Getting Angry The Jamesian Theory of Emotion in Anthropology
Robert C. Solomon
The Tahitians say that an angry man is like a botlle. When he gets filled up he will begin to spill over. (Havana, quoted in Levy 1973: 285)
The metaphor is so pervasive, it so dominates our thinking about our feelings, that we find ourselves unable to experience our emotions without it. We find it in philosophy and medicine as well as in our poetry, and we find it too in other cultures. Consequently, we believe what the metaphor tells us instead of recognizing it as a metaphor, a cultural artifact that systematically misleads us in our understanding of ourselves and, in anthropology, our understanding of other peoples.
The metaphor, captured succinctly in the Tahitian simile that an angry man is like a bottle, is the hydraulic metaphor. It presents the image of emotion as a force within us, filling up and spilling over. Rendered as science, the same metaphor is made respectable in physiological garb. The medieval physicians theorized at length on the various "humours" that determined the emotions. And in this century, the metaphor has been elegantly dressed in neurology and presented as a scientific theory—indeed, the only theory that has thoroughly dominated the subject over the past century. The theory is that an emotion is an "inner experience," a "feeling" based on a physiological disturbance of a (now) easily specifiable kind plus, perhaps, some outward manifestation and an interpretation according to which we identify this feeling as an emotion of a particular kind.
The theory received its classic formulation by William James (1884), in "What Is an Emotion?" James answered his question with his theory: An emotion is the perception of a visceral disturbance brought about by a traumatic perception, for example, seeing a bear leap out in front of you or coming across a bucket filled with blood. The theory (developed simultaneously by C. G. Lange in Europe) is now appropriately called the "Jamesian (James-Lange) theory of emotion." It is, I shall argue, as misleading as it is pervasive.
- 197 -
198 • The Emotions Emotions in Anthropology
Emotions as biological events are ihe same the world over. (Lindzey 1954; also see 1961)
The Jamesian theory has special appeal, and is particularly damaging, in anthropology. There is an obvious problem, given the nature of tile theory. An emotion as an "inner feeling" is unobservable and inaccessible to the anthropologist, thus leaving any attempt at describing emotion in other peoples at the mercy of obviously an-fhropocentric "empathy." And yet, the theory (scientific or not) has been accepted as apparently useful for interpreting not only the emotional life of other peoples but also the language used by other peoples to describe their own emotions, thus suggesting a kind of double confirmation. The theory—that emotions, as feelings based upon physiological disturbances, can be understood in strictly biological terms—results in this familiar but fallacious consequence: Emotions can therefore be taken to be more or less universal human phenomena, the same in everyone, making allowances for certain minimal differences in physiology and, consequently, temperament. (In fact, I would argue that there is little reason to suppose that such differences or their emotional consequences are minimal, but that is not the thesis I wish to pursue here: see, e.g., Freedman 1974.)
Even if the emotions were essentially the same in all people, however, it is evident that the language and interpretation of emotions, as well as their causes, expressions, and vicissitudes, vary widely from culture to culture. The effects of epinephrine may be identical in angry people from Borough Hall in Brooklyn to the beaches of Bora Bora, but there are, nevertheless, differences in the emotional lives of various peoples, and this is where anthropology enters the picture.
The anthropological appear of the Jamesian theory is obvious: It divides the phenomenon of emotion into two comprehensive components, a physiological feeling component, which can be presumed a priori (and falsely) to be more or less the same in all human beings, and a cultural component, which can be described by the anthropologist, using the same techniques of observation and interview that are appropriate for almost any other cultural phenomenon. Any mystery surrounding emotion is thus dispelled: the difficulty of "getting inside another person's head," without which one cannot understand another's feelings, is rendered unnecessary. Emotions are to be understood in the realm of physiology, not phenomenology, thus circumventing the hard problem of "empathy." The interpretation of emotions {including the basic interpretive act of naming and identifying one's emotion) is quite distinct from the emotion itself, thus leaving the emotion proper outside the realm of anthropology.
My argument hums on two related objections to the Jamesian theory. First, that the theory is not only incomplete but wholly mistaken. It trivializes, rather than captures, the nature of emotions.1 Second, the distinction between an emotion and
Getting Angry • 199
its interpretation is faulty and misleading in a variety of ways. The consequence of these objections is to insist that emotions themselves are the proper province of anthropology. My thesis is that emotions are to be construed as cultural acquisitions, determined by the circumstances and concepts of a particular culture as well as, or rather much more than, by the functions of biology and, more specifically, neurology. There may be universal emotions, but this is a matter to be settled empirically, not by a priori pronouncement.
The Variability of Emotions
Take aggression as an example. A distinction must be made between the instrumental acts that are indices of aggression (e.g., hitting, insulting, noncooperating) and the hypothetical "goal response" of the aggression motive (perceiving another person's reactions to injury). It is the latter (bat one would expect to find fransculturaliy. The aggressor's instrumental activities that serve to hurt someone else—and thus enable him to perceive reactions to injury in his victim—vvill differ from one culture to another. The form of an insult, for instance, depends on the values held by the insulted one. Or to take another example: automobile racing and football can be instrumental activities for competition only if the society has automobiles and knows how to play football. (Sears 1955)
The cultural specificity and variability of several dimensions of emotion are not in question. For instance, the various causes of emotion are clearly cultural in their specifics (whether or not there are also some causes of some emotions that might be argued to be universal or even "instinctual"). What makes a person angry depends upon those situations or events that are considered offensive or frustrating. A New Yorker will become infuriated on standing in a queue the length of which would make a Muscovite grateful. The same action will inspire outrage in some societies and not others; consider, as examples, failing to shake hands, kissing on the lips, killing a dog, not returning a phone call. The same objects will provoke fear in one culture but not in others, for example, snakes, bewitchment, being audited by the IRS, not getting tenure, and being too rich or too thin. Causes of emotion vary from culture to culture; it does not follow that emotions do, or do not, vary as well.
The names of emotions clearly vary from culture to culture, along with most vocabulary entries and names for virtually everything else. But this obvious point hides a subtle and troublesome one; how do we know whether it is only the names (i.e., phonetic sequences) that vary, rather than their reference? The problem here is whatW. V. 0. Quinc calls "radical intranslalability"; do the words "anger" in English and "riri" in Tahitian refer to "the same" emotion? How would we tell? Even if the causes are commensurable and the behavior seems to be similar, how do we gauge the similarity of the emotions? Names of emolions are clearly cultural artifacts, even "arbitrary" in the sense that it is now said as a matter of Paris-inspired cant that "all signs are arbitrary." But the identities of the phenomena that those names name
200 • The Emotions
are yet an open question, not obviously the same references for quaintly different vocabularies but clearly not entirely different either. We are, after all, identifying a shared reference to something.
A similar point can be made about the various expressions of emotion. Clearly some expressions, at least, differ from culture to culture as learned gestures and more or less "spontaneous" actions. Clenched fists are expressions of anger in one culture, not in another. Banging one's head Dn the wall is an expression of grief in one society, not in others. And the verbal expressions of emotion vary not only along with the language (of course) but also according to the familiar images and metaphors of the culture. (Not everyone would understand what we so easily and now clumsily refer to as "heartbroken.") There may well be emotional expressions that vary very little from culture to culhire, particularly certain minimal facial expressions, as Paul Ekman (1975) has recently demonstrated. But that there are such universal expressions, if there arc any, no more demonstrates the universality or "nature" of emotions than the wide variety of more complex expressions proves the variability of emotions. Again, this must at least start as an open question, for which the observation of emotional expressions may serve at most as a preliminary. Indeed, the more fundamental question—of what are these expressions expressive?—will have to wait for an account of the emotions themselves.
Finally, there is the series of metaphors to be found in almost evety culture with any vocabulary of psychological self-description that are essentially explanations and diagnoses of emotions, rather than merely names for them. The Tahitian gentleman quoted at the start of this essay, for example, is expressing a theory, the hydraulic theory, which has long been dominant in discussions of emotion in our culture, too, in part because of (but also culminating in) Freud's "dynamic" and "economic" models of the psyche in terms of various "forces" within. Metaphors and theories of emotion are often related and even interchangeable. They also influence the experience of the emotions themselves. To believe that anger is a force building up pressure is to experience the physiological symptoms of anger as a force "inside," just as believing that "falling in love" is bound to have a certain irresponsible influence on one's loving.
!t is a matter of no small interest that the same metaphors—the hydraulic metaphor in particular—can be found in societies of very different temperaments. But such metaphors are by no means universal. Catherine Lutz (1982) describes an emotional vocabulary among the Ifaluk that is relatively devoid of references to the hydraulic metaphor or the Jamesian theory,2 and the prevalence of the metaphor by no means proves the Jamesian theory to be true. Nevertheless, the variability of emotion metaphors and theories can be counted among the various dimensions of variability of emotion, if, that is, it is true that beliefs about emotions influence or determine the nature of the emotions themselves. (On the Jamesian theory, it is hard to see how or why this should be so; on the alternative view I shall propose at the end of this essay, the mutual influence of beliefs and emotions should be quite transparent.)
Getting Angry • 201
Names of emotions do not yet entail metaphors or theories, but even so rudimentary a psychological activity as "naming one's feelings" already stakes out a network of distinctions and foci that are well on their way to extended metaphors and crude theories. The fact that one language has a dozen words for sexual affection and another has fifty words for hostility already anticipates the kinds of models that will be appropriate. A culture that emphasizes what David Hume called "the violent passions" will be ripe for the Jamesian theory, bat a culture that rather stresses the "calm" emotions (an appreciation of beauty, lifelong friendship, a sense of beneficence and justice) will find the Jamesian theory and the hydraulic model that underlies it patently absurd. A culture that bothers to name an emotion pays at least some attention to it, and it is hard to find a culture with named emotions that does not also have theories about them, however primitive. In some cases, the theory might consist simply of the warning "anger is dangerous." In theory-enthusiastic cultures such as our own, the theories surrounding an emotion might more resemble the theology of the druids, thus prompting more or less perennial cries about emotional: simplicity and "getting in touch with your feelings." But whether the theory at stake is the labyrinth of Jungian typologies or the homilies of Joyce Brothers, the beliefs people have about emotions vary considerably, and it remains to be seen just how this reflects—or doesn't reflect—the crosscultural (and intracultural) variability of the emotions themselves.
(Not) Getting Angry: Two Examples
"My intestines were angry." (quoted in Levy, 1973: 214)
Anger is an emotion that would seem to be universal and unlearned if any emotion is, however different its manifestations in various cultures. John Watson chose anger as one of his three "basic" emotions (fear and dependency were the other two). It is one of those emotions most evident even in infants, and Watson suggested that it is one of the building blocks for all other emotions. More recently, Robert Plutchik (1962) has developed an evolutionary model of emotions and emotional development in which anger, again, emerges as one of the (this time eight) basic building blocks of emotion. Anger is one of the most easily observable emotions; we might debate its nuances (outrage or indignation) and perhaps surmise its etiology (jealousy, frustration, or moral offense). The causes of anger might differ from culture to culture, and the expressions, at least the verbal expressions, might vary too. Bur it is too easy to assume that anger itself and its basic manifestations—the reddened face, visible irritability and what William James properly called "the tendency to vigorous action"—are much the same from the Philippines to the Lower East Side, from Bongo Bongo to the more boisterous committee meetings of the Social Science Research Council. Everyone gets angry- -at least at some time and for some reason. Or so it would seem.
202 • The Emotions
Getting Angry • 203
But lei us consider two quite different accounts of anger, in two quite different societies. I want to discuss later in this chapter some of the methodological problems to which any such account is subject. But, as afirst, superficial observation, letus make clear certain grass differences—or at least claims about certain gross differences—in two cultures: the Tahitian (Levy 1973) and the Utka (Utkuhikhalingmiut) Eskimos (Briggs 1970) in the Northwest Territories. In both cases, these people do not get angry. Some of this may be emphasis rather than substance, but that too constitutes a significant difference in emotional life. It might be argued, for instance, that Americans give far more importance to the emotions of anger and moral indignation than do the Russians or Japanese, for example, whether or not the emotions themselves are so significantly different. But having pointed out this difference in emphasis, have we not already indicated vast differences in temperament and emotional constitution as such? For both the Tahitians and the Utka, however, anger is as rare as it is feared.
The Tahitians, according to Levy, place an unusual amount of emphasis on anger. They talk about it and theorize about it extensively; it is "hypercogrrixed," he tells us, in that "there are a large number of culturally provided schemata for interpreting and dealing with anger." (See Levy 1984.) Other emotions, sadness, for instance, are "hypocognized" and, Levy suggests, virtually unrecognized. Anger, however, is rare, no matter how much the object of concern. Does this mean, however, that it is indeed present but unaccounted for or, rather, that in circumstances in which we (for example) would most certainly have an emotion, they do not?
A partial answer to this crucial question can be couched in terms of the Tahitian theory of emotion, which is distinctively Jamesian. Emotions have a "place" in the body, the intestines, for example. Indeed, the language of emotion is often "it" rather than "I," although one must quickly add that this grammatical feature of the Tahitian language is not to be found only in the realm of emotions (Levy 1973:213). He quotes an informant;
"In my youth, [it was] a powerful thing, very powerful, very powerful 'it' was [sic], when 'it' came, and I tried to hold it down there was something that was not right. That was the cause of a lot of bad anger inside one... after a time ... that thing, 'it' would go away." (ibid., p. 212)
Levy adds that "people will say 'my intestines were angry'" (ibid., p. 213). This locution may seem slightly odd but certainly not unfamiliar; it indicates, however, a much deeper difference between our conception of anger and the Tahitian conception and, consequently and more important, a deep difference between Tahitian anger and our own.
Throughout the literature on Tahiti, Levy tells us, one message above all keeps repeating itself: "These are gentle people" and there are "extremely few reports
of angry behavior." Morrison noted two centuries ago that the Tahitians are "slow to anger and soon appeased." (in Levy 1973:275). Levy quotes a contemporary policeman who talks of "the lack of a vengeful spirit" (ibid., p. 276), and though Levy reports some forty-seven terms referring to anger, he adds that the Tahitian concern with, and fear of, anger and its violent effects are "in the face of little experience of such anger" (ibid., p. 285). The pairing of so much attention and theorizing with so rare an emotion points to a curious relationship between the having of an emotion and the understanding of it, but it is clear from Levy's descriptions and reports that this relationship is not to be construed (as we might be likely to construe it in ourselves) as one of "suppression" or social "control" as such. It is the gentleness, the lack of anger itself that seems to be learned, not the inhibition or suppression of it. And part of this learning experience, ironically, is the acquisition of an enormous number of myths and metaphors about anger through which this rather rare emotion is explained—and feared.
In Jean Briggs's (1970) descriptions of the Utka Eskimos, they do not, as her title Never in Anger indicates, get angry. Not only do they not express anger: they do not "feel" angry, and, unlike the Tahitians, they do not talk about it. They do not get angry in circumstances that would surely incite us to outrage, and they do nor get angry in other circumstances either. The Utka do not have a word or set of graded distinctions for anger, as we do and as the Tahitians do; indeed the word widi which they refer to angry behavior in foreigners and in children is also the word for "childish." There is no reason to suppose diat, biologically, the Utka have any fewer or more impoverished epinephrine secretions than we do, and Dr. Briggs's descriptions show that, on occasion, they get just as "heated up" as we do. But they do not get angry, she assures us. They do feel annoyed, even hostile, and they can display raw violence, for example, the beating of their dogs (in the name of "discipline," of course). But is this to be considered merely a nuance of terminology? Or something more significant?
There have been some severe objections to the observations and conclusions of this research, but the central claim remains intact, at least by way of a plausible hypothesis not yet refuted. Michelle Rosaldo (1984), for instance, has argued thai Briggs confuses Jack of anger with fear of anger, the sense—to be found in Tahitian society as well as in Filipino society and in our own—that anger is dangerous and can even destroy a society. But here again, we meet that suspicious and too-neat distinction between the essence of the emotion itself and talk about emotion, as if it can be assumed that the emotion remains more or less constant while our thoughts and feelings about the emotion alter its expression and its representation. But even if Briggs is wrong about the absence of anger as such, the context of that emotion and the peculiar absence of (what we would consider) the usual expressions and manifestations of it would have to be explained.
■
^""■■^ p^^g pw^
References
Bnggs, Jean L. 1970. Never in Anger. Cambndge, Mass.: Harvard University Press.
Ekman, Paul. 1975. Unmasking the Face. JSnglewood Cliffs, N.J.: Prentice-Hall.
Freedman, D. G. 1974. Human Infancy. Hillsdale, N.J.: Erlbaum.
James. William. 1884. What is an emotion? Mind 9: 188-205.
Levy, Robert I. 1973. Tahitians. University of Chicago Press.
Levy, Robert I. 1984. Emotion, Knowledge and Culture. In Richard A. SchwedeT
and Robert A. LeVine, eds., Culture Theory: Essays on Mind, Self and Emotion.
Cambridge: Cambridge University Press. Lindzey, Gardner. 1954. Psychology. Cleveland: Worth.
--! 961. Projective Techniques and Cross-Cultural Research. New York: Appleton-
Century-Crofts.
Lutz, Catherine. 1982. The domain of emotion words in Ifaluk. American Ethnologist 9: 113-28.
Plutchik, Robert. 1962. The Emotions. New York: Random House.
Rosaldo, Michelle Z. 1984. Toward an Anthropology of Self and Feeling. In Richard
A. Schweder and Robert A. LeVine, eds., Culture Theoty: Essays on Mind, Self
and Emotion. Cambridge: Cambridge University Press. Sears. Robert R. 1955. Transcultural variables and conceptual equivalence. In Bert
Kaplan, ed., Studying Personality Cross-Culturally. Evanston, 111,: Row, Peterson. Solomon, Robert C. 1976. Passions. New York: Doubleday. 1978. Emotions and anthropology. Inquiry 21:181-99.
Notes
This chapter was stimulated by the Social Science Research Council meeting in May 1981 on Concepts of Culture and Its Acquisition. Special thanks to Richard A. Shweder for his helpful criticism and encouragement. Aporlion of the "methodology" section has been adapted from my "Emotions and Anthropology" (Inquiry 21 [1978]: 181-99), with the generous permission of tile editors.
1. This has often been argued, and I shall not repeat the primary arguments here: see Solomon 1976, chap. 7: 197 8.
2. It is worth noting, however, that the criterion used for distinguishing emotion words in Ifaluk was whether or not they were identified as "about our insides," despite the argument that "the Ifaluk see the emotions as evoked in, arid inseparable from, social activity" rather than "internal feeling states" (Lutz 1982: 114, 124).
-15-
Thc Shame of Headhunters and the Autonomy of Self
Michelle 2. Rosaldo
My point is simple. Psychological idioms that we use in offering accounts of the activities of our peers—or our companions in the field1—are at tire same time "ideological" or "moral" notions. As ethnographers (and moral persons) we are compelled at once to use and to suspect them. The assumptions bound to our familiar forms of psychological explanation prove attractive yet inevitably problematic when confronted with a cultural account.
"Guilt" and "shame"2 are, of course, the idioms to concern me. Paired by Western theorists5 as complementary and/or alternative means for controlling selfish energies that we think belong to every human heart, these terms assume our faith that people everywhere are frustrated, repressed, rebellious, unfulfilled, or—at least—at odds with their society. Guilt and shame are seen, in short, as moral affects4 necessary to constrain the individuated self from dangerous and asocial acts of impulse, lust, and violence. Surely, this is a vision most of us find as suspect as it seems difficult to reject. Guilt and shame may everywhere be linked to firings like violence, sex, and strain, just as, in every case, they may concern the threat of circumstance or activity to undermine an ideal presentation of the self. And yet the "selves" that these, or other feelings, help defend—and so, the ways such feelings work—will differ with the culture and organization of particular societies. "Otir" view of persons as embodiments of continuing and conflicrual inner drives and needs is one which, in all likelihood, reflects important aspects of the "individualism" famous in the modern West, along with the experiences of Western "individual" suppressed by modem forms of social inequality. Considering data from the ilongots,5 a horticultural and hunting population of Northern Luzon, Philippines, my questions here concern alternatives that emerge among a people who assume that persons want to be not different but equivalent or "the same," and sec in individuality not essential self, but a persona born of conflict.
To begin. Most adult Ilongot men at the time of my research were, or had been, headhunters. That is, most had at one time joined with fellows on a raid in which they
- 205 -