Theories of Emotion

1Which comes first: the trembling or the fear?

Common sense says that you see a snake, feel afraid, and therefore your heart races and you jump back. In 1884 the American psychologist William James turned this order upside down. In his article "What is an emotion?" he wrote that "we feel sorry because we cry, angry because we strike, afraid because we tremble". More than a century of research has gone into testing that bold claim and the theories that answered it. Each theory gives a different answer to two questions: what order do the parts of an emotion come in, and how much does thinking matter?

21. The James–Lange theory (1884–1885)

William James (1884) and, independently, the Danish physician Carl Lange (1885) proposed that an event first triggers bodily changes, and that the emotion is our perception of those changes. Without the racing heart and trembling, James argued, fear would be only a cold judgement that danger is present.

• Support. Bodily states do colour emotion: people often notice a pounding heart before they can name what they feel. George Hohmann (1966) interviewed 25 men with spinal cord injuries and reported that those with higher injuries, who feel less from their bodies, described weaker emotions.
• Against. Hohmann's study was small and based on interviews. Later, larger studies (for example Chwalisz, Diener and Gallagher, 1988) found that people with spinal cord injuries reported emotions as intense as anyone else's.

32. The Cannon–Bard theory (1927)

The physiologist Walter Cannon, later with Philip Bard, attacked the James–Lange theory. He argued that:

1. the internal organs (viscera) are too slow to change: an emotion can be felt in a fraction of a second;
2. the internal organs have few sensory nerves, so we cannot sense their changes finely;
3. very different emotions (fear, anger, even excitement) produce similar arousal, so arousal alone cannot tell us which emotion we feel;
4. producing arousal artificially, for example with an injection of adrenaline, does not produce a true emotion (Marañón, 1924: people said they felt "as if" afraid);
5. animals whose internal organs were surgically cut off from the brain still showed emotional behaviour.

Cannon and Bard proposed instead that the thalamus sends signals at the same time to the cortex (producing the feeling) and to the body (producing arousal). Evaluation: the idea that feeling and arousal arise in parallel was an advance, but the thalamus is not the "emotion centre": structures such as the amygdala, hypothalamus and prefrontal cortex are involved. And arousal patterns are not completely identical across emotions: Robert Levenson and colleagues found some differences, for example in heart rate and skin temperature, although later reviews show these patterns are inconsistent from study to study.

Figure 1. Four theories of emotion applied to the same event. The theories disagree about the order: does the body come first (James–Lange), together with the feeling (Cannon–Bard), before a label (Schachter–Singer), or after an appraisal (Lazarus)?

43. The Schachter–Singer two-factor theory (1962)

Stanley Schachter and Jerome Singer proposed that an emotion needs two factors: general physiological arousal, and a cognitive label that explains it. The same pounding heart becomes fear beside a snake and excitement at a football final.

The experiment. 184 male university students were told they were testing a vitamin ("Suproxin"). Most were injected with adrenaline, which raises arousal; a control group received a placebo. Some were told the true side effects (racing heart, trembling); others were told nothing or were misinformed. Each then waited with a confederate (an actor) who behaved either euphorically (playing with paper balls and a hula hoop) or angrily (complaining about an insulting questionnaire). Participants who had no correct explanation for their arousal tended to "catch" the confederate's mood more than those who knew it was the injection.

Evaluation.

• The results were weaker than the theory's fame suggests: several differences were small, and the placebo group did not always differ from the adrenaline groups as predicted.
• Replications largely failed. Gary Marshall and Philip Zimbardo (1979) and Christina Maslach (1979) found that unexplained arousal tended to feel unpleasant, whatever the confederate did, rather than taking on any label.
• Ethics: participants were deceived and injected with a drug without true informed consent; this would not be allowed today.
• Lasting contribution: the idea of misattribution. In Donald Dutton and Arthur Aron's (1974) field study in Canada, men who had just crossed a high, swaying suspension bridge were more likely to telephone an attractive female interviewer afterwards than men on a low, stable bridge, as if they had relabelled fear as attraction. But the men were not randomly assigned to the bridges, the samples were small, and later studies give mixed results.

54. Cognitive appraisal: Lazarus

Richard Lazarus (1966, 1991) argued that appraisal comes first and determines the emotion. In primary appraisal we ask, "Does this matter to me? Is it a threat, a loss or a challenge?" In secondary appraisal we ask, "Can I cope with it?" Fear follows "threat, and I cannot cope"; challenge and even excitement can follow "threat, but I can cope".

Support: Joseph Speisman, Lazarus and colleagues (1964) showed people the same unpleasant film of a painful initiation ritual with different soundtracks. A soundtrack that stressed the pain produced more physiological stress than one that described the scene calmly and from a distance. Against: Robert Zajonc (1980, 1984) argued that some feelings arise before any thought, for example liking things simply because we have seen them before. Joseph LeDoux (1996) found a fast route for fear in rats, from the thalamus straight to the amygdala, which can trigger a fear response before the cortex has identified the danger. Most psychologists now accept both: some emotional reactions are very fast and automatic, while appraisal shapes most of our everyday emotional life.

65. Facial feedback: a replication story

The facial feedback hypothesis is a modern relative of James–Lange: making a facial expression should strengthen the matching emotion. Fritz Strack, Leonard Martin and Sabine Stepper (1988) had people hold a pen in their teeth (forcing a smile-like expression) or in their lips (preventing one) while rating cartoons; the "smiling" group rated them as funnier. But in 2016 a Registered Replication Report run across 17 laboratories (Wagenmakers and colleagues) found no effect. A large international study, the Many Smiles Collaboration (Coles and colleagues, 2022), then found that deliberately posing a smile produced a small increase in reported happiness, while the pen method again showed no clear effect. Conclusion: facial feedback probably exists but is small, and the famous pen study is not reliable evidence for it.

76. A modern view: constructed emotion

Lisa Feldman Barrett's theory of constructed emotion (2017), which you met in the last lesson's video, can be read as a descendant of the two-factor idea: the brain uses past experience and context to make sense of bodily sensations, and the meaning it constructs is the emotion. Critics reply that it underplays evidence for evolved, partly universal emotion systems. The debate is open.