Home Curriculum Vitae Publications Conference Reports Forthcoming Extramural Colloquia Expert Testimony Teaching Healthcare The Human Ecology of Memory Research Archive Publications and Reports Rants


An Introduction to Social Cognition

John F. Kihlstrom

University of California, Berkeley

Note: This paper is based on a presentation originally given to the Berkeley Social Ontology Group, Fall 2010. I thank Jennifer Hudin for the invitation, and John Searle and other members of the group for their comments. For more thorough treatment of this subject, see the Lecture Supplements prepared for Psychology 164, "Social Cognition" (write jfkihlstrom@berkeley.edu for access).

And no, it's not by Renoir -- though that's a good guess. It's In the Orchard (1891) by Edward Tarbell, an American impressionist painter, now in the Terra Museum of American Art.


The cognitive point of view in social psychology is simply this: the individual's social behavior is not determined by the situation he or she is in; rather, it is determined by the person's perception of that situation, broadly construed to include relevant knowledge and memory, reasoning, judgment, problem-solving, and decision-making.

While the point may seem obvious, it has not always been obvious to everyone in the field. The cognitive point of view has its origins in the "Thomas Theorem" that ""If men define situations as real, they are real in their consequences" (Thomas & Thomas, 1928, p. 529); Bartlett's (Bartlett, 1932,p. 3)) dictum that "The psychologist, of all people, must not stand in awe of the stimulus"; and the fundamental proposition of symbolic interactionism, that "Human beings act toward things on the basis of the meaning that the things have for them" (Blumer, 1969). However, these points were neglected during the heyday of behaviorism in psychology in general, when social psychology embraced Allport's (Allport, 1954) definition of the field as the study of social influence and a Doctrine of Situationism that was allied with the behavioristic denial of the importance of internal, mental processes (Zimbardo, 1999). The cognitive revolution of the 1960s, which began in the hands of psychologists interested in attention, learning, memory, and language (Baars, 1986; Hirst & Miller, 1988), restored the cognitive point of view to social psychology as well (E.E. Jones, 1985; Zajonc, 1980).

With all due respect to Allport, social psychology is no more the study of social influence than perception is the study of stimulus influence. Perception is the study of how people form internal, mental representations of the external world, and how they use these representations to guide their behavior. While most of psychology is concerned with the nature of mental life, and the role that mind plays in the behavior of the individual, social psychology studies the relation between mental structures and processes that reside in the mind of the individual to social structures and processes that reside in the world outside the individual. The influence is a two-way street the social environment surely shapes the individual's mind and behavior; but individuals also shape their social environment through processes of cognitive construction (Kihlstrom, 2010a).

Psychology, including social psychology, embraces the Doctrine of Mentalism -- that mental states stand in relation to action as cause to effect. As a behavioral science, psychology explains behavior in terms of the individual's mental states. But psychology is uniquely positioned as both a biological science, with an interest in uncovering the observer-independent principles that underlie mental life, and as a social scientist, concerned with the observer-dependent meanings that the individual ascribes to objects and events.


Cognition in the General Social Interaction Cycle

Fig4_GSIC.JPG (56518 bytes)The role of cognition in mediating the person-situation interaction is illustrated by the general social interaction cycle (Cantor & Kihlstrom, 1987; Kihlstrom, 2010) -- a conceptual framework for analyzing any dyadic social interaction (Figure 4).  Within this framework, the two participants are assigned the role of Actor and Target, respectively.  This assignment is of course somewhat arbitrary, because each individual is both an actor and the target of the other's actions.  For convenience, the Actor role is assigned to the individual who initiates the social interaction.

The general social interaction cycle is a variant on the general social interaction sequence initially described by Darley and Fazio (1980) and Jones (1986).  The description of social interaction as a cycle rather than as a sequence is intended to capture the bidirectional causality of reciprocal determinism, and to make clear that social exchange continues until one or the other participant terminates the interaction by leaving the situation.  An alternative depiction is in terms of a connectionist network representing the interaction of two cognitive-affective processing systems, representing the two individuals that constitute the dyad (Shoda et al., 2002; Zayas, Shoda, & Ayduk, 2002).

First, the Actor enters the situation -- the immediate context in which he or she physically encounters the Target (from this point on, for simplicity in exposition, we'll call the Actor "she" and the Target "he".  The Actor enters the situation with some goal in mind, something that she wants to accomplish, like asking the Target for a date for Friday night.  The Actor also carries into the situation a fund of social knowledge concerning herself and the target.  How badly does she want a date?  Does she know whether he is currently seeing someone else?  Does she have any reason to think he might be interested in her?  The Actor also carries a fund of more generic social knowledge relevant to her current goals:  What movies are in town?  Are there any parties?  And finally, the Actor carries a repertoire of skills to be used in the course of the interaction, such as how to start a conversation, and how to bring it around to the subject of Friday night.  Some of these skills are cognitive in nature, such as her ability to "read people"; others are motoric, such as a particular way of walking, or using her hands.  This sort of social knowledge is constitutes the individual's fund of social intelligence. 


As she begins the interaction, the Actor forms an impression of the situation -- of the target, and of the immediate environmental context: Does he still seem interested?  Is this a good time to ask?  This impression combines knowledge derived from two general sources: information about the current stimulus situation, extracted through the mechanisms of social perception; and pre-existing knowledge about herself and the target, retrieved from social memory.


Finally, the Actor acts on the basis of her impression.  She may approach the target or shy away, she may pop the question or not.  If she does not ask the Target for a date, the interaction will end shortly.  If she does, the interaction will continue.  Assuming that the Actor has asked him for a date, attention now shifts to the Target, who now has to do something in response to the Actor's initial salvo.


The Target enters the situation -- either actively, by approaching and greeting the Actor when he sees her, or passively, by being approached and greeted by her.  He too brings his social intelligence into the situation.  


The Target forms an impression of the situation in which he now finds himself -- a situation that is immediately clarified when the Actor asks him for a date.  The Target knows he's free Friday night, because the woman he's been dating is out of town, but that's not decisive.  Should he play hard to get?  Should he wait to see if he gets a better offer from someone else?  What if his current girlfriend finds out?


On the basis of the impression he's formed, the Target responds. He decides to keep his options open for Friday night, but doesn't want to spurn the Actor entirely, so he says he can't see her Friday, but proposes that they go out on Saturday instead.  


Now attention shifts back to the Actor.


The Actor must interpret the Target's response, and revise her impression of the situation accordingly.  Perhaps he's Jewish, or Muslim, and devout, and doesn't go out on Friday nights.  Perhaps he's seeing someone else.  Obviously he's got something he'd rather do on Friday, while she does not, and she has now clearly communicated this fact to him.  As it happens, she's also free Saturday night, but if she accepts his counteroffer she clearly communicates that she doesn't have a date for either night.  Should she let him have this information?  If she says "yes", is she becoming a pawn in whatever other relationship he may be pursuing?  Or is the "Friday-night woman" (because by now she is certain that he already has got a date for Friday night) a pawn in a new game that he is now playing with her?


On the basis of her impression, the Actor responds to the Target: she decides to take a chance, and accepts the date for Saturday night.  


Now the ball is back in the Target's court.


The Target must interpret the Actor's response, revise his impression, and figure out what to do next.  And so it goes, with the cycle of exchanges continuing.  Each participant is trying to make sense of what the other one is doing.  Each is trying to read the other's mind.  And each participant is planning and executing behavior in accordance with his or her evolving understanding of the total situation.


In addition, the General Social Interaction Cycle also transpires at another level, within each individual participant.  According to the Doctrine of Reciprocal Determinism, behavior does not simply affect the person toward whom it is directed; it also feeds back to affect the person who emitted the behavior.  The Actor may have wondered if she had the nerve, and the skill, to ask a man for a date.  Now she knows that she does (Bandura calls this kind of knowledge self-efficacy expectations).  Similarly, the Target may never have had to negotiate overlapping dating relationships.  Now he knows he can do this -- or else he's put himself in a situation where he has to learn how.


In any event, each participant in this social interaction is behaving in accordance with his or her construal of him- or herself, and of the other, and of the situation in which they meet.  Each of these construals is modified by the other's behavior, and his or her own.  And it's these individual construals, in the end, that lead the participants to behave the way they do.


The Scope of Social Cognition

Viewed broadly, "social cognition" embraces everything about the cognitive mediation of social interaction. But viewed more narrowly, social cognition refers to cognition of social objects -- people (including ourselves), the social situations in which we encounter them, and the interpersonal behaviors that transpire in those situations. Paraphrasing Jerome Bruner and Renato Tagiuri (J. Bruner & Tagiuri, 1954) (see also Tagiuri & Petrullo, 1958), social cognition places the "'knowing of people' in the wider theoretical context of how we know the environment generally". As such, the notional syllabus for a course in social cognition would look very much like that of a course in cognitive psychology, including perception, memory, categorization, judgment, language, learning, intelligence, development, and neuroscience (Kihlstrom & Park, 2002) -- except that the focus would be on people and what they do.

A typical experiment on perception might ask subjects to identify an object briefly presented on a computer screen. An experiment on social perception would ask subjects to recognize a facial expression of emotion.
An experiment on memory might ask subjects to study a list of unrelated words; in a study of social memory, those words might be the names of personality traits, or sentences describing social behaviors.

You get the idea. In what follows, I will attempt to give an overview of each of these areas, highlighting what I consider the salient issues in each.


Social Perception

Scientific psychology began under the influence of British empiricism, and the idea that all knowledge comes to us through sensory experience. Not unreasonably, then, the first scientific psychologists --psychophysicists like Ernst Weber and Gustav Fechner, and physiological psychologists like Hermann von Helmholtz and Ewald Hering -- focused their work on problems of sensation and perception. Not to put too fine a point on it, sensation has to do with the detection of distal stimuli in the environment, and the transduction of proximal stimulus energies into neural impulses that are transmitted to the brain; perception is the process by which the perceiver forms an internal, mental representation of the distal stimulus.

In the nonsocial domain, we perceive objects and their physical states: their form, their location, their motion, and what Gibson called their affordances -- that is, what the perceiver can do with the object.
Social perception has been studied mostly in the form of person perception, meaning persons and their psychological states -- their personality characteristics, beliefs, emotions, and motives. And, in a sense, we perceive their affordances, too -- what can I do with this person?

Bruner and Tagiuri, in an early analysis of person perception, listed a number of factors that influence perceptual organization, including the stimulus array itself, selective attention, linguistic categories, and especially the internal state of the perceiver -- his mental set, or expectations, and his own emotional and motivational state (set, emotion, and motive were the hooks on which Bruner had earlier hung is "New Look" in perception).

The stimulus array for nonsocial perception consists of the energy (light waves, sound waves, etc.) that radiates from the distal stimulus, falls on the sensory surfaces, and is transduced by receptor organs into neural impulses.
The stimulus array for person perception consists of the person's appearance and behavior, and also the language that others use to describe the person.

Traditionally, the psychology of perception has been dominated by a constructivist viewpoint that has come down to us from Helmholtz, and is represented in the modern era by such figures as Richard Gregory, Julian Hochberg, and Irvin Rock. For the constructivists, the information provided by the stimulus is vague, fragmentary, and ambiguous, and is not sufficient to support perception. Accordingly, the perceiver must draw on his expectations and knowledge of the world to fill in the gaps left by the stimulus array, and make inferences about the object and its states. This is what Helmholtz meant when he underscored the importance of "unconscious inferences" in such things as the perception of size and distance. And it is what Bruner meant when he argued that the perceiver must "go beyond the information given" by the stimulus. The study of person perception has followed suit, beginning with Solomon Asch's application of "Gestalt" principles to the problem of person perception.

In the late 20th century the constructivist view of perception was been challenged by an ecological view of direct perception offered by J.J. Gibson. According to Gibson, the stimulus array provides all the information necessary for perception, and the perceptual apparatus has evolved in such a way as to extract this information, permitting us to perceive the world without relying on "higher" mental processes of inference and judgment. This ecological view has also had an impact on the study of person perception. While most research on person perception has followed Asch's "impression-formation" paradigm in employing verbal materials (like trait labels) as stimuli, the proponents of the ecological view in person perception, such as Ruben Baron and Leslie McArthur, ecologists have argued that researchers should stick closer to the physical stimulus -- to the person's appearance and behavior. Accordingly, they have focused on such matters as the person's facial expressions; bodily orientation, movement, and posture; vocal cues; interpersonal distance; eye contact and touching; physical appearance and manner of dress; and also those aspects of the local behavioral environment -- the situational context -- that are under the target's control.

Personally, I side strongly with the constructivist view that the social perceiver must go "beyond the information given" in order to construct a mental representation of the target person. I also think that verbal cues (like trait adjectives) are underrated as stimulus information for perception -- after all, language is an important medium for knowledge representation, and we use language as a vehicle for social interaction. At the same time, I agree with Baron and McArthur that studies of social perception rely too heavily on verbal stimulus materials, and that we will not have a comprehensive understanding of social perception unless we spend more time studying how people actually present to us -- in appearance and in behavior, as well as in verbal description.

Most research o social perception has focused on describing the stimulus and the processes that operate on stimulus information to generate a mental representation of the target's mental states. An understudied issue is accuracy: to what extent are our mental representations of ourselves, other people, the situations in which we find them, and the activities that transpire there accurate reflections of social reality? The accuracy issue is usually framed in terms of judgments of personality traits: is a person judged to be neurotic or extraverted really neurotic or extraverted? This, in turn raises the question of the criterion: how do we know whether a person is really neurotic or extraverted? Typically, self-ratings by the target, or aggregate ratings by a number of judges, serve as criteria for validating an individual judge'sratings. The general adaptiveness of social behavior can be taken as evidence of the general accuracy of social perception -- or not, depending on your point of view. Still, the accuracy issue underscores the basic point: accurate or not, or behavior in a situation is determined by our perception of that situation. For a recent review, see Funder (2012).


Social Memory

Perceiving an object or event changes the contents of memory, by leaving some trace of perceptual experience that persists when the stimulus has disappeared. It is by virtue of memory that we are able to guide or behavior in accordance with past events, freeing us from reliance on the current stimulus environment -- and, by some accounts, the same faculty of memory that allows us to remember the past, also enables us to anticipate the future.

Ever since Ebbinghaus, psychologists have studied memory primarily by means of some variant on the verbal learning paradigm: the subject studies a list of words, or sentences, or a whole story, or pictures; this constitutes the encoding phase; then some interval elapses, which represents the storage phase; finally, in the retrieval phase, the subject attempts to remember what he studied. Precisely this same method has been used in studies of person memory, analogous to person perception, in which the stimulus materials are, typically, lists of trait adjectives that describe the target's personality, or sentences describing his behaviors. The basic verbal-learning paradigm has been used to explore how knowledge about specific people is represented in memory, and what the relations are between knowledge about a target in general, and his specific behaviors.

Terry Winograd, John Anderson, and others have characterized these sorts of memory contents as declarative knowledge -- factual knowledge (meaning that it can be true or false) that can be represented in sentence-like propositions such as Judy is intelligent or Judy won the chess tournament. And Endel Tulving has, in turn, distinguished between two forms of declarative knowledge stored in memory: episodic memory refers to knowledge of events that have a unique location in space and time (two things can't happen at the same time in the same place); semantic memory is more abstract and generic. The fact that Judy is intelligent is a piece of semantic knowledge, because it's about Judy in general; the fact that Judy won the chess tournament is a piece of episodic knowledge, because it refers to an event that occurred at a particular place and a particular time.

An important issue concerns how knowledge about persons is represented in social memory. Within an associative-network model of memory, persons can be represented as nodes, linked to other nodes that represent their typical features (semantic knowledge) and specific behaviors (episodic knowledge). One intuitively appealing model suggests that specific behavioral episodes are clustered around the generic pesonality characteristics that those behaviors exemplify. Another model, grounded in self-perception theory, holds that only episodic information is stored in memory, and semantic information is generated online, as it were, as needed. However, evidence from priming studies, and from studies of amnesic patients, strongly suggests that semantic knowledge about a person is stored in memory, and that it is stored independently of episodic knowledge.

For personality and social psychologists, a particularly interesting form of episodic memory is autobiographical memory -- a person's memory for his or her own life. Technically, all episodic memories -- including my memory for the episode in which I learned that Judy is intelligent -- are autobiographical in nature, because they refer to events and experiences in the individual subject's life. And, conversely, all autobiographical memory is episodic in nature, because it is a record of discrete events and experiences. But autobiographical memory isn't merely episodic: there's more to it than just a list of personal events and experiences (Kihlstrom, 2009a). In the first place, autobiographical memory has a very explicit self-reference. In the standard verbal-learning experiment, the items to be remembered are words or pictures that the person has studied, and so each item in episodic memory implicitly contains some degree of self-reference. But autobiographical memories very explicitly refer to the rememberer, as the agent or patient of some action, or the stimulus or experiencer of some state; they also will refer to the internal mental context of the event in question: what the subject was thinking, feeling, and wanting at the time. Autobiographical memories also have what might be called an Aristotelian plot structure: first, they are organized in a sort of temporal sequence, with each one following another; more important, autobiographical memory represents the causal relations among individual events.

The social knowledge stored in memory is not limited to our declarative knowledge of particular people. We also store declarative knowledge about people in general -- and this knowledge, too, forms part of the cognitive background for processing particular events and experiences. Bruner and Tagiuri referred to this as the perceiver's implicit personality theory -- the assumptions that the person makes about human nature (e.g., the role of nature vs. nurture), fundamental dimensions of individual differences (e.g., the Big Five personality traits -- or, as I like to think of them, "The Big Five Blind Date Questions"), assumptions about their central tendencies and variability (are most people neurotic, or extraverted?) and the correlations among them (do socially desirable characteristics tend to go together, as the "halo effect" would suggest?).

In addition to this declarative knowledge, people store in memory their procedural knowledge of the skills and rules which they use in the course of adaptive behavior. Some of these skills are motoric, in that they guide our actual overt behavior as we tie our shoelaces or drive a standard-shift car; others are cognitive, in that they guide our covert, mental activity when we engage in inductive or deductive reasoning, or find square roots without using a calculator. In much the same way, our repertoire of procedural social knowledge enables us to navigate in the social world. How do you behave at a funeral? How do you liven up a party? How do you handle an obstreperous customer in a restaurant, or a bully in a bar, or a disruptive student in class? How do you ask someone out on a date, and how do you say no to someone who asks you out? How strong should your grip be when you shake hands with a stranger? How close should you stand when you talk to someone?


Social Intelligence

Taken together, our fund of declarative and procedural social knowledge constitutes our repertoire of social intelligence (Cantor & Kihlstrom, 1987; Kihlstrom & Cantor, 2000, 2010). Traditionally, social intelligence has been construed as a sort of "social IQ", meaning that some people are socially smart while others are social geniuses, and that these individual differences can be measured by the social equivalent of standard intelligence tests. It is this view of social intelligence that has been revived by Daniel Goleman (Goleman, 2006), as an extension of his best-selling books on emotional intelligence. And it raises the question of whether social intelligence is a set of distinctive abilities -- or just, as David Wechsler himself claimed, "just general intelligence applied to social situations".

Goleman's view is what Nancy Cantor and I have characterized as the "psychometric" or "ability" view of social intelligence. As an alternative, we have proposed a "knowledge" view of social intelligence, which simply refers to the individual's fund of knowledge about the social world. In contract to the ability view, the knowledge view does not characterize social intelligence as a trait, or group of traits, which can be measured, on which individuals can be ranked from low to high, and thus compared one to the other. Rather, the knowledge view of social intelligence simply assumes that social behavior is intelligent, in that it is mediated by what the person knows (or believes to be the case) about the social world. Individual differences in social knowledge lead to individual differences in social behavior. But it does not make sense to us to construct measures of social IQ. The important variable is now how much social intelligence the person has, but rather what he knows about himself, other people, the situations in which he meets them, and the behaviors they exchange.


Social Learning

One consequence of psychology's emergence from British empiricism was its early emphasis on sensation and perception. Another was its emphasis on learning -- the fact that knowledge is acquired through experience. For much of its history, the psychology of learning was the psychology of animal learning, and the animals learned in isolation -- Pavlov's dogs in their harnesses, Thorndike's cats in their puzzle boxes, Skinner's pigeons in their operant chambers, etc. And most of what they learned was about how to get food, or how to predict and control avoid shock.

As it happened, and despite the distractions of radical behaviorism, we learned quite a bit about how animals learn, and much of that proved to be generalizable to the human case. But at the same time, Neal Miller (a psychologist working in the traditional of Clark Hull's drive-reduction theory) and John Dollard (a sociologist) advanced the concept of social learning to account for complex human behavior.  They asserted that most learning is about social behavior, and most learning takes place in a social context.  Later, Albert Bandura expanded on their view, distinguishing between learning via direct experience and vicarious learning via observation.

Learning by direct experience is the usual "trial and error" learning, familiar from studies of classical and instrumental conditioning.  Pavlov, Thorndike, and practically everyone else who studied learning well into the 20th century, all thought that the animal was passive during conditioning -- which is where the term "conditioning" comes from.  And that learning was a simple matter of forming associations between stimuli and responses, following the principles of the association of ideas originally laid down by John Locke and David Hume, and others affiliated with British empiricism. One exception was Edward C. Tolman, who argued that even animals were actively engaged in the learning process, trying to figure out what is going on in their environment, and what to do about it -- a position ridiculed by some who characterized Tolman's maze-running rats as "lost in thought at the choice point".  But we now know that Tolman got it right.  The learning organism is actively generating, testing, and revising expectations and hypotheses about its world; when it learns, it learns to predict and control events in its environment.  That's true for dogs and cats, rats and pigeons -- and it's also true for humans. We are always trying to make sense of the social world, and we do that, in large part, by generating hypotheses based on our current understanding, testing them against reality, and revising our understanding accordingly.

Bandura, for his part, was more interested in a different type of learning, one that had also been described by Miller and Dollard -- observational or vicarious learning -- not by direct experience, but by observing other people.  We do learn by direct experience, but we also learn from other people, and learning from other people is by far the more efficient.  If you stick your finger in an electrical socket, you'll eventually learn not to do that; but if I tell you not to do it, you'll learn a lot more quickly.  It's observational learning, learning from other people, which lies at the heart of social learning theory.

Bandura further distinguished between two major forms of social learning. Learning by example is learning by observing other people.  This includes various forms of imitation, some of which occur automatically and unconsciously, as well as conscious deliberate, modeling. Young children learn gender roles by observing other people who look like them (and, for that matter, others who don't). Learning by precept includes both informal and formal, sponsored teaching.  Many social interactions, involve one person in the role of teacher and the other in the role of learner.  And society has developed a wide variety of institutions to generate and conserve knowledge, and transmit it to the next generation -- not the least of which is the college and university.  There is no more efficient way for students to learn, I think, than from a well-organized course with a well-written textbook. Unfortunately, there's also no more efficient way for people to acquire maladaptive social knowledge in the form of prejudices and stereotypes. As Rogers and Hammerstein wrote in South Pacific, "You've got to be [carefully] taught to be afraid of people whose eyes are oddly made, and people whose skin is a different shade".


Social Judgment

From a constructivist point of view, both social and nonsocial percepts and memories are influenced by what the person thinks. The role of reasoning, problem-solving, judgment, and decision-making in perception and memory -- even, according to signal-detection theory, down to the most elementary sensory process of detecting the presence of a stimulus -- moots the classical distinction between "lower" and "higher" cognitive processes. But judgment, reasoning, problem-solving, and decision-making are particularly prominent in two critical aspects of social cognition, categorization causal attribution.


Impression Formation

One important class of social judgments has already been discussed, under the rubric of person perception. Whereas some "neo-Gibsonian" advocates of the ecological view of perception insist that person peception merely unpacks information provided by the stimulus, the classic view, grounded in constructivist and Gestalt theory, argues that the perceiver makes his own contribution to the perceptual process, by supplying generic and specific knowledge retrieved from memory. This memory-based knowledge is then combined with stimulus-based knowledge in what is, ultimately, an act of judgment. Memory-based and inferential processes are arguably more important when stimulus information is itself represented linguistically -- as in typical forms of personality description.


Social Categorization

Bruner noted that (I'm paraphrasing here) every act of perception involves an act of categorization: in the course of identifying an object we naturally make inferences about its category membership, and decisions about category membership allow us to make inferences about unobserved features of the object. This is no less true in social perception than it is in nonsocial perception, and probably more so -- because social stimuli are so vague, fragmentary, and ambiguous. The importance of categorization is especially clear in the case of person perception, which also goes by the name of impression formation. For example, we speak of personality types such as melancholic or sanguine, extravert and introvert, represented linguistically by nouns. And we speak of personality traits, which are essentially categories of behaviors, represented linguistically by adjectives as extraverted and conscientious. People who display a lot of extraverted or conscientious behaviors may in turn be labeled as highly extraverted or highly conscientious, which are continuous dimensions rather than discrete pigeonholes, but nevertheless perform the cognitive function of categories. When we form an impression of a person, we do so by classifying his characteristic behaviors in terms of traits, and then assigning him to a type based on the traits he displays in his behavior.

Most studies of nonsocial categorization have been concerned with the structure of categories in the abstract -- as in the debate between the classical Aristotelian view of categories as proper sets, and the revisionist view of categories as fuzzy sets, the debate between the prototype view and the exemplar view, and the dispute over the role of similarity in categorization. Content has been almost irrelevant to these psychologists -- as evidenced by their willingness to use artificial as well as natural categories to test their theories. But this begs the question: what are the natural categories of persons? Following the suggestions of Roger Brown (Brown, 1980) we can list a number of possibilities, including sex (gender), kinship, age, socioeconomic status, sociopolitical affiliation, national, racial, and ethnic origin, and personality -- as well as categories specific to local culture, such as the jocks, preppies, and goths who can be found on any college campus. When we perceive people, we perceive them through these and other categorical filters; and once we've assigned them to a category, we have some idea of what they're like.

Of course, some of these represent stereotypes, but that's the point -- a stereotype is a category. Walter Lippman (1922) defined a stereotype as "an oversimplified picture of the world, one that satisfies a need to see the world as more understandable than it really is", and that's exactly what a category is. A stereotype is a set of beliefs, shared by members of some ingroup, about the character of some outgroup, and it's applied to all members of that outgroup, regardless of their actual characteristics. The belief is that members of a certain group share features in common, so that identifying a person's group membership permits us to make inferences about that individual's characteristics. Stereotypes are insidious because they serve as a basis for emotional prejudice and behavioral discrimination by ingroup members toward outgroup members, so they have negative implications that don't arise when we categorize objects as tools or animals, but that just shows that social categories are special in some ways.

Psychiatric diagnoses are also social categories. Here, symptoms such as anxiety or impaired reality-testing are features that allow us to classify a patient as neurotic or psychotic, and then into subcategories such as schizophrenic or manic-depressive, and sub-subcategories such as hebephrenic vs. paranoid schizophrenic. This is not to get into a debate about the biological reality of various forms of mental illness. It's simply to note that whatever else they are, psychiatric diagnoses are social categories that one group of people use to label another group of people, and that these labels have consequences for the people so labeled.


Causal Attribution

Causal attribution has to do with the explanation of behavior -- not the explanations that scientific psychologists come up with, on the basis of controlled experiments, but what Fritz Heider referred to as phenomenal causality -- how causation appears to the ordinary person in the street. In Heider's view (which you can also find in the developmental theories of Jean Piaget, and elsewhere) the person acts as a kind naive scientist, generating and testing hypotheses about the way the world works. One prominent early theory of causal attribution, Harold Kelley's covariation calculus of causal attribution,, made this analogy explicit. Kelley proposed that, over many observations of interactions between an actor and a target, people extracted information about the consensus of behavior across actors, the consistency of behavior across targets, and the distinctiveness of the actor's behavior toward the target; and then they entered this data into a kind of informal analysis of variance in order to determine whether the cause of the behavior in question to something about the actor (for example, his personality traits), something about the target (again, typically his personality traits), or to the situation in which the actor and the target encountered each other. Just what a scientific psychologist would do, only without SPSS.

Because the target is part of the situation, studies of causal attribution quickly focused on a dichotomy between internal, personal causes of behavior, such as personality traits, and external, situational ones, such as the presence of others. And, just as quickly, researchers began to discover that people didn't make causal attributions according to the rules. There then developed a rather large literature documenting "errors and biases" in causal attribution and other aspects of social judgment, the most famous of which were the fundamental attribution error (people attribute behavior to the actor's personality dispositions, and ignore or downplay the role of the situation), the self-other difference (people make dispositional attributions about others' behaviors, but situational attributions about their own) and the egocentric bias (people make personal attributions about their successes, but negative attributions about their failures. (A Google search on "errors and biases" will retrieve a Wikipedia list of more than 100 cognitive errors and biases, from the "actor-observer bias" to the "zero-risk bias".)

Unfortunately, this entire classic literature on causal attribution got off on the wrong foot. Part of this was due to a misinterpretation of Kurt Lewin's "grand truism", B=f(P, E), as justifying a dichotomy between personal and environmental causes of behavior (and, for that matter, an institutional separation between personality and social psychology) -- when, in fact, Lewin was an interactionist who insisted that the person and the situation were inextricably intertwined (Kihlstrom, 2010a). Just as important, there was a misunderstanding of Heider's own seminal work -- ignoring his distinction between intentional action and unintentional behavior, and his construal of "personal dispositions" as mental states rather than as personality traits (Malle, 2008).

Malle and his colleagues have now gone back to the drawing board, and produced a theory of causal attribution that avoids the Procrustean bed of the person-situation dichotomy, and more faithfully represents what actually goes on when people make causal attributions (Malle, 2004). One result is that the actor-observer difference in causal attribution takes quite a different shape from what is presented in textbooks. Another is a complete re-evaluation of the fundamental error. From a psychological point of view, intentional behavior is always caused by the person's mental states. Thus, the fundamental attribution error is not an error; but it is fundamental.


Moral Judgment

Among the many judgments that are studied in social cognition, perhaps the most social are those that involve morality -- is a person good or bad, is a behavior right or wrong? Early accounts of moral judgment emphasized rational processes, as exemplified by Lawrence Kohlberg's neo-Piagetian theory of moral reasoning. More recently, reflecting the "affective counterrevolution" in social psychology, an approach known as social intuitionism has emphasized the role of irrational, or at least non-rational, "gut feelings" (e.g., Haidt, 2001). Link to a paper on "Reason and Emotion in Moral Judgment".


The Language of Social Interaction

Language permeates the study of social cognition, and not just because social cognition researchers favor verbal stimulus materials. Noam Chomsky correctly argued that language was not just a tool for communication, but also a tool for thought. But that doesn't mean that language isn't also a tool for communication. We use language to represent our thoughts about other people, and language is an important medium of social interaction -- which is why Roger Brown (Brown, 1965) devoted a major section of his classic textbook on social psychology to the subject.

First, as any inspection of the "personals" ads in your favorite periodical will confirm, we tend to describe ourselves, and others in the language of traits. Allport and Odbert famously counted 17,953 words in an unabridged English dictionary that could be used to capture individual differences between people (there were actually 17,954 such words, but they miscounted). A long tradition of factor-analytic research has reduced this mass to the "Big Five" dimensions of neuroticism, extraversion, agreeableness, conscientiousness, and openness to experience; or perhaps, the "Big Two" dimensions of intellectual and social desirability. Lewis Goldberg (Goldberg, 1982), among others, has proposed that the Big Five is a universal structure for representing individual differences in personality, applicable to people of any age, any epoch, or any culture. Even if that's an overstatement (and I'm not sure that it is), the language of traits and types gives us an important framework for social categorization, and for identifying just what it is about people that cause them to behave the way they do.

Studies of the language of personality description have focused on the semantic relations among various trait terms. A good example is the work of Seymour Rosenberg and his colleagues, who identified two large dimensions running through the personality trait lexicon: intellectual and social good-bad. These two dimenions are themselves related, creating a "good-bad" superdimension that accounts for the "Halo Effect" the tendency of people to perceive socially desirable traits as correlated with each other.

Another favorite representation of the semantic relations among personality characteristics has been the circumplex, introduced by Timothy Leary (he of psychedelia fame) in 1957. A circumplex is a circular arrangement of items in which the angular distance between terms indicates the semantic distance between them. Capitalizing on the greater computational power of the modern high-speed compuer, Jerry Wiggins (1980) generated a modified ciricumplex structure of personality.

Circumplex models have also been popular for representing the relations among different affective states, yielding two competing models. One, offered by Russell (1980), is based on two bipolar dimensions, positive-negative and strong-weak (Wundt offered the first such representation in the late 19th century). An alternative, offered by Tellegen and Watson (1985), represents positive and negative affect as independent of each other, lying on orthogonal dimensions, rather than diametrically opposed. At conventions, fistfights have been known to break out over whether positive and negative affect are opposite poles of a single dimension, or two indepenent dimensions. Note that the one circumplex is, essentially, a 45-degree rotation of the other.

These linguistic relations capture what is known as implicit personality theory: people's intuitive beliefs, derived from social learning, concerning the nature and scope of individual differences in personality.

Another connection between language and social cognition takes the form of the Sapir-Whorf hypothesis -- that, in some important way, the language we use constrains the thoughts we can think. For example, the very existence of so many trait terms in English may encourage us to characterize people, and explain their behavior, in terms of stable dispositions. Or, it may be the other way around -- that so many trait terms exist because we think of people that way. Which is why the Sapir-Whorf hypothesis is such a provocative idea, and why it has remained controversial to this day.

My favorite example of the language of social cognition comes from a study by Roger Brown and Deborah Fish of the causality implicit in language (Brown & Fish, 1983). If I tell you that Ted charms Paul and ask you to speculate why that might be, you'll probably attribute the liking to the subject of the sentence, Ted-- that he's charming. But if I tell you that Ted loathes Paul, you'll probably attribute the loathing to the object of the sentence, Paul -- that he's loathsome. Based on Charles Fillmore's case grammar, Brown and Fish suggested that causality is implicit in the semantic role: behavioral actions are generally attributive to the agent who instigates the action, not the patient who is the recipient; but mental states are generally attributive to the stimulus who gives rise to the experience, not the experiencer who has it. At first blush, this looks like a Whorffian result -- that linguistic semantics constrains the causal attributions we make. But in fact, Brown and Fish argue the opposite, noting that English could have adjectives like charmable that would be attributive to Paul, and like loatheful that would be attributive to Ted -- it just doesn't, perhaps because we don't think of things that way. So, to Brown and Fish, this looks like a "contra-Whorffian result, of thought containing language.


Social-Cognitive Development

Cognitive and social psychologists are interested in cognitive and social development, and theorists of social cognition are interested in the phylogenetic and ontogenetic origins of social cognition. And so social cognition has dealt with debates over nature and nurture (rationalism vs. empiricism), concrete vs. abstract thought, the Piagetian stages (especially egocentrism), novices and experts, and the like -- everything that our developmental colleagues have gone through for the past century.

Interestingly, the prevailing theory of cognitive development is expressly a theory of social cognition. I'm thinking of the idea of the theory of mind, which argues that, as development proceeds, children come to recognize that their mental states are their own, and that other people may have different beliefs, feelings, and desires. The theory of mind is part of a broader theory theory which holds that development consists of the generation, testing, and revision of theories about how various aspects of the world works -- children elaborate theories of physics and biology as well as theories of mind; and, for that matter, they develop theories of society as well (although this has not been studied all that much). Among these is an intuitive theory of personality.

The theory of mind is often operationalized in the false belief test: Sally and Jane hid a puppet in an oatmeal container; after Jane has left the room, Sally and the experimenter move the puppet to a shoebox; then Sally is asked where Jane will look when she returns and searches for the puppet. Before about age 3, Sally will say that Jane will look in the shoebox, because "That's where it is". But by about age 5, Sally will say that Jane will look in the oatmeal container, because "That's where she thinks it is". The older Sally has a theory of mind: she is now able to make inferences about another person's mental states -- and, in the final analysis, that's what social cognition is all about. One of the more interesting developments in the theory of mind has been the use of nonverbal versions of the false-belief task to push the milestone back -- leading to the discovery that even very young children have a rudimentary theory of mind; but, so far as we can tell, even adult chimpanzees don't.


Computational Modeling of Social Cognition

These days, cognitive psychologists like to write their theories in the form of operating computer programs -- an exercise that John Searle refers to as "Weak Artificial Intelligence", because it does not imply that the machine actually has mental states. And social cognition has been no exception. I'll spend relatively little time on this, except to say that the computational modeling of social cognition looks pretty much like the computational modeling of nonsocial cognition. There are, for example, a number of "symbolic" or "locationist" models of person memory, where individual people, their traits, and their behaviors are all represented as nodes in a vast associative network. And there are also some "connectionist" or "neural network" models, in which knowledge is represented by the pattern of activation in the network as a whole. They both work, and they both have their advantages and disadvantages.

Connectionist networks are often favored because they are "neurally plausible", but they have a serious liability in "catastrophic interference", meaning that they forget what they know as soon as they learn something new. Since this is something that human brains don't do, it's not clear how "neurally plausible" connectionist models are. But something like a connectionist model is probably necessary to represent the bidirectional causal influences inherent in social interaction, so they're well worth continued exploration.


Social-Cognitive Neuroscience

Cognitive psychology is interested in the neural substrates of cognition, leading to the development of cognitive neuroscience as a new interdisciplinary field. Social psychologists have followed suit, leading to the development of what is variously called social-cognitive neuropsychology (my favored term) or social-cognitive neuroscience (Kihlstrom, 2009b). Social-cognitive neuroscience had its deep origins when social psychologists began to take an interest in psychophysiological methods, and also in the lives of brain-damaged patients -- not to mention the textbook case of Phineas Gage. But it really took off when the advent of brain-imaging techniques such as positron-emission tomography (PET) and functional magnetic-resonance imaging (fMRI) which allowed investigators to view the operations of the brain while subjects performed various tasks -- including tasks involving social perception and judgment.

Social-cognitive neuroscience, like cognitive (and, for that matter, affective) neuroscience is predicated on Jerry Fodor's (Fodor, 1983) Doctrine of Modularity, which postulates that the mind is not a general information-processing system, much like a computer, but rather an assembly of interconnected modules, each of which performs a specialized task; in addition, the Doctrine assumes that each of these modules is associated with a "fixed neural architecture" -- that is, that each mental module is associated with a module in the brain. Like cognitive neuroscience, social-cognitive neuroscience proceeds by engaging subjects in a task designed to activate this module, and then identifies which brain area (if any), is specifically activated during that task. A review by Matt Lieberman published only 6 years after the field's formal debut in the pages of American Psychologist, had already identified 21 such areas, each apparently dedicated to some aspect of social cognition.

Of course, the identification of such modules is only as valid as the tasks used to identify it -- and particularly the control task employed. For example, something of a consensus has emerged that a particular area of the medial temporal lobe, known as the fusiform gyrus, is dedicated to the identification of individual faces: Nancy Kanwisher has christened this area the fusiform face area. It's a very appealing idea: the face is the primary social stimulus, and if the brain has evolved a module dedicated to any aspect of social cognition, it's got to be face recognition. On the other hand, Isabel Gauthier and others have gathered evidence that the fusiform area is activated when subjects identify birds and snowflakes as well as when they identify faces. So, it's possible that the fusiform gyrus is a "flexible fusiform area", that mediates the identification of all sorts of objects at a subordinate level of classification, and that faces are just particularly familiar examples of this general point. Or, I suppose, it might be the case that the fusiform gyrus is specialized for faces after all, and that people become experts in identifying birds or snowflakes by co-opting that area for this purpose.

The important point here is not that Kanwisher or Gautier is wrong, and the other one right. It is that neuroscientific data cannot settle questions of psychological theory. A lot of cognitive and social neuroscientists, whether explicitly or implicitly, adhere to what I have called the rhetoric of constraint -- the idea that, somehow, data about brain structure and function will tell us which psychological theories are right and which are wrong. I think this idea is wrong, and not just because, if it were true, psychology really would be, in Ulric Neisser's (Neisser, 1967) immortal phrase, just "something to do until the biochemist comes". Going more deeply, I agree with the philosopher Gary Hatfield (Hatfield, 2000) that psychology is the science that studies mental functions directly; and to the extent that neuroscience wishes to identify the brain functions that correspond to mental functions, neuroscience depends on psychology, not the other way around.

But none of this means that social neuroscience isn't a promising enterprise. Right from the beginning, psychologists have been interested in "mind in body", or the connections between what goes on in the mind and what goes on in the brain and the rest of the nervous system. For a long time, social psychologists thought that social interaction, including social cognition, was too complex to be approached with the blunt "slash and burn" methods that characterized the earlier physiological psychology. And they were right. But now, with PET, fMRI, and all the other tools of modern neuroscience, were now in a position to reveal the biological substrates of mental life, as we never were before.


The Current Scene

The turn toward neuroscience is one of the most interesting trends in the contemporary study of social cognition, but there are others.

Judgment Heuristics

One trend, which I alluded to earlier in my discussion of causal attribution, has been the documentation of various errors and biases in social cognition. The origins of this trend lie in the work of Herbert Simon, and especially Daniel Kahneman and Amos Tversky, who documented the use of various judgment heuristics. Previous work on judgment and decision-making, as exemplified by neoclassical economics, had been based on assumptions of normative rationality -- the idea that people employed logical procedures known as algorithms to optimize the outcomes of various choices -- algorithms that, when employed correctly, are guaranteed to yield the right answer to the problem at hand. Kahneman and Tversky showed that people relied instead on heuristics, or shortcuts, that bypassed the algorithms of normative rationality and allowed people to make decisions under conditions of uncertainty -- which, after all, is pretty much the definition of real life. Simon, Kahneman and Tversky, and their confreres substituted a description of how people actually think for the logicians' prescription of how people did think. And in so doing they worked a revolution in economics, and in the psychology of judgment and decision-making, that continues to rack up Nobel Prizes.



Just as early approaches to social cognition assumed, at least implicitly, that social cognition followed the principles of normative rationality, so it was it was assumed, again at least implicitly, that social cognition involved conscious, deliberate, thought. Beginning in the 1970s, however, cognitive psychology began to articulate a distinction between automatic and controlled processing. Controlled processing is, indeed, conscious and deliberate; it consumes cognitive resources, and involves serial processing. Automatic processes by contrast, are inevitably evoked by the appearance of an effective environmental stimulus; once evoked, they are incorrigibly executed, in a "ballistic" fashion; they consume few or no cognitive resources; and they don't interfere with each other, or with controlled processes. Automatic processes are reflex-like, in some respects, but they are not innate: in principle, any process, no matter how complex, can be automatized if it is practiced enough.

Whether they are innate or acquired, automatic processes are unconscious in the strict sense of the term: they operate outside conscious awareness, and independent of conscious control. In fact, automaticity has been dubbed "the new unconscious" (Hassin, Uleman, & Bargh, 2005) -- the "old" unconscious being the "monsters from the id" envisioned by Sigmund Freud and other proponents of psychoanalysis. Indeed, widespread acceptance of the concept of automaticity has helped legitimize the concept of unconscious mental life, but it is not the entirety of the psychological unconscious. In theory, automatic processes operate on conscious mental contents -- percepts, memories, thoughts, and the like -- to generate other conscious mental contents. We are aware of what we think, but we're not aware of the automatic processes that generate what we think. But, beginning with the study of implicit memory in amnesic patients, it's become clear that mental contents themselves -- percepts, memories, thoughts, feelings, motives -- can also be unconscious, yet exert an influence on the person's ongoing experience, thought, and action. Implicit cognition, emotion, and motivation join automatic processing to comprise unconscious mental life (Kihlstrom, 2008b, 2010b).

Within cognitive psychology, there is a general consensus that every task has both automatic and controlled components, and considerable effort has been devoted to measuring their differential contributions to performance. In social psychology, however, a view has developed that social cognition and behavior is overwhelmingly governed by automatic processes -- what I have called "the automaticity juggernaut" (Kihlstrom, 2008a). Somewhat ironically, John Bargh (J. A. Bargh, 1984; J.A. Bargh & Ferguson, 2000) has expressly linked automaticity to Skinnerian behaviorism. His position is not exactly a revival of behaviorism, because he adopts the central dogma of the cognitive revolution, that cognitive and other mental states and processes intervene between environmental stimulus and organismal response. But when the intervening processes are automatically evoked by environmental stimuli, the embrace of automaticity looks more and more like (with apologies to Susan Sontag) behaviorism with a cognitive face -- leading to wonder whether, in the end, we had a cognitive revolution only to learn that Skinner had it right the first time.

              (50148 bytes)Some proponents of automaticity have linked themselves to Copernicus and Darwin in revealing some unpleasant truths of human existence: Copernicus demonstrated that the Earth is not the center of the universe, while Darwin showed that Man is just another animal. Freud, who also positioned himself in this line of "naturalization", sorrowfully informed his readers that man was irrational after all, driven by seething, if unconscious, sexual and aggressive urges. Bargh and Dan Wegner (Wegner & Bargh, 1998), another driver of the automaticity juggernaut, have suggested, with apologies to Freud, that free will is severely compromised, if not simply illusory, and that the psychological truth is that we are automatons after all.

There is no doubt that automaticity plays some role in social cognition and behavior; the question is whether, as Bargh put it, behavior is "99.44%" automatic. In fact, what we have in the literature are mostly demonstration experiments, that show that automaticity plays some role in social interaction. But many of these experiments involve a very loose operationalization of automaticity, and the few comparative experiments published to date reveal more of a balance between automatic and controlled components of processing -- except in special circumstances, like very narrow response windows, where controlled processing simply cannot come into play. Nothing in the literature supports the idea that social behavior is wholly, or even largely driven by automatic processes -- or that, as some have concluded, conscious will is an illusion and consciousness only gets in the way of adaptive behavior.


An Affective Counter-Revolution

The cognitive revolution in psychology began with the recognition that internal mental states mediated between stimulus and response. And we call the cognitive revolution a cognitive revolution because the analysis of these intervening states focused on cognition -- expectations, attention, short-term memory, the syntax of language, and the like. Some cognitive scientists use the term "cognitive" to refer to all mental states, including emotional and motivational states. And, as the cognitive revolution began to spread, some social psychologists adopted the view that emotional and motivational states were themselves cognitive constructions -- that is, beliefs about what one was feeling or desiring, depending on one's perception of the situation. That point of view took social psychology a long way, but fairly soon some psychologists began to object to the "cold, rational" view of social interaction implicit in early theories of social cognition, with their focus on algebraic rules for impression formation, the analysis of variance for causal attribution, and the like. And they also began to promote the idea that feelings and desires were to an important extent independent of beliefs.

The issue began to be joined with arguments for hot cognition, or the view that emotion and motivation influenced cognition -- a viewpoint that was foreshadowed by Bruner's (J.S. Bruner & Klein, 1960) the "New Look" in perception. And it came to a head in a debate between Robert Zajonc, who argued that "preferences need no inferences", and Richard Lazarus, who argued instead for "the primacy of cognition". Research by Paul Ekman on the facial expression of emotion supported the hypothesis that certain emotional reactions are innate, universal, and essentially reflexive in nature. These "basic emotions" do not require cognitive mediation, nor is any required for us to "read" these emotional states in the faces of other people. The prevailing view now is that at least some aspects of emotion, and probably motivation as well, are indeed independent of cognition -- not least because they are evoked automatically by particular environmental stimuli.

In fact, the emerging view is that affect is not just independent of cognition, and deserves special status, but rather that affect is more important than cognition. This view can be seen clearly in the current emphasis on the role of intuition and emotion in decision-making. Antonio Damasio has written of "Descartes' error" in elevating reason above emotion -- and in separating mind from body (Damasio, 1994). And Joshua Greene cites the "Trolley Problem" as evidence that feelings, rather than reasons, guide our moral judgments. But intuitions are at least as "cognitive" as they are emotional: they play a role in recognition judgments and problem-solving, when emotion has nothing to do with anything (Dorfman, Shames, & Kihlstrom, 1996; Kihlstrom, Shames, & Dorfman, 1996), for example . And the fact that people fall back on emotion when an experiment has been deliberately constructed that reason necessarily fails them, does not mean that cognition is irrelevant to moral judgment in the world outside the laboratory. Paraphrasing Paula Niedenthal (Niedenthal, 1992), affect is information for cognition. But emotion doesn't rule cognition; rather, along with motivation, they rule side by side.

It is something of a mystery why, once the cognitive revolution took hold, it should have taken psychology so long to become serious about emotion. Part of the reason surely lies in the roots of psychology in British empiricism -- that is, with problems of knowledge acquisition. The implication of Cartesian dualism is that emotions are part of our animal nature, closely tied to the body, and that what is distinctive about humans is that we can think. The same sort of implication can be found in MacLean's concept of the triune brain, including the "reptilian" brainstem, concerned primarily with biological motives; the "mammalian paleocortex" of the limbic system, generating emotions; and the neocortex, sitting on top of everything, and making thought possible. The behaviorists, of course, threw emotion and motivation out with cognition. As Robert S. Woodworth (Woodworth, 1929) put it:

First psychology lost its soul, then it lost its mind...

It has behavior, still, of a kind.

Psychology eventually got its mind back, but it came back first with a focus on cognition. The revival of interest in emotion is a necessary corrective to what has been the hegemony of cognition. But independence doesn't mean superiority. Just as we didn't have a cognitive revolution only to find out that Skinner got it right the first time, we didn't evolve a neocortex to discover that the paleocortex is enough. And we don't need another revolution. What we need is peaceful coexistence, with each element -- cognition, emotion, and motivation -- occupying its rightful place in the mental economy of behavior. Still, the affective counterrevolution in psychology has set the stage for the development of new graduate groups in affective psychology, paralleling cognitive psychology, and an "affective neuroscience" independent of cognitive neuroscience.



Sometimes, it seems as if the "errors and biases" movement, the automaticity juggernaut, and the affective counterrevolution have come together to generate a new approach to psychology, which I have come to call the "People Are Stupid" school of social psychology (Kihlstrom, 2004), which is organized around a small number of related propositions:

People are fundamentally irrational: In the ordinary course of everyday living, we do not think very hard about anything, preferring heuristic shortcuts that lead us astray; and we let our feelings and motives get in the way of our thought processes [e.g., \Nisbett, 1980 #17881;Ross, 1977 #21628].
We are on automatic pilot: We do not pay much attention to what is going on around us, and to what we are doing; as a result, our thoughts and actions are inordinately swayed by first impressions and immediate responses; free will is an illusion [e.g., \Bargh, 1995 #6251;Wegner, 2002 #19543;Gilbert, 1991 #21632].
We don't know what we're doing: When all is said and done, our behavior is mostly unconscious; the reasons we give are little more than post-hoc rationalizations, and our forecasts are invalid; to make things worse, consciousness actually gets in the way of adaptive behavior [e.g., \Nisbett, 1977 #17107;Wilson, 2002 #11237].
We don't know what we want: We're extremely poor at predicting how we will feel about various eventualities (Wilson & Gilbert, 2005), and we're so poor at making choices that we might just as well let others choose for us (Iyengar, 2010) -- largely because, again, we don't have introspective access to our beliefs, feelings, and desires. One is reminded of the joke about the two behaviorists who had sex: one said to the other: "It was good for you, but was it good for me?".
We don't even know how stupid we are: Because of the limitations on our cognitive abilities, we fail to appreciate when our judgments and behaviors are less than optimal (Dunning, Heath, & Suls, 2004).  

The "People Are Stupid" School -- to the extent that it's not just a figment of my imagination -- had its origins in a very reasonable program of research that employed evidence of errors to produce a more realistic description of how people actually make judgments and decisions. But there are even deeper roots of social psychology's preference for the thoughtless, the unconscious, and the automatic. Somehow, fairly early on, social psychology got defined as the study of the effect of the social situation on the individual's experience, thought, and action. Think, for example, of the classic work on the "Four As" of social psychology: attitudes, attraction, aggression, and altruism; think, too, on the history of research on conformity and compliance, from Asch and before to Milgram and beyond. In each case, the experimenter manipulates some aspect of the environment, and observes its effect on subjects' behavior. Sometimes there were inferences about intervening mental states, but not very often -- otherwise, the cognitive revolution in social psychology wouldn't have been a revolution. Almost inevitably, the emphasis on how people are pushed around by situational factors led to a kind of "Candid Camera" or "People Are Funny" rhetorical stance in which social psychologists' lectures and textbooks focused inordinately on just how ridiculous -- how stupid -- people can be, depending on the situation -- a situation that, in many cases, has been expressly contrived to make people look ridiculous and stupid.


Is Social Cognition "Special"?

In the study of social cognition, the default assumption is that the cognitive processes used to perceive, remember, and think about social objects such as people are the same as those used to perceive, remember, and think about nonsocial objects. After all, you have to start somewhere, and in fact, in a wide variety of areas the default assumption appears to be reasonably valid. At the same time, there are important differences between social and nonsocial cognition. Some of these are quantitative in nature -- differences in degree; others are qualitative in nature -- actual differences in kind.

With respect to quantitative differences, the first and most important one is the poverty of the stimulus. Helmholtz noted that the stimulus environment was too vague and ambiguous, and contained too many conflicting cues, to support perception alone -- which is why perception required the perceiver to make unconscious inferences. This is all the more so, it seems to me, in the case of the social environment. Similarly, context seems to be even more important in social perception than in nonsocial perception. And so do emotion and motivation. Perception is inherently constructive, and memory is inherently reconstructive, but, again, constructive processes seem even more important in the social case -- where, to come full circle, the stimulus environment is especially vague, ambiguous, and full of conflicting cues.

Some Examples of Ambiguous Social Stimuli

9-11a.jpg (63642 bytes) 9-11b.jpg (60215 bytes) 9-11c.jpg (51540 bytes)
Muskie.jpg (34517 bytes) HClinton.jpg (48720 bytes)

Here's another example.  This photograph by Spencer Platt is titled "Beirut Residents continue to Flock to Southern Neighborhoods".  As explained by Ken Johnson, art critic  for the New York Times ("Poignant Images, with Posterity the Ultimate Winner", 11/15/2013):

In a photograph shot by Spence Platt in Lebanon in 2006, the spectacle of five attractive, fashionably dressed young people in a glossy red convertible occupies the foreground.  By surrealistic contrast, the immediate background is filled with the smoking wreckage of bombed buildings, where a few pedestrians pass by....  [T]he impression you get is of obnoxious rich kids out for a sensation-seeking drive.

But the truth of Mr. Platt's picture, which won the 2006 World Press Photo of the Year award, was not what it seemed.  In response to widespread criticism, the car's driver and passengers protested to news reporters that they were not disaster tourists but residents of the neighborhood returning to recover their belongings.

VancouverRiotKiss_CloseUp.gif (38542 bytes)Perhaps the most dramatic illustration of the importance of context to social cognition is the famous photograph of the "Vancouver Riot Kiss", taken by Rich Lam during a riot that broke out in Vancouver, British Columbia, in June 2011, after the Canucks lost in the Stanley Cup hockey finals.  Stripped of its context, the photo appears to show a couple engaged in foreplay.  VancouverRiotKiss_Original.jpg (18500 bytes)But the original photograph shows that "The Kiss" actually took place against a background of a street protest, with a police officer wielding a truncheon in the foreground.  But that only makes the photo more ambiguous.  What's going on here? Is this a piece of performance art?  An ironic act of social protest?  A Photoshopped hoax?  Is he taking advantage of her?  It turned out that the couple involved, Scott Jones and Alexandra Thomas, were caught up in the mayhem.  She was hurt in the melee, and he was trying to comfort her. 

Bruner (J. S. Bruner, 1957) wrote (and here I'm paraphrasing) that the purpose of perception is action, and that's certainly true for the nonsocial case. In order to grasp a mug of coffee, you have to identify it as a mug, note its location and distance, and also apprehend its precise shape, so that you know mold your hand around it. Gibson was onto something when he discussed perception in terms of affordances -- what you can do with the object of perception. Action is even more important in the social case. As discussed at the outset, the cognitive point of view in social psychology is that a person's social interactions are (largely) determined by cognitions -- by his perception of the situation, memories of previous encounters, and reasoning about what to do next. In most of cognitive psychology, the concept of action is rather impoverished -- pointing at a target, grasping a mug, typing at a keyboard. But in social psychology the concept of action is very broad, encompassing the person's actual behavior in a real world populated by other people. Social psychology shouldn't be defined as the study of attitudes, and it shouldn't be defined as the study of social influence. But if biological psychology is the study of mind in body, social psychology might well be defined as the study of mind in action.

What about qualitative differences? Are there any features of cognition that are unique to the social case? One possibility comes from social neuroscience, and the Doctrine of Modularity. That is, it is entirely possible that there are some aspects of social cognition that are performed by dedicated mental modules that differ from those active in the nonsocial case. We have seen hints of this -- as well as the difficulties inherent in making the case. But it's a possibility, and cognitive scientists like Ray Jackendoff (Jackendoff, 1992) have convincingly argued that there are some aspects of social cognition that have no counterparts in the nonsocial case, and which are so universal that there might well, indeed, be specialized mental modules for them, associated with a fixed neural architecture.

A second qualitative difference has to do with the self, which from the cognitive point of view may be construed as a mental representation of oneself (Kihlstrom, Beer, & Klein, 2002; Kihlstrom & Cantor, 1984). Usually, there is a clear distinction between the knower and the object of cognition, but this is dissolved when it comes to the self in a way that simply does not occur in the nonsocial case. Viewed more broadly, the world of nonsocial cognition is mostly a world of observer-independent facts: features of the world that would exist even if there were no one to experience them. But the world of social cognition is a mostly world of meaning, a world of observer-dependent facts. It's true that I can think the moon is beautiful tonight, and this is an observer-dependent fact about the social world, so with respect to social cognition the issue of observer-dependence and --independence may be a quantitative, not qualitative one. But when it comes to the self, the distinction is qualitative: people may have a view of me that differs from my view of myself, but my self-concept is mine alone, nobody has it but me, and the existence of a self depends utterly on my existence as a sentient being.

Indefinite.jpg (59282 bytes)Which brings me to the last qualitative difference between social and nonsocial cognition, and it's a big one. In social cognition, the object of cognition is itself a sentient being, possessed of intelligence and consciousness, aware of being cognized (as it were), and trying to shape the cognitions of the cognizer (I'm sorry, but I had to keep it going). In social cognition, impression management by the object is counterpoised to impression formation by the subject. And, of course, the subject knows this; and the object knows he knows it, raising the possibility of an infinite series. It's not completely clear how far the series goes (Cargile, 1970), but one thing is clear: there's nothing at all like it in the nonsocial world.

In nonsocial cognition, the objects of cognition have an existence that is independent of the observer's mind. But to some extent, at least, the objects of social cognition depend on the mind of the observer. In social cognition, belief can create reality, in the form of variants on the self-fulfilling prophecy (E. E. Jones, 1986; Miller & Turnbull, 1986; Snyder, 1984); and, because social reality consists of other persons, these behavioral and perceptual confirmation effects, imposed on the object by the perceiver, are opposed by self-verification effects imposed on the perceiver by the object (Swann & Ely, 1984). Belief can create reality in the nonsocial world, but there is nothing like this "battle of wills" outside of social interaction.


A Step to a Cognitive Social Science

Social psychology links psychology to the other social sciences, and all the social sciences share a focus on observer-created reality: that is, a reality that is a product of individual and collective mental activity, as people seek, and give, meaning in what they find in the world around them. This is yet another aspect of social cognition: the idea that cognition is not just about individual minds trying to understand and cope with external reality. Rather, cognition is also a matter of minds working together, trying to understand a reality that is, at least partly, the creation of their own minds. Social cognition, then, may be an essential first step toward a cognitive social science. Already, economics has felt the influence of the cognitive point of view -- that economic systems cannot be understood simply in terms of the operation of some "invisible hand", but that what happens in an economy is determined by individuals trying to understand, and control, a system that is of their own making. So too, perhaps, social cognition, and the cognitive viewpoint in social psychology generally, is coming to be felt in fields such as sociology, anthropology, history, and political science.



Allport, G. W. (1954). The historical background of social psychology. In G. Lindzey & E. Aronson (Eds.), Handbook of social psychology (Vol. 1, pp. 1-46). New York: Random House.

Baars, B. J. (1986). The cognitive revolution in psychology. New York: Guilford Press.

Bargh, J. A. (1984). Automatic and conscious processing of social information. In R. S. Wyer & T. K. Srull (Eds.), Handbook of social cognition (pp. 1-43). Hillsdale, N.J.: Erlbaum.

Bargh, J. A., & Ferguson, M. J. (2000). Beyond behaviorism: On the automaticity of higher mental processes. PSYCHOLOGICAL BULLETIN, 126(6 Sp Iss), 925-945.

Bartlett, F. C. (1932). Remembering: A study in experimental and social psychology. Cambridge: Cambridge University Press.

Blumer, H. (1969). Symbolic interactionism: Perspective and method. Berkeley, Ca.: University of California Press.

Brown, R. (1965). Social psychology. New York: Free Press.

Brown, R. (1980). Natural categories and basic objects in the domain of persons. Katz-Newcomb Lecture, U. Mich.

Brown, R., & Fish, D. (1983). The psychological causality implicit in language. Cognition, 14, 237-273.

Bruner, J., & Tagiuri, R. (1954). Person perception. In G. Lindzey (Ed.), Handbook of social psychology (Vol. 2, pp. 634-654). Reading, Ma:: Addison-Wesley.

Bruner, J. S. (1957). On perceptual readiness. Psychological Review, 64, 123-152.

Bruner, J. S., & Klein, G. S. (1960). The function of perceiving: New Look retrospect. In W. Wapner & B. Kaplan (Eds.), Perspectives in psychological theory: Essays in honor of Heinz Werner (pp. 61-77). New York: International Universities Press.

Cantor, N., & Kihlstrom, J. F. (1987). Personality and social intelligence. Englewood Cliffs, N.J.: Prentice-Hall.

Cargile, J. (1970). A note on 'iterated knowing'. Analysis, 30, 151-155.

Damasio, A. R. (1994). Descartes' Error: Emotion, Reason, and the Human Brain.

Dorfman, J., Shames, V. A., & Kihlstrom, J. F. (1996). Intuition, incubation, and insight: Implicit cognition in problem solving. In G. Underwood (Ed.), Implicit cognition. (pp. 257-296). Oxford, England UK: Oxford University Press.

Dunning, D., Heath, C., & Suls, J. M. (2004). Flawed self-assessment: Implications for health, education, and the workplace. Psychological Science in the Public Interest, 5, 69-106.

Fodor, J. A. (1983). The modularity of the mind. Cambridge, Ma.: MIT Press.

Funder, D. C. (2012). Acccurate personality judgments. Current Directions in Psychological Science, 21(3), 177-182.

Goldberg, L. R. (1982). From Ace to Zombie: Some explorations in the language of personality. In C. D. Spielberger & J. N. Butcher (Eds.), Advances in Personality Assessment (pp. 203-234). Hillsdale, N.J.: Erlbaum.

Goleman, D. (2006). Social intelligence: The new science of human relationships. New York: Bantam Books.

Hassin, R. R., Uleman, J. S., & Bargh, J. A. (Eds.). (2005). The new unconscious. New York: Oxford University Press.

Hatfield, G. (2000). The brain's "new" science: Psychology, neurophysiology, and constraint. Philosophy of Science, 67(Proceedings), S388-S403.

Hirst, W., & Miller, G. A. (1988). The Making of cognitive science : essays in honor of George A. Miller. Cambridge ; New York: Cambridge University Press.

Jackendoff, R. (1992). Is there a faculty of social cognition? In R. Jackendoff (Ed.), Languages of the Mind: Essays on Mental Representation (pp. 19-31). Cambridge, Ma.: MIT Press.

Jones, E. E. (1985). Major developments in social psychology since 1930. In G. Lindzey & E. Aronson (Eds.), Handbook of social psychology (2nd ed., Vol. 1, pp. 47-107). Reading, Ma.: Addison- Wesley.

Jones, E. E. (1986). Interpreting interpersonal behavior: The effects of expectancies. Science, 234, 41-46.

Kihlstrom, J. F. (2004). Is there a "People are Stupid" school in social psychology? [Commentary on "Towards a balanced social psychology: Causes, consequences, and cures for the problem-seeking approach to social behavior and cognition" by J.I. Krueger and D.C. Funder]. Behavioral & Brain Sciences, 27, 348-349.

Kihlstrom, J. F. (2008a). The automaticity juggernaut. In J. Baer, J. C. Kaufman & R. F. Baumeister (Eds.), Psychology and free will (pp. 155-180). New York: Oxford University Press.

Kihlstrom, J. F. (2008b). The psychological unconscious. In O. John, R. Robins & L. A. Pervin (Eds.), Handbook of Personality: Theory and Research. (3rd ed., pp. in press). New York: Guilford.

Kihlstrom, J. F. (2009a). "So that we might have roses in December": The functions of autobiographical memory. Applied Cognitive Psychology, 23, 1179-1192.

Kihlstrom, J. F. (2009b). Social neuroscience: The footprints of Phineas Gage. Social Cognition, in press.

Kihlstrom, J. F. (2010a). The person-situation interaction. In D. Carlston (Ed.), Oxford handbook of social cognition (pp. in press). New York: Oxford University Press.

Kihlstrom, J. F. (2010b). Unconscious processes. In D. Reisberg (Ed.), Oxford Handbook of Cognitive Psychology (pp. in press). Oxford: Oxford University Press.

Kihlstrom, J. F., Beer, J. S., & Klein, S. B. (2002). Self and identity as memory. In M. R. Leary & J. Tangney (Eds.), Handbook of self and identity (pp. 68-90). New York: Guilford Press.

Kihlstrom, J. F., & Cantor, N. (1984). Mental representations of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 17, pp. 1-47). New York: Academic.

Kihlstrom, J. F., & Cantor, N. (2000). Social intelligence. In R. J. Sternberg (Ed.), Handbook of intelligence. (pp. 359-379). New York, NY, US: Cambridge University Press.

Kihlstrom, J. F., & Cantor, N. (2010). Social intelligence. In R. J. Sternberg (Ed.), Handbook of intelligence. (3rd ed., pp. xxx-xxx). New York, NY, US: Cambridge University Press.

Kihlstrom, J. F., & Park, L. (2002). Cognitive psychology: Overview. In V. S. Ramachandran (Ed.), Encyclopedia of the Human Brain (Vol. 1, pp. 839-853). San Diego: Academic.

Kihlstrom, J. F., Shames, V. A., & Dorfman, J. (1996). Intimations of memory and thought. In L. M. Reder (Ed.), Implicit memory and metacognition (pp. 1-23). Mahwah, N.J.: Erlbaum.

Leary, T. (1957). Interpersonal diagnosis of personality: A functional theory and methodology for personality evaluation. Oxford: Ronald Press.

Malle, B. F. (2004). How the mind explains behavior: Folk explanations, meaning, and social interaction. Cambridge, Ma.: MIT Press.

Malle, B. F. (2008). Fritz Heider's legacy: Celebrated insights, many of them misunderstood. Social Psychology, 39(3), 163-173.

Miller, D. T., & Turnbull, W. (1986). Expectancies and interpersonal processes. Annual Review of Psychology, 37, 233-256.

Neisser, U. (1967). Cognitive psychology. New York: Appleton-Century-Crofts.

Niedenthal, P. M. (1992). Affect and social perception: On the psychological validity of rose-colored glasses. In R. Bornstein & T. Pittman (Eds.), Perception without Awareness (pp. 211-235). New York: Guilford Press.

Rosenberg, S., Nelson, C., & Vivekananthan, P. S. (1968). A multidimensional approach to the structure of personality impressions. Journal of Personality & Social Psychology, 9(4), 283-294.

Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39, 1161-1178.

Snyder, M. (1984). When belief creates reality. In Advances in experimental social psychology (pp. 247-305): Academic Press.

Swann, W. B., & Ely, R. J. (1984). A battle of wills: Self-verification versus behavioral confirmation. Journal of Personality & Social Psychology, 46, 1287-1302.

Tagiuri, R., & Petrullo, L. (Eds.). (1958). Person perception and interpersonal behavior. Stanford:: Stanford University Press.

Watson, D., & Tellegen, A. (1985). Toward a consensual structure of mood. Psych. Bull., 98, 219-235.

Thomas, W. I., & Thomas, D. S. (1928). The child in America: Behavior problems and programs. New York: Knopf.

Wegner, D. M., & Bargh, J. A. (1998). Control and automaticity in social life. In D. Gilbert, S. T. Fiske & G. Lindzey (Eds.), Handbook of social psychology (Vol. 1, pp. 446-496). Boston: McGraw-Hill.

Wiggins, J. S. (1980). Circumplex models of interpersonal behavior. In L. Wheeler (Ed.), Review of personality & social psychology (Vol. 1, pp. 265-293). Beverly Hills, Ca.: Sage.

Wilson, T. D., & Gilbert, D. T. (2005). Affective Forecasting: Knowing What to Want. Current Directions in Psychological Science, 14(3), 131-134(134).

Woodworth, R. S. (1929). Psychology: A study of mental life (2nd ed.). New York: Holt.

Zajonc, R. B. (1980). Cognition and social cognition: A historical perspective. In L. Festinger (Ed.), Four decades of social psychology (pp. 180-204). New York: Oxford University Press.

Zimbardo, P. G. (1999). Experimental social psychology: Behaviorism with minds and matters. In A. Rodrigues & R. V. Levine (Eds.), Reflections on 100 years of experimental social psychology (pp. 135-157). New York: Basic Books.


This page last revised 11/16/2013.