The Omnibus Homo Sacer

by Giorgio Agamben

timiōtaton estin]” (ibid., 938b34–35). According to this testimony, the oath is the most ancient thing, no less ancient than the gods, who are in fact subject to it

  in some way. Yet this does not mean that it must be thought of as a “sacred sub-

  stance”; on the contrary, the context of the passage, which is that of the recon-

  struction of the thought of Thales within the brief history of philosophy that

  opens the Metaphysics, leads one to situate the oath among the “first principles”

  THE SACRAMENT OF LANGUAGE

  315

  ( prōtai aitiai) of the pre-Socratic philosophers, as if the origin of the cosmos and of the thought that understands it implied the oath in some way.

  The entire problem of the distinction between the juridical and the religious,

  in particular as regards the oath, is thus poorly put. Not only do we have no rea-

  son for postulating a prejuridical phase in which the oath belonged solely to the

  religious sphere, but perhaps our entire habitual way of representing to ourselves

  the chronological and conceptual relationship between law and religion must

  be revised. Perhaps the oath presents to us a phenomenon that is not, in itself,

  either (solely) juridical or (solely) religious but that, precisely for this reason, can

  permit us to rethink from the beginning what law is and what religion is.

  א When law and religion are placed in opposition to each other, it is necessary to

  remember that the Romans considered the sphere of the sacred as an integral part of law.

  The Digest opens with the distinction between ius publicum [public law], which concerns the status rei publicae [status of public things], and ius privatum [private law], which concerns the singulorum utilitatem [utility of individuals]; but, immediately after, the ius publicum is defined as that law “which consists in sacred things and rites, in priests and in magistrates” ( ius publicum quod in sacris, in sacerdotibus, in magistratibus consistit [Ulpian, 1.1]). In the same sense Gaius ( Inst. 2.2) distinguishes things according to whether they belong to the ius divinum [divine law] or to the ius humanum [human law], specifying that divini iuris sunt veluti res sacrae et religiosae [of the class divini juris are things sacred or religious]; but this summa divisio [chief division] of things is obviously internal to law.

  10. Two texts will allow us to take up the analysis of the oath on new foun-

  dations. The first is a passage from Philo’s Legum allegoriae (204–8), which, dis-

  cussing the oath that God makes to Abraham in Genesis 22:16–17, puts the oath

  in a constitutive relationship with the language of God:

  You mark that God swears not by some other thing, for nothing is higher than

  He, but by Himself, who is best of all things. Some have said, that it was inap-

  propriate for Him to swear; for an oath is added to assist faith [ pisteōs eneka] and only God . . . is faithful [ pistos]. . . . Moreover, the very words of God are oaths

  [ hoi logoi tou theou eisin horkoi] and laws of God and most sacred ordinances;

  and a proof of His sure strength is that whatever He says comes to pass [ an eipēi

  ginetai], and this is specially characteristic of an oath. It would seem to be a

  corollary from this that God’s words are oaths receiving confirmation by accom-

  plishment in act [ ergōn apotelesmasi]. They say indeed that an oath is a calling

  God to witness [ martyria] to a point which is disputed; so if it is God that swears, He bears witness to Himself, which is absurd, for he that bears the witness must

  be a different person from him on whose behalf it is borne. . . . If once we take

  “by Myself have I sworn” in the right way, we shall quit this excessive quibbling.

  Probably then the truth of the matter is something like this. Nothing that can

  316

  HOMO SACER II, 3

  give assurance [ pistoun dynatai] can give positive assurance touching God, for to

  none has He shown His nature, but He has rendered it invisible to our whole

  race. . . . Nay he alone shall affirm anything regarding Himself since He alone

  has unerringly exact knowledge of His own nature. God alone therefore is the

  strongest security first for Himself, and in the next place for his deeds also, so

  that He naturally swore by Himself when giving assurance as to himself [ ōmnye

  cath’heautou pistoumenos heauton], a thing impossible for anyone but God. It

  follows that men who say that they swear by God should be considered actu-

  ally impious; for naturally no one swears by Him, since he is unable to possess

  knowledge regarding His nature. No, we may be content if we are able to swear

  by his name, which means (as we have seen) the interpreting word [ tou ermeneōs

  logou]. For this must be God for us the imperfect folk, but, as for the wise and

  the perfect, the primal being is their God. Moses, too, let us observe, filled with

  wonder at the transcendency of the Uncreated, says, “And thou shalt swear by

  His Name” (Deut. 6:13), not “by Him,” for it is enough for the created being

  that he should be accredited and have witness borne to him by the Divine word:

  but let God be His own surest guarantee [ pistis] and evidence. (Trans. altered)

  Let us try to summarize in five theses the implications of this brief treatise on

  the oath:

  (1) The oath is defined by the verification of words in facts ( an eipēi ginetai,

  precise correspondence between words and reality).

  (2) The words of God are oaths.

  (3) The oath is the logos of God, and only God swears truly.

  (4) Men do not swear by God but by his name.

  (5) Since we know nothing of God, the only certain definition that we can

  give of him is that he is the being whose logoi are horkoi, whose word testifies with absolute certainty for itself.

  The oath, defined by the correspondence between words and actions, here per-

  forms an absolutely central function. This happens not only on the theological

  level, in that it defines God and his logos, but also on the anthropological level, since it relates human language to the paradigm of divine language. If the oath

  is, in fact, that language that is always realized in facts and this is the logos of God (in De sacrificiis [65] Philo writes that “God spoke and it was done, with no interval between the two [ ho theos legōn ama epoiei]”), the oath of men is thus the

  attempt to conform human language to this divine model, making it, as much

  as possible, pistos, credible.

  In De sacrificiis (93) Philo confirms this function of the oath. “Now men,”

  he writes, “have recourse to oaths to win belief, when others deem them untrust-

  THE SACRAMENT OF LANGUAGE

  317

  worthy; but God is trustworthy [ pistos] in his speech as elsewhere, so that his

  words in certitude and assurance are no different from oaths. And so it is that

  while with us the oath gives warrant for our sincerity, it is itself guaranteed by

  God. For God is not trustworthy because of [ dia] the oath; but it is God that

  assures the oath” (trans. altered).

  One should reflect on the reciprocal implication between God and the oath

  contained in the last phrase, which closely follows a rhetorical model frequent not

  only in Judaism, which works by inverting a sanctioned truth (of which Mark

  2:27—“The sabbath was made for [ dia] humankind, and not humankind for

  the sabbath”—is a good example). Just as in the classical tradition the horkos is pistos par excellence, so also in the Judaic tradition pistos ( eman) is the a
ttribute of God par excellence. Developing this analogy (perhaps following in the path of

  the verse from Aeschylus—fragment 369—in which one reads that “it is not the

  oath that makes us believe the man, but the man the oath”), Philo establishes an

  essential connection between God and oath, making of the latter the very word

  of God. In this way, however, not only human language but even God himself

  is irresistibly drawn into the sphere of the oath. On the one hand, in the oath

  human language communicates with that of God; on the other hand, if God

  is the being whose words are oaths, it is completely impossible to decide if he is

  reliable because of the oath or if the oath is reliable because of God.

  11. The second text is the celebrated passage from the De officiis (3.102–7),

  from which I have already cited some lines, which we must now restore to their

  context. What is in question is the behavior of Attilio Regolo, who, sent to

  Rome by the enemies of whom he had been a prisoner with the oath that he

  would return, decides to return knowing that he will be put to death. The ques-

  tion that Cicero asks concerns the origin of the binding power of the oath.

  “‘What significance, then,’ someone will say, ‘do we attach to an oath? It is not

  that we fear the wrath of Jove, is it?’” (3.102). And yet, he responds, all the phi-

  losophers affirm that the gods do not become angry at or harm men. It is at this

  point that he formulates the celebrated definition of the oath that I have cited:

  “But in taking an oath it is our duty to consider not what one may have to fear

  in case of violation but wherein its obligation lies [ non qui metus sed quae vis sit

  debet intellegi]; an oath is an assurance backed by religious sanctity [ affirmatio religiosa]; and a solemn promise given, as before God as one’s witness, is to be

  sacredly kept” (3.104).

  What is decisive here is the reasoning with which Cicero goes on to establish

  the vis of the oath. It is not a matter of the anger of the gods, which does not

  318

  HOMO SACER II, 3

  exist ( quae nulla est), but of trust ( fides). Contrary to the opinion very often repeated by modern scholars, the obligatory nature of the oath does not derive

  from the gods, who are called only as witnesses, but from the fact that it is sit-

  uated in the sphere of a more far-reaching institution, the fides, which regulates relations among men as much as those between peoples and cities. “Whoever,

  therefore, violates his oath violates trust [ Quis ius igitur iurandum violat, is fidem violat]” (3.104). In the passage previously cited from the first book of the work, the fides, “foundation of law,” was defined etymologically, exactly as in Philo, by the verification of what is said: quia fiat quod dictum est appellatam fidem [“good faith” is so called because what is promised is “made good”] (1.23). Faithfulness

  is thus essentially the correspondence between language and actions. Regolo,

  Cicero can therefore conclude, has done well in observing his oath: if it is lawful

  not to observe an oath with pirates, with whom, as hostes omnium [enemies of

  all], it is not possible to have a common trust, it would be unjust “to confound

  by perjury the terms and covenants of war made with an enemy [ condiciones

  pactionesque bellicas et hostiles perturbare periurio]” (3.108).

  א It is advisable to specify the meaning of the term religiosus in Cicero’s definition of the oath. A res religiosa is, in Rome, something that has been devoted to the infernal gods ( religiosae quae diis manibus relictae sunt [things religious are those which are given up to the Gods below], Gaius Inst. 2.4); in this sense the religiosus par excellence is the grave, the place in which a corpse ( corpus, which the Romans distinguished from the cadaver, which designates a dead body deprived of a grave) has been buried. The res religiosa is removed from profane use and commerce and can be neither transferred nor burdened

  with servitude nor given in usufruct or pledge nor made the object of any stipulation

  whatsoever (Thomas, 74). More generally, the religious thing, like the sacred thing, is

  subject to a series of ritual prescriptions, which render it inviolable and which must be

  scrupulously observed. One can understand, then, in what sense Cicero can speak of

  the oath as an affirmatio religiosa. The “religious affirmation” is a word guaranteed and sustained by a religio, which removes it from common use and, consecrating it to the gods, makes it the object of a series of ritual prescriptions (the formula and gesture of

  the oath, the calling of the gods as witness, the curse in case of perjury, etc.). The double sense of the term religio, which according to the lexicons means both “sacrilege, curse” and

  “scrupulous observation of formulas and ritual norms,” can be explained in this context

  without difficulty. In a passage of the De natura deorum (2.11) the two senses are at the same time distinct and juxtaposed: the consul Tiberius Gracchus, who had forgotten to

  take the auspices at the moment of the designation of his successors, prefers to admit his error and annul the election that has taken place contrary to religio rather than allow a

  “sacrilege” ( religio) to contaminate the State: peccatum suum, quod celari posset, confiteri maluit, quam haerere in re publica religionem, consules summum imperium statim deponere, quam id tenere punctum temporis contra religionem [he preferred to make public confession

  THE SACRAMENT OF LANGUAGE

  319

  of an offence that he might have concealed rather than that the stain of impiety should

  cling to the commonwealth; the consuls preferred to retire on the spot from the highest

  office of the state rather than hold it for one moment of time in violation of religion].

  It is in this sense that, when putting together the two meanings of the term, Cicero,

  just like Caesar and Livy, can speak of a “religion of the oath” ( religio iusiurandi). In a similar way Pliny, referring to the rules against looking at certain parts of the body, can speak of a religio inherent to the knees, the left hand, and even urine ( Hominum genibus quaedam et religio inest observatione gentium . . . inest et aliis partibus quaedam religio, sicut in dextera: osculis adversa adpetitur, in fide porrigitur [The knees of a human being also possess a sort of religious sanctity in the usage of the nations. . . . There is a religious sanctity belonging to other parts also, for instance in the right hand: kisses are imprinted in the back of it, and it is stretched out in giving a pledge]; Natural History 11.250–51).

  And when, in a text of a magical character, we read the formula against sore throat— hanc religionem evoco, educo, excanto de istis membris, medullis [I evoke, lead out, and bring forth by incantation that religio from this limb, down to the marrow] (Mauss, 54/76)— religio represents both a “curse” and the collection of ritual formulas to be observed in order to produce (and remove) the incantation.

  When, anachronistically projecting a modern concept onto the past, one often speaks

  of a “Roman religion,” it must not be forgotten that, according to the clear definition that Cicero puts in the mouth of the pontifex maximus Cotta, this was nothing but the sum

  of the ritual formulas and practices to be observed in the ius divinum: cum omnis populi Romani religio in sacra (consecrations) et in auspicia (the auspices to be consulted before every important public act) divisa sit (“The religion of the Roman people comprises ritual

  [and] auspices” [ De natura deorum 3.5]). For this reason he could point to its etymology (which, moreover, is shared by modern scholars) in the verb relegere, to observe scrupulously: qui autem omnia quae ad cultum deorum pertinent diligenter retractarent et tamqu
am relegerent, sunt dicti religiosi ex relegendo (Those on the other hand who carefully reviewed and so to speak retraced all the lore of ritual were called ‘religious,’ from relegere [2.72]).

  12. The proximity between faith and oath has not escaped scholars and is at-

  tested by the fact that, in Greek, pistis is synonymous with horkos in expressions of the type pistin kai horka poieisthai (to take an oath) or pista dounai kai lambanein (to exchange an oath). In Homer oaths are what are pista (trustworthy) par excellence. And in the Latin sphere, Ennius, in a verse cited by Cicero, defines

  fides as “an oath of Jove” ( ius iurandum Iovis). And it is significant that there are attested not only formulas of an oath “by the pistis of the gods” but also “by one’s own pistis”— kata tēs heautōn pisteōs diomosamenoi (Dionysius of Halicarnassus, 11.54)—and that, in fact, the “pistis of each person” ( idia ekastōi pistis) counts as the megistos horkos (greatest oath; ibid., 2.75; see also Hirzel, 136).

  Dumézil and Benveniste have reconstructed, beginning from linguistic

  data above all, the originary features of that most ancient of Indo-European

  320

  HOMO SACER II, 3

  institutions that the Greeks called pistis and the Romans fides (in Sanskrit, it is sraddha): “personal loyalty.” “Trust” [ fede] is the high esteem in which someone is held as a consequence of our having confidently given ourselves over to him,

  binding ourselves in a relationship of loyalty. For this reason trust is both the

  confidence that we accord to someone—the trust that we give—and the high

  esteem in which we are held by someone—the trust or credit we have. The

  old problem of the two symmetrical meanings of the term faith, active and

  passive, objective and subjective, “guarantee pledged” and “trust inspired,” to

  which Eduard Fränkel drew attention in a famous article, is understood in this

  perspective without difficulty: “the one who holds the fides placed in him by a

  man has this man at his mercy. This is why fides becomes almost synonymous

  with dicio and potestas. In their primitive form, these relations involved a certain reciprocity, placing one’s fides in somebody secured in return his guarantee and his support. But this very fact underlines the inequality of the conditions.