Contribution
A perspective article that argues the major theories of emotion are complementary once emotion is seen as multi-componential. It then proposes to organize affective neuroscience around five functional networks, with the appraising brain as the driver. It updates the authors’ earlier syntheses (Sander, Grandjean & Scherer 2005; Grandjean et al. 2008; Sander 2013) and reports no new data.
Key insight
If emotion is a pattern across components, then what needs locating in the brain is not an emotion but each component’s network, and the coupling between them. On this account the coupling is the emotion’s signature.
Assumptions
- That there is “consensus among major theories of emotion that the many facets of emotion can be grouped into five components” (p. 221).
- That appraisal is constitutive of emotion, not only antecedent to it, as encoding is to memory (Figure 1 caption, after Sander 2013).
- That communication through coherence (Fries 2005) is the right model of how distributed networks share information.
Key results
A framework, not results:
- Table 1 maps theory families to levels of explanation and candidate neural mechanisms (basic, dimensional, appraisal, adaptational, constructivist).
- Synchronization is split into within-component synchronization (neuronal populations serving one component) and between-component synchronization (the distant networks of different components) (p. 222). “Synchronization of neuronal activity can take place at different levels including close neuronal assemblies interacting at high frequencies, and more distant regions synchronized at lower frequencies (see Grandjean et al., 2008)” (p. 222).
- Each network is reviewed. Expression has three levels (reflex, habit, voluntary) that can inhibit one another through the subthalamic hyperdirect pathway (p. 223). The action tendency network includes the approach–withdrawal asymmetry, and the feeling network the insula, somatosensory cortex and vmPFC.
- Sequential appraisal is supported by EEG studies of novelty, pleasantness, goal conduciveness and coping (p. 227).
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Emotion has five components, on which major theories agree | moderate | quotations from Scherer, Russell, Matsumoto & Ekman |
| C2 | The emotional brain consists of five functionally defined networks | weak | proposed; “brain data are still missing” (p. 228) |
| C3 | Component synchronization rests on neuronal phase or cross-frequency synchronization | weak | proposed by analogy with Fries (2005) |
| C4 | Feeling emerges when synchronization passes a threshold | weak | attributed to Scherer (2009b) and Dan Glauser & Scherer (2008), not tested here |
| C5 | Appraisal checks unfold sequentially | moderate | EEG, EMG and autonomic studies cited (p. 227) |
Concepts
- within- and between-component synchronization: coupling inside one component’s network, and coupling across the networks of different components.
- elicitation network: the appraising brain, centred on the amygdala for relevance detection.
- control precedence: the priority some action tendencies take over others during an episode (Frijda).
Connections
A restatement for neuroscience of the CPM that Scherer (2009) gave for modellers (LIT-1451, NOTE-1113). Its five networks are what Mohammadi, Van De Ville and Vuilleumier try to recover from data (LIT-1444, NOTE-1107), and their latent variables do not line up one per component. Takes control precedence from Frijda (THEORY-055).
Bearing on the record
A per-episode measure of synchronization? No. The paper says what synchronization should be in the brain: phase synchrony within a band, or one region’s theta phase organizing another’s gamma bursts. It gives no index, no data and no procedure, and it closes by calling the architecture “and how they interact and synchronize” “a major challenge” (p. 228). It names the methods that would be needed, “excellent resolutions at the spatial and temporal scales” such as intracranial recording (p. 225). For QUESTION-268 that is a sharper lead than the 2009 paper gave: a phase-locking or cross-frequency index computed over an episode’s window is a per-episode quantity. But it would still be unsigned. It would say how coupled the networks are, not which of them is in charge.
Is it signed, or does it say which tendency prevails? The coupling is not signed. What does say which tendency prevails is a different sentence: “some action tendencies with control precedence take priority over other potential action tendencies” (p. 223). That is the record’s potential (TERM-076) in the CPM’s own words: a tendency winning the conflict with its rivals. The paper does not connect it to synchronization or measure it. In the CPM, then, potential and synchronization are separate notions. Synchronization is a degree of coherence across components, and precedence is a ranking among the tendencies within one of them. That supports the reading in NOTE-1113 that synchronization is at most the analogue of the term’s intensity clause, and it adds that the CPM has its own name for the ordering clause, without a measure for it.
Threshold for feeling. On Scherer’s suggestion, “feeling is dependent on the processes’ degree of synchronization in the other components surpassing a certain threshold” (p. 225). This is a threshold on an unsigned degree, so it is a hypothesis about when an episode is felt, not about how it is resolved. It sits beside CLAIM-439, felt when the system models its own potential, as a rival with a different variable: degree of coupling rather than a self-model. It is the same identity-with-degree view that NOTE-1113 set against the self-model reading (closer to THEORY-101). It also does not bear on CLAIM-435’s continuity question, which is about trade-offs in choice, not about entry into consciousness.
Reciprocal inhibition in expression. The three expression levels “can be subject to reciprocal inhibition”, with the subthalamic hyperdirect pathway inhibiting prepotent responses “to shift into a more controlled behavior” (p. 223). This is a mechanism for one response overriding another, the override regime in CLAIM-435, but here it works inside one component. It is mentioned in passing.
Candidly, the paper is a map of where to look. It does not help the record measure anything.
Limitations
Programmatic and largely a literature synthesis. The synchronization proposal is not operationalized, and its key citations (Grandjean et al. 2008; Dan Glauser & Scherer 2008) were not read here. The claim of consensus on five components is argued from quotations, and the constructivist and basic-emotion authors quoted might not accept that their “components” are the CPM’s.
Open questions
Whether Grandjean, Sander and Scherer (2008, Consciousness and Cognition 17:484–495) or Dan Glauser and Scherer (2008, Brain Topography 20:224–231) compute a synchronization index per episode, and whether either says which component leads. Whether control precedence, as a ranking of action tendencies, has ever been measured in the CPM programme.