Jump to ContentJump to Main Navigation
Three Faces of Desire$

Timothy Schroeder

Print publication date: 2004

Print ISBN-13: 9780195172379

Published to Oxford Scholarship Online: October 2011

DOI: 10.1093/acprof:oso/9780195172379.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2015. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 21 October 2016

Pleasure and Displeasure

Pleasure and Displeasure

(p.71) 3 Pleasure and Displeasure
Three Faces of Desire

Timothy Schroeder

Oxford University Press

Abstract and Keywords

Pleasure and displeasure are phenomena so familiar that there seems no need for a summary of everyday knowledge of them. This chapter describes the folk psychology of hedonic tone and the evidence on neuroscience of pleasure. In addition, the four incorrect theories of pleasure are shown. It also provides the three brief arguments to defend the thesis that pleasure and displeasure are distinctive types of conscious events rather than behavioral styles. Moreover, a representational theory of pleasure and the objections to the theory are discussed. Lastly, pleasure and theories of desire are explained.

Keywords:   pleasure, displeasure, theory of desire, hedonic tone, neuroscience, psychology

The structure of chapter 2 was a straightforward one: introduce the everyday notion of reward, supplement it with scientific results, and bring these sources of knowledge together to produce a theory of the phenomenon. That structure will be repeated here, though it may sound a little odd in the present context. Pleasure and displeasure,1 unlike reward and punishment, are phenomena so familiar that there seems no need for a summary of our everyday knowledge of them. All that is needed for present purposes, one might think, is a simple description of the neural basis for what everyone already knows. Yet things are not so simple. Philosophers have managed to become so confused about the nature of pleasure that they argue about whether it is a distinctive sort of conscious feeling or a style of behavior, and scientists are confused about how best to study it. Even everyday thinking about pleasure and displeasure contains puzzles that everyday thought does not readily solve. A delicate touch will be needed to untie these various intellectual knots. The ultimate result will be a theory of pleasure that makes clear the role of this face of desire.

1. The Folk Psychology of Hedonic Tone

According to common sense, pleasure and displeasure are opposite feelings, found on a continuum that ranges from the unbearably unpleasant, through neutrality, up to ecstatic heights. These feelings have many different causes and many different effects. Both causes and effects are variable interpersonally and intrapersonally, though common sense recognizes some limits to this variability, and it is far from random in nature. Pleasure and displeasure are also held to run over varied time courses, and come to an end during dreamless unconsciousness.

All this is straightforward, but worthy of elaboration. To start with, common sense calls pleasure and displeasure ‘feelings’ to distinguish them from the five (p.72) senses. Vision is obviously a sense modality, in that it allows one to sense the (apparently) objective properties of objects around one, and so on for audition and the other senses, but pleasure and displeasure are different. They do not appear to tell us about the objective properties of objects: they appear rather to be purely subjective responses we have to objects. This is the only significant gap common sense puts between hedonic tone and the senses, however. In other respects, pleasure and displeasure are held to be just like experiencing warmth or coldness, seeing bright light or deep gloom, or having other sensory experiences. In particular, pleasure and displeasure are like sensory experiences in that they make a difference to consciousness: there is “something it is like” to be pleased or displeased, to borrow from Nagel (1974). This “something” is like the “something” it is like to see or hear, in that feeling pleased is a qualitatively distinctive conscious event. Some philosophers and scientists have held otherwise (see section 3), but folk psychology is more or less unanimous on this point.

Comparisons to sensations of warmth and coldness, or to seeing light and darkness, are held to be especially apt because feelings of pleasure and displeasure are thought of as opposites. One does not generally feel pleasure and displeasure at the same time: they drive one another out. A man having a good day until he finds that his car has been burgled does not continue to feel pleasure while also feeling sudden displeasure. Rather, the displeasure drives out the pleasure. Likewise, a woman having a lousy day will stop feeling displeasure, at least for a time, if she has a very pleasant experience, such as learning that her research program has received funding. It is sometimes said that the opposite of pleasure is pain, instead of displeasure, but this is not the considered view of common sense. Most people would agree that while all pain involves displeasure, not all displeasure involves literal pain. Though I may be pained when I learn my students are ignorant of key events in Canadian history such as the October crisis, this is not a matter of literally being in pain (unless a headache or a surge of gastric juices into my ulcer is caused by this unfortunate discovery), but it is literally unpleasant to learn such a fact. Pains are just one of many ways in which one may suffer. Hence, the proper counterpart of pleasure is displeasure, not pain.

An interesting minor fact is that pleasure and displeasure, though opposites, are occasionally experienced simultaneously. Especially when one's emotions have been particularly intense over a prolonged time, one can find oneself feeling both pleased and displeased: inclined to laugh and cry simultaneously. It has also been reported to me that the thrill of a roller coaster can involve simultaneous pleasure and displeasure, in the form of literally mixed joy and fear. Such experiences are rare for most people,2 but poets have seized upon their occurrence in romantic contexts and have made much of sweet sorrow, sweet pain, bittersweet feelings, and the like. Because pleasure and displeasure are opposites, such experiences are always a little odd, but they are nonetheless held by common sense to be real.

Again like sensations of warmth and coldness, or light and darkness, pleasure and displeasure are said to fall on a continuum. There are very slight pleasures, (p.73) somewhat more intense pleasures, unbridled joy, and so on—and similarly for displeasure. And just as, for any experience of heat, one could always have a more intense experience (though the change in intensity might be slight), so experiences of pleasure and displeasure, however intense, can always be surpassed (perhaps by a vanishingly small margin at the extremes), according to everyday thought. Minimally intense pleasure and minimally intense displeasure meet at a neutral point, itself neither a feeling of pleasure nor of displeasure. This makes hedonic tone more like experiences of warmth and coldness than experiences of light and dark, since the former also has a neutral midpoint whereas the latter does not.

Given that pleasure and displeasure come on a continuum and are opposites, it naturally follows that when pleasure and displeasure clash, the result is something like the addition of positive and negative numbers. If a person feels wonderful, and then has an experience that would typically induce slight displeasure (say, a shoelace breaks), the person is likely to feel only slightly less wonderful than before. If the same person has an experience of a sort typically inducing great sorrow, however, then that person will shift from feeling pleasure to displeasure, though often not as deep displeasure as might have been felt had the person started off with a neutral feeling, or (even worse) a bad mood. But these trends are only trends: as everyone knows, feelings of pleasure and displeasure are more complex than this. Sometimes, bad news not only reduces the pleasure one was feeling but in fact makes the pleasure crumble to dust: one feels even worse than one would have, had one begun the day feeling just average. Similarly, if one has been feeling down for some time, good news that would normally cause pleasure might not have the least effect. It can also happen that pleasure renders one immune to petty concerns, and a former bad mood can be completely relieved simply by a pleasant conversation or a mildly pleasant run of good luck in some trivial domain. So while pleasure and displeasure trade off against one another, everyday observation suggests that this trade-off is often a complex one.

What about the causes and effects of pleasure and displeasure? As already mentioned, common sense holds them to be variable both between people and within a single person over time, though this variation has its limits. Pleasure, it is widely believed, is sometimes felt in response to characteristic sensations such as massage, the scent of warm brie, the sight of graceful movement, and the like, and is other times felt in response to more cognitively sophisticated mental events such as thoughts about the success of one's child in her chosen profession, witnessing an elegant move in a chess match, or hearing a radio advertisement that strikes one as humorously devoid of irony. Thus, there are sensuous sources of pleasure and intellectual sources. Displeasure, like pleasure, is also held to be caused by both simple sensations and complex cognitions: consider the unpleasantness of an overfull bladder, of the sound of very mechanically played swing music, of the thought of paying one's taxes to support an unjust social structure. Physical pain, such as that caused by dropping a heavy piece of lumber on one's finger, is generally taken to be paradigmatic as an unpleasant experience.

(p.74) Individual variation in more cognitive pleasures is widely believed to be greater than variation in more sensuous pleasures, and similarly for displeasure. Everyone3 feels displeasure when a bone breaks, a heart attack occurs, or the like, but only a comparative handful of people were crushed when Ireland was eliminated from the World Cup in 2002. Yet even sensuous pleasure and displeasure are known by common sense to admit variation: not everyone enjoys sweets, some people get used to the smell of human excrement, and some people can withstand heat without suffering. I even know a person who finds extreme bodily need for food to be enervating but not actually displeasing: quite an unusual case, but not unbelievable. These variations in causes of sensuous pleasure and displeasure are not widely discussed in everyday contexts, but long training, upbringing, constant exposure, and the like are held to be at least one source of variation. One can become hardened to the cold, to deprivation, to the sound of children screeching, and so on, according to common sense, and one can learn to like bitter greens, the smell of very strong cheeses, very hot baths, and the like. Innate differences are held by common sense to be another source of variation in sources of sensuous pleasure and displeasure: some people suffer in the heat because their “inner thermostats” are just set differently from those of other people, and some children prefer grapefruits to oranges from an early age.

Variation in sources of pleasure and displeasure of a less sensuous, more cognitive nature are dealt with differently by common sense: they are generally attributed to preferences or desires. It is the fact that you wanted Ireland to win, while I was indifferent, that explains why you were displeased by Ireland's elimination while I was not, and so on. From this fact follows an interesting lemma of common sense: that pleasure and displeasure can provide useful information about one's preferences or desires. “What do you want to eat?” asks Tom. “I don't know,” Susan answers. “Let me look into the fridge and see.” Looking into the fridge allows Susan to survey the options, and whatever induces pleasant interest will be chosen, Susan's pleasure being a better guide to her current preferences than her intellect.4 Likewise, a person can claim to want a promotion to sales manager, then get it and realize that he didn't want the promotion after all—what he really wanted was simply recognition of his abilities—and his insight into what he wants can be delivered exactly by noting his pleasure in the round of congratulations and displeasure at the prospect of actually taking on the new duties. And in general, a person confused about what she wants is advised to consider the likely options and await the faint anticipatory pleasure or displeasure (not always so faint!) that follows, in order to learn what she wants.

Additional sources of variation in the impact of events upon hedonic tone are factors often called “confidence” and “resignation”: in essence, forms of subconscious estimates of likelihood. Two people alike in every way, going on an outing, will be expected to be differently affected by the success of the outing depending upon the expectations they have at the beginning. If both people are normally confident that things will go well, but one has seen heavy clouds presaging (p.75) rain, and so goes on the outing with low expectations, while the other has seen blue skies, and so expects beautiful weather, then the former will be expected to feel a little less sorrow than the latter if the outing is rained upon. The former person will be displeased, naturally, but, “knowing it was coming,” feel that things went “no worse than expected,” as might be said. The latter person will be more displeased, for the rain “took him aback” he will feel let down. Likewise, one is more delighted to get good weather when one had feared bad than when one expected good weather all along. Folk wisdom frequently recommends using this source of variation in order to shield oneself from hedonic shocks: “don't get your hopes up too high” is advice often given to someone in an effort to cushion the person from the displeasure of disappointment, and “appreciate what you've got: you may not have it tomorrow” is a common admonishment to one who is ceasing to take pleasure in something he feels to be certain, but which really is not.

The effects of pleasure and displeasure are just as varied as their causes. Pleased people often sigh, relax, or smile; displeased people often grunt, tense up, or frown. But there are people who register satisfaction with an unhappy-sounding grunt, or who frown at their own enjoyment of something, and people who giggle when deeply uncomfortable with embarrassment. For my own part, I have the unfortunate tendency to a nervous smile when delivering awkward or bad news, but the smile reflects no inner glee. An interesting systematic phenomenon noted by common sense is that some sources of pleasure are consistently soothing or satisfying, while others are exciting and arousing. Orgasm, eating a large meal, resting after a hard day—these are sources of pleasure that, for most people, are linked to a certain laziness. The effect of pleasure in other cases is quite different. A first kiss, news that one has been accepted to the graduate program of one's choice, taking possession of a sporty new car—these are sources of pleasure that are stimulating for most people, inducing jumping for joy, trembling with excitement, urges to act, and so on. In just the same way, sources of displeasure can cause a defeated-feeling lassitude, or they can arouse the will to fight back against the adverse circumstances. These effects are often independent of the actual usefulness of the responses in question. The energy that good news about a graduate school application can provide has no obvious use, just like the energy that bad news about such an application can bring in a person who tends to anger, but an energized state can result nonetheless.

An interesting further phenomenon surrounding pleasure is that common sense recognizes the possibility of “unreal” pleasure. This is not apparent but non-actual pleasure, but rather pleasure that is real but in some sense inappropriate. The pleasure of euphorigenic drugs, including alcohol, is often held to be unreal in this sense, for example, and the same would go for pleasure induced by electrical stimulation of the brain or by psychosurgery or neural insult. The other sort of context in which talk of unreal pleasure comes up is that in which an individual seriously misrepresents her situation and feels pleasure as a result. For instance, there is thought to be something illusory about pleasure felt because one (p.76) misrepresents a hostile takeover as a mutually beneficial merger. How common sense would justify the notion of an unreal pleasure is far from clear, but the label is applied in both sorts of contexts.

Pleasure and displeasure may last for a short time, as in the unpleasantness of sitting in an awkward position, which disappears almost as soon as one changes posture. It may last for a moderate term, as in the case of the afterglow of sexual pleasure following orgasm, which persists for some time after the initial cause of pleasure. Or it may be felt intermittently for a term of days or weeks, as romantic infatuation and depression are both thought to do. Unconsciousness (of the sort found in dreamless sleep) is generally held to cause pleasure to cease, at least for the duration of the unconscious state, and to likewise give relief from displeasure.5

In addition to its confident assertions about pleasure and displeasure, folk psychology also confesses to a certain amount of uncertainty. Two sorts of uncertainty are especially noteworthy. Common sense is not particularly clear on where, if anywhere, pleasure and displeasure are felt, or on how many sorts of hedonic tone there are. It is only common sense to say that the pain of a toothache is found in one's tooth and jaw, but it would also be absurd to say that the “pain” from a failed romance is literally in one's heart, or anywhere else, and between these cases fall very unclear ones, such as the displeasure felt in smelling rotten meat. Is it literally in the nose? Or is it, like heartache, found nowhere in particular? Or is it found throughout the body? Such questions have no settled, common sense answers. Not unrelated, the question of whether pain is different in kind from heartache in respect of its unpleasantness, or whether the pleasure of a skillful checkmate is different in respect of its pleasantness from the pleasure of a pedicure is also unanswered by common sense. There is enough confusion on these topics to leave room for the philosophers.

2. The Neuroscience of Pleasure

From common sense to neuroscience: our best scientific evidence at present supports the idea that there is a neural seat of pleasure, and an adjacent seat of displeasure, found in a phylogenetically old region of the cerebral cortex known as the perigenual region of the anterior cingulate cortex (or PGAC, a circumscribed subregion of Brodmann's area 24, found around the “knee,” or bend, in the anterior cingulate cortex; see appendix, fig. 2).6 The thesis that the PGAC is the seat of pleasure and displeasure is not universally accepted, but it has been advanced by a number of scientists (see, e.g., Devinsky et al. 1995; KSJ 2000, ch. 61; Vogt et al. 1992; Whalen et al. 1998), and in this section I will argue that the evidence warrants their assertion.7

To begin, it must be made clear that saying there is a neural seat of hedonic tone is not the same as saying that pleasure and displeasure just are happenings in neural structures. Whether they are such events, or whether they are, say, functional patterns instantiated by neural structures in humans but potentially instantiated (p.77) by other structures in other creatures, is a question for another branch of the philosophy of mind, and one that need not be addressed here. The claim being made by our present best scientific evidence is simply that there is a clearly identifiable, localized region of the brain that is the biological realization8 of hedonic capacities in humans and other animals.

It is also important to be clear that the PGAC is just a small part of the anterior cingulate cortex as a whole, and the anterior cingulate is not functionally homogenous by any means. It includes an important pre-motor region located caudal to the PGAC9 that is involved in controlling behavior; a structure exerting control over the viscera;10 and probably other functional subdivisions including one related to attention (Vogt et al. 1992). Thus, not every study examining the anterior cingulate is relevant to drawing conclusions about pleasure and displeasure: attention must be focused upon the perigenual region.

With these clarifications in mind, it is time for the obvious question. What sort of evidence would justify the claim that activity in a discrete region of the brain such as the PGAC is the biological realization of pleasure and displeasure? The best evidence that the PGAC realizes pleasure and displeasure would be evidence that it is the unique structure whose activity is, in a suitably restricted sense, necessary and sufficient for pleasure and displeasure in creatures like us.11 Consider first sufficiency. It is obvious that there will be many sites in the brain sufficient, in normal organisms, for the production of pleasure. Stimulate my optic nerve in a manner corresponding to the stimulation induced by a painting by Chagall and I will be pleased; stimulate my olfactory center in a manner corresponding to that induced by steaming hot chocolate, and I will be pleased; and so on. Intuitively, what one is seeking is the last such location in the chain of neural events sufficient for producing pleasure, but this is tricky. After all, no part of the brain is sufficient for an experience of pleasure if surgically removed from the organism and stimulated in vitro.12 Sufficiency must be understood only as sufficiency in normal, healthy organisms. As for necessity, it may be understood in a somewhat stricter sense. Even if an individual were to lose all normal sensory and imaginative capacities, it might still be possible for him to experience pleasure and displeasure if some other region of his brain were directly stimulated by chemical or electrical means. Such facts would obviously be relevant to questions of whether or not these sensory and imaginative capacities are the neural basis of pleasure. So what the scientist seeks, in looking for the neural basis of hedonic tone, is a unique discrete structure such that it is the last structure in the normal causal sequence whose activation is sufficient for pleasure and displeasure, and also such that its activation is necessary for pleasure and displeasure.

The obvious methods for testing sufficiency and necessity are invasive, causal studies, in which particular structures in the brain are chemically, electrically, and surgically manipulated. Such studies have, happily, rarely been performed on humans. But humans who can describe their experiences are the best source of information about pleasure and displeasure. The overall quality of information about the neural basis of pleasure is, as a result, poorer than it might be.13 Still, (p.78) what evidence does exist appears to strongly support the view that the PGAC plays this role.


Direct stimulation of a number of regions in the human brain is known to cause pleasure. Stimulation of sub-cortical structures such as the lateral hypothalamus, ventral pallidum, nucleus accumbens, and septum, for example, has been thought to cause pleasure (reviewed in Berridge 2003a, 2003b). But causally downstream of these structures is the PGAC, and stimulation of the PGAC also suffices to cause pleasure and displeasure in human subjects. In studies by Meyer et al. (1973) and Laitinen (1979), stimulation of the cingulum (a bundle of nerves reaching out to the various parts of the cingulate cortex) and the perigenual cingulate itself was found to have a number of effects in human subjects, corresponding to the various functions of the cingulate, including (especially when the PGAC was the specific target) hedonic effects. Some subjects reported displeasure in various forms (fear, agitation, oppression, specific and generalized pain, nausea) while others felt pleasure (including simple pleasure, relaxation, and a sense of well-being).


Damage to the PGAC (KSJ 2000, ch. 61) and cingulumotomy (Foltz and White 1962) can cause recipients to report a complete loss in subjective ability to experience displeasure, or any sort of hedonic tone, from either sensuous or more cognitive sources. It might help to recall some of the observations from Foltz and White (1962) quoted in chapter 1. Of their cingulumotomy recipients, they write:

The patient with a good result simply is not as precipitously reactive to his own environment and his own situation as he was prior to operation. The anguished facies and evidence of suffering are modified markedly. The perception of pain as such does not appear to be modified, but the patient's total reaction to pain and the threat to existence that it represents is modified markedly. Most of the patients stated they continued to have pain but it was “not distressing,” “not particularly bothersome,” “doesn't worry me anymore,” etc.

Such descriptions indicate a radical change in the nature of pain experience, which pain theorists such as Melzack (1973) and philosophers such as Aydede (2000)14 have interpreted as the feeling of pain losing its unpleasant character while retaining its character as nociception, sensory perception of bodily injury. Also important in showing the unique role of the PGAC is the fact that, while other forms of psychosurgery (such as full frontal lobotomy) are also effective in eliminating the capacity for hedonic experiences, destruction of the PGAC appears to be the most limited procedure that is effective in this regard, and the one that has the least dramatic consequences beyond loss of reported hedonic tone. (p.79) Destruction15 of the PGAC thus appears to be the most limited procedure that eliminates capacities for pleasure and displeasure.

Brain imaging techniques have provided evidence that both pleasure and displeasure cause increased activity in the PGAC, providing further support to the claim that the PGAC is the neural seat of hedonic tone. Beginning with displeasure, a number of pain studies have found that painful stimuli activate the PGAC (Coghill et al. 1999; Craig et al. 1996; Derbeyshire et al. 1998; Mertz et al. 2000; Vogt et al. 1996). There are many, many more studies than these attempting to capture images of the brain's response to painful stimuli,16 but these studies often fail to distinguish between activation of the caudal anterior cingulate and the PGAC, or have focused upon the caudal anterior cingulate exclusively. Since the caudal anterior cingulate is now believed to be a motor center, and direct stimulation of it has been found to cause no displeasure (Hutchinson et al. 1999), it would be a mistake to draw conclusions about the neural seat of displeasure from studies that do not distinguish between the PGAC and other portions of the anterior cingulate. It has been speculated that, because most pain imaging studies require subjects to remain still while experiencing pain, motor resources are called upon to inhibit avoidance behavior, fidgeting, and so on, and this activates the caudal anterior cingulate, creating a signal correlating with displeasure but not helping determine the neural localization of the feeling. This may well explain why few scientists have noticed the importance of distinguishing sub-regions of the anterior cingulate cortex from one another. Fortunately, these problems are not endemic to all work on the anterior cingulate. Further evidence regarding the neural basis of displeasure comes from studies that induce displeasure without also creating an incentive to avoidance behavior: studies inducing psychologically based displeasure. Shin et al. (2000) found that when subjects experienced guilt (and related emotions such as sadness and shame, all involving displeasure), activity increased in the PGAC. Paying attention to the emotional content (both pleasant and unpleasant) of pictures activates the PGAC as well (Lane et al. 1997), and likewise both pleasant and unpleasant tastes cause increased activity in PGAC (Zald et al. 1998).

Turning now to pleasure, most studies imaging the human brain under pleasure-causing conditions have used drugs to cause the pleasure. Correlations between pleasure and activity in the PGAC have been found in studies on alcohol (Ingvar et al. 1998), cocaine (Breiter et al. 1997), THC (Matthew et al. 1999), MDMA (Gamma et al. 2000), and amphetamines (Vollenweider et al. 1998). Another study, this one looking at pleasure and excitement induced through stories subjectively rated as strongly inducing these feelings, also found activation of PGAC during pleasure (Rauch et al. 1999).

A few studies exist that do not accord with the present thesis. One fMRI17 study dissociating pain from the anticipation of pain (Ploghaus et al. 1999) found that the anticipation of pain activated a region encompassing the PGAC and other mediofrontal areas, as expected (since anticipation of pain is itself generally unpleasant), but pain itself did not activate the PGAC to a statistically significant (p.80) degree, though it did activate the caudal anterior cingulate cortex. However, the region of analysis of the anticipation of pain did not differentiate between the rest of the mediofrontal cortex and the PGAC, leaving open the possibility that the PGAC was activated near or below statistical thresholds in both conditions, while the orbitofrontal cortex was only activated, or was much more activated, in anticipation of pain. (Recall, from chapter 2, that the orbitofrontal cortex sends information about known and predicted rewards and punishments to other regions of the brain.) Another study, conducted by Iadarola et al. (1998) using PET,18 found that the anterior cingulate gyrus was significantly activated by light brushing of the forearm and by injection of capsaicin (the compound that makes chili peppers hot—yes, that really was injected into the forearms of human volunteers!) but not by the allodynia (the pain that non-harmful touching can cause when the touch is to an area that was previously the source of painful stimulation) that followed capsaicin, once the effect of brushing the arm to induce allodynia was factored out. Here, it might be pointed out that allodynia is considerably less painful than capsaicin, and so failure to find significant activation of the PGAC might be the result of a substantially reduced signal. Perhaps most challenging, Porro et al. (1998) found that the temporal profile of the subjectively experienced pain of a painful acid injection (again, human subjects volunteered for this!) correlated with deactivation of the PGAC in an fMRI study, directly contradicting numerous other studies. Likewise, Smith et al. (1999) found that inducing tryptophan depletion to bring on depression in formerly depressed subjects caused a deactivation of the PGAC. How to reconcile these results with the dominant thesis favoring the PGAC is not clear. However, this level of disagreement, even in well-conducted scientific experiments, is far from exceptional, and is not, on its own, powerful evidence against the view under consideration.

It might be argued that the PGAC is activated by any arousing condition, and so its activity merely correlates with pleasure or displeasure without actually realizing hedonic tone. However, this would not explain why certain studies, such as Breiter et al.'s (1997) study on the pleasure produced by cocaine, found strong activation of the PGAC during the arousing “rush” period of the drug's effect, but not during the also arousing “craving” period that follows later. It would also not explain why direct stimulation of the PGAC was found to cause pleasure and displeasure (depending on site of stimulation), rather than feelings or behaviors related to neutral arousal, and it would not explain the loss of hedonic capacities that can follow damage to the PGAC. This style of argument is an important one: taken as a whole, the brain-imaging, neurosurgery, and brain-stimulation studies support much stronger conclusions than any single portion of that evidence taken on its own.

One further line of evidence supports the claim that the PGAC is the neural seat of pleasure and displeasure: no other plausible contender for this role exists within the brain. Most important, the dopamine-releasing reward system discussed in chapter 2 is not a plausible candidate for the neural seat of pleasure. (p.81) This reward system, whose core is the output from the VTA/SNpc, is not a ludicrous candidate for the neural seat of pleasure by any means. After all, rewards tend to activate it, and rewards are typical causes of pleasure. Likewise, studies on non-human animals have shown that the VTA/SNpc is a powerful site for self-stimulation (reviewed in Stellar and Stellar 1985), meaning that animals with electrodes planted in the VTA/SNpc, or in areas projecting to the VTA/SNpc such as the medial forebrain bundle, will do whatever it takes to cause these electrodes to stimulate the VTA/SNpc again, sometimes to the exclusion of eating or drinking. It was just these scientific results that led philosopher Carolyn Morillo (1990) to the conclusion that pleasure is seated in these structures. But evidence reviewed in Berridge and Robinson (1998) suggests that the scientific findings are more likely to be explained by the hypothesis that the VTA/SNpc and medial forebrain bundle can indirectly activate the pleasure center than by the hypothesis that either is such a center. For instance, as discussed in chapter 2, Wolfram Schultz and colleagues found that VTA/SNpc activation bursts briefly when an organism first learns that it will get a reward, but then immediately (in less than a second) drops off; most sources of pleasure, however, provide much longer-lasting pleasure than this. If VTA/SNpc firing were the neural basis of pleasure, and not merely an indirect cause, then this discrepancy between the time-course of firing and the time-course of pleasure would be inexplicable. Berridge and Robinson (1998) also report that human and animal experiences of food palatability (measured either by facial expression or self-report) increase with morphine, but morphine administration is not thought to cause VTA/SNpc activity. Benzodiazepines, also thought not to cause VTA/SNpc activity, have similar effects. Finally,Berridge and Robinson (1998) describe a series of experiments in which it was found that rats with massive lesions specific to dopamine releasing neurons (mainly the VTA/SNpc) were still capable of experiencing pleasure in response to sugar and displeasure in response to quinine, as evidenced by the fact that they still displayed the facial responses rats make in response to pleasant and unpleasant tastes in general. As a whole, then, the evidence is strongly against the idea that the VTA/SNpc is the neural seat of hedonic tone.

In a pair of recent reviews (2003a, 2003b), Kent Berridge surveys other prominent candidates for neural centers of pleasure and displeasure and finds most of them wanting as well. Destruction of the orbitofrontal cortex alters emotional response, but nonetheless leaves the full spectrum of hedonic tone available, and so the OFC fails to be necessary to feelings of pleasure and displeasure. Likewise, damage to the amygdala, while blunting or eliminating emotional responses (especially fear responses) to many stimuli, leaves open the possibility of pleasure and displeasure in response to other stimuli. Stimulation of the septum is reputed to be a source of pleasure, but not of displeasure, and in any case it is unclear (according to Berridge) that the results purporting to show this effect in humans are caused by stimulation of the septum. Even if stimulation of the septum can produce pleasure, it can be pointed out that it is generally reported to produce sexual pleasure (Heath 1961, 1963), and sexual pleasure is hardly the (p.82) whole of human pleasure. Perhaps what has actually been caused is sexual arousal, which in turn is a source of pleasure rather than identical to it.

It should be said that Berridge is also skeptical as to the role of the anterior cingulate cortex in pleasure and displeasure, but here his evidence is somewhat weaker. The region he identifies as cingulate cortex in Berridge (2003b) corresponds to what are now thought to be motor regions of the anterior cingulate rather than affective regions. The PGAC, thus, has not been the center of his attention in reviewing work on the anterior cingulate, and this may have allowed the introduction of some misleading evidence. Berridge's main complaint against the anterior cingulate cortex is that there is a dearth of evidence that it is necessary for pleasure. In particular, there is a study on rats (Bussey, Everitt, and Rob-bins 1997) showing that rats with damaged anterior cingulate cortices responded as strongly to rewarded stimuli as did control rats. Yet Berridge has been a leader in keeping a very clear distinction between experiments revealing modifications of hedonic tone and experiments affecting motivation. Bussey, Everitt, and Rob-bins appear to have demonstrated some motivational similarities between lesioned rats and control rats (though there were also striking differences in behavior), but this hardly entails similarities in hedonic tone, and in fact when one attempts to lesion a candidate pleasure center without lesioning the leading candidate for a reward system (the one described in chapter 2), one can expect to see complex results, including reward-based responding, without necessarily observing hedonic responses.

Berridge prefers the hypothesis that the neural realization of what he calls “core ‘liking’” is found in the circuit linking the nucleus accumbens, ventral pallidum, and brainstem parabrachial nucleus, all subcortical structures. But according to Berridge, “conscious liking may or may not accompany a given instance of core ‘liking’,” and he adds that, for his purposes, “it will be enough to identify brain systems that at least cause ‘liking’—whether or not accompanied by conscious liking” (2003a, 118). Hence, it is not completely clear that Berridge's target is the same as the target of this section, that is, the neural realization of what people commonly denote by ‘pleasure’ and ‘displeasure’. Berridge's targets are the primitive, deep-brain structures common to many vertebrates that form the biological core processes upon which conscious human pleasure and displeasure depend. Berridge's awareness of the potential discrepancy here between the core processes and the feeling of pleasure itself is what leads him to call his target ‘liking’, with the quotation marks always scrupulously attached. In fact, given the way Berridge writes about the connection between these core processes and conscious feelings, it appears that he strongly suspects that cortical structures are required for conscious pleasure—a not implausible suspicion, given the importance of cortical structures to consciousness in general.19 For instance, he describes a neural pathway via which ‘liking’ could influence “feelings of pleasure that might be instantiated by limbic regions of neocortex” (2003a, 121). The PGAC, it should be noted, is a region of limbic neocortex that is one target of the pathway Berridge mentions.20 Hence, even if Berridge is right that there is a phylogenetically (p.83) old core process that underlies feelings of pleasure and displeasure, this need not be incompatible with the view that the feelings themselves are found in somewhat phylogenetically younger structures such as the PGAC.

The evidence that the PGAC is the home of pleasure and displeasure is thus diverse and, while not flawless, impressive. What to make of it is a subject to which we will return.

3. Four Incorrect Theories of Pleasure

With the data of common sense and neuroscience at hand, it is time to find out what pleasure and displeasure really are: to have a useful theory of hedonic tone. There are four answers familiar to philosophers: pleasure and displeasure are modifications of behavioral dispositions and nothing more; they are forms of other sorts of experiences; they are distinctive qualia; and they are players of particular functional roles. This section will argue that each answer is unsatisfactory, and the next will propose an alternative.

Gilbert Ryle (1949, 1954) famously held that pleasure and displeasure are not types of experiences but rather something like modifications of one's dispositions. To engage in an activity with pleasure is to engage in that activity attentively and eagerly, says Ryle, nothing more. The view is hardly exclusive to Ryle: variations on it were widely held by his contemporaries,21 and a contemporary instance of it is found in Tye (1995). To tell the truth, I have never understood philosophers' capacity to defend the Rylean position. Is it not obvious that pleasure and displeasure are more than styles of behavior? But this is not much of an argument. Therefore, I offer three brief arguments to defend the thesis that pleasure and displeasure are distinctive types of conscious events rather than behavioral styles. An extended defense of this thesis is given by Aydede (2000), and the reader who remains unconvinced is referred to this work.

First, consider instances of intense pleasure (induced by orgasm, peak physical performance resulting in competitive victory, cocaine or heroin, or otherwise) and intense displeasure (induced by toothache, heartache, the death of a loved one, or the like). Such experiences seem to involve a distinct type of experience, pleasure or displeasure, which is made salient by its intensity. I see no reason not to take this interpretation at face value: the burden of proof is on the person who wishes to tell the majority that common sense is misleading us.

Second, common sense suggests that it is possible to feel pleasure and displeasure while dreaming. Sexual dreams and anxiety dreams are especially obvious examples. But if pleasure is a matter of behavioral style or dispositions to behave, then what is to be made of common sense? After all, no actual behavior occurs in normal dreams, and in fact the body's very capacity to move is normally disabled during dreaming (which is why sleepwalking is rare rather than commonplace). If pleasure and displeasure were essentially tied to behavior, it would be impossible for them to occur in dreams.

Third, common sense also suggests that people who are behaviorally disabled (p.84) are just as capable of feeling pleasure and displeasure as the rest of us. A person unable to control his voluntary muscles, and who has thereby lost the tendency to respond avidly to some stimuli and to shun others, is not ipso facto a person who cannot be made to feel pleasure or to suffer: it would be ferocious cruelty to deliver electrical shocks to a permanently paralyzed person, or to a person with a severe motor-control disorder, says received wisdom, and it is hard to disagree, exactly because we imagine that such shocks would be very unpleasant even though they would not affect tendencies to behave. Recent work in neuroscience concurs for non-humans: recall the rats described in Berridge and Robinson (1998), whose motivational systems were lesioned. Measures of facial reactions showed that rats continued to produce characteristic facial expressions of pleasure in response to sucrose solutions and displeasure in response to quinine solutions, though the rats no longer had any tendency to consume the sugar or avoid the quinine. Rats, like human beings, seem to be capable of the normal range of enjoyment and suffering whether or not they are capable of interesting behavior. A devout Rylean could hold that these facial responses are all there is to pleasure and displeasure in such rats, but this strains credulity to the breaking point.

Ryle (1971, 326–27) seems to be motivated to take the view he does on pleasure largely by the worry that pleasure, if a distinct sort of event in consciousness, must either be something experienced (a sort of sense datum, accessible to an “inner eye”), which he takes to be inadmissible, or else it must be a type of experience with no distinctive object, which is also intolerable. Likewise, it seems to me that Michael Tye (1995) has a reason to hold pleasure and displeasure to be ways of responding to other experiences only because he is committed to the view that every type of consciousness is an experience of something, and he can think of nothing for pleasure or displeasure to be experiences of. But Ryle and Tye fail to consider the possibility that pleasure and displeasure are types of experiences that do have distinctive objects: that hedonic tone, like vision, “says” that the world is this way or that. This is the approach I propose to take. Such an approach, a representational (or, if one prefers, intentional or informational) approach to pleasure and displeasure, is precisely one that offers a distinctive object for these sorts of experiences, and so addresses Ryle's worry, and Tye's possible worry, by making pleasure and displeasure of a piece with vision and audition. The only intelligible motivation I know for holding a Rylean position, then, is one which I will answer in the next section.

What of the view that pleasure and displeasure are not distinctive events in consciousness but rather ways of having other experiences? This sort of view is less implausible than the Rylean position, since it at least ties pleasure to consciousness, and it also has some distinguished advocates. Perhaps the best version of this theory available at present is that offered by Antonio Damasio (see, e.g., Damasio 1994). According to Damasio, pleasure and displeasure are “particular body landscape[s] that our brains are perceiving” (1994, 263). Exactly what this means is clarified in Damasio's discussion of pain. Pain, according to Damasio, involves both sensation of tissue damage (nociception) and displeasure, (p.85) which is distinct. “The innocent processing of body change [i.e., nociception] rapidly triggers a wave of additional body-state changes that further deviate the overall body state from the base range. The state that ensues is an emotion, with a particular profile. It is from the subsequent body-state deviations that the unpleasant feeling of suffering will be formed” (263, italics in original). Thus, the displeasure component of pain is one's perception of the changes in one's body that follow from being injured. Damasio is not overly specific about which body-state changes one must experience in order to experience displeasure, but details elsewhere suggest that he has in mind things such as elevated pulse rate, tightening of the stomach, sweating, twitching, rapid breathing, loss of blood to the skin, and the like. Damasio has less to say about pleasure, but it too is held to be one's awareness of changes in one's overall body state.22

To put Damasio's view in its baldest form, then: to be pleased or displeased is to feel certain clusters of bodily states. The displeasure of fear is the feeling of one's muscles tightening, one's stomach knotting, one's heart pounding, and so on; the displeasure of depression is the feeling of profound lassitude and tiredness, a sinking sensation in one's stomach when activities are contemplated, a slowness of breath and pulse, and the like. The pleasure of orgasm is a feeling of muscular tension peaking and then dropping off, contractions of specific muscles around the genitals, a sudden change from an energized to a lazy state, and so on. Or, if these identifications of types of pleasure and displeasure with types of feeling changes are not quite right, they are at least decent first attempts. These are the sorts of things Damasio holds the various pleasures and pains to be.

Put in this form, Damasio's theory appears almost as implausible as Ryle's. Everyday thought would have it that pleasure and displeasure are distinct types of feelings, not merely collections of experiences of one's bodily states. As a result, the theory runs afoul of the first criticism made of Ryle's theory: that it diverges from common sense without justifying the divergence. In addition to mere implausibility, it can also be pointed out that Damasio's theory appears to get particular cases wrong. Imagine two people about to go skydiving. One is a self-described “adrenaline junkie” who loves similar activities; the other is adrenaline averse, hating most “exciting” activities. As the moment to jump approaches, both people will have powerful adrenal, and so visceral, responses: both will feel their hearts pound, their stomachs tighten, their breathing change, and so on. If both step up and jump in the same way, their experiences of their body landscapes will be quite similar around the time of the jump. Not the same, but in gross outline quite a bit more similar to one another than to the body state of a person sitting quietly with an amusing book and a cup of coffee. On Damasio's theory, the two should have similar hedonic tones, given the similarity of body landscape. Yet one is likely to describe these feelings as “a rush,” “the thrill of anticipation,” and “elation,” while the other is likely to describe them as “a surge of anxiety” and “terror.” The difference, it seems to me, is not well explained by the subtle differences in experienced bodily landscape. The difference between the two skydivers is simply the difference in whether one is pleased or displeased (p.86) by skydiving, or even certain body landscapes. A powerful autonomic response that is enjoyed is described as a thrill. Very similar responses, not enjoyed, are described as fear and trembling.

This brings us to the third and fourth widespread philosophical views of pleasure and displeasure, namely, that pleasure and displeasure are simply qualitatively distinct types of events in consciousness, or that they are events playing particular functional roles (see, e.g., Aydede 2000; Block 1995a,b; Searle 1992 for the former sort of view, and any functionalist for the latter). Common sense concurs to at least some extent with both views. Episodes of pleasure and displeasure have distinctive phenomenologies, most would agree, and they have characteristic effects and causes, and both of these allow them to be distinguished from other events in consciousness.

I do not want to say that either view is completely false. In fact, I think both are importantly right in certain respects. Clearly, nothing would be pleasure unless it had the qualitative character pleasure has for us, no matter what functional role it would play. The “raw feel” of pleasure cannot be left out of any proper theory. More controversially, I am also inclined to the view that nothing can be pleasure unless it plays (or perhaps is supposed to play) a particular role, though I think that this role is not well specified by simply gesturing to the familiar roles played by pleasure within us. But both of these views of pleasure are missing something that is also vital to a proper theory of pleasure. What is lacking in these views of hedonic tone is any room for explaining the familiar facts about pleasure and displeasure. Why is it abnormal to experience both at once? Why do pleasure and displeasure trade off against one another? Why do pleasure and displeasure come in degrees? Why do our desires determine whether a piece of news is pleasing or displeasing? Why do gut-level confidence and resignation make a difference to hedonic tone? Such questions turn out to be very difficult to answer on either view, if either is taken as the last word on pleasure.

Consider the answer offered by the simple functionalist position that pleasure and displeasure are essentially defined by their functional roles. For something to feel like pleasure, for something to be pleasure, on this view, is just for it to be a mental thing with an opposite, not normally experienced together with it, which trades off against it, which has certain common causes and other common effects, and so on. Thus, pleasure has all of these roles exactly because it is pleasure: if it had different roles, it would not be pleasure, but something else. Analogously, one could say that someone is a baseball pitcher exactly because she is the person who throws each pitch, stands on the pitcher's mound, and so on: if she were to stand elsewhere on the field and act differently, she would be a shortstop or an outfielder or the like, and not a pitcher.

The functionalist's answer presupposes a controversial view of consciousness that has been attacked elsewhere (e.g., Block 1978; Chalmers 1996), but I would like to raise a complaint that is more specific to the issue at hand. The complaint is that, on this view of pleasure and displeasure, there is no explaining why it is that pleasure plays the role it does in organisms like us. This is an unfortunate (p.87) feature of the theory. Intuitively, it seems reasonable to ask why pleasure and displeasure trade off against one another, or why it is rare to feel pleasure and displeasure simultaneously, but not impossible. Similarly, prior to any deep theorizing about consciousness, it seems like a sensible philosophical question to ask why pleasure inclines us toward maintaining the feeling, while displeasure inclines us toward ending the feeling. But such questions are ruled out of court by the functional-role theory of pleasure. All it can say in answer is that, had our brains worked differently, feelings other than pleasure and displeasure would have been responsible for these different operations. That is, pleasure has all of these features simply because it is pleasure, and if something had not had every one of these features, it would not have been pleasure. This seems surprisingly little to say. After all, it does not appear obvious that these features of pleasure and displeasure are all equally essential to the very nature of these feelings. All else being equal, a theory with more to say by way of explanation ought to be preferred, if one can be found.

Now consider the answer offered by the view that pleasure and displeasure are essentially defined by their qualitative “feels,” by the difference they make to consciousness, and nothing more. It could be held that pleasure and displeasure are opposites just because of the way they feel, and that they trade off against one another as a result. It could also be held that pleasure is, simply by virtue of its feel, motivationally attractive, while displeasure is repulsive, and so on.

The problem for such a view is that the claims about what follows directly from the qualitative character of pleasure just do not hold up. Take the claim that pleasure and displeasure are opposites. Given the common sense observation that it is possible for them to be had at the same moment (in bittersweet moments and the like), in what sense does phenomenology support the view that they are opposites? It seems rather to support only the weaker view that they are distinct. If pleasure and displeasure are only distinct, how does it follow that they must trade off against one another? Given the phenomenology, it seems that pleasure and displeasure do not need to trade off against one another: they could both exist at the same time, each waxing or waning on its own schedule. In the same vein, consider the claim that pleasure and displeasure have their motivational roles entirely in virtue of their qualitative character. This view seems to be dealt a serious blow by the commonplace observation that people can seek out displeasure exactly for its quality as displeasure. People whose self-esteem is so low that they consider it right that they suffer, and are thereby motivated to cause themselves suffering, do not appear to derive overall pleasure from their suffering.23 In what sense, then, can it be held that their displeasure is something motivating avoidance purely by virtue of its qualitative character? Left without answers to such questions, the phenomenological view of pleasure and displeasure is reduced to saying that these feelings just do play the roles they do, and so ends up with arbitrary answers to our questions. It might be thought that the defender of the phenomenological view could just describe the phenomenology of pleasure and displeasure more carefully and so answer these questions, but this is a problematic (p.88) route for the phenomenological theorist to take. For instance, one could attempt to hold that pleasure and displeasure generally trade off against one another because they are valenced, that is, because pleasure is positive and displeasure is negative in quality. But consider what this could mean. It cannot be a mere redescription of the patterns pleasure and displeasure display in trading off against one another (for then it explains nothing) and it cannot be a redescription of the fact that people are drawn to pleasure and repelled by displeasure (then it would not be part of the phenomenological character of pleasure and displeasure). To say that pleasure and displeasure are valenced in quality is, in this context, to say that there is a further way of describing the felt quality of pleasure and displeasure in terms of some independent quality, goodness or badness. And since goodness and badness trade off, so must pleasure and displeasure, at least normally. But to say this is to give up on the pure phenomenological approach and to begin to explain the facts surrounding pleasure and displeasure in terms of the qualities (goodness and badness) to which pleasure and displeasure might putatively relate us. This is to move from a pure phenomenological approach to one closer to representationalism (more on which in a moment).

The phenomenological and functional views of pleasure and displeasure thus present a number of problems. These problems are, perhaps, not as severe as the problems presented by other well-known philosophical accounts of hedonic tone, but they are genuine problems nevertheless, and a theory of hedonic tone without them would have its attractions. It is to the development of such a theory that I now turn.

4. A Representational Theory of Pleasure

Consider, for a moment, vision. It, like hedonic tone, has distinctive phenomenological features, plays a certain functional role and exists in virtue of dedicated neural structures. But it can also be said of vision that it is a tool for giving organisms a representation of the shapes and locations of objects around them, along with a representation of the light-reflecting properties of these objects (Marr 1982). That is, in addition to our ability to think of vision in terms of what it is like to see, and in terms of the neurological means by which we see, we can also think of vision in terms of what it represents. Vision involves representing objects in the world as being here or there, red or black, steady or moving, and so on. Thinking of vision as involving representations (or, if one prefers, information or intentional content) has a number of benefits.24 First, it accords with the common sense observation that vision tells us something about the world, that there is a content to visual experience that we can accept or reject. “My eyes tell me that the two lines bend away from one another, though I know they don't really” and similar expressions give voice to this sort of thought. Second, it lets us say that certain beliefs are or are not justified by virtue of their contents being in logical accord with the contents of experience. The person who apparently (p.89) sees the milk carton on the counter is, normally, required to believe that the milk carton is on the counter, because to believe otherwise would be to contradict “the evidence of his senses.” This presupposes that the senses have contents that can be contradictory to the contents of belief. Third, thinking of vision as involving representations allows us to provide informative and powerful explanations of certain facts about vision. For instance, consider the following:

The reason that human visual depth perception can be tricked by virtual reality glasses is that the portion of the brain representing depth relies on information about coincidence in the retinal images of the two eyes (among other things), processing such information on the implicit assumption that widespread coincidence in retinal image corresponds to a unique distal cause. Virtual reality glasses provide the eyes with distinct retinal images that happen to systematically coincide, causing a misrepresentation of a common distal cause.

Such an explanation makes good sense only on the assumption that one can talk about representational or informational or intentional contents in the context of human vision. This last consideration, more than any other, has been the primary motivation for neuro scientists to take up talk of representations in the senses, to the extent that they have.

Common sense does not hold that pleasure and displeasure have contents telling us about how things stand in the world: this is at least part of the reason pleasure and displeasure are called ‘feelings’ rather than ‘senses’.25 Common sense does, however, hold that hedonic tone justifies certain beliefs—beliefs about what one wants or does not want—and there is a powerful account of the diverse features of hedonic tone that can be given in representational terms. So insofar as there is a case to be made for vision involving representations, there is also a case, if not quite as strong, for holding pleasure and displeasure to involve representations. None of this is to deny a place to the phenomenology of pleasure, or to say that representations can be formed without regard for the functional role that they play. Rather, it is being suggested that such accounts need to be supplemented by a discussion of the representational features of pleasure, just as any complete account of vision needs some discussion of its representational character.26

If one wanted to give a representational account of pleasure and displeasure, what would one have them represent? The most obvious choice seems to me to be something relating to changes in desire satisfaction and frustration. In representation, there is often a co-occurrence of the representation and the represented, and changes in desire satisfaction and frustration co-occur with distinctive hedonic experiences. By and large, getting what we want pleases us, and being pleased is a sign that things are going our way. Similarly, having our desires frustrated is unpleasant, and being displeased is a sign that things are going against our wishes. Perhaps, then, pleasure and displeasure allow us to perceive whether our desires are being satisfied or not. Here is a first attempt:

(p.90) Representational Theory of Hedonic Tone (RTHT) 1: To be pleased is (at least) to represent a net increase in desire satisfaction; to be displeased is to represent a net decrease in desire satisfaction. Intensity of pleasure or displeasure represents degree of change in desire satisfaction.

I take net desire satisfaction or frustration to be a function of the satisfaction or frustration of individual desires, with stronger desires being weighted more heavily than weaker desires in the global evaluation. I also take it that net desire satisfaction may increase (i.e., one's desires may be better satisfied, overall, than they were before) or decrease (i.e., be worse satisfied) as a result of events occurring, and a positive or negative change in one's state of desire satisfaction may be greater or lesser.27

RTHT1 is a theory with many advantages. Consider first the findings of common sense surveyed in section 1. Common sense holds pleasure and displeasure to be opposites. RTHT1 agrees and can explain: pleasure and displeasure have representational contents that are contraries. Common sense holds pleasure and displeasure to come in degrees, with a neutral point between them. RTHT1 agrees and can explain: the neutral point is a representation of desire satisfaction neither increasing nor decreasing, and hedonic tone comes in a continuous range of intensities because desires can be held with a continuous range of strengths, entailing a continuous range of possible increases or decreases in net desire satisfaction.

RTHT1 can even explain the fact that pleasure and displeasure are not normally experienced together, but can be on occasion. Experiences analogous to simultaneous feelings of pleasure and displeasure can be created by what are known as ‘opponent processes’. Though it happens rarely, one can visually experience an object as both moving and stationary—an effect sometimes known as the ‘waterfall illusion’ because it can be created by staring for a long time at a waterfall and then gazing at the stationary landscape near the waterfall (Frisby 1979). Such an illusion is made possible when mutually exclusive properties are represented by the states of distinct representational structures, rather than by mutually exclusive states of a single representational structure. In such circumstances, design considerations would suggest that the representational structures should be connected so that they will not simultaneously represent the world as evincing an impossible combination of properties (i.e., that they be opponent processes), and indeed this seems to be the case with the representation of motion—under normal conditions. Under unusual conditions, however, we find that the normal controls break down, and experiences with contradictory contents result.

Experience, together with the theory presented so far, suggests that the same sort of illusion can take place in the experience of pleasure and displeasure. Pleasure says that our desires are, on balance, being better satisfied; displeasure says that they are, on balance, more frustrated. Since RTHT1 posits that pleasure and displeasure are two distinct types of representations, and since scientific evidence (p.91) suggests that they are found in adjacent regions of the perigenual anterior cingulate cortex, we need only go on to hold that these representational structures fallibly inhibit one another in order to accommodate the simultaneous representation of our desires as both satisfied and frustrated. This account does more than fit the phenomenon: it also makes sense of it. Why should it be odd to experience pleasure and displeasure at the same time? Because the two sorts of experiences say contradictory things—their contents are mutually exclusive. Why should it be possible? Because the human psyche is capable of mutually inconsistent pairs of representations, though the brain works to avoid them.

The everyday observation that pleasure and displeasure trade off against one another in something like the manner of positive and negative numbers is also endorsed and explained through RTHT1. Representations of net increases in desire satisfaction can be turned into representations of net decreases if the subject suddenly comes to believe that a powerfully desire-frustrating state of affairs obtains, and likewise a representation of a negative trend in desire satisfaction can be overthrown by becoming aware of a single very deeply desired state of affairs.

Similar progress can be made with the observation that pleasure and displeasure have both intellectual and sensuous sources, for people have both more intellectual and more sensuous desires. Some people want to see brilliant chess played or want to avoid spending time “unproductively”: many people want to be warm and are averse to overfull bladders. These intellectual and sensuous desires naturally translate into intellectual and sensuous sources of pleasure and displeasure.

The ubiquity and apparent innateness of some desires (such as desires to be warm, to be touched gently, to be loved, not to be damp, not to be unlike everyone else, etc.) is the explanation RTHT1 can offer for the corresponding ubiquity and apparent innateness of the corresponding dispositions to pleasure and displeasure. Similarly, the fact that many desires are acquired by experience is held by RTHT1 to explain how there comes to be so much individual variation in what pleases people as adults.

The category of unreal and illusory pleasures is another common-sense phenomenon that can be given theoretical underpinnings within RTHT1. To see this, begin by noting that capacities for pleasure and displeasure exist downstream of other representational capacities: we feel happy when we seem to get our way, not simply when we actually get our way. Hedonic representations do not have special access to the facts about desire satisfaction, but depend upon other representational systems to say whether or not satisfying or frustrating states of affairs obtain. As representational capacities dependent upon others for their operation, pleasure and displeasure are capable of misrepresentations of two main types: those resulting from failures (or simple non-representation) in other perceptual representation systems, and those resulting from failures intrinsic to the structure producing hedonic representations.

Consider, for instance, the person who is happy that General Electric has purchased his corporation, because he has foolishly concluded that GE shares his corporation's vision of its mission and will help the corporation to achieve it, (p.92) thanks to GE's superior resources. In fact, GE intends to ruthlessly exploit the company's resources to promote its existing line of products. The naïve employee's pleasure at the situation will be taken by his wiser co-workers to be illusory in a sense. “This is no time to be happy,” he might be told. But where is the illusion? In this case, the illusion is in the employee's cognitive system. His happiness is inappropriate, but its inappropriateness derives from no failing on its part: it derives from its dependence upon an error elsewhere in the employee's mind.

A second type of hedonic illusion would seem to stem from the system that produces hedonic representations itself. Consider the sadness that follows the end of a romantic relationship. For a time, the sad individual will be subject to hedonic illusions because the loss will cause the individual to become hedonically insensitive to changes in net desire satisfaction. Friends may rally around the person, demonstrating a degree of loyalty not previously suspected, without ameliorating the displeasure suffusing the grieving person. An unexpected promotion at work may be met with no change of mood as well, even though the promotion is found by the individual to be a significant good. Similarly, a dent or scratch given to the sufferer's car may cause no increase in displeasure. This hedonic insensitivity to changes in how well things are going for the individual is a form of hedonic blindness, on the present account. The fact that we use a language of epistemic failing in such situations—“she just can't take it in,” “he isn't seeing things clearly right now”—reminds us that interpreting these phenomena as involving some sort of representational failing accords with common sense.

As for the unreality of pleasures induced through chemical, electrical, or surgical disruption of the brain, the truth behind this thought is demonstrated by RTHT1 by holding that the use of substances such as heroin and cocaine induces a representation of a net increase in desire satisfaction, when in fact no such increase exists (in the simplest case, at least, in which withdrawal symptoms are not an issue, and in which the user has no intrinsic desire for the drug).

That the pleasure generated by euphorigenic drugs genuinely involves some sort of representational error is independently plausible on neurological grounds. As Jon Elster points out in his recent study of emotion and addiction, euphorigenic drugs “hijack” the brain's reward system (1999, 53). Most euphorigenic drugs directly or indirectly stimulate the VTA/SNpc, causing downstream effects on structures such as the PGAC, and it is believed that this is the primary route through which they have their effects on their users (KSJ 2000, ch. 51). As described in chapter 2, the VTA/SNpc is normally stimulated, not by directly infused chemical agents, but by structures carrying information about whether unexpectedly rewarding states of affairs or unexpectedly punishing states of affairs are being represented at present. Hence, in the normal course of affairs, representations of how the world is activate the VTA/SNpc, which in turn leads to pleasure or displeasure. Activation of this system by euphorigenic drugs, that stimulate it at the midpoint of its operation through biologically abnormal means, (p.93) is indeed a hijacking of the system, and naturally can be expected to lead to misrepresentation. According to RTHT1, the pleasure itself is the site of misrepresentation. Hence the unreality of the pleasures of euphorigenic drugs.28

One final noteworthy feature of RTHT1 is that it can endorse and explain the commonsensical idea that pleasure and displeasure sometimes impose epistemic constraints on what can reasonably be believed about what one desires, just as visual experiences can impose constraints on what can reasonably be believed about surrounding light-reflecting surfaces. Everyday thinking holds that a person who is otherwise in normal hedonic conditions (not depressed, not using euphorigenic drugs, etc.) and who finds herself very much enjoying it being the case that P, or very much distressed that it is the case that P, should give some credence to the thought that P satisfies or frustrates some desire (s) of hers, whether or not she believes she has such a desire. A person who has always thought of herself as an unathletic bookworm, but who finds that she gets delight from playing badminton and lifting weights, ought to think that perhaps she desires physical activity more than she knows, for example. According to RTHT1, this is so because her pleasure says that her desires are better satisfied than before, and since this superior desire satisfaction coincides with various forms of physical activity, the conclusion that a desire for physical activity is responsible for the pleasure is reasonable.29 To ignore her feelings of pleasure and insist that she does not really have any desire related to athletic activity would be akin to denying the evidence of her senses.

The virtues of RTHT1 are thus substantial. With functionalist and phenomenological theories of pleasure, it shares the intuitively plausible claim that episodes of pleasure and displeasure are distinctive types of conscious events. This makes it a more plausible theory than Ryle's behaviorist theory or Damasio's “body-state” theory. RTHT1 improves upon its closest rivals, however, in having the virtues of a representational theory. RTHT1 allows the data of common sense to be given an explanation, something functionalist and phenomenological views struggle with, and RTHT1 also makes sense of the observation that pleasure and displeasure constrain what a reasonable person ought to believe about her desires.

In spite of all these virtues, RTHT1 has at least one serious fault, for it fails to explain one important datum of common sense. Back in section 1, it was noted that one source of variation in pleasure and displeasure comes from confidence and resignation: if a person who desires that P is completely confident that P, then she will tend to be less powerfully pleased by turning out to be right, and more powerfully displeased by turning out to be wrong, than someone who was less certain that P; likewise, a person who is averse to it being the case that P, but who is resigned to it, will tend to feel less displeased if it happens that P, and more pleased if not-P, than someone not so resigned. By invoking only represented net change in desire satisfaction, RTHT1 leaves no room for confidence and resignation to play a role.

A modification of RTHT1 should take care of this problem, however.

(p.94) RTHT 2: To be pleased is (at least) to represent a net increase in desire satisfaction relative to expectation; to be displeased is to represent a net decrease in desire satisfaction relative to expectation. Intensity of pleasure or displeasure represents degree of change in desire satisfaction relative to expectations.

RTHT2 retains all of the benefits of RTHT1, but also makes room for confidence and resignation to make a difference to hedonic tone. Exactly what difference they make, however, is worthy of further discussion, for it is not completely obvious. After all, a person with a dog phobia can be completely convinced that the dog before him is harmless while nonetheless feeling terrible: why does this confidence not translate into reduced displeasure? Furthermore, being stung by a bee would seem to be unpleasant no matter how convinced one is that the bee is going to sting and one's desires are going to be frustrated. Yet the phenomenon of confidence and resignation making a difference to hedonic tone is a robust one, in need of some account. And it can be added that common sense allows people to become hardened to displeasing stimuli, even to pain, although this has never struck anyone as a matter of confidence or resignation.

A pair of distinctions must be made in order to untangle this mess: the distinction between intellectual expectations and “gut-level” expectations, and the distinction between more intellectual and more sensuous sources of pleasure. First, expectations. Some expectations about what will happen are the product of conscious reasoning, conscious probability calculation, reliance upon authorities for likelihood estimates, and so on. These are paradigms of what I am calling “intellectual expectations.” If one talks about what one believes to be the likelihood of an event, one is generally talking about one's intellectual expectations. These can be contrasted with what one believes “in one's heart” or “in one's gut” about the likelihood of various things, a likelihood estimate that need not match the estimate one makes “with one's head.” As an illustration, think of two people playing roulette, both of whom bet a small amount of money on black, both of whom win. Both have full knowledge of the objective probability of winning (17/6), which is transparent from the structure of the game. At the level of conscious deliberation, both agree about how likely the win was, but one may accept the win practically as a matter of course, being supremely confident in her luck, while the other may be quite surprised, having expected herself “at some level” to have bad luck. These differing expectations, expectations that are sometimes disclaimed at the level of rational deliberation, and that are somewhat independent of rational deliberation in any case, are the sorts of cognitive states I have in mind in talking of “gut-level expectations.” Statements such as “Of course I knew she would be late, but somehow it still surprises me” are expressions of the effects of such gut-level expectations. The same sort of phenomenon is found in parents feeling sure that a missing child is alive even while accepting the rationality of holding that there is a great risk that the child is dead, and in people who “can't quite believe” they got the great new job, romantic partner, or piece of financial luck they did. Generally, one's gut-level confidence in a given proposition matches one's consciously held level of confidence, (p.95) but the fact that these are separate cognitive states is shown by the ease with which they come apart under various familiar conditions.

The second distinction, between sensuous and intellectual sources of pleasure, should already be clear enough from its discussion in section 1. There are feelings of pleasure that stem from sexual activity, emptying one's bladder, smelling freshly brewed coffee, and so on, and feelings of pleasure stemming from contemplation of the vastness of the universe, or the intricacy of a fugue, or even from realizing that a less well funded team has finally beaten the New York Yankees; and similarly for displeasure.

When considering cases, I think the reader will find that pleasure and displeasure tend to vary with confidence and resignation when the confidence and resignation are felt in the “gut” or “heart.” Reasoned estimates of likelihood believed with one's “head” sometimes correspond to what is felt in one's gut or heart, but when the two come apart, tendencies to pleasure or displeasure follow the viscera. Examination of cases also reveals that gut-level estimations of likelihood for intellectual pleasure and displeasure are calculated differently from estimations of likelihood for more sensuous pleasure and displeasure.

Consider first more intellectual sources of pleasure and displeasure. The person who had always, in his heart, expected to be denied the job for which he applied is the person who will be less hurt by the bad news than the person, otherwise the same in temperament, desire, and so on, who had believed in his secret heart of hearts that he could not fail to get the job. RTHT2 can explain this phenomenon, so long as its term “expectation” is read as “visceral expectation” or “gut-level expectation.” Of course, there are complications in real cases. For instance, people with low confidence in the occurrence of good things are often people who are hurt worst by bad news, while people with high confidence in the occurrence of good things are often people who are most resilient to bad news. But this is because such people tend to be rather different in a large number of ways. People who tend to have low confidence that good things will happen to them are also people who tend to see each negative event as evidence that they are, more than ever, doomed to unhappiness, or as evidence that other bad things will soon happen, while people who tend to be confident that the future bodes well for them do not place such interpretations upon bad news. Naturally, this makes a difference to hedonic tone, but not one that conflicts with RTHT2. Likewise, people who have low confidence in the occurrence of good things often tend to depression, and depression is a condition in which one's hedonic tone is largely unresponsive to one's actual circumstances—a condition of systematic misrepresentation, in other words, akin to numbness in the skin or tinnitus. People who systematically misrepresent deviation from expected desire satisfaction, naturally enough, are not going to have displeasure that is readily predicted by RTHT2. A certain amount of caution is therefore required in comparing everyday cases, looking for the effects of gut-level confidence and resignation. But when this caution is employed, RTHT2 fits the evidence of common sense fairly well—at least when one considers only intellectual pleasure and displeasure.

(p.96) Pleasure and displeasure stemming from more sensuous sources are somewhat different. It is far from obvious that gut-level expectations make a difference to the pleasure or displeasure caused by a massage or a scratch on one's thumb. Yet there are well-known sources of variability in openness to sensuous pleasure and displeasure that I think are best interpreted in terms of gut-level expectations specific to them. The fact that exposure to various specific forms of hardship can inure one to them, so that one no longer feels displeasure under the same conditions one once did, is widely recognized. Likewise, it was known to the ancient Greeks as well as it is known today that sufficient indulgence in sources of sensuous pleasure leaves one jaded, so that one no longer takes the same pleasure in sensuous stimulation that one once did. Almost equally familiar are the opposing phenomena: people with particularly hard lives are sensuously pleased by what others take for granted, and people with absurdly easy lives are sensuously displeased by what passes unnoticed by others. These phenomena suggest that experience somehow sets a baseline, against which new experiences are measured, which makes a difference to felt pleasure and displeasure. I suggest that this baseline be thought of as a gut-level expectation of sensuous desire satisfaction. But these gut-level expectations are unlike gut-level expectations of more intellectual pleasure and displeasure in that they are much more resistant to change. It takes many more exposures to frigid arctic air, or to the heat of a smelter, to become inured than it takes exposures to a friendly smile from a new colleague before one comes to take it for granted. Yet experience seems to show that such changes can eventually come to one's sensuous pleasure and displeasure.

Talk of two distinct systems of gut-level expectations is rather speculative. Nonetheless, there is evidence to support it. As was shown in chapter 2, the biological reward system begins with inputs about how things stand in the world and ends with the release of a signal that expresses the current deviation from expected reward. Between the input and the output, a calculation must be made of expected rewards. If the argument at the end of chapter 2 was leading in the right direction, and desires are what make things rewards or punishments, then predictions about rewards and punishments entail predictions about desire satisfaction and frustration. Hence, the existence of biological systems predicting rewards and punishments in a manner independent of conscious reasonable judgment will be powerful evidence for the existence of biological systems creating gut-level expectations about desire satisfaction. All that is lacking is evidence that there are two distinct systems, one more concerned with intellectual desires, the other more concerned with sensuous desires. But this is also a reasonable speculation. As I noted in chapter 2, reward signals begin in two very different places: the hypothalamus, which monitors basic states such as blood sugar and hydration, and the cortex, with its higher representational capacities. Though the idea remains rather speculative, it seems plausible to me that there would be two distinct routes to the calculation of expected rewards, and hence two distinct routes to calculation of expected desire satisfaction: one stemming from the “lower” functions of the hypothalamus, the other stemming from the “higher” functions of the cortex.30

(p.97) while the notion of expected desire satisfaction is a tangled one, there thus is a way of sorting out the tangle that meets the requirements of common sense, is not biologically implausible, and can be used to improve the match between the claims of RTHT1 and common sense. This is the notion of expected desire satisfaction I will adopt for RTHT2.

To summarize the discussion so far: RTHT2 holds that a feeling of a certain degree of pleasure is a perceptual representation of one's desires being on balance more satisfied, to a corresponding degree, than expected, and an experience of a certain degree of displeasure is a perceptual representation of one's desires being on balance less well satisfied, to a corresponding degree, than expected. The fact that pleasure and displeasure represent our desires allows us to understand the epistemic value of pleasure and displeasure in learning about what we want, and the details of how desire satisfaction and frustration are represented make sense of and systematize many phenomenological facts about pleasure and displeasure. Thus clarified and illustrated, the theory is now ready to face some objections.

5. Objections

A likely objection at this point is that there seems to be pleasure and displeasure not stemming from the satisfaction or frustration of any intrinsic desire. For instance, being stroked or massaged is often pleasurable, even when it is the furthest thing from one's mind. But if one would be pleased were one massaged, then it must normally (statistically or normatively, depending on one's view of representation) be the case, on the present line of thinking, that one wanted to be massaged. Likewise, if being forced to scrub floors with a comatose terrier would be unpleasant for one, then on the present theory it appears that one must have wanted not to be forced to so scrub floors, no matter how far this possibility was from one's mind. These difficulties, however, are not quite as serious as they may initially seem, for the content of these desires needs careful characterization. The fact that one would be displeased if one were forced to scrub floors with a comatose terrier does not entail that one has a lurking desire specifically not to do that sort of thing, even on the present theory: it could be that one's desire is for not doing any of a broad class of things, into which the particular sort of act falls. For instance, one might be averse to maltreating animals or getting dirty, and scrubbing a floor with a terrier might frustrate either desire. Thus, if some sort of event is pleasing or displeasing, the present theory does not require that one has a desire regarding that particular sort of event.

Even if it is accepted that care must be taken in assigning content to the desires whose existence is implied by experiences of pleasure or displeasure, it still may seem that we are left with too many desires. Do we really have desires not to have damaged bodily tissues, desires to ingest certain classes of foods and not ingest others, desires to be massaged, to inhale odors of cedar or sandalwood (however described) but not urine or mold (however described), and so on? That (p.98) is, do we really have a large class of sensuous desires, one that can explain the many sensuous sources of pleasure that exist? I will argue that the answer is “yes.” In common parlance, babies want full stomachs, dry bottoms, cuddling, colorful sights, and soothing sounds, and these basic desires do not vanish as we grow older, though they gain a certain degree of sophistication. Why not, then, say that we have these desires? They are not particularly cognitively complex, of course, but if we are willing to say that babies and many non-human animals have desires (as I presuppose in this work), then inevitably it will prove that there are many desires lacking such sophistication. Furthermore, desires regarding sensuous states would be just the sort of thing one would expect evolution to provide infant human beings, were it capable of doing so—after all, infants have limited conceptual resources, but many sensuous states are reliable indicators of harms or benefits in the environment, and making infants desire some of these (the sound of familiar voices) and averse to others (sharp pricks to the skin) would seem only natural.31

One might also add that it is no great innovation to discover desires where none were thought to exist before: though the claim of discovery is always a little suspect, and bears some burden of proof, that burden should not be taken to be one we cannot meet. The fact that a putative desire is having no effect at the moment on one's consciousness, such as a lurking desire that gives one a liking for being massaged, is no reason to deny its existence—we surely desire love, fame, and fortune to the same extent while asleep as while awake, for instance (imagine someone saying “he doesn't care whether I exist or not” about a loved one, simply because the loved one is asleep)—and so the mere fact that one is not thinking of how nice it would be to have a massage need be no reason to conclude that no such desire exists. Likewise, the fact that one would not mention something in a list of one's desires does not show that it does not belong on such a list: perfect introspective access is something we clearly lack for most of our propositional attitudes, desires included.

If the terms ‘desire’ and ‘want’ still seem inappropriate when explaining why, for example, certain tastes please and others do not, consider the language of the pro attitudes more generally. We have likes and dislikes, hankerings, tastes, aversions, and preferences. To be told that one would find one has a taste for torn kah kai if one were to sample it is to be told that one is disposed, under normal conditions, to be pleased by eating torn kah kai; so much is uncontroversial. Such a disposition presumably has some ground. Consider, then, the nature of that ground. Tastes (and aversions, preferences, hankerings) can be expressed as tastes for particular states of affairs. Having a taste for torn kah kai is not a matter of being pleased by the existence of the Thai dish, or being pleased when others eat it, but being pleased by a type of state of affairs—one's own tasting of torn kah kai at that instant. In this respect, tastes have something propositional about them. That this something propositional is a mental representation of some sort can be seen from the fact that descriptions of one's taste preferences exhibit intensionality: if I have a taste for black beans and rice, and that is what Pancho (p.99) Villa ate before dying, though I know nothing of this connection, then it is just as incorrect or infelicitous to say that I have a taste for what Pancho Villa ate before dying as to say that I have a desire for what Pancho Villa ate before dying. Further, just as in the case of desires, one can seem to get food that is to one's taste without it really being such—for instance, under the influence of certain drugs, even marginally palatable foods can seem very much to one's taste. Desires and whatever grounds our tastes also have substantial functional features in common. Both are poised, under appropriate conditions, to move one to action (to have a second spoonful of Thai soup, for instance), both lead to pleasure or displeasure when particular states of affairs are represented by the organism, both have a legitimate role in practical reasoning, and both can be enlisted in operant conditioning.32 Given this tremendous wealth of commonalities, why not then identify the ground of a taste with a desire for stimulation of the relevant sort?

One might wish to deny that tastes involve the concepts articulated in language, while desires essentially do, and this cannot be dismissed out of hand. Yet, at the same time, such an objection strikes me as more of a terminological dispute than a dispute of genuine interest. Even if we distinguish desires from tastes, if it turns out that they have essentially the same roles in our mental economy, the sole difference being whatever it is distinguishing perceptual content from belief content, then such a distinction seems to be as much terminological as substantive. I am not one who would hold that any desire must have a content of the fully conceptual, as opposed to non-conceptual, sort (if this distinction is even legitimate), but for those philosophers who prefer to use the word ‘desire’ in this way, I would suggest that this book must be understood to be a work on the nature of desires and tastes, which I collectively call ‘desires’. The philosopher wishing to make the distinction is welcome to do so, as is the philosopher wishing to distinguish “true desires” from wishes, say, but the distinction distinguishes amongst phenomena that are more similar to one another than differing, and that form a unified group for most psychological purposes, and so I will continue to deal with them without drawing the distinction.

One more point should be raised, this stemming from the neuroscience of pleasure. At least some sources of sensuous pleasure, such as particular smells and tastes, are processed in regions of the brain very close to the reward-processing region of the OFC and send projections there (Rolls 2000). If chapter 2 was correct, then such projections to OFC are probably what constitute these smells and tastes as rewards. But if chapter 2 was correct, then exactly similar connections to OFC are probably used to constitute complex states of affairs such as the approval of one's peers, or the timely launch of a space shuttle, as rewards. That is, as rewards, smelling violets and getting a new bicycle are on a biological par. In the case of getting a new bicycle, it is natural to hold that it is a reward because one wants a new bicycle. Given that the biological underpinnings are very much the same in the case of smelling violets, why not once again say that the person for whom this is a reward is a person who desires to inhale such scents? And once this has been said, there is no bar to saying that the pleasure (p.100) felt as a result of inhaling such a scent represents an improvement in desire satisfaction.

A powerful argument is still available to the critic. Suppose that the critic accepts, as just argued, that the grounds of our tastes (aversions, etc.) for various forms of immediate sensory stimulation are similar in structure to our standing desires. The critic may then pose a difficult choice. Tastes are the sort of thing that tempt one to say “I want it because it pleases,” while for desires “it pleases because I want it” is the natural order of explanation. If the similarities between desires and other pro attitudes are so great, then one order of explanation must triumph over the other. My claim has been as follows: most of us enjoy, say, chocolate because most of us have a standing desire for a broad class of taste experiences into which the taste of chocolate falls (and of which the taste of good chocolate is the perfect exemplar, perhaps!), just as we enjoy knowing that our loved ones are flourishing because we desire that they flourish. The alternative available to the critic is to argue that we desire not only chocolate because it pleases but also the welfare of our loved ones because that also pleases. If correct, this would refute the thesis of this chapter, for one cannot make use of the concept of pleasure in saying what it is to be a desire, if one also wishes to appeal to the concept of a desire in explaining what pleasure is—ontological circularity looms. In short, if we are to treat the relation of pleasure to all pro attitudes equally, we are forced to choose between a counter-intuitive thesis about tastes and psychological hedonism. Given such a choice, I favor the unlikely conclusions about tastes—I am far more certain that I would willingly care for my father in his old age for his sake, and not for the sake of my own pleasure or avoided displeasure, than that I do not have desires for all sorts of sensuous states. Given that desires and the grounds of tastes are so similar, a decision is called for, and the decision to call the grounds of our tastes, aversions, and so on ‘desires’ seems best.

These considerations are not decisive, admittedly, but they add some further measure of plausibility to the picture being painted of pleasure and displeasure, and further measures of plausibility will have to suffice. One final point needs to be made, and then other objections will be considered.

Even if we grant that pleasure is not the only intrinsically desired end, it surely is one end that is often desired intrinsically. Sometimes, I want to have fun, and am pretty open to what other ends I pursue so long as I will have fun pursuing them. Or sometimes I want to eat something that tastes good and am pretty open to what I eat so long as it will bring pleasure. According to the present theory, the content of an experience of pleasure is that one's intrinsic desires are being better satisfied than expected; what then to say when the pertinent desire is for pleasure? There is a hint of circularity here, but only a hint; what actually exists is not circularity but hierarchy (recalling a similar hierarchy from chapter 2). To seek pleasure is to seek, not simple improvement over expected net desire satisfaction, but the representation of such improvement. Suppose one goes looking for pleasure, finds it in playing a good game of pool, and is pleased to get it. (p.101) then one represents an improvement over expectations in net desire satisfaction in being pleased, and (if one does not misrepresent) one does so because one's desires are, on balance, being better satisfied. One desire, for pleasure, is not satisfied until the pleasure begins, and so this desire cannot be the one to trigger the representation of superior desire satisfaction. That first desire to be satisfied would be the desire to play pool well (in this example) or some other desire (such as to show off to one's friends or to be dextrous) that is satisfied by good pool playing. That desire's satisfaction leads one to represent improved desire satisfaction and so to being pleased. Being pleased satisfies one's further desire to be pleased, and so one represents one's desires as satisfied to an even greater degree—one enjoys having a good time, and has a better time for it. (Compare this with the phenomenon of not noticing one is having fun, and so missing out on some of the fun one might have had, or the phenomenon of having a good time but finding it objectionable to do so in that context, and so not enjoying oneself as much as one might.) There is nothing in the present theory that forbids intrinsic desires for pleasure or aversions to displeasure, so long as these desires are not taken to be the only sort that exists.

There is quite a different objection I want to take up concerning the anticipation and recollection of desire satisfaction and frustration, and the related phenomenon of daydreaming. It often happens that the mere anticipation or recollection of desire satisfaction or frustration is pleasant or unpleasant, even when it is quite evident that actual satisfaction or frustration is still some time in the future or the past. Likewise, daydreaming can be very pleasant, even when one does not really think that the daydreamed scenario will ever be realized (and not merely in an intellectual fashion; one may pleasantly fantasize about winning a Booker Prize while knowing in one's heart that one will never even start a novel). How can the present theory accommodate these forms of pleasure and displeasure, holding as it does that they stem from represented deviations from predicted net desire satisfaction or frustration, and not from representation of merely possible changes?

One important point to make is that we may have present desires regarding these things: for example, we may presently desire that our desires will, in the future, be satisfied, and so believing that our desires will be satisfied in the future may also satisfy desires now. Indeed, to the extent that we now care about our future, the pleasure and horror of anticipation will be fully justified.33 Likewise, a person may desire to exercise her imagination, or that it be conceivable that she receive a literary prize, or that she always recall a certain special day, or similar such things. Thus, anticipation, recollection, and daydreaming may all directly satisfy desires. This is certainly not the only mechanism by which these processes modify hedonic tone, however.

Another, equally important explanation of how recollection and daydreaming (and some cases of anticipation) can lead to pleasure or displeasure assimilates this fact to the fact that misrepresentation of the world can cause pleasure and displeasure, though there are special subtleties involved. Consider what is involved (p.102) in anticipation, dread, reminiscence, daydreaming, and the like. One uses one's imagination, or even engages in overt behavior, calling to mind the desire-satisfying or desire-frustrating scenario. One visualizes relaxing on the beach, or hears in one's “inner ear” the applause of the crowd—or its humiliating laughter. Those with good olfactory imaginations can actually “smell” the home-cooked bread before it is begun, and the memory of a lover's kiss may recall a tingle to the lips. Likewise, in imagining, one may alter one's bearing, tighten or relax particular muscles, make gestures, mumble phrases and otherwise mime, recreate or “pre-create” some event. These are voluntary behaviors, both inner and outer, and we can hardly fool ourselves with them into thinking that the imagined scenario is really happening. In this sense, they involve no misrepresentation on the part of the person. Yet I would suggest that, to the extent that they evoke hedonic responses, they have fooled some internal mental system, just as the visual system may be fooled by optical illusion-producing designs even though the person is not misled. Kosslyn (1994) tells us that many of the same cortical areas involved in perception are used in imagination. If so, then it should be no surprise that voluntary stimulation of the perceptual system should lead the brain to respond, to some degree, as though perception were occurring—after all, the neural activation created by the imaginative process is going to try to follow the channels it usually follows, even if the fact that one is voluntarily producing this stimulation changes the context in which this activation flows. Just as one can cause one's heart to race by imagining a frightening scenario vividly or cause one's muscles to tense by imagining vigorous activity, so, I would suggest, one can cause internal systems to misrepresent desire satisfaction or frustration by imagining scenarios that would be satisfying or frustrating, and so experience pleasure or displeasure. To the extent that one is not for a moment fooled by one's imagination, one can expect the pleasure or displeasure to be weaker, for not being fooled is at least partly a matter of one's voluntary stimulation of one's sensory system being imperfect, and this imperfection has weakened “downstream” effects as a natural consequence. To the extent that one's imaginative state approaches a reverie or method acting, however, one may expect a concomitant increase in emotional impact, for the imaginative state comes closer to fooling, not just a limited selection of neural subsystems, but oneself.

Finally, there are the puzzles raised by common sense that must be addressed. Where, if anywhere, is pleasure located? And how many sorts of experiences make up pleasure and displeasure?

First, the location of pleasure and displeasure. One might think RTHT2 should recognize pleasure and displeasure as having locations, at least on some occasions, as common sense does (pleasure is in the genitals during orgasm, displeasure is in a wound during pain, and so on).34 This would certainly be a problem for RTHT2 if the demand were reasonable. However, I think close phenomeno-logical attention will reveal that pleasure and displeasure are not themselves localized in such cases: it is their sources that are. One is aware of the location of a pleasant touch, but when the touch ends, the pleasure may continue without either (p.103) being found in the touched body part or being forced to relocate itself. Likewise, the unpleasantness of a scraped knee can disappear without anything seeming different about one's knee, as is shown by the results of cingulumotomies for pain control and the like (discussed earlier). Once again, hedonic tone seems not to be localized so much as to have a very salient source. And the non-localization of more intellectual pleasure and displeasure is obvious upon examination. That pleasure and displeasure are not localized makes them an oddity among the senses, but since their contents are that one's desires are on balance better or worse satisfied relative to expectations, their having felt locations would be even odder.

Second, and more challengingly: How many sorts of experiences make up pleasure and displeasure? More carefully, it can be asked whether pleasure and displeasure are what might be called ‘natural phenomenological kinds’, or whether they are heterogeneous groupings of diverse experiential phenomena. The person raising this sort of question typically has in mind the dissimilitude of having an orgasm, on the one hand, and enjoying a play by Brecht, on the other—or, equivalently, the vast gulf between finding oneself displeased with a child's hygiene and quaking with fear at the prospect of summary execution. Pleasure and displeasure appear in so many forms, it is sometimes thought, that there is nothing in common to all of them which can be called ‘pleasure’ or ‘displeasure’ and treated as a discrete entity. This is a puzzle raised, not as a result of arcane prior philosophical commitments, but as a straightforward result of noting that different experiences of pleasure and displeasure are genuinely very different.

The worry that pleasure and displeasure might not be a unified phenomenon is not wholly unreasonable phenomenologically, but it is ultimately unwarranted. Three possible sources of this worry come to mind: the fact that pleasure and displeasure are different from one another (call this ‘valence incomparability’), the fact that some forms of pleasure and displeasure seem not to admit of comparison as greater than, equal to, or less than other forms (call this ‘quantitative incomparability’), and the fact that the total character of some (un)pleasant experiences is so different from the total character of other (un)pleasant experiences (call this ‘qualitative incomparability’).

I will answer each in turn, beginning with the first and most readily answered worry—that pleasure and displeasure are not unified types of experiences because they differ from one another, and so should be thought of as being at least two distinct types of experiences, one with positive valence and the other with negative. This is, of course, true in a perfectly good sense, but it is no impediment to holding pleasure and displeasure to also be unified types of experiences in another, equally good sense. Just as experiences of warmth and cold are one sort of experience,35 so it can be said that experiences of a range of pleasure and displeasure make up one kind of experience. No one, I take it, is inclined to deny that experiences of temperature are of a kind because they come in two types, experiences of warmth and of coldness, and hedonic tone should be no different. Displeasure is instantly recognized as the counterpart of pleasure, just as cold is (p.104) instantly recognized as the counterpart of warmth, and this would not be the case unless there were some obvious phenomenological relation between them, in addition to whatever differences there are. Valence incomparability is not really in-comparability after all.

Perhaps the troubling phenomenology is not that hedonic tone comes in positive and negative types, but that they are not always quantitatively comparable. How, one might ask, can one compare the degree of pleasure of knowing one is loved to the degree of pleasure one gets from sinking into a hot bath? But think of the feeling of freezing at a bus stop in January in Winnipeg, say, and of having an ice cube pressed to one's belly while sunbathing in July, and you will probably find it hard (as I do) to imagine the two experiences in a way that makes their felt coldness just the same. Yet this would hardly convince anyone that there are really many different feelings, different in kind, which we lump together under the word ‘cold’. The fact that it can be hard to compare types of experiences in respect of their variation along some quantitative dimension is common to many classes of experiences, and not unique to pleasure and displeasure. As a result, it can hardly be sufficient grounds for holding pleasure and displeasure to be disunified phenomena. Furthermore, it is clear that, however many hard cases of quantitative comparison there might be, there are many more easy cases. For any (un)pleasant experience one has had, one can always (as a matter of empirical fact) recall many (un)pleasant experiences that were intuitively very different sorts of experiences, and yet were clearly either more or less (un)pleasant. I find it hard to compare the displeasure I experienced sledding into a brick wall to the displeasure I experienced the day my best friends ridiculed me, but I can compare them both quite readily to the displeasure I experienced eating a very inferior crème caramel (that displeasure was less) and to the displeasure I suffered upon learning of a death in the family (that was greater). Once again, it seems that the difficulties of quantitative comparisons reveal nothing very deep about the nature of hedonic tone. Comparing the redness of paint swatches or the pitch of various sounds presents all the same problems, which just goes to show how insignificant such problems are to a theory of hedonic tone.

As for the third worry, it certainly must be granted that, say, feeling cheerful because it is a sunny day and savoring the sublime elegance of Creation are different total feelings, with different qualitative characters, but I would resist the conclusion that they are such because they involve qualitatively different types of pleasure. Compare the total experience one has while hearing a teething baby cry at a certain pitch to the total experience one has while hearing a recording of a mass sung by the Vienna Boys' Choir when the choir hits the exact same pitch. There is a world of qualitative differences between the two experiences, and yet there is no temptation to attribute the difference to a difference in the sort of pitch experienced. Rather, the difference is clearly due to volume, timbre, context, pleasure, one's experience of one's bodily response to both experiences, and so on. Without special motivation to do otherwise, there seems every reason to treat the hedonic case in just the same way. One could also wonder what it would (p.105) be like to have the exact experience of enjoying the sublimeness of Creation except for the hedonic component, substituting instead the hedonic component of being cheerful on a sunny day. Holding feelings of energy, of the viscera, of muscle tone, of having particular thoughts, and so on constant, would there really be a way of enjoying the sublimeness of Creation with sunny-day pleasure rather than the usual sublime-Creation pleasure? I cannot imagine it myself.

Finally, the results of studies on the PGAC further confirm that pleasure and displeasure are unified phenomena. Studies on displeasure identify the PGAC as a displeasure center whether they use disgusting images (Lane et al. 1997), pain (Coghill et al. 1999) or guilt (Shin et al. 2000), and studies on pleasure similarly stimulate the PGAC whether through alcohol (Ingvar et al. 1998) or fond memories of competitive victory (Rauch et al. 1999). Though one could resist such evidence if one had excellent reasons to think that pleasure and displeasure are deeply disunified (one could use such evidence to show that the PGAC must not be a dis/pleasure center), in the absence of such reasons the scientific data provide yet another confirmation of the view being presented here.

6. Pleasure and Theories of Desire

If RTHT2 is correct, then no hedonic theory of desire can be correct. Any hedonic theory of desire holds that to be a desire is to be a thing which is X, where X is some property involving pleasure and displeasure. A few possible variants of the theory were mentioned in chapter 1, but whatever form a hedonic theory of desire might take, it will hold that pleasure and displeasure are part of the onto-logical ground of desire: desires are made of pleasure and displeasure, which are more basic ontological entities, though not necessarily maximally basic entities.

Given this, it is clear that no hedonic theory of desire can accept that pleasure and displeasure represent an organism's own desires. For a representation of X is less ontologically basic than an X: if pleasure represents desire, then it must be possible to say what desire is without mentioning pleasure, while it will be impossible to explain pleasure without mentioning desire. This finding is a simple one, and yet it deals a very serious blow to the hedonic theory of desire.

On the other hand, the standard theory of desire and the reward theory of desire are both left as viable theories. If pleasure represents changes in desire satisfaction, and desire satisfaction amounts to the obtaining of states of affairs the organism was inclined to bring about, that is all well and good. Likewise if desire satisfaction amounts to the obtaining of states of affairs constituted by the organism as rewards. In either case, there will be no ontological circularity in the nature of pleasure or desire.

There is one point of interest to note here, however. As described earlier, the PGAC receives input from the reward system, which appears to be a normal proximal cause of pleasure. And the reward system releases a signal carrying information about the difference between expected and actual reward. If the reward theory of desire is correct, then the reward system's signal also carries information (p.106) about the difference between expected and actual desire satisfaction. So the capacity of the PGAC to represent changes in expected desire satisfaction is straightforward. On the other hand, if the motivational theory of desire is correct, then the PGAC must somehow be drawing conclusions about the attainment of states of affairs the organism was motivated to obtain based upon information about the difference between expected and actual reward. This is not impossible, since in many cases rewards are things an organism is motivated to obtain, and things the organism is motivated to obtain are rewards for it. Reward information is thus indirect information about the obtaining of states of affairs the organism is moved to bring about. Yet this would seem to be an odd way of arranging to represent change in desire satisfaction. If the present chapter has been right, then, the hedonic theory of desire is in serious trouble, and the standard theory of desire raises an odd puzzle. Only the reward theory of desire escapes unscathed. Once again, the consideration is not decisive by any means. Yet once again, it is certainly of interest.


(1.) Also known as positive and negative hedonic tone, positive and negative affect, positive and negative emotional valence, and more. Scientists, especially, have felt compelled to add new jargon to refer to pleasure and displeasure, but I think the simplest terms are equally clear and unequivocal, and will generally prefer them. One exception: for lexical variety, I will sometimes refer to pleasure and displeasure together as ‘hedonic tone’.

(2.) Though I am acquainted with a few individuals in whom they are much more common.

(p.188) (3.) Every biologically intact person, at least.

(4.) A related phenomenon (pointed out to me by William Lycan) is the use of a coin toss to determine what one wants. One narrows the options to two, assigns each to one face of the coin, and tosses the coin. If one is pleased with the way the toss has come up, one can conclude that that is what one most wanted to do all along. If one is displeased (and perhaps wishes to make some excuse to “try again”) then one has good reason to believe one prefers the other option. Again, pleasure and displeasure provide evidence for what we want.

(5.) There may be unconscious pain and other unconscious forms of pleasure and displeasure, but these are not generally recognized as unambiguously painful, pleasant, or unpleasant by common sense.

(6.) The brain region under discussion here is quite a small one, at most a few square centimeters on the inner surface of each cerebral hemisphere.

(7.) There is also a limited amount of philosophical agreement. The only philosopher who, to my knowledge, has investigated the neural realization of displeasure has come to a similar view. See Hardcastle (1999, ch. 5).

(8.) If one prefers an identity theory of pleasure and displeasure, one can read ‘realizer’ in the degenerate sense in which everything realizes itself.

(9.) Further from your eyes, closer to the back of your head, in Brodmann's area 24′

(10.) In Brodmann's area 25.

(11.) Although there are deep waters here regarding knowledge of other minds and the like, I propose to skirt them and deal simply with the scientific approach to finding biological bases for mental events.

(12.) Looking for a last link in a neural chain is also rendered difficult by the fact that a great many neural regions are reciprocally connected and mutually influential. Strict chains are a rarity, though functional approximations to them are somewhat more common.

(13.) Berridge (2003a, 2003b) clearly maps out the failings in existing research.

(14.) But see Dennett (1978, ch. 11) and Hardcastle (1999).

(15.) Bilateral destruction, to be precise.

(16.) E.g., Hsieh et al. (1995), Rainville et al. (1997), Coghill et al. (1999), Casey (1999).

(17.) Functional magnetic resonance imaging: a technology for forming images of regions of the brain that are being used at the time.

(18.) Positron emission tomography: a technology with the same function as fMRI, though a different method of operation.

(19.) See, e.g., Weiskrantz (1986) on cortical blindness.

(20.) The ventral pallidum projects to the mediodorsal nucleus of the thalamus (Berridge 2003a), which, in turn, projects to the PGAC (Neafsey et al. 1993).

(21.) E.g., Kenny (1963).

(22.) Damasio's view of the emotions in general is one he takes to be derived from the work of William James (1890), and indeed the parallels are obvious.

(23.) Chapter 4 will also have a good deal to say about the extent to which the link between pleasure and motivation has been overestimated.

(24.) Discussed by Block (1995b), Carruthers (2000), Dretske (1995), Lycan (1987), McDowell (1994), Peacocke (1983), and Tye (1995), among others.

(25.) But it must also be said that common sense hints there might also be something more: if pleasure can really be illusory, then it must have content after all.

(26.) There are some, such as Dretske (1995), Lycan (1987), and Tye (1995), who will hold that a representational account of pleasure will entail everything important about the functional role of pleasure and everything about the qualitative character of pleasure. Though I find this view appealing, I am not committed to it here.

(p.189) (27.) It may be the case that Plato holds a similar view in the Philebus 33e ff, and that contemporary emotion theorist Nico Frijda also supports a related view in his “Emotions and Hedonic Experience” (1999). It is also interesting to compare this view to that found in Millgram (1997), in which Millgram holds that pleasure is a “rock bottom judgment of desirability” (115).

(28.) Note that calling such pleasures ‘unreal’ is not to pass any non-epistemic judgment upon them. I profit from and enjoy the illusions created by filmmakers on a regular basis. Whether the illusions produced by euphorigenic drugs are good or bad is quite independent of their status as illusions.

(29.) Several other conclusions might also be reasonable, of course, depending upon the individual.

(30.) It may interest the reader to compare the idea that there are two fundamentally different routes to the felt pleasure caused by desires to Davis's (1986) thesis that there are two different types of desires, one more linked to the appetites, the other more linked to higher cognitive processes.

(31.) Thanks to an anonymous referee of Schroeder (2001b) for this point.

(32.) This is the primary role given to desires by Dretske (1988).

(33.) Thanks to an anonymous referee of Schroeder (2001b) for this point.

(34.) Thanks to Andrew Melnyk for raising this point.

(35.) But see Akins (1996) for an account of this sense that makes it more problematic than one might have thought.