Affect Control Theory Across Cultures

David R. Heise
Department of Sociology
Indiana University
Bloomington, IN 47405

Paper presented at "Theory Section Miniconference: New Directions in Sociological Theory: Growth of Contemporary Theories," Annual Meetings of the American Sociological Association, Washington, DC, August, 2000.

Affect control theory culminates sociologically in what I call the principle of affective rationality. The principle is at work in the hotel lobby downstairs. Sociologists mill around and do all sorts of actions that are predicted theoretically--like talking to one another, confiding, consulting, toasting, treating, complimenting, congratulating, and so on. However, let these sociologists encounter a poorly dressed stranger begging for money, and they change to different actions, also predicted by affect control theory--like observing the stranger, perhaps addressing him, placating him, even giving him money for a meal. The actions of sociologists with each other are rational in the sense that these are just the kinds of behaviors required of individuals in the colleague role, as they go about their professional work with each other. The actions toward the beggar also are rational in constituting a sensible way for privileged individuals to control exchanges with underprivileged individuals.

Our professional activities are functional and rationally analyzable. Yet we rarely are involved in functional or cost-benefit analyses while engaging in these behaviors. Our actions emerge from our hearts. That is why these kinds of behaviors can be predicted from affect control theory--the actions are affectively derived. This anomaly--that our sensible actions usually are affectively based--is the principle of affective rationality. We often unfold our rational actions intuitively, rather than by analysis.

This principle is a pivot for many kinds of discussions, and I will consider only two in this brief talk today. First, I'll sketch the affect control model that predicts the actions of people in different roles. Second, I'll consider how components of this model vary cross-culturally and what some implications of those variations could be.

The Theoretic Formulation

The essence of affect control theory (Heise, 1979, 1999; Smith-Lovin and Heise, 1987; MacKinnon, 1994) can be summarized in three ideas. (1) Individuals create events to confirm the sentiments that they have about themselves and others in the current situation. (2) If events don't work to maintain sentiments then individuals re-identify themselves and others. (3) In the process of building events to confirm sentiments, individuals perform the social roles that operate society--the principle of affective rationality.

Social processes of concern in affect control theory are shown in Figure 1. The numbered paragraphs below elucidate these processes, with paragraph numbers corresponding to the sequence numbers in Figure 1.

On entering a social situation, an individual recognizes self and others in terms of social identities. Sentiments associated with these identities and with the setting itself organize the fundamental affective meanings of the scene.
The individual compares fundamental meanings in the situation with feelings available from recent experience. If experience has failed to confirm fundamental meanings, the individual conceptualizes an event to actualize the fundamental meanings. The event becomes a goal for the individual.
The individual initiates the goal event, or else prompts another actor to begin the event. Alternatively, the individual may construe happenings in progress as the goal event.
The individual perceives activities that are self constructed, activities carried out by others at the instigation of the individual, and activities produced by others acting independently of the individual.
The individual assesses perceived activities relative to the goal event. Incompletion of the goal event instigates effort to complete the event through the observer's own behavior, by prompting others, or by re-interpreting prior behaviors.
- This is the "test" phase in the Test-Operate-Test-Exit loop of Miller, Galanter, and Pribram, 1960.
- Persistent incompletion of a goal may produce stress, according to Burke (1991). This cognitive stress adds to affective generation of stress when sentiments are disconfirmed persistently--at step 2.
If it is impossible to complete the goal event, then the individual may re-conceptualize the event for actualizing affective meanings in current experience. For example, the individual may decide to perform an action herself rather than let someone else do it.
The completed event changes feelings about entities in the event. This is the impression-formation process studied qualitatively by Goffman (1959) and quantitatively in affect control theory. After impression formation, the process loops back to step 2 and:
- Feelings about self produced by current events combine with self sentiments to generate emotions that reflect one's predicament in current events (Averett and Heise, 1987; Heise and Thomas, 1989; Heise and Weir, 1999).
- Current feelings about all elements in the situation are compared with sentiments. If experience disconfirms fundamental meanings, then the individual conceptualizes a new event, and the cycle begins anew.
If the individual cannot reduce large deflections by implementing events, then the sentiments to be confirmed may be changed through attributions or labelings that change people's identities.
If events have realized the relevant affective meanings in the situation, then the person is free to move to a new setting or to another agenda item.

Nothing in this formulation is concerned with rational choice or social analysis. Instead humans are viewed as meaning-maintainers, who continually reconstruct the world to fit intuitive knowledge generated from sentiments within cognitive and logical constraints. In this perspective, rational analysis is rare rather than routine. When successful problem solving does occur, it quickly is assimilated into the affective meaning system and replayed thereafter as intuitive knowledge.

Cultural Variations

Cultural factors might shape any of the processes represented in Figure 1. However, current research is focusing on two matters--sentiments, and generation of feelings--and those are the topics I consider next. First, though, I need to introduce the measurements that are used in this cross-cultural work.

Osgood, May, and Miron (1975) laid the foundation for cross-cultural studies of affective meaning with research verifying three universal dimensions of affective response. Indigenes in more than two dozen societies were presented with a list of concepts that exist in every culture--like father, mother, girl, water, moon--and asked to respond to each concept with a modifier. Later they were asked to name the opposite of each modifier. For example, some individuals in the U.S.A. might respond to mother with the word sweet, and give the word sour as the opposite. The modifier opposites were formed into rating scales, and indigenes in each culture used the scales to rate each concept. Ratings were averaged to get numbers indicating how raters from that culture typically positioned the concepts on each scale. Then pan-cultural correlations were computed comparing mean ratings of the concepts, on scales administered in the same society and in different societies. For example, a pan-cultural correlation can be computed comparing mean ratings of the concepts on a sweet-sour scale used in the U.S.A. and on a buen-malo scale used in Mexico.

Statistical analysis of the pan-cultural correlations showed that scales clustered into three major groups--Evaluation, Potency, Activity--and every culture contributed scales to each group. For example, concepts rated as sweet by Americans tended to be rated good by Americans and buen by Mexicans, so all three of these scales--sweet-sour, good-bad, and buen-malo--contributed to the Evaluation cluster. Evaluation concerns a sense of approval or disapproval that can elaborate into judgments of morality, aesthetics, functionality, hedonism, or other standards. Potency relates to an entity's impact, and might elaborate into assessments of physical magnitude, strength, forcefulness, social power, significance, etc. Activity indexes an entity's spontaneity, which can elaborate into judgments of agency (propensity for being an actor), speed, perceptual stimulation, age, keenness, etc.

Scales for measuring sentiments are semantically anchored at either end by affectively-loaded adjectives. Adverbs characterize rating positions along the scale. The custom is to code the scale mid-point as zero, and use plus units to measure goodness, powerfulness, and liveliness; and minus units for bad, powerless, or quiet. For example, something that is rated "slightly good" gets coded +1 on Evaluation; something that is rated "quite good" gets coded +2; something rated "extremely good" gets coded +3; and something rated "infinitely good" gets coded +4.3. An EPA profile is a list of three such measurements: the first number represents Evaluation, the second is Potency, and the third is Activity.

A cultural sentiment about an entity is measured as the average EPA rating of the entity outside of the context of any event from multiple indigenes. A contextual feeling about an entity is measured as the average EPA rating of the entity in an event context. For example, we might ask respondents to rate the policeman in the sentence, "the policeman clubbed the man."

Sentiments

Pan-cultural factors could not have emerged unless people in different cultures have similar sentiments regarding basic concepts. However, the similarity begs the question, Don't cultures differ in sentiments?

Osgood, May, and Miron (1975) analyzed cross-cultural sentiments about colors, and they found cultural uniformities as well as variations. I myself (Heise, 1987) used their data to examine sentiments about seven age-sex roles across 16 cultures. I found that in most--but not all--of the 16 societies, male roles were more potent than female roles, and identities of the youngest individuals were felt to be more active than identities of the oldest individuals. Among the interesting exceptions: female roles were more potent than male roles in Brazil, and in Malaysia babies were felt to be inactive.

In the 1970s, social psychologists working in the United States began measuring hundreds of sentiments about social identities and social behaviors, providing rich empirical materials relating to the affective organization of social life. The possibility of cross-cultural analyses with such materials began with a data set obtained from Northern Ireland. Then, opportunities for cross-cultural analyses grew with Neil MacKinnon's acquisition of data in Canada, with Andreas Schneider's data collection in Germany, and with Herman Smith's data collection in Japan, and very recently in China. I now report some analyses of these data sets.

Figure 2 shows cultural sentiments for father, mother, and child as measured among indigenes of the U.S.A., Canada, Japan, China, Germany, and Northern Ireland. Father is at the topmost end of each line, child is at the bottom end of each line, and mother is the middle point of each line. This chart is based on female sentiments, but it would look about the same were sentiment measurements from males used instead.

The chart reveals that female raters in all six cultures agree that fathers, mothers, and children are not bad, and mothers are the nicest of the three. Additionally all agree that parents are powerful and children are powerless. However, aside from these generalities, major differences arise.

Japanese evaluate family members less positively than people in the other cultures, and a child actually is felt to be neither good nor bad in Japan.
Chinese evaluate family members most positively; and the Chinese are different from people in the other five cultures in feeling that mothers are more powerful than fathers.
Generally, parents are evaluated more positively than children, but not in Germany where fathers are felt to be less good than either children or mothers.
Mothers generally are felt to be nicer than fathers, but this difference is negligible in the U.S.A. The power difference between fathers and mothers also is negligible in the U.S.A.

These results typify cross-cultural analyses of sentiments. While people everywhere share some general perspectives, specific patterns of sentiments vary from one culture to another. In this particular case, we see a similar family structure in all six cultures--parents are more powerful than children, and mothers have highest status. However, the differences in sentiments between, say, China and Japan are enough to create different styles of parenting. According to affect control theory simulations[1], for example, both relationships are nurturing but Chinese fathers slip readily into a coaching sub-role, whereas Japanese fathers slip readily into a disciplinarian sub-role.

Correlation analysis provides another way to assess cross-cultural patterns of sentiments. For example, matching Canadian and Japanese identities item by item can yield a formula for predicting Japanese evaluations from Canadian evaluations--Figure 2 suggests that one might predict the Japanese evaluation of a family identity by subtracting about 1.2 from the Canadian evaluation. Correlation coefficients measure how accurate such predictions are.

Table 1 presents the correlations showing how well U.S., German, Japanese, Canadian, Irish, and Chinese sentiments predict sentiments in the other cultures. The table treats males and females separately, but I will not focus on sex since male-female results mostly are parallel.

Evaluative structures are remarkably similar in these six societies, with a mean cross-culture correlation coefficient of 0.82 for social identities and 0.85 for behaviors. That means that people brought up in Asian, European, and North American cultures largely agree about who is relatively good and who is relatively bad, and which actions are relatively right and which are relatively wrong. Japan is somewhat distinctive in the way it assigns esteem to people, and China is somewhat distinctive in its conceptions of morality. However, these Asian idiosyncrasies mainly are departures from the standards of European nations, and in any case none of the inter-cultural correlations involving evaluations go below 0.67.

Notions of who is relatively powerful and who is relatively powerless also are similar across societies, though Germany departs a bit from the international congruence. However, consensus drops dramatically in judgments of which acts have potent impact, particularly for the two European cultures, which disagree even with each other.

Ideas about who is relatively active or passive--who is likely or unlikely to exercise agency--are moderately shared across the cultures, with an average inter-culture correlation of 0.61. Ideas about the spontaneity of actions also are somewhat shared, with an average inter-culture correlation of 0.59, excluding China. China is very distinctive with regard to assessing behavioral spontaneity, correlating essentially zero with other cultures. In part, this is because Chinese subjects rate interpersonal malevolence as non-spontaneous. For example, Chinese subjects rate 25 social behaviors more than 2.0 units lower in activity than do U.S.A. subjects, and all but one of these are bad actions[2]. By contrast, only four behaviors--all good--are more than 2.0 units higher in activity.

Overall, these results indicate that an international order circumscribes judgments of morality and the allocation of honor and stigma, power and oppression in interpersonal relations. Individual societies adjust this order in only limited ways. Greater diversity between cultures emerges with regard to the amount of agency allocated to different roles[3]. However, the primary realm of cultural diversity is in definitions of the weightiness and spontaneity of social actions.

On the whole, U.S.A. sentiments give better predictions of sentiments in other cultures than do sentiments of any other society. This perhaps reflects the hegemony of U.S.A. culture. Or it may be more accurate to speak of North America's hegemony, since correlations involving Canada are almost as high as for the U.S.A.

Impression-Formation Processes

An event changes pre-event feelings about actor, behavior, object, and setting into new feelings. The pre-event feelings could be drawn from our sentiments about the event's elements, or they could be feelings that developed as a result of recent happenings.

For example, suppose we observe an employer cheating an employee. Previously we might have felt positive toward the employer and employee. However, seeing the employer cheat the employee makes the employer seem very bad. The event causes neutralization of the employee, too, as if we allow for the possibility that this employee earned victimization. Even the act of cheating is affected by the event: it is bad in this context, but not as bad as usual, as if cheating is less wicked as a workplace happening.

Social psychologists have developed equations for accurately predicting outcome impressions from the set of EPA profiles for pre-event feelings. Each of the many terms in these equations represents a mental process that occurs while interpreting events. I'll mention just three.

Every equation has a stability term. That fact means that the mind always transfers some pre-event feeling toward an event element to the post-event feeling involving the same event element. For example, we are inclined to see actors as good after events if the actors were good to begin with, and we tend to see actors as bad after events if they were bad before the event.

In what might be called a morality effect, our evaluation of an actor's behavior strongly influences the impression of the actor's goodness or badness. For example, anyone rescuing another gets evaluative credit for engaging in a noble act. Anyone killing another is discredited for engaging in a horrible act.

At the same time, though, we also are influenced by how evaluatively consistent the behavior is with the object of action. For example, an actor who performs a bad action on a good person violates a consistency principle--that good objects deserve good treatment--so the actor seems bad not only because of the morality effect but additionally because of behavior-object inconsistency.

Right from the beginning researchers in the affect control theory tradition wondered whether processes involved in impression formation were the same across cultures. Smith-Lovin (1988) reported results of a meta-analysis based on studies of subjects from the U.S.A., Ireland, and the Middle East. She found that the three processes I just mentioned--stability, morality, and behavior-object consistency--significantly influence evaluation of an actor in all of the subject groups she considered. Overall, Smith-Lovin found considerable cross-cultural similarity in the equations predicting assessment of an actor's goodness and activity. The universality of core processes affecting evaluation of actors also has been confirmed in Canada by MacKinnon (1994) and in Japan by Smith, Matsuno, and Umino (1994).

On the other hand, Smith-Lovin (1988) reported interesting differences in how Arabic speakers assess an actor's potency, as compared to English speakers. Subjects in the U.S.A. and Ireland (and also it turns out in Canada and Japan) feel that actors are especially powerful when they engage in potent actions, while the powerfulness of the object of action has little impact. For Arabs, though, resorting to strong actions implies that an actor is powerless. Meanwhile, object potency influences how potent an actor seems: engaging powerful others makes an actor seem more potent, and acting on weak others costs an actor potency. Smith-Lovin forwent speculation about this difference because of methodological limitations in the studies she considered, but it is interesting to employ the technology of affect control theory in order to work out what such a difference could mean for international understanding.

Here is an example[4]. Among people in the non-Arabic world it is reasonable for a master nation like the U.S.A. to overwhelm a villain nation, like Libya or Iraq, and doing so would make the winners feel satisfied and think of themselves as shrewd, while people in the overwhelmed villain nation should feel shocked, and should start being more agreeable. However, this may not be how the same event plays for Arabs, even if the Arabs accept the same identifications of winner and villain nations. For Arabs, a winner overwhelming a villain is feeling anxiety, and such an actor would have to be labeled as cowardly, whereas the object of action will be viewed as brave, even though shocked by the action. Such an interpretation seems to resonate with some rhetoric that comes out of the Middle East, and helps explain why East-West disagreements are so intractable in the Middle East!

Herman Smith has been exploring impression formation processes in Japan and China. As I just mentioned, Smith, Matsuno, and Umino (1994) found that Japanese actually are similar to Americans in how they construct outcome evaluations about an event's actor, behavior, and object, though overall Japanese interpretations of events might be somewhat simpler. On the other hand, Smith, Matsuno, and Ike (Forthcoming) found Japanese-American differences in interpreting an individual's state of being, as designated by phrases like "angry admiral" or "tactless doctor" or "rich professor," and, overall, Americans seem to process states of being more simply than Japanese. First of all, Americans employ the same principles to assess how an identity combines with an emotion, as opposed to a trait, as opposed to a status characteristic. For example, Americans feel about the same whether encountering an angry admiral or a suspicious admiral. This is not so in Japan where the admiral with a negative emotion would be evaluated more negatively than the admiral with a negative trait. Secondly, Americans average feelings about the modifier and identity to arrive at an evaluation of a modifier-identity combination and additionally employ a consistency principle in their assessment of the combination. Japanese also average and employ consistency; but, in addition, Japanese attend to intricate considerations such as whether a potent or impotent modifier is describing a good weak person as opposed to a bad strong person. For example, this consideration leads Japanese to feel that a meek daughter is better than just a daughter, and a meek gangster is more contemptible than just a gangster.

Smith's work in China is still in progress, so it is too early to review how Chinese impression formation processes differ from processes in other cultures, or how China's major regional cultures differ from one another. However, Smith's preliminary analyses suggest that Chinese impression formation processes may be different than both Japanese and American processes.

Implications

Returning to the principle of affective rationality, we can ask how cross-cultural congruence and diversity impact on the growth of world order.

Schneider and Heise (1995) presented an example showing that directives from an international corporation's American headquarters could cause emotional havoc in Germany, if prescribed behaviors befitting American sentiments cause disconfirmation of German sentiments. Here is a personal example, from academia rather than commerce.

Some years ago while on a lecture tour in Japan I had a meeting with an eminent Japanese professor and his assistant to discuss some methodological issues. I took my normal American seminar orientation: eagerly confrontational, though in a side-by-side manner rather than eyeball-to-eyeball. This mildly friendly but very potent and active seminar role is wonderfully successful in collaborative intellectual work. Or it is in America, anyhow. In the Japanese setting I soon was passionately writing mathematical equations to disprove a claim by the Japanese professor about some statistical matter. Of course, I expected him to write his own equations proving me wrong, or at least pinpointing our disagreement. Instead he abruptly left the room, and his assistant stood looking at me as if I were a rare specimen of rotten fish! Some painful discourse with the assistant, and with the professor after he returned, revealed that the professor had isolated himself in his office to deal with the humiliation and anger I had caused him. I was behaving far too active for Japanese academic roles, I was confronting the professor directly with no respect for his potent office, and I was obnoxiously spewing emotion. My American sentiments and interpretive processes had produced an affective rationality that subverted our work rather than furthering it.

Cultural diversity in sentiments and interpretive processes can lead to conflicting views about what is rational, as well as to menacing emotions in international encounters. We must deal with this problem as we try to build a global village that is rational and orderly and that simultaneously respects diversity. This could be a matter of life or death for some people, such as those caught in ethnic wars.

A desperate need does exist, and for that reason if no other, I believe that affect control theory will become a crucial component of internationalism in the future. It can offer understanding of diversity through an appropriate theory. It can offer factual knowledge about foreign views via cross-cultural data sets. And it can help in escaping quagmires of misunderstanding through computer simulations that show how events are understood in cultures other than one's own.

References

Averett, C. P. and D. R. Heise 1987 "Modified social identities: Amalgamations, Attributions, and Emotions." Journal of Mathematical Sociology, 13: 103-132.

Burke, Peter. 1991 "Identity Processes and Social Stress." American Sociological Review, 56: 836-49.

Goffman, Erving 1959 The Presentation of Self in Everyday Life. New York: Doubleday.

Heise, D. R. 1979 Understanding Events: Affect and the Construction of Social Action. New York: Cambridge University Press.

_____ 1999 "Controlling affective experience interpersonally." Social Psychology Quarterly, 62: 4-16.

_____ and Lisa Thomas 1989 "Predicting impressions created by combinations of emotion and social identity." Social Psychology Quarterly, 52: 141-148.

_____ and Brian Weir 1999 "A Test of Symbolic Interactionist Predictions About Emotions in Imagined Situations." Symbolic Interaction, 22 (1999): 129-161.

MacKinnon, Neil J. 1994 Symbolic Interactionism as Affect Control. Albany NY: SUNY Press.

Miller, George A., E. Galanter, and K. H. Pribram. 1960. Plans and the Structure of Behavior. New York: Holt, Rinehart, and Winston.

Osgood, C. H., W. H. May, and M. S. Miron 1975 Cross-Cultural Universals of Affective Meaning. Urbana: University of Illinois Press.

Schneider, Andreas, and D. R. Heise. 1995. Simulating symbolic interaction. Journal of Mathematical Sociology, 20: 271-287.

Smith, Herman W., Takanori Matsuno, and Michio Umino 1994 "How similar are impression-formation processes among Japanese and Americans?" Social Psychology Quarterly. 57: 124-139.

_____ , Takanori Matsuno, and Shuuichirou Ike. Forthcoming. "The Affective Basis of Attributional Processes among Japanese and Americans." Social Psychology Quarterly.

Smith-Lovin, Lynn. 1987a "Impressions from events." Journal of Mathematical Sociology, 13: 35-70.

-- and D. R. Heise 1988 Analyzing Social Interaction: Advances in Affect Control Theory. New York: Gordon and Breach.