Catharsis and Other Heresies: A Theory of Emotion
Catharsis and Other Heresies: A Theory of Emotion
Thomas J. Scheff (9,127)
Abstract: This article first discusses the development of concepts and reviews the literature on catharsis. A new theory is proposed integrating social, psychological, and neurological components. It builds upon a socio-biological model of emotions by John Dewey and G. H. Mead. They proposed that the initiation and climax of bodily emotion process are instinctive, and that conscious feeling occurs to the extent that climax is delayed. This article further develops Shibutani’s application of the Dewey-Mead model. The climax idea, together with the concept of an optimal distance for experiencing emotions, seems to explain otherwise puzzling phenomena. It also points toward the conditions under which certain types of laughing and crying become cathartic. The new theory shows the relationship between distancing, role-taking, pendulation in somatic therapy, and recent studies of mirror neurons. One implication of the new theory is that in modern societies emotion process may be frequently initiated but seldom finds resolution. Collective repression of emotions may cause larges-scale violence.
After seeing his obituary in newspapers, Mark Twain responded: “The reports of my death are greatly exaggerated. “ Similarly, announcements of the death of catharsis seem to be premature. In the literature on emotion, except for clinical studies, it is usually dismissed. After a large number of articles in the period 1960-79, very few studies have been appeared under the rubric of catharsis. Recent clinical studies continue to support the catharsis hypothesis. It is possible, therefore, that the dismissal of catharsis in social science was unwarranted.
Before discussing the heresy of catharsis, however, it will be necessary to discuss another one: the best way to begin a study is not always systematic and empirical. I have already been labeled a heretic in my own discipline, because my work is both psychological and sociological. I offer a different heresy for academic psychology: if we are to understand emotions and relationships, it will be first necessary to move away from vernacular words to single-meaning concepts.
At this time scholarship and research in social science is much more dependent on vernacular words than on clearly defined concepts. Most research is entrapped within a cultural web of words. Most theories, methods and data sets are formed by the unnoticed web of words in which they are encased. To arrive at an understanding of human experience and behavior, it might be necessary to first develop clearly defined concepts.
Although the vernacular language of any society can convey valuable insights into human conduct, it is also a repository for preconceptions and ambiguities. For example, ordinary usage in the English language, tied as it is to the first urban/industrial society, emphasizes individuals much more than relationships, and cognition much more than emotion. Traditional societies, to the extent that they are small, localized, and static, go the other way: relationships are more important than individuals, and emotion more important than cognition. This section seeks to explain how the assumptions that go without saying in the language of modern societies becloud social science.
Perhaps the first task facing social science is conceptual, and therefore, pre-scientific. The scientific approach is organized in terms of theory, method and data. But these tools are virtually useless until basic concepts have been clearly defined, because scientists, like everyone else, live in the assumptive world of their own culture. The ocean of assumptions that each culture reflects and generates is virtually invisible to its members. Unless these assumptions are avoided, the tools of science lead nowhere.
Metaphors vs. Concepts
The way that cultural assumptions impede science has formulated by the philosopher Quine:
The neatly worked inner stretches of science are an open space in the tropical jungle, created by clearing tropes [ruling metaphors] away” (1979, p. 160).
Before scientific procedures are applicable, a ruling trope may have to be overthrown. Quine’s formulation captures the radically intuitive element necessary for scientific advance. Tropes are linguistic/mental routines that both reflect and hide cultural assumptions.
The history of physical science reveals many examples of obstructive tropes. Tycho Brahe, the Danish astronomer, spent his adult life trying to determine the orbit of Venus. He made extraordinarily accurate observations of the position of the planet during his lifetime, but he assumed, like everyone else, that the planets revolved around the earth. For this reason he was one frustrated scientist.
Kepler, Brahe’s assistant, inherited the data after Brahe died. For years he made no progress. In his exasperation, Kepler developed a bizarre model in which orbits were determined by transparent solid polyhedrons. The model itself was mere fantasy, but in his play he had unthinkingly placed the sun, rather than the earth, at the center. Although Kepler’s scientific skills were far inferior to Brahe’s, he quickly solved the problem (Koestler 1967).
Similarly, Einstein began the theory of relativity with a joke concerning persons passing each other on trains, trying to determine their speed relative to each other. He realized intuitively that this situation challenged the ruling trope: time and motion are absolute. Although he had a doctorate in physics, Einstein knew little mathematics. He had to get help to put his anti-trope into mathematical form.
Quine’s formulation captures the primitive, intuitive element necessary before scientific methods can be useful. Scientific method, no matter how scrupulously applied, is helpless in the face of ambiguous or misleading tropes. Since social science often is based on tropes, rather than precise definitions, we have a long way to go.
Out of many possible examples, just one will be mentioned: the vast domain of self-esteem studies. Fearon and I (2004) found that 11 earlier reviews of self-esteem scales research agreed: Almost 20 k studies have reported findings that are either trivially small, contradictory, or null. Fearon and I propose that the most important error is conceptual. The scales assume that self-esteem is cognitive: self-evaluation. But they all include items that generate self-feeling responses, i. e., emotions. Most dictionaries include both definitions, even though they are orthogonal. There are ways of reducing a scale to a single dimension, but the overwhelming mass of studies has not gone this route.
The failure of self-esteem studies to produce results should be speaking to us. It says that being systematic and empirical does not guarantee results. No matter how “scientific” and no matter how large the data set, if the variable being studied conflates two or more dimensions, the findings will be nil. All empirical studies require the development of single dimensional concepts.
Resurrecting Catharsis from the Dead
What follows is a brief review of the treatment of catharsis in past and current literature. It was originally defined by Aristotle (Third Century B.C.) as the purgation of pity and terror in theatre audiences, and since then has figured largely in theories of drama in the humanities (Belifiore 1992). The word terror obviously refers to the emotion of fear, but the word pity is not an adequate translation of Aristotle’s usage. He was referring not to a specific emotion in the modern sense, but to empathic identification with one or more of the characters in the drama. It is important to note that the whole definition concerns only a very narrow domain, emotional reactions of members of an audience to drama. A large part of the problem with social science research on catharsis involves an unwarranted broadening of the definition of catharsis, as will be discussed below.
One part of Aristotle’s treatment of catharsis has been especially puzzling to modern readers: his insistence that catharsis in drama was not only a source of entertainment and pleasure, but also a necessity for the survival of a society in the long run. This article will seek to make his proposal understandable.
The dismissal of the idea of catharsis in social/behavioral science has been largely a result of Freud’s influence in psychoanalysis, on one hand, and many studies of what has been called “aggression catharsis” in experimental social psychology, on the other. In both cases, the difficulty arises out of confusion about concepts.
The great majority of psychoanalysts have uncritically accepted Freud’s dismissal of catharsis, even though it was brief and casual, compared to his careful documentation of it’s effectiveness in the case studies that make up Studies on Hysteria (Freud and Breuer 1895). Neither Freud nor any other psychotherapist provided evidence showing that catharsis was ineffective. Nor did they note that the talk-heavy version of psychoanalysis that Freud created after 1905 was lengthy, costly, and of uncertain outcome. Like Freud and Breuer’s early study, more recent clinical reports suggest that catharsis may be the main cause of success in psychotherapy (Symonds 1954; Kosmicki and Glickauf-Hughes 1997; Bemak and Young 1998).
In recent years, the effectiveness of a cathartic technique for pathological bereavement (unresolved grief), called “re-grief therapy,” also supports the validity of catharsis (Volkan, 1975). Volkan’s later publications on grief therapy, together with the earlier work of Parkes, imply that crying is necessary in order to complete mourning. (Volkan and Zintl 1993; Parkes 1988). However, Volkan, Parks, and other experts on mourning are somewhat cagey about crying. They don’t state a thesis, but instead offer clinical examples of cases of successful mourning in which crying is obviously an important part of the process.
In experimental social psychology, there are two branches to the study of the catharsis rubric: “aggression catharsis,” on the one hand, and vicarious aggression, on the other. Berkowitz (1962), Feshback and Singer (1971), and others have produced a large body of systematic research about the effects of aggressive behavior on subsequent levels of hostility. In his review of studies of aggression catharsis, Quanty (1976) concluded that the weight of evidence suggests that aggression doesn’t help. As will be discussed below, however, equating aggressive behavior with catharsis is an invalid use of the term.
Among three reviews of vicarious aggression, the other branch, there are disparate conclusions. Liebert and Schwartzberg (1977) stated that the weight of evidence suggests that the viewing of vicarious violence leads to increased violence, but Kaplan and Singer (1976) concluded that the results are so faulty or ambiguous that no conclusions can be drawn. The third reviewer, Geen (1976), is more cautious in drawing definite conclusions, but agrees more with Liebert and Schwartzberg than with Kaplan and Singer, suggesting that results of existing studies support social learning theory more than the catharsis hypothesis. The problem with these studies and the studies of aggression is certainly not the lack of systematic data.
Why do clinical and experimental findings differ so completely? Because they are not referring to the same thing. Berkowitz and others have tested the hypothesis that catharsis occurs through aggressive behavior. In a typical study, he has shown that active retaliation against an aggressor not only does not lower the level of hostility of the person who is retaliating but may actually raise it. This is an important finding, because most people fervently belief that “venting” is a good way to resolve anger. The findings discussed above have made no difference with the public, as far as I can tell, but they should have.
However, these findings are completely unrelated to any legitimate theory of catharsis. Neither Aristotle’s doctrine nor Freud and Breuer’s technique defined catharsis in terms of behavior. Both dramatic and psychotherapeutic approaches involve the re-experiencing of past emotional crises in a context of complete security: in the safety of the theatre or the therapist’s office. Catharsis in these contexts is analogous to Wordsworth’s definition of poetry: emotion recollected in tranquility. The extension of catharsis to include aggressive retaliation is utterly without foundation.
Studies of vicarious aggression appear more relevant to the theory of catharsis, especially studies of the effects of viewing television violence, since the contexts studied are similar to dramatic or psychotherapeutic settings envisioned by theorists of catharsis. Even in these studies, however, there are two serious conceptual problems. First, the treatment variable in studies of the effects of television is the viewing of dramatic violence. However, it is probably a mistake to equate the process of catharsis with the stimuli that may cause it to occur. Presumably catharsis involves a particular type of emotional response within the individual. The viewing of dramatic violence may or may not give rise to this response, depending on the characteristics of the stimuli, the viewers, and other conditions.
Feshback and Singer were able to show that high school students who viewed violent T V fare committed fewer aggressive acts than those students who didn’t view it, a phenomenon presumably demonstrating the effects of catharsis. The correlation they found, however, was weak. According to our theory, if the researchers had had some way of separating the viewers of violent fare into those who experienced a cathartic emotional response, such as laughter, from those who didn’t, the data would have shown still stronger support for the theory of catharsis.
Among the viewers of violent fare who experienced catharsis, the drop in aggressive behavior would have been greater than the amount reported in the study. Among the viewers of violent fare who did not experience a cathartic response, there would have been no reduction of aggressive behavior at all, thus isolating the causative process: the occurrence of a cathartic response in the context of viewing violent TV fare. The theory of catharsis would be more amenable to testing if catharsis were defined independently of the arousing stimuli.
The second issue concerns distancing, to be discussed at length below. Most studies of the effects of viewing violence rate only the quantity of violence, not the quality. For example, some studies include comic and cartoon violence, others arbitrarily exclude it. I would argue that most comic violence is properly distanced, and should therefore lead to catharsis, and that more realistic violence in drama is under-distanced, and would not have cathartic effects. Until some control over distancing of the stimulus materials is introduced, it will be difficult to interpret the relation of most studies of vicarious aggression to the catharsis hypothesis.
In addition to studies under the rubric of catharsis, more recently studies of the expression of emotions, particularly laughing and crying, have appeared. Several studies of adult crying are reported in Vingerhoet et al (1997). Billig (2005) has written about laughter, and there are also many medical studies of laughter and/or crying.
What unites all these studies is that they make no attempt to name and describe types of laughter and crying. Lumping all types together, they report no benefit. But laughter and crying are complex couplings of physical, cognitive and emotional systems. Faked laughter, for example, would likely have little or no effect on the body, since it is more like a speech act than a complex emotional reaction. Laypersons who cry easily differentiate between what they call “a good cry” and a bad one.
The medical literature seems particularly distorted in this respect. In virtually all studies, emotions are the enemies. This orientation is understandable with respect to rage, but laughing and crying also are almost always treated as pathological. There are many studies of a new pathology called emotional lability (EL), or even a more extreme label, emotional incontinence. “Shame on you.”
It seems not to have occurred to any of the authors that the absence of emotional expression might be a far wider malady, and also a much more damaging one. I have found only one article that touches even indirectly on this issue. Scoppetta et al (2005) showed that the SSRI class of anti-depressants suppresses crying even in normal persons. They admit a doubt about the wisdom of the widespread use of these drugs:
SSRI are among the most used drugs in the world, every day they are consumed by millions of people including politicians, businessmen, soldiers, army commanders, policemen and criminals. The idea is… worrying that the control of the emotions and behavior of these millions of people can be quickly modified by one SSRI for a few days….
The management of grief provides one example of over-, rather than undercontrol of emotions. It may be that the inability to mourn/unresolved grief (Mitscherlich 1975; Parkes 1988) particularly among men, is a social institution in modern societies. To the extent that the theory outlined here is true, then the widespread use of drugs that further inhibit crying and other forms of emotional expression would be extremely damaging rather than helpful.
This issue, when taken together with others discussed in this section, suggests that the discrediting of catharsis may have been a mistake. The section that follows will outline an approach to a version of catharsis updated from my earlier (1979) book.
A New Theory of Catharsis
The theory proposed here takes a step toward integrating biological, social, and psychological components. The first half of the theory concerns the biological bases of emotion, reviewing and extending work by John Dewey, G. H. Mead, and Tamotsu Shibutani. The second half explores some social and psychological dimensions of emotion. It describes the distancing of emotion, growing out of theories of drama. Another of Mead’s ideas, role-taking, seems to be relevant to distancing. A current clinical idea, the somatic approach to emotions, makes much the same point using different terms, such as “pendulation. “ Finally, the idea of mirror neurons, based on recent work in neuropsychology, points toward a neurological analogue to the role-taking/distancing idea.
Combining social psychology and biology in this way points toward a new theory of catharsis. I will seek to re-instate this idea, long out of favor. The emphasis throughout will be on the importance of developing concepts rather than using vernacular words, and generating hypotheses rather than relying on “common sense.”
Biology and Emotion
This paper outlines socio-biological models for four emotions: grief, fear, shame, and anger, based on an approach suggested by Dewey (1894) Mead (1934), and Shibutani (1961). The present paper modifies and extends Shibutani’s ideas and those presented in my own earlier book (1979) on emotions.
Dewey’s early statement (1894) proposed that emotions begin as bodily drives with universal, genetically determined beginning and end points. Shibutani adopted Mead’s parallel idea (1934) to Dewey’s formulation: acts of any kind, including emotions, have four stages: impulse, perception, manipulation, and consummation. Mead and Dewey furthermore proposed that feeling emotions depends on blockage or delay of completion of this sequence. Although the beginning and end of emotion acts are physiologically determined, the two middle stages, perception and manipulation, allow for cultural and personal elements.
Furthermore, the Dewey/Mead approach implies that emotions are merely instinctive signals of threat, the greater the threat, the more intense the emotional signal. One implication of their signaling approach is that normal emotions are extremely brief, a few seconds, at most. If that is the case, how do we account for long-lasting emotions, episodes of fear or anger that last minutes, hours or days? Mead and Dewey took no notice of this problem nor have subsequent studies. One reason that Mead and Dewey didn’t explore this problem and others was that their theory remained at a highly abstract level, they didn’t try to apply it to particular emotions.
Shibutani went beyond Dewey and Mead’s abstractions by applying them to three of the emotions mentioned above. In particular, his treatment of grief (sorrow) might serve as a model for developing concepts for all emotions. However, my application of this model to fear and anger is different than the one that he offered. I also add shame to the list, an emotion that Shibutani didn’t consider.
Shibutani began with the Dewey/Mead idea that delay or blockage gives rise to feelings. That is, feelings occur when a predetermined emotion sequence is triggered, but blocked before completion. First he applied this idea to “sorrow”:
…one feels sorrow when there is readiness to cry. [One} feels less sorry when [one] starts to cry, and crying that is violent enough often eliminates the feeling (1961, 72).
The biological basis of grief (sorrow) is bodily preparation to cry. One feels grief to the extent that the completion, crying, is delayed. This idea can form the basis for a general approach to emotion.
A few sentences later, Shibutani introduced a parallel with sexual arousal (p. 73): “Lust [that is, sexual desire] arises in situations where one cannot consummate sexual impulses [i. e, orgasm] at once.” Foreplay involves bodily arousal in preparation for sexual orgasm, but with the “consummation” (orgasm) delayed. In the realm of sexuality, foreplay connotes an intentional action, since it increases or at least prolongs pleasure. However, it appears that in modern societies there are many, many situations of unintentional delay of completion of emotion sequences, as will be discussed below. It may be that most people in modern societies are locked into patterns of what might be called emotion foreplay, rather than completion.
Although Shibutani, unlike Dewey and Mead, connected a specific emotion, grief, to a predetermined endpoint, crying, he didn’t name another part of the Dewey-Mead model: the beginning stimulus. To construct concepts for all four emotions, it will be necessary to name both the initial stimuli as well as the outcomes. In the case of grief/sadness/sorrow there is a large clinical literature suggesting that the initial stimulus that generates grief is the loss of a loved one. If this is the case, then grief is occasioned by loss, and completed by crying, that is, sobbing and tears.
In applying the idea of endpoints of emotional acts to anger and fear (72-73), however, Shibutani followed Cannon (1929), who named “fight” as the consummation of anger, and “flight” as the consummation of fear, That is, Shibutani, like Cannon, treated fight and flight as instinctual responses to anger and fear. These two ideas now seem outdated. It is generally recognized that outer behavior is not determined by instinctual drives. In this respect, humans differ from other mammals, whose behavior is largely instinctual. Although we still embody instinctive inner responses, the behavioral response that occurs is indeterminate.
This issue has been recently been treated specifically with respect to Cannon’s fight or flight. Taylor and her colleagues argue (2000) that especially for women, there is a culturally learned element in response to anger and fear. Taylor et al call this outer response “tend and befriend.” That is, among the many possible behavioral responses that humans may have to anger and fear, women, particularly, have evolved a pattern of taking care of themselves and others, and networking with others, rather than fight or flight.
Nevertheless, the idea that angry aggression is a predetermined instinctive response is still prevalent among the public at large, and among many researchers. Even the self-help psychotherapy movement that got me started into my own emotions erred in this respect. Re-evaluation Counseling doctrine, cunning about laughing and crying, had me hitting a pillow with a tennis racket for resolving anger. It never worked with me, and with only a tiny percentage of my students. Most either wouldn’t do it (“I’m not angry now”) or complained of feeling worse afterwards.
My experience with RC was richly rewarding to me, but after some 4 years I realized that the doctrine and practices were narrow and somewhat rigid. After considerable catching up on my backlog of unresolved emotions, it turned out that the next thing I needed was not only more catharsis, but to learn other skills, particularly social skills like intimate communication. Since these skills were not available in RC, I moved on.
Returning to the idea that aggression is cathartic, this idea is contradicted by clinicians. Angry aggression, which they call “acting out anger,” is only one of many possible responses to anger, and is usually an undesirable and unnecessary response. Their message, however, appears not to have been widely heard. This issue will be further discussed later in this essay.
If fight and flight are not predetermined responses to anger and fear, what are the predetermined inner endpoints? There is no consensus as the answers to these two questions in our society, neither among experts nor the public. Certainly Shibutani didn’t suggest any. I know of only two sources that speak to this issue: Nichols & Zax (1977) and my (1979) book on emotions.
Nichols and Zax (1977) have distinguished, as I do here, neurophysiological components of catharsis and described their climatic states: "The somatic—emotional aspect consists of the motoric discharge of emotion in expressive sounds and actions such as the tears and sobbing of grief, or the trembling and sweating of fear." They also indicate that laughter is cathartic, but fail to identify the associated emotion. The answer to this question has been suggested by Helen Lewis, who notes that although laughter may have many different functions, one function is to dispel shame. (1981, pp. 203, 318.)
Lewis further shows that Freud's analysis of wit and humor, which on the surface appears to involve only anger, also involves shame. (Lewis, 1984, chap. 6.) The naming of laughter as the catharsis for shame is very much in accord with the popular belief that laughter may bring relief for embarrassment, a shame variant.
There is a large self-help psychology literature that assumes that all laughter is beneficial. However, as Billig (2005) has strongly suggested, it is not the case. Most humor, he points out, involves ridicule of others, rather than acknowledgment of one’s own shame. A good laugh in the theatre, for example, occurs only when one identifies with the character that is making a mistake, rather than identifying with those pointing a finger at that character. Most racist, sexist, and ethnic humor is therefore not cathartic.
The situation with respect to anger is even more complex and confusing. According to popular lore, violent actions and speech dispel anger. This belief is shared by many clinicians and researchers, including Nichols and Zax. They identify "angry yelling" as the signal of the catharsis of anger. As already indicated, this idea seems erroneous. It is important to establish that emotion response cycles areinternal sequences, otherwise one is back to the now discredited concept of human instincts. Angry yelling is not a reflex like crying or spontaneous laughter, but a complex sequence of external behavior. Yelling or violent hostile acts may occasionally lead to the catharsis of anger, although studies of overt aggression and of family conflict suggest that they usually don’t. Furthermore, these studies also suggest that yelling and aggression may increase, rather than decrease, the level of anger.
The stimulus that leads to the impulse of anger might be frustration. One is prevented from getting what one thinks one wants, needs, or it entitled to. For the endpoint, the idea of the expression “the heat of anger” might be a hint at the orgasm of anger: a rise in body temperature, even a small one, could signal rapid metabolism of the adrenaline that is generated in anger (p. 65). I will refer to this event as “good anger.” Unfortunately, good anger seems to be rare. Certainly in my own life the proportion of good anger to acting out seems to have been quite small, probably much less than one time in a hundred. Anger orgasms seem to be difficult to achieve. The discussion of optimal distance, below, may help understand why this is the case.
With respect to fear, the stimulus might be physical danger to life and limb. The consummatory phase could be shaking and sweating, as Nichols and Zax (1977) have suggested. Although there are clinical and biological literatures that associate shaking and sweating with panic attacks, they don’t connect these responses to completion. In my contacts with soldiers during my own stint in the US Army, with students reporting fear experiences, and indeed in my own life, I have become aware of situations involving “good fear,” that is, shaking and sweating that resolves fear rapidly and completely. As with the case of anger, good fear is probably also rare, though more frequent than good anger.
The idea that there is an optimal bodily response to fear, signaled by shaking and sweating might explain an otherwise unsolved problem in the physiology of both fear and anger. These two emotions are especially mobilizing, preparing the body for intense exertion, as in the examples of fight or flight. The bodily chemical that prepares for exertion is adrenalin. Excessive un-metabolized adrenalin not only causes long-lasting consciousness of unwanted arousal, but it also appears to cause physical damage to the body.
In the flight response to fear, it seems likely that much of the excess adrenaline is metabolized by extreme exertion. But many animals respond to danger not by flight, but by freezing. Surely there must be a genetically determined physiological response that metabolizes the excess adrenalin in both animals and humans.
Finally, the impulse and consummatory phases of shame/embarrassment are even less consensual. One possibility: a threat to the social bond, feeling disconnected from another person or persons, might be the occasion for the impulse of shame (Lewis 1971), and good natured laughter at self would signal orgasm, the completion of the shame cycle (Scheff 1990; Scheff and Retzinger, 1991; Retzinger 1991; Scheff 1994; 1997; 2006).
I have heard many accounts about embarrassing situations that were resolved by laughter. This excerpt is from a report by a female student on a roller-coaster ride at the age of thirteen:
My best friend, Nicole, was next to me as we clicked up to the very top, and then flew backwards at full speed. At this point in the ride I had no control over my body and I peed in my pants! At first, I was in shock. As the ride ended, a million things were running through my head. I could not believe what just happened. What were my friends going to think about me? How was I supposed to explain to my boyfriend why my pants were soaked? Was everyone going to make fun of me? This was going to be the most embarrassing night of my life!
When the ride stopped, I turned back to my other friends that were sitting behind me. “You guys…I just peed my pants!!!!” I didn’t know what to do! I was so embarrassed that I had actually peed in my pants. I felt like my life was over.
Nicole: “Oh my god, you really did pee!!!” We both looked at my seat and at my soaking wet jeans, and then at each other. There was nothing to do but laugh. We BURST into laughter. We walked towards the exit laughing so hard that we could barely walk straight. My other friends, who were still trying to understand what had happened, were amused. “Oh my gosh!” “You poor thing!” “How did that happen?”
This story, and many others like it, suggests that good-natured laughter is the orgasm for embarrassment/shame. Orgasm ends arousal very quickly. After a genuine climax, emotional arousal decreases in seconds to zero, the mind is clear, and one feels transformed to the point being fully alive.
Although the general definitions proposed here are only speculative, they could at least provide many hypotheses to be tested. One obvious hypothesis concerns long lasting emotions. At a very abstract level the model of emotions proposed here implies that long lasting emotions would arise when an emotion sequence is initiated but not completed. There are faint hints in the literature that might support this explanation.
Retzinger’s (pp. 49-53) analysis of facial expressions that were recorded on video suggests that unless there is completion, signs of anger arousal diminish only very slowly. It seems possible that the completion of anger might be only asymptotic: approaching zero arousal but infinitely slowly. Although each incompleted anger sequence would last only in a tiny, ever decreasing amount of arousal, the lamination of thousands of such paper-thin arousals could be the equivalent of continuous anger.
At a lower level of abstraction, another process prolonging arousal would be recursion, such as being ashamed of being ashamed. Panics in theatre fires may arise in this way: seeing others are afraid works back and forth reciprocally until the audience members stampede attempting to escape the fire. There is a literature also on shame-anger sequences that result in aggression: an insulted person retaliates in anger against the other person, who retaliates back, etc.
This kind of recursion seems to be occurring both within and between each person (Lewis 1971; Scheff 1990; Scheff and Retzinger, 1991; Retzinger 1991; Scheff 1994; 1997; 2006: Tangney and Dearing. 2002). Retzinger (1991) provides a particularly detailed exploration of this sequence. She shows that unacknowledged shame preceded every one of 18 incidents of angry escalation in the quarrels of four different couples.
The models of the four emotions are repeated in the table below.
Emotion Name Stimulus Climax
Grief Loss Crying (Sobbing with tears)
Fear Physical danger Shaking and sweating
Anger Frustration Body heat
Shame/Embarrassment Threat to bond Laughter
After a genuine climax, emotional arousal decreases in seconds to zero, the mind is clear, and one feels transformed to the point being fully awake and alert.
But these models don’t resolve the other problem mentioned above, the way in which emotions that are ordinarily painful can be experiences as pleasurable. To explain this phenomenon, it will be necessary to discuss the idea of distancing/role taking, which forms a crucial part of a new theory of catharsis.
Many of the critics of the idea of catharsis have noted that not all crying and laughing are helpful. However, these critics have been unable to find variables that make for helpful and non-helpful crying or laughing. The following discussion may point in this direction.
Discussions of catharsis in the humanities in the last hundred years have been based on one central idea, the optimal distancing of emotion. Bullough (1912) was the first to argue that a work of art, such as theatre, needed to be at an aesthetic (optimal) distance from emotion, neither too far (overdistanced) nor too close (under-distanced). This idea has been central to many theoretical discussions of theatre.
Theatre people know, or at least act as if they know, that scripts that don’t arouse emotions in the audience will fail. That is, members of the audience must identify with some of the characters. Furthermore the identification must be to the point that they react to the play as if what is happening to the characters is happening to them. For example, if the protagonist is facing danger to life and limb, members of the audience will feel fear as if they themselves were in danger. If such identification does not occur, then the play is overdistanced.
However, theatre people seem to also know that there is such a thing as too much identification, leading to overwhelming emotions in the audience. I have heard from students that some recent films, such as Saw and its sequels, come very close to this point for them. Some of the scenes arouse fear (and/or disgust) to the point that they come close to being, or actually are, unpleasant. In the scheme that is being described here, these films are under-distanced.
Finally, there is the idea in the making of drama that there is an optimal or aesthetic distance in the theatre. At this distance, emotions such as fear, anger, grief and shame, which are ordinarily painful, can be experienced as increasing not only the interest of the audience, but also providing pleasure. How could this be?
It is possible that Mead‘s idea of role-taking (1934) can be used to develop an explicit theory of distancing. Society is possible, Mead argued, because humans can momentarily take the viewpoint of the other. That is, children learn very early how to take on the point of view of others. Ordinary language is complex and often so fragmented as to be wildly ambiguous. By the time children are three or four, they can move rapidly between their own point of view and that of the speaker, in order to understand what is being said to them. This process is learned so early, and becomes so automatic that we forget that we are doing it. As adults, individuals can even see matters from the point of view of imaginary others (“the rational man” in classical economics) or all others (“the generalized other.”).
Mead argued that human societies are based on flexible coordination between its members, unlike the rigid, instinct determined coordination that characterizes the social activity of all other creatures. But flexible coordination, he proposed, was based on role-taking, of seeing one’s activities not only from one’s own point of view, but also from the point of view of others.
As he stated it, Mead’s theory of society was profound, but also extremely abstract. He gave very few examples of instances of role-taking, so that it is difficult to apply his theory to actual events. Most importantly, he didn’t spell out the actual steps involved in the process of role taking. In order to better understand the concept of distancing, it will necessary to visualize the meaning of role-taking second by second in social interaction and in internal dialogue as well.
The Role-taking Process
How does role-taking play out in discourse and in internal dialogue? It seems clear that there must be a process of back and forth between one’s own viewpoint and the viewpoint of others. But neither Mead nor anyone else has considered this issue in sufficient detail.
Koestler’s theory of laughter (1967) provides an application to a specific emotion, embarrassment. Koestler’s idea is that binocular vision, seeing the same thing in two different ways, leads to laughter. According to Koestler, we laugh at a pun because it surprises us with two contradictory meanings of a word. Ordinarily, he says, we are virtually enslaved to each of the separate meanings of the word, unaware of the conflict. The pun or joke is a situation that brings the conflict to awareness, which leads to embarrassment over an “error.” For a moment, we are liberated from embarrassment through laughter. The idea of role taking, of visualizing a word from two different points of view, underlies Koestler’s treatment of puns.
Audience reaction to plays allows a way of visualizing the process of role-taking by the members of the audience. For example, there is a type of drama that does not employ optimal distancing. In Nietzsche’s terminology it is called Dionysian (as against classical drama, which he called Appolonian). In the case of this type of tragedy, the aim seems to be to maximize the suffering of the audience. In lighter fare, the aim may be the experience of chaos and/or ecstasy.
But in the business of tragedy that brings tears and comedy that gets laughs, the script is ENTIRELY dependent on optimal distance. Bertrand Evans (1960) used the concept of discrepant awareness as the focal point for his analysis of Shakespeare’s comedies and romances. He argued that awareness control was the basis of the appeal and power of drama. The well made drama, he argued, moved audience awareness back and forth between the awareness of characters in the drama and their own point of view.
Since the publication of my book on catharsis (1979) there have been several lines of study that parallel the idea of distancing in catharsis. The idea of “pendulation” (Levine 1997) seems to be identical to the process of distancing/ role-taking. Psychotherapists use alternation between what they call de-regulation and regulation (of emotions) to help patients deal with trauma. For example, the victim of an auto accident is encouraged to relive the moment, but coming and going between it and the present (Heller and Heller 2004). The right rhythm represents optimal distance, a balance between painful and comforting moments.
Another recent discovery that may relate to distancing/role-taking involves what is called “mirror neurons” (Goleman 2006). The first studies were done with primates, showing that merely seeing another monkey’s action caused the initiation of the bodily preparation for same action in the monkey who was merely watching the other (Di Pellegrino et al 1992). For example, the first monkey’s body automatically prepares to smile by seeing the other monkey smile. The action is not necessarily repeated, however, but only prepared for: the body responds by bringing up what Mead called the “attitude” of smiling (bodily preparation for smiling). The original researchers found that the initiating response by the viewing monkey occurred in the neurons in the brain. They spoke of this reflecting process as “mirror neurons.”
Later studies found evidence for the same kind of mirroring in humans (such as Fadiga et al 1995). These studies suggest that empathic preparation is genetically wired into the human brain. To prove the relevance to the role-taking/distancing idea, alternation between the brain structures that represent self and other would have to be demonstrated.
Until such a demonstration, it seems to me that Goleman (2006) overstates the case. He proposes that empathy is always automatic. However, there are also many processes, such as denial and repression, working against empathy, even if it is physiologically available. I will need to further review these studies in order to describe both sides of the equation, processes that work for and against empathy. Another issue is the relationship between limbic resonance (Lewis, et al. 2001) and mirror neurons. I suspect they are closely connected, but haven’t found this link mentioned in the literature.
The idea of pendulation, although couched in somewhat different terms, is identical to the idea of optimum distance: one goes in and of emotions that would be otherwise unbearably painful or threatening. Perhaps the main basis for the importance of this kind of alternation is that it provides control, that is, if the feeling gets too intense, one can pendulate out of it. This is the also the way distancing works in audiences to a drama; they know they are free to leave if they wish. I recall from my own childhood that I did in fact leave on several occasions.
Two Instances of Emotion Climaxes from My Own Life
I also had a similar experience of control in a situation in real life, rather than in the theatre. It occurred many years ago, after my first group psychotherapy session. I was telling my then girlfriend about the session, how I was amazed that many of the participants were crying. I also told her that I didn’t cry, and felt somewhat impotent because I couldn’t. I became very sad while telling her this last part.
To my surprise, I began crying at that point. It was an intense cry, but had no mental content. I was crying with my whole body, like a baby. It felt great, after many years of not crying. After some 15 minutes, the crying stopped, and I no longer felt sad.
As it turned out, the crying was the introduction to the next episode, which was obviously about anger. Again with no mental content, I began snarling and biting at the empty air while my body twisted and writhed. The movements were so extreme that I fell on to the floor that was covered with a shag rug. Seeing the shag, I begin biting and chewing on it like an animal.
Up to this point, the anger movements seemed spontaneous. I felt that it wasn’t really me acting ridiculous, but that my body had taken over. After only a few moments on the floor however, I wondered how my girlfriend was taking all this. So I stopped, saying to her in an ordinary tone of voice: “Rachel, are you OK?” In the hippy style of the time, she answered: “Yes, just do your thing.” So I immediately went back to snarling and chewing on the rug, as if there had been no interruption.
As with the crying episode, this one lasted about 15 minutes. When I stopped, I felt all tension had drained away. I felt better than I could remember ever feeling before. It is possible that it was the distancing, the knowledge that I could stop if I wished, that made both episodes possible. The concept of optimal distance, of pendulating between unresolved emotions and safety, might serve to explain many contradictions in the field of emotions.
How does this idea apply to practical problems of emotion management? To illustrate their application, consider once more the problem of anger management discussed in the introduction. Is there a middle way between too much and too little anger? One possible approach is the formula “I am angry at you because…” This statement should be made, as nearly as possible, in an ordinary conversational tone of voice: firmly, but without undue vehemence.
This approach serves two main ends. First it allows the angry person to voice his anger verbally, rather than hiding it. That is, anger can be expressed verbally rather than acting it out vehemently. Verbal expression informs others and even self that we are in earnest, but it avoids the kind of excess that makes trouble.
Secondly, in my own experience and from reports by students, verbal expression seems to be the best vehicle for the kind of anger orgasm described above: body heat. That is, virtually all of the “heat of anger” climaxes that I know of have occurred in the course of verbal expression of anger, rather than acting out or complete suppression. Verbal expression would seem an ideal way to alternate between the anger that one feels, and stepping outside it into ordinary conversation.
One of the drawbacks of verbal expressions of anger is that if you are too polite, the other person may not realize that you are in earnest. The remedy for this problem is to be prepared to restate your position until they get the idea. One can be relentless without being rude. This approach is just a restatement of Virginia Satir’s notion of “leveling:” being direct but respectful. However, it is advisable to use slightly different words each time you restate your position, lest the other person feel insulted by the repetition.
To provide a final example, here is another from my own life. The next day after the episode above, I was on a long airplane flight. By chance I was sitting next to a colleague from another department on my campus. I had always felt intimidated by him, partly because he was famous, but mostly because I knew that he had a sharp tongue. Nevertheless, I felt impelled to tell him about the exhilarating experience I had had the day before. After only a few sentences, he interrupted my story in order to “psychoanalyze me,” as my students would say.
I surprised myself by quickly interrupting him in turn. In an ordinary voice, and without planning what I was going to say, out came “David, you are trying to reduce my experience to yours without remainder, and I won’t stand for it.” At this point I noticed what I thought of at the time as a rise in the cabin temperature. Next came the really big surprise. He spent most of the rest of the flight apologizing for his rudeness. Far from being intimidating, he was wonderfully respectful and courteous. It appears that in the right format, anger can be instructive rather than disruptive.It might seem impetuous to move from the personal to the societal with the same model, but one of the implications of the theory proposed here is that is necessary to do so.
An example of the need for resolving hidden emotions comes from clinical experience with unresolved grief, already mentioned above. Experienced clinicians have long noted that most adults in our society find it difficult to resolve the emotions associated with profound lost. In modern societies there is a social institution that has been called the denial of death. Most of those who have suffered the lost of a loved one appear to be unable to finish their mourning. For most losses of this kind, working through the grief from a loss requires considerable time and sympathy, certainly more than a year. But the social network of those who should provide support, the immediate family and friends, usually tolerate mourning for a much shorter time. After a month, at best, the supporters tend to offer advice instead: see a shrink, take a pill, get over it, etc. Bereavement counselors find that they can help with the resolution of grief merely telling the bereaved that what they are going through is completely normal.
As indicated earlier, the repression of emotions might be the key to understanding the repetition compulsion at the collective level. The masking of the vulnerable emotions (fear, grief, and shame) behind a façade of self-righteous aggression is no stranger to the arena of international politics. The way in which the attack on Iraq served to cover over the tragedy of 911 is a case in point.
Classic Greek drama, following Aristotle, taught that the health of body politic is dependent on collective catharsis. In particular, the classic plays made it clear that nations that don’t mourn their tragedies are doomed to repeat them. The uncanny resemblance between the wars against Vietnam and Iraq is only the latest installment of this prophecy. Persons, groups, and nations need to resolve their hidden emotions rather than cover them over with self-righteous anger and aggression.
This paper first reviewed the issue of catharsis of emotions. It seems now that almost all of the commentary for the last 40 or so years, excepting clinical work, has been based on erroneous premises. Clinical studies, on the other hand, have steadily upheld the idea. With this review as a background, the paper introduced a new theory in two sections. The first section further extended the Dewey/Mead/Shibutani approach to the socio-psychobiology of emotions. It proposed a stimulus and climax for four emotions: grief, fear, anger and shame. This approach might explain emotional states of long duration. One implication is that in modern societies, emotions are usually repressed rather than resolved. The second section expands the idea of distancing by relating it to Mead’s theory of role-taking, recent work on somatic treatments of traumatic emotional stress, and current findings on mirror neurons. This part of the theory may help explain the enigma of safe, even pleasant resolution of painful emotional states. A further application may be needless collective violence: the failure to mourn past mistakes gives rise to hiding emotions under the cover of anger and aggression. Needless to say, this and other ideas suggested here are only hypotheses. But they may provide incentive for new directions in research and practice involving emotions.
Aristotle. Poetics. Third century B.C. New York: Theatre Communications Group. (1999)
Belifiore, Elizabeth S. 1992. Tragic Pleasures: Aristotle on Plot and Emotion. Princeton: Princeton University Press.
Bemak, Fred and Young, Mark. 1998. Role of catharsis in group psychotherapy International Journal of Action Methods: Psychodrama, Skill Training, and Role Playing. Vol 50(4), 166-184.
Berkowitz, Leonard, 1962. Aggression: A Social Psychological Analysis. New York: McGraw-Hill.
———————. 1970. Aggressive humor as a stimulus to aggressive responses. Journal of Personality and Social Psychology 16: 710-717.
Billig, Michael. 2005. Laughter and Ridicule: Towards a Social Critique of Humor. London: Sage.
Bushman, Brad J. 2002. Does venting anger feed or extinguish the flame? Catharsis, rumination, distraction, anger and aggressive responding. Personality and Social Psychology Bulletin. Vol 28(6). 724-731.
Dewey, John. 1894. The Theory of Emotion. Psychological Review. 1; 6 (553-568) and II; 1 (13-32).
Di Pellegrino, G., et al. 1992. Understanding Motor Events: A Neurophysiogical Study. Experimental Brain Research 91, 176-80.
Evans, Bertrand. 1960. Shakespeare’s Comedies. London: Oxford University Press
Fadiga, L., et al. 1995. Motor Facilitation during Action Observation. Journal of Neurophysiology 73 2608-26.
Feshback, Seymour and Jerome Singer. 1971. Television and Aggression. San Francisco: Jossey-Bass.
Geen, Russell G. 1976. “Observing violence in the mass media: Implications for basic research.” Pp. 193-234 in Russell G. Geen and Edgar C. O’Neal (eds.), Perspectives on Aggression. New York: Academic Press.
Goleman, Daniel. 2006. Social Intelligence. New York: Bantam
Heller, Diane and Larry Heller. 2004. Somatic Experiencing in the Treatment of Automobile Accident Trauma. US Association for Body Psychotherapy Journal, 3, 42-52.
Kaplan, Robert M. and Robert D. Singer. 1976. “Television violence and viewer aggression: A reexamination of the evidence.” Journal of Social Issues 32: 18-34.
Koestler, Arthur. 1967. The Act of Creation. New York: Dell
Kosmicki, Frank X. and Glickauf-Hughes, Cheryl. 1997. Catharsis in Psychotherapy. Psychotherapy: Theory, Research, Practice, Training. Vol 34(2), 154-159
Liebert, Robert M. and Neala S. Schwartzberg. 1977. Effects of mass media. Annual Review of Psychology 28: 141-173.
Levine, Peter A. 1997. Waking the Tiger. Berkeley: North Atlantic Books.
Lewis, Thomas, et al. 2001. A General Theory of Love. New York: Vintage Books.
Mead, George H. 1934. The Philosophy of the Act. Chicago: University of Chicago Press.
Mitscherlich, A. and M. 1975. The Inability to Mourn. Principles of Collective Behavior New York: Grove Press, 1975.
Nichols, M. P., & Zax, M. (1977). Catharsis in psychotherapy. New York: Gardiner.
Parkes, Colin. 1988. Bereavement: Studies of grief in adult life (3rd ed.)Madison, CT: International Universities Press, Inc.
Quanty, Michael B. 1976. “Aggression catharsis: Experimental investigations and implications.” Pp. 99-132 in Russell G. Geen and Edgar C. O’Neal (eds.), Perspectives on Aggression. New York: Academic Press.
Quine, William. 1979. A Postscript on Metaphor. In S. Sacks (Editor), On Metaphor. Chicago: University of Chicago Press.
Retzinger, Suzanne. 1991. Violent Emotions: Shame and Anger in Marital Quarrels. Newbury Park: Sage.
Scheff, Thomas. 1979. Catharsis in Healing, Ritual, and Drama. Berkeley: U. of California Press.
________ l990. Microsociology: Discourse, Emotion and Social Structure. Chicago: U. of Chicago Press.
________1994. Bloody Revenge: Emotions, Nationalism, War. Boulder: Westview
_________ 1997. Emotions, the Social Bond, and Human Reality. Cambridge: Cambridge University Press.
_________2006. Goffman Unbound. Boulder: Paradigm Publishers.
Scheff, T. J. and S. M. Retzinger. 1991 Emotion and Violence: Shame and Rage in Destructive Conflicts. Lexington, MA: Lexington (2000 iUniverse).
Scheff, T. J. and David Fearon. 2004 Social and Emotional Components in Self-Esteem. Journal of the Theory of Social Behavior. 34: 73-90.
Scoppetta, M., G. Di Gennaro and C. Scoppetta. 2005. Selective serotonin reuptake inhibitors prevent emotional lability in healthy subjects. Eur Rev Med Pharmacol Sci. 9 (6): 343-8.
Shibutani, Tomatsu. 1961. Society and Personality. New York: Prentice-Hall.
Symonds, Percival. 1954. A comprehensive theory of psychotherapy. America Journal of Orthopsychiatry 24: 697-712.
Tangney, June, and Rhona Dearing. 2002. Shame and Guilt. New York: Guilford.
Taylor SE, Klein LC, Lewis BP, Gruenewald TL, Gurung RA, Updegraff JA. 2000. Biobehavioral Responses to Stress in Females: Tend-and –Befriend, not Fight-or-Flight. Psychological Review. 107: 411-429.
Vingerhoets, A., F. Bussel, and A. Boelhouwer. 1997. The (non)expression of Emotions in Health and Disease. Tilburg: Tilburg University Press.
Volkan, Vamik D. 1975. “Re-grief therapy.” Pp. 334-350 in Bernard Schoenberg et al. (eds.), Bereavement: Its Psychosocial Aspects. New York: Columbia University Press.
Volkan, Vamik D. and Zintl, Elizabeth (1993). Life After Loss: Lessons of Grief. New York, NY: Charles Scribner's Sons.