What is clear is that such an experience of fear involves two aspects: Belshazzar’s fear in the sense of the fear-state that he was in, and also his fear in the sense of the fear-disposition – the trigger that initiates these states as the appropriate responses to his perception of a supernatural hand writing on the wall. When we seek to provide an analysis of this illustration of an experience of fear, we need to analyse fear both as the occurrent states that constitute a response and as the disposition that initiates these states as a response to the object of fear. So a full account of a fear-experience would have to tell us about fear as the phenomenology of the fear-state and as the underlying fear-disposition that makes such states an appropriate response to the object.²

If, like Belshazzar, our ordinary experiences of fear involve both occurrent responses and dispositions that precipitate the responses, we can ask whether either or both aspects of the ordinary experience of fear might, in some sense, form part of other experiences that we have. When we turn to our experience of art, for instance, we might ask whether either or both of the two elements of the ordinary experience of fear can in some way lend a sense of fear to the experience of a painting such as Rembrandt’s Belshazzar’s Feast.³ In this painting, Rembrandt captures the suddenness of the shocking vision through chiaroscuro and various other techniques that manipulate light and shadow to evoke a mood of shock, awe and fear.⁴ If the work succeeds in expressing Belshazzar’s fear, we might compare the components of an ordinary experience of fear, such as Belshazzar’s, with those involved in the experience of fear that attends the experience of Rembrandt’s work of art. In this way, we will be able to provide an account of what distinguishes the emotional experience of art from other emotional experiences.

Whereas Rembrandt seeks to offer us an experience of the moment of fear when the imminent destruction of an empire is anticipated, in Guernica, ⁵ Picasso offers us an experience of the moment of suffering felt in the immediate aftermath of the destruction of the Basques’ ancient town and cultural centre by German bombers during the Spanish Civil War. This monumental painting does not attempt to depict a scene or narrate a story. It is simply an expression of suffering. Picasso does not refer directly to the town or the aeroplanes, the explosion, geography or time when this happened. He uses personal imagery of the girl, the bull and the horse to depict suffering. John Berger regards the painting as a protest even though there is no depiction of what it is a protest against:

We would want a philosophical account of our experience of art to explain what it means for Rembrandt to create an image in which the spectator sees fear, and for this to engage with the spectator’s fear. And we would want the same analysis to account for how Picasso’s very different painting can achieve a similar effect in which the spectator sees suffering and this engages with the spectator’s suffering. Such an account would explain how the artists seek to engage either or both of the spectator’s emotional states and emotional dispositions. The analysis of emotional experiences of art in terms of the engagement with emotional states and emotional dispositions enables us to compare the experience of looking at Belshazzar’s Feast with that of looking at Guernica by identifying what they both aspire to, and evaluating which is more successful. Armed with this theory, we are able to explain why Picasso’s picture might be regarded as more successful than Rembrandt’s, if it is better able to engage with both our emotional states and dispositions. It also allows us to compare how fear-states and fear-dispositions attend our experience of looking at Belshazzar’s Feast with how fear-states and fear-dispositions might have attended Belshazzar’s own experience of the hand writing on the wall at the feast. Indeed, it shall be a recurrent theme of mine that we can only properly understand the emotional experience of art when our analysis allows us to compare the experience of Belshazzar’s Feast with both other artistic and non-artistic experiences of emotion.

This suggests that what is required is an analysis of fear-experiences in terms of fear-states and fear-dispositions and an account of the different ways in which these mental states and mental dispositions attend artistic and non-artistic experiences. A quite distinct project would begin by asking which of the components of Belshazzar’s fear-experience we should identify as the fear-emotion. Are we able to say whether his fear-emotion is more fundamentally the fear-state or the fear-disposition that gives rise to the fear-state? It might be thought that this is in need of resolution before we can study the relationship between art and emotion: first we determine what emotion is, and then we consider whether it is ever a component of our experience of art. Such conceptual analysis of emotion, however, might not benefit our study of art. So we must begin by considering not only what is the best way to theorize about emotion, but also which approach to emotion will most enhance our study of art.

The contemporary literature on emotion in analytic philosophy is essentially a response to (which often includes a radical departure from) William James’s analysis of emotion in The Principles of Psychology.⁷ What was regarded as radical about James’s approach to emotion is his definition of emotion in terms of our feeling of bodily changes rather than in terms of cognitive activity. Following James, the question of whether emotion is to be identified with somatic or cognitive states has occupied much of the recent philosophy of emotion. However, that is not my interest. What concerns me is the explanatory use James makes of dispositions and states in his account of emotional experiences such as fear.⁸

It is well to begin by asking which of the three approaches to analysis of states and dispositions sketched out at the beginning of this section James takes. Depending how one interprets the relationship between the two relevant chapters of the Principles, James might be interpreted as analysing emotion in any of these three ways. Critical attention has focused on Chapter 25, “The Emotions”. This has led James’s interpreters to locate his full analysis of a fear-experience in the analysis he offers of emotion in terms of the felt experience of a bodily state. On this reading, James argues that one component of the experience (the fear-state) is fundamental to the experience and all that is required is the proper analysis of this component. However, when we appreciate that the preceding Chapter 24, “Instinct”, deals with another aspect of experiences such as fear (the fear-impulse), it becomes apparent that James might be read as identifying two components of the experience and offering a discrete analysis of each. If James does identify two components of the experience, we might then wonder whether a better interpretation would not involve reading the two chapters together as providing a unified account of two components of a certain kind of psychological experience, rather than as two discrete accounts of distinct psychological concepts.

When James is read as providing an account of two discrete psychological concepts, Chapter 24 is concerned with instinct as an impulse to action and Chapter 25 is concerned with emotion as the feeling of a bodily change. For James, instinct is the faculty of acting to produce ends without foresight of those ends, or without previous education. Instinct is a reflex action: an impulse. However, these impulses are not blind. An individual’s early experience of the impulse creates expectations that either reinforce the impulse or inhibit it. In this way, action is a function of instinct as modified by the life that the individual leads, and the impulse can be seen to evolve over the course of an individual’s life history. To suggest that impulses are reflexive is also misleading for another reason. Contrary instincts might act upon the same object, and the resulting conflict might block one of the instincts, thus diminishing the predictability of the reflex. Instincts can also be inhibited by habit. Partiality to the first specimen might make an agent unresponsive to subsequent specimens. Furthermore, some instincts are transitory and, having matured at a certain age, fade away unless a habit is formed. What all of this tells us is that the impulses that determine how we respond to situations are instinctual; but they also run a course that is influenced by the vicissitudes of the agent’s life.

In Chapter 25, James provides his famous account of emotion as the feeling of bodily changes. He argues that “the bodily changes follow directly the perception of the exciting fact, and that our feeling of the same changes as they occur is the emotion”.⁹ In making this claim, his principal concern is to refute the commonsense intuition that bodily changes are consequences of emotions. Thus, although we think that we weep because we are sad, in fact we are sad because we weep. It is a fact, for James, that a pre-organized mechanism enables perceived or imagined objects to excite bodily changes, and every one of these changes is felt when it occurs. That these feelings are the emotions is thought to be apparent from the fact that, when we abstract the feelings of bodily symptoms from an emotion, we find that there is nothing left: there is no residue of “mind-stuff”.¹⁰ This theory of emotion has been submitted to close scrutiny over the last century. While there are some glaring errors (such as James’s erroneous claim that all bodily changes are felt), the current neuroscience seems to suggest that James is correct in thinking that bodily changes precede our registration of an emotion.¹¹

I introduced my initial illustration of fear as an example of a fear-experience. James’s accounts of instinct and emotion might then be understood as two independent phenomena that are part of this particular kind of experience: the fearsome object excites the fear-instinct and the fear-emotion. Each of these components of the experience can be studied separately. The prevailing approach among James’s interpreters, however, has not been to treat this as a study of “emotional experiences” (in which one asks “What are the components of an experience of fear?”), but to treat him as enquiring into the concept of emotion (in which one asks “What is the essence of the concept of fear?”). When read in this way, what matters is that he identifies emotion with the feeling of a bodily change, and this is regarded as a definition of emotion. By interpreting Chapter 25 as a definition of emotion, the disposition identified in the previous chapter ceases to feature in an account of what constitutes an emotion. The impulse might be causally responsible for the experience of an emotion at a particular point in time, but it is not relevant to understanding what an emotion is. When James is read in this way, the issue becomes one of whether he has correctly identified the emotion with the feeling of the somatic state, and, if not, what modification needs to be made to the account in Chapter 25 in order to identify the emotion correctly. Whereas my first reading of James regards the project as analysing a kind of psychological experience that involves two discrete psychological concepts, this second reading understands the project as one in the conceptual analysis of emotion.

Both of these interpretations assume that the two chapters are to be treated discretely (either as studies of discrete aspects of a particular kind of experience, or as studies of discrete concepts). The question is whether the subject of one or both chapters is relevant to the issue for which we seek an account. An alternative approach, however, is to read the two chapters as providing a sustained account of the same issue. Wollheim argues that the correct way to understand James is to read the two chapters as providing a single theory of the concept of emotion.¹² He maintains that we should really understand the impulse identified by James in Chapter 24 as providing the emotion’s dispositional context, and the felt state in Chapter 25 as the feeling state in which the emotional disposition manifests itself. James’s error lies in his terminology: Wollheim maintains that James should have called the felt state described in Chapter 25 “feeling” rather than “emotion”, and he could then have called the impulse discussed in Chapter 24 by the more accurate term “emotion”, rather than the misleading term “instinct”. When we read Chapter 25 in isolation, as James’s interpreters have done, Wollheim believes that we get a distorted picture of James’s concept of emotion. Wollheim’s own theory of emotion identifies it with a mental disposition rather than a mental state, so it suits him to argue that, when James is read accurately, he too identifies emotion with a mental disposition and feeling with a mental state.

Whether or not we subscribe to Wollheim’s own theory of emotion as a mental disposition, I suggest that there is still good reason to read the two chapters together. It is clear that, for James, a fear-experience must be analysed in terms of a fear-impulse that disposes the subject to respond in a particular way to a perceived, imagined or remembered object, and in terms of a fear-state that the subject is caused to feel on account of the fear-disposition.¹³ If our primary concern is in understanding the ontology of the fear-emotion, it matters that we can resolve whether to identify the fear-emotion with the fear-state or the fear-disposition. If, however, our primary concern is to offer an analysis of the fear-expe...