Hogg2001 (JudeYew): Hogg, M. A. (2001). Social categorization, depersonalization, and group behavior. In M. Hogg. & T. S. Tinsdale. Blackwell Handbook of Social Psychology: Group Processes (pp 57-85). Malden, MA: Blackwell.

Discussion leader/summarizer: JudeYew

Research Question

This chapter aims to "explore the effects of social categorization on intra-group phenomena." (p. 74) Hogg adopts Social Identity Theory in order to show how social categorization within/amongst groups influences the intra-group behavior, collective self/social identity, social processes and structures that takes place in groups.

Key Points/Claims

Social Categorization:

Hogg's basic assumption is that groups are "social categories" of people which "transforms perception, thought, feeling, and action so that self and others are assimilated" (p. 74) into the group. Social categorization accentuates perceived similarities of people in the same categories and differences between those from other categories. People tend to categorize mainly because they want to render the world more predictable and also increasing one's self-esteem through acquiring the attributes of the categories they want to be associated with. Some of the effects of Social categorization are:

• Homogenizing of ingroups and outgroups.
• Illusory correlation effect where individuals tend to exaggerate the correlation of distinctive features and often results in the stereotyping of group members

Social Identity:

Hogg uses the term "category prototype" throughout this chapter to describe the social identity that is formed in association with groups. What he means by this is that social categories are often based on a set of abstractions to which no real member may embody. [For Hogg, one of the problems with this is that each group has a psychological essence that is reflected in this prototype which may be based on (i) attribution error, (ii) correspondence bias. This can be seen in the essentialized notions of gender and racial differences that emerges as a result of attributing such prototypical attributes onto individuals/groups.

[PaulResnick: I don't think the previous sentence is quite right. First, Hogg merely states that people tend to attribute prototypical properties to immutable psychological properties of the group and its members. Hogg doesn't agree with that immutability-- the contingency of the prototype on other groups being contrasted with is one of his key points. Second, your number (iii) above is used as an illustration of (i) and (ii) in the text; they are not three alternatives.]

Prototypes are typically formed according to the principle of "meta-contrast" which is the "maximization of the ratio of intergroup differences to intragroup differences" (p.60). This often accentuates the similarities within groups and highlights the differences between groups. As such, "prototypes can often be extreme or polarized relative to the central tendency" of the group.

Self-Categorization:

Hogg postulates that individuals associate themselves with particular social categories through a process called "depersonalization" where individuals are perceived in terms of how they match the group prototype.

"Since prototypes capture any and all features that define category membership (i.e. attitudes, feelings, and behaviors) depersonalization makes people in groups appear attitudinally, affectively, and behaviorally relatively homogeneous" (p.61).

This process makes group behavior possible as it transforms self conception so that individuals think of themselves in terms of the group prototype. Individuals tend to associate themselves with group prototypes that they positive and that enhances their self-esteem and reduces uncertainty. As a result, one's behavior modifies and transforms accordingly to the relevant group to which one identifies with. Thus, self-categorizaton depersonalizes one's behavior in terms of the group prototype.

Social Categorization effects within groups:

Conformity:
Self-categorization has the effect of depersonalizing the individuals attitudes, feelings, and behavior, and making them conform to group prototypical behaviors and attitudes.

The social process of "referent informational influence" (Hogg & Turner 1987) takes place after categorization and has the effect of creating normative behavior within the group. This process motivates individuals to attend to the group prototype values, attitudes/behaviors and heightens "ingroup identification, a sense of belonging, self-definition in group terms and ingroup loyalty and favoritism." (p.62) Some of the effects can be seen in the following phenomena:
[PaulResnick: referent informational influence is not the process of self-categorization. It's what happens after categorization-- people are motivated to learn about the prototype and to attend to the behavior of ingroup members, especially those who are prototypical.]

• crowd behavior:

Crowd behavior exemplifies instances of de-individuation where "social identity is highly salient and thus behavior is carefully regulated by well-established ingroup norms." (p.63).
[PaulResnick: the claim is that in unusual circumstances norms don't prescribe what to do, so people are particularly attuned to what other people are doing, and try to imitate it, producing even more conformity of action than you'd get just from following norms (which allows more latitude than imitating specific behaviors.] AndrewBabson: The "bystander effect" is an instance of this- someone is taken out of the bar and beaten on the sidewalk; people stand and do nothing. I've read a study, however, that when something I might call "re-individuation" (for not remembering the original term) happens- e.g. the guy getting beaten cries for help- another person, and then other people, are far more likely to do something.
This a situation where people in the group converge on a group position that is more extreme than that the average of the individual members. From the perspective of social identity theory, this tends to happen when individuals identify with and categorize themselves in terms of the group prototype which represents the central tendency for the group.
[PaulResnick: the key point here is that group prototype gets defined in terms of other groups being compared to, based on the principle of meta-contrast. So, if a group is defined as liberal, in contrast with conservatives, then the prototype for the liberals will be more extreme than average member. On the other hand, if a group is defined as liberal, in contrast with conversatives and communists, then the prototype for the liberals will be more like the average member.]
The relationship between attitude and behavior has traditionally been problematic. This position has been challenged by Terry & Hogg (1993) who argue that "the attitude behavior relationship is stronger when people self-categorize in terms of a salient group membership." (p. 64) This suggests that if we can get people to identify with a group, then the relationship between the attitude and one's behavior becomes stronger.

Group cohesion & social attraction:
Breaking away from the traditional view of group cohesiveness, Social Identity theory posits that group cohesiveness is the result of self-categorization and depersonalization. Depersonalization produces social attraction amongst group members in the following ways:

• it imbues group members with positive group attributes and renders them prototypically attractive
• it accentuates the prototypical similarity between individuals in the group, thus encouraging similarity-based liking.
• and it extends positive self-regard to fellow members who are prototypically closely linked to oneself

Differentiation within groups:
Groups are often internally differentiated based on the degree of prototypicality that the members possess. Some of the manifestations of this differentiation are:

• deviance:

In every group, there exists a number of deviants, or individuals who are perceived to only weakly match the prototypical properties of the group. This individuals are often seen as "black sheep" as they threaten the integrity and distinctiveness of the ingroup. This view of such marginality can be extended to (sub)groups as well.
[PaulResnick: Positive deviance (i.e., deviant, but in evaluatively positive ways) plays out in more complex ways. Where solidarity is important to the group, and where it's difficult for the group to claim "ownership" of the deviant behavior, then postive deviants will be rejected by the group.]

• leadership

Hogg associates leadership with prototypical centrality within a group. Essentially leaders are individuals who are more prototypical. As such, these individuals are well liked, able to extend their influence, and are socially attractive to the other members. These individuals are thus able to embody public endorsement and are imbued with prestige and status. For Hogg "prototypical leaders do not need to exercise power in order to have influence; they are influential by virtue of their position and the depersonalization process that assimilates members to the prototype. They and their suggestions are intrinsically persuasive because they embody the norms of the group." (p. 70) [PaulResnick: However, there may be a natural progression from leadership to power, where increasing status differentiation creates an intergroup relationship. This progression can be curbed by antyhing that regrounds leadership in prototypicality, such as an external threat that enhances identification and solidarity within the group.] [PaulResnick: Also, note that leaders may employ various strategies to maintain their prototypicality, for example, by trying to manipulate the set of outgroups against which comparisons are made and hence how the prototype is constructed.]

• structural dedifferentiation within groups

One aspect of this differentiation is the creation of marginal and sub-groups. Optimal distinctiveness consideration (Brewer 1991) explains how such subgroups exists with or are assimilate by the super ordinate group. The theory states that "people strive for a balance between conflicting motives for inclusion/sameness (satisfied by group membership) and for distinctiveness/uniqueness (satisfied by individuality)." (p. 73) This theory explains why the larger a group gets, the more necessary it is for such groups to "departmentalize" so that its members feel over-included.

Another way in which groups can encompass social categories is through roles. The roles one plays within a group determines one's identification and commitment to the group. Often these roles are determined by status, or the degree to which members have skills or qualities specific to the group's purpose.

Critique

Some questions:
- One of the questions that has been raised by Hogg in his article is the point at which individuals begin to self-categorize. This is a question that he did not answer fully for me and is perhaps a crucial point to raise for a class focused on virtual communities/groups. At what point do we start to see ourselves as a part of the group?

[KathyLee] Along similar lines, while the chapter provided a thorough review of intra-group phenomena, I found a lot of it abstract and wanting of a bit more qualification. Like Jude's question above, at what level of aggregation (or group size) do we start observing these behaviors and phenomena? I realize that explicit numbers like 20 +/- 5 may not be meaningful, but when I think of a group, it usually doesn't include social aggregations larger than--let's be arbitrary--a class of students, lets say. He does go on to make concluding remarks about the promises of this line of research for organizational studies. In short, at what levels of social aggregation is his analyses applicable and useful?

[PaulResnick] I think the idea is that social categorization happens at all levels of social aggregation (small groups, institutions, religions, nations) and that there's some underlying similarity in how it works at all those different levels. There may be important differences at the different levels of aggregation, but I find it works pretty well for me in thinking about categories at levels of social aggregation as diverse as my family, the 684 students, our entire class, HCI students, SI master's students, SI, the University of Michigan, Americans, Democrats, etc.

Connections with other readings, ideas, etc.

[JudeYew] I tend to agree with the points made in this chapter, as I feel that individual behavior is very group context dependant. If one identifies strongly with a certain group, then one's behavior in that group will be more aligned with the norms of that group. This would perhaps explains students class participation in some classes and not others.

baseball
[KathyLee] excerpts from a new york times article∞ on group dynamics in the red sox and yankees.




Is common purpose different from common identity? How do the Yankees corroborate (or refute) the social attraction hypothesis or social identity theory of leadership? According to Hogg, common identity (i.e., social identity) enforces adherence to prototypicality as a desirable trait. Does common purpose somehow make that requirement less salient? (BTW, common purpose sounds very much like Wenger's shared enterprise)

Class Discussion

CaRichardson
This article along with the summary by JudeYew made some pretty abstract concepts remarkably clear despite only briefly mentioning specific examples. However, I agree with KathyLee that some actual detailed examples would have been nice. As I was reading the article and all of the one sentence references to various published research, I kept wondering how all of the constructs would be operationalized in a research study. Considered the Yankes example above. How do you measure a group's level of decentralization, degree of connectedness, and resiliance. It is one thing to use these terms in a newpaper article, but it seems like measuring them for a research study on ecommunities would be pretty difficult. One particularly important construct to define if using the theory presented in this article would be the group's prototype. Would you ask ingroup memebers (and maybe outgroup memebers) to describe the characteristics of the group and then do some kind of qualitative analysis to find themes and say that the themes describe the prototype. You might also want to nail down the comparative context as that seems to be important in interpreting prototypes and changes in prototype. Assessing or manipulating comparative context seems even more difficult than capturing a group prototype.

I wonder if the Yankee example is not an example of a decentralized as much as it is an example of a common identity group as described in Sassenberg's article. Would the Red Socks be considered a common bond group? How do two baseball teams, presumably constructed by similar rules and in a similar context end up as two different kinds of groups. Perhaps it is not a simple dichotomy but in fact two separate continuums.

Finally, I am not clear on whether or not subgroups have different prototypes than the main group. And I don't understand the comment on page 73 "Some research suggests that an effective strategy for managing inter-subgroup relations within a larger group is to make subgroup and super-ordinate group identity simultaneously salient." What exactly does it mean to make two group identities "simultaneously salient". The example in the paragraph does not help me understand this concept at all.



LaurieBuis: Wow... what a dense reading. Even after reading through Jude's summary, I'm still quite fuzzy on a lot of the points made in this chapter :) To respond to KathyLee and PaulResnick on group size, I tend to agree with Paul that I can see a lot of these phenomena existing on all levels of aggregation. After reading through this chapter, I can see a lot of these phenomena existing in my classes, in my social circle, in the workplace, and in larger levels such as the university, etc.

To respond to CaRichardson on her last point regarding the above printed statement on page 73, this was the example that I thought of to attempt to gain clarification... I may be wrong, but, this is how I read it... Think of a high school where there is a super-ordinate group identity, say the East Lansing Trojans... and subgoup identities such as the drama club, the jocks, etc. Having all been through high school, we know that there is a lot of strife between the different subgroups. People are very proud to be a jock or in the marching band. In order to diminish the strife between the subgroups, we need to make "subgroup and super-ordinate group identity simultaneously salient". So, rather than having people focuse on the fact that they are an athlete or a band member, we need to have them focus on the fact that they are a Trojan athlete or a Trojan marching band member. It seems to me by tying this commonality in to group identity, there will be less social strife... Again, I might be completely on the wrong path, but this is how I interpreted it...