Contribution
A short statement of the component process model (CPM), written to recommend it as the blueprint for computational models of emotion over basic-emotion and constructivist theories. It ends by calling for nonlinear dynamic models. The evidence is deferred to the 2009 review (LIT-1446, cited as “in press a”).
Key insight
An emotion is not an output of appraisal. It is a state of the whole organism, many subsystems pulled into synchrony by appraisal results, and its existence is a matter of degree of coherence (§1, §2c).
Assumptions
- Four uncontroversial features of emotion, after Frijda and Scherer (2009): elicitation by relevant events, action readiness, involvement of several components that “tend to cohere”, and control precedence (§1).
- That the order of appraisal checks is fixed by “systems economy and logical dependencies”, each check needing its predecessor’s result (§3c).
- That subsystems behave as coupled oscillators, so that synchronization is the right description of their coherence (§4).
Key results
A model, not results:
- Components: appraisal (cognitive), autonomic (ANS), somatic motor (SNS), motivation (action tendencies), and feeling (§3, §3e).
- SECs: novelty, intrinsic pleasantness, goal conduciveness and their successors, grouped under the four objectives, run in parallel but closing in sequence. The checks are recursive, repeated “until the monitoring subsystem signals termination of or adjustment to the stimulation that originally elicited the appraisal episode” (§3c). They are integrated “through context-dependent weighting functions” (§4).
- Patterning: each check’s result drives efferent changes in the other components, “differentially and cumulatively” (§3d).
- Feeling: integrates a representation of all components. “Quality, intensity and duration of the feeling are determined by these integrated inputs” (§3e).
- Dynamics: attractor basins, hysteresis and a cusp catastrophe. “A hysteresis function predicts that the intensity of anger will change abruptly for specific degrees of frustration, whereas a linear function assumes continuously rising anger with increasing frustration” (§4).
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Emotion is an episode of component synchronization aimed at control precedence | moderate (definitional) | the four features of Frijda & Scherer 2009 |
| C2 | Appraisal checks reach closure in a fixed order | moderate | argued from logical dependency; evidence deferred to the review |
| C3 | Feeling monitors the process by representing all components and their coherence | weak | argued from function |
| C4 | The degree of synchronization generates conscious experience | weak | stated as a hypothesis |
| C5 | Emotion dynamics are nonlinear (attractors, hysteresis) | weak | illustrated, not tested |
Concepts
- stimulus evaluation check (SEC) — one appraisal criterion, run at sensory-motor, schematic or conceptual level.
- component synchronization — the coherence of subsystem changes in an episode, the criterion of an emotion’s existence.
- monitoring — the feeling component’s integrated representation of the other components and their coherence.
- control precedence — Frijda’s: “claiming (not always successfully) priority in the control of behaviour and experience”.
Connections
Takes action readiness and control precedence from Frijda (THEORY-055). Cites Lewis (1996) on circular causality. Lewis (2005, LIT-1455) builds an attractor model of the same kind, and the two exchanged sharply in its BBS commentary, where Lewis replied that “Scherer’s model and mine share an emphasis on synchronization in emotion episodes” (p. 225).
Bearing on the record
Is synchronization a measurable analogue of potential? It is a correlate potential would produce, not potential itself. The record’s term (TERM-076) defines potential by which ranking wins the negotiation’s conflicts, persists and resumes. Scherer’s synchronization is coherence among response subsystems, the same changes appearing in appraisal, body, face and motive. One ranking prevailing many perspectives would produce that coherence, since all the perspectives it governs would be pulled the same way. But coherence does not say which ranking does the pulling, or whether it is a ranking of the system’s own states, or what sign the state has. Synchronization is unsigned, like the potential’s strength, so at best it is an analogue of the term’s intensity clause. Scherer himself has the integrated inputs fix the feeling’s quality, intensity and duration, not synchrony alone.
Two points tie the CPM closer to the record than this suggests:
- The definition already contains control precedence, glossed as “claiming (not always successfully) priority”. That parenthesis is CLAIM-283’s distinction between the precedence a motive claims and the authority it is granted.
- The integration rule is a negotiation. “Weighted integration in the form of dynamically adaptive differential activation and inhibition thresholds are likely to be more appropriate modelling principles as compared with fixed weights in regression or decision structures” (§4). Context-dependent weights are compensatory; activation and inhibition thresholds are not. The CPM does not choose between CLAIM-435’s two regimes. It expects both.
An operational measure? None here. The paper says “all we can currently measure is the individual’s verbal account of a consciously experienced feeling” (§3e), and in their commentary on Lewis, Sander and Scherer concede that synchronization models “have not progressed beyond a preliminary stage” and that nonlinear theorizing “does not lend itself readily to designing appraisal experiments and analyzing multimodal data” (Lewis 2005, p. 216). Whatever indices the 2009 review uses, such as cross-component correlation or coherence of time series, are its business, and it is unread. For QUESTION-268 this is a lead, not an answer.
Is the monitoring component a version of “felt = the system’s model of its own dominance”? In part. Scherer makes two separate claims, and the record should keep them apart.
- (A) Feeling represents the coherence. The feeling component “needs to integrate and centrally represent all information about the continuous patterns of change and their coherence in all other components.” That is close to the proposal. Feeling is a model, held by one subsystem, of the coordination of the others. It differs in two ways. What is modelled is coherence among components, not which ordering wins. And the model covers everything, the appraisal profile, the body and the action tendencies, not the potential alone.
- (B) Synchronization generates consciousness. “It is hypothesized that it is the degree of synchronization of the components … that will generate conscious experience.” Here consciousness is tied to the degree of the regime itself, not to a model of it. That is closer to Veit’s identity (THEORY-101) than to a self-model view. It also faces the counterfactual eraser (THEORY-252) if degree is read dispositionally.
Scherer’s Venn model puts the monitoring representation (circle A) only partly inside consciousness (circle B). So on his account a system can model its own coherence without that model being felt. If the record adopts “felt = model of own potential”, Scherer is a precedent for the model and a warning that the model is not sufficient. That is QUESTION-059’s gap.
On CLAIM-435. Scherer’s hysteresis is a third pattern beside the continuous curve and the threshold, and the claim does not place it. If anger jumps at one frustration level going up and drops at a lower one coming down, choice depends on the path, not only on the current cost–resource pair. The relation on pairs is then not well defined. A trade-off experiment should sweep cost up and down, and a fold would be evidence of potential changing regime with history. The claim’s second defeater should name it alongside stochastic switching.
Limitations
Programmatic. The dynamics are illustrated with figures, not fitted; the evidence is elsewhere; the components and checks are specified, but the transfer functions that would make synchronization computable are not.
Open questions
What synchronization index the 2009 review uses, and whether it can be computed per episode. Whether the feeling component’s representation of coherence is a ranking-sensitive model (which subsystem leads), or only a degree.