Baumeister and colleagues argue for the indispensability of groups in human life. Yet, in positing individual differentiation as the key to effective group functioning, they adopt a Western-centric view of the relationship of the individual to the group and overlook an alternative social identity account in which depersonalisation, not individuation, is central to understanding many group phenomena.

We show that, under some circumstances, identification and differentiation in the form of competition and individual rewards may undermine, rather than improve, group performance. The key factor for successful group performance seems to be whether or not group members share common goals and whether or not they have aligned incentives.

Our commentary articulates some of the commonalities between Baumeister et al.'s theory of socially differentiated roles and Strategic Differentiation-Integration Effort. We expand upon the target article's position by arguing that differentiating social roles is contextual and driven by varying ecological pressures, producing character displacement not only among individuals within complex societies, but also across social systems and multiple levels of organization.

The idea that differentiated selves almost always improve group outcomes is overly simplistic. We argue that it is essential to distinguish between two distinct elements of differentiated selves – identifiability and specialization – and to identify conditions under which they influence group outcomes. Adopting a group-by-situation perspective, in which group and situation variables are considered jointly, is recommended to generate novel hypotheses.

The authors argue that group performance depends on the degree to which group members identify with the group as well as on their degree of differentiation. In this commentary, I discuss results from agent-based simulations, suggesting that group performance depends, at least in part, on features orthogonal to agents' caring about group performance or about how they are perceived by other group members.

To understand a group's (dys)functionality, we propose focusing on its members' concerns for their reputation. The examples of prosocial behavior and information exchange in decision-making groups illustrate that empirical evidence directly or indirectly suggests that reputational concerns play a central role in groups. We argue that our conceptualization fulfills criteria for a good theory: enhancing understanding, abstraction, testability, and applicability.

Does joining groups trigger a cascade of psychological processes that can result in a loss of individuality and lead to such outcomes as social loafing and poor decision-making? Rather than privileging the self comprising primarily individual qualities as the “true self,” a multilevel, multicomponent approach suggests that, in most cases, personal and collective identities are integrated and mutually sustaining.

On the basis of research in the social identity tradition, we contend (a) that identification and differentiation are not mutually exclusive, (b) that a sequence in which identification gives way to differentiation is not necessarily associated with superior organizational outcomes, and (c) that social identification, and leadership that builds this, is generally a prerequisite for group success.

It may be true that “groups need selves,” as Baumeister et al. contend. However, certain types of selfhood and too much selfhood can both be detrimental to group functioning. I draw on theory and research on dual selves in work groups and teams to outline boundary conditions to the hypothesis that emphasizing individual selves yields positive effects for groups.

Humans are continually diverging and converging with respect to each other. Research across many domains suggests that differentiation and integration are aspects of a more complex set of dynamics, and are not step-wise but interdependent and continuous. Research on conformity in particular reveals that divergence and dissent are forms of cooperation, reflecting concerns for both individual and group integrity.

All groups are differentiated into more or less group-prototypical members. Central members readily influence and lead the group, and they define its identity. Peripheral members can feel voiceless and marginalized, as well as uncertain about their membership status – they may engage in extreme behaviors to try to win acceptance. These relative prototypicality dynamics sometimes benefit group performance but sometimes compromise performance.

The target article resuscitates an old but outdated dichotomy: a theoretical dualism between group belonging and intragroup differentiation. A convergence of evidence – including that within the social identity tradition – shows that intragroup differentiation is not incompatible with strong group identity. Indeed, when norms encourage autonomy, dissent, and individual freedom, intragroup differentiation occurs through group commitment, not despite it.

In this commentary, I focus on several problems that the authors' understanding of group identity raises: the legality of avoiding background diversity, the problem of effectively unshareable knowledge, the practical quality of some outcomes arrived at by groups with homogeneous backgrounds, and moral issues about fairness. I note also that much recent research challenges the view that background diversity is more likely to be a detriment than a benefit.

We agree with aspects of Baumeister et al.'s view that shared identities are necessary during initial stages of group formation. In contrast to their analysis, however, we provide evidence that the value of self-differentiation depends more on the task itself than on the stage of group development and challenge the authors to focus on the functions of the group.

Analyzing how various forms of member similarity and difference affect group performance is a worthwhile task. I argue that the authors' analysis would be improved by distinguishing between subjective and objective forms of member differentiation and by utilizing a different typology of group tasks.

The optimal functioning of male coalitionary behavior in a military context may run contrary to some of the arguments about the importance of individual differentiation in Baumeister et al. Incentives become institutionally inverted within military contexts. Because the history of combat exerted powerful and sustained selection pressures on male groups, individual identification can work against the successful completion of collective action problems surrounding in-group defense in military contexts.

Notwithstanding the appeal of the “one size fits all” approach that Baumeister et al. propose, we argue that there is no panacea for improving group performance. The concept of “differentiation of selves” constitutes an umbrella term for similar seeming but actually different constructs. Even the same type of “differentiation of selves” can be beneficial for some and harmful for other tasks.

Whereas differentiation is overestimated – it more often hurts than helps group performance – identification is underestimated. A more viable perspective sees identification and cooperative motivation as the sine qua non of group functioning, with differentiation helping in a relatively narrow set of cognitively complex tasks that require creativity and deep and deliberate information processing by individual members.

Strong social identity does not lead to lack of accountability and “bad” behavior in groups and crowds but rather causes group behavior to be driven by group norms. The solution to problematic group behavior is therefore not to individualize the group but rather to change group norms, as underlined by the relational dynamics widely studied in the SIDE tradition.

Baumeister et al. propose that individual differentiation is a crucial determinant of group success. We apply their model to processes lying in between the individual and the group – vicarious processes. We review literature in four domains – attitudes, emotions, moral behavior, and self-regulation – showing that group identification can lead to vicarious contagion, reducing individual differentiation and inducing negative consequences.