Automatic and Controlled Processes in Social Cognition 265more abstract and stable representations emerge. Park, Judd,
and Ryan (1991) proposed that representations of groups to
which one oneself belongs (i.e., in-groups) tend to contain
more information about specific exemplars, whereas repre-
sentations of out-groups tend to consist mainly of generaliza-
tions (see also Sherman, Klein, Laskey, & Wyer, 1998).
Hybrid representational models such as these provide even
more theoretical flexibility because they can account for both
the flexibility of social cognition that is emphasized in the ex-
emplar approach and the stability (sometimes even rigidity)
of social cognition that is emphasized in models relying on
more generic forms of representation such as schemas.
Distributed Memory Models
As Smith (1998) notes, one potentially important distinction
between exemplar models and earlier approaches to repre-
sentational mechanisms lies in the fact that the schema and
associative network approaches suggest the existence of a
discrete, stable, enduring cognitive entity—a thing that is
stored, accessed, used, and stored away again. In contrast, the
exemplar approach suggests the dynamic construction of rep-
resentations on the spot, depending on which particular ex-
emplars happen to get recruited for processing in a particular
context. After such a representation is formed from its con-
stituent exemplars, it does not remain as a stable entity;
rather, it is used and then deconstructed back into its underly-
ing elements, which may never come together in quite the
same way again. From this perspective, exemplar-based
representations are more like a transitory state than like an
enduring entity. This idea is taken even further in the latest
representational mechanism to catch the fancy of social-
cognitive researchers: parallel distributed memory (often
called PDP) models.
Based on models of distributed cognition developed by
cognitive psychologists, this approach to mental representa-
tion has been developed in a rather elaborate manner. An ex-
cellent summary geared toward social-cognitive researchers
was provided by Smith (1996). A detailed presentation of the
assumptions of this approach is beyond the scope of this re-
view, but a good general sense for the ideas embodied in this
way of thinking is provided by a simple analogy used by
Smith (1998). In distributed representations, a concept is rep-
resented as apattern of activationacross a set oflow-level
processing units,in much the same way that a television
screen image arises from a pattern of electrical activity occur-
ring across a set of pixels. Any individual pixel does not con-
vey meaningful information, but a very complex and richly
meaningful representation can emerge from the pattern cre-
ated by the activation’s occurring across multiple pixels.
Moreover, just as with the pixels in a television screen, the
same processing units can be involved in numerous, very dif-
ferent representations; meaning never resides in the process-
ing units per se, but instead emerges from the overall pattern
of their activity. According to this view, representations are
clearly momentary states rather than enduring entities. In fact,
whereas the other representational approaches make a dis-
tinction between representational structures and the processes
that operate on these structures, such a distinction is mean-
ingless in the PDP approach because the process of activation
itselfisthe representation from this perspective. According
to the PDP approach, representations are not stored and re-
trieved; instead, they are constructed and (approximately)
reconstructed based on the relevant input cues.
Extremely new to social psychologists, distributed-
memory mechanisms have not yet won widespread applica-
tion. Nevertheless, interest in their theoretical potential is
growing, and PDP-based accounts for a diverse range of phe-
nomena such as attribution, cognitive dissonance, and person
perception have begun to appear (e.g., Kashima, Woolcock, &
Kashima, 2000; Schulz & Lepper, 1996; Smith & DeCoster,
1998; Van Overwalle, 1998). As Smith (1998) describes in
careful detail, the PDP approach can provide an impressive
degree of theoretical coherence because it can account for vir-
tually all of the phenomena previously explained by asso-
ciative network, schema, and exemplar models. Ultimately,
Smith makes a convincing case that these alternative versions
of mental representation are not really competitors, but in-
stead are complementary windows, each with its own particu-
lar theoretical usefulness.AUTOMATIC AND CONTROLLED PROCESSES
IN SOCIAL COGNITIONA great deal of social cognition theory and research is con-
cerned with questions about the degree to which social infor-
mation processing involves active, conscious analysis of the
social environment. Historical models of person perception
and attribution regarded the perceiver as operating as a “lay
scientist” (e.g., Heider, 1958; Kelley, 1967), examining evi-
dence and reasoning about its logical implications; research
in this tradition was largely mute, however, with respect to
whether these putative mental processes involved the con-
scious application of deductive principles or processes of a
more preconscious variety. As Gilbert (1998) observes, it is
quite possible for a mental system to follow a reasoning
algorithm without requiring that the conscious mind know or
consciously apply the relevant principles. Mental processes
that do not involve active, conscious ratiocination have come