|Home | About | Journals | Submit | Contact Us | Français|
Although typically defined as the study of how people and groups interact, the field of social psychology comprises a number of disparate domains that make only indirect contributions to understanding interpersonal interaction, such as emotion, attitudes, and the self. Although these various phenomena may appear to have little in common, recent evidence suggests that the topics at the core of social psychology form a natural group of domains with a common functional neuroanatomy, centered on the medial prefrontal cortex. That self-referential, attitudinal, affective, and other social phenomena converge on this region may reflect their shared reliance on inexact and internally-generated estimates that differ from the more precise representations underlying other psychological phenomena.
A common definition of social psychology suggests that the field represents “an attempt to understand and explain how the thought, feeling, and behavior of individuals are influenced by the actual, imagined, or implied presence of others”1. However, as practiced today, social psychology as often focuses on the cognitive workings of an individual in isolation as on those specific to interpersonal interaction. Beginning with the social cognition movement in the 1970s, social psychology has emerged as the primary headwater for the study of three intrapersonal phenomena that rely little on the “presence of others”: (i) the structure of knowledge about the self; (ii) attitudes and their influence on one’s choices; and (iii) the subjective experience of emotion. Indeed, the 4th edition of The Handbook of Social Psychology – widely considered the definitive encyclopedia of the field – devotes its first two topical sections to such intrapersonal cognition, postponing its review of phenomena that occur in social contexts until the second of its two volumes.
How have these intrapersonal topics emerged as the central province of social psychology, ostensibly the science of understanding humans in interpersonal contexts? Why instead have these topics not formed a core part of cognitive psychology, which explicitly attempts to model the mental operations that support other such within-person abilities such as perception, attention, and memory? Moreover, why have several phenomena with clear implications for interpersonal behavior, such as face identification and language, become central pursuits within cognitive science while remaining comparatively peripheral to social psychology? Although a coherent sense of self, stable attitudes, and a rich repertoire of emotional experience doubtlessly play vital roles in interpersonal interaction, it is unclear how they bear more directly on social behavior than some of the abilities that have been relatively neglected by social psychology.
Over the past decade and a half, studies using neuroimaging and neuropsychological patients have provided a surprising but consistent answer to the question of what, if anything, binds these disparate topics within social psychology: a common neural basis. This work has demonstrated that four seemingly-distinct cognitive phenomena – thinking about oneself, accessing one’s attitudes, the experience of emotion, and inferring the contents of another person’s mind – all converge on a single brain region, the medial prefrontal cortex (MPFC). Such observations suggest that contemporary social psychology, far from being a patchwork of unrelated research questions, is the science of a set of closely related phenomena with a common functional neuroanatomy. Indeed, the neural confluence of self, attitudes, emotional experience, and mental state inference implies that these phenomena may pose a common cognitive challenge to the human mind, met by a common processing solution2. Rather than being the result of historical accident, social psychology appears to be a “natural kind” – a genuine set defined by deep and nonarbitrary characteristics.
Here, I review findings that suggest the ubiquity of MPFC involvement in four topics of central interest to social psychologists: the self-concept, attitudes and evaluation, emotional experience, and understanding the minds of others. The goal is to provide an impressionistic – rather than exhaustive – overview of the surprising convergence of such ‘social’ abilities in the MPFC, and, accordingly, discussion of other brain regions known to contribute to these phenomena is deferred (see Box 3).
Although the MPFC plays a critical role in subserving several phenomena of interest to social psychologists, several other brain regions also contribute importantly to the self-concept (e.g., medial parietal cortex), evaluation (the orbitofrontal cortex and ventral striatum), mentalizing (the medial parietal cortex and bilateral temporo-parietal junction), and emotional experience (e.g., the amygdala and anterior insula). In many cases, little is known about the independent contributions made by these other brain regions to such phenomena; for example, although the medial parietal cortex is frequently observed in conjunction with MPFC during self-referential processing and mentalizing, little is known about how the processes subserved by this region specifically contribute to such phenomena. At the same time, a critical but unresolved issue in cognitive neuroscience is how the particular computations performed by a brain region may vary as a function of the other regions active during a particular cognitive task. For example, in what way might the particular computations performed by the MPFC differ when interacting more prominently with medial parietal cortex in contrast to the anterior insula? An important direction for future research will be to understand how networks of interacting brain regions together subserve our cognitive abilities, rather than focusing on single brain regions in isolation.
More than a dozen neuroimaging studies have examined the neural basis of the self-concept, as traditionally operationalized by social psychologists (Box 1), and these studies have ubiquitously linked self-referential processing to activity in MPFC (Fig 1a). In the preponderance of such studies, participants have been asked to introspect about their own personality characteristics, either by reporting how well they are described by a trait adjective (curious, intelligent, impatient) or by responding to questions about their dispositions (e.g., I have a quick temper). Reflecting on one’s own dispositions in this manner consistently prompts greater MPFC activity than a variety of control conditions, including judging the personality traits of another person3–10; judging the social desirability of a personality trait (is being “curious” generally considered positive or negative?)9, 11–14; answering questions based on semantic knowledge (is 10 seconds longer than a minute?)15; or judging lexical/orthographic features of words16, 17. Moreover, MPFC activity correlates with successful memory for information processed in a self-referential manner18, suggesting that this region supports the well-documented mnemonic benefits of linking information to the self-concept19. Consistent with these neuroimaging observations, patients with frontotemporal dementia – a progressive disorder associated with disproportionately high atrophy in MPFC – demonstrate profound changes in the self-concept, including an impaired ability to judge their own personality traits20.
Research on the self encompasses a variety of phenomena, from self-referential thought and self-initiated behavior to self-regulation and self-esteem. Although there is no single definition of the self, social psychology has paid special attention to one particular aspect of selfhood: one’s concept of self82. The self-concept refers to person’s understanding of what she “is like” as a person, that is, what personality characteristics she manifests, what idiosyncratic abilities and proclivities define her as an individual, and to what extent she regards herself positively (i.e., has high or low self-esteem). Social psychological research on the self-concept has included (i) determining what information people use as a primary basis for judging what they are like; (ii) documenting the consequences of processing events in a self-referential manner; (iii) demonstrating the extent to which people distort information to maintain a consistent self-concept; and (iv) detailing individuals’ attempts to maintain high regard for the self.
A rough sense of the importance of this topic to social psychology can be estimated from the prevalence of the word “self” in titles of articles published in the field’s journals. Between 1965 and 2008, “self” appeared in the title of roughly every eighth article published in the Journal of Personality and Social Psychology (1,056 of 8,539 total) and the Journal of Experimental Social Psychology (255 of 2,235 total). In contrast, over the same span of time, all subtitles of the Journal of Experimental Psychology have included a mere 82 articles with “self” in the title (of more than 16,000 articles published; ~0.5%), and Cognition has published only 15 such articles (~0.7%).
One important line of research on the self has focused on the demonstration of enhanced processing for information encoded self-referentially. This “self-reference effect”19 has typically been observed as better memory for stimuli that participants initially judge in relation to the self (e.g., “how well does the word curious describe you?”) than those they initially judge in relation to another person (“how well does the word honest describe Bill Clinton?”) or about which they make semantic judgments (“was Bill Clinton the 40th president of the United States?”). Tellingly, although by definition this work focuses on an inherently intrapersonal phenomenon (the self) and examines a phenomenon of central interest to cognitive psychologists (memory), it has overwhelmingly appeared in social psychology (and, to some extent, clinical psychology) journals, rather than in mainstream cognitive psychology outlets.
A central concept in social psychology has been that of attitude, defined as “a psychological tendency that is expressed by evaluating a particular entity with some degree of favor or disfavor”21 (Box 2). As for the self-concept, a series of recent neuroimaging studies has suggested that the MPFC plays a critical role in the ability to access and explicitly report one’s attitudes (Fig 1b). A fairly circumscribed region of ventral MPFC is preferentially engaged when individuals respond to questions about their own preferences (e.g., “I enjoy doing laundry over going grocery shopping”) than about those of another person6, 10, 22, 23. Likewise, explicitly evaluating a stimulus as positive or negative produces greater response in this region than judging semantic24, 25, perceptual26, 27, or other nonevaluative28 aspects of a stimulus. Consistent with these neuroimaging observations, patients with damage to the ventral MPFC show impairments in reporting their preferences in a consistent manner. For example, such an individual might evaluate one person as more positive than a second and that second person as more positive than a third, but then also judge, incompatibly, the third to be more positive than the first29.
How do humans form opinions about novel objects and people? How do we store and access such evaluations? And how do we update our attitudes in response to new or contradictory information? The origins of social psychology are so closely linked to these questions that some early commentators equated the entire field with the study of attitudes and evaluation83. By 1935, the concept of an attitude had been proclaimed the principal foundation on which social psychology rests, and throughout the 1960s and 70s, the study of attitudes and attitude change dominated research in social psychology. Even today, the lead section of the Journal of Personality and Social Psychology is devoted to “attitudes and social cognition.” Consistent with their position at the center of the field, a wide-ranging array of evaluative responses have been studied by social psychologists. Although much of this work has focused on respondents’ attitudes toward social groups, individuals, and social policies (e.g., Communism, the right-to-life, or affirmative action), researchers have also commonly examined less overtly interpersonal attitudinal responses, such as those regarding personal behavior or idiosyncratic preferences (the value of tooth-brushing, the taste of anchovies, etc.).
Although social psychologists have traditionally relied on what respondents explicitly report about an attitude object, more recent work has focused on unconscious or implicit forms of evaluation that unfold automatically and without subjective awareness. Whether such implicit evaluations arise from the same processes as one’s conscious opinions and preferences continues to be a matter of some debate within the social psychological literature, although recent neuroimaging findings have suggested that implicit forms of evaluation may be distinct from explicitly-reported attitudes in relying on a neural system centered around the amygdala84. Such dissociations are reminiscent of the distinctions between explicit and implicit memory, which likewise appear to be two different systems of memory that rely on distinct neural systems.
Whereas social psychologists have often relied on what respondents explicitly articulate about an attitude object – that is their reported preferences – several other research traditions have studied preferences as they are revealed by an individual’s observable choices (Box 2). These literatures confirm the functional importance of ventral aspects of the MPFC for evaluating the desirability of a stimulus. Activity in this region has been observed to correlate with participants’ preference for one taste over another as revealed in a blind “taste test”30, as well as with their relative preference for immediate over delayed monetary rewards in an intertemporal choice paradigm (e.g., opting to receive $20 now versus a larger amount in a week)31. Moreover, beginning with the well-known case of Phineas Gage, neuropsychological research has repeatedly demonstrated that damage to ventral MPFC impairs one’s abilities both to evaluate competing courses of action32 and to revise earlier evaluations of a stimulus33.
Both social and clinical psychology have been centrally concerned with understanding the nature of emotional experience, that is, one’s subjective awareness of affective states and the consequences of such experience on behavior. The topics addressed by such researchers have ranged from the source of subjective emotional experience to the maladaptive effects of emotion that define many clinical disorders to the relation between emotion and “colder”, less affectively-based, mental operations. In addition, a sizeable literature has examined emotional expression, cataloguing the discrete types of facial expressions that accompany different emotions and examining how perceivers recognize the emotions of others34.
Although several brain regions make well-characterized contributions to the experience and recognition of particular emotions (such as the anterior insula to disgust and the amygdala to fear and anxiety)35, a somewhat underappreciated finding has been the generality of the response in MPFC during emotional experience. In a review of 55 neuroimaging experiments through 200236, the MPFC was the brain region most commonly associated with affective processing, regardless of the specific emotion being targeted (disgust, fear, sadness, anger, happiness). Interestingly, manipulations that induce particularly rich emotional experience were those most likely to engage MPFC37–48 (Fig 1c). For example, MPFC modulation is particularly likely when a person engages in extended affective processing that allows for genuine, subjective experience of emotion, such as by recalling an evocative autobiographical memory or viewing emotionally-charged films (in contrast to passively viewing affective words or still photographs). This observation suggests that the MPFC may specifically contribute to emotion by subserving the subjective experience of one’s affective state. More recently, neuropsychological research has confirmed the important role of MPFC in emotion, demonstrating that lesions to this region impair emotional experience as well as the recognition of emotional expressions49. Together, these results suggest that, although specific brain regions like the amygdala and insula may play critical roles in specific emotions, the MPFC plays a broad – albeit incompletely specified – role in emotional experience more generally.
Notwithstanding frequent forays into purely intrapersonal phenomena, social psychologists have long examined the question of how perceivers make sense of the behavior of others. For more than three decades starting in the 1960s, a sizeable literature developed around questions of attribution, such as how one determines whether an individual’s behavior (Mary is biting her finger nails) is better ascribed to her internal mental states and dispositions (she must be a nervous person) or to situational influences and constraints (she is waiting for the results of an important exam; for a comprehensive review of social psychological research on attribution, see Ref 50). More recently, researchers have begun to concentrate on one aspect of attribution, examining how perceivers generate their initial inferences about others’ mental states in the first place (e.g., how does one infer that Mary is feeling nervous to start with?), an ability referred to as “mentalizing” or “theory-of-mind”.
For all intents and purposes, neuroimaging studies have unanimously implicated MPFC in tasks that require perceivers to mentalize about the thoughts or feelings of others (Fig 1d). Recent reviews of the functional neuroanatomy underlying social cognition51, 52 have catalogued the wide range of contexts in which MPFC activity accompanies mentalizing. Greater response in this region has been observed when (i) perceivers regard stories or cartoons whose comprehension requires inferring the mental states of their protagonists (compared to understanding physical causality)53–55; (ii) answer questions about another person’s knowledge56–59; (iii) watch abstract cartoons that imply the presence of a mental agent60–62; or (iv) play a competitive game against a human (compared to a computer) opponent63, 64. Moreover, similar MPFC modulation has been associated with tasks originally developed within the social psychological literature on attribution, such as those designed to favor dispositional over situational attributions65 or during explicit attempts to form an impression of another person’s personality66–68. Neuropsychological results also confirm that, at least for nontrivial theory-of-mind tasks, damage to the MPFC impairs the ability to apprehend others’ mental states69, 70. And autism, which is marked by severe impairments in understanding others’ mental states, has been linked by at least two studies to abnormal activity in MPFC71, 72 (although the functional neuroanatomy underlying this disorder is far from completely understood).
To the extent that shared functional neuroanatomy implies shared cognitive processing2, the overlapping MPFC basis for the self-concept, attitudes, emotional experience, and mentalizing suggests that these seemingly diverse phenomena all draw on a common set of underlying mental operations. But what does the fact that the MPFC in particular subserves these social phenomena – and not some other brain region – imply about the nature of the processes underlying them? Interestingly, the MPFC has been implicated in a number of additional abilities that call for nonliteral, counterfactual, or probabilistic processing, such as understanding figurative linguistic constructions like metaphor and analogy73, 74, simulating hypothetical future events75, and reasoning about ambiguous moral conflicts76, 77. In sharp contrast, the MPFC has not only been only rarely implicated in most other cognitive activities but routinely demonstrates reduced response (i.e., “deactivation”) when participants engage in tasks involving semantic memory, executive function, perception, and many of the other types of processes studied by cognitive psychology78. Such deactivations have been argued to mark the suspension of an internally-focused mode of processing that would otherwise interfere with attention to the external environment75, 78.
Together, these neural observations support the view that ‘social’ phenomena can be distinguished from other kinds of cognitive processing by their dependence on a qualitatively distinct class of mental representation. Most cognitive abilities require exact representations that correspond veridically to the external world: people are generally surprised and consternated when they generate inexact or fallacious representations of the outside world; for example, misreaching for a wine glass and knocking it over, intending one word but blurting out another, or feeling confident in memories that prove to be false or distorted. In contrast, when it comes to our self-concept, attitudes, emotional experience, and understanding of others minds, we readily handle – indeed, may insist upon – considerably less exactitude and accuracy. Although we know roughly what defines us as a person, how much we like or dislike something, the strength of our current emotional experience, or what is going on inside the head of another person, the functional utility of these social processes does not rely on the ability to pinpoint an exact representation that corresponds precisely to an actual “fact of the matter” in the external world. Instead, social phenomena demand an ability to operate over ‘fuzzy’ mental estimates that are inexact, probabilistic, internally-generated, and subject to revision. Whereas abilities like motor control, language, and perception require the generation of discrete, specific representations, we typically experience our selves, our attitudes, our emotions, and the minds of others more like continuously shifting and indefinite approximations. Reifying these fuzzy experiences by assigning them specific labels (through language, Likert scales, etc.) either acutely disrupts normal functioning, as in the case of affective processing79, or else provides flawed or inadequate insights into their workings, as for our self-concept, attitudes, and social inferences80.
A possible exception might be our inferences about mental states, which can sometimes pertain to specific information that another person may or may not know. Indeed, a good deal of research in social cognition has examined tasks that imply “correct” answers about another person’s knowledge (such as the “Sally-Anne” false belief task). Interestingly, these tasks are most closely associated with activity in a region outside the MPFC, the temporo-parietal junction81. In contrast, many of our mental state inferences may center around fuzzier, more probabilistic estimates of others’ experience. For example, we might infer that someone is sad, but rarely need to estimate exactly how dysphoric. Or we might consider someone to possess a certain personality trait (intelligence), but rarely consider exactly to what extent.
By increasingly adopting the methods of cognitive neuroscience, social psychologists have discovered a previously unsuspected correspondence among many of the important phenomena at the core of the field. Such observations underscore the unique power of functional localization methods, such as neuroimaging, to uncover links among researchers who once believed themselves to be studying disparate empirical issues, but we now understand to have been probing different manifestations of a common underlying system. This neurally-inspired ‘lumping’ of seemingly disparate phenomena promises not only to help underscore what makes social psychology distinctive, but suggests the need to rethink the assumption that the field studies phenomena at a “higher” or more “macro” level than cognitive psychology. Rather than equating the study of social phenomena with a particular level of analysis, these findings suggest a view of social psychology as a unique branch of cognitive science, specialized for examining a distinct and natural grouping of approximate, shifting, and internally-generated – in other words, ‘fuzzy’ – cognitive operations.
The preparation of this article was supported by NSF and NIA grants to the author. Thanks to Mahzarin Banaji, Wendy Berry Mendes, Randy Buckner, Dan Gilbert, Abby Klima, Neil Macrae, Lindsey Powell, Rebecca Saxe, Diana Tamir, Dan Wegner, and Jamil Zaki for helpful comments and discussion and Dave Johnson and Jessica Schirmer for their assistance in preparing the manuscript.