that, as in the oath, the verbal act brings being into truth. This is not, as we have
seen, a magico-religious stage but a structure antecedent to (or contemporane-
ous with) the distinction between sense and denotation, which is perhaps not,
as we have been accustomed to believe, an original and eternal characteristic of
human language but a historical product (which, as such, has not always existed
and could one day cease to exist).
How, in fact, does the performative function? What permits a certain syn-
tagma to acquire, solely by its utterance, the efficacy of fact, casting off the
ancient maxim that would have it that words and things are separated by an
abyss? What is essential here, certainly, is the self-referential character of the
performative expression. This self-referentiality is not exhausted in the fact that
the performative, as Benveniste notes (ibid., 274), takes itself as referent, insofar
as it refers to a reality that it itself constitutes. Rather it is necessary to specify
that the self-referentiality of the performative is constituted always by means of
a suspension of the normal denotative character of language. The performative
verb is in fact necessarily constructed with a dictum that, considered in itself, has a purely denotative nature and without which it remains void and ineffective
( I swear does not have any force if it is not followed—or preceded—by a dictum that fills it in). It is this denotative character of the dictum that is suspended and called into question in the very moment it becomes the object of a performative
syntagma. Thus the denotative expressions “yesterday I was in Athens” or “I will
not fight against the Trojans” cease to be such if they are preceded by the per-
formative I swear. That is to say, the performative substitutes for the denotative relationship between speech and fact a self-referential relation that, putting the
former out of play, puts itself forward as the decisive fact. The model of truth
THE SACRAMENT OF LANGUAGE
343
here is not that of the adequation between words and things but the performa-
tive one in which speech unfailingly actualizes its meaning. Just as, in the state
of exception, the law suspends its own application only to found, in this way, its
being in force, so in the performative, language suspends its denotation precisely
and solely to found its existential connection with things.
Considered in this perspective, the ontological (or onto-theological) argu-
ment simply says that if speech exists, then God exists, and God is the expression
of this metaphysical “performance.” In it, sense and denotation, essence and ex-
istence coincide, the existence of God and his essence are one sole and identical
thing. That which results performatively from the pure existence [ darsi] of lan-
guage exists purely and simply ( on haplōs). (Paraphrasing a thesis of Wittgen-
stein, one could say that the existence of language is the performative expression
of the existence of the world.) Ontotheology is, therefore, a performance of
language and is in solidarity with a certain experience of language (that which
is at issue in the oath), in the sense that its validity and its decline coincide
with the strength and decline of this experience. In this sense metaphysics, the
science of pure being, is itself historical and coincides with the experience of
the event of language to which man devotes himself in the oath. If the oath is
declining, if the name of God is withdrawing from language—and this is what
has happened beginning from the event that has been called the “death of God”
or, as one should put it more exactly, “of the name of God”—then metaphysics
also reaches completion.
There remains, in any case, the possibility of perjury and blasphemy, in
which what is said is not really intended and the name of God is taken in vain.
The co-originarity of the performative structure and denotative structure of
speech ensures that the “Indo-European scourge” is inscribed in the very act
of speaking, which is to say, is consubstantial with the very condition of the
speaking being. With the logos are given both—co-originarily, but in such a way
that they cannot perfectly coincide—names and discourse, truth and lie, oath
and perjury, bene-diction and male-diction, existence and nonexistence of the
world, being and nothingness.
א This performative power of the name of God explains the fact, which is at first
glance surprising, that the polemic of the Christian apologists against the pagan gods did not concern their existence or nonexistence but only their being, in the words that Dante
puts in the mouth of Virgil, “false and lying” ( Inferno 1.72). The pagan gods exist but are not true gods; they are demons (according to Tatian) or human beings (for Tertullian).
In correspondence with a potentially infinite multiplication of their names, the pagan
gods are equivalent to false oaths, are constitutively perjurers. On the contrary, the in-
344
HOMO SACER II, 3
vocation of the true God’s name is the very guarantee of every worldly truth ( Augustine:
Te invoco, deus veritas, in quo et a quo et per quem vera sunt quae vera sunt omnia [Thee do I invoke, God, Truth, in whom and by whom and through whom are all things true
which are true] [Augustine [3], 1.3]). Once the performative power of language was con-
centrated in the name of the one God (which had become, for this reason, more or less
unpronounceable), the individual divine names lose all efficacy and fall to the level of
linguistic ruins, in which only the denotative meaning remains perceptible (in this sense, Tertullian can mention sarcastically Sterculus cum indigitamentis suis [Some Sterculus, I suppose]— Apol. 25.10).
24. In this perspective the sharp distinction between assertorial oaths and
promissory oaths corresponds to the loss of the experience of speech that is in
question in the oath. This is neither an assertion nor a promise but something
that, taking up a Foucauldian term, we can call a “veridiction,” which has as the
sole criterion of its performative efficacy its relationship to the subject who pro-
nounces it. Assertion and veridiction define, that is to say, the two co-originary
aspects of the logos. While assertion has an essentially denotative value, meaning that its truth, in the moment of its formulation, is independent of the subject
and is measured with logical and objective parameters (conditions of truth, non-
contradiction, adequation between words and things), in veridiction the subject
constitutes itself and puts itself in play as such by linking itself performatively to
the truth of its own affirmation. For this reason the truth and consistency of the
oath coincide with its performance, and for this reason the calling of the god as
witness does not imply a factual testimony but is actualized performatively by
the very utterance of the name. What we today call a performative in the strict
sense (the speech acts “I swear,” “I promise,” “I declare,” etc., which must, sig-
nificantly, always be pronounced in the first person) are the relics in language of
this constitutive experience of speech—veridiction—that exhausts itself with
its utterance, since the speaking subject neither preexists it nor is subsequently
linked to it but coincides integrally w
ith the act of speech.
Here the oath shows its performative proximity with the profession of faith
( homologia, which in Greek also designates the oath). When Paul, in Romans
10:6–10, defines the “word of faith” ( to rema tēs pisteōs) not by means of the correspondence between word and reality but by means of the closeness of “lips”
and “heart,” it is the performative experience of veridiction that he has in mind:
“‘The word is near you, on your lips and in your heart’ (that is, the word of faith
that we proclaim); because if you confess [ homologēsēis] with your lips that Jesus is Lord and believe [ pisteusēis] in your heart that God raised him from the dead,
you will be saved.”
THE SACRAMENT OF LANGUAGE
345
If one pretends to formulate a veridiction as an assertion, an oath as a deno-
tative expression, and (as the Church began to do from the fourth century on
by means of conciliar creeds) a profession of faith as a dogma, then the experi-
ence of speech splits, and perjury and lie irreducibly spring up. And it is in the
attempt to check this split in the experience of language that law and religion
are born, both of which seek to tie speech to things and to bind, by means of
curses and anathemas, speaking subjects to the veritative power of their speech,
to their “oath” and to their declaration of faith. The ancient formula of the
Twelve Tables, which expresses the performative potentiality of speech in law,
uti lingua nuncupassit, ita ius esto (as the tongue has said—has taken the name,
nomen capere—so the law is), does not mean that what is said is constatively
true but only that the dictum is itself the factum and that, as such, it obliges the person who has pronounced it. It is necessary once again to reverse, in this
sense, the common opinion that explains the efficacy of the oath by reference
to the powers [ potenze] of religion or of archaic sacred law. Religion and law do
not preexist the performative experience of language that is in question in the
oath, but rather they were invented to guarantee the truth and trustworthiness
of the logos through a series of apparatuses, among which the technicalization
of the oath into a specific “sacrament”—the “sacrament of power”—occupies a
central place.
א A loss of the understanding of the performative character of the experience of
language in question in the oath is evident in the philosophical analyses of perjury of
which we already have testimony in Aristotle. Discussing the oath of the Trojans in the
Iliad (3.276ff.), Aristotle observes that it is necessary to distinguish between breaking the oath ( blapsai ton horkon), which can only apply to a promissory oath, and epiorkesai, to perjure, which can refer only to an assertorial oath (Aristotle [1], frag. 143). In the same way Chrysippus distinguishes between alethorkein/pseudorkein, to swear the true /
to swear the false, which are in question in the assertorial oath, according to whether the affirmation on which one swears is objectively true or false, and euorkein/epiorkein, which are applied to the fulfillment or nonfulfillment of a promissory oath (Diogenes Laertius,
7.65–66; cf. Hirzel, 77–78; Plescia, 84–85). Here one sees how the model of logical truth
founded on the objective adequation between words and things cannot give an account
of the experience of language implicit in the oath. Insofar as the oath performatively
actualizes what is said, the epiorkos is not simply a false oath but implies the evacuation of the performative experience that is proper to the horkos.
Logic, which watches over the correct use of language as assertion, is born when
the truth of the oath has already waned. And if logic and science are born from the
management of the assertorial aspect of the logos, from veridiction there proceed, even if through crossings and superimpositions of every kind (which have their highest place
346
HOMO SACER II, 3
precisely in the oath), law, religion, poetry, and literature. In the middle is philosophy, which, abiding in both truth and error, seeks to safeguard the performative experience
of speech without renouncing the possibility of lying and, in every assertorial discourse, experiences the veridiction that takes place in it.
25. The performative efficacy of the oath is evident in the ancient trial, which,
in both Greece and Rome, had the form of a conflict between two oaths. The
civil trial opened with the oath of the two parties to the suit: the oath with which
the plaintiff affirmed the truth of his claims was called a proomosia (etymo-
logically: the oath pronounced first), that of the defendant an antomosia (that
is, an oath pronounced in opposition to the first), and the exchange of oaths
was called amphiorkia. Analogously, in penal law “the accuser swears that his
adversary has committed the crime and the accused that he has not committed
it” (Lysias, 46; see also Glotz, 762). The law of Gortyna shows that the Greeks
sought to limit the oath to cases in which testimonial evidence was impossible
and consequently to establish which of the two parties (usually the accused) had
a preferential right to the oath. In every case the judge decided who had “sworn
rightly” ( poteros euorkei) (Plescia, 49). Glotz has rightly observed, against the opinion of Rohde, that the declaratory oath that is in question in the Greek
trial, “far from constraining to perjury and from proving, as one says, that the
Athenians were not a Rechtsvolk, and far from being a purely religious institution that was destined eventually to substitute the justice of the gods for the fallible
justice of men” (Glotz, 761), was a properly juridical procedure in which the
declarative oath of the suit was clearly distinct from that requested as evidence.
In the Roman trial the procedure called legis actio sacramenti [the bringing for-
ward of the legal oath], which Gaius describes to us in book 4 of his Institutiones, was similar. Each of the two parties affirmed his right—in the case Gaius uses as
an example, the vidicatio of the ownership of a slave—with the formula: Hunc
ego hominem ex iure Quiritium meum esse aio, secundum suam causam sicut dixi
ecce tibi vindictam imposui [I assert that this slave is mine by Quiritary title, in accordance with his status, as I have declared it. Look you, I lay my wand
upon him], accompanied by the laying of a wand ( vindicta) on the head of the
contested slave ( Inst. 4.16). Subsequently, the one who had pronounced the first
declaration provokes the other to the sacramentum of a certain sum of money
( quando tu iniura vindicavisti, D aeris sacramento te provoco [Inasmuch as you
have made a claim without right to support it, I challenge you in a deposit of five
hundred asses]). Festus, commenting on the word sacramentum in the passage cited above, explains that it is a matter of a genuine oath that entails a sacratio: sacramento dicitur quod iusiurandi sacratione interposita factum est (one calls a
THE SACRAMENT OF LANGUAGE
347
sacrament what is done with the sanction of the oath). Only at this point did the judge pronounce his decision: “This sacramentum is the central point, the crux
of the trial, which gives its name to the whole. The role of the judge is in fact
limited, after the examination of the case, to declaring which is the sacramentum
justum and which is the sacramentum injustum” (Noailles [1], 276).
Once more, the historians of law, though re
alizing that what is here in ques-
tion is a genuinely performative efficacy, tend to explain the function of the
oath in the trial by recourse to the sacral paradigm: “It seems that the most
ancient forms of obligation make the oath effective. They provoke a change of
estate among the parties, and they create something between them in the world
beyond. In order to create, they bring forces into play. . . . These are so-called
religious forces” (Gernet [1], 61/172–73). What is thus presupposed in the form of
religiosity is just the experience of language that takes place in veridiction. In this
sense the opposition between faith and religion, so important in modern culture,
in reality corresponds point by point to the opposition between two co-originary
characteristics of the logos, which are veridiction (from which law and positive
religion proceed) and assertion (from which logic and science derive).
26. Let us attempt to understand, in the perspective of our investigation, the
“forces” that are really in question here. One of the terms about whose meaning
historians never stop debating is vindicta (and related terms like vindex, vin-dicere), which in the trial seems to designate the wand with which the parties
touched the disputed property. It is the merit of Pierre Noailles to have clarified
the original meaning of this term. It comes, according to the traditional etymol-
ogy, from vim dicere, literally: “to say or show force.” But what force? Among
scholars, Noailles observes, there reigns the greatest confusion on this point:
“they oscillate perpetually between the two possible meanings of the word: either
force or violence, which is to say force materially put to use. In reality they do
not choose. According to the occasion, it is one or the other meaning that is put
forward. The vindicationes of the sacramentum are presented now as manifestations of force, now as symbolic or simulated acts of violence. The confusion is
still greater with regard to the vindex. For it is not clearly determined whether the force or violence that the name indicates is his own, which he puts at the
service of the law, or if it is the violence of his adversary, whom he denounces as
contrary to justice” (Noailles [2], 57). Against this confusion Noailles shows that
The Omnibus Homo Sacer Page 52