Carlston, D. E., & Smith, E. R. (1996). Principles of mental representation. In: E. T. Higgins & A. W. Kruglanski (Eds.) Social psychology: Handbook of basic principles. New York: Guilford.
Summary: (introductory paragraphs of chapter)
Psychologists have long believed that people's mental representations of the social world are important (e.g., Kelly, 1955). Such representations are implicated in the phenomenological "life space" of Kurt Lewin's field theory (1951a), the clashing cognitions in Leon Festinger's dissonance theory (1957), the conceptual triads in Heider's balance theory (1958), as well as in the beliefs and evaluations in other attitude theories (e.g., Fishbein & Ajzen, 1975) and the perceptions and attributions underlying attribution theories (e.g., Jones & Davis, 1965). However, for much of this history, mental representations were simply "intervening variables" (MacCorqudale & Meehl, 1948), assumed and defined for the convenience of particular theories (Ostrom, 1984). In less formal accounts, mental representations were sometimes described colloquially in terms of introspectively available feelings and memories (e.g., Heider, 1958).

With the advent of modern cognitive science, mental representations began to be viewed as empirically tractable concepts with an existence outside of a particular, isolated theoretical approach. One consequence of this new realism was a series of debates over the mental "codes" that actually underlie the representation of different forms of information. Although these debates have not thus far had a major impact on social cognition, the field has embraced the underlying premise that mental representations are more general than individual theories and more tangible than either intervening constructs or introspective allusions. Social psychologists now speak of mental representation in terms closely paralleling those used in cognitive psychology, speculate about the existence of multiple forms of social representation, invent or adapt new methods for examining social representations, study the spontaneity with which different kinds of representations are generated, and explore the relationships of various cognitive processes to mental representation (Wyer & Carlston, 1994).

The result of all this activity is an extensive literature on social representation that touches base not only with enduring social psychological concepts such as attitude and attribution, but with cognitive constructs such as iconic storage and episodic memory. This literature helps to illuminate the phenomenology of the social perceiver, but it also relates social representations to cognitive representations more generally. The goal of this chapter is to extract from this literature five basic principles that seem to us to characterize and to clarify contemporary thinking about the nature of social representations. Some of these principles (e.g., the accessibility principle) are so central and obvious that they merit their own chapter in this volume, whereas others (e.g., the mental mush principle) have not been framed before in quite the way that we do here. Overall, however, we believe that these 5 principles capture the essence of what social psychologists have learned about mental representation thus far.

The first principle--that there are multiple forms of cognitive representation--is so central that once it is acknowledged, the need for the remaining four principles seems obvious. That is, if multiple forms of representation exist, this raises general issues regarding their origin (Principle II: they are by-products of different cognitive processes), their interrelationships (Principle III: they are interconnected), their influence (Principle IV: accessibility plays a key role), and the manner in which they are used and experienced by perceivers (Principle V: they blend together inseparably in a kind of mental mush). In a sense, then, these 5 principles constitute a coherent statement about the multi-faceted nature of people's cognitive representations.

In addition to the notion of multiple forms of representation, two other themes cross-cut our discussion of the five basic principles. First, it appears that new, more sophisticated models of representation are needed to accommodate many of the complexities raised here. Social cognition theorists have tended to selectively exploit relatively simple cognitive models, without wishing to delve too deeply into the less accessible facets of more recent models of cognitive representation. From the broader perspective that we adopt here, however, these recent models seem to have a great deal to offer. Second, it appears to be generally important how perceivers interpret the mental mush that constitutes their perception, memories and judgments. In other words, in recalling memories and making judgments, perceivers not only retrieve a mental mush composed of different cognitive representations, but they make attributions about how and why those particular representations came up. To fully understand the implications of the mush for a perceiver, it is often necessary to explore the perceiver's meta-cognitions regarding the source or import of that mush. We shall return to these themes several times in our discussion.