes: "seen historically it is not it s end but its end," i ts ter-mination-it wou ld be perfect English. But it wou ld indicate tha t theseparation which is here undertaken by Benjamin is hidden in th isword "end" in English, wh ich sub stitutes for "aim" th e word "end,"the tw o things which Benjamin asks us to keep rigorously apart.

It cannot be posited as its aim; seen historically it is not its aim but itsend, its termination; therefore the order of the profane cannot be con-structed in terms of the idea of the sacred. Therefore theocracy doesnot have a political but only a religious meaning.

And Ben jamin adds:To have denied the political significance of theocracy, to have deniedthe political significance of the religious, messianic view, to have de-nied this with all desirable intensity is the great merit of Bloch's bookThe Spirit of Utopia.Since w e saw th at w ha t is here called political and historical is dueto purely linguistic reasons, we can i n th is passage replace "political"by llpoetical," in th e sense of a poetics. For we now see that th e no n-messianic, nonsacred, that is the political aspect of history is the re-su lt of th e poetical stru ctu re of language, so th at po litical and poetica l

here are substituted , in op position to the no tion of the sacred. To theextent that such a poetics, such a history, is nonmessianic, not atheocracy bu t a rhetoric, it has no room for certain historical notion ssuch as the n otion of mo dernity, wh ich is always a dialectical, th at isto say an essentially theological notio n. You will remem ber t ha t westarted ou t from Gadamer's claim to m odernity, in term s of a dialecticwh ich was explicitly associated with the word "Spirit," w ith th e spir-ituality in the text of Hegel. We have seen , and it is for m e gratifying tofind, th at Hegel himself-when, in th e sec tion of th e Aesthe t ics onthe sublime, he roots the su blime in this separation betwe en sacredand profane-is actu ally m uc h closer to Benjamin in "Th e Task of theTranslator" than h e is to Gadamer. I will end on that note, and I will beglad to answ er questions if you want. Than k you very mu ch.