John Sallis On Translation 1

On Translation
Studies in Continental Thought

J oh n S a l l i s , ge n e r a l e d i tor
consulting editors
Robert Bernasconi William L. McBride
Rudolph Bernet J. N. Mohanty
John D. Caputo Mary Rawlinson
David Carr Tom Rockmore
Edward S. Casey Calvin O. Schrag
Hubert Dreyfus † Reiner Schürmann
Don Ihde Charles E. Scott
David Farrell Krell Thomas Sheehan
Lenore Langsdorf Robert Sokolowski
Alphonso Lingis Bruce W. Wilshire
David Wood
On Translation
JOHN SALLIS
This book is a publication of
Indiana University Press
601 North Morton Street
Bloomington, IN 47404-3797 USA
http://iupress.indiana.edu
Telephone orders 800-842-6796
Fax orders 812-855-7931
Orders by e-mail iuporder@indiana.edu
© 2002 by John Sallis
All rights reserved
No part of this book may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and recording, or by any infor-
mation storage and retrieval system, without permission in writing from the pub-
lisher. The Association of American University Presses’ Resolution on Permissions
constitutes the only exception to this prohibition.
The paper used in this publication meets the minimum requirements of American
National Standard for Information Sciences—Permanence of Paper for Printed Li-
brary Materials, ANSI Z39.48-1984.
Manufactured in the United States of America
Library of Congress Cataloging-in-Publication Data
Sallis, John, date
On translation / John Sallis.
p. cm. — (Studies in Continental thought)
Includes bibliographical references and index.
ISBN 0-253-34156-6 (alk. paper) — ISBN 0-253-21553-6 (pbk. : alk. paper)
1. Translating and interpreting—Philosophy. 2. Semantics (Philosophy) I. Title.
II. Series.
B840 .S25 2002
107′.2—dc21
2002003001
1 2 3 4 5 07 06 05 04 03 02
For Charles Scott
In Honor of Friendship
“Bless thee, Bottom, bless thee! Thou art translated.”
Contents
Preface xi
1. The Dream of Nontranslation 1
2. Scenes of Translation at Large 21
3. Translation and the Force of Words 46
4. Varieties of Untranslatability 112
Index 123
Preface
Translation goes astray.

It happens almost inevitably. It happens both with the word,
itself a translation, and with the operation (and product) named
translation.
Both go, almost inevitably, astray.
To be sure, the word appears stable and its signi¤cation well-
de¤ned. But nothing could be further from the truth, at least from
a certain truth that would consist in coincidence, even in self-
sameness. The senses that translation would signify and hold in
check prove to multiply, spreading and drifting across an exten-
sive, none-too-stable ¤eld. Translation itself—though in a sense
there is no translation itself—goes likewise astray: gesturing to-
ward the production of sameness (it is called sameness of meaning in
the classical determination of translation), it cannot but breach the
sameness in the foray that must be made into alterity, for instance,
into the alterity of another language. It is as if an ineradicable er-
rancy belonged intrinsically to the very truth of translation.
This text, On Translation, addresses the theme directly. Yet it does
so in such a way as to sustain, almost from the beginning, the
doubling of this theme back upon its very thematization; there
will be indeed multiple respects in which writing on translation
becomes a translating of translation. One could say that translation
is intrinsically double, since its movement is always across a dif-
ferential ¤eld. In any case, in On Translation, translation becomes a
double theme, corresponding to the difference between treating it
in its unrestricted spread, as translation at large (Chapter 2), and
in its restriction to translation of words (Chapter 3). Treatments of
two negative or privative possibilities frame the double treatment
of translation: nontranslation, the impossibility of which is almost
xi
perpetually countered by a dream (Chapter 1); and untranslatability,
which borders on the unspeakable (Chapter 4).
For each of the four themes to which On Translation extends, a
kind of topology is put into play. Each is referred to a locus, a
place, a site, and is interrogated as it takes shape at that site. What
drives the interrogation most forcefully is that each of these places
is also, in a certain way, noplace, nowhere: what occurs in a dream
is nowhere; just as what occurs in the theatre is nowhere, not
even in the theatre where it is played; and just as what one sees
in a painting or hears in music is nowhere. To say nothing of
words, nothing but words, which will at best only open up the
difference—one could call it the difference of all differences—
between words and the place—the place of all places—where
everything comes to pass.
My work on the theme of translation goes back to a lecture pre-

sented in 1998 at the Collegium Phaenomenologicum at the invi-
tation of Günter Figal. Subsequently I had opportunities to develop
various aspects of this theme in lectures at Trinity College (Con-
necticut), Tartu University (Estonia), Vassar College, the University
of Kansas, and Thammasat University (Bangkok). Some aspects
are also developed in a paper, “Hermeneutik der Übersetzung,”
which appeared in Hermeneutische Wege: Hans-Georg Gadamer zum Hund-
ertsten. I am grateful to Nancy Fedrow and Eric Sanday for assis-
tance with production and to my editor and friend Janet Rabino-
witch for her generous encouragement and expert advice.
Tübingen
April 2001
xii Preface
On Translation
One The Dream of Nontranslation
What would it mean not to translate? What would it mean to be-

gin thinking beyond all translation? Or, since one will always al-
ready have begun, what would it mean to begin again, to launch
one’s second sailing, beyond all translation? What would it mean
to have suspended all translational operations, to have suspended
them in the radical sense of having reached a point where even the
traces otherwise left by such operations would ¤nally be effaced
and rendered ineffective? What would it mean, having reached
such a point, to begin again thinking from that point?
If such thinking were possible, if thinking could be situated
beyond all translation, it would still not be capable of eluding dis-
course as such. Neither could it escape entanglement in—to de-
ploy the ancient ¤gure—the fabric of discourse. Its very possibility
would require, then, a discourse itself situated beyond translation,
a discourse free of translation, a nontranslational discourse.
At least such a bond of thinking to discourse is attested by Kant.
Even if there are certain moments in the critical philosophy that
seem resistant to such a bond and even though language seldom
becomes thematic in Kant’s work, it is little wonder that the proj-
ect of establishing the possibility and limits of pure reason should
eventually have come to take up, at least in a supplementary or
marginal way, the question of the delimitation of thought by lan-
guage. It is in a remarkable passage in the Anthropology that Kant
poses the bond of thinking to language. He begins with the most
classical of connections, offering an account that—if one concedes
the need, though perhaps not the necessity, of translation—may be
rendered thus: “All language is signi¤cation of thought, and, on
the other hand, the supreme way of signifying thoughts is through
language, the greatest means of understanding ourselves and oth-
ers.” Then, most remarkably, he outlines a circulation of speech in
and as which thinking comes to pass: “thinking is speaking with
1
ourselves.” Kant could hardly have echoed more clearly—whether
intentionally or not—Socrates’ celebrated declaration to Theaete-
tus: thinking (dianoe.syai) is “discourse [l1gow] that the soul
itself goes through with itself about whatever it is examining. . . .
The soul, as it appears to me, in thinking does nothing other than
converse [dial)gesyai] with itself, asking and answering itself,
and af¤rming and denying.”1 Kant reinforces the point by add-
ing, parenthetically, an exotic example: “The Indians of Tahiti call
thinking: language in the belly.” He then rounds out the passage,
adding the complementary side of the circuit: “So it is also listen-
ing to ourselves inwardly (by reproductive imagination).”2
Kant thus attests that thinking is speaking to oneself and in-
wardly listening, by imagination, to what one says to oneself.
Thinking is imaginally listening to oneself as one speaks to one-
self; it is speaking to oneself as one imaginally listens to oneself
speaking. To the extent that thinking is thus always already drawn
into speech, that is, enacted as speech, thoughts will already have
been voiced (even if in silence), signi¤cations will already have
been translated into words. There will always have commenced a
translation, not between words within the same language or in
different languages, but rather the translation, the circulation, be-
tween thought and speech, between meaning and word, that con-
stitutes the very operation of linguistic signi¤cation.
If thinking is speaking with oneself, then it will never have
outstripped such translation. Thinking will never have been able
to begin beyond such translation. In other words, for thinking
to begin beyond such translation would mean its collapse into a
muteness that could mean nothing at all; incapable of signi¤cation,
it would have ceased—if thinking is speaking to oneself—even to
be thinking. It would have risked a captivation that falls short even
of silence, if indeed silence is possible only for one who can speak.
But, granted the bond of thinking to discourse, the con¤ne-
ment of thinking to translation would seem to have followed only
because of the excessive drift of the sense of translation; once trans-
1. Plato, Theaetetus 189e–190a.

2. Immanuel Kant, Anthropologie in Pragmatischer Hinsicht, in vol. 7 of Gesammelte
Schriften, ed. Preussische Akademie der Wissenschaft (Berlin, 1902–), 192.
2 On Translation
lation is extended to cover the very operation of signi¤cation as
such, it will contaminate, as it were, whatever is bound to dis-
course. On the other hand, one may, with some legitimacy no
doubt, insist on limiting the drift of translation, on restricting
the sense of the word such that it applies only to certain linkages
between signi¤ers in different languages and perhaps also be-
tween signi¤ers in a single language. With this limit in place, one
could then propose a point of nontranslation, a zero-degree point
where discourse would contract into a purely monolingual and
nonmetaphorical operation (assuming that the word metaphorical
can appropriately—or even by a certain drift—cover all cases of
translation within a single language). It is in such a contracted dis-
course that thinking would venture—were it possible—beyond
translation. Set at the point of nontranslation, speaking to oneself
in the pure discourse that becomes possible only at this point,
one’s thinking would—were such possible—launch itself anew. No
longer distracted by diversions from meaning as such, no longer
called upon to detour through a speech even slightly deviant,
thinking in this pure discourse could set about forming its ency-
clopedia. Somewhat as the ancient Babylonians, gifted with a com-
mon language, undisturbed by any foreign tongue—or rather, lip,
as the Hebrew says, broaching metaphor or rather metonymy in
the very name of discourse—set about to build—translating this
nontranslation—“a city and a tower with its top in the sky.”3
Genesis tells of how Yahweh came down to mix up their lan-
guage so that they could no longer understand one another, of how
he scattered them over the face of the earth, bringing to an end
that project so monumental that Hegel took it to mark the begin-
ning of architecture. The Genesis story concludes by telling of the
name given to this place where the mixing up of languages oc-
curred. The name Babel, the proper name that ought properly to be
untranslatable like other proper names, is translated by confusion
into what we translate as the common noun confusion. The confu-
sion of tongues/lips imposes the necessity of translation. It im-
poses also a certain impossibility of translation, or, more precisely,
3. Genesis 11:4.
The Dream of Nontranslation 3

a limit that prevents translation from overcoming the mutual dis-
jointedness of languages and thereby reclaiming in effect the com-
mon language, compensating without residue for the confusion of
tongues.4
Is the thinking that would take place as pure discourse beyond
all translation destined to be undone by a corresponding confu-
sion? Is it, too, properly—and so also improperly—named Babel?
Is it, too, inevitably to be scattered, speaking to itself in metaphors
and in foreign tongues, submitted to the necessity of translation
but also to the impossibility of recovering through translation
what seemed a point of nontranslation? And, if thinking thus
proves incapable of sustaining its conversation with itself other-
wise than by an engagement in translation that opens it to what is
foreign, does this mean that the prospect of nontranslation disap-
pears entirely? Or, even against mounting odds, even in the face
of apparent impossibility—or rather, because of this—does one
still dream of nontranslation?
Still? Even against mounting odds? Against, on the one hand,
theoretical odds? Consider the purview that has opened on the
words that, in the modern European languages, convey the tradi-
tionally basic philosophical concepts and principles, regardless
of whether they are taken over and renewed or put in question
and displaced in their sense. These words (substance, accident, subject,
cause, essence, etc.) are preponderantly translations, if not simply
transliterations, of Latin philosophical terms. Though one can of
course undertake to redetermine the sense of such words so as, in
the end, to render the translational traces no longer effective, there
is no telling how widely the attempt to do so would have to ven-
ture in redetermining words whose values are so thoroughly inter-
related among themselves and in the language as a whole; neither,
then, can there be much assurance of reaching the end. Even if,
instead of trying to neutralize the translation from Latin, one were
to hold back from this translation, attempting—without translat-
ing back from one’s own language—to write philosophically in
4. See my discussion in “Babylonian Captivity,” Research in Phenomenology 22

(1992): 23–31. Also Jacques Derrida, “Des Tours de Babel,” in Psyché: Inventions de
l’autre (Paris: Galilée, 1987), 203–35.
4 On Translation
Latin, as was still possible in eighteenth-century Germany, one’s
text would still be compromised by translation, still bearing, as
it would, traces of the translation from Greek. The compromise
would be all the greater as ever more decisiveness came to be ac-
corded to the translation of Greek philosophical words into the
Latin terms that one would either seek to renew—most likely in
vain—or else simply carry over into the modern languages. In re-
cent engagements with Greek thought, there is much—the mount-
ing odds—to urge according the utmost decisiveness to this trans-
lation. Beginning with Heidegger’s work on the Greeks, it has
become increasingly evident that the translation of Greek philo-
sophical words into Latin terms was anything but a series of substi-
tutions of one signi¤er for another over against a selfsame, persis-
tent meaning that both equally would signify. Even if one were to
insist on a certain reticence, on suspending any totalizing evalua-
tion, there is still no denying the decisiveness of the transition in
which philosophy came to be written in Latin rather than Greek.
That the translation of Spoke,menon into subiectum was a transfor-
mation of sense is con¤rmed perhaps most de¤nitively by the
decisive shift that the very site named by the word eventually un-
derwent so as to allow the word ¤nally to name what modern
philosophy calls the subject. One could—without exaggeration—
speak of an abysmal leap rather than a transition in the case of the
translation of xQra, as this word (that borders on not being a
word) was determined in the Timaeus, its translation by Chalcidius
as locus (hence into English by Thomas Taylor as place). This leap,
not over but away from the abyss, this retreat before the abyss,
named (insofar as it can be named) by xQra, effected a transfor-
mation not just of sense but rather of the very sense of sense.
These examples, which bear upon the very possibility and op-
eration of exemplarity, could be multiplied so as to show again and
again that any history of translation is something quite other than
a story of a series of signi¤ers successively taking over the function
of signifying one and the same signi¤cation. It is precisely because
of the density, manifoldness, and complications of such histories
that access to Greek thought requires the careful and persistent
work of separating the multiple folds and breaking up the sedi-
ment of translational operations. Only by way of a countertrans-

lating that translates back from one’s own language to Greek with-
out translating one’s own language back into—back upon—the
Greek can one, as Heidegger proposed, translate Aristotle back into
Greek, gaining an access to the Greek text that, far from being
independent of translation, would depend on an ever renewed
translational—or countertranslational—strategy that could hardly
be more demanding.
In no case, it seems, could one reach a point either of nontrans-
lation or of such translation/countertranslation as could effec-
tively master the disjunction between languages so as to cancel the
effects of translation and effectively restore perfectly conjoined
points of nontranslation. And yet, in the face of such apparent im-
possibility, even against such mounting odds, does one still dream
of nontranslation?
Still? Even against mounting odds on the practical side? Against,
for instance, the mounting odds accumulating around the phe-
nomenon called—and constituted—by the name globalization? Con-
sider that the very name designates a kind of unlimited translation,
translation of everything across all borders. Rather than effacing
these borders, globalization only renders them permeable; if noth-
ing else does so, the persistence of linguistic differences guaran-
tees that borders, however permeable, remain and remain effec-
tive. As since Babel, translation remains necessary and yet in a
sense impossible, in the sense that its success can never consist in
effacing the borders and establishing a virtual reign of nontrans-
lation. In the wake of globalization, whether one is drawn along
or remains resistant, one is translated ever more into translation,
if not between languages then at least in the negotiations between
cultures that take place with regard to everything that crosses the
borders, whether artworks or customs or agricultural products. In
this very connection there arises the danger of a certain accom-
modation in which one would be prone to fail to recognize the
necessity and the effects of translation. A certain complicity be-
tween the spread of English almost everywhere and the dream of
nontranslation threatens to render translational effects and the
borders to which they attest less and less perceptible. And yet, the
threat has not gone simply unrecognized as the odds have mounted
6 On Translation
against the assumption that speaking English, speaking even in the
style and idiom of an American, is speaking without translation.
Yet the dream, it seems, persists against all odds. Its sequences
unfold, ®owing into one another yet without ever quite cohering.
The dream continues, as in every case, in a way that is neither
simply indifferent to one’s participation nor, on the other hand,
merely dependent on one’s conscious and deliberate intention. It
®ows on, engendering hope, renewing ever again the very will
that, even if by a barely decipherable mechanism, it ful¤lls. Even
as, in what as waking life one would contrast with the dream, one
confronts and perhaps even acknowledges the impossibility of
nontranslation, the dream persists and ®ows over into waking life,
even in a sense takes over the day, instills a certain madness of the
day, this daydream that is con¤nable neither just to the night nor
just to the day.
The dream of nontranslation is no more transparent than in
most other cases. What the dream is about is not manifest in the
dream. What is manifest, what in this sense one actually dreams
about, is something else the formation of which must be exposed
by an appropriate analysis in order to bring to light what the dream
is—as one will say—truly about, its latent content. Certain mani-
fest contents suggest themselves: for instance (and it is most likely
not just an instance), the contents of dreams of complete mas-
tery, of such dreams as those that thoroughly inform the drive of
technology. But only by way of a kind of psychoanalysis (one
twisted free of the theoretical constraints that, for all its force,
limit Freud’s work) could one expose, beneath the technological
mastery manifestly dreamed of, the operation of a will to non-
translation.
Thus does psychoanalysis have a bearing on nontranslation,
just as, conversely, the concept or schema of translation bears on
the very articulation of the framework of psychoanalysis. Refer-
ring to the latent and manifest dream-contents, respectively, as
the dream-thoughts and the dream-content, Freud writes: “The
dream-thoughts and the dream-content lie before us like two pre-
sentations [Darstellungen] of the same content in two different lan-
guages, or rather, the dream-content appears to us as a translation

[Übertragung] of the dream-thoughts into another mode of expres-
sion, and we are supposed to get to know its signs and its laws of
construction by comparing the original and the translation. . . .
The dream-content is given, as it were, in pictographic script
whose signs are to be translated [übertragen] individually into the
language of the dream-thoughts.”5 In short, the production of the
dream-content is a translating of the dream-thoughts, and the task
of psychoanalysis is to countertranslate from the language of the
dream-content back into that of the dream-thoughts. Or rather,
this countertranslating is what Freud in later texts calls the prac-
tical task of psychoanalysis, in distinction from the theoretical
task of explaining the process—the dream-work—by which the
dream-thoughts come to be translated into the dream-content.6
Freud insists that the interpretation of a dream, that is, the counter-
translating of manifest into latent content, can never be declared
¤nished and in itself complete: “actually one is never certain of
having completely interpreted a dream; even when the solution
seems satisfying and without gaps, it remains always possible for
a further meaning to announce itself through the same dream.”7
Thus countertranslation can never be assured of having decisively
undone all that translation would have accomplished; it can never
be certain of having arrived at a point of nontranslation. The pos-
sibility always remains that what seems simply a dream-thought
may prove to be a still unrecognized translation, the meaning of
which—the concealed dream-thought behind which—has still
to be deciphered. Even in the analysis of the dream of nontransla-
tion, there would be no assurance of ever having reached a point
of nontranslation. Regressing from the content, for instance, of
dreams of complete mastery, countertranslation could never it-
self master the latent content exposed, could never display that
5. Sigmund Freud, Die Traumdeutung, vol. 2 of Studienausgabe (Frankfurt a.M.:

S. Fischer, 2000), 280.
6. Neue Folge der Vorlesungen zur Einführung in die Psychoanalyse, in vol. 1 of Studienaus-
gabe, 453. See my discussion of the hermeneutics of the resulting circularity in
“The Logic and Illogic of the Dream-Work,” in Freud’s Unconscious Ontology, ed. Jon
Mills, forthcoming.
7. Freud, Die Traumdeutung, 282.
8 On Translation
content as a signi¤cation free of all translational effects. Not even
the dream-thought of nontranslation could be secured at a point
of nontranslation. To say nothing of the further complication, the
further translation, involved in the very production of the dream-
thoughts: for, granted Freud’s thesis regarding the dream as wish-
ful¤llment, the production of the dream-thoughts would translate
a certain determinate will—would translate it into the dream-
thoughts—and, through the ful¤llment achieved in the dream,
would translate that will, as it were, back to itself.
There are certain rare cases in which no disguise intervenes.8
In these cases the dream-thoughts are not submitted to a distortion
productive of another content that would serve to conceal the
dream-thoughts themselves. Rather, in these cases what one actu-
ally dreams about coincides with what the dream is—as one will
say—truly about.
While still in his youth Leibniz came upon the thought of a
kind of universal alphabet of human knowledge. In De Arte Combi-
natoria, published in 1666, he proposes to search for this universal
alphabet, which, employing mathematical signs, not only would
found all discovery and judgment but also would allow communi-
cation with others independently of their particular word-language.
The text from 1677 generally known as Foundations of a Universal Char-
acteristic9 (though untitled in the original) is perhaps most explicit:
Leibniz writes of “a kind of language or universal characteristic in
which all concepts and things would be brought into the proper
order and with the aid of which it would become possible for vari-
ous peoples to communicate their feelings and thoughts and to
read in their own language what another has written in his lan-
guage.” Such a language—for which “one must go beyond words”
—would compensate de¤nitively for the disorder, the deviations
and noncorrespondences, of word-languages, and it would intro-
duce a reign of unlimited communication, eliminating both the
necessity of translation and the limits that prevent its success from
in effect canceling that necessity. Through the introduction of such
8. See ibid., 136–46.

9. G. W. Leibniz, Die philosophischen Schriften, ed. C. J. Gerhardt (Hildesheim:
Georg Olms, 1965), 7:184ff.

a language, translation—both as the metaphorical transfer that
would compensate for the deviations and noncorrespondences in
a language and as transferal from one language to another—would
be replaced by operations of analysis and synthesis: “Through the
connection of its letters and the analysis of the words composed
from them [words that would be beyond word, replaced by mathe-
matical signs] everything else could be discovered and judged.”
What is perhaps most remarkable about this text is that it remains
purely a proposal, that it offers not the slightest element of such a
language, though Leibniz assures his readers that “it would not re-
quire any more work than that currently applied to encyclope-
dias”; he is con¤dent that the work needed for this new ency-
clopedia, as one could call it, the true encyclopedia that would
involve and require no translation, will require no more than ¤ve
years. Leibniz even reveals the colonialism that is implicated—as
in the Genesis story—with the claim to a universal language: “if
this language is introduced by the missionaries, then also the true
religion, which is most perfectly uni¤able with reason, will be es-
tablished on ¤rm ground.” And yet, this dream of nontranslation,
its sequence unfolding from the vision of universal communica-
tion on through the prospect of converting savages to the true re-
ligion, can persist as such only by concealing something, so that
even in this seemingly most transparent dream of nontranslation
distortion and disguise are not entirely lacking. What must remain
disguised is that, while excluded within this universal language,
translation would nonetheless be required between this language
and the word-languages that humans speak and that—perhaps in-
de¤nitely, in any case well beyond the ¤ve years that writing the
true encyclopedia would require—they will continue to speak. If
all peoples are “to read in their own language what another has
written in his language,” then translation must take place between
the universal language (legible to all) and each people’s own lan-
guage.10
10. In the development of formal logic along lines anticipated by Leibniz’s

proposal, the character of the formalizing transferal from so-called natural lan-
guages to a formal system has been a theme of recurrent debate. A recent dis-
cussion by Robert Wardy appeals to a warning by Christopher Kirwan: “ ‘one
10 On Translation
Leibniz not only envisaged a virtual end of translation but also,
at quite a different level, assigned a positive, critical function to
translation between certain existing languages. In 1670, only four
years after De Arte Combinatoria announced for the ¤rst time the
project of universal language, Leibniz prepared an edition of a
work by the Italian humanist Marius Nizolius that had ¤rst ap-
peared in 1553 under the title On the True Principles of Philosophy, against
Pseudo-Philosophers; for this edition Leibniz wrote an Introduction,
referred to as “On the Philosophical Style of Nizolius.”11 It is in
this Introduction that he discusses the critical function that trans-
lation is capable of performing. The discussion occurs in the con-
text of a broad critical rejection of Scholasticism, though, charac-
teristically and in distinction from most such critics of the time,
Leibniz sets apart those philosophers of sound and useful learn-
ing “who draw from the springs of Aristotle and the ancients
rather than from the cisterns of the Scholastics.” Leibniz’s point of
departure is provided by Nizolius’ insistence that whatever can-
not be named in the vernacular is to be regarded as nonexistent,
¤ctitious, and useless. Leibniz endorses Nizolius’ position in this
respect: whatever cannot be explained in popular terms, in the
words of some living and popular language, is nothing and should
be exorcised from philosophy unless it is something that can be
known by immediate sense experience. Leibniz thus attributes the
might be tempted to think of the whole process of formalising as a kind of

translation from words into symbols; but because the steps in it do not have to
preserve sameness of meaning they are translations of a special kind, and in par-
ticular schematising is far from coming under the ordinary idea of translating.’ ”
Wardy himself writes: “But crucially, I do not regard any such formal represen-
tations as translations of ‘natural’ sentences. Why would one think in the ¤rst place
that depicting ‘all animals are mortal’ as ‘(‘∀x)(Ax ⊃ Mx)’ is on all fours with
translating it as ‘omnia animalia sunt mortalia’?” Taking a more unquali¤edly
negative position even than Kirwan, he goes on to refer to “the delusive con-
ception of ‘translation’ into a formal system” (Robert Wardy, Aristotle in China:
Language, Categories and Translation [Cambridge: Cambridge University Press, 2000],
35–37).
11. Leibniz, Die Philosophischen Schriften, 4:138–76. English citations are adapted
from Leibniz, Philosophical Papers and Letters, ed. Leroy E. Loemker (Dordrecht:
Reidel, 1969), 121–30.

persistence of the exaggerated Scholastic style of philosophy in
Germany to the failure to progress from Latin to German. In con-
trast to their counterparts in England and France, German philoso-
phers of the time had hardly even begun to cultivate philosophy
in their own tongue. In Leibniz’s words, translated from the Latin
in which he himself wrote: “In Germany the Scholastic philoso-
phy is more ¤rmly established because, among other reasons, a late
start was made in philosophizing in German, and even now we
have hardly made an adequate beginning.” What is especially lack-
ing in German philosophy is the exercise of the critical function
that translation of Latin philosophical terms into the popular lan-
guage can perform, the function of testing and explaining—or in
some cases exorcising—the terms of the Latin discourse. Though
in Germany this critical function has hardly begun to be carried
out, Leibniz privileges the German language in this regard: “But I
venture to say that no European language is better suited than Ger-
man for this testing and examination of philosophical doctrines
by a living tongue.” The privilege accorded to German is the result
of its remoteness from Latin. Whereas “many terms of Scholastic
philosophy have been retained in some way in French transla-
tion,” German is so removed from Latin that it will not accept
such terms. Whatever would be translated from Latin into German
would also, because of the extreme difference, be submitted to a
critical test, more critical than in the case of Latinate languages.
Leibniz expresses no concern that such translation might result in
signi¤cational loss; on the contrary, he seems fully con¤dent of the
inevitable gain—if not in signi¤cation, then at least in clarity.
One would presume that this critical function of translation
would indeed have been tacitly operative as Leibniz continued to
write largely in Latin and in French. On the other hand, there are
German texts in which he explicitly begins the task of translating
the Latin philosophical vocabulary into German. In these texts he
not only writes in German but skillfully introduces German terms
to replace those of Latin Scholasticism, for instance, Selbstbestand
for substantia and Unwesen for materia.12 Yet the dream persists; its
12. See Leibniz, Philosophical Papers and Letters, 367.
12 On Translation
thought, nontranslation, will be announced more de¤nitively a
few years later in Foundations of a Universal Characteristic. As long as the
dream persists, the labor of translating Latin philosophical terms
into German can only be—can only have been—regarded as pre-
paratory to the institution of the universal language, as critical,
clarifying work in anticipation of the time—Leibniz thought it
only a few years hence—when philosophy would be in a position
to put aside every particular language, replacing them with a uni-
versal characteristic perfectly matched with thought as such.
Even with Kant, who rarely addresses the theme,13 the dream
13. One text in which this theme does arise, though brie®y, is a short piece
from 1785 entitled On the Injustice of Counterfeiting Books. In this text Kant makes his
case against counterfeiting books by insisting on the distinction between a work
(Werk) and an act (Handlung); indeed, as if to stabilize the distinction and protect
it from such erosion as living and popular language might produce, Kant appeals
to—translates back into—Latin, that is, he translates the distinction as that be-
tween opus and opera. Kant grants that an artwork, because it is an opus and not
an opera, may be copied by anyone who has rightfully acquired it and without
the consent or mention of the producer; the work and its copies may be put up
for sale without the original producer’s having any right to complain of inter-
ference in his affairs. But it is otherwise with the writing of another. In Kant’s
words: “The writing of another is the speech of a person (opera); and whoever
publishes it can speak to the public only in the name of this other and can
say nothing more of himself than that the author makes the following speech
through him (Impensis Bibliopolae),” Kant again, it seems, with the parenthesized
words seeking to stabilize the meaning by translation back into Latin. The rea-
son, then, that books, unlike artworks, are not to be imitated, counterfeited, is
that they are not works (opera) but rather acts (Handlungen) (operae), which can
have their existence only in a person, which belong therefore inalienably to the
person of the author, who has thus an inalienable right always to speak himself
through every other that puts forth the book. On the other hand, if a book is
abridged, augmented, or retouched, then it would be wrong to put it forth in the
name of the author; presenting the alterations in the proper name of the editor
would not be counterfeit. It is likewise, says Kant, with translation: “Translation
into another language cannot be taken to be counterfeit; for it is not the same
speech of the author, though the thoughts may be exactly the same” (Von der
Unrechtmässigkeit des Büchernachdrucks, in vol. 8 of Gesammelte Schriften, 86f.). Kant does
not pursue this question of translation further. He does not, for instance, con-
sider whether it would be imperative to include the name of the original author
along with that of the translator, as indeed one might well suppose on the
ground that thinking is no less act (opera) than is speech. Is one’s own thought

of nontranslation persists virtually without disguise, though lim-
ited to a very speci¤c connection. The connection is the same as
that addressed by Leibniz in his Introduction to Nizolius’ work,
namely, that between Latin (the dead and scholarly language) and
living and popular languages. Yet the translation from the former
to the latter, which to Leibniz constituted a critical gain, appeared
to Kant to produce a certain loss that, in a certain area at least,
needed to be forestalled or reversed. In the establishing of a bul-
wark against certain effects within living and popular languages,
the dream of nontranslation, at least one brief sequence, is renewed.
The pertinent discussion is linked to what Kant writes in the
Critique of Judgment about the role of genius in art: not only that ge-
nius gives the rule to art but also that this talent, as a natural gift
(Naturgabe), is that through which nature gives the rule to art. A
decisive consequence is that the rule issuing from this double giv-
ing (nature’s endowment to the artist, which, in turn, gives the
rule to art) cannot be encapsulated in a formula that could then
serve as a precept for judging and for producing artworks. All that
is available are models, which other artists may imitate in testing
their own talents. Though Kant grants that it is dif¤cult to explain
how such imitation (Nachahmung) is possible, he insists that models
to be thus imitated are the only means of transmitting an artist’s
ideas to posterity. Kant adds a rigorous requirement in the area of
the arts of speech: “in these arts only those models can become
classical that are written in the ancient, dead languages, now pre-
served only as scholarly languages.”14 The basis for this require-
ment is explained in a note added to another, somewhat parallel
passage in which Kant refers to models through imitation of which
one can manifest, though certainly not acquire, taste. The note
reads: “Models of taste in the arts of speech must be composed
in a dead and scholarly language: dead, so that it will not have to
undergo the changes that inevitably affect living ones, whereby noble
not still in some measure one’s own even when it is expressed in another voice,
even in a foreign voice? Or does the requirement that thought be enacted as
speech entail a connection so intimate that the thought, too, would be alienated
in being expressed in an alien voice?
14. Kant, Kritik der Urteilskraft, in vol. 5 of Gesammelte Schriften, §47.
14 On Translation
expressions become ®at, familiar ones archaic, and newly created
ones enter into circulation for only a short while; scholarly, so that
it will have a grammar that is not subject to the whims of fashion
but has its own unchangeable rule.”15 At least in the arts of speech,
in the models to be imitated in poetry and oratory, the living and
popular character of current languages is counterproductive. For
such models dead, scholarly languages that are spoken natively by
no one are superior.
This reversion to the dead and scholarly language, even if only
within a very limited area, is in effect a retreat before a kind of
translation that occurs, not between languages, but within a lan-
guage. It is a kind of diachronic translation that living and popular
languages undergo as if in and of themselves. It is a translation in
which semantic elements undergo displacement (“whereby noble
expressions become ®at, familiar ones archaic, and newly created
ones enter into circulation for only a short while”), as also do syn-
tactical rules (“a grammar . . . subject to the whims of fashion”).
Models of taste in the arts of speech can endure only by being
withdrawn from this translation at work in all living and popu-
lar languages, only by being composed in a dead and scholarly
language.
Such is, then, the reversal that Kant, writing in German, pro-
poses. Whatever advantages might—at least within this very lim-
ited area—be gained by translation to German (as Leibniz, writ-
ing in Latin, proposed) seem completely outweighed by the gain
in defense against the uncontrollable translation at work in all
living and popular languages.
This very translation, the uncontrollable change within any liv-
ing language, is one of the focal points of Benjamin’s re®ection in
“The Task of the Translator.” But for Benjamin the transformation
within living languages is not something that calls for defense and
reversion but rather is the expression of the afterlife of an original
work, the afterlife for which the translations of the work are pri-
mary vehicles. In its translation into another language, a work lives
on. Surviving is undergoing transformation and renewal: “There
15. Ibid., §17.

is a further ripening [Nachreife] even of words with ¤xed meaning.
What may in the author’s time have been the tendency of his po-
etic language can later be worn out; immanent tendencies can
arise anew from what has taken shape. What once sounded fresh
can later sound hackneyed; what at the time sounded normal can
later sound archaic.”16 For Benjamin recognition of these uncon-
trollable transformations, far from prescribing retreat and defense,
suggests that translation cannot be determined on the basis of a
static likeness to the original. Recognition of these transformations
points thus to the redetermination that Benjamin undertakes of
the task of the translator.
But if translation from the dead and scholarly languages to and
into German exposes what is written to the uncontrollable trans-
formations at work in every living and popular language, it also
can bring about an opening to what has been written, to what was
once written in the ancient, classical languages and has retained,
if only covertly, a certain force as origin of Western thought. To
attempt to bring about such an opening is what Hegel proposed
when, in a letter written in 1805 to the classicist J. H. Voss, the
translator of Homer into German, he said of his own endeavor that
he wished “to try to teach philosophy to speak German.”17 The
opening to the ancients would not be a matter primarily of trans-
lating oneself back to the ancients but rather of a return to self
through which the ancients would be appropriated. Hegel explains
why such translation is the “greatest gift that can be made to a
people”: “For a people remains barbarian and does not view what
is excellent within the range of its acquaintance as its own true
property so long as it does not come to know it in its own lan-
guage.” It was precisely for the sake of such appropriation of the
wealth of signi¤cations informing the works of the ancients that
Hegel took up and extended in the direction of Greek antiquity
the program of translation into German that Leibniz had already
broached with critical intent.
16. Walter Benjamin, “Die Aufgabe des Übersetzers,” in Illuminationen (Aus-

gewählte Schriften 1) (Frankfurt a.M.: Suhrkamp, 1977), 53–54.
17. Briefe von und an Hegel, ed. J. Hoffmeister (Hamburg: Felix Meiner, 1952),
1:99f. (#55).
16 On Translation
Across all the differences, abysmal though they be, there is a
certain solidarity between Heidegger and Hegel as regards transla-
tion and tradition. In The Principle of Reason Heidegger writes about
translations that at the appropriate time render a work of poetry
(Dichten) or of thinking. In such cases, says Heidegger, translation
is not only interpretation but Überlieferung, tradition in the sense of
handing-down (not just what is handed down, say, in the sense of
the “content” of tradition, but the handing-down itself, that by
which the “content” of tradition gets handed down from one ep-
och to another). As such, translation “belongs to the innermost
movement of history.”18 Still further: “An essential translation cor-
responds [entspricht] . . . to the way in which a language speaks in
the sending of being [wie im Geschick des Seins eine Sprach spricht].” It is
because such translations inscribe responsively the saying within
the sending of being (the saying of being as cd)a, as \n)rgeia, as
actualitas, . . . as will to power) that they belong to the innermost
movement of history, constituting nodal points, points of jointure,
where tradition (handing down from the sending of being) takes
place. This is the connection in which to consider Heidegger’s pre-
occupation with the transformation wrought by the translation
from Greek to Latin: “Roman thought takes over the Greek words
[die griechische Wörter] without a corresponding, equally originary
experience of what they say, without the Greek word [ohne das
griechische Wort].” 19 This translation inscribes a muted saying, that
of a sending that also decisively withholds. Because it is not just
momentous but decisively epochal, Heidegger declares, in words
otherwise astonishing: “The groundlessness of Western thought
begins with this translation.”
Yet such translational inscription cannot be only a matter of ap-
propriation. Certainly not in the sense proposed by Hegel: in what
Heidegger delimits as the end of philosophy, such appropriation
will already have occurred, indeed with such force that, short of
the most radical measures, we—the we who belong to this closure
—will continue inde¤nitely circulating within the system of sig-
18. Martin Heidegger, Der Satz vom Grund (Pfullingen: Neske, 1957), 164.
19. Heidegger, “Der Ursprung des Kunstwerkes,” in Holzwege, vol. 5 of Gesamt-
ausgabe (Frankfurt a.M.: Vittorio Klostermann, 1977).

ni¤cations that translation will have made our own. What, then,
are the radical measures needed? In the course of a discussion of
how such fundamental words as On and eqnai are to be under-
stood, Heidegger declares precisely what measures are needed:
“that, instead of merely bringing the Greek words into words be-
longing to the German language, we ourselves pass over from our
side into the Greek linguistic domain . . . of On and eqnai.” 20 It is
not only a matter of taking over from the Greeks what they have
thought and said but of setting ourselves back into the Greek, into
the domain from which Greek thinking and saying issued. In his
lecture course Parmenides, in connection with a discussion of the
translation of #l}yeia as Unverborgenheit, Heidegger explains the
translation that in this connection is required of us as transla-
tors: “If we merely replace the Greek #l}yeia with the German
‘Unverborgenheit,’ we are not yet translating. That happens only when
the translating word ‘Unverborgenheit’ translates us into the domain
and mode of experience from out of which the Greeks and in
the present case the originary thinker Parmenides said the word
#l}yeia.” 21 The task of the translator is a certain abandonment,
as is bespoken by the translation within the word Aufgabe itself. The
task of the translator is, ¤rst of all, to be translated into the domain
in which what is to be translated was originarily said.
Yet these translations—translations of Greek by translation back
into Greek—are not the only ones belonging to thinking. Heideg-
ger observes that “we continually translate also our own language,
our native language, into its own words.”22 He declares even that
“translation of one’s own language into its ownmost word” is
more dif¤cult than translating from another language. In any case
he insists that the speaking in which thinking is enacted “is in it-
self a translating,” that in it “an originary translating holds sway.”
Thus does Heidegger declare the utterly translational character of
thinking; in its various modes and directions translation is always
operative, and there is no thinking beyond translation. In the fab-
20. Heidegger, Was Heisst Denken? (Tübingen: Max Niemeyer, 1954), 138.
21. Heidegger, Parmenides, vol. 54 of Gesamtausgabe (Frankfurt a.M.: Vittorio Klos-
termann, 1982), 16.
22. Ibid., 17–18.
18 On Translation
ric of discourse in which thinking enacts itself, there is not a
single thread that has not been spun and woven by translation.
Thus dissolving the dream, Heidegger proclaims a reign of trans-
lation.
The reign of translation is a disjointed gathering that not only
is to be thought but also can be played out in dramatic poetry, in
the theatre. It is played out with an appropriateness that one could
most likely never have imagined possible in A Midsummer Night’s
Dream.23 What the play presents is, above all, a dream of translation,
a dream that is itself enacted as a translation from Athens to the
nearby wood, a dream in which the four lovers, thus translated,
undergo the effects of certain translations and countertranslations
of fancies, a dream in which simple mechanicals are translated
into actors, and one of them, declared and shown to have been
even monstrously translated, is in turn installed in the domain of
the fairy queen, that is, translated from the world of humans (in
and from the gross form he has assumed) into the tiny world of
the fairies. As Titania extends her promise to him, she summons
those slight creatures who serve her:
Therefore go with me.
I’ll give thee fairies to attend on thee;
And they shall fetch thee jewels from the deep,
And sing, while thou on pressed ®owers dost sleep:
And I will purge thy mortal grossness so,
That thou shalt like an airy spirit go.
Peaseblossom! Cobweb! Moth! and Mustardseed!
(III.i.149–55)
The play is a dream, then, of translation and countertranslation.
It is played out as a gathering of translations, which, however
much they seem, with the return to Athens, to have been no more
than
. . . the ¤erce vexation of a dream[,]
(IV.i.68)
23. Citations from William Shakespeare, A Midsummer Night’s Dream, are from
the Arden text edited by Harold F. Brooks (London: Routledge, 1979).

nonetheless leave a certain disjointure. They leave, most notably, a
disjointure of that which disjoints everything (even eventually
from itself ): time itself. For if one counts the days or nights of the
dream, it turns out that at least one has been utterly dreamed
away, the four days spoken of at the outset of the play having been
thus spanned by three days.24 Disjointing time from itself, the
play—in other respects too—forestalls resolution into nontransla-
tion, remains to the very end a play of translation.
24. At the outset of the play Hippolyta says:

Four days will quickly steep themselves in night;
Four nights will quickly dream away the time.
(I.i.7–8)
This is in a sense just what happens: the time is dreamed away, but not in four
nights, not in the four days that will steep themselves in night. A count of the
days comes up short. On the ¤rst day Hermia, Lysander, and Demetrius appear
at Theseus’ court, and Egeus proclaims his interdiction. Lysander and Hermia
plan to ®ee “tomorrow night” (I.i.164; I.i.209). Thus, the second day is the one
on the night of which they ®ee and the various nocturnal happenings take place
in the wood (see, e.g., the reference to “yonder Venus” [III.ii.61]), all four lovers
¤nally lying asleep on the ground. On the third day they are awakened; it is the
day of the new moon, the day on which the wedding of Theseus and Hippolyta
takes place.
A similar contraction seems to occur in Act V, in which the play within the
play allows the time to be dreamed away. At the beginning Theseus refers to
this long age of three hours
(V.i.33)
that has to be worn away before bedtime. Along with Theseus and his company,
we too, the spectators of the play itself, watch the play within the play that is to
help pass the time. One thing is certain: “Pyramus and Thisbe” does not last for
three hours. Yet in his last speech immediately following the play, Theseus pro-
claims:
The iron tongue of midnight hath told twelve.
Lovers, to bed. . . .
...........
This palpable-gross play hath well beguil’d
The heavy gait of night. Sweet friends, to bed.
(V.i.349–50, 353–54)
20 On Translation
Two Scenes of Translation at Large
Writing on translation inevitably becomes entangled. For nothing

quite suf¤ces to keep such writing from getting mixed up with
what is written about. Even if one touches ever so lightly on trans-
lation, one will not be able to prevent a certain adherence of trans-
lation to the discourse on translation. Nothing can quite facilitate
producing a discourse on translation that would remain entirely
distinct from translation. Writing on translation cannot but get en-
tangled in translation; inevitably it gets caught up in translating
translation and to this extent cannot but take for granted precisely
that which the discourse would as such ¤rst delimit. Even simply
to explain what translation is, simply to interpret the meaning of
the word—assuming such simple explanation and interpretation
to be possible—is, in a sense, in a primary sense of the word, to
translate. There are no means by which to delimit a discourse on
translation that would be entirely free of translation. There are no
means by which to limit, as it were, the contamination of such
discourse by translation.
To venture a discourse on translation is thus to invite complica-
tion as such. Translation insinuates itself in discourse on transla-
tion: most directly perhaps when such discourse is made to engage
in translating translation, translating it in the sense of transposing
it into other words, explicating or explaining it. To say nothing of
the fact that translation—the word, which, following a familiar
schema, one will presume to distinguish from the thing itself and
from its meaning—is itself a translation mediated by an extended
and complex history. This history is not simply constituted by a
sequence of words by which an invariant meaning would be pre-
served and reexpressed in successive languages. Not only is the se-
mantic ¤eld of the word bound up with its translation, but, still
more decisively, both word and translation are fundamentally en-
gaged with the metaphysical determination of language as such
21
and with its development beyond—yet on the basis of—the Greek
beginning. As soon as one utters the word translation, one has al-
ready resumed a history of translation and installed what one
would say within the parameters of that history.
Translation is performative. Or at least under a certain provoca-
tion the word can be induced to perform what it says. With only
the slightest energizing of its polysemy, its slippage between vari-
ous meanings will come into play, its slippage across a remarkably
extensive ¤eld of meanings. By way of this slippage, the word
undergoes what also it says. In other words—and here already I
have begun to translate, here already translation insinuates itself
into the discourse on translation, reducing the separation between
saying and said—in other words, translating into other words, the
word translation can thus be induced to undergo translation from
one meaning to another across an extensive semantic ¤eld. In
other words—again I am translating—once its polysemy is ener-
gized, translation is released into a play of translation and endowed
with a semantic mobility that will not prove readily controllable.
One could say that translation is like a translation, that in its per-
formance the word resembles especially a not very good transla-
tion. Its slippage between various meanings is similar to that by
which a not very good translation typically shifts between dis-
parate words or phrases in a less than successful effort to render a
semantically uni¤ed original text. In reading such a translation,
one reaches a certain point of intolerance with respect to the in-
decisive semantic shifts, and then, if one has the competence to
do so, one turns to the original in order to determine more pre-
cisely what is meant. For the word translation there is also a kind of
original. Some measure can be gained against the word’s other-
wise indetermining mobility by turning back to that original, by
untranslating or countertranslating translation back across its his-
tory.
The range of the word translation is enormous. One can speak of
translating words and sentences belonging to one language into
the corresponding words and sentences of another language. Yet
one can also speak of translating ideas into action, hence of trans-
lation as mediating the difference between yevr,a and pr&jiw.
There would seem to be virtually no limit to the extension of
22 On Translation
which the word, by generalization, is capable. From its more lim-
ited senses it readily slides toward the unlimited sense of move-
ment or change across some kind of interval—that is (again, as
almost always, I am translating), it slides toward signifying tran-
sition as such.
This most general, almost unlimitedly general, signi¤cation is
guaranteed by the word’s etymology. The word derives, by way of
the Middle English translaten, from the Latin translatus, which was
used as the past participle of transfero. Composed from the roots
trans (across) and fero (carry, bear), transfero is preserved in the
modern English transfer. Thus regarded, to translate is to transfer, to
carry or bear across some interval. In Latin a translator is one who
carries something over, a transferer. One of the speci¤c things that
can be transferred is meaning, as when the meaning of one word
is transferred to another. If those words belong to different lan-
guages, then there is translation in the speci¤c sense of translating
something in one language into the words of another language.
But there can also be such translation, such transfer of meaning,
within the same language, for example, between what are called
synonyms.
Thus Jakobson differentiates between interlingual translation,
which consists in “an interpretation of verbal signs by means of
some other language,” and intralingual translation, which is “an
interpretation of verbal signs by means of other signs of the same
language.” And though restricting translation to systems of signs
and to translation of verbal signs, thus reducing its generality con-
siderably, Jakobson does at least grant also a third kind of transla-
tion more extensive that the other two; this third kind, termed
intersemiotic translation, consists in “an interpretation of verbal
signs by means of signs of nonverbal sign systems.”1
Under certain conditions translation within the same language
produces what is called a metaphor. Thus the Latin noun translatio
means not only transfer but also metaphor or ¤gure.2 Indeed the
1. Roman Jakobson, Language in Literature, ed. Krystyna Pomorska and Stephen

Rudy (Cambridge, Mass.: Harvard University Press, 1987), 429.
2. Thus Cicero writes of translatio, by which he translates metafor$: “Translatio
occurs when a word applying to one thing is transferred to another because the
Scenes of Translation at Large 23

two words translation and metaphor are, etymologically considered,
virtually identical. Or rather, translation is linked to metaphor by way
of the ancient translation of metaf)rv as transfero. In this combi-
nation met$ has the sense of an interval, a between, an across, and
is thus accurately translated by the Latin trans-, while the Greek
f)rv is, as one says, literally the same word as the Latin fero. Thus,
metaf)rv, to which is linked the noun metafor$, means to carry
or bear across an interval, to transfer. But the Greek words also
carry a second sense that carries these words—translates them—
in another direction, a sense that disturbs the otherwise smooth
transition across an interval: to transfer something is also to change
it, that is—and here I am retracing a series of translations within
translation—to alter it, and hence—this is the second sense—to
pervert it. To translate or metaphorize is to bear something across
an interval at the risk—perhaps even inevitably at the price—of
perverting it.
In this sense the word translation could aptly serve to translate
an exorbitant passage in Nietzsche’s early, unpublished text “On
Truth and Lies in a Nonmoral Sense.”3 I refer to the passage in
which Nietzsche declares that the creators of language do not
aim at any pure truth, at things in themselves, but rather merely
express—Nietzsche uses, without marking any reservations, the
word Ausdruck—the relations of things to humans. For such expres-
sion these creators, according to Nietzsche, lay hold of the boldest
metaphors (die kühnsten Metaphern). Here is Nietzsche’s account—in
translation: “To begin with, a nerve stimulus transferred [über-
tragen] into an image! First metaphor. The image, in turn, copied
[nachgeformt] in a sound! Second metaphor. And each time there is
a complete overleaping of one sphere, right into the middle of an
entirely new and different one.” Thus, both images and words,
both what one sees of things and what one says of them, arise by
similarity seems to justify this transference” (Ad Herennium: De Ratione Dicendi,

IV.xxxiv).
3. Friedrich Nietzsche, “Über Wahrheit und Lüge in einem aussermoralischen
Sinn,” in vol. III 2 of Werke: Kritische Gesamtausgabe, ed. Giorgio Colli and Mazzino
Montinari (Berlin: DeGruyter, 1973), 373.
24 On Translation
transferal across the interval separating one sphere from another,
by a transferal that is a complete overleaping from one to the
other. As such the genesis of perception and of speech consists in
translations that utterly pervert what gets translated. In an impos-
sible declaration, declarable only by an operation of spacing that
keeps it apart from what it declares, Nietzsche declares: “We be-
lieve that we know something about the things themselves when
we speak of trees, colors, snow, and ®owers; and yet we possess
nothing but metaphors of things, which correspond in no way
whatsoever to the original entities.” But such metaphors would
seem to transfer virtually nothing, to carry almost nothing from
one sphere to the other. They would seem to be translations in
which almost nothing—perhaps even, as Nietzsche suggests, noth-
ing at all—gets translated. They would be bad metaphors, it seems,
bad translations, so bad as almost not to be metaphors or transla-
tions at all. And we humans would seem to have—at least are de-
clared to have—in our possession nothing but these bad transla-
tions. In place of things themselves, mistaken indeed for things
themselves, at least for their truthful expression, there would be
available to us humans only bad translations of these things, trans-
lations so bad as not even to be translations of the things them-
selves, translations that would translate next to nothing, transla-
tions that would verge on not being translations at all.
Yet, short of this extreme of abysmal perversion, translation is
otherwise. Short of this limit, its alterity with respect to itself lies
only in its polysemy and mobility. As in the scene in A Midsummer
Night’s Dream in which the mechanicals assemble in the woods out-
side Athens to rehearse the play “Pyramus and Thisbe” that they
are to present at the celebration of Theseus’ marriage to Hippolyta.
Their director Quince the carpenter has called them together at
the green plot that is to be their stage and has steered them through
a most comical discussion of such problems as that of presenting
on stage such fearful things as a lion (Snout the tinker exclaims:
“Will not the ladies be afeard of the lion?”); and the problem of
bringing moonlight into a chamber, since Pyramus and Thisbe
meet by moonlight (Quince instructs: “one must come in with a
bush of thorns and a lantern, and say he comes to dis¤gure or to

present the person of Moonshine”); and the problem of present-
ing the wall separating Pyramus and Thisbe (Bottom the weaver
declares: “Some man or other must present Wall; and let him have
some plaster, or some loam, or some roughcast about him, to sig-
nify wall”) (III.i.26, 55–57, 63–65). All these problems are, need-
less to say, problems of translation.
Just as the rehearsal begins, Puck enters and says to himself:
What hempen homespuns have we swaggering here,
So near the cradle of the Fairy Queen?
What, a play toward? I’ll be an auditor;
An actor too perhaps, if I see cause.
(III.i.73–76)
He is not long assuming a role at the margin of the play by which
to confound it thoroughly and interrupt the rehearsal. Bottom the
weaver, playing Pyramus, recites some lines and then, as Quince
had instructed, exits into the brake at the edge of the green plot.
As he leaves the rustic stage, reciting the promise,
And by and by I will to thee appear,
(III.i.82)
Puck declares to himself:
A stranger Pyramus than e’er played here!
(III.i.83)
Invisible to the humans, Puck follows Bottom into the brake and
there does his mischief. For when Bottom steps out again on stage,
the ass’s head that Puck has placed on him is there for all to see,
except of course for Bottom himself. Quince exclaims:
O monstrous! O strange! We are haunted!
Pray, masters! Fly, masters! Help!
(III.i.99–100)
The mechanicals ®ee at the sight of Bottom with the ass’s head,
Snout the tinker reappearing only long enough to proclaim:
O Bottom, thou are changed! What do I see on thee?
(III.i.109–10)
26 On Translation
Quince then also reappears momentarily, and his exclamation
marks the high point of the scene. He exclaims:
Bless thee, Bottom, bless thee! Thou art translated.

(III.i.113–14)
Bottom’s state is designated by the very same word when in the

next scene Puck reports his mischievous deeds to his master Oberon,
describing how he created utter confusion among the mechanicals
and adding, to Oberon’s enormous delight, that it was precisely
upon this translated creature that Titania came to gaze when she
¤rst awoke:
I led them on in this distracted fear,

And left sweet Pyramus translated there;
When in that moment, so it came to pass,
Titania wak’d, and straightway lov’d an ass.
(III.ii.31–34)
In this scene of translation virtually the entire range of senses

of translation is traced—indeed not just traced but in most respects
presented as in theatre, presented by way of what is presented in
the play itself. Most obtrusively presented is the sense of transla-
tion as change in form, condition, appearance, or substance, trans-
lation as transformation, as transmutation into the form of an
otherwise human character with an ass’s head, in this case, then,
translation as joining together what by nature does not belong to-
gether, translation as monstrous transformation or deformation.
Yet such monsters do not exist. Humans do not in reality have the
heads of asses: the scene is as in a dream or a trance. One will
not readily escape being entranced as one beholds the scene; in-
deed one will likely be so entranced as to suspend, for instance,
one’s awareness of the incongruity between life-size Bottom and
the tiny world of the fairies, as if Titania could, as she promises
Bottom,
. . . Purge thy mortal grossness so,

That thou shalt like an airy spirit go.
(III.i.153–54)

Thus transformed, thus further translated, Bottom is himself en-
tranced, enraptured, as he is led away to the bower of the queen
to be entertained by the fairies and doted on by the queen herself.
Translated into the enchanting world of the fairy queen, its en-
chantment intensi¤ed by the effect of Cupid’s ®ower, Bottom is
doubly enraptured as,
. . . upon this ®owery bed,

(IV.i.1)
he ¤nally declares:
I have an exposition of sleep come upon me.

(IV.i.38)4
Here, then, is another—if archaic—sense of translation manifest

in and around Bottom’s—seemingly multiple—translation: trans-
lation as entrancing, enrapturing, enchanting.
The scene of Bottom’s translation is prescribed in a text, a
script, a scroll. As such it can be submitted to translation in the
sense of reinscription in another language—as in Schlegel’s Ger-
man translation of the play.5 In translating what is said into Ger-
man, Schlegel cannot avoid also translating the names of the char-
acters speaking, even though, insofar as their reference is singular
and they lack meaning, proper names are, strictly speaking, un-
translatable. Yet not just any proper name can be simply carried
over unchanged from one language to another, especially if the
name happens to coincide with or even just to suggest a common
name to which a speci¤c meaning corresponds. Little wonder,
then, that the proper names in Schlegel’s translation appear prob-
lematic, Peter Quince becoming Peter Squenz, the name Robin Goodfellow
disappearing entirely in favor of Puck, which is translated as Droll,
and perhaps most notably the name Nick Bottom being rendered as
Klaus Zettel.
It is remarkable how Schlegel translates the exclamation that
4. Here exposition is a malapropism for disposition.

5. Shakespeare, Ein Sommernachtstraum, trans. August Wilhelm Schlegel, ed.
Dietrich Klose (Stuttgart: Philipp Reclam, 1972).
28 On Translation
forms the climax of the translation scene, Squenz’s exclamation at
the sight of Bottom’s—that is, now Zettel’s—translation:
Gott behüte dich, Zettel! Gott behüte dich!

du bist transferiert.
Here Schlegel translates Shakespeare’s word into German by trans-

lating (or rather, countertranslating) the word translated (as in
“Thou art translated”) back into its Latin original transfero. He
translates the word precisely by countertranslating it.6
Both as inscribed and as enacted, Quince’s exclamation offers
itself to translation in the sense of explicit interpretation or eluci-
dation. Interpretation of this kind is to be distinguished from the
broader, more implicit kind of interpretation that hermeneutics
regards as operative in virtually all translation. One can put what
Quince says into other words, explaining thus—as here—the vari-
ous senses that translation can be taken to have in the exclamation.
One can elucidate the complexity of meaning in the exclamation
by situating it within the scene and the play as a whole. Such in-
terpretation would allow one to show how the declaration that
Bottom is translated has also a transferred or metaphorical sense,
6. On the other hand, Schlegel avoids the translation in the subsequent pas-
sage in which Puck, reporting to Oberon, says:
I led them on in this distracted fear,
And left sweet Pyramus translated there.
(III.ii.31–32)
Schlegel’s translation:
In solcher Angst trieb ich sie weiter fort,
Nur Schätzchen Pyramus verharret dort.
Thus, in this instance translation, that is, translated, simply goes untranslated. The
avoidance is only slightly less in connection with the passage in the opening
scene in which Helena, referring to Hermia, speaks of her desire
. . . to be to you translated.
(I.i.191)
Schlegel translates:
. . . ich liess damit Euch schalten.

that is, how the ¤gure of Bottom functions as a trope within the
play as a whole. In this regard one would need to be attentive to
the way in which, after the opening scene, the very fabric of the
play consists in translations or nontranslations, most frequently
into a dream, into the scene of a dream, so that ¤nally, when he
is relieved of the ass’s head—countertranslated, one could say—
Bottom speaks of nothing but the dream he has had:
I have had a most rare vision. I have had a dream, past the
wit of man to say what dream it was. Man is but an ass if
he go about to expound this dream. . . . It shall be called
“Bottom’s Dream,” because it hath no bottom.
(IV.i.203–206, 214–15)
Schlegel translates the ¤nal sentence as:
sie soll Zettels Traum heissen, weil sie so seltsam angezet-

telt ist.
Here one sees from a particular angle how Schlegel’s translation of

the name Bottom as Zettel is both peculiarly appropriate and yet lim-
ited: for, while it is no doubt the case that such a dream is only
seldom instigated (angezettelt), its having no bottom is something
quite other than the infrequency of its instigation; and while no
doubt it can be said in various regards that Bottom/Zettel is an
instigator, this quality is not among those alluded to in the name
Bottom.
Whatever the connections may be through which an interpre-
tation would contextualize and mark the metaphorical transfers
operative in speci¤c lines of the play such as Quince’s exclamation
that Bottom is translated, it is imperative that the interpretation
take into account the fact that it has to do not just with a written
text but with drama to be performed, with a text that is not pri-
marily to be read but to be performed in the theatre. Precisely be-
cause Bottom’s translation belongs to theatre, it exhibits the fur-
ther sense of translation as change into another medium or sphere,
as with the translation of ideas into action; but here it is a mat-
ter of the more manifold translation of a dramatic script into spec-
tacle, action, speech, and even at certain points music.
30 On Translation
Thus does the scene of Bottom’s translation serve to present a
broad range of senses of translation: as change in form, condition,
appearance, or substance; as entrancing, enrapturing, enchanting;
as reinscription in another language; as interpretation, elucida-
tion; and as change into another medium or sphere. There is at
least one other sense presented: translation as carrying or convey-
ing to heaven, even (in more archaic usage) without death. It
would by no means be entirely out of the question to regard what
follows in the wake of Bottom’s translation, his being carried off—
further translated—to the bower of the fairy queen, as translating
into a comedic presentation this remarkable—if archaic—sense of
translation. But beyond this scene it is unmistakably broached in
another, in the scene in Quince’s house (IV.ii) in which the me-
chanicals lament that because of Bottom’s disappearance and be-
cause there is no replacing him they will not be able to present
their play before the duke and his company. The scene begins in-
deed with Starveling reporting Bottom’s absence:
He cannot be heard of. Out of doubt he is transported.

(IV.ii.3–4)
As to another world. As from this world to the next. As prepared

by the translation he underwent in the forest.7 Thus presenting the
sense of translation as conveyance to a beyond, this scene supple-
ments the scene of Bottom’s translation.
If such scenes from theatre can thus serve to present concretely
the manifold senses of translation, the converse also holds: bring-
ing certain senses of translation to bear on theatre can serve to ex-
pose the very constitution of drama. It is, in part at least, because
translation ¤gures so prominently in A Midsummer Night’s Dream that
this play is, above all, one that folds back upon itself so as to dem-
onstrate dramatically what goes to make up drama as such. Its very
title bespeaks this demonstration. For there is no more appropriate
¤gure of drama than the dream in which images are set forth by
an imagination that, as Theseus proclaims in the play,
7. On the debate concerning the more speci¤c sense to be attributed to the

word transported in this passage, see the note in the Arden edition, p. 100.

. . . gives to airy nothing
A local habitation and a name.
(V.i.16–17)
Set forth by forces that exceed what reason could ever fabricate,
. . . that apprehend
More than cool reason ever comprehends,
(V.i.5–6)
the images that haunt dreams and the theatre enchant, enrapture,
entrance one who gives himself up to them, who lets himself be
translated into the midst of their play. In the theatre virtually
everything conspires to ensure that one dreams on, caught up ec-
statically in the spell of what is said and in the shining of what
appears.
Like a dream, drama has its own time. Its time is such as to
suspend—while also in a sense mimicking—everyday time. Often
at least it is a magical time, like that of the festival.8 It is a time
like that of midsummer night, which the Elizabethans associated
with the midsummer madness brought on after days of intensive
summer heat, a state characterized by a heightened receptiveness
to the delusions of imagination. It is a time in which, as in the
time of a dream, one lets oneself be captivated by the enchant-
ment and magic of the scene; it is a time in which appearance is
neither less nor other than being. Puck’s words, addressed to the
audience at the end of the play, bespeak theatre as dream and its
time as the time of a dream:
If we shadows have offended,

Think but this, and all is mended,
That you have but slumber’d here
While these visions did appear.
And this weak and idle theme,
No more yielding but a dream,
8. Gadamer observes that the suspension of everyday time is a signi¤cant link

of theatre to festival (Hans-Georg Gadamer, “Über die Festlichkeit des Theaters,”
in vol. 8 of Gesammelte Werke [Tübingen: Mohr Siebeck, 1993], 297–98).
32 On Translation
Gentles, do not reprehend:
If you pardon, we will mend.
(V.i.409–16)
If the full bearing of translation on the constitution of drama
is to be gauged, drama must be referred to a delimitation of the
artwork as such as being in itself disclosive. Such delimitation
constitutes a decisive breakthrough in that it breaks with the clas-
sical determination of the artwork as mimetic as well as with the
modern redetermination of art as representation. Though by no
means ventured only by what is usually called hermeneutics, such
a delimitation can be sketched perhaps most economically by de-
veloping two points central to Gadamer’s discussion of the art-
work. (1) An artwork is not a vehicle of mimetic repetition; it does
not operate by presenting in a mimetic image something other
that is already there prior to, independently of, the artwork.9 The
artwork does not re-present something that would already simply
have been present; it is in no sense an allegory, which would say
something in order thereby to bring one to think something else.
Rather than merely setting something in view again, the artwork
brings to view something hitherto unseen, something even un-
foreseen. In this way the artwork intensi¤es one’s vision, lets one
see what one would otherwise not see. (2) And yet, in order to see
that upon which the artwork opens, what is required is not that
one leave the artwork behind for the sake of the vision but rather
that one engage it insistently. Its opening to something unforeseen
takes place in and from the work itself. The vision the artwork
evokes is not a vision that passes beyond the work; rather, “one
can ¤nd what it has to say only in it itself.”10 The truth of the
work, its disclosive opening, is secured and sheltered precisely in
the work itself. In Gadamer’s hardly translatable phrase, the work
achieves “die Bergung von Sinn ins Feste.”11
9. See Gadamer’s discussion of what he terms Nachahmung and contrasts with

both the ancient and the modern determinations of mimesis (ibid., 302).
10. Gadamer, “Die Aktualität des Schönen. Kunst als Spiel, Symbol und Fest,”
in vol. 8 of Gesammelte Werke, 128.
11. Ibid., 125.

But, then, it is imperative that the work be there differently. If
the disclosive opening takes place in and from the work, if, as
Gadamer attests, the work is “in itself there as meaningful,”12 then
everything will hinge precisely on how it is there. The artwork is
not simply present in the way that things of nature are present,
nor even as mere artifacts or other persons are present. The art-
work is there in such a way as to open disclosively as nothing else
does, drawing whoever would engage it into a vision the intensity
of which will never be matched by mere perception of things.
How, then, is its presence so constituted that it can open upon
such a vision and, in the case of the dramatic work, can draw one
into the theatrical dream? The question of the dramatic work is
thus one of presence, a question of the work’s peculiar presence—
not of presence, however, in the sense that went for so long un-
questioned, not of presence as the insuf¤ciently thought being of
things, not, therefore, of presence in the sense effectively decon-
structed through the work of Heidegger and Derrida, but rather
another sense that deconstruction will have served precisely to free.
The presence of the dramatic work is at the same time a pro-
duction of presence. This is what is distinctive about presence in
the theatre: the work is present in and as a production of presence.
The presence thus produced is no sheer selfsame positivity to
which one would simply have added a genesis (as in describing a
t)xnh). In the theatre nothing is present in the way in which
something simply made (a mere artifact) is present once it has
been released from the process of fabrication. Rather, the presence
produced in and as theatre is more like that of the living present,
which requires for its very upsurge a complicity with an imme-
diate past that is radically not present. As time is constituted across
this difference, so is the dramatic work constituted across the dif-
ferences at play in it. Speci¤cally, the dramatic production of pres-
ence takes place as a variety of translations across these ¤elds of
difference. What gives coherence to this variety is the broad sense
of translation as transition across some kind of interval.
Theatre abounds in translation. There is translation, ¤rst of all,
12. Gadamer, “Dichtung und Mimesis,” in vol. 5 of Gesammelte Werke, 85.
34 On Translation
within language, within what is called a language, intralingual
translation, in Jakobson’s terminology. Such translation occurs not
only in the metaphorical transfer of sense, in metaphorical expres-
sions as such, but also in the transitions between one’s own lan-
guage and the language of an author such as Shakespeare. Transla-
tion across the difference separating the language of Elizabethan
England from the English of today (in which various kinds of
differentiations—and not only of dialects—continue to operate)
must be carried out even simply in reading Shakespeare’s text, and
all the more so in entering into a performance of one of his works,
either as actor or as spectator.
As soon as there is performance, another translation will have
come into play, a translation of the language of the text, not into
another language, but into the scenes of the play and the action
carried out on those scenes. Thus enacted, the language of the
written play is translated into a spectacle of action, and the trans-
lation would be across the very difference between word and deed
(l1gow and Trgon), were the deeds not themselves actions only
within the theatrical dream. And yet, the drama is not merely this
spectacle of deeds but rather words and deeds together. But, in
turn, this conjunction is made possible by still another translation,
that of the written text of the play into living, sounding speech.
Furthermore, the dramatic presentation as a whole, that is, as
unity of speech, action, and scene, is carried out precisely as trans-
lation. The actors must translate themselves into the characters,
without of course actually becoming those characters, producing
a presence that belongs to the play but that is not their own, yet
producing it precisely by means of their own presence. None of
the characters depicted in A Midsummer Night’s Dream are actually
present there on stage, none even exist as such and most have
never existed; yet by way of the presentation and the translation
operative in it, by way of the visions thus engaged, it comes about
that, in the words of Quintilian, “things absent are presented . . .
in such a way that they seem actually to be before our very eyes.”13
13. Quintilian, Institutio Oratoria VI.ii.29–30. In this passage Quintilian is refer-

ring to the rhetorical use of phantasies (fantas,ai) or visions (visiones), espe-
cially in gaining power over the moods of the audience. Yet, if the phantastical

Yet, however vivid the seeming, the spectators too must translate
themselves, or, more likely, let themselves be translated. They must
be moved from a state in which they merely perceive the stage and
those appearing upon it to a disposition in which they behold
those persons imaginatively as the characters enacted and regard
what is seen on stage as the scene of action of those characters. The
spectators must give themselves over, must let themselves be en-
tranced, enraptured, enchanted by the presence produced before
their very eyes.
In the theatre nothing that one beholds is actually present as
such. In the theatre one beholds only phantoms, and even what-
ever truth might be accorded to the words they speak is utterly
compromised by their being the words of phantoms about phan-
toms, by their being words addressed to a scene where the pres-
ence of what seems to be is only a presence produced by and dif-
ferent from what actually is. Within the orbit in which being is
nothing other than presence and truth is the presentation of be-
ing, theatre cannot but be declared remote from truth; it is for this
reason that it has, since the ancients, never ceased to provoke sus-
picion. And yet, to everything one beholds in the theatre there be-
longs a presence the intensity of which is rarely, if ever, matched
in the everyday perception of things and persons; and the inten-
sity of this presence is only ampli¤ed by all that one hears, by
the words of phantoms speaking to phantoms. The differentiation
could not be more pronounced: on the one hand, the presence of
what actually is, that is, the presence that is nothing other than
being; on the other hand, the presence that is always other than
being, the presence produced in the theatre. The very texture of
this theatrical presence is translation. In its production everything
is borne across a space of difference. All that makes up theatre—
words, actors, spectators, etc.—all are carried away across various
spaces of difference, along various itineraries of translation.
Such translations are inseparable from the force of imagination,
visions are taken to be such as can also be presented from the dramatic stage,
what he says applies no less to the theatre.
36 On Translation
the force that draws across intervals of difference, the force that is
the very drawing by which something or someone is borne across
such a space. As the spectator’s vision, for instance, must be drawn
beyond the persons and things actually present on stage to the
phantastical scene being presented. If this tractive operation is it-
self pictured as the spectator’s seeing through the one to the other,
then this ¤gure will have begun to communicate with one of
the oldest and most decisive determinations of what comes to be
translated as imagination; the ¤gure will be no less than an inversion
of what the Platonic texts call eckas,a.14
But how is the operation of translation played out concretely in
the play? As A Midsummer Night’s Dream folds back upon itself, how
does it indeed show in its own fabric the constitutive operation of
translation? It does so, above all, by playing out certain kinds of
translation and of nontranslation, by playing between these in the
mode of comedy, or, more speci¤cally, in that mode of comedy in
which (as often in the Platonic dialogues) the very obliviousness
to something serves, as it is played out, to let what has remained
out of account be disclosed all the more forcefully.
Again, then, attention needs to be focused on the play within
the play, not only because comedy reaches here its highest pitch
but also because it is precisely with this turn that the play most
openly folds back disclosively upon itself. Yet the play “Pyramus
and Thisbe,” which the mechanicals ¤nally perform before Theseus
and his company, is ¤rst performed in part, or rather, is rehearsed,
in the scene in which Quince and his troupe assemble in the
woods outside Athens. This is the scene of Bottom’s translation.
But what is decisive is that, with the exception of Bottom’s
very special translation, there is in the case of the mechanicals al-
most no translation. Each of them is who he is, even when he is
supposed to be playing some character—to be translated into a
character—in the play “Pyramus and Thisbe.” This simplicity is
what renders so comical—indeed farcical—both their performance
14. See Force of Imagination: The Sense of the Elemental (Bloomington: Indiana Uni-
versity Press, 2000), 46–52.

of the play and their discussion, just before the rehearsal, of just
how it should be performed. Their simple identity is indicated
even by their names, by the fact that their names say exactly who
they are, say their trades. For—with the exception of their perfor-
mance of “Pyramus and Thisbe”—they are men who have never
done anything but their trades—
Hard-handed men that work in Athens here,
Which never labour’d in their minds till now.
(V.i.72–73)
Thus Bottom the weaver has his name from an object on which
yarn is wound. Quince the carpenter is named after quines, blocks
of wood used for building. The name of Snug the joiner means:
close-¤tting. The name of Flute the bellows mender refers to the
®uted bellows used for church organs. The name of Snout, who is
a tinker, that is, a mender of household utensils, refers to the spout
of a kettle. And the name of Starveling the tailor bespeaks the pro-
verbial thinness of tailors.15
In their simple self-identity, the mechanicals show virtually no
understanding of the translation operative in theatre. Bound by
their respective trades, embodying in themselves the limits de¤ni-
tive of each t)xnh and of t)xnh as such, each of them is bound
to himself. Being who they are and nothing more or other, they
cannot imagine being translated into something other, much less
carry out such a translation by means of imagination. It is this en-
tirely unimaginative, nontranslational outlook that is expressed
when in the woods outside Athens they discuss the play they are
about to rehearse.16 What the discussion reveals is their oblivious-
ness to the translations that actors and spectators must undergo in
theatrical presentation.
On the one hand, they fear that the audience will take what is
seen as actually present rather than as presenting a vision to be
apprehended imaginatively, to be translated into a scene of the
15. See The Complete Signet Classic Shakespeare, ed. Sylvan Barnet (New York: Har-
court, Brace, Jovanovich, 1963), 533 n.
16. See R. W. Dent, “Imagination in A Midsummer Night’s Dream,” Shakespeare Quar-
terly 15 (1964): 126.
38 On Translation
drama. They worry that Theseus and his company will regard
what appears on stage simply as it is, simply as present rather than
as the vehicle for a production of presence. Thus, they set about
devising ways to prevent this from happening. Bottom initiates the
discussion:
There are things in this comedy of Pyramus and Thisbe that
will never please. First, Pyramus must draw a sword to kill
himself; which the ladies cannot abide. How answer you that?
(III.i.8–11)
Starveling is ready to leave the killing out. But Bottom, who is to
play the role, proposes instead that Quince write a prologue in
which he can explain that Pyramus is not really killed and, even
better, as Bottom says,
tell them that I, Pyramus, am not Pyramus, but Bottom the
weaver. This will put them out of fear.
(III.i.19–21)
Bottom would thus have the audience told that he, Bottom, is Bot-
tom and no one else, that he, Bottom, is not Pyramus. Such a dec-
laration, delivered in a prologue, would in advance reestablish Bot-
tom’s simple self-identity and cancel his translation into suicidal
Pyramus.
Snout, Starveling, and Bottom agree, too, that the ladies will be
. . . afeard of the lion.
(III.i.27)
Snout suggests another prologue. But Bottom retorts:
Nay, you must name his name, and half his face must be seen
through the lion’s neck. . . . and there, indeed, let him name
his name, and tell them plainly he is Snug the joiner.
(III.i.35–36, 43–44)
Bottom’s proposal is thus that Snug do as he, Bottom, will do, that
he declare plainly and also show that he, Snug the joiner, is Snug
the joiner and not a lion; thus, too, would he reassert his simple
self-identity and retract his translation into a fearful lion.

But then, on the other hand, the mechanicals are concerned that
the audience will not be able to imagine anything that is not ac-
tually seen as such on stage. For being simply themselves, they sup-
pose the same condition to hold for what is presented on the the-
atrical stage: that whatever is presented must be actually present as
such on stage. Thus, noting that Pyramus and Thisbe meet by moon-
light, they consider how moonlight can be brought into the cham-
ber where they will perform. They consult a calendar and con¤rm
that the moon does shine that night; and then Bottom proposes:
Why, then may you leave a casement of the great chamber

window, where we play, open; and the moon may shine in
at the casement.
(III.i.52–54)
Thus would moonlight be presented by being let in through the

window so as to be actually present as such there where the play
is to be performed. But moonlight could be presented otherwise;
it could be presented without being actually present as such. It
could be presented by way of something else on the basis of which
one could envisage moonlight without its actually being present
as such. Such a presentation is what Quince proceeds to propose:
Ay; or else one must come in with a bush of thorns and a lan-
tern, and say he comes to dis¤gure or to present the person
of Moonshine.
(III.i.55–57)
It is not merely fortuitous that Quince, though a carpenter, plays

the role of playwright or at least of editor and director of the play
within the play. For what he proposes at this juncture could not
be more decisively different from what Bottom, being who he is,
has proposed. Instead of arranging for moonlight to be actually
present as such on stage, Quince proposes that it be ¤gured or pre-
sented by bringing on stage a lantern carried by a man bearing
also the proverbial attribute of the man in the moon. Moonlight is
to be presented without being actually present as such; it is to be
presented by way of something else the presence of which is ca-
pable of transporting those with imaginative powers to a vision of
40 On Translation
moonlight. Quince’s malapropism—his substitution of dis¤gure for
¤gure—is not without its appropriateness: for one who, like the me-
chanicals, is bound to actual presence as such, all presentation will
amount to a dis¤guring. Even Quince, doubling as carpenter and
playwright, remains to some degree bound: for even as he pro-
poses that someone come in to present Moonshine, he declares
also that this ¤gure is to
. . . say he comes to dis¤gure or to present the person of
Moonshine.
(III.i.56–57 [italics added])
He is to explain to the audience that he comes to present Moon-
shine, as if his mere appearance would otherwise be taken simply
as what it is, as if his mere appearance could not in and of itself
evoke an imaginative vision of Moonshine.
In this play Pyramus and Thisbe are to talk through the chink
of a wall. The same concern that arose about the moonlight arises
also about the wall. Snout says:
You can never bring in a wall. What say you, Bottom?
(III.i.61–62)
What Bottom says betrays who he is, in the double sense of the
word: it is inconsistent with his being who he is, that is, amounts
to his going astray from simply being who he is; and yet, it reveals
who he is, who he will prove to be, namely, one who can, in the
most remarkable ways, be translated. Here is what Bottom says:
Some man or other must present Wall; and let him have
some plaster, or some loam, or some roughcast about him,
to signify wall; and let him hold his ¤ngers thus, and
through that cranny shall Pyramus and Thisbe whisper.
(III.i.63–67)
Wall need not be actually present as such; like moonshine, wall
can be presented by someone. Bottom explains how this presenta-
tion is to operate: the one who is to present wall is to be adorned
in such a way as to signify wall. Here signify has the sense: offer
some sensibly manifest indication of that which is to be signi¤ed.

Or, more to the point, it means: offer to vision some sensible con-
tent on the basis of which it can be imaginatively translated to a
vision of that which is to be signi¤ed. It is perhaps a mark of Bot-
tom’s secret af¤nity to translation that he identi¤es this signify-
ing that is at the heart of theatrical presentation and that, having
identi¤ed it as the way by which someone may present wall, he—
thus surpassing even Quince—forgoes prescribing that the one
who comes to present wall must say he has come to present wall.
Here, at the threshold of his own remarkable translation, Bottom
broaches translation as it takes place in the theatre.
In the actual performance of “Pyramus and Thisbe” it is other-
wise, and the resulting incongruity contributes a great deal to the
comedic, not to say farcical, character of the play within the play.
To be sure, the performance incorporates those who, as the re-
hearsal discussion found necessary, are to present moonshine and
wall; and though indeed there is also added, as was prescribed, a
prologue, the prologue completely forgoes telling what it was to
have told, that Bottom is not Pyramus but Bottom the weaver. For
by the time of the performance Bottom has been translated into
all manner of guises other than that of Bottom the weaver.
These additions made to the play are made at considerable price:
in order to include presentations of moonshine and wall as well
as the Prologue, the play has to be skewed, the roles shifting such
that half the actors are transposed into roles other than those ini-
tially assigned. Quince, who was to have played Thisbe’s father, de-
votes himself instead to delivering the Prologue. Starveling, ini-
tially cast as Thisbe’s mother, ends up presenting moonshine. And
Snout, initially assigned the role of Pyramus’ father, is transposed
into the signi¤er of wall. These translations displace entirely the
three parents, who do not appear at all in the actual performance.
With Quince, as he delivers the Prologue, it is otherwise than
in what Bottom says—and, especially, does not say—about signi-
fying wall. For Quince follows his artisanal compulsion to identify
the signi¤ers, if only indexically and not by name. He says:
This man, with lime and rough-cast, doth present

Wall. . . .
(V.i.130–31)
42 On Translation
And then a few lines later:
This man, with lantern, dog, and bush of thorn,
Presenteth Moonshine. . . .
(V.i.134–35)
As if these words did not suf¤ce, both persons who come forth as
signi¤ers also identify themselves as such. Snout identi¤es himself
by name:
In this same interlude it doth befall
That I, one Snout by name, present a wall.
(V.i.154–55)
Starveling’s self-identi¤cation by way of the personal pronoun is
only slightly more discreet:
All that I have to say is, to tell you that the lantern is the
moon; I the Man i’ th’ Moon; this thorn-bush my thorn-bush;
and this dog my dog.
(V.i.247–49)
On the other hand, no character is more insistent than Snug in
identifying himself as signi¤er in order effectively to interrupt the
frightful signifying that might otherwise be carried out:
You ladies, you whose gentle hearts do fear
The smallest monstrous mouse that creeps on ®oor,
May now, perchance, both quake and tremble here,
When lion rough in wildest rage doth roar.
Then know that I as Snug the joiner am
A lion fell, nor else no lion’s dam.
(V.i.214–19)
Thus would Snug, reasserting that he is who he is, forestall trans-
lation.
What, then, is played out in the rehearsal and performance of
“Pyramus and Thisbe” and what is disclosed thereby? In their
simple self-identity the mechanicals carry on a discourse about the
theatre that is governed by their tacit allegiance to undivided pres-
ence and its corollary, the dif¤culty in distinguishing between

theatrical presentation and simple presence as such. On the one
hand, they believe that if something is presented on stage, it will
be taken to be actually present as such—as with the killing and
the lion. On the other hand, they believe that in order for some-
thing to be presented, it must be actually present—as with the
moonlight shining in through the casement. And yet, what occurs
in the course of the play is not simply an enunciation and enact-
ment of this view; rather, in and around the enunciation and en-
actment of it, there is played out an exceeding of it. The opening of
the rehearsal scene already broaches such a move, Quince present-
ing the place in the forest as a theatre, hinting at a theatrical pre-
sentation of the theatre itself:
Pat, pat; and here’s a marvelous convenient place for our

rehearsal. This green plot shall be our stage, this hawthorn-
brake our tiring-house; and we will do it in action, as we
will do it before the Duke.
(III.i.2–5)
Here already “this green plot” is translated into “our stage” and
“this hawthorn-brake” into “our tiring-house” (that is, dressing
room). The green plot is not simply what it is; its simple presence
as such is already breached by the translation. As soon as the green
plot is a stage and the hawthorn-brake is a dressing room, there is
violation of simple identity and of undivided presence, as well as
a retracting of that identi¤cation of presentation with presence
that is the presupposition for the entire discussion that is about to
commence. Within this discussion itself there is continual exceed-
ing of this presupposition: even to foresee and hence fear that the
ladies in Theseus’ company might fail to distinguish the presenta-
tion of a lion from its actual presence requires that in a sense they
translate themselves into their audience-to-be and into the future
in which the performance will be held. But what is most striking
is the way in which the exceeding of undivided presence is played
out in the transition—broached by Quince—to the presenting,
¤guring (or dis¤guring), signifying of moonshine and of wall;
and, above all, when the one most susceptible to translation re-
leases the signi¤er from the need to attest verbally that it is such.
44 On Translation
In all these respects what is played out is the way in which, in and
around the simple presence to which the mechanicals would ad-
here, differences break out and across these differences transla-
tions occur. By the way these are played out in the play, by be-
ing allowed to reopen on the ground of the mechanicals’ very
obliviousness to them, these differences and these translations are
dramatically—indeed comically—disclosed in a more forceful
and wondrous way than one could perhaps ever have imagined.

Three Translation and the Force of Words
Suppose, now, one were to resist the word’s polysemy and mobility,
which appear in theatre to be given full rein. Suppose, now, one
were to take a certain distance from the tangle of senses and from
the entangling, enrapturing dream of the theatre. Suppose, now,
one were to restrict the word translation to a single meaning, to the
single meaning that, currently at least, it would most readily be
taken to have. Suppose, now, within a certain discursive interval,
one were to translate translation only as transposition from one lan-
guage to another—even while leaving aside the question of limits,
of the limit to which resistance to the word’s polysemy and mo-
bility can be sustained, of the limit to which distancing from its
multiplicity and slippage is possible, of the limit to which restric-
tion of the word to this simple meaning can be effectively main-
tained.
It is to translation in this sense that Jakobson awards the desig-
nation translation proper. In order to distinguish other senses from
this proper sense, Jakobson resorts to translating translation, though
doing so in what would—by his own designations—have to be re-
garded as an improper sense. Thus he improperly translates the
improper senses of translation, rewording intralingual translation as re-
wording and intersemiotic translation as transmutation, thus setting off the
would-be proper sense of translation precisely by putting in play an
improper sense. To be sure, Jakobson stops short of making the
determination of the proper dependent on this operation of the
improper on itself, though the independence of the proper would,
it seems, have ¤nally to be based on rigorous differentiation be-
tween intralingual and interlingual. If the singularity of languages
were to be compromised, if translating within a language could
be, at the same time, a translating between languages, then the
sense and delimitation of proper translation would to that extent
become problematic.
46
There are other entanglements, too, other kinds of complica-
tions that will be dif¤cult to hold at bay inde¤nitely but that need
to be left aside at least for a time. One of these has to do precisely
with the singularity of languages, with the dif¤culties that arise
as soon as one undertakes to specify what constitutes a single,
proper language (and, hence, translation between two such lan-
guages, translation proper, in Jakobson’s designation). Even if one
brackets all historical, developmental considerations and maintains
a strictly synchronic point of view, the fact remains that languages
typically include certain foreign words and phrases that function
as foreign elements precisely as they function within the language
itself. In some cases they are explicitly treated as foreign additions,
as when they are printed in italics. In other cases, especially in
speech, the mark of their foreignness may be more subtle, and in
instances where, from a diachronic point of view, a process of as-
similation could be traced, it may be almost entirely effaced. Yet
in every case the foreign word or phrase functions as if it belonged
to two different languages; that is, from the point of view of the
would-be singular language, such words and phrases function as
if they both belonged and did not belong to the language. The op-
eration of such words and phrases within a language has an effect
on the limit that otherwise would determine the language in its
singularity: it is as if the limit that would encircle the would-be
singular language had split into two concentric circles outlining
a parergonal band of undecidability. If the functioning of such
words and phrases is ampli¤ed in the direction of a polylingual
text—one thinks of Finnegans Wake1—then severe complications
confront translation. How is one to translate a text that is written
in more than one language or at least in what is not a singularly
determinable language? There are also cases in which, as with
Presocratic texts, an interpretation may be offered precisely as an
extended translation of the text or at least as serving only to pre-
pare the translation. How is one, then, to translate such transla-
tions?
The singularity of languages is also complicated by the way in
1. See Jacques Derrida’s discussion in “Des Tours de Babel,” in Psyché, 207–208.
Translation and the Force of Words 47

which proper names function. It is not uncommon for a proper
name to belong in common to more than one language, though
proper names are not, on the other hand, simply indifferent to the
alterity and diversity of languages. Most notorious are the compli-
cations that arise from the fact that proper names as such do not
signify a meaning but rather name something singular (even where
there is a certain multiplication of the singulars). To the extent
that the very concept of translation is linked to the signi¤cation
of meaning, it remains problematic whether and in what sense
proper names can be translated. Yet translations there are. As when,
in Schlegel’s translation of Shakespeare’s play, Peter Quince becomes
Peter Squenz, Robin Starveling becomes Matz Schlucker, Nick Bottom be-
comes Klaus Zettel, and Robin Goodfellow, untranslated, is omitted al-
together. Translations there are, not only such instances as these,
which can hardly fail to raise questions about translation of proper
names, but also instances in which the translations are thoroughly
established and taken for granted, as when Peter becomes Pierre,
Elizabeth becomes Elsebet, John becomes Jean or Johann, and Richard re-
mains untranslated or, rather, remains the same in translation.
Still another complication is broached by the series of connec-
tions outlined by Aristotle in a passage that was to prove decisive
for the way in which language came subsequently to be taken up
as a philosophical problem. The passage is from On Interpretation,
though it is also in a sense detached by Aristotle from this text. In
translation it reads: “Spoken words are symbols of affections in the
soul, and written words are symbols of spoken words. As writing
is not the same for all men, so likewise speech is not the same for
all. But the affections of the soul, of which these words are pri-
mary signs, are themselves the same for all, as are also the things
of which these affections are likenesses.”2
It should be noted that it is the translation that interposes the
word word in all its occurrences in this translation. In the Greek
text there occurs no word for word but only, in the one instance,
writing, inscription (gr$mata) and the written or drawn (that which
has been written or drawn: graf1mena). In the other instance
2. Aristotle, On Interpretation 16a.
48 On Translation
what is designated is that which is in the voice (t% \n t_ fvn_). It is
neither an accident nor a shortcoming that Aristotle speaks of
speech as in—and, hence, as coming forth from and as—the
voice. The operation of the word fvn} in speech about speech is
found likewise in the Platonic texts, one of which will be exam-
ined below. It is found even in such contexts as that of the Cratylus,
where other words that one could take as words for word are also
operative.3 The chief candidate is Onoma, which, however, ranges
over a broad spectrum of senses. At one extreme it can mean an
expression or a saying. At the other extreme its sense diverges in two
different directions: on the one hand, it can mean noun in the
grammatical sense as opposed to verb (]+ma, which, however, also
ranges over the same broad semantic ¤eld), while, on the other
hand, it can mean name, not only just as proper name but also in
the sense of the name one may have made for oneself by one’s
deeds, hence also good name or fame. The other most pressing can-
didate, the word l1gow, is such as to exceed the word word in such
manifold ways—not only by its semantic range but also by its
manifold of concurrent senses—that Aristotle’s avoidance of it in
the passage would have been virtually inevitable.
Strictly speaking, it is not just the word for word that is missing
in the passage. In a sense the passage is not about words at all, at
least not as they function normally in language, being connected
to other words according to certain syntactical rules so as to ex-
press a coherent meaning. If one were to venture to translate the
virtually untranslatable word l1gow as discourse and if one were
also to adhere to what could then be called the ancient ¤gure of
discourse as weaving, then one could say that the Aristotelian pas-
sage is not about the weaving together of words into discourse but
rather about the various folds both in and of the fabric of dis-
course. These folds belong to discourse no matter how it may
3. At the beginning of the Cratylus Hermogenes reports to Socrates, who is

just joining the conversation, what his interlocutor Cratylus has been maintain-
ing, namely, that there is a natural correctness of names (Onoma). Hermogenes
explains that this means that a name is not merely “a piece of their own voice
[fvn}] that people utter,” that being merely voiced does not suf¤ce to guaran-
tee that something is truly the name of a thing (Crat. 383a).

happen to be woven. They are produced, not by weaving words
together into expressions, but rather through a kind of transla-
tion. Or rather, through a series of translations: writing translates
speech, speech translates the affections in the soul, and these af-
fections translate things. Because this manifold is produced by
translation—speci¤cally, by translations that Aristotle regards as
conveying the essential—each fold is a likeness or a symbol of that
upon which it is folded. This multiple folding, this manifold that
translation produces in and of the fabric of discourse, is the origi-
nary complication.
Yet none of the translations involved in producing the originary
complication coincides with translation between languages, with
translation in what Jakobson considers the proper sense. Inasmuch
as the originary complication comes into play in and with the op-
eration of a language, translation across the interval between two
languages would presuppose the originary translations between
writing, speech, affections, and things. In any case this originary
manifold needs, for the moment at least, to be disregarded in or-
der to focus on translation between different languages. What is
needed is to resist—even in this originary direction—the mobility
of translation, to restrict it to the sense of translation from one lan-
guage to another, and thus to keep a certain distance from the
originary complication.
Translation is inseparable from measure. In translation from
one language to another, a measure must govern the transference
that occurs across the interval separating the languages. It is in ref-
erence to this measure that a translation can be judged good or
bad or even not a translation at all. What is the measure? The trans-
lation produced is supposed to be true to the original, true to the
text (or speech) from which it is produced and of which it is al-
leged to be a translation. But what is this truth of translation?
What does truth mean in this connection? Presumably it consists
in the translation’s corresponding to the original, in its being like
the original. But what sense does correspondence have here? Cor-
respondence in what respect? And how can a word, phrase, or sen-
tence in one language be like a word, phrase, or sentence in an-
other language?
50 On Translation
It is to this question of the measure, the truth, of translation
that the classical determination of translation responds. This clas-
sical determination is prepared in the Platonic dialogue Critias, the
fragmentary sequel to the Timaeus. In the Critias the promise made
in the Timaeus would be made good: now Critias would tell in de-
tail the story he had only brie®y outlined in the Timaeus, the story
of the great and wonderful deeds of the original Athens, the Ath-
ens of 9,000 years ago, in its struggle against the expansionist de-
signs of Atlantis. Already in the Timaeus Critias relates how the
story has come down to him from his grandfather Critias, who
was told it by his father Dropides, who, in turn, was told it by his
relative and friend Solon. Solon, in his turn, had been told the
story when he traveled to a foreign land, speci¤cally, when he
visited Saïs in Egypt, a city said also, like Athens, to have been
founded by Athena, but by Athena under another name, the for-
eign name Neïth. Thus, Solon’s story of Athens as it was indeed in
the beginning was brought from a foreign land, from a foreign
city whose founding and constitution had so much in common
with Athens as to make it a kind of foreign double of Athens. It
was there, in that foreign city, that the story had been preserved
in writing, surviving thus the loss to which living memory is sub-
ject, especially when, as in Greece, much of the population has
been repeatedly destroyed by natural calamities. But as preserved
in this foreign place, the story had itself become foreign—foreign
being understood by the Greeks primarily in reference to speech,
the foreigner being precisely one who did not speak Greek. In
other words, the story had been preserved, not in Greek, but in a
foreign speech, in the speech of the foreign place where it was
sheltered from destruction. Thus, in bringing the story back to
Athens from this foreign place, perhaps even in order to recover
the full story as such, Solon was faced with the problem of trans-
lation.
It is in the Critias that Critias describes how Solon dealt with
this problem. The passage on Solon’s translation occurs, most ap-
propriately, at that point in the dialogue where, having spoken of
ancient Athens, Critias is about to tell of Atlantis; in the narrative,
as in translation as such, it is a matter of transition between one’s

own and the foreign. It is precisely to forestall a certain confusion
regarding these that Critias interrupts the narrative and speaks
brie®y of translation.
Here—in (my own) translation—is how he begins: “Brie®y,
before the account [l1gow], it is necessary to explain something,
lest perhaps you wonder at hearing Greek names of foreign men
[‘Ellhnik% barb$rvn #ndr3n <n1mata]. The cause of these
you will now learn.”4
What the explanation is to forestall is the wonder or astonish-
ment that might be provoked by hearing names that, though they
are the names of foreigners, are Greek rather than foreign. Al-
though, from this point on, the story is even more thoroughly for-
eign, not only recovered from a foreign place but largely about still
another foreign place, Critias is to narrate it in Greek, saying even
the names of foreign men (the rulers of Atlantis) entirely in Greek.
Or rather, almost entirely in Greek: for there is one notable excep-
tion near the beginning of Critias’ account of Atlantis. The account
begins by referring again, as at the beginning of the account of
ancient Athens, to the gods’ portioning out of the whole of the
earth. Critias relates that Poseidon took for his allotment the island
of Atlantis. Not only did he form and shape the island, surround-
ing the acropolis with circular belts of sea and land enclosing one
another alternately, but also he begat of a mortal woman, Cleito
(daughter of one originally sprung from the earth itself ), ¤ve
pairs of twin sons. Having then divided the island of Atlantis into
ten portions, Poseidon set about assigning to each son two things:
¤rst, a portion of the island over which to rule, and, second, a
name (Onoma). The ¤rstborn of the eldest twins was assigned
the acropolis and its surroundings, and there he was to reign as
king over the others; his name, as Critias states it, was thoroughly
Greek. And yet, from Critias’ earlier explanation one knows that
the name of the king almost certainly cannot have been (as Critias
says) Atlas; it must, rather, have been a foreign name, which sub-
sequently came to be translated as Atlas. But then the island and
the ocean, which share his name, cannot have had the names At-
4. Plato, Critias 113a.
52 On Translation
lantis and Atlantic but must have had—from Poseidon—other, for-
eign names. One realizes that the names Critias is using in his
account—names that will have resulted from translation—replace
the original, foreign names. The story of the island of Atlantis
is thus in fact the story of an island that the Greeks called Atlan-
tis but that itself almost certainly bore (natively, as it were) an-
other, foreign name that remains unknown to—or at least unsaid
by—Critias. The translational replacement of the original, foreign
names has the effect of undoing the assignment of names carried
out by the god himself. Hence, such translation represents, within
this perspective at least, a subversive and excessive venture on the
part of mortals.
Yet it is not the name of the king that is the exception but rather
that of his twin brother. To this second-born of the eldest twins
Poseidon assigned a portion of the island extending from an ex-
tremity near the pillars of Heracles up to the region now called
Gadeira. As in every case, Poseidon assigned him also a name. But
in his case, unlike the others, Critias says the name not only in
Greek translation but also in its original, foreign form: his name,
says Critias, was “Eumelus in Greek, but Gadeirus in the native
[speech].”5 The mention of the original, foreign name—foreign
to Greeks but native to inhabitants of the island that Greeks call
Atlantis—serves at the very least as a reminder that all the other
names that occur in Critias’ discourse about the island kingdom
and its exploits are results of translation. The survival of this origi-
nal, foreign name in a discourse otherwise entirely in Greek is
presumably to be attributed to its consonance with the still cur-
rent place name Gadeira. This particular connection serves to point
up the signi¤cance that the connection between name and place
has throughout the Critias and especially in Poseidon’s assignments
to his sons. In Critias’ speech about their names, there occur the
two words xQra and t1pow, around which the most abysmal dis-
course of the Timaeus circles, a discourse (chorology) that ventures
even beyond what will come to be called place. And when Critias
says the foreign name of the son called Eumelus in Greek, the word
5. Ibid., 114b.

he uses to identify the name Gadeirus as other than Greek, as be-
longing to the indigenous speech of Atlantis, is simply \pixQrion:
native, belonging to the country, to the xQra.
The beginning of the passage on translation is oriented pri-
marily by the word Onoma: the concern (which makes an expla-
nation necessary) is with the effect of certain names, of Greek
names used for foreign men, whose names would not have been
Greek but foreign. Not only does the word Onoma orient the pas-
sage from the beginning, orienting indeed the entire passage on
translation, but also the word imparts to the entire passage its
peculiar polysemy. Its most apparent sense at the beginning of the
passage is that of name, indeed proper name. The concern is with
the effect of the Greek names used for foreign men, as—to take
the exceptional case—with the use of the Greek name Eumelus for
the foreigner whose name, as both he and Poseidon would have
said it, is Gadeirus. Yet as long as Onoma has the sense of proper
name, translation will remain problematic, in particular the very
translations that Critias goes on to consider in the passage. How
can a proper name in the speech indigenous to the island that
Greeks call Atlantis be translated into a Greek proper name? Per-
haps only insofar as the former is more than a proper name, in-
sofar as it also functions as a common noun that does not just
name a singular but also signi¤es a meaning. Is it at such a tran-
sition that the Platonic text hints by ascribing to the one person
whose foreign name is stated a Greek name that very transpar-
ently doubles as a common noun? As if the Greek name Enmhlow,
which means rich in sheep, could have been arrived at via this mean-
ing. As if G$deiron might have had some such meaning and thus
have been appropriately translated by Enmhlow.
But then once Onoma assumes the sense of common noun, the
expressed concern shifts. What now might prompt wonder are,
for instance, the various words by which, in Greek, foreign men
might be characterized—that certain ones are heroes or statesmen
or philosophers. But then the sense of Onoma will easily extend to
whatever is said, in Greek, of foreign men, assuming thus the
broad sense of an expression or a saying, approximating (at least in
this dimension) to the sense of l1gow, the word used by Critias
to designate the entire discourse on Atlantis, easily extended, in
54 On Translation
turn, to Critias’ entire discourse as such. The slippage of the word
Onoma would thus serve to broaden the passage on the translation
of names into a description of how the entire story narrated by
Critias came to be translated.
Critias continues with a sentence that may be translated as fol-
lows: “As Solon was planning to make use of the story [l1gow] in
his own poetry, he found, on investigating [diapunyan1menow]
the force of the names [t|n t3n <nom$tvn d4namin], that those
Egyptians who had ¤rst written them down had translated them
into their own voice [ecw t|n aSt3n fvn|n metenhnox1taw].”
In reporting that Solon was planning to make use of the story
in his own poetry, Critias is reiterating what he said in the Timaeus.
In the earlier report Critias indicated that Solon did not in fact suc-
ceed in carrying out his plan. Critias cites two reasons: ¤rst, Solon
pursued his poetry only as something ancillary (p$rergon), and,
second, he was compelled to put it aside on account of the evils
in Athens with which he had to contend upon his return from
Egypt. Critias is of the opinion that if it had been otherwise, if
Solon had been able to carry out his plan of rendering the story
in Greek poetry, “then neither Hesiod nor Homer nor any other
poet would ever have proved more famous than he.”6 In a word,
he would have made a name for himself as a poet.
In order to have rendered the story in Greek poetry, he would
have had to translate the writings in which it had been recorded,
the writings that he was shown while in Egypt.7 He would have
had to translate these writings at least to the extent necessary for
retelling the story in Greek; in other words, he would have had to
produce a translation at least in the sense of a retelling of the story
in Greek. What about the proper names of the foreigners who ¤g-
ure in the story, of those from the island and empire that Greeks
call Atlantis? Perhaps, from considerations of prosody, Solon would
6. Plato, Timaeus 21c–d.

7. Critias reports that the Egyptian priest under whose tutelage Solon was
taken told him brie®y about the laws and the noblest of deeds performed by the
ancient Athenians. Yet the priest promised that at their leisure they would go
through the full story in exact order and detail, taking up the writings them-
selves (ibid., 23e–24a).

have preferred to translate these names into Greek. Or perhaps, lest
auditors wonder at hearing Greek names of foreign men, he would
have chosen to leave these names untranslated.
Yet he had no such choice. This is what he found out once he
set about investigating the matter. He found that those Egyptians
who had ¤rst written the names down had translated them. Hence,
in the writings that Solon was shown in Egypt, the foreign names,
the names foreign to both Greek and Egyptian, have already been
replaced by translations. In these writings there is already reason
for concern about the effect such names might have, about the
astonishment that could be prompted by hearing Egyptian names
of foreign men. Yet in the Egyptian writings their own names, the
names they would have called themselves and would in the case
of his sons at least have been called by Poseidon, are completely
effaced, with, it seems, the exception of Gadeirus. Short of trans-
lating back from the Egyptian names, that is, countertranslating,
there is no means of retrieving their original names, the names
native to them but foreign to Greek and Egyptian. The loss of their
names is hardly less de¤nitive than that of Atlantis itself, which
“sank into the sea and vanished,”8 leaving no trace of itself except
the shoal that made the ocean at that spot impassable, just as there
was left from the original names only the traces provided by their
Egyptian translations.
But what is the sense of translation here and how does it take
place? Critias says that the Egyptians translated the names into
their own voice (fvn}). Or, if one translates translation back, as it
were, into Greek: they carried them over, transferred them (the
word is a form of metaf)rv) into their own voice. Here, too, as
in Aristotle and in other Platonic texts, what one might otherwise
call speech is spoken of as the voice, as taking place in and as the
voice, as phonation. In translating, the Egyptians carried the for-
eign names over into their own voice. Having done so, or as a way
of effecting the transfer, they wrote down the names.
But how did Solon, as Critias says, discover that the Egyp-
tians who had ¤rst written down the names had translated them
8. Ibid., 25d.
56 On Translation
into their own voice? Critias attests that Solon did so by investi-
gating, by carrying out a thorough search through questioning
(diapuny$nomai). Nothing dictates of course against assuming
that Solon may have addressed questions to certain Egyptians he
met in Sa•s, for instance, to the old priest under whose tutelage he
was taken. Yet Critias’ account states unequivocally that the object
of Solon’s investigation was the force of names. Whatever questions
may have been addressed to the Egyptians would have been di-
rected precisely to this goal, to searching out the force of names.
He could, assuming a common speech, have asked someone about
the status of the names. He could have asked the old priest, for
instance, about the voice in which the names of the various leaders
of the now-sunken island were inscribed. He could have asked the
old priest whether these names, as they had once been written
down by Egyptian scribes, were Egyptian names or not, assuming
that if they were not Egyptian they must have been in the voice
of those who inhabited the island. Yet, even if the old priest had
presumed to answer and had informed Solon that indeed all the
names were Egyptian, the breach of singularity belonging to each
voice forestalls all certainty in this regard. For there is nothing to
prevent a name from belonging to more than one voice, most no-
tably, but not exclusively, in the case of proper names. There is
nothing to guarantee that a name inscribed in Egyptian is not also
a name in the voice—now presumably extinct—indigenous to the
island called (by Greeks) Atlantis. That the names by which certain
leaders of the island are called in the Egyptian writings appear
Egyptian—and a native speaker can presumably determine this al-
most unfailingly—does not establish conclusively that these names
result from translation and not from repetition. Even if, merely re-
peated at the time of inscription, they had once seemed foreign to
the Egyptian voice, the assimilation that the antiquity of the writ-
ings would have permitted would have served to efface their al-
terity.
One could suppose, then, that this inevitable uncertainty is what
led Solon not just to ask the Egyptians about the names in the
Egyptian writings but to investigate the force of these names. The
force (d4namiw) of a name lies in its being capable (d4namai),
in its being capable of accomplishing that which it is proper to

a name to accomplish. What is proper to a name as such is that
it announce something or someone, that it announce that which
it names. In announcing what it names, the name presents it,
makes it present in a certain way, in a way that philosophy—from
Plato on—distinguishes from the way in which sense percep-
tion (arsyhsiw) makes things present. A name is capable, then,
of making present in a certain way that which it names even
when what it names is not at all present to sense perception, even
when it could not be present to sense perception, even when it has
passed away so as never again to be able to come before sense per-
ception, even when it has not yet come to be so as to be able to
come before sense perception, even when it is such that it could
never come before sense perception. The force of a name is thus
its capacity to make manifest that which it names, to draw it forth
into a certain manifestness that is, in a certain way, independent
of perceptual manifestness. This is why names, especially when
they are preserved in writing, are the repository of memory. Just
as he is about to begin his discourse in the Critias, a discourse in
which nothing ¤gures more prominently than names, which are
themselves to be recalled in order that the Athenian beginning be
remembered—as he is about to begin Critias calls upon all the
gods, but most of all upon Mnemosyne, “for nearly all the most
important part of our account [l1gow] depends on this goddess.”9
In investigating the force of names, Solon would, then, have
searched out—with or without assistance from the Egyptians he
met in Sa•s—the capacity of the names in question to make mani-
fest that which they name. And, in investigating that force, Solon
would—with or without assistance—have put the force in force,
that is, would have let the names exercise their capacity to make
manifest that which they name. Investigating them in their exer-
cise of this capacity to make manifest, he would, in turn, have
gained a measure of this capacity, a measure of their force. How,
then, was it that by investigating the force of names Solon discov-
ered that the names had been translated, that the names written
down long ago by the Egyptian scribes were translations, not the
9. Plato, Critias 108d.
58 On Translation
original, native names of those told of in the account? It can only
have been by way of the measure he gained of the force of the
names. It can only have been through his discovering a certain in-
capacity of the names in their translated form, through his ¤nding
the force of the names inscribed in the Egyptian writings to be
weak, as measured, most immediately, against the force of such
names as he would have said in his own voice. That those in-
scribed names were translations can have been attested only by
their relative incapacity to make manifest that which they name,
by their leaving that which they name still in some measure con-
cealed, resistantly closed off, precisely as though it were foreign.
Here one ¤nds indicated for the ¤rst time what later—and espe-
cially in modern times—will be ever more insistently declared:
that translation cannot occur without loss, that in a translation the
force of names will always have been diminished, that in transla-
tion names undergo a loss of force. Again and again it will be said
that a translation is always less forceful than its original.
In translating into their own voice the proper names of certain
inhabitants of the island that Greeks call Atlantis, the Egyptian
scribes could hardly have avoided also translating what, in the
voice of those inhabitants, had been said of them. The scribes can
hardly have avoided translating, for instance, the name that certain
of them made for themselves, that is, the fame, the reputation, that
would of course have been declared, not merely by citing proper
names, but by words, common nouns and verbs, describing their
qualities and their deeds. Thus, there is good reason to suspect that
virtually everything the Egyptian scribes wrote down would have
been a translation, that the writings Solon was shown in Egypt
were nothing but translation.
In any case, since Solon’s intent was to use the story for his own
poetry, it was his task to translate these writings into Greek, that
is, to convey them into his own voice: “So he himself, in turn,
retrieved the thought [di$noia] of each name and leading [Wgvn]
it into our own voice wrote it out.”
Beginning in each case with an Egyptian name, Solon’s ¤rst
move was to retrieve the thought of the name; then, in a second
move, he led, directed, drew, this thought into his own voice, into
the voice of Greeks; then, as a ¤nal move or as the consummation

of the second move, he wrote down the Greek name. Here for the
¤rst time the structure, the basic constitution, of translation is de-
termined; here the Platonic text declares what may be called the
protoclassical determination of translation. But in this determina-
tion everything depends on how the single word di$noia is un-
derstood. Very few words prominent in the Platonic texts appear,
on the one hand, so readily translatable and yet are, on the other
hand, so resistant to translation into all those modern words that
one would attempt, as it were, to translate back into di$noia.
Following the itinerary of this word in the Republic, for instance,
could hardly not have the effect of persuading one to leave the
word literally untranslated and to translate it only in the sense of
surrounding it with a discourse in which its sense could be ad-
umbrated.10 But even if—conceding, hesitantly and with reserva-
tions, to a certain tradition of translation—one were to translate
di$noia as thought or intention (limiting these words, as would be
necessary, by reference to a well-de¤ned philosophical sense, as in
phenomenology), it remains ambiguous. For, thus translated, the
word could designate either thinking or that which is thought
through the thinking, either the intending or that which is in-
tended by it. Even aside from considerations of the correlativity of
the noetic and noematic sides, there is good reason to retain the
ambiguity, but in the form of a duality: for Solon’s task was to re-
trieve the thought, which he could have done only by carrying out
to some degree the thinking through which the thought (that
which is thought) is thought, that is, by enacting the thinking of
the thought, the intending of the intended (intentum). Speci¤cally,
then, Solon’s retrieval would have taken the form of an enactment
in which he would have come to intend that which is intended
through the name. But that which is intended, that which the
name names, is nothing other than that which the name, through
its force, can make manifest. Solon’s retrieval of the di$noia of a
name would thus have taken the form of an enactment in which
the force of the name would be released, would be put in force in
10. See Being and Logos: Reading the Platonic Dialogues, 3rd ed. (Bloomington: Indiana
University Press, 1996), 424–43.
60 On Translation
such a way that what is named would be made manifest. The re-
trieval of the di$noia of a name, which would constitute the ¤rst
move of translation, would already have been carried out in the
investigation of the force of the name, the investigation by which
Solon found out that the Egyptian text was in fact a translation.
Indeed the retrieval of the di$noia of a name would have been
carried out even short of the explicit gauging of this force that is
required in order to expose it as a translation.
What would not, however, have been carried out is the decisive
second move that translation requires, the move in which the
di$noia would be drawn into his own voice. It is in this second
move that all the complications and dif¤culties of translation will
prove to be concentrated. For what it requires is that one discover
in one’s own voice a name that not only names the di$noia—the
very same di$noia named in the other voice—but also is of such
force as to be capable of making manifest that which it names, of
bringing it to presence in that certain way appropriate to names.
Though this second move brings one from the foreign back to
one’s own, it is decidedly not a matter merely of retrieval, of grasp-
ing again what has been said, but of saying what has never yet
been said in one’s own voice.
As the passage on translation proceeds, it shifts ever more de-
cisively from focusing merely on proper names to consideration
of discourse as such. Nothing points this up so clearly as Critias’
remark immediately following the passage: “And these very writ-
ings [resulting from Solon’s writing down the names in transla-
tion] were in the possession of my grandfather and are now in
mine, and as a child I learned them by memory.” There is every
reason to believe that what was translated and written down in
and as these writings handed down to Critias himself from his
grandfather was the entire story as such and not merely the proper
names of those who ¤gured prominently in it.
One could conclude, then, that the passage from the Critias con-
stitutes a discourse on translation as such. From it issues the in-
augural determination of translation, the protoclassical determina-
tion, which subsequently comes to be stabilized by reference to
the distinction that itself comes to be stabilized as fundamental to
philosophy (so radically so as to determine the very sense of fun-

damental). Through this double stabilization the classical determi-
nation of translation is constituted. In those few instances in the
history of philosophy where translation again surfaces as a ques-
tion, only scant deviation from this classical determination is to
be found. Even when Nietzsche declares certain would-be funda-
mental translations to be utter perversions of what would be trans-
lated, when he describes them as overleaping from one sphere into
another without carrying anything over, as translations that trans-
late nothing, the parameters of this discourse are still governed
by the classical determination of translation precisely as this dis-
course overturns that determination, inverts it, and gestures, per-
haps, beyond.
In the protoclassical determination that issues from the Critias,
the force of names ¤gures prominently. Retrieving the di$noia
of a name requires that the name not be simply and emptily
repeated in the manner all too familiar from everyday speech; re-
trieving the di$noia requires, rather, that the force of the name
be released so that, through its force, the name comes to make
manifest that which it names. Drawing the di$noia into one’s
own voice requires a search within one’s own voice that could be
carried out only by voicing various names, not in mere repeti-
tion, but in such a way as to release and at the same time to mea-
sure their force. Only through such a search could one—if at all,
for untranslatability is not ruled out—¤nd a name that names
the very same di$noia with force suf¤cient to make it manifest;
and only if a name has such force is it possible, by releasing the
force, to con¤rm that the name names the very same di$noia
named in the other voice. Thus, as conveyance of the di$noia
across from one voice to another, translation is engaged with the
force of names, with enactments that release and measure such
forces. Indeed it is the force of names that, above all, enables trans-
lation.
In the stabilization that produces the classical determination
of translation, the di$noia is secured as noht1n over against the
acsyht1n. Or if, for the moment, one merely resumes what is
perhaps the most decisive and questionable translation in the his-
tory of the West, the translation of the basic words of Greek phi-
losophy into Latin, then one may say: in the classical determina-
62 On Translation
tion the di$noia is secured as intelligible over against the sensible
character of the name.11 With the di$noia thus secured as intelli-
gible being, the orientation is shifted away from the manifesta-
tion effected by the name, away from the force by which a name
makes present that which it names. If what is named, that is, the
di$noia, is perpetually present as such, then the capacity of a
name to let it be present has diminished signi¤cance and ¤nally
comes to be constitutively linked to merely human limitations. For
a name to name something no longer entails that in some measure
it makes manifest what it names, that it names it in and as making
it manifest, bringing it to presence. In place of such naming as
making manifest, what becomes de¤nitive is the abstract relation
between the name and the di$noia named, which is determined
as intelligible and eventually as signi¤cation or meaning. In a cu-
rious reversion to a position not unlike one of those that in the
Cratylus was unhinged in and into comedy,12 everything comes to
depend on the abstract relation of the name to its signi¤cation or
meaning. Once names thus become signi¤ers, the very sense of the
force of names will have been lost.
The classical concept of translation thus makes no mention of
the capacity of names to make manifest; in this determination all
reference to the force of names has disappeared, or if a trace is
still indicated, it is no more than a vestige now quite ineffective.
The schema that constitutes this determination is correspond-
ingly simple: translation consists in the movement from a unit in
one language (word, phrase, sentence, etc.) to a corresponding
unit in the other language, this movement being carried out by
way of circulation through the signi¤cation, the meaning. Begin-
11. It proves necessary to repeat this securing with respect to the name itself.
For a name is not just a singular sensible occurrence, not just, for instance, a
singular sound or series of sounds uttered by a speaker at a certain time. A name
can be repeated at various times and uttered by various speakers, and there will
be considerable variation among these instances, no one of which can be iden-
ti¤ed as the name itself. This “ideality” of the name requires, then, that a dis-
tinction be drawn between the name itself, which can never be uttered or heard
as such and which is thus stabilized as intelligible, and the various instances in
which there is a sensible utterance or inscription of the name.
12. See Being and Logos, chap. 4.

ning with, for instance, a word in one language, one passes to the
meaning in order then, from the meaning, to pass to the corre-
sponding word in the other language. In this determination the
sense of correspondence, the truth of translation, is also deter-
mined: a translation is true to its original if it has the same mean-
ing. The measure of translation is restitution of meaning.
At least with units larger than a single word, translation will
usually require syntactic as well as semantic strategies. It is well
known that what is expressed by certain words in a certain syn-
tactic structure in one language often can prove expressible in
another language only through a very different syntactic struc-
ture. In some cases certain words in the one language can be con-
veyed in the other language, not as such, but only through cer-
tain syntactic strategies; such is often the case, for instance, with
Greek particles. But whatever the syntactic transformations re-
quired, they are entirely in service to the restitution of the mean-
ing of certain linguistic units. Even if, for instance, in translating
poetry, there may be reason to preserve certain syntactical struc-
tures along with meter and rhyme, all such strategies are, in the
end, to be subordinated to the ideal of saying in the translation
precisely what the text—in smaller or larger segments or even just
as a whole—means.
In this classical determination there is reference neither to the
manifestive force of names nor to any enactment that would re-
lease and measure such force. To the extent that the Greek name
for name—that is, Onoma—has its very sense determined by ref-
erence to manifestive force, one will be compelled to admit that
none of the translations proposed for it, not even the translation
name, to say nothing of word, phrase, expression, etc., are true to it, not
at least insofar as all these modern names for what the Greeks
called Onoma are governed by the classical determination of trans-
lation or at least by a determination of language as such that cor-
responds to it. Even the word language is itself something less than
a true rendering of that by which one could take the Greeks to
have designated the same phenomenon. It is not just that language
would render both t% graf1mena and t% \n t_ fvn_ (as we say,
both writing and speech), but, more signi¤cantly, that with lan-
guage one says abstractly what the Greeks said with remarkable phe-
64 On Translation
nomenal concreteness: for what is called language comes to pass
phenomenally precisely as inscription or as voice. What is called
language occurs, happens—indeed is—only as an inscription or a
voicing of names.13
The classical determination of translation is in force in Cicero’s
re®ections and no doubt in his practice of translation. Its schema
is clearly discernible in The Best Kind of Orator (De Optimo Genere
Oratorum), a work dated 46 b.c. though not published in Cicero’s
lifetime. The work was to serve as an Introduction to Cicero’s
translations of Demosthenes’ On the Crown and Aeschines’ Against
Ctesiphon, though these translations were never published nor per-
haps ever completed. In any case they provided an occasion for
Cicero to re®ect on what is at stake in translation and on what his
speci¤c intentions were.
In the ¤rst of the two passages devoted to this re®ection,14
Cicero begins by referring to the two orators he has translated,
characterizing them as the two most eloquent Attic orators and
noting that the orations translated were speeches that Demos-
thenes and Aeschines delivered against each other. Cicero then in-
dicates his speci¤c intent, the capacity in which he went about
translating these orations: “And I did not translate them as an in-
terpreter but as an orator. . . .” Two things, he says, had to be re-
tained: “. . . keeping the same thoughts [sententia] and the forms,
or as one might say, the ‘¤gures’ of thought. . . .” Retention of the
same thoughts or meaning is required for translation as such in its
classical determination; retention of the ¤gures of thought is sec-
ondary in that it is prescribed by the speci¤c character of the
works, that they are orations, and by Cicero’s speci¤c intent to
translate them as an orator. While thus keeping the same thoughts
13. The Greek designation of speech as fvn} or as t% \n t_ fvn_ is not, then,

just another instance in which speech as such is designated by the name of an
anatomical part indispensable to its production. It is not simply an alternative to
the ancient Hebrew designation of speech as lip or the Latinate and modern des-
ignation of speech as tongue, which is retained etymologically in language. For
unlike the lip and the tongue, voice is not itself an anatomical part but rather
the very guise in which speech occurs. It is only as voice, only in the sounding
of a voice, only in the voicing of names, that speech occurs and so is.
14. Cicero, De Optimo Genere Oratorum V.14 and VII.23.

and ¤gures, he does so, as he adds, “. . . in words that conform to
our usage.” The schema of the classical determination, supple-
mented by the oratorical speci¤city, is clearly in place: the transla-
tion consists primarily in saying in Latin words the same thoughts
as were said in the Greek words of the orators.
The second passage reiterates and extends what is said in the
¤rst. Cicero refers to the virtues of the speeches he has translated
and expresses his hope that the translation retains these virtues.
Three virtues are named: ¤rst, the thoughts; second, the ¤gures of
thought; and third (extending the oratorical speci¤city), the order
of topics. He says that, while retaining these three virtues of the
original, he has proceeded by “following the words only insofar
as they are not abhorrent [non abhorreant] to our usage.” His point
is, on the one hand, the same as in the ¤rst passage: translation is
a matter of carrying the three virtues over into words belonging
to the native way of speaking. But, on the other hand, it is also a
matter of determining just which words to translate as such and
how exactly to translate them, of determining these speci¤cs of
the translation by reference to usage, to what is “not abhorrent to
our usage.” Thus Cicero goes on to say speci¤cally: “if all the
words are not directly translated from the Greek, we have at least
tried to keep them within the same kind [genus].” This reference
to speci¤c translational strategies that would diverge from direct
transfer (from one word to an exact equivalent) is ampli¤ed in the
¤rst passage, which concludes: “And in so doing, I did not hold it
necessary to render [reddere] word for word [verbum pro verbo], but I
preserved the general kind and force of the words [sed genus omne
verborum vimque servavi]. For I did not think I ought to count them
out to the reader like coins, but to pay them by weight, as it were.”
A word-for-word rendering is not necessary as long as the same
thoughts and ¤gures (and perhaps order) are retained, and such
retention, Cicero suggests, is possible provided one preserves the
general kind and force of the words. A word-for-word rendering
is not only super®uous but, he further suggests, not even very de-
sirable (assuming that it is possible). What counts is that the words
be rendered in a way that sustains, not the individual words, but
the sense of what is said, the thoughts. To this end the same gen-
eral kind of words need to be used in the translation, and words
66 On Translation
need to be used that retain the force that the words of the origi-
nal have.
Cicero’s re®ection on translation thus inscribes the classical
schema by which translation consists primarily in carrying the
meaning of a unit in one language over to a corresponding unit
in the other language. The schema does not exclude requirements
speci¤c to the text or to the intent of the translator; neither does
it exclude the various strategies by which units and their limits
would be shifted and syntactical structures transformed in the
course of translation. Yet all these supplementary moments would
continue to be rigorously linked to or governed by the require-
ment that the meaning of what the original says, its thoughts, be
rendered in the translation. Cicero grants that one of the things
needed for such restitution is that the force of the words be pre-
served. In the word translated as force, the word vis, Cicero’s text
retains a trace of the protoclassical determination. But vis is not
d4namiw, even if it translates—yet without translating—d4namiw;
it is only a trace marking the absence of what had once been
thought in the Platonic discourse. With Cicero the word for force
has lost the force that d4namiw once had.
The classical determination of translation is nowhere more
clearly and succinctly presented than in a passage in Locke’s An
Essay Concerning Human Understanding. The passage occurs in Book III,
entitled “Of Words,” and is thus set within the context of Locke’s
general theory of language. The primary moments of this theory
are expressed when Locke, noting that man was by nature fash-
ioned so as to be capable of producing articulate sounds, observes
that man could then make these sounds “stand as marks for the
ideas within his own mind, whereby they might be known to oth-
ers, and the thoughts of men’s minds be conveyed from one to an-
other.”15 Thus, for Locke, two connections are de¤nitive of lan-
guage: it consists of words, which, ¤rst, stand for ideas and which,
second, make it possible for these ideas to be communicated from
one man to another. These two functions of words, to signify
15. John Locke, An Essay Concerning Human Understanding, in vols. 1 and 2 of The
Works of John Locke (London, 1823; reprint, Aalen: Scientia Verlag, 1963), Book III,
chap. i, §2 (in this edition Book III is found in vol. 2).

and to communicate, are intimately linked: because the ideas that
words signify are invisible, internal, they could not be communi-
cated otherwise than by means of the signifying words. Since so-
ciety requires communication, “it was necessary that man should
¤nd out some external sensible signs, whereof those invisible ideas,
which his thoughts are made up of, might be made known to oth-
ers.” Thus it was that words came “to be made use of by men, as
the signs of their ideas.”16
What words announce are thus neither the things nor the
meanings (in the Greek sense, as noht1n) spoken of but rather the
invisible ideas interior to our minds. Locke insists on this connec-
tion, on its exclusivity, even suspending for its sake the question
of representation: “words in their primary or immediate signi¤ca-
tion stand for nothing but the ideas in the mind of him that uses
them, how imperfectly soever or carelessly those ideas are col-
lected from the things which they are supposed to represent.”17
Indeed Locke goes so far as to declare that to make words stand
for anything but the ideas in our minds is a perversion of the use
of words and a cause of obscurity and confusion.
Yet, for all his insistence, Locke privileges the names of simple
ideas so as to allow that they “intimate also some real existence,
from which was derived their original pattern.”18 Locke does not
explain how such intimation would operate, how it not only would
presumably link an idea to some real existence but also, presum-
ably by way of the idea, would link the name of the idea to such
existence.19 Instead, he goes on to develop the thesis that the names
of simple ideas are not capable of any de¤nition. By a de¤nition he
means: “the showing the meaning of one word by several other
16. Ibid., III.ii.1.

17. Ibid., III.ii.2.
18. Ibid., III.iv.2.
19. Locke himself acknowledges in the Essay the dif¤culty of establishing any
connection between ideas and the reality of things. In its most succinct form: if
the mind perceives nothing but its own ideas, how is it to know that these ideas
agree with things themselves? In the case of simple ideas he attempts to address
this dif¤culty by having recourse to nature, to the naturalness of the way in
which things operate on the mind to produce simple ideas. See my discussion
in Force of Imagination, 86–87.
68 On Translation
not synonymous terms.”20 Yet meaning, for Locke, is synonymous
with idea, so that to show the meaning of a word is to elicit by
other words the idea signi¤ed by the word. To give a de¤nition of
a word is, then, to offer a series of words not synonymous with it
that are capable of eliciting the very idea that the word stands for.
Hence, Locke’s demonstration that the names of simple ideas are
incapable of being de¤ned: “the several terms of a de¤nition, sig-
nifying several ideas, they can all together by no means represent
an idea, which has no composition at all: and therefore a de¤ni-
tion, which is properly nothing but the showing the meaning of
one word by several others not signifying each the same thing, can
in the names of simple ideas have no place.”21 Since a de¤nition
must consist of several words, each signifying an idea, a de¤nition
can signify only a composite of the several ideas, not an incom-
posite, simple idea. Locke seems to have no doubts about the one-
to-one correlation assumed to hold between word and idea; one
wonders whether his con¤dence might have been disturbed by, for
instance, the de¤nition of shape (sx+ma) that Socrates offer in the
Meno, that (in translation) “shape is the only thing found always
following color.”22
Locke’s con¤dence would seem, on the other hand, to be bol-
stered by the examples that he goes on to mention, examples in-
tended to demonstrate the futility of attempting to give de¤nitions
of simple ideas. It is precisely in this context that he comes to refer
to translation. In his ¤rst example only the word is lacking: he
supposes a situation in which a Dutchman is asked what beweeginge
means—that is, one may say, a situation in which the Dutchman
is asked to de¤ne beweeginge, a de¤nition that cannot but be a trans-
lation, indeed even if it should fail to be a proper de¤nition. The
translation—one will note how questionable it is, especially if one
recalls the Greek original—might be in English: “ ‘The act of a be-
ing in power, as far forth as in power.’ ” Or it might be in Latin:
“ ‘actus entis in potentia quatenus in potentia.’ ”23 Locke has only scorn
20. Locke, Essay, III.iv.6.

22. Plato, Meno 75b.
23. Locke, Essay, III.iv.8.

for such alleged de¤nitions, such “exquisite jargon.” Suppose that
someone were to receive such a de¤nition: “I ask whether any one
can imagine he could thereby have understood what the word
‘beweeginge’ signi¤ed, or have guessed what idea a Dutchman or-
dinarily had in his mind, and would signify to another, when he
used that sound.”24 These cannot, in Locke’s terms, be de¤nitions
of this name of a simple idea, of the name beweeginge, though they
are translations, even if bad ones.
The most signi¤cant passage concerning translation occurs as
Locke turns to another example in which alleged de¤nitions prove
to be, not de¤nitions at all, but only translations. He refers to
the atomists’ alleged de¤nition of motion as a passage from one
place to another. Here the problem is that one word (passage) re-
places another (motion) with which it is synonymous, with which
it shares the same meaning. In the alleged de¤nition of motion,
what occurs is motion from one word to another with the same
meaning, that is, circulation from one to the other by way of the
common meaning. Thus Locke declares: “This is to translate, and
not to de¤ne, when we change two words of the same signi¤ca-
tion one for another.”25 One notices that Locke does not restrict
translation to transferal between different languages: whether one
substitutes for motion the Latin motus or the English passage, it re-
mains a matter simply of translation. In both cases it is a matter
of movement across a difference, either within a language or be-
tween languages; in this movement from one word to another, the
meaning—for Locke, the idea—is both what is carried over and
what makes the movement as such possible.
Locke thus reiterates quite precisely the classical determina-
tion of translation as transition or transfer from one word to an-
other by way of circulation through the common meaning. This
reiteration is especially remarkable in view of the philosophical re-
moteness of Locke’s work from the Greek context in which this
determination was forged; there is perhaps no better index of this
24. Ibid.
70 On Translation
remoteness than the difference separating what the Greeks called
di$noia—and also cd)a—from those internal objects of thought
that Locke called ideas. What is remarkable is that the classical de-
termination could continue to govern the concept of translation
even across this enormous difference.
Gadamer’s hermeneutical discussion of translation reaf¤rms the
classical determination yet also underlines the limit, the incom-
pleteness, of translational restitution of meaning, as well as a cer-
tain inevitable distortion produced by translation. The discussion
is restricted to translation from one language to another, to trans-
lation as passage of meaning across the differential interval be-
tween two languages. The discussion is also largely strategic: it is
oriented to elucidating the conditions of understanding as such, to
elucidating these conditions by focusing on situations in which
understanding is disrupted or made exceptionally dif¤cult, as, for
instance, in the case of linguistic difference. Just as a broken tool
can serve to light up the situation in which tools otherwise func-
tion normally, so can the breakdown of communication and the
resulting need for translation serve to illuminate the situation in
which, otherwise, one converses with another or reads a text. Yet,
while thus drawing out the parallels with translation that serve to
illuminate the character of a conversation in which two persons
come to an understanding, Gadamer also, if more subtly, lets this
orientation to conversation recoil upon translation in such a way
as to elucidate it along the lines of the classical determination
while also exposing the limits and the distortion that have the ef-
fect of compromising this determination, of beginning to under-
mine it.
To an extent Gadamer grants the restitution or preservation
of meaning that is central to the classical determination. Yet he
stresses equally that translation, in preserving the meaning, trans-
poses it into a different context. Here is his formulation in Truth
and Method: “Here the translator must carry the meaning to be un-
derstood over into the context in which the interlocutor lives. This
is not of course to say that he is at liberty to falsify the meaning
intended by the speaker. Rather, the meaning is to be preserved,
but, since it is to be understood in a new language world, it must

establish its validity therein in a new way.” Gadamer concludes:
“Thus every translation is already interpretation.”26 One could say:
the translator not only must intend the meaning and keep that in-
tention in force so that the meaning is preserved in the translation
but also must interpret the meaning so as to be able to set it in the
context of the other language, to express it in the new language
world in such a way as to establish it as a valid meaning within
that world. Because the meaning must be ¤tted to the new context,
installed within that context, it can never suf¤ce for the would-be
translator of a text only to reawaken the original psychic processes
of the writer, that is, the complex of meaning-intentions borne by
the original text. Rather, as Gadamer says, the translation of a text
is a text formed anew, eine Nachbildung. Only through such Nachbilden
can what is meant in a text be carried over into the context of
another language.
And yet, every translation is like a betrayal; it is a kind of trea-
son committed against the original text. This is what Gadamer says
in a text from 1989 entitled “Lesen ist wie Übersetzen,”27 in which
his theme is not so much translatability as untranslatability. Yet this
theme is already broached, if less emphatically, in Truth and Method:
the translator, Gadamer says, must make a “constant renunciation”
because, however faithful his translation may be, he cannot over-
come the gulf between the two languages so as to close completely
the gap between original and translation. There are always junc-
tures where no smooth transposition is possible, where in order
to emphasize one feature of the original—that is, to carry it over
to the translation—other features must be played down or even
suppressed. Translation occurs, then, says Gadamer, as a highlight-
ing (eine Überhellung). Hence, on the one hand, a translation that
takes its task seriously is always clearer than the original: expres-
sions that in the original remain ambiguous, that bear manifold
meanings, must be resolved by the translator into univocal expres-
sions in all but those few fortunate instances in which the language
26. Gadamer, Wahrheit und Methode, in vol. 1 of Gesammelte Werke, 387–88. In link-
ing translation to interpretation, Gadamer echoes Heidegger, who writes: “But
every translation is already interpretation” (Was Heisst Denken?, 107).
27. Gadamer, “Lesen ist wie Übersetzen (1989),” in vol. 8 of Gesammelte Werke, 279.
72 On Translation
of the translation offers expressions comparably ambiguous or mani-
fold. But such resolution entails, on the other hand, that the trans-
lation is also ®atter than the original, that—in Gadamer’s phrase—
“it lacks some of the overtones that vibrate in the original.”28
Even in Schlegel’s translation of Shakespeare such ®atness is not
entirely lacking; even in such masterly translation, highlighting
cannot but have occurred with a resulting loss of some of the
overtones that vibrate in Shakespeare’s original text. For example,
near the beginning of A Midsummer Night’s Dream there is an ex-
change between the lovers Lysander and Hermia, whose love has
just been interdicted by Theseus at the urgings of Hermia’s father.
The exchange has to do with the dif¤culties that true love ever
encounters. It issues in a certain resolve, expressed by Hermia:
If then true lovers have been ever cross’d,
It stands as an edict in destiny.
Then let us teach our trial patience,
Because it is a customary cross,
As due to love as thoughts and dreams and sighs,
Wishes and tears, poor fancy’s followers.
(I.i.150–55)
Here is Schlegel’s translation:

Wenn Leid denn immer treue Liebe traf,
So steht es fest im Rate des Geschicks.
Drum lass Geduld uns durch die Prüfung lernen,
Weil Leid der Liebe so geeignet ist
Wie Träume, Seufzer, stille Wünsche, Tränen,
Der armen kranken Leidenschaft Gefolge.
In the translation three instances of highlighting can be marked.

In each instance the manifold sense of the original is resolved into
a more nearly univocal sense, rendering the translation thus ®atter
than the original, robbing the text of some of its overtones.
The ¤rst instance has to do with that which true lovers ever
encounter. Shakespeare’s text calls it being “cross’d.” No doubt the
28. Gadamer, Wahrheit und Methode, 390.

word alludes to cruci¤xion, hence to pain, to suffering. But to be
“cross’d” also means to be opposed by someone or by some force,
to meet opposition that hinders what one desires or intends. It
means also, consequently, to be frustrated by such hindering of
one’s intent. Such opposition and the resulting frustration are pre-
cisely what Lysander and Hermia have experienced. Yet Schlegel
resolves these manifold senses into a single one: what true lovers—
or rather, true love (treue Liebe)—encounter or undergo is pain and
suffering (Leid). And what Shakespeare’s text calls “a customary
cross” becomes in the translation the suffering that is proper to
love. In another passage, just a few lines earlier, in which Hermia
refers to the cross:
O cross! too high to be enthrall’d to low[,]
(I.i.136)
the word is rendered as Qual (agony, suffering). Here too, then, the
translation produces a ®attening and even a certain literalizing of
the original.
In the second instance the ®attening is produced in a different
way. It occurs in the rendering of the names of those other things
that are as proper to love as is being cross’d. Shakespeare’s text
lists ¤ve such things: thoughts, dreams, sighs, wishes, and tears.
Schlegel’s translation lists only four, omitting thoughts; it also si-
lences the wishes of the lovers (stille Wünsche). Consequently, it ori-
ents the description of love’s deeds to mute passion; this is one
possibility in Shakespeare’s text but by no means the only one.
The reductions culminate in the third instance. It is a matter of
naming that which all those things proper to love follow, that
which leads and governs them. In Shakespeare’s text its name is
fancy. It is called poor fancy, though not because it is weak or im-
potent but rather quite the contrary: because it is so powerfully
operative and yet is deprived of that which it envisions, cross’d in
its imagined intent. Schlegel’s translation, on the other hand, calls
this leader of love’s deeds by the name already suggested by the
previous reductions: it is a Leidenschaft (passion) that not only is
poor (arm) but also is ailing, ill (krank), presumably because of its
lack of satisfaction. The reduction is here especially decisive: by
replacing fancy with the translation of passion, Schlegel obliterates
74 On Translation
the all-important reference to phantasy and imagination, which
¤gure so thoroughly in the play as a whole. It is because lovers’
deeds are governed by phantastical vision that their loves can be
cross’d, as when Puck squeezes the magic juice on their eyelids. It
is precisely such vision that becomes thematic in Theseus’ speech
on imagination at the beginning of the ¤nal Act. In the operation
of the lover’s fancy, passion is no doubt involved, is generated in
the sight—actual or imaginary—of the beloved. But fancy thor-
oughly exceeds mute passion. It is also the creative vision that, as
Theseus says,
. . . bodies forth
The forms of things unknown. . . .
(V.i.14–15)
And it can command a power of speech that
. . . gives to airy nothing

A local habitation and a name.
(V.i.16–17)
The risk of inordinately ®attening and distorting a text in its

translation is all the greater when the difference between the two
languages is more extreme. In such cases the difference between
various translations of an original into one and the same language
or into closely related languages is also likely to be greater. This
difference between various translations is especially striking when
it is found in translations of classical texts that have undergone
multiple retranslation over a considerable span of time as well as
the stabilizing effect that extensive commentary and interpreta-
tion can have.
The ¤rst sentence of Plato’s Phaedo is sounded in the voice of
Echecrates, a citizen of Phlius, a city in the Peloponnesus to which
the news about the details of Socrates’ death had not yet traveled.
Echecrates puts his question to Phaedo, who has recently arrived
from Athens. Echecrates wants to know whether Phaedo was pres-
ent at the scene of Socrates’ death or whether he heard about it
from someone else. When Phaedo responds that he was there him-
self, Echecrates asks him to tell about what was said and done

there. It is this narrative by Phaedo that constitutes almost the en-
tire dialogue.
Echecrates’ opening question reads as follows: “A[t1w, o
Fa,dvn, pareg)nou Svkr$tei \ke,n+ t_ =m)rf, " t2 f$rmakon
Tpien \n t! desmvthr,~, [ Wllou tou Rkousaw;” Hackforth’s
English translation is fairly typical: “Were you there yourself,
Phaedo, with Socrates on the day when he drank the poison in the
prison, or did you hear the story from someone else?”29 Also fairly
typical is Schleiermacher’s German rendering: “Warest du selbst,
o Phaidon, bei dem Sokrates an jenem Tage, als er das Gift trank
in dem Gefängnis, oder hast du es von einem andern gehört?”30
In these translations of this sentence there are two somewhat dis-
tinct kinds of highlighting at work; both have the effect of ®atten-
ing the text, of closing off possibilities, of resolving multiplicities
that remain intact as such in the original.
One such translational operation is exercised on the phrase
" t2 f$rmakon Tpien (“when he drank the poison,” according
to the typical translations). What is at stake here is the word
f$rmakon. Certainly the word can mean poison. Typical transla-
tions take it for granted that since at the end of the dialogue Soc-
rates dies from having drunk the f$rmakon the word simply des-
ignates poison. But in fact the word also can mean drug, medicine,
remedy; f$rmakon n1sou means a medicine or remedy for a disease.
If one is attentive to Socrates’ ¤nal words about a debt owed to
Asclepius, the physician god to whom it was customary to sacri¤ce
a cock upon recovering from an illness or disease, then it hardly
seems outrageous to keep this second signi¤cation (as medicine,
29. Plato, Phaedo, trans. R. Hackforth (Indianapolis: Bobbs-Merrill, n.d.), 27.

Fowler’s version is similar: “Were you with Socrates yourself, Phaedo, on the day
when he drank the poison in prison, or did you hear about it from someone
else?” (Plato, Euthyphro, Apology, Crito, Phaedo, Phaedrus, trans. H. N. Fowler, Loeb Clas-
sical Library [Cambridge, Mass.: Harvard University Press, 1914], 201).
30. Platon, Phaidon, trans. F. Schleiermacher, in vol. 4 of Sämtliche Werke (Frank-
furt a.M.: Insel, 1991), 191. Zehnpfennig’s version is similar: “Warst du selbst,
Phaidon, bei Sokrates an jenem Tag, als er im Gefängnis das Gift trank, oder hast
du es von einem anderen gehört?” (Platon, Phaidon, trans. Barbara Zehnpfennig
[Hamburg: Felix Meiner, 1991], 3).
76 On Translation
remedy) in play, as the word f$rmakon does but as the typical
translation poison does not. Indeed one can easily ¤nd translations
that—based on the con¤dence that the main point is just Socrates’
execution—obscure both senses. Thus Tredennick simply renders
the phrase " t2 f$rmakon Tpien as: “when he was executed”31—
dropping all reference to the f$rmakon and to Socrates’ drinking
it, ®attening the phrase to the point of nonrecognition. But there
is still a third sense of the word f$rmakon: an enchanted potion or
philtre and, linked to this, a charm or enchantment. If one anticipates
the preoccupation, expressed later in the dialogue, with charming
away the fear of death, then it is less than outrageous to suppose
that even this third sense is in play when the word f$rmakon oc-
curs in the opening sentence. What is needed is a translation that
retains all three senses that are in play in the Greek word. The
translation by Brann, Kalkavage, and Salem comes closest to ¤lling
this need by rendering f$rmakon as potion.32
The second kind of highlighting reduction that one ¤nds in
translations of the opening sentence of the Phaedo has to do, not
with the multiplicity of meanings of words, but with syntax, spe-
ci¤cally with word order. In general it is well known that in trans-
lating a sentence from classical Greek into a modern European lan-
guage, one is usually compelled to alter the word order quite
thoroughly in order to produce a translation that not only is ®uent
but also expresses the meaning of the sentence as a whole. But it
is also known—if less widely—that in most Platonic dialogues the
very ¤rst sentence is among the most signi¤cant, in many cases
announcing a theme, a question, or a directionality that governs
the entire dialogue. In some cases this announcement is borne
primarily by the very ¤rst word, as with the word kat)bhn (“I
went down”) at the very beginning of the Republic.33 The Phaedo is
31. Plato, Phaedo, trans. Hugh Tredennick, in The Collected Dialogues of Plato, ed.
Edith Hamilton and Huntington Cairns (Princeton: Princeton University Press,
1961), 41.
32. Plato, Phaedo, trans. Eva Brann, Peter Kalkavage, and Eric Salem (Newbury-
port, Mass.: Focus Publishing/R. Pullins Company, 1998), 27.
33. See my discussion in Being and Logos, 313–20.

also such a case. With its very ¤rst word, a[t1w, it announces the
question that will occupy the entire dialogue. At the outset the
question is enacted in this initial word: What is spoken of when,
addressing Phaedo, Echecrates says, or rather asks about, yourself?
What is the self of Phaedo or of any person? Is it the soul or the
body or both? More generally, what does it mean for something
to be itself? In the most rigorous sense it means not being any-
thing other than itself, being identical with itself, being the same
as itself; indeed the word a[t1w can also mean, in a certain syn-
tactical connection: same. The ¤rst word of the dialogue thus al-
ludes to a kind of being that is the same as itself; when the dia-
logue, at several crucial junctures, comes to speak of such beings,
the names that will be used for them are cd)a and eqdow.
With the ¤rst sentence of the Phaedo, at least with the beginning
of the sentence, there is need, then, to preserve the word order,
even if, as in the typical translations cited above, altering the word
order produces a smoother, more ®uent sentence in English or
German. Again it is the translation by Brann, Kalkavage, and Salem
that ful¤lls this need: “You yourself, Phaedo—were you present
with Socrates on that day when he drank the potion in the prison,
or did you hear from somebody else?”
Once Phaedo has explained what caused the long delay of
Socrates’ execution, Echecrates is keen to hear all about what
took place at the scene of Socrates’ death. He implores Phaedo to
give a full and exact report: “Ta6ta d| p$nta proyum}yhti qw
saf)stata =m.n #pagge.lai, . . .”34 Fowler translates: “Be so
good as to tell us as exactly as you can about all these things, . . .” 35
Hackforth renders it: “Well, please do your best to give us a reli-
able report, . . .”36 Schleiermacher: “Alles dieses bemühe dich doch
uns recht genau zu erzählen, . . .”37 However, with phrases such
as “Be so good as to,” “Do your best to give us,” and “Bemühe
dich doch,” these translations all fail to keep open the sense of
34. Plato, Phaedo 58d.

35. Plato, Euthyphro,etc., trans. H. N. Fowler, 203.
36. Plato, Phaedo, trans. R. Hackforth, 28.
37. Platon, Phaidon, trans. F. Schleiermacher, 193.
78 On Translation
proyum}yhti, from proyum)omai. While the word can indeed
be said to mean to be ready, willing, to do a thing, such a render-
ing leaves out of account the inclusion in this word of the word
yum1w, which can mean heart, soul, spirit and which in the Republic
is the name given to that part of the soul that mediates between
calculation and desire. Only the translation by Brann, Kalkavage,
and Salem manages to retain this component in what otherwise
seems a rather commonplace word: “Well, put your heart into giv-
ing us as sure a report as you can about all these things, . . .”38
If even the most masterful translations will always have been
compelled by the force of linguistic difference to choose between
signi¤cations that in the original are intact in their multiplicity,
then translation will always involve loss.39 By highlighting certain
signi¤cations, translation will necessarily—with a necessity en-
forced by linguistic difference—reduce or even obliterate others,
reducing them to mere overtones or silencing them altogether. In
translation something of the original is lost; this is why it never
suf¤ces to translate a text into one language by translating its trans-
lation into another language. The loss incurred by translation is
not, however, pure expenditure but rather, at the very least, will
be situated within an economy in which the loss is compensated
for by certain gains in another dimension. In translating a text one
may come to a more adequate and detailed understanding of it de-
spite the necessary reduction, or rather, in many instances, be-
cause the necessity of reduction, the operation of linguistic differ-
ence, puts in relief features of the original that would otherwise
go unnoticed. There are indeed some texts, for instance, the frag-
ments of the early Greek philosophers, that one could not interpret
without also engaging in translational operations. Yet there is a re-
turn, a compensatory gain, not only for the translator but also for
the reader capable of reading the text only in translation. In this
38. Plato, Phaedo, trans. Brann, Kalkavage, and Salem, 28.

39. It goes almost without saying that the extent and signi¤cance of the loss
depend on the character of the text. With technical and business communica-
tions the loss is minimal and may be of no signi¤cance at all. With literary and
philosophical texts, on the other hand, the loss is seldom insigni¤cant.

connection the loss of signi¤cation would be the price of extend-
ing the range of communication.40
Yet loss there is. Indeed declarations and attestations of such loss
through translation are ubiquitous. The case of poetry is most fre-
quently and most vehemently invoked, even to the point of its be-
ing simply declared untranslatable.41 Such a declaration is recorded
40. While it is necessary, especially as a bulwark against commodi¤cation, to

place all considerations of the relation of translation to the reader within the
framework of the economy just sketched, Benjamin goes too far in dismissing
this relation as irrelevant to the understanding of translation and even as mis-
leading. Observing that what is essential to a poetic work (Dichtung) is neither
communication (Mitteilung) nor statement (Aussage), Benjamin argues that a trans-
lation determined as conveying (vermitteln) something to the reader would—at
least in the case of a poetic work—convey nothing but a communication (Mit-
teilung), hence something quite unessential. What is essential to a poetic work,
according to Benjamin, is what it contains in addition to mere communication:
“the unfathomable, mysterious, ‘poetic’.” He concludes that as long as translation
is considered as serving the reader, that is, as merely conveying a communica-
tion to the reader, it will remain bad translation, that is, “a vague transmission
[ungenaue Übermittlung] of an unessential content” (Benjamin, “Die Aufgabe des
Übersetzers,” 50). In this account there are two assumptions that, if questioned,
put also in question Benjamin’s thesis that translation is not for the reader, a the-
sis that he extends also to the original of a poetic work, thus maintaining that
“no poem is for the reader [kein Gedicht gilt dem Leser].” What is, ¤rst of all, as-
sumed in the entire discussion is an abstract distinction between communica-
tion (i.e., what can be communicated or conveyed) and the poetic, as if a com-
munication could not be conveyed precisely in such a way as also to bear the
poetic along with it. This points in turn to the second assumption, namely, a
very traditional and highly sedimented concept of the artwork as basically a
thing upon which a poetic moment is grafted, that is, in the case of Dichtung,
communicable statements endowed with a mysterious, poetic signi¤cance.
41. The great exception is Hegel, who maintains not only that it is a matter
of indifference whether a poetic work is read silently or heard aloud but also
that such a work “can even be translated into other languages without essential
detriment to its value.” It can even—without detriment, he implies—be “turned
from poetry into prose” (Hegel, Ästhetik, ed. Friedrich Bassenge [West Berlin:
das europäische buch, 1985], 2:331).
Hegel’s assertion of such a reign of translatability precisely where it would
least be expected, in poetry, is a direct consequence of the position he takes
regarding the proper sensible element of poetry. As a form of art— even as the
highest form—poetry must have a sensible element; for art as such is the sensible
80 On Translation
by Boswell as having been put forth by Samuel Johnson in a con-
versation of 11 April 1776.42 Boswell initiates the discussion by con-
fessing his own inability to de¤ne translation or to illustrate what
it is by means of a similitude; he suggests nonetheless that its ap-
plication to poetry is limited, that “the translation of poetry could
be only imitation.” Johnson responds by granting that books of
science can be translated exactly and that history too, except inso-
far as it is poetical, admits of translation. He continues: “Poetry,
indeed, cannot be translated; and, therefore, it is the poets that
preserve languages; for we would not be at the trouble to learn a
presentation of the true. But whereas one might take spoken or written words
as comprising this sensible element, corresponding to the stone, color, and tone
of architecture, painting, and music, Hegel insists that the proper element of
poetry is inner representation and intuition itself (das innere Vorstellen und An-
schauen selbst). As the painter uses color in order to present something, so the
poet shapes one’s inner representational powers so that one comes to intuit in-
wardly that which the poet would present. Speech, which might otherwise be
taken as the sensible element in poetry, Hegel considers a mere sign from which
one withdraws at the very start; speech exhausts itself in its capacity as a mere
sign, and the sensible character of speech is not carried over to the poetic work
itself; as mere sign, speech does not determine—but only communicates—the
poetic work. Thus the work remains unaffected by shifts from one system of
signs to another, that is, by translation.
Hegel’s concept of the poetic work is grounded in the thesis, central to the
Aesthetics, that art as such is essentially past. Indeed this essential pastness is pre-
eminently displayed in poetry, in which the proper sensible element becomes a
spiritual form (intuition), while the apparent sensible element (speech) proves
to be a dispensable, external sign. In Hegel’s words: “Precisely at this highest
stage, art now transcends itself, in that it forsakes the element of a reconciled
embodiment of the spirit in sensuous form and passes over from the poetry of
representation to the prose of thought” (ibid., 1:94).
Any confrontation with Hegel’s assertion of the unlimited translatability of
poetry, confronting this thesis with the almost ubiquitous testimony to the con-
trary, would have to engage the fundamental position of Hegel’s Aesthetics as a
whole. Here it must suf¤ce merely to formulate a question from which such an
engagement might commence—namely: Can the power of inner representation
or intuition, which Hegel identi¤es as imagination (Phantasie), operate in essen-
tial detachment from speech? Or is its allegedly spiritual character necessarily
contaminated, as it were, by the sensible character of speech?
42. James Boswell, Life of Johnson (London: Oxford University Press, 1966), 742.

language, if we could have all that is written in it just as well in a
translation. But as the beauties of poetry cannot be preserved in
any language except that in which it was originally written, we
learn the language.” Clearly Johnson has in mind the classical lan-
guages, since it is only the so-called dead languages that require
such preservation, that can live on only through those who learn
these languages in order to read the poetry, which in turn is pre-
served in its beauty only within the original language.
And yet, Johnson’s seemingly categorical denial of the possi-
bility of translating poetry is moderated by what he says in other
connections. For instance, in a conversation dated 9 April 1778,43
there is mention of a recent translation of Aeschylus, a translation
that is praised by one of Johnson’s interlocutors but that he thinks
is little more than verbiage. Asked to reconsider by reading one of
the plays in this translation, Johnson enunciates the standard by
which he will judge it: “We must try its effect as an English poem;
that is the way to judge of the merit of a translation.”
Johnson’s further remark that “Translations are, in general, for
people who cannot read the original” could suggest that he places
little value on the enterprise of translation. And yet, two decades
earlier Johnson had written in a rather different tone about trans-
lation. In The Idler44 he introduces his history of translation with
the declaration that of all the studies undertaken in the past three
centuries “none has been more diligently or more successfully
cultivated than the art of translation.” Though he mentions some
Roman translations of Greek poetry and grants that the Arabs “felt
the ardour of translation,” he regards translation as primarily
something modern. He traces brie®y the history of translation in
England, marking the opposite extremes that for the most part had
prevailed. From the time of Chaucer until that of the Restoration,
the translations produced were, with few exceptions, strictly lit-
43. Ibid., 920–21.

44. Samuel Johnson, The Idler, in vol. 2 of The Yale Edition of the Works of Samuel
Johnson, ed. W. J. Bate, John M. Bullitt, and L. F. Powell (New Haven: Yale Univer-
sity Press, 1963), 211–17 (no. 68, Saturday, 4 August 1759; no. 69, Saturday, 11 August
1759).
82 On Translation
eral. Johnson refers to the case of Caxton, whose translations from
the French are said to have followed the original so scrupulously
that they were barely English: “tho’ the words are English the
phrase is foreign.” With the Restoration, translators threw off the
yoke of such “servile closeness”; yet the freedom, even licentious-
ness, of their translations did little more than veil “their want of
learning behind the colours of a gay imagination.” At best, these
translations proved—for all their mistakes and negligence—more
delightful to the reader.
Johnson’s history of translation, displaying the opposite ex-
tremes of literalism and freedom, serves primarily to set the stage
for the appeal to a mean, with which Johnson concludes: “There
is undoubtedly a mean to be observed. Dryden saw very early that
closeness best preserved an author’s sense, and that freedom best
exhibited his spirit; he therefore will deserve the highest praise
who can give a representation at once faithful and pleasing, who
can convey the same thoughts with the same graces, and who when
he translates changes nothing but the language.” Thus, following
the classical determination, Johnson in effect identi¤es the mea-
sure of translation as restitution both of meaning (“thoughts”)
and of form or style (“graces”). It is because a translation ought
to change nothing but the language that the merits of an English
translation of a poem are to be judged by trying its effect as an
English poem.
Even if one takes Johnson at his word, that “Poetry, indeed,
cannot be translated,” virtually all else that he says of translation
appears to construe this impossibility as an impossibility of com-
plete restitution. The restitution of thoughts and of graces will in-
evitably be limited, and the translator of poetry will always nec-
essarily have changed more than just the language. Still, granted
the limit, there can be excellence, even greatness, in the transla-
tion of poetry. In the conversation of 9 April 1778 concerning En-
glish translations of Greek classics, Johnson is asked about Pope’s
translation of Homer and declares: “Sir, it is the greatest work of
the kind that has ever been produced.”
From the perspective of modern linguistics, Jakobson is equally
insistent that poetry cannot—at least by a certain measure—be

translated. In his words: “Poetry by de¤nition is untranslatable.
Only creative transposition is possible.”45 Heidegger goes still fur-
ther and extends such virtual untranslatability to thinking as well
as poetry—in a remark itself barely translatable, requiring trans-
position: “Thinking can no more be translated than can poetry.”46
One can agree that most poetry and many philosophical texts are
untranslatable if untranslatable means precisely not translatable
without loss of signi¤cation, even perhaps very signi¤cant loss of
signi¤cation. One can agree that such kinds of texts are untrans-
latable if untranslatable means that the translator will never suc-
ceed in changing nothing but the language, in effecting a pure
transition from one language to the other. Yet translations there
are, and though they may never be without loss, the loss and the
reduction and distortion it can produce are not such as to dis-
entitle these translations altogether. Not, at least, in the most
fortunate cases: for Schlegel has produced translations of Shake-
speare’s poetry, as Schleiermacher and others have produced trans-
lations of Platonic dialogues, and as Heidegger himself has pro-
duced translations of passages from Greek philosophical texts. For
the most part—and certainly in such exceptional instances—it
is not a matter of untranslatablilty in an unconditional sense, as
though any attempt at translation would inevitably fail to produce
anything that could even be deemed a translation. Rather, in every
instance it is a matter of a reexpression that can—and often does
—succeed to some degree, but—at least in the case of poetry—
always only to some degree. Reexpressing a text in another lan-
guage world requires resolving certain multiple meanings, trans-
posing various syntactic structures, and shifting from particular
45. Jakobson, Language in Literature, 434.

46. “So wenig wie man Gedichte übersetzen kann, kann man ein Denken
übersetzen” (Heidegger, “Spiegel-Gespräch,” in Antwort: Martin Heidegger im Ges-
präch, ed. G. Neske and E. Kettering [Pfullingen: Günther Neske, 1988], 108). An-
other passage simply declares poetry untranslatable: “Translation and translation
are not the same if it is a matter, on the one hand, of a business letter and, on
the other hand, of a poem. The one is translatable; the other is not” (Heidegger,
Der Satz vom Grund, 163).
84 On Translation
metaphorics valid in the one language to other metaphorics valid
in the other language, hence reexpressing certain ¤gures of the
original within an alien metaphorics appropriate to the other lan-
guage. Such reductions, transpositions, and shifts result in loss of
signi¤cation and of syntactical and metaphorical force, and this is
why translation of poetic or philosophical texts can succeed only
to some degree. It is in this sense that such texts may be called
untranslatable.
Such texts involve, then, a certain untranslatability, which is at-
tested by the loss sustained in translating them. In some cases a
loss can be discerned with remarkable clarity in the translation of
proper names. Such names are as such meaningless, and for the
most part they function in discourse without becoming meaning-
ful. Even when, as in A Midsummer Night’s Dream, a name such as
Theseus invokes a historical or mythical character, it does so by
referring to something singular rather than by signifying a general
meaning. It is for this reason that a proper name does not belong
to a particular language with the same insistence that other words
do; and it is also for this reason that a translator between languages
as closely related as English and German can simply carry over
many of the proper names unaltered. Except for their historical or
mythical associations, these names have virtually no relation to
meaning and thus have little bearing on the meaning, the complex
of meaning, that the translator is to preserve in the translation. Yet
there are exceptions, certain names that both name and signify
and that have therefore a marginal or oblique bearing on the com-
plex of meaning. In some cases the proper name signi¤es by co-
inciding with some generic designation: as with the characters
Mustardseed, Wall, Moonshine, and Lion. In such instances the
translation can straightforwardly render the generic designations:
as does Schlegel in rendering these as Senfsamen, Wand, Mond-
schein, and Löwe. But there are other names that are related to
meaning in more subtle and complex ways; these are the tag-
names that Shakespeare weaves into the fabric of his plays in such
masterly ways. In such cases the name not only indicates an indi-
vidual character but also through its meaning can bear, for in-
stance, on the very character that it also names. In A Midsummer

Night’s Dream a prime instance is the name Bottom. Bottom is a
weaver, and his name in effect signi¤es his profession, since in
Elizabethan English the word bottom designates a kind of frame
used in weaving. Both the identi¤cation of Bottom as a weaver and
the allusion of his name to weaving anticipate his deeds in the
play, not only that of weaving words together in a peculiar way
(as when he says: “the eye of man hath not heard, the ear of man
hath not seen, man’s hand is not able to taste, his tongue to con-
ceive, nor his heart to report, what my dream was” [IV.i.209–12])
but also that of weaving together the human world and the fairy
world, as when he is whisked off to the bower of the fairy queen.
Through another of its meanings the name Bottom associates the
character it names with the ass whose head he acquires through
Puck’s mischief, with the ass’s head that, atop Bottom’s body, pro-
vokes Peter Quince’s exclamation:
Bless thee, Bottom, bless thee! Thou art translated.
(III.i.113–14)
Translated he is indeed in Schlegel’s translation of the play. When
Nick Bottom becomes Klaus Zettel, his name retains, to be sure,
its reference to his profession as a weaver (in the various ways he
carries it on in the play), for the word Zettel, like Bottom, alludes as
a common noun to weaving. On the other hand, the translation
effaces the direct association of the name with the ass’s head. The
double meaning with which the proper name Bottom is associated
when taken as a common noun does not survive the translation
and the resolution it requires. In this regard a signi¤cant semantic
component is lost in the translation.
Other losses can occur in translation, loss with respect to other
moments of the original discourse. Nietzsche stresses that one of
the things most easily lost in translation is the tempo of the style of
the original. Nietzsche says that “there are honestly meant trans-
lations that . . . are almost falsi¤cations of the original, merely be-
cause its bold and merry tempo (which leaps over and obviates all
dangers in things and words) could not be translated.” In this con-
nection Nietzsche celebrates, at the expense of Germans and their
language, the very language in which nonetheless Nietzsche writes
86 On Translation
the celebration—Nietzsche celebrates Machiavelli: “How could
the German language . . . imitate the tempo of Machiavelli, who in
his Principe lets us breathe the dry, re¤ned air of Florence and
cannot help presenting the most serious matters in a boisterous
allegrissimo. . . .” Above all, Nietzsche celebrates Aristophanes, in-
deed to such an extent that he forgoes even broaching the question
of the impossible translation of Aristophanes’ tempo into German
(which he calls, in a performative contradiction, “ponderous, vis-
cous, and solemnly clumsy”); instead he veers off toward another,
very heterogeneous kind of translation—if one can still call it
that—by Aristophanes’ great contemporary and rival: “And as for
Aristophanes—that trans¤guring, complementary spirit for whose
sake one forgives everything Hellenic for having existed, provided
one has understood in its full profundity all that needs to be for-
given and trans¤gured here—there is nothing that has caused me
to meditate more on Plato’s secrecy and sphinx nature than the
happily preserved petit fait that under the pillow of his deathbed
there was found no ‘Bible,’ nor anything Egyptian, Pythagorean, or
Platonic—but a volume of Aristophanes. How could even Plato
have endured life—a Greek life he repudiated—without an Aris-
tophanes?”47
There are, then, various kinds of losses, losses that can be—
and often are—undergone with respect to various moments of
discourse. Several have been marked: loss of multiple meanings
through resolution that retains only some while excluding others;
loss through transposition of syntactic structures, which can ob-
literate, for instance, the signi¤cance that a certain word order
has for the discourse; loss of metaphorical forcefulness as a result
of the necessity of shifting from metaphorics valid in the origi-
nal language to those valid in the other language; and loss of the
tempo of the style of the original. These various kinds of loss do
not for the most part operate independently. Loss of semantic com-
ponents, for instance, may prove to be precisely what necessitates
47. Nietzsche, Jenseits von Gut und Böse, in vol. VI 2 of Werke: Kritische Gesamtaus-
gabe, §28.

a shift of metaphorics and the loss this entails. And the transposi-
tion of syntactic structures required by the language into which a
translation is made can result in a loss of the tempo of the original.
Loss there is indeed, or rather, various kinds of loss, loss with
respect to various moments of the original discourse. And yet, the
question imposes itself, a question that, it seems, needs especially
to be posed today so as to mark the limit of—if not to undo—a
certain pathos of the end of philosophy, the question whether in
translation there is only loss, whether even in the translation of
Greek philosophical language into Latin and hence into the mod-
ern European languages there was only loss. Or whether, at the
very least, this translation served to open possibilities in the Latin
language that would otherwise never have been offered, possibili-
ties in turn passed on in some degree to the modern European
languages. It is a question of whether, in Schlegel’s translation of
Shakespeare, there is not something gained for the German lan-
guage, possibilities of sense that the language would otherwise not
offer, that hitherto it did not offer—indeed in a way parallel to
that in which through the poetry of Hölderlin and Goethe new
possibilities of sense were opened up. Benjamin’s af¤rmation of
such translational gain for the language as such is emphatic: con-
sidered as translators, “Luther, Voss, Hölderlin, and George have
extended the boundaries of the German language.” In this regard
Benjamin cites Pannwitz, for whom translation’s transformation of
one’s own language takes the form of an imperative to which few
translators have measured up: “The basic error of the translator is
that he holds onto the chance condition of his own language in-
stead of letting his language be powerfully moved by the foreign
language. . . . He must extend and deepen his language by means
of the foreign language.”48
Yet, even beyond what can be gained for a language as such by
the effect that translation can have upon it, is it possible for trans-
lation to bring about—even if without simply canceling the loss—
a certain gain in the work translated? Gadamer grants that there
48. Benjamin, “Die Aufgabe des Übersetzers,” 60–61.
88 On Translation
are rare instances in which the translator succeeds in compensat-
ing for the loss by balancing it with a comparable gain or even by
producing such gain as exceeds the loss. He mentions, though
without elaboration, the case of George’s translation of Baudelaire’s
Les Fleurs du Mal.49 Yet one cannot but wonder about the possibility
of such gain, or rather, about what precisely could constitute such
gain. If indeed translation preserves the meaning expressed in the
original, merely transposing that same meaning into another lin-
guistic context, then the gain could not be a gain in meaning,
could not be an accrual of additional meaning to the original
meaning. The gain, it seems, could only be one of expression; it
could only be a matter of expressing the meaning to a greater de-
gree in the translation as compared with the original. But what is
this greater degree of expression? And does such expression to a
greater degree remain, as it seems it must, distinct from the mean-
ing that it, to a greater degree than the original, expresses?
Expression to a greater degree may be achieved by virtue of the
metaphoricity of the translation, by an enhancement of the ¤gures
of expression. Translating any poetic text requires engagement
with the metaphorics of the text if the forcefulness and expressive-
ness of the ¤gures are to be carried over to the translation. In
many instances what is required is a shift of metaphorics, or rather
the unfolding or composition of a metaphorics that in the lan-
guage of the translation can come near matching in expressiveness
the corresponding metaphorics in the original. For example, in the
exchange between Lysander and Hermia near the beginning of
A Midsummer Night’s Dream, the exchange in which they speak of the
dif¤culties that true love ever encounters, there is a passage in
which Lysander surveys the various ways in which love can all too
soon be brought to an end. The passage concludes as follows:
And, ere a man hath power to say ‘Behold!’,

The jaws of darkness do devour it up:
So quick bright things come to confusion.
(I.i.147–49)

Doch eh’ ein Mensch vermag zu sagen: schaut!
Schlingt gierig ihn die Finsternis hinab:
So schnell verdunkelt sich des Glückes Schein.
The metaphorical shift is evident in the ¤nal line: that which,
according to Shakespeare’s text, comes to confusion is said—
countertranslating Schlegel’s rendering—to grow dark. Thus, in
place of the metaphorics of confusion, that is, of a disturbance that
weakens or destroys the composition of something by mixing
up, jumbling together, what belongs to it, Schlegel inscribes a
metaphorics of darkening, setting this off against Shakespeare’s ex-
pression bright things (and his own word Schein) but also carrying it
over from Shakespeare’s expression jaws of darkness in the previous
line. Even though it would not have been impossible to retain the
original’s metaphorics of confusion, translating this as Verwirrung
(Sprachverwirrung means confusion of tongues, as in the story of Babel),
Schlegel’s shift of the metaphorics links the ¤nal line to the pre-
vious one and allows him to interpret Shakespeare’s bright things as
des Glückes Schein, as the light or shining of good fortune or of love’s
happiness, since something that is lighted or that shines can sub-
sequently grow dark but cannot as such come to confusion. Thus,
on the one hand, it could be said that Schlegel’s shift of the meta-
phorics has the effect of ®attening the text metaphorically by re-
ducing what form two metaphorics in Shakespeare’s original to the
single metaphorics of darkening. Yet, on the other hand, the shift
can be regarded as consolidating and hence strengthening the
metaphorics and as making it possible to enhance the speci¤city
of Shakespeare’s bright things. At least in this regard there is good
reason to say that the translation makes a gain over the origi-
nal, even if, regarded otherwise, there is loss with respect to the
original.
In the shifting of metaphorics, whether enforced by linguistic
difference or prompted by other concerns, something that always
must be taken into account is the resonance of the relevant ¤gure
with other metaphorical ¤gures in the same discourse. Such reso-
nances between ¤gures become all the more signi¤cant for the
discourse as a whole when these ¤gures are rigorously bound to a
90 On Translation
conceptuality; in no other discourses is this more thoroughly the
case than in the Platonic dialogues, at least if allowance is made
(and this, too, is a problem of translation) for the fact that the con-
cept, as it were, of the concept is not yet intact or decided in these
discourses. One could suppose, then, that the resonances are no-
where more signi¤cant than in those passages in the dialogues
where certain ¤gures are rigorously bound not just to a conceptu-
ality but to the very origination of conceptuality as such. One such
passage occurs in the Phaedo at the point where Socrates has just
¤nished telling of his efforts to investigate nature, efforts that
failed and that left him bereft of even the knowledge he had for-
merly thought secure. It is against this background that Socrates
then proposes to tell his interlocutors how he went on to venture
a quite different kind of inquiry, the kind of inquiry that had
eventually provoked such opposition that Socrates found himself
condemned to death. As he is awaiting death there in the prison
cell with his closest friends, he tells of his de4terow plo6w.
Tredennick’s translation, “makeshift approach,”50 simply de-
metaphorizes the phrase, ignoring the fact that de4terow plo6w
designates the kind of sailing that one must venture when there is
no wind and it becomes necessary to resort to the oars. Even if
one can say that such a means of sailing is in a sense makeshift,
this does not entail that it is something just randomly taken up,
for it is a means always available and always to be relied upon
in such situations. Although Hackforth’s rendering, “second-best
method,” 51 avoids suggesting randomness, it still strips the expres-
sion of its metaphoricity. Schleiermacher’s rendering, “zweitbeste
Fahrt,” Zehnpfennig’s “zweite Fahrt,” and Fowler’s “second voy-
age”52 adhere a bit more to the metaphoricity of the expression.
Yet still, the word plo6w (linked to pl)v, to sail, and plo.on, ship)
does not designate just any kind of voyage (Fahrt) but only a sea
voyage, a voyage by ship, sailing. Hence the most accurate render-
ing, the one that adheres to the metaphoricity of the expression,
50. Plato, Phaedo, trans. Hugh Tredennick, 81.

51. Plato, Phaedo, trans. R. Hackforth, 127.
52. Platon, Phaidon, trans. F. Schleiermacher, 299; Platon, Phaidon, trans. Barbara
Zehnpfennig, 121; Plato, Euthyphro, etc., trans. H. N. Fowler, 343.

is that found in the translation by Brann, Kalkavage, and Salem:
“second sailing.”53
Yet what counts most is not just the accuracy of this translation
but the fact that it leaves intact the resonances of this expression
with numerous others. There is, for instance, an earlier passage
in which Simmias expresses his reservations about what Socrates
has just declared concerning the immortality of the soul. Simmias
suggests that perhaps in this life one cannot know anything sure
about such matters and so “must sail through life in the midst of
danger, seizing on the best and the least refutable of human dis-
courses [l1goi], at any rate, and let himself be carried upon it as
upon a raft.”54 Yet, above and beyond this and other passages built
around the metaphorical value of sailing, the most decisive reso-
nance sustained by de4terow plo6w is with the sailings set out at
the beginning of the dialogue, the mythic sailing of Theseus to
Crete to slay the Minotaur and the sailing of the ship to Delos in
ful¤llment of a vow made to Apollo to assure Theseus’ success,
the sailing that has the effect of delaying Socrates’ death and in
this sense opening the very interval in which the discussions in
the Phaedo take place. To say nothing of the manner in which the
course of those discussions, the way followed by the dialogue it-
self, has the character of a nautical course.
Even in the rendering of proper names translation can produce
a gain, though it does so only rarely and chie®y in the case of
tag-names, which not only name but also signify something about
the very thing or person named. Such a gain is registered when
Nick Bottom becomes Klaus Zettel. The gain is marked in the ¤rst
scene (I.ii) in which the mechanicals appear, the scene in Quince’s
house where they meet to be assigned the roles they are to play in
the performance before Theseus and his company. As he reads off
the assignments, Quince is playing already, as later, the role of di-
rector. Yet it is Bottom who instructs him how to proceed and who
urges him on:
53. Plato, Phaedo, trans. Brann, Kalkavage, and Salem, 79.

54. Plato, Phaedo 85c–d.
92 On Translation
You were best to call them generally, man by man, according
to the scrip.
(I.ii.2–3)
Quince replies:
Here is the scroll of every man’s name which is thought ¤t
through all Athens to play in our interlude before the Duke
and the Duchess, on his wedding-day at night.
(I.ii.4–7)
As soon as Quince has stated the title of the play, Bottom again
urges him to read out the assignments:
Now, good Peter Quince, call forth your actors by the scroll.
(I.ii.14–15)
The ¤rst name called is that of Nick Bottom:
Answer as I call you. Nick Bottom, the weaver?
(I.ii.16)
In Schlegel’s translation, on the other hand, Bottom’s name has
already been mentioned before Peter Quince reads it off from the
list and assigns to Bottom the role of Pyramus. Or rather, the word
Zettel has already been used as a common noun before Quince/
Squenz uses it as the proper name of the character who is to play
Pyramus. In Schlegel’s translation Bottom/Zettel refers ¤rst to die
Liste, which translates the scrip:
Es wäre am besten, ihr riefet auf einmal Mann für Mann auf,
wie es die Liste gibt.
Quince/Squenz then refers to die Liste by another word, der Zettel:
Hier ist der Zettel von jedermanns Namen. . . .
Then Bottom/Zettel himself, urging Quince/Squenz on, speaks of
der Zettel, calling his own name yet not as such, not as a proper
name, calling his name before it has yet been called in the play:
Nun, guter Peter Squenz, ruf ’ die Acteurs nach dem Zettel auf.

It is at precisely this point that Quince/Squenz then calls the
name, the proper name, of Bottom/Zettel:
Antwortet, wie ich euch rufe!—Klaus Zettel, der Weber.
Thus, the passage is translated by Schlegel in such a way as to

link Bottom/Zettel to the list, the scrip, the scroll, from which
the director, Quince/Squenz, reads off the name of each of the
mechanicals/actors along with the name of the character each is
to play. Bottom/Zettel thus comes to be associated with writing,
with the writing and production of plays and in particular with
casting, which, in this very passage and as it continues, Bottom/
Zettel tries to take over from Quince/Squenz. By the time he is
¤nally called by name and by profession, Klaus Zettel, der Weber,
an association has been woven that calls up the ancient ¤gure of
discourse as weaving. The association does not occur in Shake-
speare’s English text; yet it enhances the metaphorics of that text,
compensating to a degree for the loss of association of Bottom’s
name with the beast with whose head he comes to be endowed.
The enhancement is even more striking in a passage near the
beginning of the ¤nal Act. The passage comes just after Theseus
has delivered his extended discourse on imagination as impel-
ling the lunatic, the lover, and the poet. In this discourse Theseus
draws a connection—or rather, redraws a connection operative
since antiquity—between fantasy and imagination:
Lovers and madmen have such seething brains,

Such shaping fantasies, that apprehend
More than cool reason ever comprehends.
The lunatic, the lover, and the poet
Are of imagination all compact[.]
(V.i.4–8)
Schlegel’s translation af¤rms the same connection:
Verliebte und Verrückte

Sind beide von so brausendem Gehirn,
So bildungsreicher Phantasie, die wahrnimmt,
Was nie die kühlere Vernunft begreift.
94 On Translation
Wahnwitzige Poeten und Verliebte
Bestehn aus Einbildung.
The passage in question constitutes Hippolyta’s response to what
her newlywed husband has just said. She refers to what the four
lovers have told of their night—or dream—in the forest:
But all the story of the night told over,
And all their minds trans¤gur’d so together,
More witnesseth than fancy’s images,
And grows to something of great constancy;
But howsoever, strange and admirable.
(V.i.23–27)
Doch diese ganze Nachtbegebenheit,
Und ihrer aller Sinn, zugleich verwandelt,
Bezeugen mehr als Spiel der Einbildung.
Es wird daraus ein Ganzes voll Bestand,
Doch seltsam immer noch und wundervoll.
What is to be noted is the subtle shift that Schlegel intro-
duces by rendering fancy’s images as Spiel der Einbildung. In this trans-
lation there is restitution of meaning. Indeed Einbildung alone pre-
serves the sense both of image and of fancy (that is, of fantasy), since
it incorporates Bild and has already, in Theseus’ discourse, been
linked to Phantasie. Hence, in the translation of fancy’s images as Spiel
der Einbildung, Spiel functions purely as a signi¤er of a surplus of
sense. Especially at this stage of the play, where re®ection is car-
ried out on the preceding events of the play and preparation then
commences for the play within the play with which the play vir-
tually concludes, the surplus of sense produced by the introduc-
tion of Spiel enhances the discourse signi¤cantly.55 For one can say,
55. The production of a surplus of sense through the introduction of Spiel is

perhaps even more conspicuous in Schlegel’s translation of a later passage. The
passage occurs in one of the conversations in which Theseus and his company,
in the course of the performance of “Pyramus and Thisbe,” exchange comments
about it. In particular, when Hippolyta exclaims:

countertranslating: beyond its opening, the entire play has up to
this point been a matter of play and of play of imagination, indeed
a play within the play even before the commencement, in the ¤nal
Act, of the play within the play that is performed by the mechani-
cals, the play “Pyramus and Thisbe.” So thoroughly has everything
become a matter of play that even the mechanicals,
A crew of patches, rude mechanicals,
That work for bread upon Athenian stalls,
(III.ii.9–10)
have become players in a play.
In bad translations, too, there is often a surplus of sense pro-
duced; but in such cases the surplus is at odds with the sense of
the original and has the effect of distorting and destabilizing the
translation. With Schlegel’s rendering of fancy’s images as Spiel der Ein-
bildung it is quite otherwise. Here the surplus enhances the meta-
phorics that governs this entire portion of the play, the meta-
phorics of fantasy and imagination, the metaphorics that Theseus
puts in play so as to say what he has to say about fantasy and
imagination. This is a metaphorics that turns toward its very ori-
gin, toward those forces capable of the originary displacement of
sense through which metaphors and systems of metaphors (meta-
phorics) are—even if never from a simple beginning—constituted.
The surplus produced by Schlegel’s introduction of Spiel into his
This is the silliest stuff that ever I heard[,]

(V.i.207)
Theseus responds:
The best in this kind are but shadows; and the worst are no worse, if imagi-
nation amend them.
(V.i.208–209)
Schlegel’s translation introduces Spiel, corresponding to nothing in Shakespeare’s
English text, and couples it again, if less directly, with Einbildung, or rather, now,
with Einbildungskraft:
Das beste in dieser Art ist nur Schattenspiel, und das schlechteste ist nichts
schlechteres, wenn die Einbildungskraft nachhilft.
96 On Translation
translation enhances this metaphorics by bringing the ¤gure of
play to bear on the operation of imagination. More broadly, the
introduction of the ¤gure of play contributes a decisive moment
to the metaphorics within which the play and the play within the
play are determined as such, as plays, as enactments in which
everything is only as it seems and not as it actually is, in which
everything is translated into something else and everyone into
someone other. Even in what one would like to consider play in a
proper, nonmetaphorical sense, for instance, the play of children,
translation and metaphor are the decisive constituents. In play even
in its would-be proper sense, translation and metaphor are already
in play.
The insight behind Schlegel’s translation of fancy’s images as Spiel
der Einbildung, or rather, the insight that, perhaps covertly, guides
that translation, is the following: that bringing the ¤gure of play
to bear on imagination brings about an enhancement of a dis-
course that would turn back toward the origin of metaphoricity
and translation. Translation, in particular, would seem to have no
other recourse than play of imagination. From a word in one lan-
guage or from the meaning of this word, one cannot infer the
word in another language that will convey the same meaning. Nei-
ther is translation a matter of judgment in the classical sense of
subsuming a particular under a universal; for the relation between
particular and universal is a quite different relation from that be-
tween a signi¤er and its meaning. Because translation deals with a
relation neither simply between meanings nor between meanings
and singular things but rather between meanings and words, it
requires a power of another kind than reason and judgment. Be-
cause translation engages a movement neither simply from word
to meaning nor from meaning to word but rather, as Figal has
shown,56 a double movement from the sphere of one’s own lan-
guage into that of a foreign language and back from the foreign to
one’s own, translation requires something more than simply the
56. Günter Figal, “Seinserfahrung und Übersetzung: Hermeneutische Über-

legungen zu Heidegger,” Studia Philosophica 57 (1998): 184.

power of intention and expression. Translation cannot but have re-
course to imagination, for it is imagination—especially as it came
to be determined by Kant, Fichte, and the German Romantics (to
whose circle A. W. Schlegel belonged)—that has the capacity to
mediate between alien spheres without reducing the foreignness
of their relation. Imagination is the force of holding them together
in their difference, of holding them together by its movement be-
tween them, by its hovering between them.57
Gadamer’s discussion of translation, oriented to the analysis of
conversation or dialogue, becomes perhaps most incisive at those
points where a certain analysis of conversation is turned back
upon translation so as to clarify it. Thus, in Truth and Method Gada-
mer deals with the back and forth, the to and fro (Hin und Her),
that is characteristic of conversation and then indicates how this
character pertains also to translation: “And, as in conversation,
when there are such unbridgeable differences, a compromise can
sometimes be achieved in the to and fro of dialogue, so the trans-
lator will seek the best solution in the to and fro of weighing and
considering [im Hin und Her des Wägens und Erwägens]—a solution that
can never be more than a compromise.”58 It is to such a to and fro
movement, a hovering, that imagination is peculiarly suited. And
it is in this regard, as a free oscillation to and fro between various
different terms, that imagination deploys its force as play of imagi-
nation. It is to such play of imagination that translation cannot but
have recourse.
Even though, according to the classical determination, it is im-
perative that translation circulate through the meaning of the lin-
guistic unit being translated, it is not the meaning that gets trans-
lated but the linguistic unit (the word, the phrase, the sentence).
Because it is a matter of translating a word (for instance) in a
foreign language into a word in one’s own language (or, less com-
monly, the opposite), translation requires a spanning of the lin-
guistic difference, a persisting in the alterity. Yet it requires a to
and fro not only between languages but also within the language
57. See Force of Imagination, esp. chap. 2.

98 On Translation
into which one is translating, a to and fro in which one hovers
between various translations that freely and imaginatively offer
themselves, for they cannot be inferred or in any such way deter-
mined. On the other hand, they must be weighed and considered,
tested in their capacity to express the meaning signi¤ed by the
word being translated. This testing, this measuring of their force
of expression, requires that the free play of imagination be bound
to the intentional and expressive powers, to the very power of
speech as such. For if the playful hovering of imagination between
the various translations that offer themselves is to issue in a trans-
lation, these possibilities must be measured; the expressive force
of each, its capacity to express the meaning expressed by the word
being translated, must be measured against the others and, above
all, against what is expressed by the word being translated. It is
because of this measuring and the intentional-expressive powers
it involves that the play of imagination in translation may be called
a lawful play and that one may refer in this connection to the free
lawfulness of the imagination, a lawfulness without a law.59 In-
deed, turning within the circle, one will insist that translational
possibilities can offer themselves only to the free lawfulness of
imagination, only to a free play of imagination that also is bound
to the intentional animation of the sense of the word being trans-
lated.
Schlegel’s translation of fancy’s images as Spiel der Einbildung pro-
duces a surplus, a certain gain over the original, which, however,
enhances the metaphorics of the original so as to make the trans-
lation in this respect more expressive. But what about such gains?
Do they indeed go no further than to enhance the expressiveness
of the translation, to make it express better—more forcefully—the
same meaning that the original expresses? Is this the limit of the
advance that Schlegel achieves by translating fancy’s images as Spiel der
Einbildung, by introducing into the translation the surplus of sense
conveyed by Spiel? Is it in this case only a matter of expressing bet-
59. The sense of lawful play is discussed already in Plato’s Republic (424e–425a)
(see Being and Logos, 21–22). The designation “the free lawfulness of the imagina-
tion” and its characterization as “a lawfulness without a law” are developed in
Kant’s Critique of Judgment (Kritik der Urteilskraft, 240–41).

ter in German a complex of meanings that de¤ne imagination/
fantasy and that would remain completely invariant as such in the
translation? In a certain respect it is unquestionable that what
Schlegel achieves here goes beyond mere enhancement of the ex-
pression of an invariant meaning: by bringing Spiel into play in the
translation Schlegel makes the translation say something about
imagination that is not said in Shakespeare’s text. And yet, how can
this surplus be produced in the translation, how can something
be said that is not said in the original, without altering the mean-
ing of the original, that is, without distorting it, that is, without,
in the end, simply producing a bad translation? Is it possible to
sequester the invariant original meaning, to immunize it against
the intrusion of a surplus, to consign this surplus to a sphere of
mere expression exterior to the meaning that translation is to pre-
serve? Or is it perhaps the very concept of meaning that needs
here to be put into question? Can translation—especially when it
is a case of such a genius of translation as Schlegel—be under-
stood as the mere recontextualizing of one and the same mean-
ing? Does the primary responsibility of the translator lie only in
the preservation—in the pure reexpression—of such meaning?
What about meaning as such? What about the concept of mean-
ing? There is perhaps nothing more resistant to being put into
question. For precisely what one intends as a question putting
meaning in question—as in the questions just posed—will almost
invariably prove to be nothing more than sheer tautology, which
thus fails entirely to open what one calls (tautologically) meaning
as such to questioning and, instead, merely repeats the same, as if
compulsively, indeed under a compulsion stemming from the be-
ginning of philosophy. For meaning, what, as such, and concept all say
the same, and any con¤guration of them results only in tautology.
Yet, even short of such abysmal complications, there are others
that are pertinent, complications pertaining speci¤cally to transla-
tion, complications that occur in the circulation in and as which
translation takes place. For the question is whether translation is a
pure circulation through meaning. Two kinds of entanglements
may be mentioned that have the effect of complicating the would-
be pure circulation.
The ¤rst is a certain nonreciprocity that sometimes occurs in
100 On Translation
translation. The pertinent instances are not those in which a word
or phrase is so poorly translated as to make it dif¤cult to counter-
translate, to translate back from the translation to the original.
Rather, it is a matter of instances in which the fecundity of the
translation makes possible a development that would never have
been possible starting from the original. In such an instance the
development has the effect of making successful countertransla-
tion impossible—that is, the original word or phrase will prove
to be inadequate as a translation of the very word or phrase into
which it was translated. For example, in his essay on l1gow in
Heraclitus, Heidegger ventures to translate the verb l)gein by lesen
in the sense of sammeln.60 In turn, the English translation renders
this sense by the word to gather or gathering.61 Not only is this trans-
lation appropriate but, most signi¤cantly, the semantic and meta-
phorical possibilities that it opens up allow the re®ection that
Heidegger begins in his essay to be developed to the point where
gathering could no longer be translated, without further ado, back
into Sammeln.62 What the translation enables here quite exceeds
merely preserving one and the same meaning. Here it is not a mat-
ter of one and the same meaning simply persisting intact through
the translation; rather, it is a matter of a translation that, without
falsifying the original, enhances the meaning and opens it to mu-
tation and transformation.
The second kind of entanglement is what might be called over-
translation. This can occur in texts in which a certain basic word
or phrase undergoes a mutation of sense as a result of theoretical
developments carried out in the text. Overtranslation occurs when
such a word or phrase is, from the very beginning of the text,
translated in such a way that in its translated form it signi¤es
the mutated sense reached only through the developments in the
60. Heidegger, “Logos (Heraklit, Fragment 50),” in Vorträge und Aufsätze (Pful-
lingen: Günther Neske, 1954), 207–29. See esp. 209–10.
61. The translation is by David Krell, in Heidegger, Early Greek Thinking (New
York: Harper & Row, 1975), 59–78. See especially 61–62.
62. See The Gathering of Reason (Athens, Ohio: Ohio University Press, 1980) and
especially the Translator’s Note to the German translation: Die Krisis der Vernunft:
Metaphysik und das Spiel der Einbildungskraft (Hamburg: Felix Meiner, 1983).

text. One result of overtranslation is to render this development
incoherent, trivial, or even imperceptible, since the development
is already tacitly built into the translation from the beginning. On
the other hand, if overtranslation is recognized as such, it can—
like a broken tool—serve to illuminate something, namely, a cer-
tain operation of textuality that exceeds mere transition between
stable, preconstituted meanings, an operation by which a text can
produce a mutation in the very meanings that it is engaged in sig-
nifying.
In the course of Plato’s Sophist, the Stranger arrives at a kind of
de¤nition of being. What he says at this point might be translated
as follows: “For I set up as a limit [^row] by which to delimit
[Vr,zein] beings [t% Onta] that they are nothing but d4namiw.” 63
But if, from the beginning of the dialogue, one had already built
this de¤nition into the translation, if, for instance, one had trans-
lated t% Onta as capable beings or potential beings, then it would not be
possible even to translate this passage without rendering it tauto-
logical, to say nothing of the effect on earlier passages in the text.
Much the same bind can easily arise in translating certain of
Heidegger’s texts, perhaps most notably the unpublished treatises
from the late 1930s and early 1940s. The radicality of these writings
is such that their language is formed and deformed within the text
itself, often in a kind of mutational repetition that sets words apart
from the constituted language of metaphysics, that recomposes
them at or beyond the limit. Heidegger himself gives an important
indication in a marginal note to Vom Wesen der Wahrheit, in which
he lays out the stages of the mutation that the sense of Wesen under-
goes in and through that text.64 Nothing could be less true to
Heidegger’s text than to incorporate that development preemp-
tively in the very translation of Wesen.
Nonreciprocal translation and overtranslation serve to expose a
mutation of meaning brought about either in and through the
63. Plato, Sophist, 247e.

64. The marginal note is to the third edition, 1954. It reads: “Wesen: 1. quidditas
—das Was—koin1n; 2. Ermöglichung—Bedingung der Möglichkeit; 3. Grund
der Ermöglichung” (Heidegger, Wegmarken, vol. 9 of Gesamtausgabe [Frankfurt a.M.:
Vittorio Klostermann, 1976], 177).
102 On Translation
translation or in and through the operation of the text itself. In
turn, the exposure of such mutation puts in question still more
decisively the requirement that translation preserve meaning, that
it be limited to pure circulation, to pure reexpression of meaning.
In this connection it is not only a question of translation sustain-
ing a more complex relation to meaning but also a question of
whether the very concept of meaning (with all its complications)
can ever suf¤ce for a discourse on translation.
What is remarkable about Gadamer’s discourse on translation is
the way in which, framed within the orbit of the classical deter-
mination, it carries that determination to the limit and thereby
gestures toward another determination. The general formulations
in Truth and Method establish the frame: translation is to preserve the
meaning, though, because of the necessity of setting the mean-
ing within a new context, translation is already interpretation and
its product eine Nachbildung. Gadamer writes even in a way reminis-
cent of Hegel’s stress on the appropriative power of translation: as
in reading a text, so in translation, says Gadamer, it is a matter
“of alienness and its conquest [vom Fremdheit und Überwindung dersel-
ben].” 65 Through translation one would make an otherwise alien
meaning one’s own. Yet in this process one’s own thoughts would
come into play, not simply to cancel the alienness of the text so as
to appropriate it, but—as Gadamer says, drawing out the parallel
between reading a text and translating it—to engage in the re-
awakening of the meaning of the text (in die Wiedererweckung des
Textsinnes). But still, even if in need of being reawakened, it is as
though the meaning were intact so that the fusion of horizons—
which Gadamer explicitly mentions in this connection—would
serve, in the end, only for its preservation and reexpression in an-
other linguistic context.
And yet, even in Truth and Method there are other indications that
begin to push the classical determination toward the limit. One
65. Gadamer, Wahrheit und Methode, 365. While retaining the double movement
between the foreign and one’s own, Figal emphasizes the engagement of this
movement with alterity rather than its appropriation of the foreign: “One could
say that translation as such is characterized by a double alterity” (Figal, “Seins-
erfahrung und Übersetzung,” 184).

such indication lies in Gadamer’s introduction of the concept of
spirit (Geist). Speci¤cally, he characterizes the gap across which
translation moves, not just as a gap between the original words
and the restitution achieved in the words of another language, but
rather as a gap between the spirit of the original words and the
spirit of the restituting expression. But what is the spirit of words,
of wording (Wortlaut), literally, of word-sounds? Is it anything
other than their meaning, that which animates the sounds so that
they are words and not just sounds? Even if, alternatively, one
hears in Gadamer’s appeal to spirit an echo of von Humboldt’s idea
of the spirit of a language, of a spirit unique to each language, one
would still need to ponder whether meaning could remain intact
apart from the various spirits of various languages, merely ex-
pressed and reexpressed according to those spirits and in those
languages.
Even Gadamer’s characterization of translation as highlighting
(Überhellung) works against the classical determination by which,
in Truth and Method, it is framed. For one could hardly suppose that
those features of the original that are emphasized and those others
that are played down or even suppressed have only to do with the
expression of a selfsame meaning that remains completely un-
touched by such highlighting.
Yet it is after Truth and Method that a stronger thrust is exerted
toward unsettling the classical determination of translation. Most
notably, in the text “Lesen ist wie Übersetzen”—indeed right from
the beginning of this text, with the opening citation from Croce:
“Every translation is like a betrayal.”66 For a translation that purely
reexpresses, if less powerfully, the meaning of the original could
never be charged with betrayal, with treason, but only with weak-
ness, imperfection, ®atness. In order for every translation to be
like a betrayal, translation must be such as inevitably to violate
the meaning of the original; it must be such that it could never
reexpress the meaning of the original. Even in this case one might
of course continue to measure translation by the classical norm,
66. “Jede Übersetzung ist wie ein Verrat” (Gadamer, “Lesen ist wie Über-
setzen,” 279).
104 On Translation
demanding that it ought to restitute the meaning though it al-
ways fails to do so, fails even necessarily and not just in fact. Yet if
translation necessarily fails to achieve the result prescribed by its
norm, there will be good reason to question that norm, to ques-
tion whether it is the norm appropriate to translation, to question
whether translation ought to be measured by the demand for res-
titution of meaning.
Gadamer begins to venture such questioning in the text “Lesen
ist wie Übersetzen.” On the one hand, he acknowledges that there
are certain texts (especially of a more technical sort) where all
that is important is to grasp what is meant; in such cases transla-
tion would legitimately be determined by the norm of restitution
of meaning. But now Gadamer stresses how different certain other
kinds of texts are, most notably, though not only, poetry. In trans-
lating, as in reading, such texts, one is engaged—says Gadamer—
in “an interpretation through tone and tempo, modulation and
articulation—and all this lies in the ‘inner voice’ and is there for
the ‘inner ear’ of the reader.”67 Such texts are not mere expressions
of meaning, but rather their very operation as texts, their textu-
ality, also involves tone and tempo, modulation and articulation.
Therefore, a translation that only restituted the meaning of such a
text without recovering its inner voice would utterly betray the
original.
And yet, beyond even this questioning, it remains to be asked
whether in the case of such texts the distinction between the
meaning and the so-called inner voice can remain intact, whether
what one would presume to distinguish as the meaning is not al-
ready so infused with the tone, tempo, modulation, and articula-
tion as to be inseparable from them. In this case, if the inner voice
belongs to what one would have called the meaning, it would not
suf¤ce to reconstitute the task of the translator as one of preserv-
ing both meaning and the various moments that go to make up
the inner voice. One could not keep the classical determination of
translation, of its norm or ideal, intact simply by adding the re-
quirement that other moments supplementary to meaning be pre-
67. Ibid., 284.

served along with the meaning. And just as the inherence of the
inner voice in what comes to be said has the effect of putting
the classical determination of translation in question, it also sus-
pends the very determination of discourse as expression of mean-
ing, as expression that is distinct from—that does not itself belong
to—the meaning it expresses. What would also be put in question
—that is, reopened to questioning—are those translations that
constitute the manifold within which discourse and translation in
its alleged proper sense operate. It would be a matter, then, of tak-
ing up again the original complication outlined by Aristotle, of
doing so in such a way as to reopen those questions that have al-
most always seemed to have been settled by Aristotle. It would be
a matter of unsettling and reinterrogating the way in which words
are signs of what the Greeks called affections in the soul and the
latter, in turn, likenesses of things.
In his deconstructive studies of Husserl’s Logical Investigations,
Derrida has demonstrated how meaning as the correlate of a pure
intention is necessarily contaminated by what belongs inseparably
to its expression, by what Derrida calls indication (l’indice).68 To
this extent he reopens questioning at the level of the original com-
plication outlined by Aristotle; Derrida pursues such questioning
most openly—though by no means exclusively—in his decon-
struction of phonocentrism, calling into question—indeed invert-
ing and displacing—the relation that Aristotle posits in declaring
that written words are symbols of spoken words. Derrida also fo-
cuses quite precisely on what the contamination of meaning en-
tails with regard to translation: “Within the limits in which it is
possible, in which at least it appears possible, translation practices
the difference between signi¤ed and signi¤er. But if this differ-
ence is never pure, no more so is translation, and for the notion
of translation it would be necessary to substitute a notion of trans-
formation: regulated transformation of one language by another, of
one text by another. We will never have, and in fact have never
had, to do with some ‘transport’ of pure signi¤eds from one lan-
68. Derrida, La Voix et le Phénomène (Paris: Presses Universitaires de France,

1967), chaps. 1–3; see my discussion in Double Truth (Albany: State University of
New York Press, 1995), chap. 1.
106 On Translation
guage to another, or within one and the same language, that the
signifying instrument—or ‘vehicle’—would leave virgin and un-
touched.” 69
Meaning is not left virgin and untouched either as the pure
correlate of an intentional act or as sequestered from what belongs
to its expression (for instance, the inner voice). Translation cannot,
then, consist in the transposition of meaning from one language
to another; it cannot consist simply in the movement from a sig-
ni¤er in one language, through the meaning, to a signi¤er in the
other language. Even if translation and the lawful play of imagina-
tion that animates it must still in a sense go through this move-
ment, the movement alone as pure transposition will no longer, as
in the classical determination, suf¤ce to de¤ne translation. Rather,
in passing between languages, the play of imagination will pro-
duce a regulated transformation. As when, with the translation of
Bottom as Zettel, the connection with weaving remains invariant
while the association with the creature with whose head Bottom
is endowed is exchanged for an association with the writing and
production of plays.
For all its radicality, Benjamin’s analysis stops short of drawing
these conclusions. Benjamin fully recognizes that mere restitution
such as would produce likeness (that is, similarity of meaning)
between translation and original is insuf¤cient: “it can be demon-
strated that no translation would be possible if in its ultimate es-
sence it strove for likeness [Ähnlichkeit] to the original.”70 While
there are of course good reasons for this, the “demonstration” that
Benjamin offers, at least immediately, is almost beside the point:
he refers to the changes in tenor and signi¤cance that the original
undergoes over time as a result of general changes in the language
and to the comparable changes sustained by the translation as a
result of transformations occurring in the language of the transla-
tor, transformations that may even be determined in part by great
translations. While one will grant that consequently the relation
between the original and the translation will undergo a change, it
69. Derrida, Positions (Paris: Les Éditions de Minuit, 1972), 31.

70. Benjamin, “Die Aufgabe des Übersetzers,” 53.

will nonetheless need to be maintained that their relation can
change only if there is a determinate relation, some accord be-
tween translation and original that—whatever future changes may
occur—will have been aimed at in producing the translation.
In any case Benjamin’s point is only that the relation between
translation and original is not one of likeness or similarity. What,
then, is the character of their relation? Benjamin sometimes pro-
poses a phenomenological strategy in order to explicate this rela-
tion. The task of the translator would be to recover the intention
of the original so as to reenact that intention within the language
in which the translation is to be produced: “The task of the trans-
lator consists in ¤nding that intention in the language into which
he is translating that awakens in that language the echo of the
original.” 71 Yet an echo is a kind of imperfect likeness, and it is
thus on the intention, not the echo it produces, that Benjamin
must insist. Thus: “A translation, instead of making itself similar
to the meaning of the original, must lovingly and in detail form
within its own language the way of meaning [Art des Meinens] be-
longing to the original. . . .”72 Up to this point Benjamin’s posi-
tion does not differ essentially from Gadamer’s in Truth and Method:
translation is a matter of reenacting the meaning of the original
in another language, recontextualizing one and the same meaning
or at least the intention to which it corresponds. But Benjamin
turns in a very different direction in the continuation of the sen-
tence: “. . . thus making both the original and the translation rec-
ognizable as fragments of a greater language, just as shards are
fragments of a vessel.” What connects original and translation
Benjamin calls kinship (Verwandtschaft), a kinship between the lan-
guage of the original and that of the translation. It is in his expli-
cation of this kinship that Benjamin explains his reference to a
71. Ibid., 57.

72. Ibid., 59. The recourse to the intention rather than directly to the meaning
is expressed perhaps most succinctly in the following passage: “On the other
hand, over against the meaning, the language of the translation can—indeed
must—let itself go, in order to give voice to the intentio of the original, not as
reproduction [Wiedergabe] but as harmony, as a supplement to the language in
which the original is expressed, as its own kind of intentio” (ibid.).
108 On Translation
“greater language” of which original and translation would be rec-
ognizable as fragments. He declares that this kinship of languages
consists in the fact “that in each of them as a whole one and the
same thing is meant [gemeint], which however is attainable by no
single one of them but only by the totality of their intentions sup-
plementing each other: pure language.”73 Translation would, then,
actualize this kinship, integrating languages, bringing them into a
certain correspondence in their way of meaning, ripening the
seed of pure language.
From this point on, Benjamin’s analysis breaks radically with
the classical determination of translation and even with its herme-
neutic modi¤cation. He is acutely aware of the danger of breaking
all bonds of translation to the restitution of meaning. He writes
of how the task of ripening the seed of pure language “seems
never resolvable.” He asks—not just rhetorically—whether the
very ground of translation does not withdraw “once the restitu-
tion of meaning ceases to provide the measure.” And he concludes:
“Viewed negatively, this is indeed the meaning of all the forego-
ing.”74 Thus, Benjamin does not conceal the consequences that
must be faced once one abandons restitution of meaning as the
measure of translation and instead dedicates translation to the task
of ripening the seed of pure language.
And yet, for all its radicality, this move does not break entirely
with the classical determination of translation. In the end, as the
end to which all translation would—even if without determinate
measure—be referred, Benjamin’s analysis invokes pure meaning
such as would remain uncontaminated by signifying operations.
This pure meaning, virgin and untouched, is no longer (as in the
classical determination) that which can be said (signi¤ed) in any
particular language but rather is, in a very classical sense, the ideal,
that which all languages together, with their mutually supplemen-
tary intentions, would say if that totality of signi¤cation were, at
the limit, to be realized. Thus, Benjamin’s analysis posits at the
ideal limit a totality of meaning that would have escaped contami-
73. Ibid., 54.

74. Ibid., 58.

nation by signi¤cation, a realm of meaning in which all commu-
nication and even all intentions are extinguished, a pure language
in which there remains only the expressionless word.
Benjamin declares that “the task of the translator is to release
[erlösen—hence also: save, rescue] in his own language that pure
language that is captivated by another, to liberate the language im-
prisoned in a work in his recasting of the work.”75 One cannot
but wonder: Is language imprisoned in works so as to be in need
of being liberated? Or does it only seem imprisoned because the
language of and in the work is regarded from the vantage point of
the ideal of pure meaning, of pure language? Is it not rather in a
work that language is liberated, freed to itself in such a way that
what has not—and perhaps could not have—been previously said
comes to be said? In any case there is every reason to wonder—as
indeed does Benjamin himself—how translation governed only by
reference to such an ideal of pure meaning and by the imperative
of liberating it could be regulated. Could ripening the seed of pure
language ever become—or be assured of becoming—a regulated
transformation?
The radical critique of the classical determination of transla-
tion, which with Nietzsche takes the form of mere inversion, be-
comes with Benjamin a rupture, a break. Or rather, it would have
been a break, an utter break, had it not reconstituted the classical
determination at the limit, as an ideal, retaining—even if at an
unbridgeable distance—a vestige of pure meaning. If translation
is, then, from this distance, liberation of pure meaning, giving it
back to itself—that is, still, restitution of meaning—mere reference
to this ideal does not suf¤ce to prevent translation from becoming
capricious (or, at best, creative) transformation.
If one is to persist in the deconstruction of the classical deter-
mination while also redetermining translation as regulated trans-
formation, as engaged by lawful play of imagination, then it is im-
perative to turn back from the dream of pure language, which is
also a dream of nontranslation. It is imperative to be attentive, not
to a remote vestige of meaning by which translation would—but
75. Ibid., 60.
110 On Translation
cannot—be bound, but to what bounds and is bound to language
as it operates in human comportment. Once language is no longer
construed simply as signi¤cation, as a totality of signi¤ers signify-
ing meanings themselves independent of signi¤cation, then what
bounds and is bound to language proves to be nothing other than
what comes to be said—das, was zur Sprache kommt. What comes to
be said in speech or in a linguistic work is nothing other than that
which the speech or the linguistic work makes manifest, that
which it lets—in the unique way proper to language—show itself,
come to presence. This is what, bound to language and its mani-
festive force, also bounds language, determines its bounds, re®ects
back to it its very determinateness. If translation of speech or of a
linguistic work is to be a transformation that is regulated, if it is
to be engaged by a play of imagination that is lawful, then refer-
ence to—that is, being bound by—that which the speech or work
makes manifest is imperative.
It would be a matter of attending again to the force of words
and of bringing translation again to rely on the force of words—
on their force of making manifest—in order to bring into play a
measure by which to regulate translation. By being bound by what
the original makes manifest, translation would carry out a regu-
lated transformation of the original.
As in the rendering of fancy’s images as Spiel der Einbildung in which
Schlegel will have been attentive to the way in which Shake-
speare’s play makes imagination manifest and will, by introducing
Spiel, have enhanced the manifestive force in the translation in a
way that accords with the original.

Four Varieties of Untranslatability
Attestations to untranslatability abound. Poetry especially, many

have declared, is untranslatable. Yet translations of poetry also
abound. Even the poetry of those whose poetic gift and artis-
tic mastery would seem to make their work—if any is indeed—
untranslatable has found translators, and the best among them
have produced translations that no one would deem unworthy of
the title. At least with only rare exceptions: Hölderlin’s translations
of Sophocles were dismissed by his contemporaries, largely be-
cause of what they saw as a monstrously literal rendering of the
syntax of the original. But such an exception only serves to point
to a depth of translatability that could be fathomed only by such
a poet as Hölderlin, that could only have remained otherwise con-
cealed. While Hölderlin remained unheeded as a translator by
those whom one would most have expected to have recognized
his gift and his mastery as a translator (Hegel, for instance), his
incomparable achievement was eventually to come to light. More
than a century later Benjamin writes of Hölderlin’s translations of
Oedipus Tyrannus and Antigone: “In them the harmony of the lan-
guages is so profound that sense is touched by language only as an
aeolian harp is touched by the wind. Hölderlin’s translations are
prototypes of their kind; they are to even the most perfect render-
ings of their texts as a prototype is to a model [als das Urbild zum
Vorbild].” 1
Poetry is not, then, simply untranslatable, not even the poetry
of Sophocles, of Shakespeare, of Hölderlin himself. It is not un-
translatable in an unconditional sense: there is no poem of which
one can say in advance that every would-be translation of it will
prove disentitled as such, will prove to be unentitled to be called
a translation. If, nonetheless, attestations to the untranslatability of
1. Benjamin, “Die Aufgabe des Übersetzers,” 61.
112
poetry abound, what is attested can only be a more limited un-
translatability. In whatever way the attestations are framed, regard-
less of how unconditionally they may be stated, the untranslat-
ability of poetry thus attested can consist only in poetry’s not
being translatable without loss and without the ®attening and dis-
tortion generally that is produced by such loss. Yet, at least in the
case of the best translations, a certain gain also is brought about,
certain enhancements of the original, which may offset or other-
wise balance what is lost through the reductions, transpositions,
and shifts that translation requires. Even—perhaps most of all—in
the case of poetry, translation is a matter of exchange; it operates
within an economy geared to the difference between the two lan-
guages, an economy that is not ¤xed but is open to the initiatives
of an adventurous translator, one willing to venture a certain ex-
penditure with con¤dence of what will be returned. Such initia-
tive may open up a hitherto unsuspected depth of translatability,
as in the case of Hölderlin’s translations of Sophocles. But it may
also fail abysmally. For it is not just a matter of willingness to ven-
ture nor of con¤dence in one’s venture; in the case of one such as
Hölderlin the utmost reticence may be in play. It is a matter of the
translator’s gift, of his genius for opening a favorable transforma-
tion between the two languages. But it is equally imperative that
what occurs through the play of imagination be lawful, that the
translation be a regulated transformation. Short of reconstituting
the classical determination of translation, regulation can be opera-
tive only through attentiveness to what the original makes mani-
fest, to the force of its words.
As already in the attestations regarding poetry, untranslatability
can itself become a theme of speech, even to such an extent that
the speech may identify, name, or mark something that cannot be
said in that very speech, something untranslatable not only into
that speech but into any speech whatsoever. One may, for in-
stance, write about how what one would say escapes what one does
—indeed can—say. In such a case one writes about the untrans-
latable, supplementing what one writes with a writing that testi-
¤es to a certain untranslatability operative in what one writes: as,
for instance, writing in a letter of one’s deep feelings for another,
one writes also of the untranslatability of those feelings into writ-
Varieties of Untranslatability 113

ing; that is, one attests in the writing to their untranslatability into
writing.
The con¤guration of such untranslatability is quite different
from the untranslatability of a text in one language into another
language. For what is attested when one writes of the untranslat-
ability of certain feelings is an untranslatability that, while neces-
sarily not total, is nonetheless unconditional. It is not merely that
if one’s feelings were to be translated, to be expressed in words,
there would be a certain loss, as when Plato’s translators render
f$rmakon as poison, losing sight of the medicinal effect for which
in his ¤nal words Socrates expresses his debt to Asclepius. In the
case of untranslatable feelings it is such that—or at least it is at-
tested to be such that—any would-be translation of those feelings
into speech would be not just a bad translation, not just a transla-
tion in which much would be lost, but rather no translation at all.
Any alleged expression of those feelings in language would prove
to be an imposter, something quite other than an expression of
those feelings. And yet, the very possibility of attesting to such un-
translatability requires that it not be total: in order to identify,
name, mark the feelings as untranslatable into speech, it must be
possible to say something about them and so to mark a limit of
their untranslatability. It must be possible at least to name them,
to call them by some such name as feelings; and it must be possible
at least to say of them that they are untranslatable. Because at least
their untranslatability is translatable into speech, it cannot be total.
Yet attestation of untranslatability, of unconditional untranslat-
ability, is not limited to writing or to language generally. It is not
only in language, not only by way of language, that one can attest
to an untranslatability into language as such and into such mean-
ings as can be signi¤ed by linguistic signi¤ers. There are indeed
exceptional cases in which that which is attested to be untranslat-
able into linguistically signi¤able meanings comes to provide the
very medium of the attestation. In such cases there is a pecu-
liar coincidence of attestation with that to the untranslatability of
which it attests. As in the case of a painter who paints in such a
way as to attest in the painting to its untranslatability into lan-
guage, into what can be said.
Focusing now on such a case, it will be a matter of again trans-
114 On Translation
1. Mimmo Paladino. Senza
titolo [Untitled], 1989. Oil
on canvas and wood, 108
× 88 × 15 cm.
lating translation beyond translation within a language or between

languages, of translating translation and the questions it prompts
to the interval, the difference, separating language and all that lan-
guage can say from painting and all the elements of visibility as
such that painting can visibly present.
In 1990 the celebrated Italian artist Mimmo Paladino exhibited
at Villa delle Rose in Bologna a cycle of seven paintings along with
a cycle of eight closely related drawings.2 The two cycles were
given the single title EN DO RE. One of the paintings (see ¤gure 1)
has on its surface an inscription that is almost the same as this
title; the title given to the two cycles, EN DO RE, is, by Paladino’s
own testimony,3 merely a mutation of the inscription in the paint-
ing, EN DE RE. In addition to clarifying the relation between the
title and the inscription, Paladino made available a small book en-
titled EN DE RE, a very strange book about which it suf¤ces here
to say only that it disrupts every attempt to translate—and so to
2. There is a catalogue to the exhibition: Paladino (Bologna: Nuova Alfa Edito-

riale, 1990). See also the volume of reproductions and essays entitled EN DO RE
(Siracusa: Tema Celeste Edizioni, 1990).
3. Mimmo Paladino, letter to author, 30 August 1994.

make sense of—these enigmatic marks EN DE RE.4 Thus, when EN
DE RE is actually inscribed in the painting, one should not assume
that Paladino is producing some kind of synthesis of word and
image; for what look like words are not words, that is, are not
translatable into words. Paladino refers to them—undecidably—as
ciphers (cifra).
Though the cycle of paintings has the cipher-title EN DO RE,
sharing it with the cycle of drawings, the individual paintings in
the cycle are untitled, are designated as Senza titolo. This disentitle-
ment could be regarded as a shield meant to secure a pure visi-
bility, to guard the paintings against any intrusion by language, to
forestall all translation of the visible into the word. Addressed to
the viewer as a kind of warning, the disentitlement would be, in
this case, the painter’s way of prescribing that one behold the
paintings as they display themselves before one’s vision, leaving
aside all translation of the spectacle into discourse, which could
only contaminate the pure visibility of the paintings. By withhold-
ing titles from the paintings, the painter would seem to proclaim
a reign of absolute untranslatability, an untranslatability by which
the paintings would be sealed off in themselves over against the
otherwise intrusive word; disentitlement would, in this case, ab-
solve the paintings in their visible presence from all relation what-
soever to language.
And yet, disentitlement is not itself totally absolved from lan-
guage: when works not only are deprived of a title but also are
presented, in exhibitions and catalogues, with the designation un-
titled, such disentitlement borders on entitling them, on entitling
them to the title untitled. Yet the very sense of the designation un-
titled requires that it not be a title: a painting can be designated as
untitled only if it has no title. One would thus have to say that the
designation untitled both is and is not a title; or rather, that it is a
title the very bestowal of which frees the painting from entitle-
ment. With this designation it is as though language only grazed
the surface of the painting, being re®ected back to itself rather
4. Achille Bonito Oliva with Mimmo Paladino, EN DE RE (Modena: Emilio

Mazzoli Editore, 1980). See my discussion of this book in Shades—Of Painting at the
Limit (Bloomington: Indiana University Press, 1998), 120–24.
116 On Translation
than adhering to the painting, letting the painting thus go free.
Here language outdistances itself. Yet, in such disentitlement, the
painting will have been grazed by the word, will have been re-
leased by self-outdistancing language. As also through the inscrip-
tion of untranslatable ciphers, the painting will have been touched
by language. But rather than importing language, as it were, into
the painting, the effect is to take up, in a painterly way, the ques-
tion of language and painting, the question of the translatability
and untranslatability of painting.
In disentitlement the painting is touched by language only to
be released to itself, to its pure visibility, its pure untranslatability.
This untranslatability is not absolute: The painting is not absolved
from all relation to language, since the disentitling word is pre-
cisely what frees the painting to itself and lets its pure visibility
be sheltered from the intrusion of language. Disentitling is a way
of saying the untranslatability of the painting, of translating its
untranslatability while leaving the painting nonetheless untrans-
lated, indeed untranslatable. One will want to ask: What is its
untranslatability?—and yet, the question is possible only if, in ask-
ing about the what, the question outdistances itself and releases the
painting from the hold of anything like a what. Only in this way,
by turning language back to itself so as to free from it that of
which one will have spoken, can one say what the untranslatability
of painting is: that the visible elements gathered in their visibility
in the painting cannot be translated into a linguistically signi¤able
meaning or complex of meanings such as could be con¤gured in
a title or in a discourse elaborating the sense of a title. To be sure,
the painting is touched by language and its signi¤cations: in being
disentitled, in bearing an inscription of quasi-linguistic ciphers,
even in being written about by the artist. And yet, it is touched
by language only to be released into a silence that can never be
matched by words. What constitutes its untranslatability is this si-
lence.
Even in the case of paintings that are not disentitled, paintings
to which Paladino gives a title, there operates still a certain re-
leasement from language, a certain reign (even though not abso-
lute) of untranslatability. For, by the artist’s testimony, his titles are
not such as to subsume the painting under a governing signi¤ca-

tion: “I never give titles that convey a particular meaning, which
could trap one into reading the work in strictly literary and sym-
bolic terms.” Instead, writes Paladino, “The title of a work always
represents for me the side that is disquieting [or: displacing—il lato
spiazzante] for the interpretation of the work.”5 The title, one could
say, serves precisely to disturb, interrupt, and displace all interpre-
tation oriented to replacing the work’s visible presentation with
linguistic signi¤cations; it serves to block—to secure the painting
against—all hermeneutical efforts to translate visibility into signi-
¤cation.
This is nowhere more transparently the case than in those in-
stances where silence or silent belongs to the title. There are several
such titles: a painting from 1977 is entitled Silence, I am retiring to paint
(Silenzioso, mi ritiro a dipingere); two others from 1979 and 1980, re-
spectively, bear the same remarkable title Silent Red (Rosso silenzioso).6
As the word silence, when uttered, breaks the very silence to which
it refers, so in these titles the word, re®ected back to itself, frees
the paintings to their silence, says and yet releases their untrans-
latability. Most forcefully so when silent is conjoined with the name
of something distinctively visible, with a name such as red: the red
that is painted—as in the two paintings entitled Silent Red—is not
the red that can be said; it is a red that is only to be seen, a si-
lent red.
As with the indescribable red of an Aegean sunset. Or the un-
speakably beautiful red of a rose. One senses somehow that such
sights are to be beheld in silence. One senses, no matter what one
may say, that they are untranslatable.
What is perhaps most remarkable about the paintings (and in-
deed the drawings too) in the cycle bearing the cipher-title EN
5. Mimmo Paladino, letter to author, January 1991.

6. The 1979 work entitled Silent Red is reproduced in Achille Bonito Oliva with
Mimmo Paladino, EN DE RE, 37; as is Silence, I am retiring to paint (ibid., 15). The
latter, as well as Silent Red from 1980, is included in the catalogue of the Villa delle
Rosa exhibition. Both are reproduced also in Paladino: una monogra¤a / a monograph
(Milan: Charta, 2001), 65, 90; this retrospective collection also includes im-
ages of several related works: two from 1979 entitled Silence (Silenzioso) (ibid., 76,
79) and another from 1978 designated Untitled (Senza titolo) (ibid., 68), which re-
sembles the 1980 Silent Red in showing an only slightly interrupted ¤eld of red.
118 On Translation
2. Mimmo Paladino. Senza
titolo [Untitled], 1989. Oil
on canvas and wood, 108
× 88 × 15 cm.
DO RE is the way in which the release, the withdrawal, that be-

longs to painting as such is set into the work itself. Each of the
paintings has a double structure (see ¤gure 2): there is a lower
surface over which is placed an obfuscating panel, which conceals
a portion of the lower surface and whatever images may be de-
picted on that surface. In most of the paintings—or rather, works,
for the double structure makes them exceed painting—there are
also images on the obfuscating panels. As a result of this double
structure and the artistic handling of it, there is in each case set
into the work a withdrawal of images, a retreating of images be-
hind other images. In other words—as if words were not precisely
the issue—the work gathers and distributes the images in such a
way as to present their disappearance, their escape from view. In
this way the works present the painterly analogue of their own
untranslatability: as the images escape from view and are pre-
sented in the work as escaping from view, so the painting itself,
that is, what is presented in the painting, escapes from words.
Even—it must be said—from the very words in which it has just
been said. By presenting the disappearance of images, the painted
work also presents—in the only way it could be presented in a
painted work, in the very element of painting and thus necessarily

only by outdistancing itself, as from the other side language out-
distances itself—the untranslatability that belongs to the painted
work.
In the case of painting, untranslatability does not have to do,
then, with linguistic difference, with the difference between lan-
guages in which one would—but cannot—say the same, that is,
speak tautologically. It has to do, rather, with the difference be-
tween language and the visible; more precisely (though still clas-
sically), it has to do with the relation—or nonrelation—between
linguistically signi¤able meanings and con¤gurations of the visible
in which the visible is brought to present its very visibility. Ulti-
mately, this untranslatability has to do with the very difference
that Socrates opened up—and opened to questioning—when he
ventured his second sailing, turning from visibly present things to
l1goi. One could call this the difference of all differences, the gi-
gantic difference, recalling that it was this difference, the question
of this difference, that provoked the contention described in Plato’s
Sophist as a gigantomax,a per- t+w o[w,aw. Western philosophy
was eventually to domesticate this difference in a series of trans-
lations of the words noht1n and acsyht1n, thereby stabilizing
also the sense of translation, giving it its classical determination.
If it is only since Nietzsche that philosophy has learned again how
to open this difference to questioning, painting can also attest—
has perhaps always attested—to its questionableness, to its utter
irreducibility, to the untranslatability that separates what, none-
theless and most remarkably, are called by the same name: sense.
Music would attest even more forcefully, if it were possible, to
the untranslatability of sense into sense: the impossibility of say-
ing in words what is sounded in a musical composition is so
patent as to be proverbial. Thus music would add its voice to that
by which philosophy would open to questioning anew the gigantic
difference between—and as the difference between—that which
is made manifest through the force of words and that which can
be seen, heard, or otherwise sensed.
And yet, the case of music is different. For however untranslat-
able music may be into words, music can be and often is put into
words; perhaps even in an originary way music is linked to song,
to the human voice. Without weakening in the least the utter un-
120 On Translation
translatability of music into words, one will need to say that none-
theless there is a certain af¤nity between music and words, espe-
cially between music and such words as can be sung, namely, po-
etry. In their untranslatable difference, music and poetry have a
profound af¤nity and can come together in an accord that is mu-
tually enhancing.
Kant caught a glimpse of this af¤nity. In the Critique of Judgment
he links ¤ne art to the expression of aesthetic ideas and as a result
divides the ¤ne arts according to the analogy between the arts and
the way expression functions in speech. Taking linguistic expres-
sion to consist of the three moments, word, gesture, and tone,
Kant divides the ¤ne arts, accordingly, into arts of speech (poetry
and oratory), arts corresponding to gesture (the visual/formative
arts—die bildenden Künste), and those corresponding to tone, which
for Kant are arts of the play of sensations. Music falls in this third
group. What music shares with speech, hence with poetry, is thus
tone, and it is in their respective tones that music and poetry have
an af¤nity. Kant writes: “Every linguistic expression has in its con-
text a tone appropriate to its meaning.” One could say: within lan-
guage, within poetry in particular, there is always already some-
thing like music, a kind of protomusic. Within language there is
always already another language, an untranslatable language of
tones. Music lets this other language sound outside poetry and its
proper language: “the art of music employs this language all by
itself in its full force.”7
There is a further consequence, one that could also be reached
by another route quite independent of Kant’s systematic considera-
tions (which in other regards have the effect of ranking music as
the lowest of the ¤ne arts or even as not quite a ¤ne art). The con-
sequence can be stated thus: music is always already mixed into
poetry, always already inherent in it, and for this reason music
can come to supplement poetry. Despite the untranslatability that
separates music decisively from poetry, the art of music can let
sound outside poetry the very music, the protomusic, the language
of tones, inherent in a poetic work; and thereby, without violat-
7. Kant, Kritik der Urteilskraft, §53.

ing the immediate untranslatability, it can supplement the poetic
work—can, in another, more profound sense of this word, trans-
late it.
As in the Overture and Incidental Music that Mendelssohn
composed for A Midsummer Night’s Dream. As in the Scherzo’s musical
presentation of the world of the fairies. As in the eerie chromatic
Andante’s setting the tone for Oberon’s and Puck’s casting of the
spell and in its later inversion as Oberon undoes the spell upon
Titania. As in the parody funeral march’s matching perfectly in
tone the scene of the performance of “Pyramus and Thisbe.” As in
the Nocturne’s sounding of the wood’s magic and of the depth of
the lover’s sleep. As in the musical interlude that transports the lis-
tening spectator from the scene of Hermia’s frantic search for
Lysander to that of the entrance of the mechanicals, when, as if to
the jocular tune heard, they arrive at the “green plot” that is to be
their stage, among them, soon to be translated, Nick Bottom the
weaver.
122 On Translation
General Index
Actualitas, 17 Gadamer, Hans-Georg, 32–34, 71–73, 88,

Aeschines, 65 89, 98, 103–105
Aristotle, 48–50, 106 Globalization, 6
Artwork, 33–34, 80n; in Kant, 13n. See also
Painting Hackforth, Reginald, 76, 78, 91
Aufgabe, 18 Hegel, G. W. F., 3, 16, 17, 80n, 103
Ausdruck, 24 Heidegger, Martin, 5–6, 17–19, 34, 72n, 84,
101–102
Benjamin, Walter, 15–16, 80n, 88, 107–10, 112 Highlighting (Überhellung), 72–79, 104
Boswell, James, 81–82
Brann, Kalkavage, Salem, 77, 78, 79, 92 Indication (l’indice), 106
Chalcidius, 5 Jakobson, Roman, 23, 46–47, 50, 83–84

Cicero, 23n, 65–67 Johnson, Samuel, 80–83
Ciphers, 116; in Paladino, 116, 118–20
Classical determination of translation, xi, Kant, Immanuel, 1–2, 13–15, 98, 99n, 121
51, 62–71, 98; in Samuel Johnson, 83; Kirwan, Christopher, 10n
limit of, 71–73; origin of, 120; and the
protoclassical determination, 59–62 Language: Greek understanding of, 64–65;
Colonialism, 10 “pure” language (Benjamin), 108–109
Comedy, 37; in Plato’s Cratylus, 63 Leibniz, Gottfried Wilhelm von, 9–13, 15, 16
Croce, Benedetto, 104 Locke, John, 67–71
Locus, 5
De¤nition: in Locke, 68–70
Demosthenes, 65 Materia, 12
Dent, R. W., 38n Measure, 50, 111; loss of (in Benjamin),
Derrida, Jacques, 4n, 34, 47, 106–107 109–10; “radical” measures, 17–18; in
Drama, 19, 33–37; in A Midsummer Night’s the Critias, 51, 58–59; and imagina-
Dream, 30, 31–32 tion, 99
Music, xii, 120–22
Figal, Günter, 97
Force of imagination, 36–37, 98–99; in Nachbildung, 72, 103
Schlegel, 94–98, 99–100, 110–11 Nature, 68n; in Kant, 14; in Locke, 68n
Force of names (t|n t3n <nom$tvn Nietzsche, Friedrich, 24–25, 62, 86–87,
d4namin), 110–11, 113; in the classi- 110, 120
cal determination, 62–65; in Plato’s Nizolius, Marius, 11
Critias, 55, 57–59; in Cicero, 65–67; in
Locke, 67–71; of Greek words, 16–18; Painting, xii, 114–20
in music, 120–22 Pannwitz, Rudolph, 88
Fowler, Harold North, 76n, 78, 91n Phonocentrism, 106
Freud, Sigmund, 7–9 Poetry: Solon’s (in Critias), 55, 59; transla-
123
tion of, 17, 64, 80–81ff., 84–85, 88, Transfero, 23, 29
105, 112–14, 121–22 Translation: as betrayal, 72, 104; as
Polysemy, 22, 25, 46; of Onoma, 54 conveyance, 31; as entrancing, 28;
Proper names, 28, 47–48, 85, 92 as handing-down (Überlieferung), 17;
Psychoanalysis, 7–9 history of the word, 22–23; as inter-
pretation, 7–9, 21, 23, 29, 118; into an-
Quintilian, 35 other language, 28; as change in me-
dium, 30; nonreciprocity, 100–101,
Restitution of meaning, 64; in Benjamin, 102–103; overtranslation, 101–103; in
107–10; in Cicero, 67; in Gadamer’s theatre, 19–20, 27–31, 37–45; and
hermeneutics, 71, 104–105; impossi- thinking, 18–19; as transformation,
bility of, 83–85; in Samuel Johnson, 15–17, 27, 88, 101, 106–108, 110–11, 113
83; in Schlegel, 95 Translatus, 23
Tredennick, Hugh, 77, 91
Schlegel, August Wilhelm, 28–30, 48, 73–
75, 84ff., 92–97, 99–100, 111 Überhellung, 72, 104
Schleiermacher, Friedrich, 76, 78, 91 Überlieferung, 17
Selbstbestand, 12 Unverborgenheit, 18
Semantic displacement, 15 Unwesen, 12
Substantia, 12
Suiectum, 5
Synonym, 70 Voss, Johann Heinrich, 16
Taylor, Thomas, 5 Wardy, Robert, 10n

Technology, 7 Will to power, 17
124 Index
Greek Word Index
acsyht1n / noht1n, 62 yevr,a, 22

#l}yeia, 18 yum1w, 79
gr$mata, graf1mena, 48, 64 cd)a, 17, 71
de4terow plo6w, 91–92 l1gow, 2, 35, 49, 101
dial)gesyai, 2 metaf)rv, 24
di$noia, 60, 71; as “meaning,” 63 On, 18
dianoe.syai, 2 Onoma, 49, 54, 64
diapuny$nomai, 57 pr$jiw, 22
d4namiw, 67 ]+ma, 49
eckas,a, 36 t)xnh, 34, 38
eqnai, 18 Spoke,menon, 5
\n)rgeia, 17 f$rmakon, 76–77
Trgon, 35 xQra, 5, 53–54
125
JOHN SALLIS is Edwin Erle Sparks Professor of Philosophy at
Pennsylvania State University. His previous books include Force
of Imagination: The Sense of the Elemental; Chorology: On Beginning in
Plato’s Timaeus; Shades—Of Painting at the Limit; Being and Logos:
Reading the Platonic Dialogues; Double Truth; Stone; Delimitations; Cross-
ings: Nietzsche and the Space of Tragedy; Echoes: After Heidgegger; Spac-
ings—Of Reason and Imagination; The Gathering of Reason; and Phe-
nomenology and the Return to Beginnings.

John Sallis On Translation 1

Uploaded by

Copyright:

Available Formats

John Sallis On Translation 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

John Sallis On Translation 1

Uploaded by

Copyright:

Available Formats

On Translation

Studies in Continental Thought

Translation goes astray.

My work on the theme of translation goes back to a lecture pre-

What would it mean not to translate? What would it mean to be-

1. Plato, Theaetetus 189e–190a.

The Dream of Nontranslation 3

4. See my discussion in “Babylonian Captivity,” Research in Phenomenology 22

The Dream of Nontranslation 5

The Dream of Nontranslation 7

5. Sigmund Freud, Die Traumdeutung, vol. 2 of Studienausgabe (Frankfurt a.M.:

8. See ibid., 136–46.

The Dream of Nontranslation 9

10. In the development of formal logic along lines anticipated by Leibniz’s

might be tempted to think of the whole process of formalising as a kind of

The Dream of Nontranslation 11

12. See Leibniz, Philosophical Papers and Letters, 367.

The Dream of Nontranslation 13

15. Ibid., §17.

The Dream of Nontranslation 15

16. Walter Benjamin, “Die Aufgabe des Übersetzers,” in Illuminationen (Aus-

The Dream of Nontranslation 17

The Dream of Nontranslation 19

24. At the outset of the play Hippolyta says:

Writing on translation inevitably becomes entangled. For nothing

1. Roman Jakobson, Language in Literature, ed. Krystyna Pomorska and Stephen

Scenes of Translation at Large 23

similarity seems to justify this transference” (Ad Herennium: De Ratione Dicendi,

Scenes of Translation at Large 25

Bless thee, Bottom, bless thee! Thou art translated.

Bottom’s state is designated by the very same word when in the

I led them on in this distracted fear,

In this scene of translation virtually the entire range of senses

. . . Purge thy mortal grossness so,

Scenes of Translation at Large 27

. . . upon this ®owery bed,

I have an exposition of sleep come upon me.

Here, then, is another—if archaic—sense of translation manifest

4. Here exposition is a malapropism for disposition.

Gott behüte dich, Zettel! Gott behüte dich!

Here Schlegel translates Shakespeare’s word into German by trans-

Scenes of Translation at Large 29

Schlegel translates the ¤nal sentence as:

sie soll Zettels Traum heissen, weil sie so seltsam angezet-

Here one sees from a particular angle how Schlegel’s translation of

He cannot be heard of. Out of doubt he is transported.

As to another world. As from this world to the next. As prepared

7. On the debate concerning the more speci¤c sense to be attributed to the

Scenes of Translation at Large 31

If we shadows have offended,

8. Gadamer observes that the suspension of everyday time is a signi¤cant link

9. See Gadamer’s discussion of what he terms Nachahmung and contrasts with

Scenes of Translation at Large 33

12. Gadamer, “Dichtung und Mimesis,” in vol. 5 of Gesammelte Werke, 85.

13. Quintilian, Institutio Oratoria VI.ii.29–30. In this passage Quintilian is refer-

Scenes of Translation at Large 35

Scenes of Translation at Large 37

Scenes of Translation at Large 39

Why, then may you leave a casement of the great chamber