to something that cannot be assumed by a subject, that the ēthos of this dis-
junction can only be testimony—that is, something that cannot be assigned to
a subject but that nevertheless constitutes the subject’s only dwelling place, its
only possible consistency.
3.22. Giorgio Manganelli has written of a special figure of heteronymy, which
he calls “pseudonymy squared” or “homopseudonymy.” It consists in using a
pseudonym that is in every respect identical to one’s own name. One day, one
of his friends tells him that he has published a book of which he knows nothing,
just as other times it had also happened to him that “sober-minded people” let
him know they have seen books with his first and last name on display in cred-
ible bookstore windows. ( Pseudonymy)2 brings the paradox of ontological heteronymy to an extreme point, since here it is not only an “I” that gives way to
another; this “other” even claims not to be “other,” but rather fully identical to
the “I,” something the “I” cannot but deny. “I had acquired and partially read a
book that an honest slanderer, an historicist, a specialist of anagraphs had called
‘mine.’ But if I had written it, if there had been an ‘I’ capable of writing a book,
that book, what could explain the absolute, irritating strangeness that divided
me from what had been written?” (Manganelli 1996: 13).
REMNANTS OF AUSCHWITZ
849
With respect to the simple “I,” the homopseudonym is absolutely foreign
and perfectly intimate, both unconditionally real and necessarily non-existent,
so much so that no language could describe it; no text could guarantee its
consistency. “So I had written nothing. But by ‘I’ I meant the person with my
name and without pseudonym. Did the pseudonym write? It’s likely, but the
pseudonym pseudowrites; it is technically speaking unreadable by the ‘I,’ al-
though it might be readable by the squared pseudonym ‘I,’ who obviously does
not exist. But if the reader is non-existent, I know what he can read: what can
be written by the degree zero pseudonym, something that cannot be read by
anyone who is not the squared pseudonym, the non-existent one. In fact, what
is written is nothing. The book means nothing, and in any case I cannot read
it without giving up my existence. Maybe it’s all a prank: as will be made clear,
I have been dead now for many years, like the friend I met, and the book I’m
leafing through is always incomprehensible; I read it, I reread it, I lose it. Maybe
one has to die several times” (ibid.: 14).
What this terribly serious joke lays bare is nothing less than the ontological
paradox of the living-speaking (or writing)-being, the living being who can say
“I.” As a simple “I” with a name but no pseudonym, he can neither write nor
say anything. But every proper name, insofar as it names a living being, a non-
linguistic thing, is always a pseudonym (a “degree zero pseudonym”). I can only
write and speak as the pseudonym “I”; but what I then write and say is nothing,
that is something that could be read or heard only by a squared pseudonym, who
does not exist in himself, if not by taking the place of the first “I,” who then gives
up his existence (that is, dies). At this point, the pseudonym’s elevation to the
second power is complete: the “I” with a name but no pseudonym disappears in
the non-existent homopseudonym.
But the next question is: Who is speaking in Manganelli’s story, who is its
author? Who bears witness to the unease of this intimate strangeness? The “I”
without pseudonym, which exists but cannot write? Or the degree zero pseud-
onym, who writes the unreadable text of the first “I”? Or rather the third, the
squared pseudonym who reads, rereads, and loses the empty and incomprehen-
sible book? If it is clear that “I have been dead for many years,” who survives to
speak of this death? In the process of vertiginous, heteronymic subjectification,
it is as if something always survived, as if a final or residual “I” were generated
in the word “I,” such that the pseudonym’s elevation to the second power were
never truly completed, as if the squared “I” always fell back onto a new “I,” an
“I” both indistinguishable from and irreducible to the first.
850
HOMO SACER III
3.23. The term “to survive” contains an ambiguity that cannot be eliminated.
It implies the reference to something or someone that is survived. The Latin
supervivo— like the equivalent superstes sum— is in this sense constructed with the dative, to indicate the person or thing with respect to which there is survival. But from the beginning, the verb also has a reflexive form when referred
to human beings, which designates the striking idea of survival with respect to
oneself and one’s own life. In this form, the one who survives and the person
to whom something survives thus coincide. If Pliny can therefore say of a public
figure that “for thirty years he had survived his glory” ( triginta annis gloriae suae
supervixit), in Apuleius we already find the idea of genuine posthumous existence, a life that lives by surviving itself ( etiam mihi ipse supervivens et postumus) .
In the same sense, Christian authors can say that Christ—and every Christian
along with him—is both testator and inheritor insofar as he has survived death
( Christus idem testator et haeres, qui morti propriae supervivit); moreover, they also can write that the sinner survives on earth on account of being in truth spiritually
dead ( animam tuam misera perdidisti, spiritualiter mortua supervivere hic tibi).
This implies that in human beings, life bears with it a caesura that can trans-
form all life into survival and all survival into life. In a sense—the sense we have
encountered in Bettelheim—survival designates the pure and simple continu-
ation of bare life with respect to truer and more human life. In another sense,
survival has a positive sense and refers—as in Des Pres—to the person who, in
fighting against death, has survived the inhuman.
Let us then formulate the thesis that summarizes the lesson of Auschwitz: The
human being is the one who can survive the human being. In the first sense, it refers to the Muselmann (or the gray zone); it therefore signifies the inhuman capacity
to survive the human. In the second sense, it refers to the survivor; it designates
the human being’s capacity to survive the Muselmann, the non-human. When one looks closely, however, the two senses converge in one point, which can be
said to constitute their most intimate semantic core, in which the two meanings
momentarily seem to coincide. The Muselmann stands in this point; and it is in
him that we find the third, truest, and most ambiguous sense of the thesis, which
Levi proclaims when he writes that “they, the Muselmänner, the drowned are the complete witnesses”: the human being is the inhuman; the one whose humanity
is completely destroyed is the one who is truly human. The paradox here is that if the only one bearing witness to the human is the one whose humanity has been
wholly destroyed, this means that the identity between human and inhuman is
never perfect and that it is not truly possible to destroy the human, that some-
thing always remains. The witness is this remnant.
REMNANTS OF AUSCHWITZ
851
3.24. Concerning Antelme’s book, Blanchot once wrote that “man is the
indestructible that can be infinitely destroyed” (Blanchot 1993: 130). The word
“ indestructible” here does not mean something—an essence or human rela-
tion—that infinitely resists its own infinite destruction. Blanchot misunder-
stands his own words when he sees infinite destruction as the place of “the
human relation in its primacy,” as the relation to the Other (ibid.: 135). The
indestructible does not exist, either as essence or as relation; Blanchot’s sentence
must be read in another sense, one that is both more complicated and simpler.
“Man is the indestructible who can be infinitely destroyed”—like “the human
being is the one who can survive the human being”—is not a definition which,
like all good logical definitions, identifies a human essence in attributing a spe-
cific difference to it. The human being can survive the human being, the human
being is what remains after the destruction of the human being, not because
somewhere there is a human essence to be destroyed or saved, but because the
place of the human is divided, because the human being exists in the fracture
between the living being and the speaking being, the inhuman and the human.
That is: the human being exists in the human being’s non-place, in the missing articulation between the living being and logos. The human being is the being that is lacking to itself and that consists solely in this lack and in the errancy it opens.
When Grete Salus wrote that “man should never have to bear everything that he
can bear, nor should he ever have to see how this suffering to the most extreme
power no longer has anything human about it,” she also meant this much: there
is no human essence; the human being is a potential being and, in the moment
in which human beings think they have grasped the essence of the human in
its infinite destructibility, what then appears is something that “no longer has
anything human about it.”
The human being is thus always beyond or before the human, the central
threshold through which pass currents of the human and the inhuman, subjec-
tification and desubjectification, the living being’s becoming speaking and the
logos’ becoming living. These currents are coextensive, but not coincident; their
non-coincidence, the subtle ridge that divides them, is the place of testimony.
4
The Archive and Testimony
4.1. One evening in 1969, Emile Benveniste, Professor of Linguistics
at the Collège de France, suffered an attack on a street in Paris.
Without identification papers, he was not recognized. By the time he was iden-
tified, he had already suffered a complete and incurable aphasia that lasted until
his death in 1972 and kept him from working in any way. In 1972, the journal
Semiotica published his essay, “The Semiology of Language.” At the end of this
article, Benveniste outlines a research program that moves beyond Saussurian
linguistics, one that was never realized. It is not surprising that the basis for
this program lies in the theory of enunciation, which may well constitute Ben-
veniste’s most felicitous creation. The overcoming of Saussurian linguistics, he
argues, is to be accomplished in two ways: the first, which is perfectly compre-
hensible, is by a semantics of discourse distinct from the theory of signification
founded on the paradigm of the sign; the second, which interests us here, con-
sists instead “in the translinguistic analysis of texts and works through the elabo-
ration of a metasemantics that will be constructed on the basis of a semantics of
enunciation” (Benveniste 1974: 65).
It is necessary to linger on the aporia implicit in this formulation. If enunci-
ation, as we know, does not refer to the text of what is uttered but to its taking
place, if it is nothing other than language’s pure reference to itself as actual dis-
course, in what sense is it possible to speak of a “semantics” of enunciation? To
be sure, the isolation of the domain of enunciation first makes it possible to dis-
tinguish in a statement between what is said and its taking place. But does enun-
ciation not then represent a non-semantic dimension precisely on account of
this identification? It is certainly possible to define something like a meaning of
the shifters “I,” “you,” “now,” “here” (for example, “‘I’ means the one who utters
the present speech in which ‘I’ is contained”); but this meaning is completely
foreign to the lexical meaning of other linguistic signs. “I” is neither a notion
nor a substance, and enunciation concerns not what is said in discourse but the
pure fact that it is said, the event of language as such, which is by definition
852
REMNANTS OF AUSCHWITZ
853
ephemeral. Like the philosophers’ concept of Being, enunciation is what is most
unique and concrete, since it refers to the absolutely singular and unrepeatable
event of discourse in act; but at the same time, it is what is most vacuous and
generic, since it is always repeated without its ever being possible to assign it any
lexical reality.
What, from this perspective, can it mean to speak of a metasemantics
founded on a semantics of enunciation? What did Benveniste glimpse before
falling into aphasia?
4.2. In 1969, Michel Foucault also publishes Archaeology of Knowledge, which formulates the method and program of his research through the foundation of a
theory of statements ( énon cés). Although Benveniste’s name does not appear in
the book and despite the fact that Foucault could not have known Benveniste’s
last articles, a secret thread ties Foucault’s program to the one the linguist out-
lined. The incomparable novelty of The Archaeology of Knowledge consists in hav-
ing explicitly taken as its object neither sentences nor propositions but precisely
“statements,” that is, not the text of discourse but its taking place. Foucault was
thus the first to comprehend the novel dimension of Benveniste’s theory of enun-
ciation, and he was the first then to make this dimension into an object of study.
Foucault certainly recognized that this object is, in a certain sense, undefinable,
that archaeology in no way delimits a particular linguistic area comparable to
those assigned to the various disciplines of knowledge. Insofar as enunciation
refers not to a text but to a pure event of language (in the terms of the Stoics, not
to something said but to the sayable that remains unsaid in it), its territory can-
not coincide with a definite level of linguistic analysis (the sentence, the proposi-
tion, illocutive acts, etc.), or with the specific domains examined by the sciences.
Instead, it represents a function vertically present in all sciences and in all acts
of speech. As Foucault writes, with lucid awareness of his method’s ontological
implications: “the statement is not therefore a structure . . . ; it is a function of
existence” (Foucault 1972: 86). In other words: enunciation is not a thing de-
termined by real, definite properties; it is, rather, pure existence, the
fact that a
certain being—language—takes place. Given the system of the sciences and the
many knowledges that, inside language, define meaningful sentences and more
or less well formed discourses, archaeology claims as its territory the pure taking
place of these propositions and discourses, that is, the outside of language, the brute fact of its existence.
In this way, Foucault’s archaeology perfectly realizes Benveniste’s program
for a “metasemantics built on a semantics of enunciation.” After having used a
854
HOMO SACER III
semantics of enunciation to distinguish the domain of statements from that of
propositions, Foucault establishes a new point of view from which to investigate
knowledges and disciplines, an outside that makes it possible to reconsider the
field of disciplinary discourses through a “metasemantics”: archaeology.
It is certainly possible that Foucault thus merely dressed up old ontology,
which had become unacceptable, in the modern garb of a new historical meta-
discipline, thereby ironically proposing first philosophy not as a knowledge, but
as an “archaeology” of all knowledges. But such an interpretation fails to recog-
nize the novelty of Foucault’s method. What gives his inquiry its incomparable
efficiency is its refusal to grasp the taking place of language through an “I,” a
transcendental consciousness or, worse, an equally mythological psychosomatic
“I.” Instead, Foucault decisively poses the question of how something like a sub-
ject, an “I,” or a consciousness can correspond to statements, to the pure taking
place of language.
Insofar as the human sciences define themselves by establishing a linguistic
stratum that corresponds to a certain level of meaningful discourse and linguis-
tic analysis (the sentence, the proposition, the illocutive act, etc.), their subject
is naively identified with the psychosomatic individual presumed to utter dis-
course. On the other hand, modern philosophy, which strips the transcendental
subject of its anthropological and psychological attributes, reducing it to a pure
“I speak,” is not fully aware of the transformation this reduction implies with
respect to the experience of language; it does not recognize the fact that language
The Omnibus Homo Sacer Page 133
