We do not die each for ourselves, but for one another, or even, who knows, one in place of another.—GEORGES BERNANOS, Dialogues des Carmélites, III, 1
The powerful originality of the thought of Emmanuel Levinas makes itself felt on each page of his oeuvre, all the more as time passes. But perhaps it never asserts itself more powerfully than with the relatively late doctrine (introduced between 1968 and 1974) of substitution. Paul Ricoeur, condemning what he called a “strategy” of “accumulation of excessive, hyperbolic expressions, destined to baffle common thinking,” saw substitution as “crowning this sequence of excessive expressions.”1
A surprising doctrine indeed, because it involves a redoubling of responsibility—put another way, “one more responsibility”2—such that I substitute myself for the other precisely in what he has that is most properly his own, his own responsibility: “the overemphasis of openness is responsibility for the other to the point of substitution.”3 What is more, at issue is not a mere hyperbole of responsibility, where I would take upon myself the burden that properly belongs to the other out of sympathy, scruple, or solidarity. Instead, at issue is first of all “a responsibility with regard to men we do not even know,”4 or even, “responsibility for the persecutor” himself.5 A strange and shocking assertion! Strange, for, if there were a man we knew well and whose face we should never forget (supposing that he has one),6 wouldn’t he be, precisely, our persecutor? Shocking, too, for how could we take upon ourselves the responsibility for a fault committed against us? We already experience great difficulty in not clamoring for revenge; how would we bear the difficulty of taking upon ourselves—of taking upon myself, the victim—the fault of the persecutor? And why must we? For in the end, wouldn’t such a demand—“In the trauma of persecution [. . .] to pass from the outrage undergone to the responsibility for the persecutor”7—betray a perversion of ethics, now become a mechanism for condemning the victims in the place and stead of the executioners? In the end, how could this demand not contradict the famous dedication of Otherwise than Being: “To the memory of [. . .] the victims of the same hatred of the other man, the same anti-semitism”?8
I will attempt to show that, in the final instance, there is no contradiction whatsoever to be found here, nor an excess of hyperbole, but instead a purely conceptual necessity, paradoxical as it may be. Indeed, in order to avoid degrading substitution to a mere rhetorical effect (whether one approves or deplores it), it is necessary to see that its brutality results from the encounter of two lexicons and of two modes of questioning, which play like two telluric plates at once opposing and supporting each other—that is, one lexicon that belongs to phenomenology and one that depends upon what one could call, for lack of a better name, the revelation of the infinite. And indeed, how could one say, as Levinas did in the conclusion of his 1974 book, that Otherwise than Being “ventures beyond phenomenology” (trans. Lingis, 183), if not because substitution fixes precisely the second focus of the ellipse, with the first focus remaining the phenomenological I (transcendental, or perhaps also daseinsmässig)? Two declarations offer unambiguous confirmation. “The overemphasis of openness is responsibility for the other to the point of substitution, where the for-the-other proper to disclosure, to monstration to the other, turns into the for-the-other proper to responsibility. This is the thesis of the present work.” And: “This book interprets the subject as a hostage and the subjectivity of the subject as a substitution breaking with being’s essence.”9 In short, what is first of all and radically at stake in substitution is subjectivity.
Thus it could not be said more clearly: substitution (which constitutes me as a “hostage,” a term still to be defined) does not first fall under an ethical horizon, because it has as its more radical task the contradiction of the primacy of the I, that is to say, the determination of subjectivity by an essence, and thus by being. At issue here is not yet, once again, ethics (which determines how subjectivity should act in order to render justice to the other), but a prior task, otherwise difficult—that of freeing subjectivity from any ontological determination (of identifying, for instance, which subjectivity acts when the question is the rendering of justice). Here we find exactly verified a paradox advanced several years later: “My task does not consist in constructing ethics; I only try to find its meaning. [. . .] One can without doubt construct an ethics in function of what I have just said, but this is not my own theme.”10 For, even and above all if one establishes ethics as first philosophy, one must recognize that this establishment itself could not fall under ethics, since ethics profits thereby. Or else, one must understand “a concept of the ethical [Begriff des Ethischen] that is separated from the tradition that derives the ethical [das Ethische] from knowledge and from Reason.”11 In which case, just as substitution exceeds ethics itself, the term subjectivity must no longer be understood here according to ontology, even when inverting it, and even less according to what metaphysics understands under the title of ethics.12
Subjectivity must itself become a question, so as to allow itself precisely to be rethought beginning from substitution, and so perhaps to conceive the ethical element non-metaphysically: “Subjectivity is being hostage. This notion reverses the position where the presence of the ego to itself appears as the beginning or as the conclusion of philosophy.”13 As soon as one ceases to hear it in its ethical sense, substitution loses its strange hyperbolic quality, because it assumes from that point forward an eminent and nonethical function—that of putting into question the “essence of being,” as “philosophy” (metaphysics and even phenomenology) presupposes it for defining man’s “subjectivity.” Substitution must thus be understood here in the extra-moral sense of a radical reversal of ontology by the bias of a destruction of all determination according to the being of what is at issue in subjectivity. Thus understood, “ethics” (the ethical element) opens a way toward an ipseity without being.
But where does Levinas’s notion of substitution come from? The point here is not to reconstitute its lexical provenance, which in itself is of little importance (supposing, of course, that it could be traced); rather, my goal is to zero in on the operation, the impact, and the stakes that could qualify substitution, to the point where it would assume such a polemical role in so vast a “destruction.”
The hypothesis I wish to suggest is the following: the question of substitution is posed to Levinas by Heidegger himself, in section 26 of Sein und Zeit. Indeed, Heidegger would not have held and retained until the end, and despite all his faults, such prestige in Levinas’s eyes14 if the existential analytic had merely missed the question of the other. However, contrary to a widely shared opinion, Sein und Zeit does not pass over in silence the question of the other, even if the other does not occupy in the book the center of the question of being. The border between Heidegger and Levinas does not pass between, on the one hand, a Dasein without alterity and, on the other, an ego (moi) determined by the other. More subtly, it separates two opposed ways of describing the relation of the ego (moi) or of Dasein to the other; and the difficulty consists in locating exactly where the line of division passes. I will attempt to show that the entire opposition turns on the possibility, or not, of a substitution. And we are all the more authorized to follow this hypothesis by the fact that Levinas himself—rather late, it is true, but nevertheless with great precision—commented upon this very paragraph from Sein und Zeit.15
After having determined the In-der-Welt-sein (chap. 2), then the worldliness of the world (chap. 3), in order to attain the In-Sein as such (chap. 5) and care (Sorge, chap. 6), to which everything leads, the transcendental analytic must along the way specify the In-der-Welt-sein as being-one’s-self (Selbst-sein) and also as being-with (Mitsein) (chap. 4). In this context, the question of the other is posed, but so too is the question of the They (das Man), as if alterity necessarily were in league with the indifferentiation of the self. Let’s limit ourselves, for the moment, to underscoring that, for the analytic of Dasein, the existential of being-with (Mitsein) strictly implies the “other,” that is to say the other as a(nother) Dasein: “On the basis of this like-with [mithaften] being-in-the-world, the world is always already the one that I share with the others. The world of Da-sein is a with-world [Mitwelt]. Being-in [In-Sein] is being-with [Mitsein] others. The innerworldly being-in-itself of others is Mitda-sein [or co-Dasein].”16 Without any ambiguity, Dasein, precisely because it opens a world of itself (as In-der-Welt-sein), opens it as a world that from the outset is offered to alterity, as a co-world. Thus being in this co-world immediately implies being-with others, others who are themselves in the mode of Dasein. In short, Dasein implies, in its very being, co-Dasein: “Only because it has the essential structure of being-with, is one’s own Da-sein Mitda-sein [or co-Dasein] as encounterable by others.”17 Thus, that Dasein implies an other Dasein (as another being) does not yet pose any difficulties. The difficulties begin only later, when the point is to determine how a Dasein encounters the other: does it go right away, simply as co-Dasein, to encounter the other?
If the other can be encountered only as a(nother) Dasein, it is properly necessary that it be encountered as a Dasein, and not as a being of the world that would not measure up to a Dasein (a nichtdaseinsmässig being). Thus it will be necessary for Dasein to relate to another otherwise than it would relate to an innerworldly being, which is to say, otherwise than through Besorgen (taking care) (of a task, or of a need), which only applies to an innerworldy being. Nevertheless, this mode of encounter cannot be called Sorge (care, cura) either, for Sorge will only intervene in view of the being-toward-death of Dasein proper. Thus Dasein can only relate to the other (as) Dasein in a particular mode, that of Fürsorge (solicitude, concern, care for . . .). Let us now describe the mode of access to the phenomenon of the other (as) Dasein. First, a remark: the German Fürsorge indicates, taken in its first instance for a “faktische soziale Einrichtung,” what French calls “assistance publique,” and English “Medicare” (America) or “social care” (Britain).18 A first meaning of this term inevitably follows: social assistance, Fürsorge, goes first and foremost, and most frequently, to occupy itself (in the mode of Besorgen) with innerworldly beings, of which an other (as) Dasein would happen to find himself deprived, and thus endeavors to provide such things for him as food, clothing, shelter, etc. Now, Heidegger condemns this Fürsorge as “deficient.” One might be astonished, even scandalized, by this because the biblical care toward “the widow and the orphan” (for example, Exod. 22:21, Deut. 10:18, or Isa. 1:18) thus finds itself downgraded to the level of a deficient mode of Fürsorge. One’s astonishment would be well justified, for even though Fürsorge cares for the other only indirectly and attends first of all directly to innerworldly beings, of which the other has urgent need (besorgen), this first-response emergency assistance nonetheless remains the existential (and not only the existentiell) condition of possibility for other modes of Fürsorge.19 This justice thus rendered to “sublime materialism,”20 one must also recognize Heidegger’s reasons for condemning its “deficiency”: precisely because it is concentrated upon the first-response emergency goods, and thus those of first usage, this assistance in fact treats only innerworldly beings-at-hand (Zuhandenes) and thus not the other (as) Dasein. The other (as) Dasein thereby remains undifferentiated, non-individuated, even anonymous, such that the same assistance could be applied to anyone without distinction. Paradoxically—and thus the pertinence of Heidegger’s analysis—the assistance (the Fürsorge as a Besorgen that is unaware of itself) does not concern itself with the alterity of the other (as) Dasein, because it does not even consider Dasein’s individualized otherness, identified as such and for itself. Assistance socializes alterity, which it thereby renders indeterminate and indifferent. The other becomes whomever and anyone; he thus takes on the faceless visage of They (das Man), and in the end remains in a deficit of alterity.21 Heidegger’s argument thus situates itself, in effect, upon the conceptual field that will become Levinas’s: the recognition of the other as the Other, without match or equal.
However it may be with this first “deficient” mode of Fürsorge, it allows by contrast a derivation of its positive mode or, more precisely, a glimpse of the “two extreme possibilities” of this positive mode.22 The first possibility comes down to “tak[ing] the other’s ‘care’ [Sorge] away from him and put[ting] itself in his place in taking care [Besorgen], [. . .] leap[ing] in for him [für ihn einspringen],”23 thus unburdening him of his care (Sorge) by taking his place in the management of his material needs (ses besoins en étants) (im Besorgen). This first mode is inadequate for the (one might say Hegelian) reason that Heidegger formulates unequivocally: such a Dasein, unburdened of its care (Sorge) by another Dasein, would immediately fall under its domination (Herrschaft), even if silent or unconscious;24 once again, the other would no longer be himself, but instead a servant of Dasein insofar as mine, and thus not yet properly another (as) Dasein. Thus, and specifically in order to respect the alterity of the other Dasein (and thus, doubtless, of the other himself), it is necessary to pass to the second “extreme possibility” of Fürsorge, “which does not so much leap in for [nicht einspringt] the other as leap ahead of him [vorausspringt], not in order to take ‘care’ [Sorge] away from him, but to [sic] first to give it back to him [zurückzugeben] as such.”25 For the first time, the assistance (Fürsorge) of Dasein as mine no longer concerns innerworldly beings, to which the other could attend (Besorgen), but the other’s existence (or potentiality-of-being) even as other, (as) Dasein. Solicitude (Fürsorge) finally becomes, literally, care for (Sorge für) the other as such, namely, as another Dasein. And this care for the other (as) Dasein consists precisely in not claiming to take his place, but instead in allowing him to take upon himself the weight of his own possibility, “the burden of being.”26 Care for the other amounts to not substituting oneself for him, and instead allowing him to carry his load, that of being, about which he cannot not decide, because it is what is most his own.27 Care of the other requires abandoning him to himself.
Heidegger thus maintains, within the frame of the analysis of Mitsein (§26), this strange paradox: that the solicitude that takes true care (Fürsorge) for the other consists precisely in not substituting for him. But on what argument, and thus through what description, is this paradox established? Obviously here the matter is not one of selfishness or indifference (which in any case are simple moral determinations, and thus ontic, lacking any existential pertinence), but instead of the strict ontological demand for a higher care for the other, which makes manifest that he, too, is defined by his care of self (Sorge). In the end, the point is to recognize in the other the originary determination of Dasein: “Because its essence lies rather in the fact that it in each instance has to be its being as its own, the term Da-sein, as a pure expression of being, has been chosen to designate this being.”28 Thus, for Dasein, to accede to the self always means either enduring mineness or deviating from it: “Mineness [Jemeinigkeit] belongs to existing Da-sein as the condition of the possibility of authenticity [appropriation] and inauthenticity [in-appropriation].”29
Now, Dasein only accomplishes its ownmost—properly attaining to oneself and attaining what is proper to oneself by appropriating being, or rather, by allowing oneself to appropriate oneself through being—by exercising its final possibility, by being in the mode of being-toward-death.30 Indeed, my possibility implies not only the possibility of my death but my death as possibility (of impossibility). Such a death, as possibility, implies that no one can excuse himself from it, and thus that no one can take up the task for me, nor substitute himself for me. Doubtless, someone else can “die in my place,” through sacrifice or devotion to me (or another). However, even in such a case, the one who sacrifices himself first of all will die his own death and not mine; and, second, he will only spare me my death for a time, because in the end there will always come a moment in which I, in person, will have to live my death. No one will ever do that in my stead. “Death is a possibility of being that Da-sein always has to take upon itself,” for “No one can take the other’s dying away from him.”31 If I, Dasein, wish to attain my ownmost and my ipseity, I must never allow an other to substitute for me, especially at the instant of my death, “essentially and irreplaceably mine [unvertretbar].”32 The complete definition of death “as the ownmost nonrelational, certain [. . .] possibility,”33 is articulated according to this ownmost and, so to speak, follows from it, as first and only superlative. In death, as a possibility that I anticipate, what is at stake is my appropriation (Eigentlichkeit) to myself (Selbstheit) as Dasein. I am as I will die (sum moribundus),34 alone, because unsubstitutable.
This conclusion is confirmed a contrario by the fact that substitution—for Heidegger does admit the possibility—always relates back to the They (das Man), in its ceaselessly repeated attempt to disappopriate Dasein of itself, which is to say, to restore “a constant tranquilization about death.”35 Of course, the possibility of substitution belongs as a matter of principle to Dasein insofar as it is open to others;36 but the substitution that follows always debases this Dasein in a public event, accessible to everyone and no one—in short, accessible to the They: “Dying, which is essentially and irreplaceably [unvertretbar, unsubstitutably] mine, is perverted into a publicly occurring event which the they encounters.”37 Substitution, whether it is for me by another, or for another by me, in every case prevents (in me or in the other) Dasein’s appropriation of itself.
Thus it becomes clear that, for Heidegger, substitution contradicts ipseity in every case (the other for me, or me for the other). I am Dasein only in the first person, and never through another, nor for another. As a consequence, Fürsorge must also always let itself fall back completely into Sorge: “Even if only privatively, care is always taking care of things [Besorgen] and concern [Fürsorge, solicitude].”38 Even here, the für- remains a mere add-on prefix,39 changing nothing in the center of gravity of the care that is Sorge. As the meaning of the being of Dasein, Sorge stays centered on Dasein as mine.40
The conflict becomes completely clear. For Heidegger, ipseity excludes the substitution of the other for me, or me for the other, and is decided exclusively and solely by my anticipatory resoluteness regarding my death. For Levinas, “subjectivity is from the first substitution,”41 and I only attain to my ipseity by substituting for the other or by allowing the other to substitute for me—for here the one comes back precisely to the other, if at least one of us, I (moi), comes back to the other and not to the self. “The fact of not evading the burden imposed by the suffering of others defines ipseity itself.”42 The justification for this reversal remains to be made, and will be accomplished in several steps.
The first step leads to rejecting the claim that ipseity can be attained on the basis of the sole I, understood as Dasein—the sole I, which is also to say, the I as solitary. Now, subjectivity is not summed up by the supposedly transparent relationship of consciousness to itself, because “consciousness, knowing of oneself by oneself, is not all there is to the notion of subjectivity.”43 Of course, this does not mean that subjectivity retreats into the unconscious, but that the circle in which the cogitatio is closed never allows access to what, in the ego, specifies it most radically as in itself. And the same impossibility exists for the cogitatio’s other privileged modality, the will: “The ego [Moi] [. . .] is this original expiation. This expiation is involuntary, for it is prior to the will’s initiative (prior to the origin).”44 I once again underscore that what Levinas does not hesitate to call an “ ‘inversion’ of intentionality,” which would go “against intentionality,”45 implies nothing less than a wholesale “inversion of consciousness,” which becomes a “consciousness countercurrentwise.”46 Far from the ego attaining itself and experiencing itself by returning upon itself, it will reach itself only by going against the current that comes upon it from elsewhere. Thus, just as Levinas retains from Descartes the idea of the infinite, so does he reject the cogito (at least according to its standard formulation),47 for an important reason: while the ego of the cogito decides itself for itself through a thinking that is centrifugal, but which always returns upon itself, the ego of the idea of the infinite discovers itself preceded by what will always remain exterior to it, the infinite itself, and which nevertheless defines the ego always more originally than does the ego itself. If one can still speak here of subjectivity, it cannot any more name itself than it can conceive itself. As a subjectivity without an I, without a name established in the nominative, it no longer carries any name other than that attributed to it from outside, as a sobriquet: me [moi]. Me, or the name that comes from elsewhere, which names me from the very place where the other sees me, as I will never see myself. Me, or the name that names me as I never will.
Henceforward, I name myself “in the accusative form, which is a modification of no nominative form.”48 I have no proper name, because even my me is “not an ego [Moi], but me under assignation”49—but which assignation, and by what? Clearly, my assignation by that which accuses me, I who have a “self, from the first in the accusative form (or under accusation!).”50 It is necessary here to keep something of a phenomenological meaning associated with accusation, beyond its patent juridical meaning: I know myself as an accused me, as when a light that is too bright shines at me and singles out (accuse) my features, rendering them more visible to my spectator, without my being able either to see them or to control them. Under luminous accusation, I no longer phenomenalize myself from my point of view, but rather from that of the other. I become me in a light from elsewhere. And here, I am seen without seeing (myself); I do not appear first through myself and for myself (as if my phenomenality could produce itself beginning from myself), but instead, I first and in fact only appear summoned before the other; against both common sense and the juridical situation, I do not appear first, and in order, then, to appear summoned before the other; instead, I only appear if I appear summoned before or in front of . . . [the other].51
Now there comes a second step: identifying or at least designating that which accuses or points out, even if only in the phenomenological sense, the elsewhere from which the other, also come from elsewhere, summons me to appear (comparaître); in short, that in relation to which (or to whom) I discover myself in this situation where one names me in the accusative. One could limit oneself to a simple, juridical response: I am named in the accusative me because an other, eventual holder of the role of the other in general, accuses me with a certain charge (crimen). But this explanation doesn’t hold, first because it only works by arbitrarily presupposing my guilt, as if it were always already the case that I am guilty; it assumes that accusation is immediately equivalent to indictment, and thus that the accusative first derives from morality, from the determination between good and evil. Now, it could be that the Levinassian accusative (as much as the Heideggerian Gewissen) must, at least here, be understood once again in the extra-moral sense, which is to say in the strictly phenomenal sense. Next, and above all, the juridical explanation of the me (moi) seems to grant that I would be, first and to begin with, the me through or by myself, such that it could only receive the charge of accusation afterward; thus my existence as I would again precede my qualification as me, if only because it renders it possible; thus the accusative would leave the ego intact in its unassailed nominative, confirming it by indicting it. But here it is precisely the point to think a me that is anterior to the I, a me that dispenses absolutely with the I and has never assumed its form. In short, the point is to think a me outside of being: “Not strictly speaking an ego [moi] set up in the nominative in its identity, but first constrained to. . . . It is set up as it were in the accusative form.”52 The me, thus taken back (repris), is not declined as the object case of a nominative I. On the contrary, this I would be declined instead from a sole and originary me (accusative), in its essentially adversative character, considered from elsewhere; the I would only appear by neutralizing, after the fact, the anarchy prior to the me. Prior, indeed, because the accusation does not make a ground upon some thing (a being), which would precede it and, supporting itself on it, eventually resist the accusation, or exonerate or exculpate itself from the accusation. Now, it is not enough that one accuse me for me to discover myself originally as a me in the accusative; on the contrary, the posture of the accused, if limited to a solely moral horizon, presupposes again and always my ontic self-sufficiency, precedence, and independence. Again, it grants me the status of a being, in order to assure me evermore of being and to appropriate me thus to myself as subject, even if only to be able later to accuse me as a moral or juridical subject. Finally, accusation, even understood in the extra-moral sense, no longer allows for the anteriority of the other over me, nor therefore the me itself, because accusation can allow itself to be reduced to a call of my conscience, by itself, in the sense where, according to the existential analytic, “Dasein is at the same time the caller and the one summoned,” such that it “calls itself.”53 At fault, conscience remains no less autistic (the ethical autism of the scrupulous), because one can very well find fault with oneself, without any access to the other. And most often, such is the case. Thus, it is necessary to refrain from thinking the accusative on the basis of moral accusation, or even the call (of conscience), lest one paradoxically reinforce the ontic primacy of an implicit nominative and, in so doing, the ego of metaphysics.
Thus we arrive at the third and final moment. How can one manage to radicalize the accusative that must constitute me without presupposing any subject prior to the accusation? We have stated it: by not making the accusative contingent upon a fault committed by an I that is prior. Doubtless, but am I not, in one way or another, always already guilty of a fault that I committed, and thus am I not always already being before that fault, by being an I in the nominative? Will it be necessary to renounce allowing myself to be named “at the outset” in the accusative, which is to say, before any accusation? Not at all, for there remains a way: the accusative can exert itself in an incontestably extra-moral mode, that is to say, one that does not imply that I be responsible in the mode of an ontic precedent. Now, for this path to become apparent, it is necessary to construct a crucial experiment in accordance with the principle that “to be oneself, the state of being a hostage, is always to have one degree of responsibility more, the responsibility for the responsibility of the other.”54 Let us be precise: at issue here is not a responsibility in addition to a first responsibility, but a degree more in the fact of responsibility; put another way, a responsibility raised to a higher power, a responsibility squared (R2). What responsibility squared can I shoulder, if not precisely a responsibility without any prior fault, without reason, counter to the principle of sufficient reason, and to which I no longer have to answer in conformity to a law or a norm? Namely, the responsibility to let myself be named responsible not for what I would have committed as an I (a pre-being) in the role of efficient cause, but for what I have not committed and, by definition, for that of which I cannot assume the cause, namely for what the other has committed: “being responsible for his brother right up to being responsible for his freedom.”55 But, this result poses a new question: how can I know for certain what the other has committed, without fantasizing or dreaming up a morbid culpability? In order to do this, it would be necessary to know with certainty what only another than I can have committed, and which I, in no case, would have committed, even if I had wanted to: what I would in no case be able to will. Now, this act, which I by definition can neither want nor be able to do (and which is excepted from the principle of sufficient reason) can be identified without difficulty: it is my persecution by the other, the persecution that he would exercise against me, “the unlimited accusative of persecution.”56 Paradoxically, but inevitably, responsibility squared, taken in its extra-moral sense, will be a third stage, truly “a degree [. . .] more.” Not responsibility in the first sense (responsibility for what I have done), nor in the second (responsibility for what the other has done), but truly in the third sense: responsibility for what the other has committed against me, and which certainly imposes itself as originating elsewhere than in me, since the other committed it precisely against me. Such a “responsibility of the ego for what the ego has not wished,” which is deepened in “responsibility for the responsibility of the other,”57 this is what unquestionably constitutes me as a me, precisely because it is unfolded without the will,58 without any prior being,59 in short without me, or rather without I and prior to me. Only this responsibility to a “degree more” and in a decidedly extra-moral sense invests me definitively in the accusative, as an originary me, which is to say without ego: “under accusation by everyone, the responsibility for everyone goes to the point of substitution. A subject is a hostage.”60 The hostage, or the me without self, at the mercy of the one who accuses.
The violence of the term “hostage” requires that we pay close attention. “Subjectivity as hostage”61—certainly, the formula signifies that my exposition to the other does not depend on me, because it precedes me and institutes me by derivation from him. But above all it signifies that my subjectivity does not depend on the other, either, at least not on the other understood according to a relation that would unite me to him as to another being: “Through substitution for others, the oneself escapes relations.”62 Between him and me, there is nothing in common, no third party, no mediator, not even a relation, only the pure “possibility of putting oneself in the place of the other,” which precedes sympathy or respect, because it alone makes them possible. “Hostage” does not define a condition (that would stabilize me on my foundations), but a non-condition, an instability that thus installs me, by derivation, in the situation of ethical becoming: “The non-condition of being hostage [L’incondition d’otage] is not the limit case of solidarity, but the condition for all solidarity.”63 In other words, the hostage (along with the substitution that elicits the hostage) does not arise from ethics, but establishes its conditions of possibility. Substitution does not belong to ethics, but exercises in relation to it a transcendental function: it renders it possible. The hostage, who discovers himself responsible for everyone and for everything that they have done (to me), has nevertheless not committed any fault (he is responsible, not guilty), and he is phenomenalized in the accusative without anyone accusing him of anything (or condemning him). But, without any fault or condemnation, he thus attains his only possible ipseity: “The ipseity, in the passivity without arche characteristic of identity, is a hostage. The word I means here I am, answering for everything and for everyone.”64 Answering for what I have not done—this defines the condition of those who have survived extermination, for all feel the duty (the responsibility) to be responsible for what has happened against them, and above all against those who were annihilated. The obligation to answer for it signifies a duty to speak in their name and to speak against those who did the annihilating: responsibility to respond to the unthinkable, for speech to respond to the unsayable. The deportees and the zeks [Gulag inmates], when they return, speak nothing other than this charge to stigmatize evil, which weighs upon their innocence.
Thus substitution belongs neither to moral philosophy nor to ethics, because it accomplishes ipseity in a nonmetaphysical and non-ontological mode.
Thus substitution asserts itself as the culmination of individuation and of ipseity, precisely by means of the heteronomy that it demands, and that it dispenses. Following in the wake of Heidegger, but also of Sartre, Levinas never gave up on the question of the self and of its mode of being (or of nonbeing), which engaged him from the beginning; but in the end, his answer contradicted in every way both Heidegger and Sartre: the individuation of the self does not pass through the I, its for-itself (pour-soi), nor its mineness (Jemeinigkeit), and in particular not through my possibility, as Dasein, of being-toward-my-death; my individuation, on the contrary, proceeds from my responsibility toward the other, before every accusation and every response, which is to say, from my absolutely unconditioned responsibility, without reason or cause. And this ipseity—contradictory, without reason—thus no longer belongs to metaphysics, whose two principles it contradicts.
So the other accuses me, but first of all in the sense in which, by highlighting (en accusant) the features of a face, one makes it clearer and easier to recognize. The other accuses me in the accusative, such that he places me for the first time in the light, gives me the depth of a face that I would not have without the light of his gaze, and shows me as myself (including to myself). I become myself and unique precisely insofar as the other summons me, thus “accused as unique.”65 What makes me me does not coincide with what I think, nor with what I think of it, nor with what I wish, nor even with what I am—here thinking remains as “indeterminate” as being—but with that for which and to whom I answer. “I am ‘in myself’ through the others.”66 We should even say that this exteriority or deportation of the me outside of itself works its paradox as a sort of phenomenological reduction. Because the issue is that of attaining an “ipseity reduced to the irreplaceable,”67 we must then ask: What is in me such that nothing can substitute for it and replace me? It is neither my thought, which can always be replaced by another, nor my resolution to be or not to be according to my final possibility, which can always be denied, but only my responsibility. My responsibility raised to the point of substitution confirms itself to be the only irreducible in this new reduction. Not in the sense of that for which I would become responsible through my (moral) decision, always subsequent to the (ontic) I, but in the sense of that for which I discover myself responsible without having either willed it or thought it, because it is the others who have in advance made me a hostage of their own responsibilities. I thus find myself, strangely by dint of substitution and in its favor, the “non-interchangeable par excellence,”68 because “nothing is unique, that is, refractory to concepts, except the I of responsibility.”69 Indeed, I become irreducibly myself, or put another way, I identify myself with that which resists every reduction of the me when I accomplish this perfect residuum—not taking up the variations and intermittences of the I (even understood as Dasein), but the fait accompli of my responsibility for that which does not depend on me, never has depended on me and never will—the responsibilities that others have taken without me but for me. For me: not in my favor, but in my place and on my account. This fact, accomplished before me and for me, qualifies me with an unequaled facticity.
Now we catch a better glimpse of the power of the re-reading (in the double sense of a recuperation and a correction) that Levinas carried out on the doctrine of the Selbstheit elaborated in Sein und Zeit. Where the Dasein reigns in the first person, the me in the accusative suddenly appears; where resoluteness decides and wills, the hostage undergoes a decision that he had not made; where Dasein anticipates its possibility, the substitute knows himself to be thrown into what it is no longer possible for him to avoid. But above all, where Dasein manages to individualize itself by itself, and thus to arrive at self-steadfastness or self-constancy (Selbst-ständigkeit), precisely because “the constancy of the self means nothing other than anticipatory resoluteness,”70 the hostage only attains the “superindividuation of the ego”71 (an individuation “to the power”), by entrusting himself to the other, for “the uniqueness of the self is the very fact of bearing the fault of another.”72 Or, what amounts to the same thing, “God loves man as an ipseity.”73
In fact, when Levinas himself commented upon Sein und Zeit section 26, he concluded his objection with these words: “This would be the I of the one who is chosen to answer for his fellowman and is thus identical to itself, and thus the self. A uniqueness of chosenness!”74 It is necessary to go that far: if I am insofar as a hostage, this is because I discover myself always already chosen. There is a “Here I am” even in the “You!” of the one who condemns me. The condition of the hostage, provided that one does not too quickly understand it in the moral sense, indicates, exactly as election, that I am only myself on the basis of the me signified to me by the other. I am not the signification that the other would impose upon me, yet the other nevertheless does indeed signify to me who I am.
Still, other questions remain: might not the “I, unique and chosen,”75 who knows himself only insofar as the other gives him his signification, correspond to the adonné (the gifted), which receives itself from what it receives? Could this be an ethical version of the adonné? Not at all, because what is at issue through and through, as we have seen, is an intrigue in the extra-moral sense, one of strict phenomenology. Does Heidegger’s Jemeinigkeit truly signify the possession of its being by the I? As if what revealed itself as mine would be that which was claimed by an I? Instead, isn’t it a matter of the burden of a decision regarding being, which falls as a responsibility on a me (moi)? And, in this case, would not Dasein also (and first of all) become a hostage—the hostage of being?
These questions thus remain. But this itself proves that, between Heidegger and Levinas, what is at issue is a combat between two types of thought.