University of California-Berkeley University of Virginia
[Full reference: Keltner, D., & Haidt, J. (2001). Social functions of emotions. In T. Mayne & G. A. Bonanno (Eds.), Emotions: Current issues and future directions. New York: Guilford Press. (pp. 192-213).]
Please address correspondence to: Dr. Dacher Keltner, Department of Psychology, 3210 Tolman Hall, University of California, Berkeley CA, 94720. Internet: firstname.lastname@example.org.
Social Functions of Emotions The primary function of emotion is to mobilize the organism to deal quickly with important interpersonal encounters (Ekman, 1992, p.171).
Emotions are a primary idiom for defining and negotiating social relations of the self in a moral order (Lutz & White, 1986, p.417).
Emotion theorists disagree in many ways, but most share the assumption that emotions help humans solve many of the basic problems of social living. For evolutionary theorists, emotions are universal, hard-wired affect programs that solve ancient, recurrent threats to survival (Ekman, 1992; Lazarus, 1991; Plutchik, 1980; Tomkins, 1984; Tooby & Cosmides, 1990). For social constructivists, emotions are socially learned responses constructed in the process of social discourse according to culturally specific concerns about identity, morality, and social structure (Averill, 1980; Lutz & White, 1986). These contrasting approaches conceive of the defining elements, origins, and study of emotions in strikingly different ways, but both ascribe social functions to emotion.
In this chapter we present a social-functional account of emotions that attempts to integrate the relevant insights of evolutionary and social constructivist theorists. Our account can be summarized in three statements: 1) Social living presents social animals, including humans, with problems whose solutions are critical for individual survival; 2) Emotions have been designed in the course of evolution to solve these problems; 3) In humans, culture loosens the linkages between emotions and problems so that cultures find new ways to solve the problems for which emotions evolved, and cultures find new ways of using emotions. In the first half of the chapter we synthesize the positions of diverse theorists in a taxonomy of problems of social living, and then consider how evolution-based, primordial emotions solve those problems by coordinating social interactions. In the second half of the chapter we discuss the specific processes according to which culture transforms primordial emotions, and how culturally shaped, elaborated emotions help solve the problems of social living.
Evolutionary theorists concern themselves with universal, biologically based, genetically encoded emotion-related patterns of appraisals and responses observed across species and cultures, which we will refer to as primordial emotions. Primordial emotions are shaped by evolutionary forces, genetically encoded, embedded in the human psyche, linked to biological maturation, and they involve coordinated physiological, perceptual, communicative, and behavioral processes that are meant to produce specific changes in the environment. They occur most typically within the context of immediate face-to-face interactions, and they are brief, lasting seconds to perhaps minutes, but not hours or days (Ekman, 1992).
Although functional analyses have been a mainstay of evolutionary theorizing about emotion (e.g., Campos, Campos, & Barrett, 1989; Darwin, 1872; Izard, 1977; Plutchik, 1980), it is only recently that theorists have begun to systematically link specific emotions to social functions. This recent theoretical development can be attributed to several sources (Barrett & Campos, 1987). Advances in ethology and behavioral ecology have illuminated the specific advantages and problems posed by group living (Krebs & Davies, 1993). Working within different traditions, theorists have begun to characterize the connections between specific emotions and attachment (e.g., Kunce & Shaver, 1987), mate selection and protection (Buss, 1992), game theoretical characterizations of altruism, cooperation, competition, and interpersonal commitment (Frank, 1988; Nesse, 1990; Trivers, 1971), and dominance and submissiveness (e.g., Keltner & Buswell, 1997; Ohman, 1986).
We believe these recent developments can be summarized in a social functional approach to emotion (Frijda & Mesquita, 1994; Keltner & Haidt, 1999; Keltner & Kring, 1998). Social functional accounts operate at multiple levels of analysis, specifying the social benefits emotions bring about for the individual, dyad, group, and culture. Across diverse methods and levels of analysis, social functional accounts share certain assumptions. Most notably, social functional accounts assume that group living, which has been characteristic of humans and other species for millions of years, confers many advantages over solitary living. These advantages include more proficient food gathering, responses to predation, and raising of offspring. Group living also creates new problems requiring the coordination of group members.
To meet these problems and opportunities, humans have evolved a variety of complex systems. Each system is organized according to a specific goal (e.g., to protect offspring or maintain cooperative alliances) that is served by multiple subsystems. These include specific perceptual processes, higher order cognition, central and autonomic nervous system activity, as well behavioral responses, both intentional and reflex-like. For example, theorists have observed that humans form reproductive relationships with the help of an attachment system (e.g., Bowlby, 1969; Kunce & Shaver, 1987). The attachment system involves perceptual sensitivities to potential mates, representations of relationships, autonomic and hormonal activity related to affiliative, sexual, and intimate behavior, behavioral routines such as flirtation, courtship, and soothing, and specific emotions as we discuss below.
Within each system, primordial emotions serve two general functions (see also Johnson-Laird & Oatley, 1992). First, primordial emotions signal that action is necessary, either because of a deviation from an ideal state of social relations, or because an opportunity presents itself. Primordial emotions therefore involve perceptual, appraisal, and experiential processes that monitor the conditions of ongoing relations, detecting disturbances (e.g., an infant’s distress) or opportunities (a potential mate). Once activated, emotion-related perception and experience interrupt ongoing cognitive processes and direct information processing to features of the social environment that allow for the restoration or establishment of desirable social relations (Clore, 1994; Lazarus, 1991; Lerner & Keltner, in press; Schwarz, 1990).
Second, primordial emotions motivate behavior that establishes (or reestablishes) more ideal conditions of social relations. Primordial emotions involve autonomic, hormonal, and central nervous system activities that are tailored to specific social actions, such as fighting, copulating, offering comfort, and signaling dominance (Davidson, 1980, 1993; Frijda, 1986; Levenson, 1994; Le Doux, 1996; Porges, 1995; Sapolsky, 1989). Primordial emotions also involve vocal, facial, and postural communication that: provide quick and reliably identified information to others (Ekman, 1984, 1993; Izard, 1977; Scherer, 1986), which shapes social interactions as we detail in a section that follows. We now consider how different emotions might help solve the different problems of group living.
The Problems of Group Living and Primordial Emotions
Functional accounts begin with an analysis of the problems emotions were presumably designed to solve, either by evolution or cultural construction (Keltner & Gross, 1999). Our review of the theorizing on the social functions of emotion identifies three classes of problems related to group living to which emotions are intimately linked, and we would argue, have been designed to solve. Table 1 summarizes the nature of these problems, the general systems that have evolved to meet these problems, as well as related emotions and their specific functions. We note that Table 1 does not include all states we consider emotions (e.g., amusement does not fit neatly into the taxonomy); rather it lists the emotions that seem well suited to solving specific social problems. Additionally, although Table 1 describes how emotions are linked to prototypical social objects (e.g., sympathy for vulnerable family members), we believe emotions generalize to related social objects (e.g., sympathy for downtrodden members of society). Several theorists have commented on the flexibility of the associations between emotions and their intentional objects (e.g., Tomkins, 1984), which we discuss in an ensuing section.
The problems of Physical Survival
First, individuals must solve the problems of physical survival, including avoiding death by predation, violence, and disease. Fear is the primordial emotion at the heart of the “fight-flight” system (Ohman, 1986), which helps individuals avoid death by predation or other physical attacks. Much is understood about the physiology of fear (LeDoux, 1996). On the appraisal side, the amygdala contains specialized areas that scan incoming sensory information for patterns that have been associated with danger. The amygdala can trigger a fear response even before the incoming information has been sent to the occipital cortex for full processing, and even before an individual knows what an object is (LeDoux, 1996; see chapter XX for further elaboration). On the output side, primordial fear involves triggering the hypothalamic-pituitary-adrenocortical axis, which pumps a quick dose of cortisol and other stress hormones into the bloodstream, to ready the organism for fight or for flight. Primordial fear can be seen as the heart of a system that includes a variety of cognitive and behavioral mechanisms that make it more effective, e.g., vicarious learning, and the preparedness of animal phobias (Mineka & Cook, 1988).
Disgust can similarly be seen as the primordial emotion at the heart of the “food-selection” system (Rozin, 1976b), which helps humans choose a balanced and safe diet. Unlike fear, disgust is not found in other animals; only a simpler precursor, distaste, can be seen in rats, and other generalist animals. Co-evolving with the tremendous expansion of cognitive ability in humans, the distaste response has expanded to become the disgust response (Rozin, Haidt & McCauley, in press). In humans, food rejections are not based primarily on the sensory properties of the object, but rather on a knowledge of what it is, or what it has touched (Rozin & Fallon, 1987). Potential foods elicit disgust if they resemble, or have come in contact with, certain powerful elicitors of disgust, such as feces and decaying animal bodies. The food selection system is further expanded by the addition of learning mechanisms, such as one-trial learning for nausea-inducing foods (Seligman, 1971), and by cultural mechanisms, such as cuisine, which marks prepared foods with a reassuringly familiar blend of spices or flavors (Rozin, 1996).
The Problems of Reproduction
Evolutionary and attachment theorists have speculated how a variety of emotions solve the problems of reproduction, which include procreation and the raising of offspring to the age of reproduction. The problems of finding and keeping a mate are in part met by emotions of romantic love and desire, which facilitate the identification, establishment, and maintenance of reproductive relations. These emotions involve appraisals, perceptions, and experiences that are sensitive to cues related to potential mate value. These include beauty, fertility, chastity, social status, and character (Buss, 1992; Ellis, 1992), expressive behaviors that signal interest and commitment (Frank, 1988) and evoke desire and love, and hormonal and autonomic responses that facilitate sexual behavior (Davidson, 1980). In table 1 we also contend that the experience and display of emotions related to the loss of a partner provoke succorance in others, and eventually helps individuals establish bonds with new mates. We label this emotion sadness (for similar analysis of distress, see Dunn, 1977; for analysis of grief, see Lazarus, 1991).
The protection of potential mates from competitors is equally critical. Jealousy, the literature shows, relates to mate protection, and is triggered by cues that signal potential threats to the relationship, such as possible sexual or emotional involvement of the mate with others (Buss, 1992). Jealousy motivates possessive and threat behaviors that discourage competitors and prevent sexual opportunities for the mate (Wilson & Daly, 1996).
Mammalian neonates are extremely dependent and vulnerable to predation, and continue to be so for much longer periods of time compared to other species. As a consequence, social species have evolved caregiving-related emotions of parental and child love and sympathy, which facilitate protective relations between parent and offspring (Bowlby, 1969; Kunce & Shaver, 1987). The caregiving system involves perceptions and experiences that sensitize parents to infantile cues (e.g., of neotany, distress) and infants to vocal and visual cues of parenthood (Fernald, 1992). Filial love and sympathy are characterized by experiences and expressive behavior such as mutual smiles and gaze patterns. These are elements of interactions that strengthen loving bonds, and physiological responses that help caretakers respond to others’ distress (e.g., Eisenberg, et al., 1989).
Finally, theorists have argued that emotions help solve two subclasses of problems related to group governance, which arise in several contexts, including the allocation of resources and distribution of work (Fiske, 1991). First, to avoid the problems of cheating and defection and to encourage cooperation, in particular among non-kin, humans reciprocate cooperative and non-cooperative acts towards one another (Trivers, 1971). Reciprocity is a universal social norm (Gouldner, 1960), and is evident in gift giving, eye-for-an-eye punishment, quid-pro-quo behavior in other species (de Waal, 1996), and the tit-for-tat strategy (Axelrod, 1984). Several emotions signal when reciprocity has been violated and motivate reparative behavior (de Waal, 1996; Frank, 1988; Nesse, 1990; Trivers, 1971). Guilt occurs following violations of reciprocity and is expressed in apologetic, remedial behavior that re-establishes reciprocity (Keltner & Buswell, 1996; Tangney, 1991). Moral anger motivates the punishment of individuals who have violated rules of reciprocity, and is defined by the sensitivity to issues of justice and unfairness (Keltner et al., 1993). Gratitude at others’ altruistic acts is a reward for adherence to the contract of reciprocity (Trivers, 1971). Envy motivates individuals to derogate others whose favorable status is unjustified, thus preserving equal relations (Fiske, 1991).
Second, humans must solve the problem of group organization. Status hierarchies provide heuristic solutions to the problems of distributing resources, such as mates, food, and social attention, and the labor required of collective endeavors (Fiske, 1990; de Waal, 1986, 1988). Hierarchies are dynamic processes, and require continual negotiation and redefinition. The establishment, maintenance, and preservation of status hierarchies is in part accomplished by emotions related to dominance and submission (de Waal, 1996; Ohman, 1986). Embarrassment and shame appease dominant individuals and signal submissiveness (Keltner & Buswell, 1996; Miller & Leary, 1992), whereas contempt is defined by feelings of superiority and dominance vis-a-vis inferior others. Awe tends to be associated with the experience of being in the presence of an entity greater than the self (Keltner & Loew, 1996). In some cases, awe is elicited by nature or art (Shin, Keltner, Shiota, & Haidt, 2000). In other instances, awe is elicited by the presence of higher status others, and thereby endows higher status individuals with respect and authority (Fiske, 1991; Weber, 1957).
Primordial emotions and social interaction
We believe that emotions most typically solve the problems of social living in the context of ongoing face to face interactions (e.g., Frijda & Mesquita, 1994; Keltner & Kring, 1998). This view has typically been espoused by those who argue that emotions are constructed within social relationships, and by implication, are not biologically based or universal (e.g., Lutz & Abu-Lughod, 1991). Evolutionary theorists, however, have long suggested that humans evolved adaptive responses to the emotional responses of others (e.g., Darwin, 1872; Ohman & Dimberg, 1978), consistent with the claim that communicative behavior of senders and receivers co-evolved (Eibl-Eibesfeldt, 1989; Hauser, 1996). From this perspective, one individual's emotional expression serves as a "social affordance" which evokes "prepared" responses in others. Primordial emotions structure social interactions in at least three ways.
First, emotional displays evoke complementary emotional responses in others (for more complete review, see Keltner & Kring, 1998). These complementary emotions are core elements of interactions such as courtship, bonding, appeasement, and reconciliation (Eibl-Eibesfeldt, 1989). Thus, soothing interactions involve sympathetic responses to another’s distress. Some socialization interactions involve angry responses to another’s transgression, followed by a shamed or guilty response to the anger (Ausubel, 1955; Gibbard, 1990). Appeasement interactions involve displays of submissive emotions, such as embarrassment and shame, that evoke reconciliation related emotions in the observer, such as amusement (in the case of embarrassment) and sympathy (in the case of shame) (Keltner, Young, & Buswell, 1997). Displays of love and desire coordinate the flirtatious interactions of potential romantic partners (Eibl-Eibesfeldt, 1989).
Second, emotional communication conveys information about the sender’s mental states, intentions, and dispositions, which are critical to social interactions (Fridlund, 1992). The empirical literature suggests that emotional displays provide rapid, fairly reliable cues of the sender’s emotion, intentions, and disposition (for review, see Keltner & Kring, 1998). Contagious emotional responses provide a more direct route to the understanding of others’ mental states, leading individuals to experience similar responses to objects or events in the environment, coordinating individuals’ perception and action (Hatfield, Rapson, & Cacioppo, 1994).
Finally, emotions serve as incentives for others’ social behavior: an individual’s expression and experience of emotion may reinforce another individual’s social behavior within ongoing interactions (e.g., Klinnert et al., 1983). For example, the display of positive emotion by both parents and children rewards desired behaviors and shifts in attention, thus increasing the frequency of those behaviors (e.g., Tronick, 1989). The literature on social referencing indicates that displays of more negative emotions deter others from engaging in undesirable behavior (Klinnert et al., 1983).
To summarize, we have argued that group living presents humans with the problems of physical survival, reproduction, and group governance. Humans have evolved complex systems to meet these problems and opportunities, and emotions serve important functions within these systems by signaling that problems or opportunities exist and by coordinating the actions of interacting individuals. We now consider how culture elaborates upon primordial emotions.
Culture and the social functions of elaborated emotions
Evolutionary perspectives on emotion have always included a role for culture. Darwin begins The Expression of the Emotions in Man and Animals (1872) with a description of a cross-cultural study in which he sent questionnaires to missionaries around the world, asking them to comment on the facial and bodily expressions of emotions shown by non-Western people. He concluded that some expressions are highly universal, whereas others are more variable, a conclusion that has withstood the test of time (Haidt & Keltner, 1999). Ekman has also always included an important role for culture in his “neuro-cultural” theory of emotions (1972). Ekman built on Tomkins’ (1963) notion of universal “affect programs”, suggesting that culture plays its role as a modulator of both inputs (what counts as an insult, or a loss?) and outputs (display rules about which emotions can be expressed in which circumstances).
While Ekman’s critics have often ignored his writings about culture (e.g., Ekman, 1972), they have also pointed out ways in which culture may play a more profound role in human emotional life. Impressed by how culture and language give humans flexibility and creativity in designing their lives and societies, social constructivists concern themselves with the total package of meanings, social practices, norms, and institutions that are built up around emotions in human societies (Lutz, 1988; Lutz & Abu-Lughod, 1990; Shweder, 1993; White, 1990). We refer to these complex meanings as elaborated emotions. Elaborated emotions are shaped by social discourse and interaction, and by concepts of the self, morality, and social order (Markus & Kitayama, 1991; Shweder & Haidt, in press). Elaborated emotions vary across cultures, they cannot be experienced by infants, and they can last for years or centuries. For example, the hatred felt towards an historical enemy, although at any one moment in time comprised of brief emotional experiences, is made up of values, beliefs, images, action tendencies, and affective dispositions or sentiments that can pass from one generation to the next and last for extended periods of time1.
In human beings, culture alters the use and expression of many evolved traits or systems, including primordial emotions, in several ways. Childhood instruction, culturally specific environmental conditions, and personal experiences determine culture specific elicitors of emotions, and shape the manner in which primordial emotions develop and are expressed. We will analyze cultural variation in emotion by applying our functionalist perspective to the sorts of objects that anthropologists and social constructivists are concerned with: meaning systems and emergent social phenomena such as institutions and practices.
Modern approaches to culture often focus on how humans create the symbolic and material worlds that then shape, constrain, and enable them, as in Shweder’s (1989) succinct formulation, ”culture and psyche make each other up”. Cultures draw on many sources to create their symbolic worlds, including the human body and the phenomenological experiences it provides (Geertz, 1973; Lakoff, 1987). All cultures have noticed that people experience hunger, fatigue, illness, and sexual arousal, and all cultures have developed ethnotheories, customs, and practices that explain the origins and govern the interpretation of these experiences. Primordial emotions, we propose, provide universally available patterns of perception, experience, and action that cultures work into their ethnotheories, values, practices, and institutions. The primordial emotion of anger, for example, might be elaborated into a codified defense of the social-moral order, linked to satisfying feelings of righteousness, and valued as a prosocial force (e.g., the Ifaluk emotion of song, Lutz, 1988). Conversely, anger might be elaborated as a childish and destructive response whose expression is suppressed, leading to the use of other methods to solve interpersonal disputes (Briggs, 1970).
Culture and the elaboration process loosen the link between a primordial emotion and the social problem it was designed to solve in two ways, as we elaborate below. First, cultures find new solutions to the ancient problems that emotions were designed to solve. Second, cultures find new uses for old emotions that have little to do with their “original” function. Once this loosening is recognized it becomes easier to reconcile evolutionary approaches (which focus on primordial emotions, and therefore find universality) and social constructionist approaches (which focus on elaborated emotions and therefore find cultural variation).