Refine articulation between cohesion and differentiation
We are not convinced that cohesion and differentiation are two sequential steps within group formation, and we would like to suggest that the model should be more dynamic. These two dimensions could be articulated along two main directions. First, the two dimensions can coexist. The same individuals can act cohesively in relation to some shared goals and also develop differentiation in relation to other tasks or projects. Furthermore the coexistence of cohesion and differentiation is precisely a core principle for some organizations (see the literature on high reliability organizations, e.g., Weick & Roberts Reference Weick and Roberts1993). Second, both processes interact permanently. In fact, a cohesive group action usually enriches and further differentiates individual agents. Differentiated agents, in turn, have stronger identities. They are thus more likely to both modify and adhere to group prescriptions. Finally, although cohesion and differentiation do not necessarily work in opposite directions, they may do so in some cases. Agents that are too highly differentiated can also lead to group disaggregation. To sum up, rather than characterizing two successive steps of group life, cohesion and differentiation coexist and permanently interact in many directions.
The proposal carries implicit value choices that are not necessarily universal
The proposed model is very abstract and tends to generalize observations made in particular contexts. However, some key aspects are not well-defined, taken as universal, or kept implicit. We observe that (1) no definition of the concept of group is provided, (2) an historical perspective shows that the concept of differentiation is far from being universal, and (3) a great part of the argument relies on the definition of success/failure, or system gain.
As a matter of fact, the authors do not provide a definition of the concept of group. The nature of the groups they refer to is unclear. For example, they quote in the same paragraph Le Bon, who refers to crowds, and Smith, who refers to hierarchical organizations, namely factories. Social theory distinguishes different processes of individual identification/differentiation according to group characteristics: whether it is based upon anonymous or face-to-face relations, formal or informal, primary or secondary, and so forth.
The concept of individual was not yet elaborated in the ancient Greece (See, e.g., Jaynes Reference Jaynes1976). The philosopher Descartes is credited with introducing the concept of differentiation in the seventeenth century. Japanese society is still very influenced by Confucian ethics, in which the process of integration/differentiation is very different from that of Western societies as the authors describe it (see Geertz Reference Geertz1974). It is thus quite unclear to what extent the conclusions of the authors can be generalized.
The same outcome can be judged a success or a failure, depending upon the assessment criteria. Because they define “system gain” in terms group productivity or performance, they implicitly value the group over the individual; this choice should be made explicit. Moreover, “system gain,” as Baumeister et al. define it, is not necessarily a primary motivation for all groups. And it is probably even less so for individuals.
The success of a group in shaping individual behaviour should refer to the values of group individual members
Since Pareto's (Reference Pareto1916) Trattato di Sociologia Generale, and more exactly a paper he published a few years earlier (Pareto Reference Pareto1913), it is quite clear to social scientists that what is best for a group generically differs from what is best for the individuals who compose it. This divergence makes it clear that individual and group objectives have to be distinguished.
Clearly, the benchmark of group behaviour should be group objectives. The paper focuses on how group behaviour (productivity) emerges from individual behaviour and group organization. It ignores completely the question of how individual objectives are incorporated into group objectives. The whole literature on Social Choice makes it clear that the relationship is almost never straightforward, except in very particular and unrealistic circumstances, such as when all individuals have a cleardefined, common goal (Arrow Reference Arrow1951/1963).
Moreover, the very fact that identification to the group and differentiation matters is a clear proof that individuals do not care for productivity only. Thus, there is, if not a contradiction, at least a great reductionism in adopting total productivity as the group objective or as a criteria for measuring its performance.
To sum up, we greatly appreciate the efforts of the authors in attempting to incorporate systematically both identification and differentiation into the analysis of group behaviour. We propose that their approach (1) be extended to allow both aspects to play a key role simultaneously, (2) could benefit from making their value choices more explicit, and (3) should account for the fact that group objectives generally differ from those of the individuals that compose it.
You have
Access
We highly appreciate the intent of the authors to underline the role of both identification and differentiation in explaining individual behaviour in groups. We would like to suggest that (1) cohesion and differentiation should be better articulated, (2) the proposal carries implicit value choices that are not necessarily universal, and (3) the success of a group in shaping individual behaviour cannot be judged using a priori criteria but should refer to the values of individual group members.
Refine articulation between cohesion and differentiation
We are not convinced that cohesion and differentiation are two sequential steps within group formation, and we would like to suggest that the model should be more dynamic. These two dimensions could be articulated along two main directions. First, the two dimensions can coexist. The same individuals can act cohesively in relation to some shared goals and also develop differentiation in relation to other tasks or projects. Furthermore the coexistence of cohesion and differentiation is precisely a core principle for some organizations (see the literature on high reliability organizations, e.g., Weick & Roberts Reference Weick and Roberts1993). Second, both processes interact permanently. In fact, a cohesive group action usually enriches and further differentiates individual agents. Differentiated agents, in turn, have stronger identities. They are thus more likely to both modify and adhere to group prescriptions. Finally, although cohesion and differentiation do not necessarily work in opposite directions, they may do so in some cases. Agents that are too highly differentiated can also lead to group disaggregation. To sum up, rather than characterizing two successive steps of group life, cohesion and differentiation coexist and permanently interact in many directions.
The proposal carries implicit value choices that are not necessarily universal
The proposed model is very abstract and tends to generalize observations made in particular contexts. However, some key aspects are not well-defined, taken as universal, or kept implicit. We observe that (1) no definition of the concept of group is provided, (2) an historical perspective shows that the concept of differentiation is far from being universal, and (3) a great part of the argument relies on the definition of success/failure, or system gain.
As a matter of fact, the authors do not provide a definition of the concept of group. The nature of the groups they refer to is unclear. For example, they quote in the same paragraph Le Bon, who refers to crowds, and Smith, who refers to hierarchical organizations, namely factories. Social theory distinguishes different processes of individual identification/differentiation according to group characteristics: whether it is based upon anonymous or face-to-face relations, formal or informal, primary or secondary, and so forth.
The concept of individual was not yet elaborated in the ancient Greece (See, e.g., Jaynes Reference Jaynes1976). The philosopher Descartes is credited with introducing the concept of differentiation in the seventeenth century. Japanese society is still very influenced by Confucian ethics, in which the process of integration/differentiation is very different from that of Western societies as the authors describe it (see Geertz Reference Geertz1974). It is thus quite unclear to what extent the conclusions of the authors can be generalized.
The same outcome can be judged a success or a failure, depending upon the assessment criteria. Because they define “system gain” in terms group productivity or performance, they implicitly value the group over the individual; this choice should be made explicit. Moreover, “system gain,” as Baumeister et al. define it, is not necessarily a primary motivation for all groups. And it is probably even less so for individuals.
The success of a group in shaping individual behaviour should refer to the values of group individual members
Since Pareto's (Reference Pareto1916) Trattato di Sociologia Generale, and more exactly a paper he published a few years earlier (Pareto Reference Pareto1913), it is quite clear to social scientists that what is best for a group generically differs from what is best for the individuals who compose it. This divergence makes it clear that individual and group objectives have to be distinguished.
Clearly, the benchmark of group behaviour should be group objectives. The paper focuses on how group behaviour (productivity) emerges from individual behaviour and group organization. It ignores completely the question of how individual objectives are incorporated into group objectives. The whole literature on Social Choice makes it clear that the relationship is almost never straightforward, except in very particular and unrealistic circumstances, such as when all individuals have a cleardefined, common goal (Arrow Reference Arrow1951/1963).
Moreover, the very fact that identification to the group and differentiation matters is a clear proof that individuals do not care for productivity only. Thus, there is, if not a contradiction, at least a great reductionism in adopting total productivity as the group objective or as a criteria for measuring its performance.
To sum up, we greatly appreciate the efforts of the authors in attempting to incorporate systematically both identification and differentiation into the analysis of group behaviour. We propose that their approach (1) be extended to allow both aspects to play a key role simultaneously, (2) could benefit from making their value choices more explicit, and (3) should account for the fact that group objectives generally differ from those of the individuals that compose it.