volume 4
october 2001

Testing the curtailment thesis

A research proposal on the curtailment of emotional expression in popular songs and novels

by Thomas J. Scheff

According to Norbert Elias, the modernization of society curtails the expression of anger and shame, while at the same time furthering sexual expression. But, how does popular music fit in with this "curtailment thesis"? Following his analysis of popular song lyrics — Words of love and isolation — Thomas J. Scheff here sketches the outlines of a research project to test the curtailment thesis on popular music and novels.

Problem. Theorists have argued that modern civilization curtails the expression of emotion. Max Weber predicted that modern societies would have rationalization as their driving force, which would mean the suppression of emotions. In a similar vein, Parsons suggested that urban industrial societies would move, in most of their arenas, toward "affective neutrality." For Freud (1930) civilization was tantamount to repression of feeling and desire, including sexual desire. In this vein, Freud had many followers, such as Reich, Marcuse, and others.

Following Freud, Elias (1984) was more specific, suggesting that modernization curtailed expression of anger and shame. Elias contradicts Freud about the expression of sexual desire, since he argued that the suppression of shame would increase sexual expression. A related, but quite different thesis asserts only a change in the type of emotion, a historical shift from shame to guilt. Finally, it has been proposed that curtailment of the expression of shame leads to violence. These ideas have been stated only in broad terms, and without convincing evidence. The most cogent argument is by Elias (1984), who showed the curtailment in the expression of shame in excerpts from advice manuals in five European languages, over a span of seven centuries. But he recognized that his analysis was unsystematic, and therefore only suggestive.

These theorists imply first, that that the open expression of emotions is curtailed by modernization. Secondly, modernization does not decrease the actual amount of emotion and feeling, only its expression. Finally, they suggest that the blockage of emotional expression has serious negative consequences for both individuals and societies. I will subsequently refer to the three hypotheses — curtailment, blocking, and consequences — combined as the theory of curtailment. Although this study will only concern the first hypothesis, testing it could be important since the theory links it to the other two hypotheses. However, if the evidence does not support the first hypothesis, curtailment of expression, there will be less risk in ignoring the remaining two hypotheses.

If however, the hypothesis that modern civilization curtails the expression of emotion is supported, then we would need to further test the complete theory. My preliminary studies on all emotional expression in popular songs (Scheff, 2001), and the Stearnses on anger — see below — suggest support for the curtailment hypothesis, in the main. However, it is fairly clear that Freud was mistaken in one particular: the expression of sexual desire has obviously increased during the seventy years under study. But this proposition and the broader curtailment hypothesis need to be systematically tested.

Examples and limitations. The larger theory of curtailment could be important in many areas, such as childrearing, psychotherapy, education, crime control and conflict resolution. Psychotherapy, for example is divided between schools of thought that consider emotion — e.g., psychoanalysis, catharsis, gestalt — and those that do not — e.g., behaviorist, cognitive, and narrative therapies. If, as the theory suggests, blocking emotions from expression leads to pathology, findings supporting the curtailment hypothesis imply possible support for the emotion-oriented therapies, as well as changes in other areas.

Another example: managing intense emotions is a key concern in resolving protracted conflict. But most current training and literature on mediation and conflict resolution gives very little attention to emotions, perhaps making mediation less effective than it could be (Retzinger and Scheff, 2000). Another hypothesis is that curtailment of the expression of shame gives rise to violence (Gaylin, 1884; Scheff, 1994; Elias, 1996; Gilligan, 1996). This idea suggests an inverse relation between changes in the yearly rates of violent crime and the expression of shame. For example: would the large changes in the violent crime rates in the Uniform Crime Reports in the 1980's and 1990's, and changes in the amount of emotional expression in the media be correlated? The curtailment hypothesis is a very general statement about social process in modern civilizations.

I have found only one study that seriously attempts to document the curtailment thesis, a historical study of attitudes toward anger in the United States over the last two centuries (Stearns and Stearns, 1986). Their study concerns how curtailment of anger might occur: through condemnation of its expression. They examined a very large number of texts, including advice manuals, diaries, and secondary studies. Their findings seem to support the hypothesis with regard to anger. They report that early in the 19th century, only excessive anger was condemned. Righteous anger was not only not condemned, but even encouraged. Toward the middle of the 19th century, however, they suggest that intolerance of any kind of anger began to arise, and continues, increasingly, into the 20th century. But are their findings sufficiently reliable to support further investigation of the theory of curtailment?

The Stearnses admit that their study had many shortcomings. For example, they mention "inability to establish the representativeness of any given source" and lack of explicit coding procedures — "Often the researcher is forced to reason from brief comments on subtopics (e.g. temper tantrums), circumlocutions, ... and from the outright absence of comment where it should be logically expected" (Stearns and Stearns, 1986: 249). The names of most of the many documents they examined are not stated or even enumerated, and the method of coding anger not made explicit. For these and other reasons, the reliability of their findings cannot be assessed.

Another limitation of the Stearnses' study is that it concerns only anger, leaving out other emotions. Unlike Elias, for whom the rising threshold against the expression of shame was a key feature of the civilizing process, the Stearnses have little to say about shame, its siblings and cousins like embarrassment, humiliation, guilt, envy, etc. or grief, fear, pride and love. Expert opinion now holds that the various emotions interact, particularly anger with the shame family, contempt and disgust. If we are to study changes in emotional expression, it would be advantageous to include several emotions, not just one. Determining simultaneous historical changes in several emotions could discriminate between various theses. For example, has there been a shift from shame to guilt, or decreasing expression of all the emotions? If a systematic test of the curtailment hypothesis is positive, then further investigation of the whole theory would be warranted.

Hypotheses. This study will put the curtailment hypothesis to an exact test, as well as several related propositions, by identifying changes in the level of expression of emotions, affects and sexual desire in the texts of popular songs and novels. The specific emotion markers that will be identified are those for anger, fear, anxiety, shame, and guilt. In addition, explicit references to sexual desire will be identified.

Testing the curtailment hypothesis with the lyrics of popular songs and the texts of best-selling novels may ?? strategic. Current thinking in media studies and in the sociology of culture suggests that modern societies are increasingly rationalized, as Weber and Parsons suggested, but even so they still provide special enclaves in which emotions may be expressed, such as the family and the ???

Hypotheses

Freud / Reich / Marcuse: Over the last seventy years, there has been a decrease in the number of all emotion markers year by year. a. There has been a decrease in indications of sexual desire.
Elias: There has been a decrease in anger and shame markers. a. Because of the suppression of shame, there has been an increase in indications of sexual desire.
Alternative to the curtailment hypothesis: there has been a decrease in shame, but also an increase in guilt markers.
Lewis / Gaylin / Gilligan / Scheff: Violent crimes will vary inversely with the number of shame markers.

Some results from a preliminary study. In the first stage of the proposed study, I will use the Top 40 songs on the charts in the U.S. for the last seventy years (1930-1999) from the website Lyrics World (2000). Since only the most popular songs make the charts, they form what sociologists call collective representations, social facts that extend beyond the desires of particular individuals. These lyrics form a bounded and replicable domain, as do the most popular films and novels that will be examined in the second stage of this study.

About 11,000 titles and 9,000 lyrics are available on Lyrics World. These lyrics represent 91% of the total lyrics — that are not duplicates: very popular songs sung by different artists in the same year. From the preliminary study that I have carried out (Scheff, 2001), it appears that lyrics for the period 1930-1960 were much more likely to mention emotion and feeling names than later lyrics — e.g.:

With each word your tenderness grows, tearing my fear apart ...

["The Way You Look Tonight", 1936]

A few lyrics from the earlier period describe the hiding of feeling ("Laughing On the Outside, Crying On The Inside", 1946), but there seems to be many more in the later period, such as the following:

People say I'm the life of the party
Cause I tell a joke or two
Although I might be laughin' loud and hearty
Deep inside I'm blue

["Tracks Of My Tears", 1965]

And:

I pretended I'm glad you went away
These four walls closin' more every day
And I'm dying inside
And nobody knows it but me
Like a clown I put on a show
The pain is real even if nobody knows

["Nobody Knows", 1996]

Seemingly absent from 1930-1960 are lyrics that renounce all feeling, like those in Tina Turner's lines:

Who needs a heart when a heart can be broken?

[What's Love Got To Do With It?, 1984]

and Simon and Garfunkel's lyrics:

I've built walls
A fortress deep and mighty
That none may penetrate
I have no need of friendship; friendship causes pain
It's laughter and it's loving I disdain

["I Am A Rock", 1965]

Furthermore, it is clear that rather than decreasing, as Freud thought, indications of sexual desire increased over this period. The references to sexuality during 1930-1960 were infrequent and indirect. During 1961-1980 they became more frequent and less indirect in Top 40 lyrics. Finally in the period 1981-1999 they have become still more frequent and much more explicit. If the evidence shows that these changes were paralleled by a decreasing level of shame expression, it will suggest that Freud's attention to sexual repression was misdirected; he should have been concerned with emotional repression, specifically with the hiding of shame.

Method. I would use first a quantitative, then a qualitative, method to identify emotion markers, as Lewis did in her magisterial study of shame and guilt in psychotherapy sessions (1971). I am familiar with every aspect of her study. First she scanned transcripts with an early version of the Gottschalk-Glaser (1969) content analysis procedure to locate emotion markers. The emotion markers identified by this procedure are anger, fear, anxiety, shame, and guilt.

The Gottschalk program is made up of very long lists of words; each word on each list has been found to be often associated with one of the six emotions or affects. This procedure is only preliminary, since it uses single words rather than sentences, and is acontextual. Lewis then employed what would today be called discourse analysis to confirm the meaning of the markers in the actual verbal context. In Lewis's study, she provided only a brief list of markers for shame and anger. The literature today provides a comprehensive list of verbal markers for shame, anger, anxiety, fear, and guilt, so that identifying emotion markers in the analysis of discourse is now repeatable.

The Gottschalk procedure is available as computer software (Gottschalk, 1995; 2000). It has been translated into 13 languages in addition to English, and shown to be reliable and valid in all of them (Gottschalk, 1995: 24-25). My confidence in this procedure is based not only on these formal checks, but also, closer to home, that it made the Lewis (1971) study possible. I will follow up the computerized Gottschalk analysis of the total database by confirming the markers for each emotion in its verbal context. Because of the large numbers, I will analyze only a 20% sample (N = 1,800 lyrics). There will be three coders, myself and two assistants that I will train for this purpose. We will each examine the Gottschalk markers in 700 lyrics, which allows an overlap of 300 lyrics to determine inter-rater reliability.

In the second stage of the study, the same method will be applied to the text of the best-selling novel of each of the last seventy years. As with the song lyrics, I will use both the Gottschalk coding procedure and discourse analysis. Again the novel texts will be discourse analyzed independently by myself and an assistant, to check on intercoder reliability. It is generally understood that the market for novels is less dominated by young people than the market for popular songs. Including novels as well as popular songs, will enable me to trace very general changes in emotional expression,

This method will be an actual test of the theses discussed above: the Stearnses on anger, Elias and others on anger and shame, and the Freud / Reich / Marcuse hypothesis on emotions and affects, and on sexual desires. It will also determine the interrelations of changes in the numbers of marker of the six emotions and affects. These interrelations would show whether guilt markers increase as shame markers decrease, as has been claimed, and whether anger and shame markers decrease together, as I (Scheff, 1994) and others (Lewis, 1971) have proposed. Finally, I will also determine the correlations between emotion markers in media texts and rates of violent crimes. These kinds of assessment have never been attempted before.

References

Elias, Norbert (1994), The civilizing process. Oxford: Blackwell, 1994.
Elias, Norbert (1996), The Germans. Power struggles in the 19th and 20th centuries. New York: Columbia University Press, 1996.
Freud, Sigmund (1930), Civilization and its discontents. London: Hogarth Press, 1930.
Gaylin, Willard (1984), The rage within. Anger in modern life. New York: Simon and Schuster, 1984.
Gilligan, James (1996), Violence. Reflections on a national epidemic. New York: Vintage, 1996.
Gottschalk, Louis A., and Goldine C. Gleser (1969), Manual for using the Gottschalk-Gleser Content Analysis Scales. Berkeley: University of California Press, 1969.
Gottschalk, Louis A. (1995), Content analysis of verbal behavior. Hillsdale, NJ: Lawrence Erlbaum Associates, 1995.
Gottschalk, Louis A. (1999), Personal correspondence, 1999.
Lyrics World (2000), retrieved from the Internet, 2000, at: http://www.summer.com.br/~pfilho/html/top40/index.html.
Lewis, Helen B. (1971), Shame and guilt in neurosis. New York: International Universities Press, 1971.
Retzinger, Suzanne M., and Thomas J. Scheff (2000), "Emotion, alienation, and narratives: resolving intractable conflict." In: Mediation Quarterly, 2000, 18: 71-86.
Scheff, Thomas J. (1994), Bloody revenge. Emotions, nationalism, and war. Boulder, CO: Westview, 1994.
Scheff, Thomas J. (2001), "Words of love and isolation. Individualism and alienation in popular love songs, 1930-1999." In: Soundscapes, 2001, 4 (Autumn).
Stearns, Peter, and Carol Stearns (1986), Anger. The struggle for emotional control in America's history. Chicago: University of Chicago Press, 1986.