Article body

I

Chapter 1 of Christine Tappolet’s Philosophy of Emotion: A Contemporary Introduction begins with a rich survey of the influences of the ancient and modern periods on contemporary theorizing about emotions. After that, we turn to the three central questions that guide the rest of the book:

a) The essence question: What is the essence of emotions?

b) The normative question: Are emotions good or bad for us?

c) The regulation question: Can we regulate our emotions, and if so, how?

Tappolet, 2023, p. 13

Other commentators in this symposium focus on these three questions as they appear in the remaining eleven chapters. In this article I’ll make some brief remarks about the essence question and the structure of part 2. There are three competing answers to the essence question: emotions are essentially feelings; they are essentially motivations; or they are essentially evaluations (e.g., with some sort of cognitive structure suggested by terms like “judgment,” “thought,” “appraisal,” “evaluative representation,” or “perception/perceptual experiences”). Keen on winning the battle of theory construction, each theory presents arguments in favour of thinking that feelings, motivations, or evaluations are the essential core of emotional phenomena.[1] From here, there are subsequent attempts at strengthening each theory by avoiding apparent counterexamples (e.g., unconscious emotions, emotions such as admiration that need not be connected to motivation, and recalcitrant emotions that conflict with the agent’s evaluative judgments or beliefs). Alternatively, other philosophers of emotion defend “hybrid” theories that posit several essential components. According to Andrea Scarantino (2014), for example, emotions are states that combine motivational and evaluative components.[2]

Notice, though, that many of these views defend positive answers to the question of whether emotions have an essence. These views assume from the start that there is some essence, or some indispensable core feature, of all emotions. Thus, what could be added to part 2 is a taxonomy of arguments for skeptical answers to the essence question, according to which there is no essence of emotions. Taking a cue from Wittgenstein, as Jon Elster (1999) does, one might develop such a skeptical view by glossing the concept of emotion as a “family resemblance” concept (see Tappolet, 2023, p. 62). To the best of my knowledge, an alternative skeptical view was first proposed in the work of Amélie Oksenberg Rorty (1980b; 1985; 2004) and Paul Griffiths (1990; 2001; 2004). More specifically, Rorty and Griffiths argue that the generic category of the emotions doesn’t designate a natural kind.[3] Lisa Feldman Barrett (2006; 2017) and Andrea Scarantino (2012) also defend similar views with respect to specific emotion categories, like fear and anger.

This gives us several theories that could be compared with the three views mentioned above that defend non-skeptical answers to the essence question. Section II of my commentary summarizes some noteworthy aspects of Rorty’s argument. I’ll conclude, in section III, by mentioning some nuances that shape this ongoing debate about the essence of emotions.

II

Rorty’s skeptical answer to the essence question can be summarized by the following slogan: “Notoriously, emotions do not form a natural kind [or class] distinguished from motives, moods, and attitudes” (Rorty, 2004, p. 269). In her article “Enough Already with ‘Theories of the Emotions,’” she defends this view by pointing out two similarities between theories of emotion that give non-skeptical answers to the essence question (see also Rorty 1980b; 1985). When presented with counterexamples for thinking that, say, feelings, motivation, or evaluations are essential to emotions, these theories turn to what Rorty calls the tactics of species qualification and gerrymandering emotions. Both of these tactics, she argues, turn out to undermine the explanatory power of theories that each try to identify some essential or core feature of all emotions.

Species qualification involves a series of heel-digging responses from advocates of non-skeptical answers to the essence question. Consider, for example, how an evaluative theory like judgmentalism might appeal to species qualification:

“Emotions are evaluative judgments.”

“What sort of evaluative judgments? What about stock market evaluations or evaluations of the state of the climate? The evaluative judgments of realtors, art dealers, or fine food and wine connoisseurs? Are they emotions?”

“Well, no. Emotions are a species of erroneous or incomplete evaluative judgments.”

“So ‘Socrates was a vulgar, ugly layabout’ is an emotion?”

“Emotion-judgments are a species of incomplete evaluations that are presumptively motivating.”

“So motivating desires that embed incomplete evaluations—for instance, ‘I want that juicy red apple’—are emotions?”

“Well, desire-emotions are accompanied by feelings of a certain sort.”

Rorty, 2004, p. 272

From here, the thought goes, there are further qualifications made that water down the judgmentalist’s initial claim that emotions are essentially evaluative judgments.[4] As this dialogue emphasizes, a wide range of psychological attitudes are reintroduced into a theory meant to give pride of place to only one of them.

As Rorty points out, this tactic is familiar from other areas of philosophy. The explanatory power of theories of imagination, perception, or the will, for example, don’t merely depend on how well each theory fits with folk psychology and folk speech. Their explanatory power also depends on their place in a more complete picture or theory of mental functioning. As Rorty says, the “meaning and import—the claims—of the views of Aristotle on pathe, Seneca on ira and passio, Spinoza on affectus, Hume on the passions, Rousseau on sentiment, Sartre on emotion are deeply embedded in their metaphysics and philosophy of mind, on the force of their distinctions between activity and passivity, their theories of the essential or individuating properties of persons” (Rorty, 2004, p. 270). To properly assess the explanatory power and plausibility of claims about passion, feeling, affect, sentiment, and emotion, then, we need a more complex story about how they may cohere with, or become opposed to, other psychological activities like sensation, perception, imagination, belief, desire, choice, and so on.

The second tactic Rorty identifies consists in gerrymandering emotions, or classifying varieties of them among several other distinctions in philosophy and psychology. Gerrymandering, as Rorty defines it, involves placing emotions within a larger framework of distinctions between (1) mental states and mental activities, (2) between cognition and motivation, (3) between perceptions and proprioception, (4) between voluntary and nonvoluntary states, (5) between physical and psychological conditions, or even (6) between psychological states primarily explained by physical processes and psychological states neither reducible to nor adequately explained by physical processes.[5] Different theories of emotions, having been gerrymandered in one or more of these ways, are then evaluated on the basis of their elegance and simplicity, as well as their ability to account for different emotional phenomena while avoiding counterexamples.

There are several issues with gerrymandering. I’ll mention just two. First and foremost, emotions seem to resist being neatly categorized one way or another with these parameters (see Rorty, 2004, pp. 272-275). As Rorty puts it, “[what] makes the placing of emotions along a schema of these various parameters problematic is that the meanings and force of these dichotomies have themselves shifted” (Rorty, 1980a, p. 2). One example she mentions is the contentious history of the distinction between active and passive responses. There is a substantive debate about the relation between, on the one hand, active and passive actions and, on the other hand, actions that are voluntary as opposed to involuntary (see Frankfurt, 1976).

Second, and relatedly, emotions involve an incredibly diverse set of phenomena exhibiting opposing characteristics within the larger framework of distinctions. Some bouts of love are active, while other bouts of love are passive. Moreover, emotions can be described more generally as being more or less voluntary, or completely involuntary. On the one hand, bouts of anger can be so involuntary they render irrelevant the question of whether the person chose to have them or not. That question is, as Rorty observes, “a question with such a tangle of counterfactual hypotheses that we need to reconstruct the person’s entire history and constitution to make sense of it” (Rorty, 1985, p. 346). On the other hand, a person’s susceptibility toward certain emotions, like anger, is sometimes what we blame them for.[6] This would be the case if that emotional susceptibility is something that arises out of policies the person adopts voluntarily.[7]

The lesson here is that the general category of emotions complicates these commonplace distinctions in philosophy and psychology. Theories of emotion, when they are gerrymandered, must be assessed for “their relative elegance, simplicity, richness, and completeness in encompassing what are (currently) taken to be the relevant phenomena of the field, as well as for the plausibility of their ingenuity in absorbing or excluding objections and counterexamples” (Rorty, 2004, p. 273).

III

To conclude, let’s return to theories that defend non-skeptical answers to the essence question. In her discussion of “theories of emotions,” Rorty wishes to draw attention to the following observations:

In short: theories of the “emotions” (1) do not “cut at the joints”: their subject matter encompasses a heterogeneous set of attitudes, not sharply distinguished from motives, moods, propositional attitudes; (2) are comprehensible only within a larger frame of a relatively complete philosophy of mind/philosophical psychology.

Specific “emotional” attitudes are individuated and identified (1) within a nexus of supportive and opposed attitudes that are characteristically (2) within the context of a narrative scenario. (3) A culture’s repertoire of “emotions” is structured by its economic, political, and social arrangements.

Rorty, 2004, p. 278

My commentary has focused on Rorty’s argument, but the arguments developed by Griffiths, Elster, Barrett, and others are worth engaging with as well.

Several outstanding issues shape this ongoing debate about the essence question. First, Robert C. Roberts (1988) has challenged Rorty’s claim that emotions don’t form natural a kind.[8] Second, setting aside any critical responses, we should note how challenging it is to successfully defend this negative claim. As Rorty observes, “this is not the sort of claim that can be demonstrated: at best it can be grounded in detailed discussions of the problems that arise in identifying, explaining, characterizing those various conditions that are commonly classified as emotions” (Rorty, 1980a, p. 3). Third, Rorty specifically tries to argue for the impossibility of capturing the essence of emotions by treating them as a natural kind.[9] However, one might agree with Rorty that emotions aren’t a natural kind, but still try to give a positive answer to the question of whether emotions have an essence.[10] Indeed, several philosophers have defended the claim that emotions are sui generis mental states irreducible to other kinds of mental states.[11]

Nonetheless, I think Rorty does a remarkable job of identifying problems for theories of emotion in the business of giving non-skeptical answers to the essence question. A survey of the various skeptical perspectives on the essence question and extant critical responses would be a valuable addition to part 2.[12]