Source
- von Scheve & Ismer (2013), LIT-696, read in full in NOTE-552, from the authors’ final draft. Page numbers are the draft’s.
- Barsade (2002), LIT-706, read in full in NOTE-551, from the August 2001 revise-and-resubmit manuscript. Every number below is the manuscript’s, not the published article’s.
What the account says
The definition. von Scheve and Ismer, LIT-696, propose “an understanding of collective emotions as the synchronous convergence in affective responding across individuals towards a specific event or object” (p. 3, restated p. 19). The definition “does not necessarily presuppose that collective emotions are qualitatively different from individual emotions”, nor that convergence happens face to face, and “in its most basic form … does not even presuppose or require mutual awareness of others’ emotions” (p. 19).
The baseline is shared appraisal. “For collective emotions to emerge, individuals have to appraise an event in similar ways, which in turn requires a minimum of shared appraisal structures or shared concerns” (p. 19). People stuck in one traffic jam, with the same goal, the same limited coping potential and the same belief that the jam will last, “might well simultaneously experience anger or frustration with only very limited mutual awareness”. This is where the account presupposes THEORY-055, the record’s appraisal account, on which an appraisal of an event’s pertinence to one’s concerns elicits the emotion: run that mechanism in many people whose concerns coincide and the same event elicits the same emotion in each. von Scheve and Ismer build from Lazarus & Smith and Scherer, the same tradition, and extend it by social appraisal (Parkinson, Manstead & Fischer).
Two processes amplify the baseline and change its quality (pp. 19–20). One is mutual awareness, through mimicry and contagion in co-presence or through mediated sharing and labelling. The other is membership in a collective, which brings shared beliefs and values, social identity and group-based emotions, norms for feeling and display, and group-based sentiments. The paper places these mechanisms on three levels: social cognition, expressive behaviour and symbolic practice (pp. 12–18).
I-mode and We-mode. “(1) socially shared appraisal structures promote ‘I-mode’ emotional convergence and (2) appraisals founded on joint, collectively intentional states based on the identification with a social collective foster the elicitation of ‘We-mode’ collective emotions” (p. 14), after Tuomela and Salmela. The illustration is a shareholder’s happiness at a dividend against a founder’s happiness when her start-up turns a profit. Contagion in a crowd may in turn create the identity that makes We-mode emotion possible, as the authors suggest for the Arab Spring marches (p. 20), and that transition is “not confined to face-to-face gatherings” (p. 21).
Bottom-up, by method. The framework is built “in the tradition of methodological individualism” (p. 11), as a choice to link disciplines, not as an argument. Nothing in it requires a group to be the subject of the emotion, and nothing in it denies that a group could be one.
What the evidence shows
Convergence has been measured. Barsade’s Study 2, LIT-706, had 113 MBA students in 26 groups do a mixed-motive negotiation with no confederate. Groups were formed at random, and the intraclass correlation of members’ pleasant mood was .00 before the exercise, “indicating absolutely no convergence”, and .24 after it: “24% of the variance in mood was due to being part of the group” (manuscript p. 35). That is von Scheve and Ismer’s definition put into a number a decade before they wrote it. They do not cite this paper.
One route has been tested causally. Barsade’s Study 1 randomly assigned a confederate’s displayed mood and found members’ mood moved with it. That result, and what it shows about group processes, is the contagion account, THEORY-123. This account extends it: contagion is one route to convergence among several, alongside shared appraisal and membership, and the account adds the claim that the routes interact. The extension is a widening of the same mechanism class, not a rival to it.
The rest is argued. The cognitive mechanisms are taken from appraisal theory. The behavioural ones, inferring shared values from similar expressions and “facial dialects” making contagion more effective within groups, are offered with “hardly any evidence yet”. The interaction between the levels is drawn in a figure the draft gives only as a caption. The authors say the key linkages lack evidence (NOTE-552, C3–C5).
What this does not say
- It does not say groups cannot have emotions. The account needs no
group subject; it does not exclude one. Whether a group can itself be in
an emotional state is the question of THEORY-125, Huebner’s
genuinely collective emotion, and the two are compatible. They answer
different questions. This account asks whether, how and why members’
affect converges on one object. That one asks whether a group bears an
emotional state that is not its members’ states in aggregation. Huebner’s
candidate case is a campaign whose members’ emotions diverged (some
hopeful, some angry, some saddened) while, he argues, the campaign was
afraid. So convergence is neither necessary nor sufficient for his
collective emotion, and his collective emotion is neither necessary nor
sufficient for convergence. Both can be true of one group at once. No
rivalsrelation is declared, because neither account excludes what the other asserts. - It does not say convergence and a group’s own emotion are the same kind of thing. List’s taxonomy of collective attitudes, LIT-700, gives the distinction a name, though List himself says only that “there can be aggregate fears and hopes, common fears and hopes, and corporate fears and hopes” (§1). Placed in it, I-mode convergence is an aggregate attitude, a summary of members’ states such as an ICC. Convergence with mutual awareness approaches a common attitude. Neither needs a group agent. A genuinely collective emotion would be a corporate one. The placement is the record’s, not List’s or von Scheve and Ismer’s.
- It does not say the traffic-jam case is a collective phenomenon in any stronger sense than coincidence. The definition admits it by stipulation, and the I-mode/We-mode distinction exists to separate it from the cases that are. Whether We-mode emotions differ in quality is untested.
- It does not settle what converges. Mood, emotion and sentiment are run together across the mechanisms. Barsade measured mood, which has no particular object, so her ICC is convergence of affect but not, strictly, of a response “towards a specific event or object”.
- It does not depend on causal appraisal for all of its routes. Contagion by mimicry and afferent feedback produces mood without an appraisal of the event. Barrett’s constructionist account, THEORY-057, which rivals THEORY-055, would locate the baseline convergence in shared emotion concepts rather than shared appraisals. That reading sits closer to the paper’s “culture and shared knowledge” strand than to its appraisal strand, and the paper does not consider it. If THEORY-055 fell, the baseline clause would need restating; the convergence definition and the contagion route would not.
Connections
- Barsade’s groups were ad hoc and competitive, so the measured convergence is at the I-mode end only. Whether contagion between members who identify as a group differs from contagion between strangers is the test the We-mode clause invites (NOTE-551, open questions).