Article body

Christine Tappolet’s Philosophy of Emotion is an admirable and important contribution to the philosophy of emotion. Among its achievements are its exceptional conciseness and theoretical richness. Tappolet guides us through a history of scientific and philosophical debates about emotion with excursions into practical rationality, epistemology, metaethics, philosophy of mind, and more. All along the way, we get glimpses into Tappolet’s own original thinking and her overall view of how all these issues fit together into a coherent and compelling theory of emotion and value.

Tappolet argues that the essence of emotions is constituted by representations of things as having specific evaluative features (ch. 6). But now we need an account of what evaluative features are. In particular, we need an account that is compatible with evaluative theories of emotions. Providing such an account is the main purpose of chapter 9. Tappolet defends a version of sentimentalism. The core claim of this group of views is that evaluative features depend in some way on emotions (or other reactions).

Tappolet plumps for the following:

Nonreductive Neo-sentimentalism: x is E-worthy (i.e., worthy of emotion E) if and only if emotion E is fitting in response to x, and E is fitting in response to x because x is E-worthy.

As Tappolet notes (2023, p. 167), nonreductive neo-sentimentalism is circular. But she is sympathetic to sensibility theory, which is interested in offering only an elucidation and not a reduction of evaluative concepts such as, for example, admirable. According to nonreductive neo-sentimentalism, x is admirable if and only if admiration is fitting in response to x, and admiration is fitting in response to x because x is admirable. The value of this circular statement is that it helps us see how the concept admirable is related to other concepts (e.g., admiration and fittingness) and so, while not reductive, it is nonetheless elucidating (Tappolet, 2023, pp. 166-168).

The fact that Tappolet lands on nonreductive neo-sentimentalism is good news for her because this theory fits well with evaluative theories of emotion. Recall that on such views the essence of emotions is constituted by representations of its objects as having specific evaluative features (e.g., admirableness), and those representations are correct when their objects have those specific evaluative features. Tappolet then concludes that “it is plausible that fittingness is nothing but representational correctness” (2023, p. 168; my emphasis). That is, what it is for an emotion to be fitting is for it to correctly represent the way that the world is.

It is this last claim that I want to look at more critically. Given its importance to Tappolet’s preferred version of sentimentalism, it is worth considering more closely. Tappolet seems to endorse something like the following:

Fittingness as Correct Representation (FCR): A’s emotion e toward object o is fitting just in case e represents o as having evaluative feature f (with intensity i) and o actually has f (with intensity i)—that is, just in case A’s emotion e correctly represents o as having f (with intensity i).

FCR makes fittingness strongly analogous to truth. To say that A’s belief that p is true is just to say that it correctly represents the way the world is—that is, just in case p.[1] Ben’s belief that George Eliot’s real name was Mary Ann Evans is true just in case her real name was Mary Ann Evans. Likewise, Fred’s fear of some dog represents that dog as dangerous to some degree and thus his fear of that dog is fitting just in case that dog is actually dangerous to that degree.

My main concern with FCR is that it fails to treat fittingness as a normative property. I will argue that fittingness seems to display certain features of normative properties and that FCR does not seem capable of straightforwardly explaining such properties.[2]

1. Normative Features of Fittingness

Reasons are uncontroversially normative entities. One interesting feature of the reason relation is that it is sometimes permissive and sometimes requiring.[3] To say that a reason to φ is permissive is to say that it recommends φ-ing and could justify an agent φ-ing (to some degree), but the agent would not be blameworthy for failing to φ even though he possessed this reason to φ. Thus, such reasons are discretionary. For example, I have such a reason to bring fresh flowers into my department’s common area because doing so would give it a more pleasant atmosphere. I would be justified in bringing in such flowers because they would give it a more pleasant atmosphere. However, my failing to do this does not make me blameworthy—even if I fail to do it simply because I don’t feel like it. But, if we are divvying up departmental responsibilities and I make a sincere promise to my colleagues to bring flowers into the department, my reason to do so is now requiring. It is no longer up to me whether to do so. My not feeling like doing it would be insufficient to justify or excuse my failing to do so and I would be blameworthy for this failure.

Fittingness seems to exhibit this same normative duality: it is sometimes a permissive relation and sometimes a requiring relation. For example, it is fitting for me to admire admirable people in the sense that it is discretionary whether I admire them. If I were to admire them for the right reasons, I would be justified in admiring them. However, I would not be blameworthy (morally, epistemically, prudentially, aesthetically, etc.) for failing to admire them.

In fact, fittingness seems even more permissive in the sense that sometimes the intensity of my emotion is discretionary. For example, let’s grant that Michael Jordan is a tremendously admirable athlete. It seems as though it would be fitting for me as a fellow basketball player to admire Michael Jordan even if I do not admire him very much despite the fact that he is, we have granted, tremendously admirable. I’m not blameworthy for failing to admire Jordan with the fully accurate degree of intensity.[4]

However, sometimes fittingness seems to be a requiring relation. That is, sometimes a person seems to be required to hold some attitude with a particular degree of intensity toward some object. In what sense is it requiring? First, failing to hold that attitude (to the right degree) while lacking an adequate excuse makes that person blameworthy. Second, in deliberating about how to feel about that object, they should exclude certain considerations against feeling that way.[5] Third, we often talk about fittingness in terms that suggest it is a requiring relation—for example, Neil deserves to feel guilty for what he has done, George owes Nikola gratitude, Charles ought to feel grief about his wife’s death, and so on.

To see that fittingness seems to sometimes be requiring, consider the following. Imagine that I need a kidney and a total stranger donates their kidney to me in order to benefit me for my own sake. It seems as though it would be fitting for me to be grateful to them for donating their kidney. But whether I am grateful to them does not seem to be a discretionary matter. In this case, fittingness seems to be a requiring relation. Arguably, if I fail to feel grateful to them (to the right degree) and I lack an adequate excuse, then I’m blameworthy.[6] Likewise, one might argue that in deliberating about how I should feel toward the donor, I should exclude certain considerations. For example, imagine that I simply am not sure how to feel toward them and so I deliberate on the matter. The fact that I do not like them, that I would be embarrassed to be grateful, that I don’t feel like being grateful, and so on should be excluded from my deliberation about whether to feel grateful to them or to try to make myself feel grateful to them. In fact, it seems accurate to use requiring language to describe gratitude to them. We would say that I owe them gratitude and that they deserve my gratitude for what they have done for me (Card, 1988; Lewis, 2024).

Other reactions that seem to be fitting in the requiring sense. Consider grieving the loss of a loved one. One might think, as Robert Solomon (2004, p. 75) argues, that “grief is not only expected, as the appropriate reaction to the loss of a loved one, but also in a strong sense is obligatory.” Likewise, it is plausible that feeling guilt, remorse, or shame for one’s own wrongdoing is also fitting in the requiring sense.

In fact, some emotions seem to be fitting in a strongly requiring way—that is, what is required is both having the emotion and having it with the correct intensity. One might think that to owe someone a lot of gratitude, but to feel only a little grateful to them makes one blameworthy. Solomon notes that “a person who does not grieve or does not grieve sufficiently at the death of a loved one is subject to the most severe moral censure” (2004, p. 78; my emphasis). Likewise, feeling only a little bit of guilt or shame for a serious wrongdoing or betrayal seems to make one blameworthy.

2. Problem for FCR

If fittingness is normative and it can be permissive and requiring, FCR is in trouble. First, if all that it means for gratitude, grief, guilt, remorse, and shame to be fitting is that they accurately represent the world, it is a mystery why fittingness seems normative at all. Arguably, whether some attitude correctly evaluatively represents it object by itself says nothing about whether an agent is permitted or required to have that attitude, just as the fact that a belief would be true says nothing about whether a believer is permitted or required to hold that belief. Rather, one needs evidence that the content of the attitude would match reality.

Second, fittingness seems to refer either to (a) one relation between an attitude and its object that has multiple dimensions (i.e., permissive and requiring) or (b) more than one relation (i.e., a permissive one and a requiring one). But FCR doesn’t seem capable of capturing this. After all, representational correctness is a single relation with a single dimension (i.e., on/off). An emotion either correctly represents an object as having a certain evaluative feature or it does not. But on the intuitive normative conception of fittingness, it can be weakly permissive (i.e., whether or not to have the emotion with a particular intensity is discretionary), strongly permissive (i.e., both whether to have an emotion and with which intensity to have it are discretionary), or strongly requiring (i.e., whether or not to have the emotion with a particular intensity is not discretionary).

Third, and relatedly, FCR does not seem capable of explaining why it seems correct to say that a fellow basketball player’s mildly admiring Michael Jordan is fitting, when it seems uncontroversial that Michael Jordan is very admirable. That is, it is sometimes felicitous to call an attitude fitting even when we have explicitly granted that it does not fully accurately represent its object. But, if fittingness is just correct representation, then a representation that is not fully accurate is simply not fitting.

3. Conclusion

I do not mean to present these problems as a knockdown argument against FCR. Rather, I have just tried to highlight some evidence that I think tells against FCR and in favour of an alternative account of fittingness. Responding to such evidence is especially important for Tappolet as fittingness is crucial to her view of evaluative features (i.e., nonreductive neo-sentimentalism) and thus to her view of emotions as representations of evaluative features.