researchers have actively worked out this view ( Onorato, 1997; Turner, 1988,
1989; Turner & Oakes, 1989; Turner, Oakes, Haslam, & McGarty, 1994). Our
aim here is to summarize the present picture, which, although largely drawn
from self-categorization theory, is consistent with the assumptions and
metatheory of social identity theory. Self-categorization theory is in a sense
the theory of the self-concept that social identity theory required but did
not itself provide.

We first summarize some major themes (assumptions, hypotheses) in the
currently dominant view of the self in social psychology. We then seek to
show how self-categorization theory qualifies, rejects, or otherwise chal-
lenges these ideas, as a way of outlining a different view or model of the
self. Finally, we consider the extent to which the self-categorization analysis
of the self deals with or is contradicted by the evidence that sustains the
existing view.

There is obviously not one dominant view of the self in social psychology,
but researchers have widely shared some assumptions about the nature of
the self. These assumptions, moreover, have a unifying core, which can be
referred to as the personality or interpersonal perspective. This core reflects
an orientation to the self in which the concept is used to make sense of and
guide research on individual differences and interpersonal relationships.
The dominant model emphasizes the uniqueness and relative stability and
consistency of a person's behavior even when there is situational variability
( Epstein, 1973, 1980; Kihlstrom & Cantor, 1984; Kihlstrom et al., 1988; Markus
& Cross, 1990; Markus & Wurf, 1987). It construes the self-concept in terms
of stable individual differences, relatively fixed cognitive and personality
structures, interpersonal orientations and styles, and enduring motives and
predispositions.

Four ideas in the literature, which appear frequently and are only rarely
questioned, illustrate the personality model. These ideas make up a model
in the sense that they function as axioms on which research is based rather
than as hypotheses to be tested competitively. Although some researchers
take a different position, we think that what follows is faithful to the main-
stream view.

These ideas are that the self-concept is, in content, a representation of
an individual's personal identity; each self-concept is a unique or idiosyncratic
property of the perceiver, belonging to only one individual and not shared
with others; the social self is a looking-glass self, a reflection and internali-
zation of others' reactions to the public self as presented in social interac-
tion; and the self-concept is a relatively enduring, stable cognitive structure.

-12-