What Is Emotion? History, Measures, and Meanings (review)

This book poses a valuable challenge to some recent trends in affective science. Kagan points out recurring methodological deficiencies in emotion research and isolates recurring inferential flaws in theorization. His arguments should provoke more rigorous experimentation and reflection. However, when Kagan makes positive arguments, his analyses often suffer from the same methodological and inferential flaws. Indeed, one might argue that Kagan’s objections often apply more forcefully to his own claims.

More exactly, Kagan takes issue with several tendencies in current work on emotion. His major targets range from the standard psychological and neurological methods of study (such as self-reports and fMRI scans) to such fundamental theoretical issues as the existence of basic emotions. He opposes these tendencies to his own preferences for seeing emotions as far more particular and as strongly shaped by culture and gender.

Some of Kagan’s arguments are already familiar in the social sciences, though he develops them lucidly. For instance, he carefully outlines the problems with self-reports. Many researchers rely on test subjects’ judgments about their emotional states. Kagan shows how unreliable these are.

Unfortunately, Kagan himself draws on testimonies that are similar to but more problematic than self-reports. For instance, Kagan is concerned to make the case for some deep cultural differences between East and West. In connection with this, he cites an anecdote wherein a Hindu “schoolmaster told an English visitor” that (here, he quotes the schoolmaster) “if some close relation dies, even if a wife or son,” then “Great Indian souls” would not be “too much distressed” (173). This is problematic in numerous ways. First, such an assertion is evidentially valueless. Second, even if it happens to be true, it does not make a comment about Indians generally, but about mahatmas (great souls), those who have achieved spiritual advancement. Despite clichés, ordinary Indians are like everyone else in becoming very much distressed at the death of loved ones. Third, such a comment should apply equally to, say, American Christians who, if truly devoted, will believe that the deceased is happily with God, and available for eventual reunion.

Kagan’s criticism of self-reports is part of a larger critique implicitly indicating that current affective scientists fail to consider an adequately wide range of alternative hypotheses. In connection with this, Kagan details some problems with fMRI scans—first, the extreme ecological distortion of fMRI studies of emotion; second, the complexity of interpreting the scans. His discussion of work on the amygdala is perhaps the best part of the book. His evidence suggests that apparent links between the amygdala and fear may be more aptly characterized as involving caution.

One way of putting Kagan’s general point is that, like their fellow humans, researchers in emotion suffer from confirmation bias. They begin with an idea—for example, that amygdala activation is involved with fear. They do research, then select the interpretations that fit their initial idea. But, here too, Kagan suffers from the same fault—most obviously, in his arguments for very deep East/West and male/female differences. It is not that Kagan is clearly wrong in his general claims. It is simply that his arguments are often highly speculative and equivocal, at least as much as the accounts he criticizes. (Consider, for instance, his idea that women will not pursue mathematics in large numbers because, for genetically determined reasons, they cannot enjoy solving math problems as much as men do.)

Central to Kagan’s approach is his advocacy of particularity in the study of emotions. For example, in Kagan’s view, fear is too broad and incoherent a category. The fear of an attacking predator is not the same as the fear of being audited by the IRS. This is for two reasons. First, the particularity of the initial feeling is not the same. Second, the subsequent appraisal of the situation is not the same.

Related to this, Kagan stresses that our experience of emotion is almost always “blended.” That is, our emotional state at...