SOCIAL IDENTITY CONTRIBUTIONS Social Identity as a Contextual Product

SOCIAL IDENTITY CONTRIBUTIONS Social Identity as a Contextual Product

One of the key differences between an identity-based definition of the group and more traditional sociological definitions is that in the former case the group itself—together with the strength and nature of individuals’ attach- ment toit—is assumed tobe context-dependent rather than in any sense given (e.g., demographically). For example, whether and how a woman acts in terms of a gender-defined identity in an organization depends not simply on her possession of ‘objective’ physical, psychological, or behavioral characteristics but also on how that social category is comparatively and normatively defined, whether it has prior meaning, and on the individual’s perceived prototypicality in relation to a particular category definition. Being ‘a woman’ is thus likely to mean something different in a ‘traditional’ organ- ization where women typically occupy junior positions and struggle to gain promotion, and where feminism is frowned upon, than it does in a more progressive organization (Ellemers, van den Heuvel, de Gilder, Maass, & Bonvini, 2004; Fajak & Haslam, 1998; Fiske, Bersoff, Borgida, Deaux, & Heilman, 1991; Schmitt, Ellemers, & Branscombe, 2003).

For example, it follows from social identity theory that in the traditional organization women are more likely to embrace strategies of social creativity which lead them to self-enhance by defining themselves positively on non- status-defining attributes (e.g., sociability, compliance) than to pursue strategies of social change which bring them into competition with men on status-defining dimensions (e.g., for pay). It also follows from self- categorization theory that in the traditional organization women who do compete with men (or who pursue strategies of personal mobility) would tend to be relatively aprototypical and, hence, should be likely (a) to be seen as less ‘womanly’ (Hopkins, 1996), (b) to receive less support from other women, and (c) to have less influence over them (Ellemers, van den Heuvel et al., 2004).

Following on from this point, it can be seen that there are no inherent, stable definitions of social categories (whether ingroups or outgroups) and no predefined, universal identities in terms of which a person will define them- selves. This point is discussed by Wharton (1992) in an extended treatment

69 of the way in which employees’ self-definition in terms of gender and ‘race’

S OCIAL I DENTITY IN I NDUSTRIAL AND O RGANIZATIONAL P SYCHOLOGY

can change across different workplace settings (see also Gioia et al., 2000; Jackson, 1992; Ridgeway, 1991). Related evidence also emerges from pro- grammatic work conducted by Levine and his colleagues which shows how social context can change the meaning of specific events for particular social identities. This work relates to issues of stress and health (for a discussion see Haslam, 2004) and shows how contextual changes in identity definition impact upon the perceived severity of particular stressors and symptoms. In an illustrative study, Levine (1999) presented female office secretaries with scenarios in which a particular constellation of symptoms was described. Pre-testing had established that some of the symptoms were particularly relevant to a secretarial identity because they impacted on secretarial work (restricted manual dexterity, back pain, flu), and some were particularly relevant toa gender identity because they affected physical appearance (a scar on the face, a broken nose). The participants’ task was to rate how distressing they would find each of these scenarios and how much each would adversely affect their lives. Significantly, though, before doing this, gender was made salient for half of the participants by telling them that the researchers were interested in comparing the responses of men and women, while the other half had their secretarial identity made salient by telling them that the researchers were interested in the responses of different professional groups. As predicted, the extent to which particular symptoms were seen as a cause for concern depended on the extent that they were threatening to the particular social identity that the experimenter had made salient. So, for example, when their identity as secretaries was salient, women saw symptoms relevant to gender identity as reasonably trivial. However, when their gender was salient, women perceived these same scenarios to be much more distressing, but saw symptoms related to secretarial work as more trivial.

Levine (1999) developed these arguments by elaborating upon the context- sensitivity of the self-categorization processes at work in symptom appraisal. In a subsequent study, male rugby players evaluated the seriousness of a range of illness scenarios under conditions where their identity as males was made salient. However, the contextual definition of this identity was varied by indicating that responses were being compared with those of either women or ‘new-age’ men. Levine (1999) reasoned that the rugby players’ understanding of what it means to be male—and hence what constitutes a male-related stressor—would vary in these two contexts. In particular, he predicted that when they were compared with new-age men, the rugby players would perceive their ‘traditional’ male identity to be under threat and that, in order to re-establish a distinct identity, they would here be particularly keen to emphasize their own masculinity by downplaying the seriousness of emotional threats (a violent temper, depression) and threats tophysical attractiveness (a facial scar, a burn on the hand). The results confirmed these predictions. In this, the findings suggest not only that salient