One of Hegel’s big ideas is that creatures with a self-conception are the subjects of developmental processes that exhibit a distinctive structure. Call a creature “essentially self-conscious” if what it is for itself, its self-conception, is an essential element of what it is in itself. How something that is essentially self-conscious appears to itself is part of what it really is. This is not to say that it really is just however it appears to itself to be. For all that the definition of an essentially self-conscious being says, what such a one is in itself may diverge radically from what it is for itself. It may not in fact be what it takes itself to be. But if it does mis-take itself, if its self-conception is in error, that mistake is still an essential feature of what it really is. In this sense, essentially self-conscious creatures are self-determining, (partially) self-constituting creatures. Their self-regarding attitudes are efficacious in a distinctive way.

For such a being can change what it is in itself by changing what it is for itself. To say of an essentially self-conscious being that what it is for itself is an essential element of what it is in itself entails that an alteration in self-conception carries with it an alteration in the self of which it is a conception. Essentially self-conscious creatures accordingly enjoy the possibility of a distinctive kind of self-transformation: making themselves be different by taking themselves to be different. Insofar as such a difference in what the essentially self-conscious creature is in itself is then reflected in a further difference in what it is for itself—perhaps just by in some way acknowledging that it has changed—the original change in self-conception can trigger a cascade. That temporal, historical process whereby what the thing is in itself and what it is for itself reciprocally and sequentially influence one another might or might not converge to a stable equilibrium of self and conception of self.

Because what they are in themselves is at any point the outcome of such a developmental process depending on their attitudes, essentially self-conscious beings do not have natures. They have histories. Or, put differently, it is their nature to have not just a past, but a history: a sequence of partially self-constituting self-transformations, mediated at every stage by their self-conceptions, and culminating in them being what they currently are. The only unchanging essence they exhibit is to have what they are in themselves partly determined at every stage by what they are for themselves. Understanding what they are requires looking retrospectively at the process of sequential reciprocal influences of what they at each stage were for themselves and what they at each stage were in themselves, by which they came to be what they now are.

Rehearsing such a historical narrative (Hegel’s “Erinnerung”) is a distinctive way of understanding oneself as an essentially historical, because essentially self-conscious, sort of being. To be for oneself a historical being is to constitute oneself as in oneself a special kind of being: a self-consciously historical being. Making explicit to oneself this crucial structural aspect of the metaphysical kind of being one always implicitly has been as essentially self-conscious is itself a structural self-transformation: the achievement of a new kind of self-consciousness. It is a self-transformation generically of this sort that Hegel aims to produce in us his readers by his Phenomenology. The kind of self-consciousness it involves is a central element in what he calls “Absolute Knowing.”

I suppose that when it is sketched with these broad strokes, this is a reasonably familiar picture. Entitling oneself to talk this way requires doing a good bit of further work, however. Why should we think there are things that answer to the definition of “essentially self-conscious beings”? What is a self? What is it to have a self-conception—to take oneself to be a self, to be a self to or for oneself? For that matter, what is it for anything to be something for one? (Such a concept was taken for granted in Hegel’s Introduction.) And how might the notion of a self-conception, or anything else, being essential to what one really is, what one is in oneself, be cashed out or explained? Hegel’s way of answering these questions, his detailed filling in and working out of the relevant concepts, is no less interesting than the general outline of the story about essentially self-conscious, historical beings those details are called on to articulate.

II. Identification, Risk, and Sacrifice

Let me address the last question first. Suppose for the moment that we had at least an initial grasp both on the concept of a self and on what it is to have a self-conception, something one is for oneself. The story I have just told about essentially self-conscious beings indicates that in order to understand the relationship between selves and self-conceptions, we would need also to understand what it is for some features of a self-conception (how one appears to oneself) to be essential elements of one’s self—that is, what one is in oneself (what one really is). A self-conception may include many accidental or contingent features—things that just happen to be (or be taken to be) true of the self in question. The notion of an essentially self-conscious being applies only if there are also some things that one takes to be true of oneself such that one’s self-conception having those features is essential to one’s being the self one is. How are they to be thought of as distinguished from the rest?

Hegel’s answer to this question, as I understand it, can be thought of as coming in stages. The first thought is that what it is for some features of one’s self-conception to be essential is for one to take or treat them as essential. They are constituted as essential (a normative status) by the practical attitude one adopts toward them. The elements of one’s self-conception that are essential to one’s self (i.e., that one’s self-conception has those features is essential to what one actually is), we may say, are those that one identifies with. Talking this way makes it explicit that essentially self-conscious beings are ones whose identity, their normative status as being what they are in themselves, depends in part upon their meta-attitudes of identification, their attitudes of identifying with some of their attitudes, some privileged elements of what they are for themselves. Of course, saying this does not represent a significant explanatory advance as long as the concept of the practical attitude of identification remains a black box with no more structure visible than its label.

So we should ask: What is it that one must do in order properly to be understood as thereby identifying oneself with some but perhaps not all elements of one’s self-conception? The answer we are given in Self-Consciousness is that one identifies with what one is willing to risk and sacrifice for. Hegel’s metonymic image for this point concerns the important case of making the initial transition from being merely a living organism, belonging to the realm of Nature, to being a denizen of the normative realm of Spirit. The key element in this index case is willingness to risk one’s biological life in the service of a commitment—something that goes beyond a mere desire.¹

It is only through staking one’s life that freedom is won; only thus is it proved that for self-consciousness, its essential being is not [just] being, not the immediate form in which it appears, not its submergence in the expanse of life, but rather that there is nothing present in it which could not be regarded as vanishing moments, that it is only pure being-for-self. [PG 187]

By being willing to risk one’s life for something, one makes it the case that the life one risks is not an essential element of the self one is thereby constituting, while that for which one risks it is. An extreme example is the classical Japanese samurai code of Bushido. It required ritual suicide under a daunting variety of circumstances. To be samurai was to identify oneself with that ideal code of conduct. In a situation requiring seppuku, either the biological organism or the samurai must be destroyed, for the existence of the one has become incompatible with the existence of the other. Failure to commit biological suicide in such a case would be the suicide of the samurai, who would be survived only by an animal. The animal had been a merely necessary condition of the existence of the samurai (like the presence of oxygen in the atmosphere, which is important to us, but with which we do not just for that reason count as identifying ourselves). No doubt even sincere and committed samurai must often have hoped that such situations would not arise. But when and if they did, failure to act appropriately according to samurai practices would make it the case that one never had been a samurai, but only an animal who sometimes aspired to be one. One would thereby demonstrate that one was not, in oneself, what one had taken oneself to be, what one was for oneself. The decision as to whether to risk one’s actual life or to surrender the ideal self-conception is a decision about who one is.

I called the sort of example Hegel uses to introduce this thought “metonymic” because I think that a part is being made to stand for the whole in this image. The point he is after is far broader. For identification in the general sense is a matter of being willing to risk and if need be sacrifice something one actually is (in oneself) for something one is merely for oneself, even if what is risked is not life, but only other commitments or entitlements. It is the sacrifice of a status for an attitude. Hegel’s arresting story of the struggle-unto-death offers a vivid image of one important dimension of the transition from Nature to Spirit. But once the realm of Spirit—all of our normatively and conceptually articulated doings—is an up-and-running enterprise, most of what we have to lose, to risk, and to sacrifice is not a matter of biology, but of culture. What we at these subsequent stages in our development are in ourselves is in large part a matter of status, commitment, authority, and responsibility. Rejecting something one already is because it collides with some commitment is identifying with the commitment one endorsed, by sacrificing something else.

So, for instance, risking or sacrificing one’s job for a point of moral or political principle is a self-constituting act of identification in the same sense that risking or sacrificing one’s life for it is. And acts of identification through risk-or-sacrifice need not be such large-scale, wholesale affairs as these. From the point of view of identification, paying taxes, though seldom a threat to biological endurance (though there is a box labeled “death and taxes”), does belong together with liability to military service (a risk of a risk of life). Both express one’s practical identification, through sacrifice, with the community one thereby defends or supports. Whenever undertaking a new commitment leads to breaking a habit or abandoning a prior intention, one is identifying with that commitment, in practical contrast to what is given up. The historical cascade of sequential self-transformations by identification with elements later sacrificed, each stage building on the previous ones, takes place largely in the normative realm opened up by the initial bootstrapping transition from the merely natural.

Indeed, I want to claim that Hegelian Erfahrung, the process of experience, ought to be understood as having this shape of identification and sacrifice. It, too, is a process of self-constitution and self-transformation of essentially self-conscious beings. Each acknowledged error calls for an act of self-identification: the endorsement of some of the mutually incompatible commitments one has found oneself with, and the sacrifice of others. Experience is the process whereby subjects define and determine themselves as loci of account, by practically “repelling” incompatible commitments. (Compare the way objects are determinately identified and individuated by the specific properties they exhibit, and hence the materially incompatible properties they modally exclude—properties themselves determinately contentful in virtue of their relation of exclusive difference from a specific set of materially incompatible properties.²) Subjects do that by changing their doxastic and inferential commitments: rejecting some, refining others, reciprocally adjusting and balancing what claims are taken to be true, what one is committed to doing, and what is taken to follow from what, so as to remove and repair discordances. This is the process by which the always somewhere colliding and competing claims of the mediating authority codified in universals and the immediate authority exercised by particulars are negotiated and adjudicated. It is accordingly the process by and in which conceptual contents develop and are determined.

III. Creatures Things Can Be Something For: Desire and the Triadic Structure of Orectic Awareness

The story about essentially self-conscious beings, elaborated in terms of identification through risk-and-sacrifice, is what forges the link between the constitution through development of selves and the constitution through development of conceptual contents in the process of experience. And that story presupposes a conception of selves, and so of self-conceptions. In order to entitle ourselves to an account of the shape I have just sketched, we must answer the questions: What is a self? What is it to have a self-conception—to take oneself to be a self, to be a self to or for oneself? For that matter, what is it for anything to be something for one?

The first and most basic notion, I think, is practical classification. A creature can take or treat some particular as being of a general kind by responding to it in one way rather than another. In this sense, a chunk of iron classifies its environments as being of one of two kinds by rusting in some of them and not in others. The repeatable response-kind, rusting, induces a classification of stimuli, accordingly as they do or do not reliably elicit a response of that kind. Because reliable differential responsive dispositions are ubiquitous in the causal realm, every actual physical object exhibits this sort of behavior. For that reason, this sort of behavioral classification is not by itself a promising candidate as a definition of concepts of semantic content or awareness; pansemanticism and panpsychism would be immediate, unappealing consequences.

Hegel’s alternative way in is to look to the phenomenon of desire, as structuring the lives of biological animals. A hungry animal treats something as food by “falling to without further ado and eating it up,” as Hegel says. [PG 109] This is clearly a species of the genus of practical classification. The state of desiring, in this case hunger, induces a two-sorted classification of objects, into those consumption of which would result in satisfying the desire, and the rest. The constellation of hunger, eating, and food has structure beyond that in play in the inorganic case of rusting (response) and wet (stimulus), though. What ultimately drives the classification is the difference between hunger being satisfied and that desire not being satisfied. But it is important that the classification of objects by that difference is subjunctively conditioned on a mediating performance, process, or response. What is classified is objects that if responded to by eating would satisfy the hunger, and those that do not have that property. Both the role played by the practical activity of the desirer—that is, what it does in response to the object—and the subjunctive, hypothetical-dispositional character of the classification in terms of the effect that that doing would have on the satisfaction of the desire are important to Hegel’s picture.

Desires and the responsive practical performances that subserve them play distinctive roles in the lived life of an animal. They are intelligible in terms of the contributions they make to such functions as its nutrition, reproduction, avoidance of predation, and so on. Because they are, they direct the orectic awareness of the desiring animal to the objects that show up as significant with respect to them in a distinctive way. They underwrite a kind of primitive intentionality whose character shows up in the vocabulary it entitles us to use in describing their behavior. Dennett considers in a related context a laboratory rat who has been conditioned to produce a certain kind of behavior in response to a stimulus of a repeatable kind, say, the sounding of a certain note.³ We can in principle describe the repeatable response in two different ways: “The rat walks to the bar, pushes it down with his paw, and sometimes receives a rat-yummy,” or “[t]he rat takes three steps forward, moves its paw down, and sometimes receives a rat-yummy.” Both describe what the rat has done in each of the training trials. What has he been conditioned to do? Which behavior should a reductive behaviorist take it has been inculcated and will be continued? Abstractly, there seems no way to choose between these coextensional specifications of the training. Yet the way in which desiring organisms like rats are directed at desire-satisfying objects via expectations about the results of performances lead us confidently to predict that if the rat is put six steps from the bar, when the note sounds it will walk to the bar and push it down with his paw, not walk three steps forward and move its paw down. We do so even in this artificial case for the same reasons that we expect that if we move a bird’s nest a few feet further out on a limb while it is away, on its return the bird will sit in the nest in its new location, rather than on the bare limb in the nest’s old location. The bird is “onto” its nest (to use a locution favored by John McDowell in this context) rather than the location. That is, it is onto the object that has acquired a practical significance because of the functional role it plays in the animal’s desire-satisfying activities. A desire is more than a disposition to act in certain ways, because the activities one is disposed to respond to objects with may or may not satisfy the desire, depending on the character of those objects.

Orectic awareness has a triadic structure, epitomized by the relations between hunger, eating, and food. Hunger is a desire, a kind of attitude. It immediately impels hungry animals to respond to some objects by treating them as food—that is, by eating them. Food is accordingly a significance that objects can have to animals capable of hunger. It is something things can be for desiring animals. Eating is the activity of practically taking or treating something as food. It is what one must do in order in practice to be attributing to it the desire-relative orectic significance of food. Eating is the activity that is instrumentally appropriate to the desire of hunger. It is subjectively appropriate, in that it is the activity hungry animals are in fact impelled to by being in the desiring state of hunger. It is objectively appropriate in that it is an activity, a way of responding to environing objects, that often (enough) actually results in the satisfaction of the desire.

This distinction between two sorts of instrumental propriety of activity to desire funds a distinction between appearance and reality for the objects responded to—a distinction between what things are for the organism (the orectic significance they are taken to have) and what things are in themselves (the orectic significance they actually have). Anything the animal responds to by eating it is being practically taken or treated as food. But only things that actually relieve its hunger really are food. The possibility of these two coming apart is the organic basis for conceptual experience, which in its fully developed form is the collision of incompatible commitments. Even at the level of merely orectic awareness, it can lead to the animal’s doing things differently, in the sense of altering which objects it responds to by treating them as having the orectic significance generated by that desire. Its dispositions to respond to things differentially as food—that is, by eating them—can be altered by such practical disappointments. If all goes well with an experiential episode in such a process of learning, the subjectively appropriate differential responsive dispositions become more reliable, in the sense of more objectively appropriate to the desire that motivates those activities.

IV. From Desire to Recognition: Two Interpretive Challenges

This account of the triadic structure of orectic awareness offers a reasonably detailed answer to the question: What is it for things to be something for a creature? It is a story about a kind of protoconsciousness that is intelligible still in wholly naturalistic terms, and that nonetheless provides the basic practical elements out of which something recognizable as the sort of theoretical conceptual consciousness discussed in the first three chapters of the Phenomenology could perhaps be understood to develop. We know that Hegel subscribes to the Kantian claim that there can in principle be no consciousness (properly so described—that is, no sapient, apperceptive awareness) without self-consciousness. So making the step from the orectic awareness of animal denizens of the realm of Nature to the conceptual consciousness of knowers and agents who live and move and have their being in the normative realm of Spirit—creatures who have achieved the status of selves or subjects—requires the advent of self-consciousness. We need to understand what this achievement consists in, and why genuine consciousness requires it. As we will see, what is required to be able to take something to be a self is to be able to attribute attitudes that have distinctively normative significances: to move from a world of desires to a world of commitments, authority, and responsibility.

The triadic structure of orectic awareness gives us a place to start in addressing this issue. We should apply the answer we have in hand to the question: What is it for things to be something for a creature? to the more specific case: What is it for selves to be something things can be for a creature? That is, what would be required for the orectic significance something had for a desiring animal to be not food or predator, but self or subject, in the sense of something things can be something for? And then, once we understand what it is to take or treat things as selves or subjects, we can ask further what one must do to take oneself to be a thing of that kind: what one must do to take oneself to be a self.

The account of the triadic structure of orectic awareness provides two sorts of resources for answering these questions. First, it tells us something about what a self or subject is. It is something things can be something for. What it offers is a construal of that status in terms of what it is to be a desiring animal, a subject of orectic awareness, an institutor of orectic significances, an assessor of the consilience or disparity of what things are for it or subjectively and what they are in themselves or objectively, the subject of the experience of error and the cyclical feedback process of revision-and-experiment it initiates and guides. This is what a (proto)self in the sense of a subject of orectic awareness is in itself. The question, then, is what it is for something to be one of those, to have that orectic significance, for some (to begin with, some other) creature. The second contribution the triadic structure of orectic awareness makes to understanding the nature and possibility of self-consciousness consists in providing the form of an answer to this more specific question. For it tells us that what we must come up with to understand what it is for something to be accorded this sort of orectic significance by some creature—to be for it something things can be something for—is twofold: an account of the desire that institutes that orectic significance, and an account of the kind of activity that is instrumentally appropriate to that desire. The latter is an account of what one must do in order thereby to count as taking some creature as itself a taker, something things can be something for, an instituter of orectic significances.

The philosophical challenge, then, is to see what sort of an account of self-consciousness one can produce by assembling these raw materials: by applying the triadic account of orectic awareness to itself. The interpretive challenge is to see to what extent one can by doing that explain the index features characteristic of Hegel’s distinctive claims about the nature of self-consciousness. Two features of his approach are particularly worthy of attention in this regard, both of them stemming from his master-concept of recognition. First is his view that both self-conscious individual selves and the communities they inhabit (a kind of universal characterizing them) are synthesized by reciprocal recognition among particular participants in the practices of such a recognitive community. For Hegel, self-consciousness is essentially, and not just accidentally, a social achievement. Second, recognition is a normative attitude. To recognize someone is to take her to be the subject of normative statuses—that is, of commitments and entitlements—as capable of undertaking responsibilities and exercising authority. This is what it means to say that as reciprocally recognized and recognizing, the creatures in question are geistig, spiritual, beings, and no longer merely natural ones. Here are some of the familiar representative passages:

Self-consciousness exists in and for itself when, and by the fact that, it so exists for another; that is, it exists only in being acknowledged.… The detailed exposition of the Notion of this spiritual unity in its duplication will present us with the process of Recognition. [PG 178]

A self-consciousness exists for a self-consciousness. Only so is it in fact self-consciousness; for only in this way does the unity of itself in its otherness become explicit for it.… A self-consciousness, in being an object, is just as much “I” as “object.” With this, we already have before us the Notion of Spirit. What still lies ahead for consciousness is the experience of what Spirit is—this absolute substance which is the unity of the different independent self-consciousnesses which, in their opposition, enjoy perfect freedom and independence: “I” that is “We” and “We” that is “I.” [PG 177]

But according to the Notion of recognition this [that a self-consciousness’s certainty of itself have truth] is possible only when each is for the other what the other is for it, only when each in its own self through its own action, and again through the action of the other, achieves this pure abstraction of being-for-self. [PG 186]

I see two principal philosophical challenges that arise in understanding the discussion of recognition and self-consciousness in these and related passages. First, how are we to understand the transition from the discussion of the concept of desire to the discussion of the concept of recognition? This corresponds to the shift from consideration of particular merely biological creatures impelled wholly by natural impulses, in relation to their species, on the one hand, to consideration of genuinely social self-conscious individuals motivated by normative relations of authority and responsibility within their communities, on the other. How one understands the relation between these, both conceptually and historically, is evidently of the first importance in understanding what Hegel has to teach us about the normative realm he calls “Geist.”

The second issue concerns the formal structure of his account of the synthesis of social substance by relations of reciprocal recognition. To recognize someone is to take or treat that individual in practice as a self: a knowing and acting subject, hence as subject to normative assessment as potentially committed, responsible, authoritative, and so on. The picture that is presented of the sort of community within which fully adequate self-consciousness is achieved is one in which recognition is an equivalence relation: everyone in the community recognizes and is recognized by everyone else (“each is for the other what the other is for it”), and so recognizes everyone recognized by anyone else. Individuals are, roughly, particulars whose exhibition of, characterization by, or participation in universals is essential to them. In the case of self-conscious individuals, this means that the norms of the community they are members of are essential equally to the individual members and to the community as a whole.⁴

In such an ideal community, each member is to be able to recognize himself as a member. To say that is to say that recognition is reflexive. Reflexive self-recognition is the form of self-consciousness. Recognition is also to be symmetrical, that is to say, mutual or reciprocal (Hegel’s “gegenseitig”). It is this aspect that is lacking in the defective forms of recognition that structure the defective forms of self-consciousness rehearsed in the Phenomenology, under the allegorical heading Mastery. The view appears to be that insofar as recognition is de facto not symmetrical, it cannot be reflexive. I cannot be properly self-conscious (recognize myself) except in the context of a recognition structure that is reciprocal: insofar as I am recognized by those I recognize. (This is the essence of Hegel’s Wittgensteinian view of self-consciousness, which by contrast to a Cartesian view sees it as a social achievement, which accordingly takes place in important respects outside the self-conscious individual. It is not a kind of inner glow.)

A big question is, then: Why? Why should it be the case that reciprocal (that is, symmetrical) recognition is a necessary condition of reflexive recognition (that is, self-consciousness, awareness of oneself as a self)?

Here is a thought about the shape of a possible answer. It is a formal fact that if a relation is both symmetrical and transitive, then it is also reflexive, and hence is an equivalence relation. That is, if ∀x,y[xRy→yRx] and ∀x,y,z[xRy&yRz→xRz], then ∀x[xRx]. For we can just apply the transitivity condition to the symmetry pairs xRy and yRx to yield xRx.⁵ So if recognition were (for some reason) de jure transitive—if it were part of the nature of recognition that one is committed to recognizing anyone recognized by someone one recognizes—then achieving de facto symmetry of recognition would suffice for achieving de facto reflexivity of recognition. That is, each community member would recognize himself—and in that sense count as self-conscious—so long as everyone was recognized by everyone they recognize (that is, so long as recognition was reciprocal). So one way to forge the desired connection between social reciprocity of recognition and self-consciousness would be to establish that recognition must by its very nature be transitive.

In what follows, we see how the triadic account of orectic awareness can be used in a natural way to build a notion of recognition that satisfies these twin philosophical constraints on the interpretation of Hegel’s notion of self-consciousnesss in terms of desire and recognition. Doing so will both clarify the nature of the transition from desire to recognition and explain why reciprocal recognition is the key to self-consciousness.

V. Simple Recognition: Being Something Things Can Be Something for Is Something Things Can Be for One

We can think of the triadic structure of orectic awareness (TSOA) as consisting of three elements and three relations among them. The three elements are

an attitude (desire), e.g., hunger;
a responsive activity, e.g., eating; and
a significance, e.g., food.

The three relations are

The attitude must motivate the activity, in the sense of activating a (more or less reliable, in a sense determined by the assessments in the following [6]) disposition to respond differentially to objects.
Responding to an object by engaging in the activity is taking or treating it in practice as having a significance defined by the attitude that motivates the activity. This is the subjective significance of the object.
The desiring attitude assesses the object, implicitly attributing to it an objective significance, accordingly as responding to it by engaging in the activity the attitude motivates does or does not satisfy the desire. If it does not, if what the object was subjectively or for the animal does not coincide with what it was objectively, or in itself—that is, if the activity was not successful in satisfying the motivating desire—then an error has been committed. In that case the desire motivates changing the reliable differential responsive disposition to engage in the associated activity when activated by the desire and stimulated by a range of objects. This is learning, as the result of experience.

We are now interested in a more complicated constellation of elements and relations, in which the TSOA enters twice. It is, of course, the structure of the whole thing: “Self-consciousness is desire,” [PG 174] at least in the sense that the most primitive form of self-awareness is to be understood as a development of the basic structure of orectic awareness. And we want to understand why self-consciousness is also a recognitive status, one that specifically requires reciprocal recognition. Within the triadic structure of desire as it applies to self-consciousness, the significance attributed to an object, what it is for the organism exhibiting the orectic awareness in question, is to be orectically aware: to be something things can be something for. That is, the significance attributed by engaging in a responsive activity and assessed by the motivating attitude must itself exhibit the TSOA. For one to have that significance for oneself—not just being in oneself something things can be something for, but being that for oneself as well—that significance must be something things can be or have for one.

The triadic structure of orectic awareness tells us that the two big questions that must be answered are these:

What activity is it that institutes this significance (namely, having the TSOA)? That is, what is it that one must do, how must one respond to something, to count thereby as taking or treating it as exhibiting the TSOA? What is to the TSOA as eating is to food?
What desire or other attitude is it that motivates that activity and assesses the success of taking something as having the orectic significance of being a TSOA, i.e., being something things can be something for? What is to the TSOA as hunger is to food?

To begin to address these questions, and to indicate an important point of contact with Hegel’s own vocabulary, we may call what I must do, the activity, whatever it is, that I must engage in, in order thereby to be taking or treating something in practice as something things can be something for, “recognizing” that other creature. So far, we should think of this term as just a label for an answer to the first question. Recognizing others is attributing to them the practical significance of exhibiting the triadic structure of orectic awareness: taking them to be takers, subjects for whom things can have a practical significance, relative to a desire, and mediated by an activity. What can we then say at this level of abstraction about the desire or attitude that is the third element completing the TSOA whose activity is recognizing and whose significance is exhibiting the TSOA? Hegel’s answer is, I think, clear, if surprising: it is desire for recognition, the desire that others take or treat one in practice as a taker, as something things can be something for, as an instituter of significances.

If we put aside for the moment the crucial question of why a desire to be recognized should be the attitude for which recognizing others is the appropriate activity, and so why it institutes the significance of being something things can be something for—making that something things can be for one, a protoconception of selves—we may ask what would happen if a being with that desire got what it wanted. If the desire for recognition is satisfied by responding to others by recognizing them, then according to the TSOA the subjective significance the recognized ones have for the recognition-desirer shows up as being correct, as what they objectively are in themselves: subjects of significance-instituting attitudes and activities. And what is required for that is just that one be recognized (for that is what it takes to satisfy the desire) by those one recognizes (for that, on the line of thought being considered, is what one must do in order, if all goes well, to satisfy the desire). So if it is a desire for recognition that completes the higher-order TSOA whose activity is recognition and whose instituted significance is exhibiting the TSOA, it follows that the recognition-desire can be satisfied only by achieving reciprocal recognition. On this construal, then, having a practical protoconception of selves—being able to take or treat things as subjects things can be something for, recognizing them—and being self-conscious in the sense of reciprocal recognition are two aspects of one achievement, two sides of one coin.

In order to give a reading of these claims in terms of the triadic structure of orectic awareness, the black box notion of recognition must be filled in so as to answer the following three questions:

Recognizing: What, exactly, is it that one must do in order to be recognizing someone? That is, what is the activity we have labeled “recognizing”? How is it that doing that is taking or treating someone as exhibiting the TSOA? What is the differential responsive disposition that is to be licensed by the instituting attitude?
Being recognized: Why should the desire to be taken or treated that way oneself (that is, to be recognized) be the one making appropriate that activity—namely, recognizing?
Self-consciousness: Why does the reciprocal recognition that results when that desire for recognition is satisfied by recognizing someone else amount to self-consciousness, in the sense of applying a (proto)conception of selves to oneself?

The challenge is to give an answer to the first question that will entail plausible answers to the other two questions.

The first point to make is that general recognition, taking someone to be something things can be something for, must be understood in terms of specific recognition: taking someone to be something things can have a specific significance for, say, being of kind K (e.g., food, a predator, a potential sexual partner). One takes someone to be a taker in general just in case there are some specific significances, values of K, for which one takes it that that individual is a K-taker, i.e., can take things to be Ks. So in order to answer the preceding questions for the more general case, it will suffice to answer those questions for specific recognition, relativized to some instituted significance K that things can have for a creature.

According to the TSOA, specifically recognizing someone as a K-taker requires responding to the other in a way that practically or implicitly attributes both an attitude and an activity related to each other and to the significance K in the three ways previously specified as (4), (5), and (6). This means:

One must attribute an activity that one takes to be what it is for the other to be responding to something as a K.
One must attribute a desire or other attitude that one takes to license or authorize responding to things as Ks—that is, by engaging in that activity.
One must acknowledge in practice a distinction between correct and incorrect responses of that sort, assessed according to the attributed attitude that authorizes responses of that kind.

We can start with the thought that in the most basic case, one can take another to be a K-taker only if one is oneself a K-taker. Taking the other to be a K-taker will then be attributing to him activity of the same sort in which one oneself engages in response to things one (thereby) takes to be Ks. That is, my taking you to be able to treat things as food is my taking it that you respond to some things with the same behavior, eating, with which I respond to food.

We are now in a position to put in place the keystone piece of this explanatory structure. What the recognizing attributor responds differentially to as the success of a desire-authorized responsive activity is the cessation of that activity. Thus no longer being disposed to respond by eating things indicates that hunger was satisfied, so the thing previously responded to as food was in itself what it was for the one recognized as a desirer of food.

What, then, is the differential response that is keyed to this difference in the one being recognized as a K-taker? This is the decisive point. My taking your K-response to have been authorized by a K-desire that serves as a standard for the success of your K-taking, and taking that K-response to have been correct or successful by that standard is my acknowledging the authority of your K-taking, in the practical sense of being disposed myself to take as a K the thing you took to be a K. Taking it that the kind of fruit you ate really was food, in that it satisfied your hunger, is being disposed to eat that kind of fruit myself when and if I am hungry, i.e., have a desire of the same kind. This is a second-order disposition, involving a change in my first-order dispositions. My specific K-recognitive response to you is to acquire the disposition: if I have the K-desire, then I will K-respond to the things to which I (thereby) take you to have successfully K-responded. My acknowledging your K-desire as authoritative in the dual sense of licensing your responsive K-activity and serving as a standard of normative assessment of its success or correctness consists in my treating it as authorizing my own K-takings, should I have a K-desire.

So in the first instance, my treating your K-desire as having the normative significance of being authoritative for K-takings is treating it as authoritative for them full stop—not just for your K-takings, but for K-takings generally, and so for mine in particular. What it is for it to be K-takings (and not some other significance or no significance at all that you are practically attributing to things by responding to them in that way) that I take your responses to be consists in the fact that it is my K-taking responsiveness (and not some other activity) that I am conditionally disposed to extend to the kind of objects that satisfied your desire. The link by which the specifically recognized one’s activity is assimilated to that of the recognizer is forged by the interpersonal character of the specific authority of the recognized one’s successful takings, whose acknowledgment is what specific recognition consists in. The only way the recognizer’s orectic classifications can be practically mapped onto those of the other so as to be intelligible as implicitly attributing specific desires, significances, and mediating responsive activities exhibiting the TSOA, is if the authority of the assessments of responsive significance-attribution on the part of the one recognized is acknowledged in practice by the recognizer. So specific recognition involves acknowledging another as having some authority concerning how things are (what things are Ks). When I do that, I treat you as one of us, in a primitive normative sense of “us”—those of us subject to the same norms, the same authority—that is instituted by just such attitudes.

VI. Robust Recognition: Specific Recognition of Another as a Recognizer

Looking back at the most primitive sort of preconceptual recognition of others from the vantage point of the fully developed, conceptually articulated kind brings into relief the crucial boundary that is being crossed: between the merely natural and the incipiently normative. In the merely orectically aware animal, desire is a state that motivates and regulates responsive activity immediately. It causally activates differential responsive dispositions to engage in activities, and its matter-of-factual satisfaction causes the creature to desist from or persist in them. But the recognizer, who is aware of the creature as aware of things, does not feel that creature’s desires, but only attributes them, implicitly and practically, by treating the creature as having them. The recognizer accordingly takes up a more distanced, mediated, abstract attitude toward these significance-generating attitudes. The recognized creature’s attitudes are treated in practice as assessing the correctness of practical responsive classifications, as licensing or authorizing the responsive activity—in the first instance in the case of the one recognized, but then also on the part of the recognizer, who merely attributes the attitude to the other. The relation between the attitude the recognizer attributes and the activity he himself engages in is a normative one. Even in the most primitive cases it is intelligible as the acknowledgment of authority rather than mere acquiescence in an impulse. In treating the attitudes of the recognized other as having authority for those who do not feel them, the recognizer implicitly accords them a significance beyond that of mere desires: as normatively and not merely immediately significant attitudes.⁶

The story I have rehearsed about what happens when the triadic structure of orectic awareness is applied to itself as significance shows how recognition develops out of and can be made intelligible in terms of desire. But it also shows why just being orectically aware is not enough to give one a conception of a self. That is something one can get only by recognizing others. For the possibility of treating attitudes as having a distinctively normative significance opens up in the first instance for the attitudes of others, for desires one attributes but does not immediately feel. The claim we have been shaping up to understand is Hegel’s central doctrine that self-consciousness consists in reciprocal recognition. It is clear at this point that recognizing others is necessary and sufficient to have a conception of selves or subjects of consciousness. But the relation between that fact and reciprocity of recognition as what makes it possible for the participants to count as applying that concept to themselves in the way required for self-consciousness has not yet been made out. To make it out, we can exploit the observation made in the previous section: that if recognition could be shown to be de jure transitive, then any case in which it was also de facto symmetrical (reciprocal) would be one in which it was also de facto reflexive. For reflexivity follows from transitivity and symmetry.

Simple recognition is not in the relevant sense transitive. For what I am doing in taking another to be a subject of orectic awareness—namely, simply recognizing that desirer as a desirer—is not what I take that desirer to be doing. The one simply recognized need not be capable of being in its turn a simple recognizer, and so something with even a basic conception of selves. For that we need to go up a level, and consider what it is to take another not just to be orectically aware, but to be aware of others as orectically aware. That is, we must consider what it is to recognize another as a simple recognizer, hence as itself the kind of thing for which things can have a specifically normative significance. I call that practical attitude robust recognition. Robust recognition is a kind of simple recognition: simple recognition of someone things can have a specific kind of orectic significance for—namely, the significance of being something things can have orectic significances for.

What is important for my story is that robust recognition is transitive. This is clear from the account already offered of recognition in terms of acknowledging the authority of what things are for the recognized one. Recognizing someone as a recognizer is acknowledging the authority of their recognitions for one’s own: recognizing whoever they recognize. If one presses the details, there are some subtleties that arise, but when they have been carefully considered, the basic conclusion about the transitivity of robust recognition remains. So I suppress them here.

Because it is a kind of simple recognition, the activity element of the orectic structural triad characteristic of robust recognition—what one must do to be taking or treating someone as (having the significance of) a simple recognizer—is practically to acknowledge as authoritative for one’s own takings the takings of the one being recognized (if they are successful, and within the range of significance of one’s simple recognition). In this case, doing that is acknowledging the authority of the recognized one’s simple recognitions. Those simple recognitions are themselves a matter of acknowledging the authority of the ground-level orectic takings of the one simply recognized. So what the robust recognizer must do to be taking someone as a simple recognizer is to acknowledge as authoritative whatever ground-level takings the one robustly recognized acknowledges as authoritative. And that is to say that the robust recognizer in practice (and so, implicitly) treats as transitive the inheritance of authority of ground-level takings that is what simple recognizing consists in.

It might seem that the hierarchy generated by acknowledging different levels of recognition is open-ended: robust recognition is taking to be (simply recognizing as) a simple recognizer, super-robust (say) recognition would be simply recognizing as a robust recognizer, super-duper-robust recognition would be simply recognizing as a super-robust recognizer, and so on. Perhaps surprisingly, the crucial structural features of recognition do not change after we have reached robust recognition. The key point is that robust recognition is a specific instance of simple recognition, i.e., recognition of something as having a special kind of orectic awareness—namely, awareness of something as being orectically aware. As we have seen, that is a particular kind of orectic significance things can have. As a result of this fact, the nascent recognitional hierarchy could be formulated this way: orectic awareness, simple recognition of something as orectically aware, simple recognition of something as simply recognizing, simple recognition of something as a simple recognizer of simple recognizers, and so on. But what one must do in order thereby to be simply recognizing someone—the activity (corresponding to eating in the paradigmatic orectic desire-activity-significance triad of hunger, eating, food) one must engage in to count as taking or treating an organism as (having the significance of being) orectically aware—is to acknowledge the normative authority for one’s own responses of their takings of things as something. Taking someone to be a simple recognizer is accordingly acknowledging in practice the authority of their takings of someone as an orectic taker, which is acknowledging the authority of their acknowledgings of authority. Whatever ground-level takings of things as something the one being robustly recognized (simply recognized as a simple recognizer) takes to be authoritative, the robust recognizer takes therefore to be authoritative. In robustly recognizing you, I must simply recognize whoever you simply recognize.

The effect is to produce the transitive closure of the acknowledgment of authority of ground-level takings in which simple recognition consists. By the “transitive closure” of a relation is meant the relation R′ that is generated from R by the two principles: (i) ∀x ∀y[xRy→xR′y] and (ii) ∀x∀y∀z[(xRy&yRz)→xR′z]. It is an elementary algebraic fact that the transitive closure of the transitive closure of a relation is just the transitive closure of that relation. (Technically, closure operations are idempotent.) All the structural work has been done the first time around. For a to recognize b in the “super-robust” way—simply to recognize b as a robust recognizer—would commit a to acknowledge as authoritative b’s simple recognitions of someone c as a simple recognizer. Now b’s simple recognition of c as a simple recognizer (which is b’s robust recognition of c), we have seen, consists in b’s practical commitment to inherit c’s acknowledgments of another’s—d’s—ground-level takings as authoritative. The effect is then that a must likewise be practically committed to inherit b’s inherited acknowledgments of those ground-level commitments as authoritative. But this puts a in exactly the position a would be in if a recognized b robustly, rather than super-robustly. Formally, once one has established that a relation is transitive, that ∀x∀y∀z[(xRy&yRz)→xRz], that has as a consequence (and hence requires nothing else to establish) that ∀w∀x∀y∀z[(wRx&xRy&yRz)→wRz].

Because robust recognition is the transitive closure of simple recognition, there is no difference between simple recognition of someone as a robust recognizer, and robust recognition (simple recognition of someone as a simple recognizer) of someone as a robust recognizer. And robust recognition is transitive: for what one is doing to be robust recognizing, it must include commitment to robustly recognize (simply recognize as a simple recognizer) whoever is robustly recognized by those one robustly recognizes. These are facts about the activity pole of the structure of simple recognition, and therefore of robust recognition. What relates them is that the significance pole of robust recognition is the whole structure of simple recognition—just as the significance pole of simple recognition is the whole triadic structure of ground-level orectic awareness. Indeed, we have seen that the significance pole of ground-level orectic awareness is the crucial element in the activity pole of simple recognition (and therefore of robust recognition). For practical acknowledgment of the authority of the ground-level significances attributed in nonrecognitional orectic awareness is what the activity of simple recognizing consists in.

If these are the relations between the activity and significance poles making up the triadic structure of recognitional awareness, what, then, about the attitude or desire pole? The story told so far lays it down both that the desire that motivates simple recognizing (and so institutes its characteristic significance) is a desire for (simple) recognition, and that the only orectic takings on the part of one recognized that a simple recognizer is obliged to acknowledge as authoritative are those that the one recognized takes to be successful. So we should ask: Which of the recognizings of a simple recognizer should a robust recognizer take to be successful? The answer is: Only those that satisfy the relevant desire. That is, a desire to be simply recognized, which is to say a desire to have the authority of the simple recognizer’s takings acknowledged by another. But that is precisely what a robust recognizer does in simply recognizing anyone as a simple recognizer. So from the point of view of a robust recognizer, all the simple recognitions of the one robustly recognized count as successful, and hence as authoritative. There is nothing that could count as taking someone to have a desire to be simply recognized, motivating that one’s simple recognitions, that fails to be satisfied.

With this observation, we have reached our explanatory-interpretive goal. For we wanted to know

how recognition should be understood to arise out of desire,
how normativity should be understood as an aspect of recognition,
how self-recognition—that is, reflexive recognition relations—should be understood to require reciprocal recognition, that is to say symmetrical recognition relations, and
how self-consciousness should be understood to consist in the self-recognition achieved by reciprocal recognition.

The answer to the first question was supplied by seeing how the triadic structure of orectic awareness could be applied to itself, so that what something was taken or treated in practice as was a desiring, significance-instituting creature. The answer to the second was supplied by seeing how simple recognizing consists in the recognizer’s achieving a mediated, distanced, relation to the immediate felt impulse of the recognized one’s desire, in the form of its significance, conditional upon the recognizer’s own desires, for the recognizer’s own practical awareness. In this way the other’s desire is practically acknowledged as authoritative, and the other’s desire shows up for the recognizer in the shape of the recognizer’s commitment or responsibility. The answer to the third question was supplied by showing how (because of the idempotence of transitive closure operations) the social authority structure constitutive of robust recognition is essentially and in principle, hence unavoidably, transitive. For it is a basic algebraic fact that wherever a transitive relation happens to hold symmetrically, it is also reflexive. The argument that a relation’s being de jure transitive and de facto symmetrical suffices for it to be reflexive shows a certain kind of sociality to be sufficient for self-recognition. What shows it to be necessary is that I cannot practically distinguish impulse from authority in my own case. It remains only to put these answers together to supply a response to the fourth and final question.

VII. Self-Consciousness

The connection between robust recognition and self-consciousness is as immediate as that between the triadic structure of orectic awareness and consciousness. For to be a self, a subject, a consciousness—for Hegel as for Kant—is to be the subject of normative statuses: not just of desires, but of commitments. It is to be able to take a normative stand on things, to commit oneself, undertake responsibilities, exercise authority, assess correctness. Recognition of any kind is taking or treating something as such a self or subject of normative statuses and attitudes. It is consciousness of something as (having the normative significance of) a self or subject. For recognition itself exhibits the TSOA—protoconsciousness. The significance it accords to the one recognized is that of exhibiting that same structure. And adopting that practical attitude toward another is taking or treating its states as having normative significance as authorizing and assessing performances—not merely producing them but making them appropriate. Eating on the part of the one recognized is now treated as something that involves a commitment as to how things are, a commitment that can be assessed by both recognized and recognizer (who need not agree) as correct or incorrect.

Self-consciousness consists in applying this practical protoconception of a self to oneself: recognizing not just others, but oneself. This is self-consciousness, or having a self-conception, in a double sense. First, it is a matter of consciousness of something as a self: treating it as having that practical significance. Second, it is an application of that conception to oneself. Having a self-conception in the first sense consists in a capacity for recognition. We might call this a “conception of selves.” For recognizing is what one must be able to do in order thereby to be taking or treating something as a self, in the sense of a subject of normative statuses of authoritative (in the sense of probative, though still provisional and defeasible) commitments as to how things are. Having a self-conception in the second sense is a matter of the reflexive character of one’s recognition: that among those one recognizes is oneself. The lowest grade of self-conception that exhibits these two dimensions would be simple recognition of oneself: being orectically aware of oneself as orectically aware of things. We might call this “simple self-consciousness.” But the two dimensions are much more tightly bound up with one another if one is aware of oneself as able simply to recognize things. In that case, the conception of selves that one applies to oneself is as something that has a conception of selves. We might call this “robust self-consciousness.”

If a robustly recognizes b, then a acknowledges the (probative, but provisional and defeasible) authority of b’s successful simple recognitions. Robust recognition, we have seen, is a kind of simple recognition: simple recognition of someone as able to take others to be simple recognizers. If b robustly recognizes someone, say c, then that recognition is successful just in case it satisfies b’s desire for robust recognition. If b’s robust recognition of someone is successful in this sense, then in virtue of robustly recognizing b, a must acknowledge b’s robust recognition as authoritative. But because by hypothesis a does robustly recognize b, b’s desire for robust recognition is satisfied, all his robust recognitions are successful (in a’s eyes). Thus if it should happen that b does robustly recognize a, then because a robustly recognizes b, we have a symmetry of robust recognition. Because, as we have seen, robust recognition is transitive, this means that a will acknowledge the authority of b’s robust recognition of a. So a counts as robustly recognizing himself. Thus robust self-consciousness is achievable only through reciprocal recognition: being robustly recognized by at least some of those one robustly recognizes. This means that a community (a kind of universal) is implicitly constituted by one’s own robust recognitions, and actually achieved insofar as they are reciprocated. That is the sort of reciprocally recognitive community within which alone genuine (robust) self-consciousness is possible. What Hegel calls the “ ‘I’ that is ‘We’ and ‘We’ that is ‘I’ ” is an equivalence class defined by means of dyadic recognitive “I” / “thou” relations that are transitive, symmetrical, and reflexive.⁷

VIII. Conclusion

I can now bring my story to a quick close. I started it with the concept of essential elements of one’s self-conception being ones that one identifies with, in the sense of being willing to risk or if need be sacrifice for them. One consequence of the transition from desire to commitment within the attitude component of the triadic structure of orectic awareness is that where the activity-motivating character of desire is extinguished with its satisfaction, the activity-licensing character of commitment need not be. In particular, desire for recognition in the form of a commitment to being recognized is a standing, structural element of self-consciousness. It persists even when fulfilled by the achievement of reciprocal recognition that is self-consciousness. Because it persists as part of the necessary background against which any other commitments are adopted and relinquished, being for oneself a recognizer is an essential element of one’s self-conception. One’s identification with it consists practically in the structural impossibility of relinquishing that commitment in favor of others. To be self-conscious is to be essentially self-conscious: to be for oneself, and identify oneself with oneself as something that is for oneself a recognized and recognizing being.

A fuller telling would continue with an account couched in the same basic terms of the specific distorted form of self-consciousness that construes itself under the distinctively modern, alienated category of pure independence (authority without correlative responsibility) that Hegel epitomizes in the form of the “Master.” It would explain how the self-conception characteristic of Mastery arises from overgeneralizing from its capacity immediately to constitute itself as essentially self-conscious—making it so just by taking it so—to yield an ultimately incoherent model of a self-consciousness all of whose conceptions are immediately constitutive, thus eliding quite generally the crucial “distinction that consciousness involves,” between what things are for it and what they are in themselves. And it would explain what Hegel elsewhere calls “die Wirkung des Schicksals”: the causality of fate. This is the metaphysical irony that undermines the Master’s existential commitment to possessing authority without correlative responsibility, to being recognized as authoritative without recognizing anyone as having the authority to do that. It is what ensures that the Master can never be in himself what he is for himself. That story will be told in Chapter 10.

Chapter 8

The Structure of Desire and Recognition

Self-Consciousness and Self-Constitution