Contribution
The paper offers a working definition of collective emotion and a vocabulary of mechanisms that lets three separate literatures be read as accounts of one phenomenon. The literatures are face-to-face contagion and effervescence, culture and shared knowledge, and social identification. What it adds is the decomposition into bottom-up mechanisms on three levels, and the claim that contagion and shared appraisal interact. Contagion makes shared appraisals visible, and that visibility can create a group identity that did not exist before.
Key insight
People can share an emotion without any group process at all. If they share concerns and beliefs, the same event produces the same appraisal in each of them, and so the same emotion, as with drivers stuck in one traffic jam. Everything social is then an amplifier or a transformer of that baseline. Seeing others respond the same way, or catching their expressions, strengthens the convergence. It also licenses an inference that they share one’s values, and that inference can turn a crowd of separately aggrieved people into a “we” whose anger is about what is happening to us.
Assumptions
- Methodological individualism (p. 11). The framework is built bottom-up from individual appraisal and behaviour, so that disciplines can meet on common ground. It is a methodological choice; the paper does not argue against group-level subjects.
- Appraisal theory for individual emotion elicitation (Lazarus & Smith 1988; Scherer), extended by social appraisal (Parkinson; Manstead & Fischer).
- Convergence, not identity, is the criterion. Collective emotions need not be “qualitatively different from individual emotions” (p. 19), though the authors expect We-mode ones to be.
- Affect is used broadly. Emotions, sentiments and moods all appear; group-based sentiments are explicitly “mood-like affective dispositions” (p. 17).
Key results
A conceptual paper. Its results are claims about how mechanisms relate.
- Review, part 1: three perspectives (pp. 5–11).
- Face-to-face: Hatfield, Cacioppo & Rapson (1992) define contagion as the “tendency to automatically mimic and synchronize facial expressions, vocalizations, postures, and movements with those of another person and, consequently, to converge emotionally”. Durkheim (1912) and Collins (2004) supply the functions of effervescence. Contagion studies are “somewhat mute” on cultural embeddedness, and Durkheimian sociology leaves the behavioural pathways unexplored.
- Culture and shared knowledge: Parkinson, Fischer & Manstead (2005) list five sources of group emotion, namely identical eliciting events, mutual influence on appraisals, shared values and norms, identification, and emotional behaviour constitutive of membership. Also Bar-Tal’s collective emotional orientation, and de Rivera’s emotional atmosphere, climate and culture.
- Identification: group-based emotions are felt “on behalf of” a group (Smith 1993), are elicited by events concerning one’s social identity (Kessler & Hollbach 2005), and can be felt in solitude, as in collective guilt.
- Cognitive mechanisms (pp. 12–14). (1) Socially shared appraisal structures promote I-mode convergence. (2) Appraisals founded on joint, collectively intentional states promote We-mode collective emotions. Aggregate accounts treat collective intention as a sum, “much like Barsade and Gibson (1998) define group emotion as the sum of its parts”. Non-aggregate accounts locate it in intentions held as a member.
- Behavioural mechanisms (pp. 14–16). (1) Perceived similar expressions promote inferences of similar appraisals and values, which amplify I-mode and possibly We-mode emotion. (2) “Facial dialects” (Elfenbein et al. 2007) make contagion more effective within groups than across them. Both are offered as plausible, with “hardly any evidence yet”.
- Symbolic mechanisms (pp. 16–18). Four: collective memory formed through discourse; group-based sentiments, which are enduring dispositions that, following Frijda (1994), carry appraisal dispositions favouring valence-congruent emotions; social norms, both as shared appraisal inputs and as feeling and display rules (Hochschild 1979); and cultural practices of sharing and representing emotion (Rimé 2009; Tsai’s “ideal affect”).
- Synthesis (pp. 18–21). The baseline (shared appraisal) is modified by two intertwined processes: mutual awareness and membership in a collective. Contagion in a crowd can create a social identity, as in protest movements. The I-mode to We-mode transition is not confined to co-presence; communication and representation can do it too, as with “imagined communities”.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Collective emotion can be defined as synchronous convergence in affective responding towards an event or object, with no requirement of mutual awareness | definitional; argued from cases (the traffic jam) | pp. 3, 19 |
| C2 | Contagion, shared knowledge and social identification are complementary routes to convergence, not rival accounts | moderate: a review argument, with overlap shown case by case | pp. 5–11 |
| C3 | Shared appraisal structures produce I-mode convergence; collectively intentional appraisals produce We-mode collective emotions, which differ in quality | weak to moderate: conceptual, relying on Tuomela and Salmela; no empirical test | pp. 12–14 |
| C4 | Contagion is more effective within groups than across them, because expression and recognition follow cultural “facial dialects” | weak: an inference from dialect research; the authors say evidence is lacking | pp. 15–16 |
| C5 | Contagion in co-presence can create group identity and so convert I-mode into We-mode emotion | weak: illustrated by protests (London 2011, the Arab Spring), not tested | pp. 15, 20 |
Concepts
- collective emotion: synchronous convergence in affective responding across individuals towards a specific event or object.
- I-mode / We-mode collective emotion: emotion from shared but private concerns, versus emotion from concerns held as a member of a group (after Tuomela 2006 and Salmela 2012).
- group-based emotion: an emotion felt on behalf of a group, elicited by events bearing on one’s social identity; possible in solitude.
- group-based sentiment: an enduring, mood-like affective disposition towards an in- or out-group (belonging, solidarity, hostility, resentment) that predisposes valence-congruent emotions.
- facial dialects: culture-specific variation in how emotions are expressed and recognised.
- social calibration: adaptation of emotional expression to a group’s cultural environment (von Scheve 2012).
Connections
- Barsade 2002 (LIT-706). Not cited. Barsade’s experiment is a causal test of the face-to-face mechanism in work groups. Both use Hatfield et al.’s account of mimicry and feedback. Barsade’s Study 2 operationalises group contagion as mood convergence, measured by the ICC rising from .00 to .24, which is this paper’s definition operationalised.
- Sanchez-Burks & Huy 2009 (LIT-699), unread. Their emotional aperture is recognising the distribution of emotions in a collective. It is the perceiver’s half of the “mutual awareness” process here.
- Collective intentionality. Gilbert is cited by the 1990 paper the record holds (LIT-477). Bratman is cited by “Shared Intention” (1993), which the record does not hold; its nearest holding is “Shared Cooperative Activity” (LIT-471). The paper uses them only to separate aggregate from non-aggregate collective intention. It does not choose between Gilbert’s plural subject and Bratman’s reductive account.
- Appraisal theory. The mechanism presupposes appraisal-caused emotion (Lazarus LIT-550; Frijda LIT-540, THEORY-055). Barrett (LIT-542, THEORY-057) rejects causal appraisal. Her account would locate convergence in shared emotion concepts and the social reality they make, which is nearer this paper’s “culture and shared knowledge” strand than its appraisal strand.
Bearing on the record
- It is the natural source for a THEORY on what a collective emotion is. Such a theory would hold that collective emotion is convergence produced bottom-up, rather than a state of a group subject. Its rival would be a genuinely-collective account: Huebner’s “Genuinely collective emotions” (LIT-702), which the paper cites, and his book (LIT-426). A candidate, not filed.
- Bears on
social-ontologyby making emotions one of the states the group-agency debate must handle. The paper takes no side on whether groups are subjects. - No ML instruction. Nothing here belongs in the anthology.
Limitations
- No data. It is a framework paper; the authors flag the behavioural linkages as lacking evidence.
- The definition admits too much by its own admission. Convergence without awareness (the traffic jam) counts as collective emotion. Whether that is a collective phenomenon or only a coincidence of individual ones is the question the I-mode/We-mode distinction is introduced to handle, and it is not settled.
- Mood, sentiment and emotion are run together across the mechanisms. That suits the convergence definition but blurs what is converging.
- Interaction between the levels is asserted and drawn in Figure 1, but not modelled.
Open questions
- Are We-mode collective emotions measurably different from I-mode ones in experience, expression or consequence?
- Is contagion more effective within groups than across them, and is that due to expressive dialects or to identification?
- Can convergence be shown to produce group identity, as the protest examples suggest, rather than only to follow from it?
Corrections
- none to a seeded skim (there was no seed)
- Citation. The draft cites Gilbert’s “Walking together” as Midwest Studies in Philosophy 25; the record’s entry (LIT-477) gives volume 15, which is correct for 1990.