Copyright Sociological Research Online, 1999


Stanley Raffel (1999) 'Revisiting Role Theory: Roles and the Problem of the Self'
Sociological Research Online, vol. 4, no. 2, <>

To cite articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 27/01/99      Accepted: 07/06/99      Published: 30/6/99


This paper looks at some of the major texts in the history of role theory. The question that is asked is whether any of these works have been able to theorize the self adequately. It is suggested that neither Parsons nor Merton has any place for the self in their respective theories. While Goffman does make a space for the self, it is only a negative space. Even ethnomethodological theory cannot imagine a role player capable of self-expression. It is argued that a solution to the problem of how to conceive of self and role can be developed from some ideas present in the work of the philosopher Lawrence Blum. The concept of self as identity that can be extracted from his work can allow social theory to imagine actors who are simultaneously expressing their selves and fulfilling their roles. Affinities between this idea and some key concepts in theories of both (Alan) Blum and Peter McHugh and Charles Taylor are suggested.

Identity; Moral Frameworks; Principles; Reflection; Role; Role-conflict; Role-distance; Self

The Problem of Self

Can people be themselves? Logically, this question sounds absurd because, if the idea of self references what one is, how can one be other than what one is? Yet, in the real world of persons' experience, descriptions of what seems illogical abound. We speak of 'not feeling ourselves today.' If this expression often just refers to physical discrepancies from our 'true' nature, other versions of this phenomenon surely make reference to the realm of human action. There is the everyday phoniness from which, if J. D. Salinger's Catcher in the Rye is to be believed, no one is exempt. There is moral hypocrisy. There are even outright impostors. Inauthentic selves are a real possibility. Being oneself can be a problem for actual people.

What is the source of this problem? Why is it difficult, some would say even impossible, to be ourselves in all situations? The sociological answer carries with it an additional paradox. Probably no sociologist believes that selfhood is even possible apart from society, society being necessary to both make and sustain us as what we 'are' and yet, at the same time, most versions of what hinders or even makes impossible the expression of real selfhood lay the blame in one way or another on society. In the most naive, totally unsociological installment of this argument, the self is imagined as what was 'originally' there 'before' society. With this way of thinking clearly our real self will start to be destroyed in so far as there is any influence whatsoever from anything 'outside' us.

While the more sociological views all share an awareness that, for one reason or another, the self actually needs society to be what it is, there still coexists with this recognition a sense of how problematic it can be fully to be oneself while meeting the demands of social life. For example, in Marxism there is the sociologically convincing idea that a person is not a full self originally because we reach self-realization only when/if we are productive. At the same time there is the caveat that if, as is almost always the case at least in capitalist society, we are forced to produce under conditions of alienated labour, it is certainly not the case that what we produce is really us. Or, while functionalism, at least in its more sophisticated variants such as the ones discussed below, sees society in the form of a stable social structure and shared cultural norms as necessary for, even constitutive of personality (self?), there always seems to be a concomitant assumption that the self must cease to be wholly itself, e. g. it is forced into compromises, it must replace what it really desires with what it admits it does after all need, if an overall order is to survive. Interactionism exhibits similar ambivalence. On the one hand there is the eminently sociological insight that there is only really selfhood via the influence of society in the form of other people. At the same time, many if not all interactionists tend to think of the encounter with others as necessarily or at least frequently requiring something like the hiding of aspects of the 'real' us.

All of these theoretical understandings as to the difficulty of maintaining selfhood in social life receive support in ordinary persons' understandings. Thus many certainly feel a sense of alienation from their work products just as Marx would expect, a conviction that often the demands of the social structure and/or the culture require abandoning what they (their selves?) really want, as functionalists would predict, and a sense that it is difficult to really be themselves in public (society?) which would not surprise interactionists.

This paper is intended as a contribution to the discussion at both the theoretical and practical level of how there can after all be something like a self in society Both to make the paper manageable and to reflect the concerns of much of the literature to be considered, it has been decided to narrow the perspective to the specific issue of the integration of self and work. There are obviously many other ways in which 'society' impinges on the individual besides via work tasks, but equally clearly the world of work must be one site where the average adult individual cannot but feel the call, the requirements of what any sociologist would see as societal forces.

So we want to ask whether it is ever possible to be oneself at work. We will eventually argue that, far from this being impossible, according to a depiction of the idea of self which we shall defend, some persons selves are fully on show in their approach to work. We shall begin by looking at how this question has been theorized in previous sociological literature. As the concept of role has been central here, we shall need to assess various versions of role theory, focusing on the issue of whether these theories have made an adequate place for the self. Having found the standard theories disappointing in this respect, we will look elsewhere for help with the problem of the self-role relation. Key to the solution proposed will be exploring and re analyzing various work that has radically reconceived the very idea of a self.

Functionalism and Conflict Theory

Parsons, of course, has bequeathed to sociology the problem of order. This can be expressed in terms of the concerns of this paper as the problem of how selves can be integrated with the needs of overall society, how selves and an orderly social structure can simultaneously be possible. Parsons' attempts to solve this problem are detailed in his two major works, The Structure of Social Action and The Social System (Parsons, 1937, 1951). It is only in the latter book that the concept of interest to us, role, makes an appearance. A role is there defined as 'what the actor does in his relations with others seen in the context of its functional significance for the social system' (Parsons, 1951, p. 25). The key example of what he is clearly thinking of here would be one's occupational 'role'. In so far as a person successfully takes up an occupational role, he or she is clearly meeting certain demands or expectations of the social structure but at the same time he or she must be actively doing something. As Parsons puts it, 'he is acting...this is what we mean by his playing a role'(Parsons, 1951, p. 25). The idea of role does seem, then, to be a solution of sorts to the problem of integrating self and society.

However, a difficulty is apparent in this orthodox Parsonian way of attempting to resolve the social order problem. In that what one does in a role is basically to fulfill shared expectations, expectations that by definition come from others, can it really be said that the role player has much or even anything of a self left? Even though the Parsonian self is active, in that his or her activity is totally a matter of complying with expectations, doesn't the orthodox Parsonian solution, then, really solve the problem of order at the cost of robbing actors of selves? It would seem to demand just mechanical compliance with expectations as one's actual social task, at least in so far as one wants to have any positive functional significance for society. Mechanical compliance hardly seems to leave much possibility for being one's self. Assume that, in good Parsonian fashion, one is conforming to one's work role. While one would certainly in so doing be contributing to social order, how is one also expressing one's self? It seems more a case of going along with what others or the whole society expect, quite irrespective of what one actually wants. Although Parsons, with his notion of 'internalization' of expectations could counter that what normally happens in practice is that social expectations actually become one's personal desires or at least one's perceived duty, this scenario is more suggestive of someone who has lost all sense of self rather than someone who is expressing themselves. Arguably, many of the other theorists who have worked in this area have been aware of this weakness in the Parsonian approach, the absence of real selfhood in his theory, and have attempted to rectify this deficiency.

Firstly we can consider Merton's contribution in this light. Though it cannot be said that Merton's overall theory is as comprehensive or profound as Parsons', to his credit he has always shown a livelier awareness than Parsons of the ways in which the social structure may produce not just order but disorder, conflict, dysfunction. With regard to a person's role, it is consistent with his differing approach that he recognizes the possibility that there can be 'processes tending to make for disturbance or disruption..., creating conditions of structural instability' (Merton, 1969, p. 424). His key concept for explaining how this instability is possible is his idea of role conflict. He begins by pointing to the fact that in virtually any imaginable occupational role, one is going to have more than one role partner. For example, the school teacher will have to deal with the structured expectations not just of pupils but also of fellow teachers, the school head, parents, the superintendent of schools and also, perhaps, local community groups. He calls all these differing groups the members of a role-set (Merton, 1969, p. 423). The significance that Merton derives from the existence of role-sets is that following a role is never likely to be the clear-cut matter of compliance that Parsons implies because there are likely to be conflicting role expectations from one's different role partners. For example, following the teacher role cannot just be analyzed, as it would be by Parsons, as a matter of doing what is expected because, according to Merton, different and even contradictory things are going to be expected by, for example, one's immediate colleagues and the superintendent of schools.

However, while the existence of role-conflict clearly makes room for some variation in how various persons will fulfil their roles, does role-conflict make room for what we might really want to call the emergence of self? We can try to answer this question by examining the typical mechanisms by which, according to Merton, role conflict is normally resolved (Merton, 1969, pp. 425-33). Examples are that one works out which of the various conflicting role-partners happens to be maintaining his/her expectations with the greatest intensity and therefore conforms to the expectations of that partner. Alternatively, one analyzes which of the conflicting role partners has the greatest power and decides it would be wise to conform to the demands of that partner.

As is clear from these (typical) examples the resolution of the conflict always seems to be down to some structural factor external to the self. Furthermore, as what the actor is doing at the point that the conflict is resolved is always going along with some group's expectations, Merton's thesis would still not be allowing much or indeed any place for self in what one does in following roles. It is as if the very problem of role-conflict really presupposes an actor who already has abandoned his/her self to society and so only faces a residual problem: to which of the various faces of that society is it most expedient to conform?

Dahrendorf's classic piece, Homo Sociologicus is worth mentioning at this point. Firstly Dahrendorf deserves some credit for understanding the potential theoretical importance of the concept of role. He realizes that it is sociology's attempt to theorize where and how the self and society can intersect (Dahrendorf, 1973, p. 5. See also Gerth and Mills, 1954, esp. p. 23). He also perceives how what has tended to prove troublesome about the idea of persons as role-followers is that this view tends to rob persons of anything like a self (Dahrendorf, 1973, p. 26). However, although he does go on to offer some interesting suggestions, anticipating ethnomethodology to an extent, as to how following role expectations does not totally rule out individuality (for example: ' one of Herr Schmidt's roles can prescribe his entire behavior in the corresponding social position. There is a range in which the individual is free to behave as he chooses.' (Dahrendorf, 1973, p. 25), it is fair to say that his main conclusion is that the very fact that we adopt a sociological perspective, the idea of role being for him at the heart of such a perspective, dooms us to never being able to theorize the self adequately. For Dahrendorf, just as say economics is a way of thinking, a mode of abstraction that will perhaps never allow us adequately to understand irrational behavior, so sociology is a perspective that via its very emphasis on social shaping in the form of roles and related ideas, will never allow us to theorize the self adequately. Here, in a way, Dahrendorf's approach could be less helpful than Parsons'. Whereas the latter at least attempts to theorize a sociologically viable self via his notion of internalization, the former seems to think the idea of selfhood is an anathema to life in society even though he recognizes that this dooms sociology to nothing more than a partial version of reality.

This seems defeatist and this paper will certainly seek to resist Dahrendorf on this point. However a more immediate concern would be that even at the time he published his views in revised form (l973), what he says would seem unjust as a characterization of sociology as a whole because it takes no account of the major alternative traditions to both structural functionalism and his own conflict theory. We therefore turn now to the version of arguments on the self-role relation that can be extracted from Goffman, from ethnomethodologists and, finally, from orthodox symbolic interactionism as represented by Mead.


It is perhaps not surprising to discover that two theorists who are, after all, closely identified with structural functionalism, have difficulty in making much place for the self in their role theories. However, let us turn now to the work of someone who we might expect to offer more in this regard, Erving Goffman. Goffman, of course, has written about the appearance of self in society in many places, most notably in his first major book, The Presentation of Self in Everyday Life (Goffman, 1959). But, as our problem in this paper is not the sheer presentation of self but the possible coexistence (or not) of self and social roles, the most directly relevant work is Goffman's extended essay, Role Distance (Goffman, 1969). Goffman has, famously, developed his notion from some casual observations of various ages of children on a merry-go-round. At the age of two or three, if they are willing to go on the ride at all, the children are visibly fearful and perhaps inadequate role performers in the physical sense of not even being able to remain on the horse. At the age of four or five, this changes. At this age they do, we could say, adequately play the role in that they are intently and competently involved in the activity, say carefully holding the reins or keeping their feet in the stirrups. But, as those familiar with his ideas will appreciate, it is really what happens next, at around the age seven or eight, that really interests Goffman. While children of this age may still be willing to go on the ride, they are also very keen to show that they are not fully involved in the activity. In one way or another, they strive to show their distance from the role. For example, they might whip and drive in a parody of a jockey, jokingly pretend to fall off, or ostentatiously ride their horse backward (Goffman, 1969, p. 64). This of course is his idea of 'role-distance'.

Goffman goes on to demonstrate, basing his conclusions on his observations of a hospital operating theater, that this concept can be very useful in depicting how adult actors comply with at least some of their occupational roles. So he suggests that as they perform their roles, medical staff do not just play them, they also show distance from them. To mention a few examples, when a nurse and surgeon confess that they may be running out of sutures, an intern, the role-distant member in this case, suggests, humorously: 'We can finish up with scotch tape' (Goffman, 1969, p. 71). Or when a chief surgeon is accidentally stabbed in the finger by his assistant, he jokes: 'If I get syphilis, I'll know where I got it from and I'll have witnesses' (Goffman, 1969, p. 75).

Now it cannot be denied that Goffman has noticed a phenomenon that undoubtedly occurs. In so doing, clearly he has managed to suggest a way in which there can be both something like selves and something like roles in the social world. I would say that Goffman has made real progress over the conceptualizing of role-behavior of both Parsons and Merton in that he shows us how it is possible for a person to both perform or play a role and maintain something like a self. But notice that, even in Goffman, there is a sense in which it can be said that the same problem remains. The self does make an appearance in Goffman, but its appearance seems to take only an oddly negative form. Thus, in a series of extremely interesting articles on Goffman's concept of the self, Travers has pointed to the paradoxical 'unknowable' character of Goffman's self or, similarly that self in Goffman always seems to be characterized by its 'non-appearance as a definable entity' (Travers, 1994a, p. l30, p. l32. See also Travers 1994b and 1995). It still seems, then, even in Goffman's work, impossible to show or be oneself in a role in that the self only appears in the form of the extent of one's distance from the role. So, the self appears only in behavior which says what one is not or not just, in these examples in the statement, via role distance, that one is not just a doctor. We would argue that, in Goffman, one's role can never be a vehicle by which one's self can be positively expressed. There can only be self in distance from one's role. It can be concluded, then, that in spite of all his significant differences from the structural-functionalists or conflict theorists, Goffman has in common with them the inability to imagine the active, involved (and so not distant) role player as someone with a visible, positive self.


The place where we shall look next for a possible solution to our problem is in the body of work known collectively as ethnomethodology. Even more than Goffman, ethnomethodologists are credited with making a place for the active, interpreting side of social behavior so there are excellent grounds for anticipating that their input will be helpful for this issue. As a typical example of the ethnomethodological approach, I have selected Zimmerman's insightful study of the receptionist role in a large social work agency (Zimmerman, 1970). Zimmerman has looked at the reality of role-following here and he does indeed discover that there is room and even the need for much active, interpretive work in order to produce adequate role performance.

Just three examples of his findings will hopefully be sufficient to suggest both the thrust and the limitations of the ethnomethodological approach. As in most organizations, there are rules (role expectations) concerning when the office closes. However, in practice, as Zimmerman points out, the receptionist had to work out (interpret) how early to actually stop receiving new clients so as to insure that case workers would not be forced to work too late beyond official closing time(Zimmerman, 1970, p. 229).Or, whereas the formal policy (role expectations) are to assign each new case to a different case-worker in turn, it remains a permanent part of the receptionist's real role to decide (interpret) when and how to make exceptions to this rule. For example, in practice, it was seen as inappropriate to give worker A, who was still seeing his first client, a third client even if it was his turn to get another client(Zimmerman, 1970, pp. 230-31).Or again, while the receptionist's official job (structured role expectation) was just to tell clients to wait to be served, in practice she had to decide (interpret) when it would be appropriate to attempt to speed the process up. So sometimes she worked out that she needed to inform a case worker that a person had been waiting 'a long time'(Zimmerman, 1970, p. 231).

Hilbert has drawn out the general implications of ethnomethodological studies such as Zimmerman's for role theory. As he puts it: 'Much of the ethnomethodological literature supports the view that actor by necessity behave without benefit of literal role prescriptions' (Hilbert, 1981, p. 215) or: '(roles) not unequivocally prescribe procedure' (Hilbert, 1981, p. 214).

Zimmerman's study and ethnomethodology in general, then, present quite a different image of the basic process of playing a role than the other theorists we have considered (with the partial exception of Dahrendorf). We see that role-following is not just a matter of going along with fixed expectations. The actual events of everyday life constitute a complex and ever changing array of contingencies such that no fixed expectations could ever be sufficiently detailed to dictate to actors what they should actually do to fulfill their roles. Therefore, there is an inevitable place, in adequate role-following, for not being a passive recipient of society's rules but an actively engaged actor, an actor with creative, negotiating, and interpretive skills. Now, granting that there is, as ethnomethodologists have observed, this inevitable interpretive element in adequate role-playing, surely this does allow a much more active role for the person in role-following than the previous theorists have implied. And note, also, that, unlike Goffman's actor, what the ethnomethodological actor must do is not a form of withdrawal or distance, it is part of the role. However, the question that would remain is whether such a form of activity is really sufficient for us to conclude that what we would want to call self is being invested in a role. In a stimulating general critique of ethnomethodology, Blum and McHugh have depicted the limits of the ethnomethodological approach:

Thus, the deep need (of ethnomethodology) for an actor who is competent, resourceful, and even masterly in his use and expression of convention. (Blum and McHugh, 1984, pp. 114-15)

This description would seem to describe Zimmerman's receptionist very well. She is certainly more resourceful and competent in her use of the conventions, e. g. the closing time convention or the wait your turn convention, than the standard sociological image of the role player would lead us to expect. Thus, she needs to imagine ways to play her role more adequately than the letter of the role, the official policy, the explicit formal expectations, might permit. However, at the same time, her basic limit would seem to be just the competent or conventionally adequate fulfillment of the task. What seems doubtful is whether anyone whose ultimate goal is competence could really be managing to express what we could want to call their self in the role. I conclude then that the ethnomethodological actor is indeed more active than his or her conventional sociological counterparts but the form of activity envisioned is only that required for competence, for going along with the conventions, and the weakness of this is that it is possible, even easy to imagine an actor who is highly competent at their task and still would not really be seen as able to express their self in their role.

As a demonstration of this feature of ethnomethodological approaches, we can cite Hilbert's summary of what he believes ethnomethodology has taught him about role behavior:

Competent attention to such prescriptions (i. e. roles) consisted of knowing their limitations, knowing how to reconstitute their intended meanings, and knowing how to abandon them completely when the situation called for it. (Hilbert, 1981, p. 215)

Hilbert thus formulates ethnomethodology as revealing what competent role behavior amounts to, e. g. reconstituting intended and so not just clearly visible meanings, but note that he is also assuming that ethnomethodological theorizing will never take us beyond depicting what is required for competence. That is the limit of ethnomethodological horizons.

Mead's Contribution

For help in resolving what can now be seen to be quite an intractable problem, we can turn to a theorist who is currently unjustly neglected even by movements which have been much influenced by him such as interactionism and ethnomethodology, Mead. Mead is helpful in providing a sociologically viable definition of what it is to have a self. To have a self is not a matter of being utterly unique, of having an 'individual' personality. As Mead recognizes, such definitions of self are not really tenable from a sociological perspective. Instead the essence of a self is that it is the one thing that can be 'an object to itself' (Mead, 1952 p. l36). This, of course, is how Mead defines consciousness. Neither Parsons nor Goffman nor even ethnomethodology really think of an actor as a self if by that we mean, with Mead, that which can be conscious or, in other words, an object to itself.

What we consider now is how to develop this Meadian perspective in a way that utilizes a core insight from role theory, namely the existence of normative prescriptions attached to positions. Habermas, one of the few theorists to seriously engage with Mead in recent times, has pointed to a weakness in orthodox Meadian accounts of the rise of the self or consciousness. In depicting this as a matter of being able to take the role of the other or even as a matter of taking the perspective of the generalized other, Mead may have a misconception concerning how society impacts on persons (Habermas, 1987, pp. 11-15). Here we build on and also further refine Habermas' reservations about Mead. Mead explains the process of becoming a self as a matter of being able to take the attitude of others. The individual:

...becomes an object to himself only by taking the attitude of other individuals within a social environment or context of experience and behavior in which both he and they are involved. (Mead, 1952, p.138)

Mead does recognize that there is more to becoming a fully formed self than just seeing the viewpoint of concrete other people. In so far as we can play games (a crucial stage of development as he sees it) we can see ourselves from the viewpoint of a whole team but, interestingly, his fundamental approach to explaining this stage is not to adopt the perhaps more straightforward Parsonian idea that we reach this level by incorporating the rules of the game. Instead Mead sees this stage as still a matter of taking the role of the other, albeit now many more than one other:

The child who plays in a game must be ready to take the attitude of everyone else involved in that game. (Mead, 1952, p.151)

He then goes on to speak of adult social development as when we can see the viewpoint of the whole society, but even here he basically understands this process of becoming a fully socialized self (fully conscious, fully an object to oneself) as a matter of taking the attitude of the other. The key difference, as he sees it, is that now it is a 'generalized' other and not just concrete other people one must take the attitude of:

But at the second stage in the full development of the individual's self, that self is constituted not only by an organization of these particular individual attitudes, but also by an organization of the social attitudes of the generalized other or the social group as a whole to which he belongs...The individual arrives at them or succeeds in taking them, by means of further organizing, and then generalizing, the attitudes of particular other individuals. (Mead, 1952, p.158)

While it is certainly true that Mead has here moved well beyond the idea that a self has no sense of society as a whole, it can be seen that he is still imaging the basic experience of self-formation as a matter of generalizing from his root experience of taking the attitude of others.

The reservation one can have about Mead is his tendency to think of the social, that which is external to the actor, originally as other people but always, even when he allows for the influence of wider social forces, as some 'other.' As we have seen, even when he goes on to imagine less concrete others, first a whole team and then a generalized other, he is still tending to think of the social as taking the form of some (perhaps now very abstract) other. To be an object to oneself, for him, is always, then, a matter of taking the role of some other. It therefore seems to me that Mead does not give sufficient weight to the idea of society as taking the form, not of the 'other' but the form of a more direct encounter with rules or norms. Habermas goes on from a criticism of Mead much like this to assume that as part of thinking of the social as rules, we need to reject the whole idea of the self as consciousness, as that which can be an object to itself (Habermas, 1987, p. 12).

This conclusion is part of Habermas believing that the whole paradigm of consciousness is outdated. However, what if, instead we attempt to theorize how Mead's idea of the self as what can be an object to itself would be transformed (rather than rejected) if it is placed within a perspective that imagines actors as influenced to be social by rules and expectations and not just, as more orthodox Meadians might have it, by the ability to take the perspective of some other, whether a concrete other or the 'generalized' other? What we can try to develop is what the Meadian process of consciousness, of making its own behavior an object for itself, might look like in the case of such an actor.

The Principled Actor

At first we might be inclined to say that what consciousness in this realm would consist of would be a matter of competently following the rules. But an interesting problem now surfaces. Competence certainly requires awareness of the rules but how is it in any sense making one's own behavior an object for oneself? Let us move more slowly by analyzing the idea of a competent rule follower. What can be said of such a person? It can be said that they are doing well whatever the rules require. They are then competent (good) at going along with expectations. But can we not say that, if this is an actor's limits, even as this actor must be very aware of the rules, there is a glaring unconsciousness, an unself-reflectiveness in stopping here? I would suggest that something akin to making their own behavior an object for themselves for such an actor would require thinking not just about how to do it well but orienting to whether it is well, whether it is right, to do it in the first place:

Prior to the question of doing it well or poorly is: whether to do it at all....One can undertake to do it because 1) it is expected. 2) it is right. If one undertakes (to do it because one thinks it is) right, it is difficult to imagine that one would want to do it poorly. That is, to undertake something because it is right seems to require that one try to do it well. And yet to undertake to do it can be for other than right reasons--because it is expected regardless of right--and here trying to do it well is more problematic. An actor, then, can endow his action with right and/or expected grounds, which means that for any actor this itself can be a problematic issue of orientation. (Blum and McHugh, 1984, pp. 155-6)

I would suggest that theorizing the rightness rather than the expectedness of any contemplated action might make sense as a version of Mead's making one's own behavior an object for oneself. This, rather than competence, would be the self-reflective way for an actor to orient to rules.

This problem of the rightness as distinct from the expectedness of what one is doing clearly has no definitive solution. No actor can be sure what he or she is doing is right or demonstrate that what he or she is doing is right. But that does not mean that orienting to rightness cannot be differentiated from orienting just to what is expected. Thus, a distinction that can capture this significant difference is the distinction between being principled and merely being rule-guided. (Blum and McHugh, 1984, pp. 151-72) To be principled need not entail certainty or the demonstrable rightness of one's action but we can still make a distinction between being principled and just going along with expectations. The principled actor has a personal commitment to the worth, the worthiness and therefore we can say the rightness of what he or she is orienting to. This actor has theorized the rightness of what he or she is doing. We imply this when we say it is a matter of principle for them. In this sense, then, it is a matter of right versus wrong and not just expectation that the act be undertaken.

There is also another way to grasp the distinction between the principled and rule-guided actors, namely in terms of whether their respective acts signify anything about them. Both the principled actor and the actor who is just doing things because they are expected would of course be doing things, i. e. engaging in actions in the real social world. But we can suggest that they would have contrasting attitudes to the significance of their acts. In that the principled actor believes in the rightness of what s/he is doing, all their (principled) acts amount to a representation in action of what they value, of what is worthwhile. On the other hand, if an actor is just doing things because the rules expect it, then it seems wrong to say what s/he does represents what they value or care about. What can be derived from this is that, unlike the rule-guided actor, in the case of the principled actor, their self is available in their actions. This is because their actions say something about, affirm, represent what they believe.


Support for the perspective emerging here comes from the version of self elaborated in Charles Taylor's work. In his earlier work, Taylor had already developed the notion, reminiscent of Mead, that the essence of selfhood resides not in phenomena like individuality, uniqueness, separate personality but the capacity for what he calls 'strong evaluation' (Taylor, 1985). This is the idea that to have a self is to have more than desires that one is conscious of and may act on. What is required in addition is to be able to evaluate those desires in a second-order way. A being with a self can make its own desires an object of evaluation, assess their worth. In his much more fully developed later work, Taylor is able to point to that which provides the basis for the self's second-order evaluative work. It is whatever 'moral frameworks' or 'fundamental orientations' a person respects (Taylor, 1989, p.29). These frameworks are or form our self, a self being for Taylor something like an identity, who we are (self) conceived of as where we stand on what fundamentally matters to us:

We speak of people as selves, meaning that they are beings of the requisite depth and complexity to have an identity. (Taylor, 1989, p. 32)

A self, then, in Taylor's sense is more than just a bundle of desires. It is a being that can work out what to do by reflexively applying standards of worth to first-order phenomena, including among such phenomena its own desires. In so doing and in the action that it undertakes in the light of this work, it is affirming its self in the sense of its identity, that is, what it really cares about, what matters to it.

The similarity of this notion to the Blum and McHugh idea of the principled actor should be apparent. But now we need to address the question neither Blum and McHugh nor Taylor explicitly discuss, how something like their version of self can be applied to problematics drawn from the perspective of role theory.

Self in Role

For help with this issue, I turn to another non-sociological source, a paper by the philosopher Lawrence Blum. (Blum, 1994. No relation to Alan Blum cited above.) Blum begins the portion of his analysis that is relevant to our problem with a true story about a famous American teacher, Herbert Kohl. Kohl was asked by the parents of one of the children in his class to give extra tutoring to their son who, at the age of fourteen, still did not know how to read. Kohl agreed to do this in the evening two days a week for several months. It is important for our understanding the full picture to appreciate that the boy in question was very angry and defiant and that, as Kohl reports, he never actually grew to like him. Therefore, it is clear that the request from the boy's parents was certainly unsought and probably unwelcome (See Blum, 1994, p. l01). Thus, we can imagine Kohl being initially and even throughout at some level quite resistant to the task he yet undertakes.

But then we need to ask how we can explain his doing it. My account will draw on Blum's account, although I will be adapting it to the question at hand. What Kohl does here is not just a personal decision. It is in some sense a product of his role. As Blum puts it:

It is significant in this situation that it is not simply as a human being or as an individual that Kohl takes on the tutoring, that he responds to the boy's plight, but specifically as a teacher. It is as a teacher that he sees the boy's need clearly, appreciates its significance, and is in a position to do something about it. (Blum, 1994, p. l02, his emphasis)

But having pointed to the sense in which this example seems congruent with role theory, Blum now points to certain problems with applying an orthodox role theory model to account for the behavior. Role theory starts from the idea of requirements or obligations that impinge on all those who occupy a position. The problem is that clearly what Kohl does is not exactly a requirement of the role. And Blum adds that Kohl never talks about himself as experiencing the help as something he was literally required to offer (Blum, 1994, p. 102 ).

But how then do we explain it if it is neither behavior dictated by the role nor something purely personal? Blum offers a couple of pointers as to how we might proceed. At one point he speaks of Kohl as not just following a role because he 'invokes the entire structure of social meaning and tradition within which the concept of teacher gets its significance. He sees that significance as involving certain rules and ideals-promoting the intellectual growth of his pupils, teaching them habits of mind and love for learning...' (Blum, 1994, p. 102). To formulate this, it would seem that 'teacher' has become much more for Kohl than a set of prescriptions he is obligated to comply with because of the position he occupies. In that 'teacher' comes to remind him of deep 'social meanings' and 'tradition' and is something 'significant' surely he is endowing that role with rightness. It is something that, as he sees it, as he orients to it, is well worth doing and so is not just something expected of him.

As it comes to have significance for him, this has certain consequences for how he understands it. In particular, note that he comes to see it as more than a set of arbitrary rules he must fulfill. Instead he sees fundamental underlying justifications for those rules, 'values and ideals,' what Blum and McHugh would call principles. We can imagine that someone who, as here, sees teaching as an embodiment of values, e. g. love of learning, would want not just to comply with it but, so to speak, to actively further it. Hence, postulating for Kohl such a (principled) commitment helps provide for how he would do more than he is merely obligated to do. At another point, Blum speaks of how people who relate to their roles as does Kohl: '...believe deeply in (these) values and ideals and must in some way choose or at least affirm them for herself'(Blum, 1994, p. 104, his emphasis). It is probably important that Blum obviously thinks twice about formulating Kohl as 'choosing.' He is not choosing if by that we mean opting to define teaching in any way he sees fit. But he is affirming by his behavior in that how he acts does more than just comply with a pre-existing role. It is as if his act, by how it so clearly shows the depth of his commitment to teaching manages to affirm the role as something desirable, something that at least one person is convinced of the worth (rightness) of and not just the obligatory character of. Affirming or reaffirming here is what Blum and McHugh refer to as the principled actor's power to 're authorize convention (in our terms conventions attached to a role) through the agency of his decision to undertake or not.' (Blum and McHugh, 1984, p. 157)

We have tried then to account for Kohl's behavior here with the idea that what could just be a set of expectations, those of the teacher role, have become more like a set of principles or moral frameworks he believes in, wishes to affirm and reaffirm, wants to undertake to represent in ways that signify the deep values that ground these expectations. Thus, helping the boy after hours signifies love of learning, commitment to intellectual growth and so on. Now if a role can become all this for someone, we would also argue this implies that something like their self can be exercised, displayed, invested in how they can and do fulfill a role. Thus, I would say we certainly feel we see Herbert Kohl (his self) in the response he makes to this situation. How is it his self? l. Not in the sense that it is just his personality in that firstly the concept of personality fails to capture the point that structural facts connected with the occupation of teacher partly account for his decision. Second, personality as an idea would find it difficult to explain why Kohl does what he does even though he reports that his personal desires and affections were never fully enlisted on the boy's side. 2. Not in the sense that what we see on show here is some utterly free and autonomous 'individual' totally separate from social influences. We see a self more in Mead's-Blum and McHugh's-Taylor's sense: He is an object to himself in that he is reflecting on the possible significance of what he does. Thus we can imagine that he has needed to work out what it would signify as to the value he attaches to some of the deep ideals of teaching had he refused to help this boy. He is principled in that only some sense that it is right or worthwhile rather than expected that he teach or, more deeply, promote intellectual growth can account for the level of commitment that he displays. We see someone with an identity in that we see someone who obviously has certain things that matter to him, things that he cares about and also that these 'frameworks' can orient him, even to the point of causing him to dismiss some of his own reactions, e. g. his initial sense of the inconvenience and unwelcomeness of the request.

Now, why is the relation to role exemplified by Kohl and articulated with the help of Blum so important to the underlying problem of this paper, the problematic relation between having a self and being an involved social actor in the sense of having a role? Our argument would be that we are seeing and conceptualizing something not really available in all the other perspectives discussed: someone who is simultaneously fulfilling social needs, realizing social values, i. e. performing a role, and realizing their self. The structural-functional approaches and even the alternative provided by conflict theory were, it was suggested, limited in that they tended to view adequate performance of social roles as requiring the sacrifice of selfhood. In contrast to their image of the options, we see no reason to say that Kohl has to give up on his self to fulfill his role.

The account offered here has in common with Goffman's the fact that something like a self is on show. A person's identity or fundamental principles is clearly at least one version of their self but identity, at least in this example and I think in this respect the example is typical, works quite differently than role distance. The major difference is that it is clearly possible to express one's identity or principles via how one carries out his or her role. Thus, Kohl is clearly expressing the kind of teacher and even the kind of person he is by how he responds to this challenge. So, one's identity can be invested in the doing of a role, not necessarily in one's distance from it. And if this is right, it suggests that we have, finally, a way in which self can appear at the same time that playing a role can occur. Kohl, passionately reaffirming his own deep values by his work stimulated decisions, in this instance his decision to help this boy, is clearly expressing his self, his identity in the sense of what deeply matters to him in life, at the same time that he is being a teacher.

How, finally, is what we are witnessing here also different from what ethnomethodologists have lead us to expect concerning role behavior? It is true, as ethnomethodologists would have predicted, that Kohl is not able to act as he would think he needs to act in this situation just by complying with some pre-ordained script. Clearly he needs to decide or work out what ought to be done when he is confronted with the parents request. However, the weakness in applying an ethnomethodological perspective to this story is that it is hard to understand Kohl's actual creative or interpretive behavior here as just an attempt to be a competent teacher. Surely he could have sustained the sense that he was a competent teacher by (politely) declining the parents' request, say by citing how there are well specified conventional limits as to the work hours and job duties even of highly competent teachers. I want to say that Kohl shows himself here to be more than a competent, even a highly competent teacher. He shows himself to be a committed teacher. We see more than competence, we see commitment, in someone who is not just resourceful, even masterful in representing the conventions of their role, but additionally seems willing to do things that could never quite be expected of anyone in the role in question. And in that it is so clearly not just a social value but their, e. g. an individual teacher's, commitment, commitment displays and engages self in a way that even an abiding interest in competence never can.

Two Versions of Reflexivity

Now, a final issue. The sort of self that has emerged in the course of this paper is a reflexive self. How does the reflexive self as we are conceiving it relate to another current argument which identifies selfhood with a kind of reflexivity, the version of self associated with Giddens? Giddens writes:

What to do? How to act? Who to be? These are the focal questions for everyone living in circumstances of late modernity. (Giddens, 1991, p. 70)

We can say right off that these questions do not sit very well with the story of Herbert Kohl that has just been narrated. Kohl seems more to know who he is (a teacher), what he needs to do (agree to help the young man) and how to act (in ways that he thinks will promote intellectual growth and the other values he associates with his profession). But then what is Giddens referring to if it is not (what we would mean by) reflexivity? Here is a further passage in the same vein. In this instance Giddens is summarizing and partially quoting from the work of an advocate of 'self-therapy,' Janette Rainwater:

Self-therapy is grounded first and foremost in continuous self-observation. Each moment of life, Rainwater emphasizes, is a 'new moment', at which the individual can ask, 'what do I want for myself.' Living every moment reflectively is a matter of heightened awareness of thoughts, feelings and bodily sensations. Awareness creates potential change, and may actually induce change in and through itself. For instance, the question: 'Are you aware of your breathing right now?', at least when it is first posed, usually produces an instantaneous change. The raising of such an issue may make the person 'aware that she is exhibiting a normal full breathing cycle and allows her body to say "whew" in relief, take a deep breath and then exhale it.' 'And,' Rainwater adds parenthetically to the reader, 'how is your breathing now after reading this paragraph?'- a question that I could echo to whoever is reading this particular text...(Giddens, 1991, p. 71)

This confirms that this sort of self-observation is what Giddens means by reflexivity ('living every moment reflectively') even as his final phrase reveals that he is also struck by the strangeness of it all. It is strange but also strangely familiar, not, I suspect because it really characterizes 'everyone' in late modernity but because, I suggest, Giddens and Rainwater are unintentionally depicting what the tourist experience feels like. (McHugh et al. 1974) Returning to the first passage, it is characteristic of tourists to be wondering 'what to do' (because nothing seems necessary), 'how to act' (because they lack requisite knowledge), and even 'who to be' (because their real everyday selves can be placed on hold to be replaced by 'flings,' new temporary personas such as lager louts, Shirley Valentine figures, etc.) Or, looking at the second passage, if we sense the familiarity of this essentially strange mode of being, I suggest it is because it imitates the 'denaturalization' of absolutely everything that characterizes the experience of being a tourist. (McHugh, 1974, et al. p.138) The tourist has just this 'heightened awareness' of 'each moment.' That is, he/she sees (notices) the distinctive, separate quality of every particular thing. Every object is observable as what it 'is': the greenness of the grass, the height of the buildings and so on; all the things that the inhabitant has long since ceased to be particularly aware of.

The tourist, of course, is not committed to anything he/she is seeing, he/she does not belong, does not think of his/her self as 'part' of something. It is this lack of relation, commitment, or feeling of belonging which produces the heightened awareness of all the sheer particulars. To put the same point from the perspective of the person with regard to anything that, unlike the tourist, they do inhabit, feel they are part of or belong to, inherent in belonging is a kind of loss of awareness, a ceasing to notice the sheer particularity. Once we belong, things start to appear natural, they inevitably lose their freshness, they get absorbed into a whole (what we belong to). (McHugh, 1974, et al. p.143)

An extremely important example of how differently things appear to the two forms of life, tourist and inhabitant, would be in terms of how much of life appears to be a choice:

Modernity confronts the individual with a complex diversity of choices and, because it is non-foundational, at the same time offers little help as to which options should be selected. (Giddens, 1991, p.80)

To an extent, this is undoubtedly accurate and unarguable but Giddens wants us to see choice in this sense as exemplified in such extreme forms of reflexivity (in his sense) as the following:

Where do you want to live?With whom do you want to live?Do you want to work? To study? Are there any ingredients from your fantasy life that you would like to incorporate into your current life? (Giddens, 1991, p. 74, quoting Rainwater)

Why are there so many choices? Is it because, as Giddens would believe, this person is simply more reflexively aware that she/he does not actually have to, for example, live anywhere in particular? I would suggest that there are only all these choices if one assumes a person utterly without commitments, acting and thinking (at home) just like a tourist. Thus, if we say, as I would, that Kohl feels he has no real choice in the circumstances, that he senses that he pretty much has to help this kid, what does that tell us? Does it tell us that he is not reflexive enough, not noticing every moment? For example, would we say that he does not notice that he does not have to be a teacher at all, in which case of course surely he could choose not to agree to the request?

But we have already said that he is committed to being a teacher. Now we need to articulate how that is significantly different from being, as it were, a tourist with regard to his role. One thing it means is that he will perhaps not notice certain (formal) options as choices not because he is (deeply) unaware of them but because, given his commitments, he has long since worked out what is part and parcel of his job, his commitment. These actions are now so much a part of what he is that he hardly notices he is choosing them. It is the opposite of the tourist who only has so many choices because his/her actions are not part of or reflective of any underlying commitments. Thus Kohl could (and we would add should) say that he has to do what he does. He should say he feels he has no real choice, not because he unreflectively lacks full awareness of all the possibilities and how he has selected the ones he opts for, but because he is fully reflexively aware of what a teacher ought to do, so aware perhaps that he has ceased to notice (as a tourist would) all the necessary decisions, all the (in Giddens' sense) choices that his work involves him in.

Arguably, then, though this paper shares with Giddens the interest in the reflexive self, we mean very different things by it. On the one hand it can refer to a kind of social rootlessness associated with not feeling a part of anything. On the other hand, it can mean someone who is socially involved but sees that involvement reflexively, meaning that they are conscious of how their actions are a significant affirmation or negation of the worth, the rightness, of those very involvements (roles) they have.


BLUM Alan and MCHUGH, Peter, (1984) Self-Reflection in the Arts and Sciences. New Jersey: Humanities Press.

BLUM Lawrence, (1994) Moral Judgment and Particularity. Cambridge: Cambridge University Press.

DAHRENDORF, Ralf (1973) Homo Sociologicus. London: Routledge and Kegan Paul.

GERTH, Hans and MILLS, C. Wright (l954) Character and Social Structure. London: Routledge and Kegan Paul.

GIDDENS, Anthony (l991) Modernity and Self-Identity. Cambridge: Polity Press.

GOFFMANThe Presentation of Self in Everyday Life. New York: Anchor Press.

GOFFMAN (1969) 'Role-Distance' in Where The Action Is. London: Allen Lane.

HABERMAS, Jurgen (l987) The Theory of Communicative Action, Volume Two. Cambridge: Polity Press.

HILBERT, R. A. (1981) 'Toward an Improved Understanding of "Role"', Theory and Society, 10, 207-226.

McHUGH, Peter, RAFFEL, Stanley, FOSS, Daniel, and BLUM, Alan (l974) On the Beginning of Social Inquiry. London: Routledge and Kegan Paul.

MEAD, George Herbert (1952) Mind, Self, and Society. Chicago: University of Chicago Press.

MERTON, Robert (l969) Social Theory and Social Structure. New York: Free Press.

PARSONS, Talcott (l937) The Structure of Social Action. New York: Free Press.

PARSONS, Talcott (l951) The Social System. New York: Free Press.

TAYLOR, Charles (l985) 'What is Human Agency?' in Human Agency and Language. Cambridge: Cambridge University Press

TAYLOR, Charles (l989) Sources of the Self. Cambridge: Cambridge University Press.

TRAVERS, Andrew (l994a) 'The Unrequited Self', History of the Human Sciences, 7, 2, pp. 12l-140.

TRAVERS, Andrew (l994b) 'Destigmatizing the Stigma of Self in Garfinkel's and Goffman's Accounts of Normal Appearances', Philosophy of the Social Sciences, 24, 1, pp. 5-40.

TRAVERS, Andrew (l995) 'The Identification of Self', Journal for the Theory of Social Behavior, 25, pp. 303-340.

ZIMMERMAN, Don (l970) 'The Practicalities of Rule Use' in Jack Douglas (editor) Understanding Everyday Life. Chicago: Aldine Press.

Copyright Sociological Research Online, 1999