Contribution
A reconciliation of two ideas of valence that have “largely lived side by side” (p. 1): appraisal theory’s several valences and the one-dimensional valence of core affect and common-currency theories of choice. Both are kept, at different levels, and the paper proposes hypotheses for each path between them.
Key insight
Valence is not one thing. A food can be pleasant, bad for one’s diet, congruent with one’s identity and morally neutral at the same moment. Each is a separate valenced appraisal. Choice still needs one net value, and that is a second, integrated valence. Which integration rule produces it is an open empirical question.
Assumptions
- The component process model’s checks and their order. Pleasantness is a relevance check, run early; goal conduciveness is an implications check, run later (p. 2, citing Grandjean and Scherer 2008).
- That “there is a logical need for the integration of complex affective experiences into a common currency to compare, rank, and choose between options” (p. 4).
- That integration rules from behavioural decision theory transfer to affect.
Key results
A framework, not results:
- Five micro-valences (pp. 2–3): pleasantness/beauty, goal conduciveness, power, self-congruence, moral goodness. Each is bivariate in itself (“more or less unpleasant and more or less pleasant”).
- Pleasantness versus goal conduciveness (p. 2). “The degree of goal conduciveness is directly determined by the degree of need satisfaction or goal attainment. In contrast, needs may set thresholds for pleasantness, but cannot predict pleasantness beyond those thresholds.” The evidence is Berridge (2003), consumer research separating hedonic from utilitarian attitudes, and Litt et al. (2010): “individuals strive harder to acquire consumer goods they like less”.
- Integration rules (pp. 8–9): weighted sum (all information, with trade-offs), and tallying, lexicographic, elimination-by-aspects and one-clever-cue rules (some information ignored, “no trade-offs between micro-valences are required”). The choice of rule depends on accuracy versus effort, on certainty, and on ecological fit.
- Instability (p. 9). “The integration of strongly positive and negative micro-valences may be unstable because it is experienced as a conflict”.
- Paths over time (pp. 9–12). Micro-valence to macro-valence; macro-valence back to micro-valence by misattribution; micro to micro by facilitation, sequence constraints, carry-over and opponent processes; micro-valences to emotions as attractors. No path from emotions to macro-valence: an emotion’s valence comes from appraising it (p. 12).
- Conscious and unconscious macro-valence can conflict (p. 12): the same micro-valences can be integrated differently at the two levels.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Several appraisal outcomes are valenced, not only pleasantness and goal conduciveness | moderate | argued from the CPM; illustrated |
| C2 | Intrinsic pleasantness and goal conduciveness are dissociable valences | moderate | cited: Berridge 2003, Voss et al. 2003, Litt et al. 2010, Booth et al. 2010 |
| C3 | Choice requires a one-dimensional macro-valence | moderate (argued) | Cabanac’s trade-off experiments, cited |
| C4 | Micro-valences are integrated by rules that may or may not allow trade-offs | weak | hypothesis, by analogy with decision research |
| C5 | Emotions have no fixed valence; it comes from appraising them | weak | argued |
Concepts
- micro-valence — the valenced outcome of one appraisal check; it has an object and can coexist with opposed micro-valences.
- macro-valence — the one-dimensional net valence used for choice; always present, possibly objectless.
- intrinsic (un)pleasantness — a “genetically determined association of particular objects with (dis)pleasure”, whose bounds needs can shift.
- goal conduciveness — the “functionality or efficiency of a situation to satisfy needs, achieve goals, or confirm values”.
Connections
Takes its checks from the component process model (LIT-1451, NOTE-1113) and its sequence evidence from Grandjean and Scherer (2008) and Aue et al. (2007). Sets itself beside the evaluative space model (LIT-1438, NOTE-1101), which integrates positive and negative evaluations, where this paper distinguishes qualitatively different ones (p. 13). Cites Berridge (2003) and Berridge and Kringelbach (2008), whose account the record holds as THEORY-893 (read in NOTE-1110), and Rangel et al. (2008) on Pavlovian, habitual and goal-directed valuation. The varieties of valence are argued in Scherer’s 2013 Emotion Review paper (LIT-1464), which this paper does not cite, citing Scherer (2010) instead. Its “arousal, dominance, and novelty” (p. 4) is a gloss on Fontaine et al. (LIT-1437).
Bearing on the record
Does “levels of valence” give a principled reply to CLAIM-438? In part. It gives the record a principled way to concede the objection where it is right. It does not give a way to keep “hedonic” for the regime the record defines.
What it gives:
- The split is the CPM’s own, and grounded in Berridge. Shuman et al. separate intrinsic pleasantness from goal conduciveness on the same evidence CLAIM-438 uses. “The sensual enjoyment of an object and the goal striving toward an object rely on different brain mechanisms (e.g., Berridge, 2003)” (p. 2). Their own example is a wanting–liking dissociation: people “strive harder to acquire consumer goods they like less” (Litt et al. 2010, whose title is “parallel counterdriving of wanting and liking”). So a reply of the form “the potential account covers goal-conduciveness valence, and intrinsic pleasantness is a separate check with its own ordering” is not ad hoc. It is the taxonomy of the theory the record already draws on.
- The record’s ordering matches goal conduciveness. TERM-076’s ordering ranks the system’s states “by their bearing on its maintenance”, and its sign is progress against a reference course. Goal conduciveness is “directly determined by the degree of need satisfaction or goal attainment” (p. 2). Pleasantness is not: needs only “set thresholds” on it. The fit is with goal conduciveness, and the paper explains why the kind clause could not rescue the account in CLAIM-438’s first attempt. Hunger’s ‘wanting’ is about maintenance because it is goal-conduciveness valence, not pleasantness.
- The anhedonia data fit the same line. “Anhedonic individuals do not differ in their reported pleasure from consummatory behaviors (e.g., eating dinner), but report reduced anticipatory pleasure regarding goal-directed activities (e.g., making dinner; Gard et al., 2007)” (p. 14). That is CLAIM-438’s suggested repair, an ordering over consummatory states distinct from one over pursuits, in the paper’s own words.
What it does not give:
- Goal conduciveness is not ‘wanting’. Goal conduciveness appraises how far an event serves a goal; ‘wanting’ is the incentive salience that makes a cue pull. The addict who gets the drug and enjoys it less is high on goal conduciveness for the wanting-driven goal. That is the case where CLAIM-438 says the sign clause gives the wrong answer, and relabelling the valence does not change that.
- It does not make pleasantness potential. A reply that wants to keep “hedonic” needs ‘liking’ to be the potential of a pleasantness ordering. The paper gives pleasantness an ordering, but no account of its winning anything. It appears in the macro-valence only as one weighted input. The paper’s one candidate for a micro-valence that “may generally dominate” is moral goodness (p. 9), not pleasantness.
- Pleasantness fails the kind clause. It is an ordering over stimuli, fixed largely by evolution, with sugar “not a particularly sensitive indicator for the nutritive value of a food” (p. 2). It is not an ordering of the system’s own states by their bearing on its maintenance. A pleasantness potential would need TERM-076’s kind clause rewritten.
So the candid reading is that the paper supports CLAIM-438’s second remedy: revise CLAIM-440 and TERM-076 from hedonic to goal-conduciveness (or motivational) valence, and treat intrinsic pleasantness as a separate check with its own ordering, outside the account. That is the owner’s decision. If taken, the CPM supplies the vocabulary.
The paper’s macro-valence is the record’s common currency. The “integrative ‘common currency’ to compare alternatives for choices” (p. 1) is CLAIM-440’s part 2: the negotiation admitting an ordering of the options. Two things carry over:
- The two regimes are the paper’s integration rules. A weighted sum trades off; lexicographic and elimination-by-aspects rules need “no trade-offs between micro-valences” (p. 8). That is CLAIM-436’s contrast between a currency and an ordering without one, and CLAIM-435’s two regimes of potential. Shuman et al. reached it from decision theory, independently.
- Potential of one micro-valence is the paper’s lexicographic case. “Moral goodness may generally dominate other micro-valences” (p. 9) is a hypothesis that one ordering wins the integration. It is potential in the record’s sense, with priority, though the paper does not test persistence or resumption.
“Perspectives”. The paper’s own word for the micro-valences is perspectives: “we focus on the qualitative difference between perspectives with which the same object may be approached (e.g., pleasantness or goal conduciveness of food)” (p. 13). That is a precedent for reading appraisal as several perspectives, each with its own ranking. It is weaker than the record’s reading: the paper’s perspectives are criteria of one evaluator with weights, not parts of an organization with authority. But the instability of integrating opposed micro-valences (p. 9), and the evaluation that “can suppress competing evaluations” (p. 10), are the negotiation’s conflicts in the paper’s terms.
On ambivalence (CLAIM-440, consequence 4). The paper’s conscious–unconscious conflict, where the same micro-valences yield different macro-valences at two levels of processing (p. 12), is a version of the record’s split potential: one ordering prevailing in one part of the system and another elsewhere. It also gives the record’s marker for it: “non-verbally avoiding and verbally approaching the partner” (p. 12).
For CLAIM-439. The paper says integration at any level “may occur consciously or unconsciously” (p. 7) and that the number of micro-valences integrated and how consciously “are independent issues” (p. 9). It has no account of when valence is felt, so it neither supports nor rivals the self-model condition.
Limitations
A framework with many hypotheses and no new data. The integration rules are borrowed from choice research and none is tested on affect. The two figures are schematic. Its use of Berridge is a citation, not an analysis; it does not ask whether goal conduciveness is incentive salience.
Open questions
Which integration rule governs macro-valence, and whether it changes with state (the paper predicts more effortful integration under negative macro-valence, p. 12). Whether a pleasantness ordering can be potential, or only weighted. Whether the physiological signatures that separate intrinsic valence from goal congruence (facial EMG versus autonomic measures, reviewed in NOTE-1106) could serve QUESTION-268 as markers of which ordering prevails.