262 Social Cognition
behavior is encountered (e.g.,won the citywide chess tourna-
ment), it is assumed to be linked directly to the central concept
(Tina) by a relatively strong node, because it fits with precon-
ceptions about this target quite well. However, when an in-
consistent behavior is encountered (e.g.,got confused trying
to figure out the subway system), it is assumed to be linked to
the central concept in a more tenuous way because it does not
really fit with the general image of the target. However, the in-
congruity embodied in the inconsistent behavior is assumed to
provoke efforts to resolve the confusion by thinking about
how the inconsistent behavior might make sense in light of
other known facts. This triggers the formation of inter-item
associations among the different behavior nodes. Thus, al-
though inconsistent behaviors are likely to be less strongly
linked to the central person concept than are consistent behav-
iors, the inconsistent behaviors are actuallymorelikely to be
linked to a variety of other behavior nodes. As a result, the in-
consistent behaviors tend to be more memorable on average
because they tend to have more associative links with other
items, producing a greater number of pathways through which
activation can spread into them and draw them into working
memory.
We have presented only the most general statement of how
associative network models have been applied in the domain
of social memory. Specific theoretical approaches have been
much more elaborate in their assumptions—although they
still share the key core assumptions that we have outlined.
This general approach has been used to predict a wide and
impressive array of empirical phenomena concerning social
memory, including the serial order of information retrieval in
free recall (e.g., Srull, 1981) and the influence of different
processing goals and levels of attentional capacity on the
probability of recalling inconsistent versus consistent infor-
mation (e.g., Srull, Lichtenstein, & Rothbart, 1983; for a
comprehensive review, see Wyer & Srull, 1989). Research in
these and several other topic areas confirm the explanatory
power of the relatively simple assumptions embodied in the
associative network approach.
Schemas
A rather different view of the nature of human understanding
emerged in Continental philosophy (particularly the ideas of
Kant). From this perspective, simple associations are inade-
quate to account for the complexity of human cognition.
Instead, it is assumed that knowledge is organized into more
elaborately structured conceptual representations. This ap-
proach to mental representation is epitomized in schema
theories.
Originally introduced prominently into psychology by
Bartlett (1932), schema theories focus on the role played
by generic knowledge structures that organize a person’s un-
derstanding of a particular domain. A schema can be thought
of as a subjective theory (Markus & Zajonc, 1985) that is for-
mulated to account for the generalities of one’s experience.
The elements of the schema are typically thought to be orga-
nized by more than simple association. For example, spatial,
temporal, logical, and causal relations constrain and provide
coherence to the schematic structure. To take a simple exam-
ple, a faceschema consists not only of a set of elements that
are associated with faces (e.g., eyes, nose, mouth), but also of
rules about the spatial relations among these elements. This
general understanding of what faces are like is assumed to
have been abstracted from experience with numerous specific
faces over time. In addition to this inductive pathway to
schema formation, it is often assumed that schemas can be
learned in a more top-down manner. For example, most
schoolchildren could, one hopes, articulate a rather detailed
mammalschema, although they have most likely not induced
its elements by observing particular instances. Instead, they
have learned directly what the core elements of the schema
are and how these elements are related to one another.
The elements contained in a schema often function like
variables that can take a variety of values, provided that they
adhere to the fundamental constraints of the schema. For ex-
ample, there is a range of acceptable colors and shapes that
eyes can take, but they must invariably be located above the
nose, contain a pupil and an iris, and so on. This observation
points to the fundamental function of schemas: They serve as
templates for understanding experience by providing preor-
ganized, general-purpose understandings that can be adapted
to the particulars of the current situation via instantiation.It
is assumed that schemas will be activated spontaneously in
situations in which they are relevant, and that this activation
occurs in an all-or-none fashion. Thus, unlike the associative
network models (in which some nodes in a network can be
active while others are not), schema models assume that if
any part of the schema has been activated, then the rest of the
schema will also be activated.
Schemas are thought to fulfill a variety of functions (for
a review, see Bodenhausen, 1992). Most notably, they provide
a basis for making inferences about unspecified elements of a
stimulus or situation, and they can guide the interpretation of
ambiguous features as well. Activated schemas also tend to
guide the processes of perception and memory toward infor-
mation that is relevant to the particular schema. One famous
demonstration of the operation of schemas was provided by
Bransford and Franks (1971), who showed that memory for
ambiguous verbal stimuli (e.g.,the notes were sour because
the seam was split) was substantially enhanced when a relevant
schema was activated that would allow for the disambiguation
of the sentence (in this example,bagpipe). As Bartlett (1932)