University of Pennsylvania Libraries

NOTICE W ARNING CONCERNING COPYRIGHT RESTRICTIONS

The copyright law of the United States (title 17, United States Code) governs the
making of photocopies or other reproductions of copyrighted material.
Under certain conditions specified in the law, libraries and archives are
authorized to furnish a photocopy or other reproduction. One of these specific
conditions is that the photocopy or reproduction is not to be used for any
purpose other than private study, scholarship, or research. If a user makes a
request for, or later uses, a photocopy or reproduction for purposes in excess of
fair use, that user may be liable for copyright infringement.
This institution reserves the right to refuse to accept a copying order if, in its
judgment, fulfillment of the order would involve violation of copyright law.
This notice is posted in compliance with Title 37 C.F.R., Chapter II, Part 201.14

Graduate Faculty Philosophy Journal

Volume 10, Number 1
Spring 1984

The Idea of Language:

Some Difficulties in Speaking
About Language
Giorgio Agam.ben

ANYONE who has been educated----or has lived-in a Christian or

Jewish environment is familiar with the word: revelation. Such familiarity, however, does not necessarily entail the ability to define its
meaning. I would like to begin my reflections with an attempt to define
the term. Naturally, I don't mean to bring up a theological problem. On
the contrary, I am convinced that a correct definition of this term is
neither irrelevant to the theme of our meeting nor alien to the study of
philosophy: to that discourse which, it has been said, can speak of everything, provided that it speak first of all ofthe fact that it is speaking
ofit.
The constant feature that characterizes every concept ofrevelation is
its heterogeneity with respect to reason. This does not simply mean
that the content of revelation must necessarily strike reason as absurd---even though the Church fathers often insisted on this point. The
difference in question is far more radical; it involves the very nature of
revelation: its very structure.
If the content of a revelation-no matter how absurd, for example,
that pink donkeys sing in the Venusian sky-were something that
human reason and language could still know and say through their
own power, it would in itself cease to be a revelation. What it teIls us
must therefore be something that not only could not be known without
such revelation, but further, it must be something that conditions the
very possibility of knowledge in general.
It is this radical difference at the level of revelation that Christian
theologians express, saying that the sole content ofrevelation is Christ
himself, that is, the word ofGod. Jewish theologians affirm, similarly,

141

GRADUATE FACULTY PHILOSOPHY JOURNAL

that the revelation of God is his name. When Saint Paul wants to explain to the Colossians the economy of divine revelation, he writes: "...
to fulfill the word of God, even the mystery which hath been hid from
ages and from generations, but is now made manifest to his saints"
(Col.I, 26-27). In these lines, "the mystery" is in apposition to "the word
of God" (6 AOYOC; 'toO eEOU). The mystery that was hidden and is now
revealed does not concern this or that natural or supernatural event,
but simply the word of God.
So if the theological tradition has always understood revelation as
something human reason cannot know on its own, this means only that
the content ofrevelation is not a truth that can be expressed in the form
of linguistic propositions regarding that which exists (even if a supreme being), but is rather a truth that concerns language itself: the
very fact that language (and hence knowledge) exists. The meaning of
revelation is that man can reveal what exists through language, but
cannot reveal language itself. In other words, man sees the world
through language, but does not see language. This invisibility ofthe revealing in what it reveals is the word of God, it is revelation.
Therefore theologians say that the revelation of God is at the same
time his concealment or, further, that in the word God is revealed in his
very incomprehensibility. It is not simply a matter ofa negative determination or of a lack ofknowledge, but of an essential determination of
divine revelation, which a theologian has expressed in these terms: "supreme visibility in the deepest obscurity" and "revelation ofsomething
unknowable." Once again, this simply means, what is here revealed is
not an object about which there would be much to be known, but that
cannot be known for lack of adequate instruments ofknowledge. What
is revealed here is the unveiling itself, the very fact that knowledge
and the opening of a world exist.
In this horizon, the construction of trinitarian theology seems to be
the most rigorous and coherent attempt to conceive the paradox ofthat
primordial statute ofthe word that the prologue ofthe Gospel according
to John expresses by saying: EV pxn ~v 6 AOYOc;, In the beginning
was the Word. The unitrinitarian movement of God that has become
familiar to us through the Nicaean symbol (Credo in unum Dominum
Iesum Christum [ilium dei unigenitum et ex patre natum ante omnia
saecula, Deum de Deo, lumen de lumine ... genitum non factum, consubstantialem Patri ...") says nothing about worldly reality, has no
ontic content, but takes into account the new experience of the word
which Christianity has brought to the world. To use Wittgenstein's
terms, the symbol says nothing about how the world is, but reveals that
the world is, that there is language. The word, which is absolutely in
142

AGAMBENITHE IDEA OF LANGUAGE

the beginning, which is therefore the absolute presupposition, only presupposes itself, it has no precursor that can explain it or reveal it in
turn (there is no word for the word!); and the trinitarian structure ofthe
'word is only the word's self-revelation. Now this revelation ofthe word,
this presupposing nothing-which is the only presupposition-is God:
"and the word was ... God."
The real meaning ofrevelation, therefore, is to show that every word
and every human cognition have roots and a foundation in an opening
that transcends them infinitely; yet at the same time, this aperture
'concerns only language itself, the possibility and existence oflanguage.
As the great Jewish theologian and leader of the neo-Kantian school,
Hermann Cohen, said: the meaning of revelation is that God is not revealed in something, but to something, and that, therefore, his revelation is simply die Schpfung der Vernunft, the creation of reason. Revelation does not mean this or that statement about the world, nor what
can be said through language, but that there is the world, that there is
language.
There is language - what can such a statement mean?
It is from this point ofview that we must take a look at the locus classicus wherein the problem of the relationship between revelation and
reason has been debated: namely, the ontological argument of Anselm.
For, as many promptly pointed out to hirn, it is not true that the mere
uttering ofthe name God, of quid maius cogitari nequit, necessarily implies the existence of God. Yet a being whose mere linguistic naming
implies existence does exist, and this being is language. The fact that I
speak and someone listens does not imply the existence of anything except of language. Language is that which must necessarily presuppose itself. What the ontological argument proves, therefore, is that if
men speak, ifthere are reasoning animals, then there is a divine word.
Which means simply that the signifying function always pre-exists.
(Provided that God is the name of the pre-existence of language, of its
dwelling in the arche - then, and only then, does the ontological argument prove the existence of God.) But this pre-existence, contrary to
what Anselm thought, does not belong to the realm of significant
speech; it is not a proposition endowed with meaning, but a pure event
of language before or beyond all particular meaning. In this light, it is
useful to reread the objection that a great but little-known logician of
the eleventh century, Gaunilo, opposes to Anselm's argument. When
Anselm declared that the uttering of the word God necessarily implies
for the person who understands it the existence ofGod, Guanilo posited,
in objection, the experience of an ignoramus (an idiot, as he says) or a
143

GRADUATE FAClJLTY PHILOSOPHY JOURNAL

barbarian who, faced by signifying speech, surely understands that

there\is an event of language, that there is, Gaunilo says, a vox, a
human voice, but can in no way grasp the meaning of the utterance.
This idiot or this barbarian, Gaunilo continues, does not conceive of
the voice itself, that is, the sound of the syllables and of the letters,
which is a thing somehow true, so much as he thinks ofthe meaning of
the heard voice; and not as it is conceived ofby those who know what
is usually signified by that voice (and who conceived of it, therefore,
according to the thing [secundum rem]); but rather, he thinks of it as
it is thought ofby those who do not know the meaning and who think
only according to the movement of the mind that tries to represent to
itself the effect and the meaning of the heard voice.

The perception no more of a mere sound but not yet of a meaning, this
"thought of the voice alone" (cogitatio secundum vocem solam, as
Gaunilo calls it) opens up a primevallogical dimension that, denoting
the pure "taking place" of language, without any specific event of
meaning, shows that there is still a possibility of thought beyond signifying propositions. The most original logical dimension that is involved in revelation is not, therefore, that of the signifying word, but
that of a voice which, without signifying anything, signifies significance itself. (This is the sense of theories like that of Roscellinus, of
whom it was said that he had discovered "the meaning ofthe voice" and
had affirmed that the universal essences were only flatus vocis. Here
flatus vocis is not the simple sound, but in the sense explained above,
the voice as a pure indication of an event of language. And this voice
coincides with the most universal dimension of meaning, with being.)
This endowment of a voice for language is God; is the divine word. The
name of God, that is, the name that names language, is he~ce (as the
mystical tradition has never tired of repeating) a meaningless word.
In the terms of contemporary logic, we could then say that revelation
means that, if such a thing as a metalanguage exists, it is not a signifying statement, but a pure non-signifying voice. That there is language
is equally certain and incomprehensible, and this incomprehensibility
and this certainty constitute faith and revelation.
The chief difficulty inherent in philosophical exposition involves this
same order of problems. In fact philosophy is not concerned only with
what is revealed through language, but also with the revelation oflanguage itself. A philosophical exposition is, in other words, one that,
whatever it speaks of, must take into account the fact that it is speaking of it; a philosophical statement is one that, in everything that it
144

AGAMBEN/THE IDEA OF LANGUAGE

says, says above aillanguage itself. (Hence the proximity, but also the
separation, between philosophy and theology, a link at least as old as
the Aristotelian definition of first philosophy as eEOAOYlK~, theological.)
All of this could also be expressed by saying that philosophy is not a
view of the world, but a view of language, and, in fact, contemporary
thought has followed this path with all too much enthusiasm. However,
a difficulty arises here from the fact that (as is implicit in Gaunilo's definition ofvoice) a philosophical exposition cannot be simply a discourse
that has language as its subject, a metalanguage that speaks of language. The voice says nothing, but shows itself precisely as logical
form, according to Wittgenstein, and therefore cannot become the subject of discourse. Philosophy can lead thought only to the boundaries of
the voice: it cannot say the voice (or at least, so it seems).
Contemporary thought is resolutely aware of the fact that an ultimate and absolute metalanguage does not exist and that any construction of a metalanguage remains trapped in a regression to infinity. All the same, the paradox ofphilosophy's intention is precisely that
of an utterance that would speak oflanguage and show its limits without having a metalanguage at its disposal. In this way, philosophy
comes up against what is represented as the essential content ofrevelation (and, perhaps, also ofpoetry): logos en arche, the word is absolutely
in the beginning, it is the absolute premise, or, as Mallarme once wrote,
"the word is the beginning developed through the negation of every beginning." And it is against this dwelling of the word in the beginning
that a logic and a philosophy (as weIl as a poetry) aware oftheir tasks
must always again be measured.
Ifthere is a point on which contemporary philosophies seem to agree,
it is precisely the acknowledgement ofthis premise. And so hermeneutics assurnes the irreducible priority of the signifying function, affirming (according to the declaration of Schleiermacher that stands as a
motto to Wahrheit und Methode) that "in hermeneutics there is only
one presupposition: language"; or else by understanding, as Apel does,
the concept ofWittgenstein's "language game" as a transcendental condition of all knowledge. This apriori is, for hermeneutics, the absolute
premise which can be reconstructed and be made self-conscious, but
cannot be overcome. Coherently with these premises, hermeneutics can
set itself up only as the horizon of an infinite tradition and interpretation whose ultimate meaning and foundation must necessarily remain
unsaid. It can question itself about how comprehension occurs, but the .
fact that there is comprehension is what, remaining unthought, makes
145

GRADUATE FACULTY PHILOSOPHY JOURNAL

all comprehension possible. "Every act of the word," Gadamer writes,

"in the act of its occurence, makes at onee present the unsaid to which
it, as reply and reference, refers." It is therefore clear that hermeneutics, though harking back to Hegel and Heidegger, tends to neglect the
very aspect ofthese philosophers' thought eoncerning, on the one hand,
absolute knowledge and the end ofhistory and, on the other, theEreignis and the end of the history of being.
In this sense, hermeneutics is opposed - but not so radically as it
might seem - to such languages as science and ideology which, while
presupposing more or less knowingly the pre-existence of the signifying function, ignore this premise and allow its productivity and nullifying power to operate without constraint. And, in truth, it is difficult to
see how hermeneutics could convince the advocates ofthese attitudes to
renounce their positions~ If the foundation is, in any case, unsayable
and irreducible, if it always anticipates speaking man, casting hirn into
a history and adestiny, then a thought that recalls and deals with this
presupposition seems ethically the equivalent of one which, abandoning itself to its fate, carries out to the end (and there is, actually, no
end) the violence and lack of foundation of such a thought.
It is therefore no accident if, according to an authoritative current of
contempory French thought, language is, indeed, maintained in the beginning, but this dwelling of the logos in the arche has the negative
structure of writing and of gramme. There is no voice for language;
rather, from the start, language is a trace and an infinite self-transcendence. In other words, language, which is in the beginning, is the nullification and the deferment of itself, and significance is only the inexhaustible cypher of this lack of foundation.
It is legitimate to ask oneselfwhether this awareness ofthe pre-existence of language that characterizes contemporary thought can really
fulfill the task ofphilosophy. It could be said that here thought considers its task done by the very acknowledgement ofwhat constituted the '
more genuine content offaith and revelation: the embeddedness ofthe
;logos in the arche. What theology declared incomprehensible to reason
is now aceepted by reason as its premise. All comprehension is founded
on the ineomprehensible. But, in this way, isn't the very thing that
should be the philosophic task par excellence abandoned-namely the
dissolution ofthe presupposition? Wasn't philosophy the utterance that
was meant to be free ofall premises, even ofthe most universal premise
expressed in the formula: there is language? For isn't philosophy a matter of comprehending the incomprehensible? Perhaps in the very abandoning of this task, which sentences the handmaiden philosophy to a
secret marriage with her mistress theology, lies the present difficulty of
146

AGAMBEN/THE IDEA OF LANGUAGE

philosophy, just as the difficulty of faith coincides with its acceptance

by reason. The abolition of the frontiers between faith and reason also
marks their crisis, that is, their reciprocal judgment.
Contemporary thought has come close to that limit, beyond wliich a
new unveiling of language seems no longer possible. The arche character ofthe logos is now completely revealed, and no new figure ofthe divine, no new historie destiny can arise from language. Language, in the
very moment it is located in the beginning, also reveals its absolute
anonymity. There is no name for the name, there is no metalanguage,
not even in the form of a non-signifying voice. If God was the name of
the language, "God is dead" can mean only that there is no longer a
name for language. The fulfilled revelation of language is a word completely aband6ned by God. Man is cast into language without having a
voice or a divine word that guarantees hirn a possibility of escape from
the infinite play of signifying propositions. And so, at last, we are left
alone with our words, alone for the first time with language, abandoned
by any further foundation. This is the Copernican revolution that the
thought of our time has inherited from nihilism: we are the first people
who have become completely aware of language. What previous generations thought of as Muse, God, Being, Spirit, Unconscious, we see
clearly for the first time for what they are: names oflanguage. Thus all
philosophies, all religions and all knowledge that have not become
aware of this turning point, belong for us irrevocably to the paste The
veils that theology, poetry, ontology, and psychology have drawn over
what is human have now fallen and, one by one, we restore them to
their proper place in language, which has dispelled from itself everything divine, everything unsayable: it stands entirely revealed, absolutely in the beginning. Just like a poet who can finally see the face of
his muse, so the philosopher now looks at language face to face (that's
why - muse being the name of the most original experience of language - Plato says that philosophy is "the supreme music").
_
Nihilism undergoes this same experience of a word abandoned by
God; but it interprets the ultimate revelation of language from the
standpoint that there is nothing to reveal, that the truth of language
attests to the nothingness of every thing. The absence of a metalanguage thus becomes the negative form of a foundation, and nothingness
the last veil, the last name fo_r language.
If, at this point, we take up Wittgenstein's image ofthe fly trapped in
a bottle, we could say that contemporary thought has finally acknowledged the inevitability of the bottle whose prisoner the fly iso The preexistence and the anonymity of the signifying function constitute the
147

GRA.DUATE FACULTY PHILOSOPHY JOURNAL

presupposition that always anticipates the speaking man and from

which there seem to be no exits of any kind. Men are doomed to understand each other in language. But, once again, the actual project that
was originally entrusted to that image is thus neglected: the possibility
that the fly could escape from the bottle.
The task ofphilosophy, therefore, must be resumed at the very point
where contemporary philosophy seems to abandon it. If it is true, in
fact, that the fly must begin by seeing the bottle within which it is trapped, what can such avision signify: What does seeing the boundaries of
language mean? (The bottle, indeed, is not a thing for the fly, but what
it sees things through.) Is it possible to conceive of an utterance which,
without being metalinguistic and plunging into the unsayable, says
language itself and shows its limits?
An ancient tradition of thought locates this possibility in the theory
of ideas. Contrary to the interpretation that sees in it the unsayable
foundation of a metalanguage, at the basis of the Platonic theory of
ideas lies an acceptance without reservations of the anonymity of language and ofthe homonymy that governs it (this is how we must understand Plato's insistence on the homonymy between ideas and things as
weIl as Socrates' rejection of all misology). This same finiteness and
ambiguity of human language opens the way to a "dialectical journey"
of thought. For, if every human word presupposed another word, that
is, ifthe presupposing power oflanguage never ended, then truly there
could be no experience of the boundaries of language. Aperfect language, on the other hand, from which all homonymy had disappeared
and in which all signs were univocal, would be a language with no ideas
whatsoever.
The idea lies entirely in the play between the anonymity and the
homonymy of language. Neither does the One exist and haue a name,
nor does it not exist and not haue a name. The idea is not a word (a
metalanguage) nor is it the vision of an object outside language: it is
the vision (t5ELV) of language itself. (This is the genuine content of every
aporia, which makes it a pattern of philosophical exposition). For language, which mediates all things and all knowledge for man, is itself
immediate. Nothing immediate can be reached by speaking men - except language itself, except mediation itself. This immediate mediation
represents for man the only possibility of reaching a beginning freed
from all presupposition, even from the presupposition of language itself; of reaching, in other words, that o:pX~ avuno8EToc; which Plato,
in The Republic, presents as the TEAOc; , as the fulfillment and end of
aUToc; Aoyoc;, of language itself, and at the same time as the "thing
itself' and the concern of man.
148

AGAMBENITHE IDEA OF LANGUAGE

No true human community can, in fact, rise on the basis of a presupposition - whether it be that of a nation or a tongue or even the apriori
of communication of which hermeneutics speaks. What unifies men is
not a nature or a divine voice or the common imprisonment in signify. ing language, but the vision oflanguage itself- and, therefore, the experience ofits boundaries, ofits end. The only true community is a community without presupposition. Pure philosophical exposition therefore cannot be the exposition of one's own ideas on language or on the
world, but an exposition of the idea of language.

149