Baumeister et al. rightly argue for the indispensability of groups in human social life. Yet, in positing individual differentiation within groups as the key to effective group functioning, Baumeister et al. adopt a Western-centric view of the individual's relationship to the group, and they also overlook an alternative social identity account in which depersonalisation, not individuation, is central to understanding many group phenomena. I focus my commentary on Baumeister et al.'s account of group performance, although I believe that similar points can be made about other group phenomena that they discuss.
Baumeister et al.'s hypothesis is that there is a universal need for people to feel individually differentiated within groups and, if that need is not satisfied, groups become dysfunctional. I believe that this conceptualisation is a view rooted in individualistic Western cultures that emphasise the independence of the self from others. Such self-construals are by no means universal. In more collectivistic societies, the self is typically seen as interdependent with others (Markus & Kitayama Reference Markus and Kitayama1991; Triandis Reference Triandis1989). These cultural differences in the ways people see themselves and their groups have implications for Baumeister et al.'s analysis of the conditions most likely to inhibit social loafing in groups and encourage its opposite – individuals working harder in a group than alone, what I have called “social labouring” (Brown Reference Brown2000, p. 190). Drawing on Karau and Williams' (Reference Karau and Williams1993) meta-analysis, Baumeister et al. conclude that social loafing is reduced in settings where individuals are identifiable, feel that their contributions are indispensable and are subject to evaluation by others. What Baumeister et al. fail to note, however, is that 90% of the samples in Karau and Williams' data set involved Western participants. Once one broadens one's cultural focus, such a conclusion may no longer be warranted.
Take, for example, a well-known study (not cited by Baumeister et al.) that compared individual and group performance in China, Israel (two collectivistic societies), and the United States (a more individualistic society) (Earley Reference Earley1993). Participants thought that they would be performing a task either alone (Individual condition), with a group of people from the same region as them (In-group condition), or with a group of people from a different region (Out-group condition). In the latter two conditions, they believed that the group's performance would be assessed only as a whole. In the US sample, consistent with Baumeister et al.'s hypothesis, participants worked harder in the Individual condition than in either of the two group conditions. In Israel and China, however, performance was higher in the In-group condition than in the Individual condition, clear evidence of social labouring even under conditions of low identifiability.
In another study (also not cited by Baumeister et al.), Earley (Reference Earley1989) compared North American and Chinese participants on an additive task, under conditions of high and low shared responsibility for the task outcome, and under differing degrees of individual accountability. More individualistic participants (mainly from the United States) worked harder under low shared responsibility and were affected by the accountability variable. However, the more collectivistic participants (mainly from China) worked harder under high shared responsibility and were unaffected by accountability – their highest performance was observed in the high shared responsibility-low accountability condition, just where Baumeister et al. would have predicted the lowest performance.
However, it is not necessary to venture beyond American shores to observe social labouring under conditions of low identifiability. According to Social Identity Theory (SIT, Tajfel & Turner Reference Tajfel, Turner, Austin and Worchel1979; Reference Tajfel, Turner, Worchel and Austin1986), in some contexts people's group memberships can acquire such psychological significance that the fortunes of the group become their fortunes too. Indeed, under conditions of maximal intergroup salience, people can become so identified with their in-group that they become “depersonalised” and start to assume many of the key group-defining attributes as their own (Turner et al. Reference Turner, Hogg, Oakes, Reicher and Wetherell1987). In such circumstances, group members may exert themselves more than they ever would when alone, particularly if by doing so they can achieve some positive in-group distinctiveness. Now, Baumeister et al. enlist SIT in support of their own analysis (sect. 1.1, para. 4) even if, curiously, they cite a nonexistent source for it (“Turner & Tajfel 1982”) and fail to mention either of its original formulations (Tajfel & Turner Reference Tajfel, Turner, Austin and Worchel1979; Reference Tajfel, Turner, Worchel and Austin1986) or one of its most important derivatives (Turner et al. Reference Turner, Hogg, Oakes, Reicher and Wetherell1987). As I hope is clear from my account of SIT above, one of its central concepts (“depersonalisation”) is directly at odds with Baumeister et al.'s favoured idea, “individual differentiation.”
As an example of empirical support for a SIT analysis of group performance, even in an individualistic cultural context, consider Worchel et al. (Reference Worchel, Rothberger, Day, Hart and Butemeyer1998, experiment 3; again not mentioned by Baumeister et al.). American students first undertook a task on their own. They then carried out the same task in a group, believing that their individual contributions to the group outcome were not identifiable. They did this either in the implied presence of an out-group or where no out-group was mentioned. Their in-group identity was also made differentially salient by wearing the same colour lab coats with their team name written on it or by wearing lab coats of different colours and with no team name. The presence of an out-group resulted in social labouring; the absence of an out-group caused social loafing. Social identity salience qualified this main effect such that most social labouring occurred when participants wore the same uniform in the out-group “present” condition, and most social loafing occurred in the absence of an out-group and with no common uniform. Notice that maximal group productivity co-occurred with minimal identifiability, in direct contradiction to Baumeister et al.'s hypothesis.
In summary, then, I conclude that Baumeister et al.'s analysis of the relationship of the individual to the group suffers from a degree of cultural and theoretical myopia. It has overlooked the possibility that not everyone is “WEIRD” (Henrich et al. Reference Henrich, Heine and Norenzayan2010b); and, in its neglect of the important account of the individual-group nexus provided by SIT, it too blithely assumes that individuals can be effective in groups only to the extent that they remain individuated.
Baumeister et al. rightly argue for the indispensability of groups in human social life. Yet, in positing individual differentiation within groups as the key to effective group functioning, Baumeister et al. adopt a Western-centric view of the individual's relationship to the group, and they also overlook an alternative social identity account in which depersonalisation, not individuation, is central to understanding many group phenomena. I focus my commentary on Baumeister et al.'s account of group performance, although I believe that similar points can be made about other group phenomena that they discuss.
Baumeister et al.'s hypothesis is that there is a universal need for people to feel individually differentiated within groups and, if that need is not satisfied, groups become dysfunctional. I believe that this conceptualisation is a view rooted in individualistic Western cultures that emphasise the independence of the self from others. Such self-construals are by no means universal. In more collectivistic societies, the self is typically seen as interdependent with others (Markus & Kitayama Reference Markus and Kitayama1991; Triandis Reference Triandis1989). These cultural differences in the ways people see themselves and their groups have implications for Baumeister et al.'s analysis of the conditions most likely to inhibit social loafing in groups and encourage its opposite – individuals working harder in a group than alone, what I have called “social labouring” (Brown Reference Brown2000, p. 190). Drawing on Karau and Williams' (Reference Karau and Williams1993) meta-analysis, Baumeister et al. conclude that social loafing is reduced in settings where individuals are identifiable, feel that their contributions are indispensable and are subject to evaluation by others. What Baumeister et al. fail to note, however, is that 90% of the samples in Karau and Williams' data set involved Western participants. Once one broadens one's cultural focus, such a conclusion may no longer be warranted.
Take, for example, a well-known study (not cited by Baumeister et al.) that compared individual and group performance in China, Israel (two collectivistic societies), and the United States (a more individualistic society) (Earley Reference Earley1993). Participants thought that they would be performing a task either alone (Individual condition), with a group of people from the same region as them (In-group condition), or with a group of people from a different region (Out-group condition). In the latter two conditions, they believed that the group's performance would be assessed only as a whole. In the US sample, consistent with Baumeister et al.'s hypothesis, participants worked harder in the Individual condition than in either of the two group conditions. In Israel and China, however, performance was higher in the In-group condition than in the Individual condition, clear evidence of social labouring even under conditions of low identifiability.
In another study (also not cited by Baumeister et al.), Earley (Reference Earley1989) compared North American and Chinese participants on an additive task, under conditions of high and low shared responsibility for the task outcome, and under differing degrees of individual accountability. More individualistic participants (mainly from the United States) worked harder under low shared responsibility and were affected by the accountability variable. However, the more collectivistic participants (mainly from China) worked harder under high shared responsibility and were unaffected by accountability – their highest performance was observed in the high shared responsibility-low accountability condition, just where Baumeister et al. would have predicted the lowest performance.
However, it is not necessary to venture beyond American shores to observe social labouring under conditions of low identifiability. According to Social Identity Theory (SIT, Tajfel & Turner Reference Tajfel, Turner, Austin and Worchel1979; Reference Tajfel, Turner, Worchel and Austin1986), in some contexts people's group memberships can acquire such psychological significance that the fortunes of the group become their fortunes too. Indeed, under conditions of maximal intergroup salience, people can become so identified with their in-group that they become “depersonalised” and start to assume many of the key group-defining attributes as their own (Turner et al. Reference Turner, Hogg, Oakes, Reicher and Wetherell1987). In such circumstances, group members may exert themselves more than they ever would when alone, particularly if by doing so they can achieve some positive in-group distinctiveness. Now, Baumeister et al. enlist SIT in support of their own analysis (sect. 1.1, para. 4) even if, curiously, they cite a nonexistent source for it (“Turner & Tajfel 1982”) and fail to mention either of its original formulations (Tajfel & Turner Reference Tajfel, Turner, Austin and Worchel1979; Reference Tajfel, Turner, Worchel and Austin1986) or one of its most important derivatives (Turner et al. Reference Turner, Hogg, Oakes, Reicher and Wetherell1987). As I hope is clear from my account of SIT above, one of its central concepts (“depersonalisation”) is directly at odds with Baumeister et al.'s favoured idea, “individual differentiation.”
As an example of empirical support for a SIT analysis of group performance, even in an individualistic cultural context, consider Worchel et al. (Reference Worchel, Rothberger, Day, Hart and Butemeyer1998, experiment 3; again not mentioned by Baumeister et al.). American students first undertook a task on their own. They then carried out the same task in a group, believing that their individual contributions to the group outcome were not identifiable. They did this either in the implied presence of an out-group or where no out-group was mentioned. Their in-group identity was also made differentially salient by wearing the same colour lab coats with their team name written on it or by wearing lab coats of different colours and with no team name. The presence of an out-group resulted in social labouring; the absence of an out-group caused social loafing. Social identity salience qualified this main effect such that most social labouring occurred when participants wore the same uniform in the out-group “present” condition, and most social loafing occurred in the absence of an out-group and with no common uniform. Notice that maximal group productivity co-occurred with minimal identifiability, in direct contradiction to Baumeister et al.'s hypothesis.
In summary, then, I conclude that Baumeister et al.'s analysis of the relationship of the individual to the group suffers from a degree of cultural and theoretical myopia. It has overlooked the possibility that not everyone is “WEIRD” (Henrich et al. Reference Henrich, Heine and Norenzayan2010b); and, in its neglect of the important account of the individual-group nexus provided by SIT, it too blithely assumes that individuals can be effective in groups only to the extent that they remain individuated.