268 Social Cognition
example, Bargh et al. showed that activating stereotypes about
elderly persons resulted in slower rates of walking. Similarly,
Chen and Bargh (1997) showed that subliminal presentation
of African American (as compared with European American)
faces resulted in more hostile behavior in a subsequent verbal
game played with an unprimed partner. Moreover, the un-
primed partner’s behavior also became more hostile as a con-
sequence, showing that self-fulfilling prophecies can emerge
in a very automatic manner—even when participants are un-
aware that stereotypical concepts have even been activated
and have formed no conscious intention to act in a manner
consistent with these concepts. Although the precise mecha-
nisms responsible for these fascinating effects have not been
isolated, the very existence of the phenomenon provides a
potent demonstration of the potential automaticity of not only
social thought, but also interpersonal interaction.
A principal advantage of automatic reactions lies in the
fact that they are largely not dependent on the availability of
processing resources. Because of the great efficiency with
which they unfold, automatic processes do not require much
investment of attentional capacity or perceiver motivation.
Whereas novice drivers can find it harrowing to coordinate all
of the requisite activities (shifting gears, monitoring traffic,
steering, braking, etc.), after the process has been automated,
not only can these tasks be easily performed, but the driver
may also have sufficient reserve capacity available for
singing along with the stereo or engaging in mobile phone
conversations. Empirical confirmation of the resource-
conserving properties of automatic mental processes was pro-
vided in a series of experiments by Macrae, Milne, and
Bodenhausen (1994). In one of their studies, they asked
participants to engage in two tasks simultaneously: a visual
impression-formation task that involved reading personality
descriptions of four different persons, and an audio task that
involved listening to a description of the geography and econ-
omy of Indonesia. For half of the participants, stereotypes
were activated in the impression-formation task (by provid-
ing information about a social group to which each target be-
longed). Some of the personality information was consistent
with stereotypes about the relevant group, and the rest was ir-
relevant to such stereotypes. One might expect that giving
these participants an additional piece of information to inte-
grate would simply make their task all that much harder—but
in fact, the introduction of the stereotype provided a frame-
work that participants could spontaneously use to organize
their impressions, making the process of impression forma-
tion much more automatic and efficient. As a consequence,
participants who knew about the group memberships of the
social targets not only recalled more information about the
targets (as revealed in a free recall measure), they also learned
more information about Indonesia (as revealed in a multiple-
choice test). The automatic reactions triggered by stereotype
activation provided a clear functional benefit to perceivers by
making the process of impression formation more efficient,
thereby freeing up attentional resources that could be devoted
to the other pressing task.
When automatic effects of these sorts occur without
awareness, intention, or much attentional investment, is there
any hope of preventing them or stopping them after they start?
In the realm of automatic stereotyping effects, Bargh (1999)
has argued that the prospects for controlling such effects are
slim to none. Indeed, the final hallmark of an automatic
process is its imperviousness to control. In line with Bargh’s
assertion, the previously described research of Devine (1989)
showed that even low-prejudice individuals who disavow
racist stereotypes are still prone to showing automatic effects
of stereotype activation. Similarly, Dunning and Sherman
(1997) found that implicit gender stereotyping occurred inde-
pendently of participants’ level of sexism. However, other re-
search has begun to suggest that at least some of the time, it
may be possible to develop control over automatic processes.
Uleman et al. (1996), for example, found that with practice,
people could learn to avoid making spontaneous trait infer-
ences. Similarly, it seems that egalitarian individuals can also
learn to control automatic stereotyping effects, at least under
some circumstances (e.g., Wittenbrink, Judd, & Park, 1997).
It is toward the processes through which mental control can be
achieved that we now turn our attention.Controlled Social CognitionThe process of controlling thought and action, at least in rel-
atively novel and unpracticed domains, requires attention.
Whereas automatic processes occur efficiently and thus re-
quire little expenditure of mental resources, effortful, con-
trolled processes come with an attentional price to pay.
Moreover, controlled processes typically require intentional
deployment, and they occur in a manner that is at least par-
tially accessible to the conscious mind. Whereas many com-
putational processes of implicit cognition are regarded to be
massively parallel, attention and consciousness represent a
processing bottleneck that results in highly selective and
serial information processing (e.g., Simon, 1994). As Simon
notes, connecting one’s motives to one’s thought processes
requires a system that can cope with the constraints imposed
by limitations of attentional capacity.
Attentional capacity has turned out to be a major theoreti-
cal construct in social cognition research (for a review, see
Sherman, Macrae, & Bodenhausen, 2001) precisely because it
plays such a fundamental role in determining whether it