But if everything is a role, doesn’t it follow that nothing is? Role theory, of whatever brand, extends to life as a whole a dramaturgical imagery. But this extension creates a puzzle from the start. The imagery of the theater draws its meaning from an implicit contrast between the goings-on onstage and off. But to treat the whole world as a stage is to dismantle the frame that keeps the theater apart, and so appears to undermine the metaphor by depriving the dramaturgical setting of its distinctness. Modeling our understanding of human beings on the theater is sometimes designed to make a point that survives the dismantling. Especially when we recall the origin of “role” in mask, the main objective of role theory may be to accentuate a dividing line between a true inner self, on the one hand, and people’s external comportment in social situations, on the other. In this vein, the actor in the theater is supposedly marked by such a division, for example, between Laurence Olivier’s self and his demeanor when playing Hamlet. But this construal of the theatrical imagery is not available within a thoroughly constructivist conception of self. Such a conception leaves no room for an asocial, “true” inner self that is engaged in enacting roles without being fused with them. Retaining the theatrical imagery within a constructivist frame is possible, however, if we draw a distinction within the domain of roles between different ways in which roles can relate to the self. Although every facet of human life can in principle be subsumed under some role, not all roles are the same.
The first step in this direction is the observation that to conceive of the self as constituted by roles is not to think of it as a mere assemblage of roles. To form a self, the roles must be integrated in some fashion. What does the integration consist in? The answer I suggest is metaphorical: the roles must form a dovetailing, interrelated, and interacting arrangement that we can imagine as possessing a certain “density” or as forming a “core.” But as this imagery suggests, people can also occupy roles that are more tenuously connected to the elements forming that core; the ties may become too distended to still count such roles as integral parts of the self. We colloquially mark this possibility by contrasting the personal and the impersonal as two different styles of demeanor and interaction. Within role-t heory this contrast is best captured in the sociological notion of role-distance, which denotes a mode of enacting a role without identifying with it and so without fully integrating it into the self.9 Identification with a role or detachment from it need not be fixed: the distance between a person and a role can shrink or expand; it can fluctuate over time. It is also not the case that some roles must be worn tightly, whereas others are kept at a distance by all their takers. Still, a certain degree of uniformity in the style of enacting different roles exists, and so some generalization regarding role-distance is possible: certain roles are more likely to be enacted at a distance than other roles. Moreover, these uniformities have a normative side; for example, it seems less appropriate to be a detached parent than a detached bank-teller. Where does this normative difference come from? The most straightforward answer points to the roles themselves: the distance appropriate for the enactment of a role is itself one of the role’s normative aspects, part of its script. So it is meaningful, though not altogether accurate, to speak in general about detached or distant as opposed to nondetached or proximate roles. We act in a personal capacity when we enact a proximate role, and impersonally when enacting a distant one.
What are the markers and incidents of proximity or distance? Of the many factors that come to mind, two are particularly salient; they concern, respectively, sincerity and motivation. An example I have used on other occasions effectively illustrates both features. Suppose that someone helps
Sam’s four-year-old child to cross the street just as Sam happens to walk by, and so he says to the benefactor, “Thank you for helping my child.” Contrast this episode with another familiar display of politeness: the AT&T operator who concludes each exchange with a customer by proclaiming: “Thank you for using AT&T.” Consider sincerity first. Even though Sam thanks his child’s helper strictly in the capacity of parent (after all, the benefactor did not render any help directly to Sam), the thanking in this case would be generally interpreted as conveying a genuine sense of gratitude. This is in keeping with a common understanding of how a norm of sincerity ordinarily undergirds our speech acts. As Professor John Searle observes, “it is linguistically unacceptable (though not self-contradictory) to conjoin the explicitly performative verb with the denial of the expressed psychological state.”10 Specifically, in our example, it would be unacceptable to say, “Thank you, but I’m not really grateful.” However, when we turn to the operator’s thanking, we discover the opposite, and equally instructive, oddity. It would be quite ludicrous for the overly zealous telephone operator to say, “Thank you for using AT&T,” and then add, “and I really mean it.” The oddity would not disappear even if the particular operator happened to experience a sense of gratitude, born perhaps of a belief that his own livelihood is secured by the customer’s patronage. Evidently, the norm of sincerity does not belong in this language game.
A similar contrast between the two roles illustrates the difference in motivation. In conveying their thanks, both the parent and the operator act in compliance with their respective roles’ requirements. Why do they do so? Why do the parent and the operator comply with their respective roles’ demands? Again, the question I consider is not in the first place psychological but normative: the claim is that different standards of appropriate motivation apply in the two cases. One way to see the difference is to notice that in the parent’s case, the thanking will be taken to reveal the parent as a polite person. It is likely, for example, that Sam would respond similarly on other occasions and in different roles; e.g., he would be careful to thank a helpful salesperson in a store. The thanking in both instances is personal and so internally motivated, just because or in the sense that the display of politeness in both instances issues from an undifferentiated core of the self, which is constituted in part by the two respective roles and is activated or displayed in their performance. Not so in the case of the operator: we would expect no consistency between his demeanor when buying in a store and the gratitude he expresses as operator. Whether rude or polite on other occasions, the operator would be sure to use the polite refrain over the phone. Correspondingly, since this role is external and impersonal, it is not expected to exert a motivating force all by itself; it must be aided by some inducement that links the role’s requirements to its holder. The most common one is of course payment: the operator appropriately performs the role’s demands in order to be paid, whereas monetary inducement would be out of place in the parent’s case.
The metaphor of distance suggests gradation, and relatedly, identification can come in degrees. So the two kinds of roles we’ve contrasted are better seen as the polar ends of a continuum than as a binary distinction. This raises a question concerning identity. If roles are constitutive of the self, then a binary opposition would seem to be more apt: shouldn’t there be a clear answer to the question what is me and what is not? I will not try to do full justice to this important query but only dull its edge by enlisting for this purpose the conclusions reached in Derek Parfit’s seminal study of personal identity.11 Like most philosophers who deal with this topic, Parfit is concerned for the most part with temporal identity, whereas our present inquiry primarily raises issues of what we may call compositional identity, concerning the composition of the self and its boundaries at any given time. The two issues are closely related, however. Now on Parfit’s view, temporal continuity is a matter of degree, so that an earlier self can be more or less connected to a later one; there is no deep fact of the matter as to whether two temporally bound selves are stages of a single one or not. The same scalar picture applies to compositional identity as well. The question whether something is a part or an aspect of me does not present a genuine binary option, and it need not have a single correct answer. Parfit suggests that we can nonetheless preserve the binary logic that he associates with the concept of identity, by legislating a clear-cut if somewhat arbitrary criterion in light of which binary determinations concerning personal identity over time can be made. Doing so, he maintains, is harmless as long as it is also held that, our ordinary beliefs to the contrary notwithstanding, “identity is not what matters.” He advocates accordingly revising the prevailing attitude toward the self by diminishing the normative significance we attach to its identity. But we can also move in the opposite direction: retaining the ordinary normative significance we attach to our identity, temporal as well as compositional, while allowing that a looser concept of identity is appropriate in this case, one that accommodates a degree of fluctuation and indeterminacy.