Emotion and Virtue advances the nuanced argument that some individual virtuous character traits are partially and necessarily constituted by ‘modified’ (having undergone a particular course of development) emotion traits (p. 26). Arguing for the disunity of the virtues (Ch. 4), Sreenivasan focuses primarily on the individual examples of compassion and courage, along with their emotion pairs, sympathy and fear. The main arguments of the book occur in chapters 2 and 6–9, and so it is on these chapters that this review will focus.

“What must an agent be like to be virtuous? How must she be constituted, psychologically, in order to do what virtue requires?” asks Sreenivasan (p. 22). The answer is contained in Chapter 2: she must possess a certain, modified character trait – and not just coincidentally or peripherally. This is Sreenivasan’s integral view.

(ctv) says that in order to be an exemplar of, for example, compassion, an agent must consistently judge correctly what to do in a variety of situations calling for compassion. Let’s see this in action. Supposing the virtuous thing to do in the previous case is to help the man, in order to qualify as virtuous, Sreenivasan’s agent must correctly make the compassionate moral judgment “Let me help this old man to reassemble his shopping” (p. 25).

So it is that, according to the integral view, the nature of compassion is such that “every exemplar of compassion’s generic ability to make correct moral judgements in situations that call for compassion is partly constituted by a morally rectified sympathy trait” (p. 33).

The first of three arguments for the view comes in chapter 6. This adverbial argument runs as follows: intuitively, there are specific requirements for how virtuous actions are performed. In the case of compassion, “in order to be performed as an exemplar of compassion_v would perform it, any compassionate_v act must sincerely express the agent’s sympathy” (p. 125). Recall the man and the cart. The manner in which the agent helps the man seems to matter: it is not enough to merely right the cart, the agent must also act in such a way “as to preserve the thought that the help is an expression of sympathy” (p. 125). This is because, again, in order for the act to be performed as the exemplar of compassion would perform it, it must be done sympathetically.

Chapter 7 argues for the more specific formulation (briefly introduced in chapter 2) that a morally rectified emotion trait, plus cleverness and supplementary moral knowledge, are together sufficient for an agent to pass (ctv). What are each of these? By cleverness, Sreenivasan means “a generic excellence in practical reasoning, broadly construed to include specificatory reasoning in addition to narrowly instrumental reasoning” (p. 143) while the supplementary moral knowledge required is, specifically, belief “in moral equality and the correct division of moral authority” (p. 164). Recall that Sreenivasan’s is a ‘bento box’ view. These, then, are the three entrées: modified emotion trait, cleverness, supplementary moral knowledge.

This ‘bento box’ view is further differentiated from ‘black box’ views in this chapter, particularly how it is to overcome the problem of salience − that is, how to overcome the difficulty that, for some agent, “either the features of her situation to which some generous action constitutes the right response are not at all salient to her or else their salience is not sufficient to elicit the correct judgment for her” (p. 133). Sreenivasan argues that the integral view is well placed to solve this salience problem, in part thanks to the work of chapter 3 (on emotions). Roughly, this is because emotions are functional − they serve the purpose of alerting us to our environment. Says Sreenivasan, they “control a subject’s ‘input salience’ and…they control her ‘output salience’” (p. 133). Thus, emotions are well-suited to explain why certain features of the world appear salient to an agent: they appear salient because emotions selectively focus the agent’s attention (p. 135).

Chapter 8 argues that, in some cases, deferring to an exemplar of virtue is the best way to determine the morally correct course of action. This argument is advanced in two steps. The first highlights the genuine puzzle of knowing what is required in various practical situations, the solution being moral deference. Consider Mouse, whose poor country cousin has asked to stay with him for a short while (p. 169). Mouse’s house is too small to comfortably put up the country cousin; the cousin would have to sleep on the floor. What is Mouse to do? Foot the bill for his country cousin at a hotel? It would make most sense, says Sreenivasan, for Mouse to consult an exemplar of generosity (p. 171).

The second step is establishing the authority of an exemplar of virtue’s advice, which Sreenivasan says “is partly constituted by the proto-authority of her emotional response to the agent’s situation” (p. 180). What does this mean? Suppose Mouse goes to his friend Guilia for advice (p. 170). What warrants Mouse’s moral deference in this case is the proto-authority of Guilia’s emotional response to his predicament (p. 180). What emotional response? In this chapter, the emotion pair for generosity is not named. (Indeed, it’s not clear whether Sreenivasan thinks each virtue has a unique emotional complement or whether some virtues share emotions). Using instead Sreenivasan’s own example of the virtue of compassion, however, we would say that Guilia’s proto-authority comes from her sympathetic response. Thus it is that Guilia is a reliable judge of what to do in the above scenario.

Essentially, (iv) claims that every exemplar of courage uses a morally rectified fear trait to pass (ctv) (p. 214). Take Achilles, for instance: an exemplar of courage, it is a modified (by which, once more, Sreenivasan means having undergone a particular course of development) fear trait that makes him sensitive to the presence of dangers in his environment (p. 219). The adverbial argument is applied; how an agent performs a courageous action matters. Thus it is that fearlessly is derived. Take then our other two entrées, cleverness and supplementary emotional knowledge, and we have our ‘bento box’ for passing (ctv): a modified fear trait, cleverness, and supplementary emotional knowledge together tell Achilles what it is he should do. Thus it is that we can reliably defer to Achilles in practical situations in which the virtue of courage may be called for.

While the argument runs neatly for courage, questions remain about other virtues. Take, for instance, kindness. Says Sreenivasan himself, “while I have no doubt that a morally rectified emotion trait is also a functionally integrated constituent of the virtue of kindness, for example, I do not know how we are supposed to refer to this emotion (or investigate it), except as ‘the emotional dimension of kindness’” (p. 111). The question was briefly raised earlier in this review: does each virtue have a unique emotion pair? One could imagine sympathy playing the integral role in kindness just as in compassion (indeed, it’s not clear what principally distinguishes kindness from compassion; this is made even more ambiguous when we consider their adverbial counterparts, kindly and compassionately).

Regardless of this identification difficulty, the book presents a compelling and neatly argued case for its conclusion that emotions play an integral role in some individual virtues.