Back to David Vessey's Homepage


On the Incompleteness of George Herbert Mead’s

Theory of the Social Self as an Account of Intersubjectivity:

Re-reading Henrich after Habermas

David Vessey

(Versions presented at SAAP ’98 and SPEP ’98)

               One of the deans of Kant scholarship, Dieter Henrich, produced two seminal essays in the late sixties on the topic of self-consciousness.[1]  In both of these essays Henrich presented a series of arguments meant to undermine a particular view of subjectivity—the Reflection Theory of self-consciousness. In brief, the Reflection Theory holds that self-consciousness is one’s self reflecting on one’s self.  On this view not only is reflection the key character of self-consciousness, but the self only emerges in the act of reflection.  That is, the self, understood as the unity of subject and object, comes to be through reflective introspection.  In these early essays, Henrich claims the view is embraced by almost all major philosophers from Descartes to Kant.  It is Fichte, according to Henrich, who first recognizes the problems created by the Reflection Theory and designs a theory of self-consciousness to avoid its dilemmas.  Why, however, introduce these themes into a paper on George Herbert Mead? 

               In the end of the first volume of the Theory of Communicative Action[2] Jürgen Habermas reviews Henrich’s arguments as well as Henrich’s own solution.  Habermas’ criticism is that Henrich could never arrive at “an alternative self-interpretation of modernity because [he] starts from the very model of self-consciousness that his opponents also take as basic.”[3]  Instead Habermas will make the standard and oft repeated Habermasian move that “the specific quality that Henrich rightly wants to bring out in the self-preservation of self-conscious subjects . . . cannot be salvaged within the framework of the philosophy of the subject.”[4]  We need to shift our basic categories to that of the “paradigm of communicative action”—a paradigm change which, three pages later on the last page of the book, Habermas will claim was “prepared by George Herbert Mead” and others.  That is to say, Habermas suggests that it is precisely Mead’s theory of the social self which leads to an account of the self which avoids any dilemmas originating with the Reflection Theory of self-consciousness.  In fact, and here we arrive at the point of this paper, it would seem that the basic structure of the argument against the Reflection Theory could be brought to bear in a slightly modified form against Mead’s theory.  Specifically, the crucial problem of the Reflection Theory—the problem of circularity where the self must pre-exist the self-relation which brings it into existence—would seem to be a problem for every theory which presents self-recognition as a key mechanism for the emergence and constitution of the self.  And Mead may have just such a theory.  

               On to George Herbert Mead.  Mead has two different philosophical aims united by a common argument.  First, Mead is concerned with showing how it could be possible for the reflective capacities of the mind to emerge in human beings as a product of evolution.[5]    Second, Mead is concerned with demonstrating the irreducible sociality of the self.  The self only becomes individuated though socializing interaction with others.  As such, the degree and character of individuation will be socially determined.  Clearly this view about the irreducible sociality of the self is independent of the claim of the emergence of the self.  For example, (if we take the power of self-reflection to be the key power for the emergence of the self) it certainly could be the case that our mental faculties of reflection evolved in such a way as to free us from our ‘social fetters’ rather then remain confined by them, and it could also certainly be the case that our mental faculties of reflection were created in us, but in such a way that their functioning is irreducibly social.  In fact, neither of these two alternatives are prima facie more implausible than Mead’s conclusion that we have emergent social selves.[6]   Of course, if Mead can show that the exact same process which results in the emergence of reflective powers also guarantees the irreducible sociality of the self, then the combination of emergence and sociality immediately follows.  The key lies in isolating, articulating and defending such a process.[7]

               Mead recognized linguistic interaction to be the medium of both the process of emergent reflexivity and the process of socialization.  We are different from all other animals in our ability to use speech, and it is through speech that we establish ourselves in relation to others.  Let us consider first Mead’s argument that the mental capacities of reflection emerge with the capacity to speak.  Language must be understood as the set of gestures which structure the expectation/response interchange of actions between two organisms.  Mead argues that there is a clear difference between linguistic (speech, writing) and non-linguistic gestures: specifically, the author of the linguistic act is distanced from the act in a way which one is not in non-linguistic gestures.  This distance reaches its extreme in writing (or recordings), but even in the case of speech we can, as we speak, listen to our voice as others might hear us.  There is an additional differentiation within speech between spoken ‘signals,’ and fully grammatically structured propositional expressions.  Examples of spoken signals extend from yelps and roars, to cries of “fire” or “attack.”   From now on I will refer to the first, pre-speech interactions as “gesture-mediated interactions”; the second stage, the first speech stage, I will refer to as “signal mediated interactions”; and I will refer to the third stage as “symbol mediated interactions.” 

               Take Mead’s favorite example of gesture mediated interaction: the dog-fight. 


Dogs approaching each other in hostile attitude carry on such a language of gestures.  They walk around each other, growling and snapping, and waiting for the opportunity to attack.  . . . The action of each dog becomes the stimulus to the other dog for his response.  There is then a relationship between these two; and as the act is responded to by the other dog, it, in turn, undergoes change.  The very fact that the dog is ready to attack another becomes a stimulus to the other dog to change his own position or his own attitude.  He has no sooner done this than the change of attitude of the second dog in turn causes the first dog to change his attitude.  We have here a conversation of gestures.[8]

Through the internalization of expectation patterns a coordination of activities emerges in which one organism, in this case a dog, responds to the other organism while at the same time anticipating the other’s response to itself.  The meaning of the gestures appears between the expectations/responses of the participants.  As such it is impossible to characterize the agency of the action, and consequently the action itself, in terms of only one organism.  The conversation of gestures is a genuinely social action.  However, at this stage there is not yet reflective awareness by the participants, only complex arrangements of stimuli, expectations, and responses.   

               When the participants are able to use vocal gestures, the interaction becomes categorically modified.  The difference derives from the ability of the participants to react to themselves as the other participant reacts to them.  As a result, the interaction need not only be an interchange of expectation/response between two participants, but may also involve an interchange of expectation/response with oneself.   Mead refers to this as taking the attitude (or stance or role) of the other.


When, however, one is making use of the vocal gesture, if we assume that one vocal element is a stimulus to a certain reply, then when the animal that makes use of that vocal gesture hears the resulting sound he will have aroused in himself at least a tendency to respond in the same way as the other animal responds.[9]

The ability to take the stance of the other towards oneself is the origin of the possibility of self-reflection. Since animals use significant symbols—and we can presume they do not have reflective abilities—, the use of significant symbols must be a necessary, but not sufficient condition for the ability to self-reflect.  What the possibility of self-reflection requires is the existence of a grammatically structured, normatively regulated language.  Only given such a developed language can social and physical situations be articulated in such a way that we can take up a perspective of possible future behaviors and reactions.  Thus, only given such a developed language can one participate in the inner conversation which constitutes thought.  Thinking, according to Mead, is a form of speaking to oneself, and speaking (meaningfully) requires such a language.  “Mind,” the object of reflection, is the internalization of the symbolic interaction gleaned from the reaction of others and ourselves to ourselves.  This internalization is strengthened through repetition.[10] Importantly it is still only through the reactions of others that our awareness of our self (as the source of the action which caused the reaction) arises.[11]

               I suggested above that the way that Mead grounds the irreducibly social nature of the self is through collapsing the process of the emergence of reflection and the process of socialization such that reflection itself is irreducibly intersubjective.  The permeating social origin and character of reflection should be clear from the argument above, we should now consider how the self arises and is individuated in the process of symbolic interaction.  The key is realizing that the self simply is that network of involvements with the generalized other. The following quotation explicitly conjoins the two issues.


The mechanism of introspection is therefore given in the social attitude which man necessarily assumes towards himself, and the mechanism of thought, insofar as thought uses symbols which are used in social intercourse, is but an inner conversation.  Now it is just this combination of the remembered self which acts and exists over against other selves with the inner response to his action which is essential to the self-conscious ego — the self in the full sense of the term — although neither phase of self-consciousness, insofar as it appears as an object of our experience is a subject.[12]

 The first source of the content of our self is the internalization of the perspectives of other towards ourselves.  When we interact symbolically with others we recognize their reactions as reactions to ourselves and from them infer ourselves as the source of their response.  As we expand our sphere of interaction we acquire a multitude of reflected ‘pictures’ of ourselves.  This purely centripetal constitution from other to self is buttressed by our reactions to hearing our own communicative expressions.  As we saw above, in such cases we hear our own expressions and have the tendency to respond to them as the other would respond to them.  This is another way of taking the attitude of the other towards ourselves, but it is no longer merely an internalization of others reactions since we are also aware of the fact that the expression to which we are reacting is our own.[13]  Consequently we are aware of two different facets of our selves, that which responds to the others and that which appears as the internalization of the perspectives of others: in  Mead’s terms, the “I” and the “me”, both of which together constitute a self.[14]  Clearly, then, through this process of symbolic interaction and assuming the role of the other a self is constituted in its individuality solely through its involvements with others.

               Let’s turn to Dieter Henrich.  Henrich makes two different arguments against the Reflection Theory—the view that the self is first constituted through self-reflection[15]—, one ontological, one epistemological. The first argument Henrich makes against this view is the obvious one; it is circular.  It attempts to explain the existence of reflection self-consciousness as a result of self-reflection. “That which is to brought into explicit consciousness through reflection must be present, at least implicitly so that it can call forth the act of reflection which is directed to it.  [I]n reflection a consciousness of the subject is presupposed; . . . the Reflection Theory can at most explain explicit experience of self, but not self-consciousness as such.”[16] In the earlier essay he adds,  “If the Subject-Self is not the Self, then neither can the Self, of which we come to have knowledge, that is, the Object-Self, ever can be identical with it.” In Fichtean terms, the I=I does not hold. Henrich takes it for granted that the I and the me must correspond in order to truly be an account of self-consciousness, but this is precisely where Henrich’s argument fails to bear against Mead’s view.  It is not that on Mead’s view the I and the me are not two facets of the same self—indeed they are—, but they are facets of the self.  More specifically, neither the Me nor the I are substances, rather they are functional terms for the way in which the self operates in its involvements with others.  It is precisely this intersubjectively derived function understanding of the self which Habermas takes as crucial for avoiding Henrich’s arguments.  In his main work on Mead, Habermas writes that Mead “leaves behind the reflection model of self-consciousness according to which the knowing subject relates to itself as an object in order to lay hold of and thereby become conscious of itself. [17]

Mead himself, claims Habermas, avoids the pitfalls of the Reflection Theory.  Yet  we have only considered Henrich’s first objection. 

               Henrich realizes that his epistemological objection to the Reflection Theory is more universal than the first claiming that “[a]ny interpretation of consciousness as the self-reference of an I can be seen to be untenable on grounds totally independent of the circularities of the Reflection Theory in its narrower sense”  He argues,


[I]n any case, regardless of how the ‘I’ comes into relation to itself, whether by an act of reflection or in some other way, the ‘I’ must grasp itself in self-consciousness.  Since the grasping of itself must specifically be a conscious apprehension, the ‘I’ must have some notion that that which it becomes acquainted in self-consciousness is itself. [18]

The process of becoming reflectively self-aware can not be the original event of acquiring self-awareness for one must be able to know when one is or is not actually reflecting on one’s self.[19] Notice that this epistemological argument—that self-consciousness can not simply be an awareness of an object, it must be an awareness of a self, which implies that there must already be some self-consciousness in place in order to direct the awareness and signal its accuracy—can be reasonably construed as a version of the learner’s paradox. In the most general terms, knowledge acquired through recognition requires a sufficient awareness of the object to be able to recognize it when one sees it. The question is whether this criticism, in its general form, applies to Mead’s theory of the self.  I believe so.

               Simply enough, it could never be the process of internalization which functions as the origin of mind for the power of internalization depends on there being already a reflexive relation to oneself.  Of course this relation need not be a conscious one, but in order for subjects to recognize themselves in another they must have implicitly differentiated their “self” not only from other things in the world, but from other selves.  To take the attitude of the other towards oneself is to recognize oneself as that towards which one is taking the attitude.  Therefore the powers of internalization can not be temporally prior to and still constitutive of the development of the powers of self-reflection.    

               Moreover, just as to internalize a perspective of another presumes an established relation to oneself, so too does it presume a relation to another person. Certainly in order to take the role of an other we must be in a position to differentiate other people from inanimate objects, —that is to differentiate those objects which have a perspective on us and those which don’t—, as well as differentiating those properties of the other person which contribute to signal their attitude. Only a small fraction of the tremendous amount of perception information we acquire in a symbolic interaction with others functions to constitute the other’s perspective and one must be aware of that distinction from the start.  For another person to appear even as a candidate for a perspective from which to reflect on ourselves necessitates that there is a pre-established relation to the other as a subject; this relation can not be established through the interactions themselves.  Thus the relation established in vocal communication could not be the originary medium of the establishment of the relation of the other, nor of the relation to oneself.

               Both of these arguments—one for a pre-reflective self-awareness and one for a pre-reflective other awareness—are derived from Henrich’s arguments against the Reflection Theory of self-consciousness. Yet it is precisely in the context of avoiding Henrich’s arguments that Habermas introduces Mead’s account of the social self.  What are we to conclude?  At the very least, Mead’s theory of the emergent, social self needs to be supplemented by an account of pre-reflective self and other awareness.   Mead fails to recognize a level of intersubjectivity which is necessarily prior to the processes internalization which on he sees as constitutive of the self.  This intersubjective medium must already be leaden with linguistically structured patterns of habituation which themselves operate as forces of individuation.  Thus the processes of individuation and socialization//habituation are more active behind the reflective mind than Mead acknowledges. Admittedly, this conclusion may not come as a surprise to Mead scholars,[20] but what should this problem mean for Habermasians?

               Habermas, after all, explicitly and repeatedly aims to avoid the aporia of reflection theories of consciousness. There are four features to Mead’s theory that attract it to Habermas.  First, the self is intersubjectively constituted. (Mead downgrades “the agency of the ego in the philosophy of consciousness into a ‘me’, into a self that first emerges in the contexts of interaction before the eyes of an alter ego—, and in doing so, [he] will shift all fundamental philosophical concepts from the basis of consciousness to that of language.”[21]) Second, this constitution occurs through language. (“The ‘me’ casts off the reifying gaze, however, as soon as the subject appears not in the role of the observer but in that of the speaker and from the social perspective of a hearer encountering him in dialogue, learns to see and to understand himself as the alter ego of another ego.”[22])  Third, this interaction establishes the equiprimordiality of individuation and socialization. (“The only promising attempt to grasp the entire significance of social individuation in concepts is, I believe, initiated in the social psychology of George Herbert Mead.  Mead makes the connection between differentiation within the structure of roles, on the one hand, and the formation of conscience and gain in autonomy by individuals who are socialized in increasingly different conditions on the other hand.”[23]) And fourth, taking the role of the other establishes a practice of self formation which at the same time develops the character traits required for moral reasoning.  (“The principle of universalization is intended to compelthe universal exchange of roles that G. H. Mead called ‘ideal role taking’ or ‘universal discourse.’”[24])

               None of these four features depend on the problematic claim that self-consciousness originates with the recognition of ourselves in the reactions of others.  Thus Habermas need not inherit the problems of Mead’s view.  Moreover, Habermas recognizes limitations with Mead’s view.   If Habermas were interested in preserving Mead’s view, he would need only to supplement the theory with a more developed account of pre-reflective self-awareness.  This account could have its inspiration in other sources—say Merleau-Ponty’s theory of the embodied self—, or it could be constructed out of some of the rare, but suggestive, claims Mead himself makes about a “biologic individual.”[25]   Then again, as Habermas’ four concerns have moved away from the problems of the origin of self-consciousness, perhaps another theory—one which takes into consideration from the start a level of embodied self-awareness—would serve his ends better.


[1] Fichtes ursprungliche Einsicht (Frankfurt: Klosterman, 1967) translated as “Fichte’s Original Insight” in Contemporary German Philosophy, Vol. 1 (Univ. Park, PA: Penn State University Press, 1982) pp. 15-53; and “Selbstbewusstsein: Kristische Einleitung in einer Theorie” in Hermeneutik und Dialectik (Tübingen: Mohr, 1970) translated as “Self-Consciousness: A Critical Introduction to a Theory” in Man and World 4 (1971) pp. 3-28.  All pages references are to the English translations.

[2] Theory of Communicative Action, Vol. I: Reason and Rationalization in Society  (Boston: Beacon Press, 1985)

[3]  TCA I, pg. 394.

[4]  TCA I,  pg. 396

[5] Such a view, if it is successful, would conjoin two paradigmatic pragmatic topi: the Darwinian model of evolution as stimulus/response interaction with one’s environment, and the ability of humans to appropriate reflectively their environment and use scientific “inquiry” to respond to and solve the problems in their environment.. Through the convergence of these projects the pragmatists exploited Darwin’s experimentalist rather than Hegel's dialectical account of history to solve some of Kant’s problems without relying on the baroque speculative claims of the post-Kantian idealists.

[6] Even though the views which together account for the emergent social self are not logically correlated, this is not to say that the two are entirely independent.  For example, if one holds the second view, that the self is irreducibly social, then this puts limits on the powers which can emerge through evolution.

[7]  The fact that Mead appeals to same process to address both the emergence of the powers of reflection and the sociality of the self tends to keep commentators from properly distinguishing the two distinct concerns.  

[8] Mind, Self and Society. ed. Charles Morris, Chicago: Univ. of Chicago Press, 1934. p. 14, 43.  Consider also the following, more general statement: “There exists thus a field of conduct even among animals below man, which in its nature may be classed as gesture.  It consists of the beginnings of those actions which call out instinctive responses from other forms.  And these beginnings of acts call out responses which lead to readjustments of acts which have been commenced, and these readjustments lead to still other beginnings of response which again call out still other readjustments.  Thus there is a conversation of gesture, a field of palaver within the social conduct of animals.” (Selected Writings ed. Andrew Reck, Chicago: Univ. of Chicago Press, 1964. p. 124)

[9] Mind Self and Society pp. 63, 65-6

[10] Mead: “In so far as one calls out the attitude in himself that one calls out in others, the response is picked out and strengthened.  That is the only basis for what we call imitation.  It is not imitation in the sense of simply doing what one sees another person doing.  The mechanism is that of an individual calling out in himself the response which he calls out in another, consequently giving greater weight to those responses than to the other responses, and gradually building up those sets of responses into a dominant whole.” (MSS pp. 63, 65-6)

[11] There is an additional process that, on the one hand, encourages the development of self-reflection, and, on the other hand, plays a central role in the constitution of the self.  Whenever the expectations are not fulfilled and there is a disruption in the normal interactive processes, the new meanings which arise as a result must themselves be taken up reflectively in order that one may, with modifications, return to the normal interactive patterns of behavior.  Such emphasis on rational reconstructions are typical of pragmatic emphases on “inquiry.”  Habermas understandably focuses much attention on this event.  He writes that “[b]y virtue of a break in the direct connection between stimulus and response, intelligent conduct arises, characterized by ‘the ability to solve the problems of present behavior in terms of its possible future consequence.’  The organism pauses and becomes aware of what it is doing when it arouses certain responses to its own gestures in another party.” (Theory of Communicative Action Vil. II: Lifeworld and System: A Critique of Functionalist Reason (Boston: Beacon Press, 1987) p. 11). Consider also the following quotation from “Individuation Through Socialization: On George Herbert Mead’s Theory of Subjectivity” (in Postmetaphysical Thinking: Philosophical Essays. tr. William Mark Hohengarten. Cambridge: MIT Press, 1992). “For Mead [individualization] depends on the internalization of the agencies that monitor behavior, which migrate, as it were from without to within.  In the process of socialization, the growing subject takes what the reference person expects of him and first makes it his own, so that thereafter he can universalize and integrate the diverse and even contradictory expectations by means of abstraction; and to the extent that this occurs, there arises an internal center for the self-steering of individually accountable conduct.” (p. 152)

[12] Selected Writings, p. 146

[13] “Any gesture by which the individual can himself be affected as others are affected, and which therefore tends to call out in him as a response as it would call out in another, will serve as a mechanism for the construction of a self.  . . . If this statement is correct the objective self of human consciousness is the merging of one’s responses with the social stimulation by which he affects himself.  The ‘me’ is a man’s reply to his own talk.” (Selected Writings, p. 140)

[14] He writes: “The ‘I’ is the response of the organism to the attitudes of the others; the ‘me’ is the organized set of attitudes of others which one himself assumes. The attitudes of the others constitute the organized ‘me’, and then the one that reacts towards them the ‘I.’” (Mind, Self and Society pg. 175.) Importantly, we are never aware of the “I” as a subject, but only as an object.

[15]  The full statement of the view is that most modern thinkers from Descartes to Kant “held [the theory] that the essence of the Self is reflection. This theory begins by assuming a subject of thinking and emphasizes that this subject stands in a constant relationship to itself.  It then goes on to assert that this relationship is a result of the subject’s making itself in its own object; in other words, the activity of representing, which is originally related to objects, is turned back upon itself and in this way produces the unique case of an identity between the activity and the result of the activity.” “Fichte’s Original Insight”, pg.  16

[16] “Self-Consciousness . . .”, pg. 11

[17]“Individuation Through Socialization: On George Herbert Mead’s Theory of Subjectivity” pp. 162-163.  Habermas continues, “As long as subjectivity is thought of as the inward space of one’s own representations, a space that is disclosed when the object-representing subject bends back, as if in a mirror, upon its activity of representation, everything subjective will be accessible only in the form of objects of self-observation or introspection—and the subject itself only as a me objectified under his gaze.  The ‘me’ casts off the reifying gaze, however, as soon as the subject appears not in the role of the observer but in that of the speaker and from the social perspective of a hearer encountering him in dialogue, learns to see and to understand himself as the alter ego of another ego. (“Individuation Through Socialization: On George Herbert Mead’s Theory of Subjectivity”, pp. 170-172).  Habermas is not alone in thinking that Mead avoids Henrich’s criticisms of the Reflection Theory of the self, see also lectures twelve and thirteen of Ernst Tugendhat’s  Self-Consciousness and Self-Determination (Cambridge: MIT Press, 1986; originally published as Selbstbewusstsein und Selbstbestimmung: Sprach-Analytische Interpretationen (Frankfurt: Surkamp, 1979)).

[18] “Self-Consciousness” pg. 11. Henrich continues, “To this end, it is not necessary that it have any sort of conceptual knowledge of itself or that it be able to give a description of itself.  But in any case, it must be able to assert with certainty that it is itself with which it becomes acquainted in self-consciousness, whether this self-acquaintance results from reflection or some other manner.” 

[19] If this sounds strange, consider how many facts about you which are available to reflection but which have no role to play in the constitution of your self.  A some point we must be aware of which facts about us count in the constitution of our self and this awareness can not be acquired first and foremost through reflection. Rather we must be always already aware of ourselves in some form or another.  What that form might be is clearly at issue here.

[20] Mitchell Aboulafia, for example, is “convinced that an indirect, prereflective consciousness of self exists, and that Mead’s model can not deal with it and it thus in need of modification” (The Mediating Self: Mead, Sartre and Self-Determination. New Haven: Yale University Press, 1986. p. 58). Aboulafia’s argument is phenomenological not transcendental.

[21] “Individuation Through Socialization: On George Herbert Mead’s Theory of Subjectivity” p. 162.

[22]“Individuation Through Socialization: On George Herbert Mead’s Theory of Subjectivity” p. 172.

[23]“Individuation Through Socialization: On George Herbert Mead’s Theory of Subjectivity” p. 151.

[24] Moral Concsiousness and Communicative Action p. 64.

[25] Sandra Rosenthal has put much weight on a few, scattered places where Mead speaks of the “biologic individual.”  See Sandra Rosenthal’s and Patrick Bougious’ Mead and Merleau-Ponty: Toward a Common Vision. Albany: State University of New York Press. 1991


Back to David Vessey's Homepage