Sofia Berinstein

This paper explores the epistemology of a particular dimension of perceptual experience—its affective character: the ‘badness’ of, for example, the smell of garbage or the pain of a stubbed toe; the ‘goodness’ of the taste of chocolate, touch of sunshine, or sound of a musical chord. I take the view that affective character is epistemically significant, putting the perceiver in touch with axiological relations in which elements (garbage, bodily harm, sunshine, chocolate, and consonance) stand to perceivers. Two representationalist approaches to accommodate the affective character of experience—cognitivist and non-cognitivist—are explained and analyzed. Considering the objections they face, I motivate a non-representationalist alternative for capturing the epistemic significance of affective character. Because affective character is ubiquitous across the senses, the view developed suggests that perception is an inherently evaluative capacity and that some empirical judgments based on affectively valenced perceptual appearances are normative judgments.

Anil Gupta’s presentationalism holds that the rational role of experience is hypothetical—it licenses transitions from views to judgments. His is opposed to the conventional view that appearances independently confer some epistemic status on perceptual judgments. Gupta further reconceives of appearances as logical entities that can vary radically from presented elements. In this paper I argue that presentationalism cannot accommodate the affective character of perceptual appearances. It makes an implausible prediction: that faultless painful experiences can relate a perceiver to benefit and faultless pleasant experiences can relate a perceiver to harm. The upshot is that appearances must be more tightly bound to their elements than on Gupta’s account.

Perceptual experiences have an affective character or ‘valence’: coffee tastes good, garbage smells bad, notes have a beautiful tone, and pains hurt. Such experiences seem to have a dual character: they are epistemically significant—disclosing information about the value of their objects to subjects—and they seem to be intrinsically motivating, giving perceivers defeasible reason to pursue or avoid their objects. I argue that what have been taken to be competing considerations on an account of valence—epistemic and motivational significance—are compatible. Further, existing accounts tend to treat them as competing considerations, thus accommodating one of them at the expense of the other. Here I argue that the phenomenon of valence gives us compelling grounds for accommodating both kinds of consideration and offer an account that is capable of doing so, which explains how valence can exhibit both a world-to-mind and mind-to-world direction of fit. It will turn out, I claim, that the difficult choice between cognitivist and non-cognitivist accounts of valence isn’t forced upon us. Finally, I recommend affective character as a topic of special interest to metaethics, because it speaks to questions in the epistemology of value that have long puzzled the field.

What is the nature of our aesthetic sensibilities or tastes? I address this question by examining the relationship between the perceptual and emotional aspects of aesthetic feeling, asking what stands in the way of the intuitive claim that aesthetic sensibilities consist in unified perceptual and emotional responsiveness to objects of experience. On a schematic notion of perception, perception is univocally receptive or representational capacity, reflecting features of the world in a purely indicative, as opposed to evaluative, manner. A counterpart schematic conception of emotion suggests that it is, rather than a discerning faculty, a reflection of how our values, interests and desires meet with the circumstances, or a matter of how the circumstances make us feel. These tropes place perception and emotion at odds with each other with respect to ‘direction of fit’; whereas perception exhibits a belief-like direction of fit—representing the world as it is take it to be—emotion exhibits a desire-like direction of fit, expressing how we wish the world was. I argue against both of these schematic conceptions. First, I provisionally motivate a form of cognitivism about aesthetic emotions—the accuracy conditions of emotions are aesthetic value relations in which objects of experience stand to subjects. Second, drawing on the increasingly popular view that perception is intrinsically valenced, I argue that perception, similarly, is capable of disclosing aesthetic value relations. On such a picture emotional and perceptual aesthetic capacities have a parallel empirical structure, making plausible the claim that their unified operation constitutes aesthetic sensibility.

The puzzle of aesthetic rationality is how non-aesthetic claims or observations given by perception can serve as justificatory reasons for aesthetic judgments. Two opposing standpoints have traditionally received support. The rationalist standpoint holds that, despite appearances to the contrary, aesthetic judgment is a species of rational judgment. The perceptualist instead claims that perception ultimately provides foundational support for aesthetic judgments—it is an ‘unjustified justifier’. This paper is devoted to the question of whether a third kind of response can remain independent of these two options: ’feelings views’. Such views claim that feeling (affect, emotion, sentiment) serve as bases for aesthetic judgments. Here I analyze three increasingly prominent types of feeling view, which have not yet been compared, asking (1) if they can maintain genuine independence of perceptualism and rationalism and (2) if they can avoid the serious objections facing their rivals, thereby offering a more attractive solution to the puzzle.

Are all voluntary actions based on practical inferences? The popular Humean view of motivation requires that beliefs and desires be united by reason to motivate action. I argue that there is a class of voluntary actions that eludes this Humean model: ‘sensual’ actions, such as sipping a coffee, hugging a friend, choosing an outfit, or making the next brushstroke on a painting. Actions like these are more plausibly motivated by direct perceptual apprehensions of value. I claim that actions which are motivated by valenced perceptions need not be based on the sorts of rational inferences required to join beliefs and desires. Rather, the relevant cognitive and motivational elements are already present in valenced experiences like pain, which simultaneously inform the perceiver of what is the case and motivate appropriate forms of response.

Here I examine epistemological consequences of good-for theories of value (GF), which claim that relational value has metaphysical priority over non-relational value. GF may have a compelling response to longstanding questions about disagreement over normative matters: that the objective values over which subjects disagree may themselves vary between subjects. On this picture we have two sorts of disagreement rather than one: weak disagreement, in which parties stand in differing objective axiological relations to the same object (e.g. x is good for me, x is bad for you, you take x to be good and I take x to be bad) and strong disagreement in which parties stand in similar axiological relation to the object (e.g. x is good for me and x is good for you, but I take x to be good and you take x to be bad). Keeping apart such different kinds of disagreement can save us from confused attempts to make sense of both types with a single account. Thus GF may help us to address the tension between (a) prolific disagreement across various forms of evaluative judgment and (b) the apparent objective validity of those judgments.

Experiences—hearing beautiful music or stubbing one’s toe—can feel good or bad. I articulate a way in which such affectively valenced experiences can disclose the value of objects to subjects. My account appeals to a normative epistemic principle I call ‘inheritance’: whether an experience is good or bad for a subject is a function of whether its objects are good or bad for that subject. If experience is governed by inheritance, it suggests that all perceivers are directly aware of value de re through their pleasant and painful experiences and that such experiences motivate intelligent action independently of the intellect’s powers. I argue against representationalist alternatives, which rely on the idea that all epistemically significant features of an experience have truth or accuracy conditions. We can avoid this core cognitivist epistemic principle if we look at value perception as a matter of being positively and negatively affected by the objects of perception, rather than one of ascribing evaluative properties. Such a view allows us to understand the epistemic and motivational significance of pleasurable and painful experiences, and thus how humans and animals alike are capable of moving from perception to action without the aid of cognitive inferences. The normative dimension of perception, understood in terms of its affective character, may be what puts us in immediate moral contact with our community and makes us naturally susceptible to the value of works of art.