Contribution
A short reply to DeYoung and Krueger’s cybernetic theory of psychopathology. It agrees that cybernetics grounds such a theory and sees the grounding differently. In passing it gives Carver’s compact statement of what he and Scheier mean by valence.
Key insight
Affect does not report where a system is relative to its goal. It reports how fast the system is getting there, against a reference: “a reference value (at which affective valence would be neutral) … these are all seen in terms of velocity of goal attainment rather than states” (§Confidence, Doubt, and Psychopathology).
Assumptions
- Feedback loops in human action have represented goals, though “not all feedback loops have represented goals” (§Goal Based Behavior).
- Goals are organized hierarchically, after Powers (1973), and many are in play at once.
- There are two broad motivational systems, incentive approach (extraversion) and threat avoidance (neuroticism).
Key results
Argument, not results:
- Problems arise when any of a loop’s four parameters is extreme or biased (§Elements of Feedback Control and Sources of Problems).
- The five factors are sensitivities to classes of goals, not loop parameters. Neuroticism is threat sensitivity, not negative affect.
- Valence comes from how well a system is doing, so each system yields both signs (§Feedback Loops, Goals, and Dimensions of Personality).
- Engagement is high near the criterion and fails when progress is far below it (Figure 3 and text).
- Disengagement and doubt are not predicted by control theory.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Valence reflects the velocity of goal attainment relative to a reference value, not the state reached | moderate | asserted, with citation to Carver & Scheier 1990, 1998, 2013 |
| C2 | The approach and avoidance systems can each produce positive and negative affect | moderate | asserted; anger as approach-related negative affect (Carver & Harmon-Jones 2009) |
| C3 | The five factors reflect sensitivities to classes of goals rather than cybernetic parameters | weak | interpretive argument |
| C4 | Disengagement from a failing goal is not predicted by control theory as such | moderate | argument from the thermostat; footnote 3 offers priority re-ranking as a possible fix |
Concepts
- reference value — the loop’s goal, “the state of affairs the loop is trying to bring into existence”.
- velocity of goal attainment — the quantity affect tracks. Valence is neutral at a reference value of it.
- functionally superordinate — the level of the Powers hierarchy currently in charge, which “can vary across situations and persons”.
- priority ranking — the order of the many goals in play, subject to “continual juggling” (n. 3, citing Carver 2003).
Connections
Responds to DeYoung & Krueger (2018), not in the record. Restates Carver & Scheier 1990 (LIT-1441, Unavailable) and draws on Powers (LIT-1179, Unavailable; his feedback papers LIT-1188 and LIT-1197 are read). Cites Carver 2003, “Pleasure as a sign you can attend to something else”, for priority management. That paper is not in the record.
Bearing on the record
The record’s proposal (CLAIM-440, TERM-076) makes hedonic valence a regime in which one ordering of a system’s own states, by their bearing on its maintenance, prevails in the negotiation among its perspectives. Its sign is the direction of movement along that ordering, and its intensity is the potential’s strength. Carver’s account bears on it in four ways.
- The sign clause is anticipated, and is less accurate than Carver’s. Carver and Scheier already make the sign of affect a function of movement relative to a goal, not of position. The proposal should say the clause is theirs. They also differ from it. Their neutral point is a reference velocity: “a reference value (at which affective valence would be neutral)”. The proposal makes neutral “potential with no movement”. On Carver’s account a system that stands still where it expected to advance has negative valence. On the proposal’s it is neutral. The two give different predictions for an animal that stops gaining. The proposal should either adopt a reference rate, so that sign is the direction of movement relative to an expected movement, or say why rest is neutral.
- Carver has a valence per goal loop. The proposal has one potential ordering. For Carver, each of the approach and avoidance systems “can be responsible for either positive affect or negative affect”, and his Figure 2 has two dimensions crossing at “neutral points on both dimensions”. This is two valence signals, one per system, not one ordering. The proposal can absorb it only by arguing that felt valence is whichever loop’s signal prevails. That is a substantive addition, and Carver does not make it.
- Potential is present only in outline. Carver speaks of competition between “two sources of guidance for behavior”, of which level is “functionally superordinate”, and of “continual juggling of the priorities afforded to each” goal (n. 3). This is the nearest the tradition comes to the proposal’s potential, and it is not tied to valence. Carver ties affect to rate and priority to the hierarchy. The proposal’s identification of intensity with the potential’s strength is not in Carver.
- A possible conflict over intensity, not checked. Footnote 3 cites Carver 2003 for priority juggling. Its title says pleasure is “a sign you can attend to something else”, which suggests that positive affect lowers a goal’s priority. If so, on Carver’s account a strongly positive goal loses precedence, while on the proposal’s, intense positive valence is strong potential. The record has not read Carver 2003. This should be checked before the intensity clause is promoted, since THEORY-055’s evidence that control precedence tracks intensity is also not read here.
Carver’s text does not touch ambivalence, the lexicographic/compensatory distinction (CLAIM-435), or felt valence (THEORY-101).
Limitations
A commentary. It asserts the affect account and cites its sources rather than arguing for it. The figures were not seen. The reference rate is named, but how it is set is not discussed.
Open questions
What sets the reference rate, and whether it is itself a potential ordering’s expectation. Whether the two systems’ valence signals combine into one valence, or are felt separately.