Four degrees of stereotype formation:
differentiation by any means necessary
Stereotype formation is a strangely neglected topic within social psychology. Research in the social cognition tradition tends to treat stereotypes as givens: cognitive heuristics, which are part of our mental repertoire (or cognitive toolbox), that are activated and then applied (e.g., Gilbert & Hixon, 1991). This two-step process already presumes the fact of formation. So, although this general approach has proved to be of considerable heuristic value, it is not clear that this tells us too much about how stereotypes are formed in the first place. This approach kicks in after the stereotypes are in place so to speak. There has also been some debate in recent years about whether stereotypes are constructs we can always take 'off the peg' in this pre-packaged way (see e.g., Spears & Haslam, 1997). In this chapter we have our cake and eat it: we accept that stereotypes sometimes represent well learned knowledge structures about social groups that are simply activated, but at other times they have to be constructed in context from the resources available. However, if we are to consider stereotype formation it is surely self-evident that we cannot rely exclusively on prior knowledge or fixed structures.
Arguably if we want to look at how stereotypes are formed, we have to consider more dynamic and social processes that govern the relations between groups: how meaning is extracted, constructed and developed over time. This does not mean to say that this always refers to the beginning of group life; stereotypes can develop and change with the comparative context. This side of stereotyping has not been neglected. Social identity and self-categorization theories, with their common theoretical tradition, have had much to say about how we extract and develop the meaning that forms the basis for group identity and social stereotyping. One strength of these approaches is that they are dynamic. They involve intergroup comparison and therefore vary with context, with the result that they do not take stereotypic content as fixed. Below we contrast this view of stereotyping and stereotype formation with the activation/application approach