One of the aims of this article is to try to define feelings according to their functional characteristics, when seen as all-purpose comparators. Iuliia Pliushch claims that my use of "affective feelings" is ambiguous, because they seem to be defined either as "feelings that possess valence", or as "feelings that express emotions". I am happy to accept the blame for not rephrasing in my own terms the subcategory of "affective feelings" discussed in emotion theory.
A similar discrepancy, however, may seem to be present between two passages of my chapter where I do express my own view:
As will be seen below, some feelings, however, do not express emotions, i.e., are not affective. (Proust this collection, p. 2)
All the feelings vary in affect in roughly the same way, because they all include valence in their informational structure. (Proust this collection p. 20)
The discrepancy is only apparent, however, and should disappear when the issue of valence in its relation to affect is properly addressed. In emotion theory, the relations between valence and affect, and even the existence of valence, are highly debated. With rare exceptions,[1] the question is ignored by theorists of somatic, agentive, or noetic feelings.[2] The proposal summarized in (2), however, posits that affect will result from valence (not the other way round). Section 7 aims to explain why affect depends on the dynamics of valence throughout the domain of feelings. These relations are modulated by the dynamic conditions that prevail in the contrast between expectancy and observation in a given domain. When observation and expectancy coincide with a predicted temporal pattern - with a small stake involved-, the corresponding feelings should not involve affect on top of valence. This is the case for the feelings of agentive success that are generated in routine actions. Hence (1) holds. When you predictably overcome a minor obstacle, you don't feel particularly thrilled. When special dynamic conditions obtain, however, (acceleration or deceleration in the rate of observed change, as compared with the expected rate of change), valence will be intensely felt, in terms of vividly positive or negative experiences. Scoring an ace in a tennis game, especially if it is a rare achievement for this player, elicits in him/her an intensive positive affect. Dynamic variations of this kind also apply to metacognition, where Archimedes' "Eureka" is affect-laden, while the felt ability to respond to a memory question in a laboratory is not.
Hence there may be affect-laden feelings beyond the domain of what is traditionally called "emotional" or "affective feelings". Reciprocally, one might suspect that in the latter domain, too, affect only appears beyond thresholds of positive or negative valence, with colder kinds of feelings occupying the lower end of the continuum.