John Sallis On Translation 1
John Sallis On Translation 1
John Sallis On Translation 1
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
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.
Tübingen
April 2001
xii Preface
On Translation
One The Dream of Nontranslation
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-
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.
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-
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
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
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
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
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
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).
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).
20 On Translation
Two Scenes of Translation at Large
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
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
26 On Translation
Quince then also reappears momentarily, and his exclamation
marks the high point of the scene. He exclaims:
he ¤nally declares:
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:
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.
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)
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:
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:
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
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
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.
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.
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)
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.
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
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.
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
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
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
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.
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
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
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
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$noia named in the other voice—but also is of such
force as to be capable of making manifest that which it names, of
bringing it to presence in that certain way appropriate to names.
Though this second move brings one from the foreign back to
one’s own, it is decidedly not a matter merely of retrieval, of grasp-
ing again what has been said, but of saying what has never yet
been said in one’s own voice.
As the passage on translation proceeds, it shifts ever more de-
cisively from focusing merely on proper names to consideration
of discourse as such. Nothing points this up so clearly as Critias’
remark immediately following the passage: “And these very writ-
ings [resulting from Solon’s writing down the names in transla-
tion] were in the possession of my grandfather and are now in
mine, and as a child I learned them by memory.” There is every
reason to believe that what was translated and written down in
and as these writings handed down to Critias himself from his
grandfather was the entire story as such and not merely the proper
names of those who ¤gured prominently in it.
One could conclude, then, that the passage from the Critias con-
stitutes a discourse on translation as such. From it issues the in-
augural determination of translation, the protoclassical determina-
tion, which subsequently comes to be stabilized by reference to
the distinction that itself comes to be stabilized as fundamental to
philosophy (so radically so as to determine the very sense of fun-
62 On Translation
tion the di$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.
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
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).
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
24. Ibid.
25. Ibid., III.iv.9.
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
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)
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)
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.
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
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.
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
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
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.
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:
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,
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.
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)
Schlegel translates:
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,
(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
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
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).
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).
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).
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-
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
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-
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.
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-
114 On Translation
1. Mimmo Paladino. Senza
titolo [Untitled], 1989. Oil
on canvas and wood, 108
× 88 × 15 cm.
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-
118 On Translation
2. Mimmo Paladino. Senza
titolo [Untitled], 1989. Oil
on canvas and wood, 108
× 88 × 15 cm.
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-
122 On Translation
General Index
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
124 Index
Greek Word Index
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.