At the center of Jurgen Habermas’s account of the development of self-iden­tity is the capacity to question and criticize conventions. This requires a capac­ity to conceptually abstract from given contexts through an appeal to principles. And this capacity is learned through the internalization of social and linguistic norms. For Habermas, the development of self-identity is pred­icated on the development of moral identity.

Habermas draws on the work of George Herbert Mead to articulate a theory of the development of self-identity through linguistically mediated interaction. For Mead, “The fact that all selves are constituted by the social process… is not in the least incompatible with, or destructive of, the fact that every individual self has its own peculiar individuality.”14 In fact, it is only through “the social process” that a “peculiar individuality” can be realized. From Mead, Habermas draws a description of how socialization, the learn­ing and internalizing of linguistic and social norms, produces not simple conformity to those norms, but true individuation. I shall take up this descrip­tion at the ontogenetic level—at the level of individual development.

The crux of the interpretation Habermas draws from Mead, and what takes it beyond Freud, and beyond the thesis of the Dialectic of Enlightenment, is the concept of the perspective of the “generalized other”: the idea that through linguistic interactions a human child develops an under­standing of social norms not simply as expressions of arbitrary choice or self – interest, but as subject to demands for and tests of validity.15

For Mead, the child’s self-identity, in both its epistemic and its practical dimensions, develops through the mechanism of taking the perspective of another, who, in an interactive relationship, takes up a performative attitude toward the self.

In the first stages of taking the attitude of the other, the child understands its interaction with its parents as a reciprocal satisfaction of interests expressed as imperatives. Once the child comes to recognize that the parents’ expression of imperatives is connected with the parent’s provision of care, the child responds to the threat of withdrawal of care by internalizing the parents’ attitude toward herself.16 At first the internalized attitudes are still tied to the concrete roles of a particular self and other. The attitudes or roles become detached from specific persons, and the transition to the perspective of the generalized other begins, with the introduction of a third-person or objective perspective. Thus far, the description of development applies as much to Freud as to Mead.

Freud and Mead realized that these patterns of behavior become detached from the context-bound intentions and speech acts of individual persons, and take on the external shape of social norms insofar as the sanctions connected with them are internalized through taking the attitude of the other, that is to say, to the degree that they are taken into the personality and thereby rendered independent of the sanctioning power of concrete reference persons.17

At first, the objective or generalized norm of action is understood only in terms of an imperative which rests on choice—in this case, on the generalized choice, or arbitrary will, of the group. It is only with the transition to what Mead calls the perspective of the generalized other—and here we move beyond Freud—that the child comes to understand group norms not in terms of arbitrary will or choice and self-interest, but in terms of claims to validity: in terms of mutual obligations and expectations. The mechanism of internal­ization is essential to the development of this understanding.

The authority of the “generalized other” differs from authority based only on disposition over means of sanction, in that it rests on assent. When A regards the group sanctions as his own, as sanctions he directs at himself, he has to presuppose his assent to the norm whose violation he punishes in this way. Unlike socially generalized imperatives, institutions claim a validity that rests on intersubjective recognition, on the consent of those affected by it: “Over against the protection of our lives or property, we assume the attitude of assent of all members in the community. We take the role of what may be called ‘the generalized other.’”18

In other words, through internalizing the attitude of the generalized other, the child comes to recognize “objective” social norms as her own. Thus, I come to recognize or constitute myself as a member or participant in the “we.” This connection between the self and the “we” is internalized, both in the sense of the development of understanding, and in the sense of the setting up of a motivational structure. In learning to orient her behavior to norms and take part in normatively regulated interactions, the child understands her self as a participant in a “we,” and comprehend the meaning of a valid norm—i. e., that it rests, ideally, on the assent of each and all and as such is subject to criticism on the grounds that there are good reasons for my not assenting. The child makes a transition from a motivational structure based on interests and imperatives to a motivational structure based on an orienta­tion to validity claims—on a recognition of shared expectations and obliga­tions. Thus, internalization is the mechanism of both a comprehension of the meaning of social norms as, in principle, valid norms and an anchoring of those norms in a motivational structure.

It must be stressed that this account of the development of self-identity only works if we stipulate that the generalized other represents not an actual community consensus, but an ideal, a standard against which any norm must be measured. Contrary to what Habermas, and Mead, often seem to suggest, there is in modern societies no given social world of conventional norms that the child simply takes over and internalize. Rather, a child internalizes many different and conflicting normative positions. In order to participate effec­tively in interactions with primary others, and with institutions, a child may take the attitude of her mother’s appeal to an “ethic of care,” her father’s appeal to an ethic of rights, her grandfather’s appeal to an ethic of particular traditional conventions, and the ethic of strategic rationality in the pursuit of self-interest upheld by her favorite TV show. (More likely, each of her refer­ence persons will represent a mixture of these attitudes and others.) She will have to deal with the differences between her mother who thinks homosexu­ality is sick and her aunt who’s a lesbian and a gay rights activist, a black teacher who teaches the principles of universal equality and black pride and white friends who are racists. There is no actual given perspective of a unified conventional “we” which is the generalized other. What is internalized is a capacity to appeal to principles, to standards of normative validity. The child is forced to individuate through taking positions with respect to given conflicts. Ideally, she learns to do so through abstracting from her particular loyalties to each of her different reference persons, to appeal to principles.19

The crucial point here is that for Habermas, Mead’s concept of taking the attitude of the other, which Mead understands as calling forth a response in oneself that one also calls out in the other, cannot be understood on the model of a simple response to a stimulus, but must be understood through a model of linguistic interaction. The response, can be understood “in the full dialogical

sense as an ‘answer’”; hence, what is internalized is not simply assent, taking the attitude of the other demands “internalizing yes/no responses to statements or imperatives.”20 Habermas takes this idea from Tugendhat, who notes that “consent only has significance against the background of the possibility of refusal.”21 What is internalized is a capacity for critique.

For Habermas, the capacity for critique grounded in linguistic communi­cation underlies the development of both moral identity and more general personal identity. In analyzing the structure of role perspectives inherent in linguistic communication, Habermas follows Mead, who systematically connects “the role-taking effective in socialization with the speech situations in which speakers and hearers enter into interpersonal situations as members of a social group.”22 One takes oyer a reflected sense of self or “me” by adopting the other’s perspective or expectations toward oneself. But the struc­ture of linguistic interaction is such that the other, in recognizing you as a participant in interaction, expects you to take a position in response to her speech act. Thus, as a participant in the interaction, one must accept the free­dom and responsibility of taking a position in response to the other’s speech act. It is this freedom—and this responsibility—demanded of participants in linguistically mediated social interaction, which is the source of individuation.

The performative attitude that ego and alter adopt when they act commu­nicatively with one another is bound up with the presupposition that the other can take a “yes” or “no” position on the offer contained in one’s own speech act. Ego cannot relinquish this scope for freedom even when he is, so to speak, obeying social roles; for the linguistic structure of a rela­tion between responsible actors is built into the internalized pattern of behavior itself.23

The demand that one “take a position” is not restricted to questions of moral principle and the justification of norms; this is a demand found in any commu­nicative interaction. In my response to another’s speech act I am necessarily taking a position. Thus, individuation is an ongoing process: the development of a sense of self-identity (of a “me”) takes place through a continual process of reflection on and assimilation of the actual positions taken, in practice, in my responses to other’s speech acts. The element of spontaneity and unpredictabil­ity is introduced by the “I” which acts, in response to the acts of others.

Thus in the socialization process an “I” emerges equiprimordially with the “me,” and the individuating effect of socialization processes results from this double structure. The model for the relation between the two agencies is the “answer” of a participant in communication who takes a “yes” or “no” position. Which answer ego will give in any instance, what position he will take, cannot be known in advance—either by him or by anyone else.24

For Habermas the development of self-identity is a response to the demands inherent in the structure of linguistic communication.

The individuation effected by the linguistically mediated process of social­ization is explained by the linguistic medium itself. It belongs to the logic of the use of the personal pronouns, and especially to the perspective of a speaker who orients himself to a second person, that this speaker cannot in actu rid himself of his irreplaceability, cannot take refuge in the anonymity of a third person, but must lay claim to recognition as an indi­viduated being

Among the universal and unavoidable presuppositions of action oriented to reaching understanding is the presupposition that the speaker qua actor lays claim to recognition both as an autonomous will and as an individuated being.25

In linguistic communication, the speaker is required to recognize and take responsibility for herself as a “me,” and to take positions in response to others as an “I.” The demand made of participants in linguistically mediated inter­action, that they accept the freedom and responsibility of taking an affirma­tive or negative position in response to an other’s speech act, is what underlies the development of moral identity as a critical relation to social norms, and of personal identity as a critical relation to oneself.

Moral identity is based on the recognition “that a norm deserves to be valid only insofar as… it takes into account the interests of everyone involved, and only insofar as it embodies the will that all could form in common, each in his own interest, as the will of the generalized other.”26 It is this orientation to the validity of a norm that provides the individual with the basis for critique: if anyone’s interests are being excluded, then the norm is not valid. If anyone presents reasons for not consenting, then the validity of the norm is called into question. And this is what makes it possible for the individual to abstract from particular norms to universal principles, to move from simple conformity to or deviance from given norms to a capacity to rela- tivize and criticize given norms in the name of universal principles, but to do so as a participant in a social world.

For Habermas, this capacity for critique which is built into linguistic inter­action, and built into the internalization of norms, is essential to not only moral but more general personal identity.

The identity of the ego can… be stabilized only through the abstract abil­ity to satisfy the requirements of consistency, and thereby the conditions of recognition, in the face of incompatible role expectations and in pass­ing through a succession of contradictory role systems. The ego-identity of the adult proves its worth in the ability to build up new identities from

shattered or superseded identities, and to integrate them with old identi­ties in such a way that the fabric of one’s interactions is organized into the unity of a life history that is both unmistakable and accountable.27

The self-identity of the adult depends on the ability to “take over and be responsible” for integrating all of the different, often conflicting, positions one takes, into a narrative that is meaningful to others and to oneself. This requires a cognitive ability to resolve conflicts among particular positions by abstracting to more complex meanings, and by reflecting on practices in a process of self-critique: by asking the questions, what kind of person am I, and is this the person I want to be? For Habermas, this cognitive process of self-evaluation calls for an “appropriative form of understanding.”28

But if the capacity for critique is essential, it is not a sufficient condition of the development of a meaningful and recognizable self-identity. There also has to be an existential commitment to the meanings you produce through your practices, and through which you critically judge and guide your practices.

What Habermas’s developmental model doesn’t answer is how we come to be able to make that commitment to a recognizable, integrated, and mean­ingful self-identity and keep it relatively open, flexible, and nondefensive. It is unable to account for varying levels of identity-competence; to account for why many—probably most—of us fail to successfully develop coherent and meaningful self-identities and typically err either on the side of rigidity and defensiveness—a failure to question and criticize—or on the side of mushy indistinctness—a failure to abstract from particulars and resolve contradic­tions. Nor can it account for why this failure often takes the form of a spir­ited resistance to identity, abstraction, resolution, and integration. To fill in this gap, I draw on the work of Julia Kristeva.


Kristeva’s work is characterized by a profound ambivalence with regard to the nature of society, and hence of language and individuation. There are two different models in Kristeva’s work of the development of identity through the internalization of linguistic and social norms. On one hand, Kristeva could be described as a Derridean poststructuralist with a stoic individualist twist. In this guise, Kristeva sees the “sociosymbolic order” as a closed structure that is essentially repressive and essentially patriarchal. To this extent she agrees with Derrideans, but she differs from them insofar as she insists that the struc­tures of language and individuation are essential for human social interaction and participation. Thus, the only solution is to stoically accept the closed, repressive, patriarchal order of language as the only means of participation in social interaction, while at the same time constantly subverting it. Self-iden-

tity is a constant oscillation between stoic acceptance and subversion of the Law:

A constant alternation between time and its ‘truth’, identity and its loss, history and that which produces it: that which remains extra-phenomenal, outside the sign, beyond time. An impossible dialectic of two terms, a permanent alternation: never the one without the other.29

But there is another track in Kristeva’s work: in many of her writings, Kristeva understands individuation in terms of a not-impossible dialectic between system and practice, as a constant process of investment in and inter­nalization of a language system that is constantly transformed through indi­vidual and social practices. Kristeva’s “subject-in-process” is a subject who develops and changes through taking up positions, or identities, through an investment in a sociosymbolic order and thereby realizing and expressing her own heterogeneity (and in turn transforming language and society).

Against Derrida’s invocation of a constant negativity in resistance to any identity, Kristeva argues that the refusal of identity renders negativity merely positive, leaving us in a space in which difference no longer exists. In its relentless subversion of Identity, Derrida’s “trace” “marks anteriority to every entity and thus to every position:… the trace dissolves every thesis—mater­ial, natural, social, substantial, and logical—in order to free itself from any dependence on the Logos.”30 The effect of this resistance position, to any thesis, or identity is a theory which “gives up on the subject, and must remain ignorant [of the subject’s] functioning as social practice… .”31 For Kristeva, the subject constitutes itself only through positing, through taking positions or identities within a social world and a symbolic order—only by engaging in a world of shared or identical meanings, through which one can realize one’s own meaning.

Like Habermas, Kristeva focuses on the need to take positions in everyday social interactions as central to the constitution of self-identity. Both Habermas and Kristeva argue that the capacity to take positions requires the development of a capacity to identify oneself with a social “we,” in a shared symbolic meaning-system. But whereas for Habermas the need to take a posi­tion means the need to relate to norms in a critical and questioning manner, for Kristeva taking a position tends to mean taking a position of identity within the symbolic order, which will allow “nonidentity” or difference to emerge—to be realized or expressed—thereby producing a new position. Whereas for Habermas the pressure to individuate inheres in the freedom and responsibility to take an affirmative or negative position in response to others’ speech acts, Kristeva analyzes the pressure to individuate in terms of the tension between unconscious drives and socio-linguistic systems, and analyzes the positing of identity as a condition for the expression or articulation of

desires—and hence, the realization of a self and its meaning—in language. Kristeva provides no mechanism for moving from a conventional to a post – conventional, critical, moral identity, because she does not recognize a connection between identity of meaning and intersubjective validity. What she does offer is an analysis of the constitution of identity in terms of the expres­sion or articulation of a self, through the expression of bodily heterogeneity and of affects, in language which is meaningful to oneself and to others. She provides an account of the development of a practical self-identity—of an identity which is postconventional insofar as it is an ability to relate mean­ingfully to differences within oneself and between self and others. At her best, Kristeva understands the tension between drives and language—both, forces beyond the individual’s control—not as barriers to, but, as Axel Honneth has put it, as “enabling conditions” of the development of self-identity.32 For the development of individual identity, or individual meaning, is only possible through the expression or realization of one’s specificity in language—in a system of shared meanings, through interaction with others.

Kristeva’s psychoanalytic analysis of the development of identity in terms of expression, rather than repression, produces some surprising results: against Lacan and Freud, who tend to argue that learning symbolic and social systems entails the repression of aggressive and pleasure-seeking drives, Kristeva argues that the development of a capacity for signification, or linguis­tic competence, emerges out of a deployment or expression of those drives. In the Freudian-Lacanian scenario:

The symbolic function is… dissociated from all pleasure, made to oppose it, and is set up as the paternal place, the place of the superego. According to this view, the only way to react against the consequences of repression imposed by the compulsion of the pleasure principle is to renounce plea­sure through symbolization by setting up the sign through the absence of the object, which is expelled and forever lost.

What this interpretation seems to rule out is the pleasure underlying the symbolic function 33

For Kristeva, the move into language and a social world—into linguisti­cally mediated interaction—is not a fall, not a renunciation of pleasure. Rather it is “a separation which is not a lack but a discharge and which… arouses pleasure.”34 The pleasure Kristeva describes here is associated with what she calls abjection: an aggressive drive (corresponding to Freud’s anal drive) for expulsion, destruction, separation, which underlies the rejection of dependence on the power of others, the separation of self and other, and the distinction between subject and predicate in language. Abjection interacts with the (oral) drive for incorporation: a drive toward both having and being, possession of and identification with others, and investment in language.

Kristeva describes the development of a capacity for signification—for linguis­tic competence and an orientation to meaning—in terms of the pleasures of abjection and incorporation.35 And it is these experiences of pleasure which, for Kristeva, motivate the internalization of symbolic and social norms.

Thus, whereas Habermas argues that norms are initially internalized only in response to a threat and initially represent only the arbitrary dictates of authority, Kristeva argues that the development of linguistic competence and the development of self-identity through the internalization of sociosymbolic norms is a pleasurable process.36 Moreover, she argues that these processes represent for the child a deliverance from utter dependence, from helplessness in the face of authority, to a means of signification—a means of participation in a social world through an orientation to meaning.

While Kristeva often advocates the “constant alternation” between iden­tity and nonidentity as the “impossible dialectic” of the self, in many of her writings Kristeva upholds a normative ideal of an integrated self—a complex self-identity based on a reflexive and affective recognition and acceptance of the difference and nonidentity within the self. It is only through the cognitive recognition and affective acceptance of the complexity or internal differenti­ation of the other that the child comes to recognize and accept both the sepa­rateness of her self from others with whom she interacts, and the internal differentiation of her own self.

The developmental condition for this recognition and acceptance is the transition from prelinguistic or drive-based, to linguistically mediated inter­action with others, and hence, the opening up of a social world. Central to this process is the internalization of social and linguistic norms. This inter­nalization is mediated by experiences of pleasure, by affective relations, and by cognitive development.

According to Kristeva, the child moves from a relation to the primary care­taker based on satisfaction of needs to a relation based on a shared orienta­tion to meaning. Too often, Kristeva describes this path of development in the terms of Lacanian psychoanalysis, as a transition from a relationship of immediacy or merging with the mother to an acceptance of the paternal Law of the symbolic order. Kristeva’s twist to this story is that the investment in the phallic symbolic order is mediated by an identification with an idealized father and that the motivation for investment in the symbolic order is the recovery of the mother (the primary object) in language, and meaning.37

But Kristeva also tells the story in another way. In this other version, the child moves from a need-based relationship with the first caretaker (who is, typically, the mother) to a recognition that the mother’s needs are not wholly satisfied by the child, that she has an other meaning in her life beyond the child. The child has to recognize, in some rudimentary way, that the mother is complex and internally differentiated. And the child is forced to recognize

this through the mother’s failure to satisfy all of her demands. In the process she learns that there is a realm of meaning that can satisfy desires, and she is able to identify with the mother’s desire for and investment in that meaning. It is this identification with the mother’s desire—with the mother’s means of participation in a social world, with her investment in a shared social mean­ing—which mediates, ideally, the child’s internalization of linguistic and social norms.38

The motivation for this internalization is not simply the threat of punish­ment but the promise of the gift of meaning—of a means for mediating rela­tionships with primary others and for participating in a larger social world where desires can be satisfied.

It is this affective investment in social meaning which underlies the capac­ity to affectively accept one’s own differentiation from others and the differ­entiation within oneself. It also underlies the capacity to develop an integrated sense of self, which will not dissolve into differences. One is able to realize one’s self through expression of one’s “nonidentity” in terms of shared mean­ings and this expression is mediated through affective investment in discourse with others.

Kristeva argues that the transference relationship—i. e., the relationship of identification with a loved other—

is a true process of self-organization. This means that once the accidents, aggressions, and errors of my discourse (of my life), have been inserted into the transference dynamic, they are no longer those failures of a final – istic linear process that anguished me before. To the contrary, in transfer­ence love, “errors” are overcompensated; they produce the libidinal self-organization that has the effect of making me more complex and autonomous. Why? Because, as they are introduced by means of discourse into transference (into love…), the death drive, or the “negative” in Freud’s sense of the term, enters the service of symbolic apprenticeship,

autonomization, and greater complexity of the individual.39


The failure to develop an integrated sense of self-identity, is characterized by Kristeva as the development of linguistic competence through the learning of linguistic norms in the relative absence of any affective investment in those norms. One is able to address oneself to a universal other, but not to a partic­ular other. Or, as Kristeva puts it, you get the “kit of representation but with­out the caboodle of drive. The caboodle remained in the emptiness of maternal fusion and/or maternal absence.”40 This happens when social and linguistic norms are experienced as primarily repressive and are only adopted in response to threat. And this happens in a social world in which too many given norms are oppressive—where, in particular, mothers, or primary care­givers, too often do not fully experience themselves as participants.

In this case, meaning comes to be understood only on the level of the universal and is nonparticularized. To affectively invest in general or univer­sal meanings one needs to be able to make meaning for another—for, in particular, a loved other. The use of language without an addressee of discourse—without someone who is spoken to—leaves it empty of significa­tion for the speaker.41

The strength of Kristeva’s account, for feminism, is her insistence that the affective relationship cannot serve as an end in itself, as a means of produc­ing individual or particular meanings. Rather, it serves as a means of invest­ing in a world of shared meanings, of constituting and experiencing oneself as a participant in that world and of making those meanings constantly open to diversity and change.

The account of the development of self-identity that I presented here is, of course, far from complete. I show only that any understanding of the devel­opment of self-identity must take both of these constitutive elements into account: both the capacity to relate to norms in a critical way and the capac­ity to express or realize one’s own meaning through affective investment in discourse with others. Both of these capacities entail the internalization of social and linguistic norms, an internalization of shared or identical meanings through relationships of identification with others. Both provide a way to unlock the paradoxes of the self: of individual identity versus social identity, of drives versus language, autonomy versus relationship to others.

In closing, I call attention to some unresolved problems in Habermas’s and Kristeva’s theories, which I have been unable to deal with adequately here. Both Habermas and Kristeva tend to conflate the learning of social norms with the learning of linguistic norms: for Kristeva, the two are conceptually conflated in the concept of a sociosymbolic order; Habermas differentiates conceptually between them, but the differentiation is not always evident in his analysis. Perhaps that’s why neither provides much of an understanding of the ways in which particular, as opposed to general or universal collective identities and affiliations, influence and interact with the development of self – identity. I tried to redress the latter omission to some extent in this paper; the former problem proves more resistant.