William James

The originating figure for the whole cluster. James’s 1884 Mind paper made peripheral bodily feedback causally primary in emotion — “our feeling of the [bodily] changes as they occur IS the emotion” — the thesis that all later theories in this wiki either extend, challenge, or reframe.

Core claim and its corollary

Strip an emotion of its bodily manifestation and only “a cold and neutral state of intellectual perception” remains. Restricted to emotions with “distinct bodily expression,” this entails discrete visceral patterns → the autonomic-specificity hypothesis and the basic-emotions tradition. See friedman-2010-jamesian-perspective.

Reappraisal (Friedman/Ellsworth)

The textbook “bear” example flattened James’s actual view. James acknowledged cognition (“Objects… take their place… as elements in total situations”), individual variation, and that “running” stood for many “invisible visceral” movements — implying inherently multivariate response patterns. He held within-emotion resemblance “trumps” individual variation, which Friedman reads as consistent with natural kinds (contra some constructionist appropriations of James — see lisa-feldman-barrett).

Line of descent

• Challenged by walter-cannon (Cannon–Bard) and by stanley-schachter (two-factor cognitive-arousal).
• Revived empirically by Ax, Ekman, Levenson (autonomic specificity) and conceptually by Damasio (somatic markers, “Neo-Jamesian”).
• Operationalized as an individual-differences claim by Schandry (1981) and the Munich psychophysiology tradition: if feelings are perceptions of bodily states, people who perceive their hearts better should feel more. See olga-pollatos, heartbeat-detection-task.
• Recast in predictive-coding terms by anil-seth (interoceptive-inference) — James’s ambiguous “perception of bodily change” reread as top-down Bayesian inference.

Where the empirical tests land, and it is not on James

Two studies in this wiki test the individual-differences corollary directly, and both report the same asymmetry: interoceptive accuracy relates to felt arousal (disputed — Pollatos et al. 2005 yes, Dunn et al. 2010 no) and to felt valence not at all (both, unanimously).

That is a problem for the strong reading of James specifically. If stripping an emotion of its bodily manifestation leaves “a cold and neutral state of intellectual perception,” then perceiving the body better should intensify the whole emotion, quality included. It does not appear to touch quality. What it touches, when it touches anything, is the arousal term — which is Schachter and Singer’s variable, not James’s.

Pollatos et al. nevertheless present their result as supporting “the James–Lange theory of emotions” in the abstract’s closing line, which their own valence null does not sustain. Dunn et al. read the same dissociation the other way and say so explicitly: it favours hybrid two-factor accounts, following Frijda (2007), for whom valence is “not bodily at all.” The wiki follows Dunn on this. See core-affect and — for the irony that the two-factor lineage reached the right prediction without thinking about bodies — dror-2017-two-factors.