The Unity and Commensurability of Pleasures and Pains


Ole Martin Moen
Philosophia (Springer), 41 (2) 2013: 527-543 (link)


Abstract: In this paper I seek to answer two interrelated questions about pleasures and pains: (i) The question of unity: Do all pleasures share a single quality that accounts for why these, and only these, are pleasures, and do all pains share a single quality that accounts for why these, and only these, are pains? (ii) The question of commensurability: Are all pleasures and pains rankable on a single, quantitative hedonic scale? I argue that our intuitions draw us in opposing directions: On the one hand, pleasures and pains seem unified and commensurable; on the other hand, they do not. I further argue that neither intuition can be abandoned, and examine three different paths to reconciliation. The first two are response theory and split experience theory. Both of these, I argue, are unsuccessful. A third path, however—which I label “dimensionalism” —succeeds. Dimensionalism is the theory that pleasure and pain have the ontological status as opposite sides of a hedonic dimension along which experiences vary. This view has earlier been suggested by C. D. Broad, Karl Duncker, Shelly Kagan, and John Searle, but it has not been worked out in detail. In this paper I work out the dimensionalist view in some detail, defend it, and explain how it solves the problem of the unity and commensurability of pleasures and pains.


Introduction: Two Opposing Intuitions

Think of these three pleasurable experiences: The taste of ice cream, the feeling of being loved, and the excitement of reading a detective story. Do these experiences share a single quality that accounts for why they are all pleasures? Similarly, think of these three painful experiences: The searing burn after having touched a hot stove, the sting of a pinprick, and the feel of a pressing headache. Do these experiences share a single quality that accounts for why they are all pains? This is the problem of pleasure and pain unity. Moreover: Are all pleasures and all pains rankable on a single, quantitative hedonic scale? This is the problem of pleasure and pain commensurability. These two problems—which, as we shall see, are closely interrelated—are jointly the topic of this paper.

When we reflect on this issue, our intuitions seem to draw us in two opposing directions. On the one hand, pleasures and pains seem unified. Looking at pleasures first, it seems that the taste of ice cream, the feeling of being loved, and the excitement of reading a detective story—although these differ in many respects—do share a certain quality (perhaps a certain kind of positive buzz), and it seems to be by virtue of sharing this quality that they are pleasures and that we are able to reliably pick them out as such. Pleasures, at least, do not seem to be an arbitrary demarcated group of experiences, and children do not need to learn rules for figuring out what is pleasurable and what is not. Children sense what is pleasurable because of how pleasures feel. The same goes for pains: It seems that burns, pinpricks, and headaches—although they also differ in many respects—share a certain quality (perhaps a certain kind of negative buzz), and it seems to be by virtue of sharing this quality that they are pains and that we are able reliably to pick them out as such.

We also seem to think of pleasures and pains as if they were in some sense commensurable, for we commonly rank them in terms ofmore and less. It makes sense to say that eating bread is less pleasurable than eating cookies, but more pleasurable than eating flour. It also makes sense to say that jamming one’s finger is painful, but less painful than surgery without anesthetics, and more painful than a pinprick. It even makes sense to say of an activity such as eating bread that it moves from being pleasurable (the first four slices), to being neutral (the fifth and sixth slice), to becoming positively painful (stuffing in bread past the seventh slice). Pleasure and pain seem to exist on a continuum, and when people are asked to fill out the McGill Pain Questionnaire, they are presumably not dumbfounded when asked to rank their pain on a scale from zero to five.[1]

As such, there is something intuitive about the view that pleasures and pains are unified and commensurable. Murat Aydede summarizes this view as follows:

Since [pleasure and pain] are opposites of each other in some sense and admit of degree, they are thought to constitute a continuum at the one end of where there is the pleasure-sensation of increasing intensity, and at the other, there is the pain-sensation of varying degrees again. As you move toward the middle, the intensity of both pleasure and pain decreases till the vanishing point which constitutes indifference. [2]

Our intuitions do not exclusively draw us toward unity and commensurability, however, and particularly among philosophers, the view that pleasures and pains are unified and commensurable has fallen into disrepute. The standard objection is that though we might speak of pleasures and pains as if they share a unifying property, they do not, and since they do not, they can hardly be commensurated in terms of this (non-existent) property. This objection is called the heterogeneity objection.

To understand the force of the heterogeneity objection, one must understand the inclusive usage of the terms “pleasure” and “pain” that is common in philosophy (and that I shall take for granted in this paper). John Locke is a proponent of this inclusive usage: Locke explains in Essay Concerning Human Understanding that he uses the terms “pain” and “pleasure” to refer “not only [to] bodily pain and pleasure, but [to] whatsoever delight or uneasiness is felt by us.”[3] Henry Sidgwick similarly explains in Methods of Ethics that “pleasure” includes “every species of ‘delight,’ ‘enjoyment’ or ‘satisfaction’ …, the most refined and subtle intellectual and emotional gratifications, no less than the coarser and more definite sensual enjoyments.”[4] Leonard Katz, on a similar note, explains that “Pleasure, in the inclusive usages most important in moral psychology, ethical theory, and the studies of mind, includes all joy and gladness — all our feeling good, or happy. This is often contrasted with similarly inclusive pain, or suffering, which is similarly thought of as including all our feeling bad.”[5]

This inclusive usage gives the heterogeneity objection momentum. Consider and compare the pleasures we get from the following activities, all of which are pleasures in the inclusive sense of the term: Being massaged, eating candy, smelling fragrance, scoring a goal in a football match, listening to Rachmaninoff, having self-esteem, reading a well-crafted philosophy paper, and being in love. These experiences appear to be qualitatively very different. So is it clear that there is a single quality running through all of them? Socrates, in the Philebus, thought not:

If one just goes by the name, then pleasure is one single thing, but in fact it comes in many forms that are quite unlike each other. Think about it: we say that a mad man gets pleasure, and also that a sober-minded person takes pleasure in his very sobriety. Again, we say that a fool, though full of foolish opinions and hopes, gets pleasure, but likewise a wise man takes pleasure in his wisdom. But surely anyone who said in either case that these pleasures are like one another would rightly be regarded as a fool.[6]

Derek Parfit would avoid being regarded as a fool:

Compare the pleasure of satisfying an intensive thirst or lust, listening to music, solving an intellectual problem, reading a tragedy, and knowing that one’s child is happy. These various experiences do not contain any distinctive common quality.[7]

So would Fred Feldman:

Consider the warm, dry drowsy feeling of pleasure that you get while sunbathing on a quiet beach. By contrast, consider the cool, wet, invigorating feeling of pleasure that you get when drinking some cold, refreshing beer on a hot day. … they do not feel at all alike.[8]

Turning from pleasures to pains, imagine and compare the following: headaches, car sickness, muscle cramps, paper cuts, nightmares, toothaches, hangovers, hunger pangs, guilt, freezing, burning, boredom, and the smell of milk gone sour. These experiences also appear to be qualitatively very different, and it seems no clearer in the case of pains than in the case of pleasures that they share a unifying quality. As writes Rem B. Edwards, “[t]he disagreeable feeling of intense grief over the death of a loved one is just not the same kind of disagreeable feeling as that of a burn, a bee sting, or toothache.”[9] Pains seem to be radically different from one another, and even simple sensory pains—pains as recognized by Crisp and Rachels (see footnote 5)—vary in ways that seem to defy strict quantification. Sensory pain is not a single feeling that, when present, varies solely in terms of more and less. Sensory pain can be pulsing, throbbing, flashing, shooting, pricking, stabbing, wrenching, sore, numb, tearing, etc., and these differences are qualitative, not quantitative.

As such, Edwards suggests a pluralist account according to which “pleasure” and “pain” have a “variety of referents rather than a single referent.” The belief that pleasures and pains are unified, he claims—echoing Socrates—stems from the naïve assumption that what goes under one name must share one unifying quality. In Edwards’ view, “the word ‘pleasure’ refers to many different inner qualities of feeling which we find interesting and desire to sustain, cultivate, and repeat; and the word ‘pain’ refers to many different inner qualities of feeling which we find objectionable and desire to terminate and avoid.” The best we can hope for, in Edwards’ view, is therefore a Wittgensteinian family resemblance relation between various pleasures and pains.[10]

If the best we can hope for is family resemblance, then unity and commensurability are threatened, since on this view, it is not true that all members of the group pleasures (or pains) share a property that accounts for why these, and only these, are members of the group pleasures (or pains). Moreover, if pleasures (or pains) do not share a unifying property, they cannot be commensurated in terms of this (non-existent) property.

One way to respond to this is to concede that pleasures and pains are in fact not unified and commensurable, and that in treating them as such, we make a mistake. If, however, we are not willing to give up our intuition about unity and commensurability, there are logically two ways to proceed. The first option is to claim that the heterogeneity objection is without merit, and that pleasures and pains are in fact homogeneous. I believe this option is off the table, since it is undeniable that there is a great diversity among pleasures and pains. A second option is to argue that, in some sense, heterogeneity is compatible with unity and commensurability. I believe this is a more tenable approach, and in the following I shall present and assess three theories that seek such reconciliation. I shall first look at response theory and split experience theory, and argue that these are unsuccessful. Thereafter I shall introduce a third theory, dimensionalism, and argue that this theory succeeds.


Response Theory

Response theory is a cluster of views according to which pleasure and pain experiences are unified and commensurable, not by virtue of sharing a certain experiential quality, but by virtue of prompting certain responses.[11] To my knowledge, the earliest formulation of response theory is found in the Henry Sidgwick’s Methods of Ethics. Sidgwick writes:

[W]hen I reflect on the nature of pleasure,—using the term in the comprehensive sense which I have adopted …,—the only common quality that I can find in the feelings so designated seems to be that relation to desire and volition expressed by the term “desirable” … I propose to define Pleasure … as a feeling which, when experienced by intelligent beings, is at least implicitly apprehended as desirable or—in cases of comparison—preferable.[12]

There are several versions of response theory. William Alston argues that pleasure is a unified experience by virtue of being “an experience which, as of the moment, one would rather have than not have.”[13] A similar view is defended by L. W. Sumner, who argues that “what all pleasures share is not a homogeneous feeling tone, but the fact that they are … objects of some positive attitude on our part.”[14] This view is also suggested by Rem B. Edwards alongside his family resemblance view. “‘Pleasures’ and ‘pains,’” Edwards writes, “are feelings which in the former case we wish to sustain and repeat and in the latter we wish to eliminate and avoid.”[15]

Response theory offers a possible way to reconcile heterogeneity with unity and commensurability. First, response theory has no problem accounting for heterogeneity. Since it locates unity not in the quality of our experiences, but in our responses to our experiences, it places no restrictions on how diverse our experiences may be. Admittedly, the response in question can be glossed in different terms (in terms of affect, want, like, desire, etc.). Regardless of what our favorite gloss is, however, response theory seems to offer a way out of the problem at hand, since all likely glosses appear to admit of unity and commensurability. Let me exemplify this using “desire.” All species of desiring have a property in common that accounts for why these, and only these, are desires: a certain attraction and repulsion. This unifying property, moreover, admits of commensuration, since every attraction and every repulsion, regardless of its other properties, has a certain strength or pull, and this strength or pull exists in terms of more or less. As such, it seems that response theory can reconcile heterogeneity with unity and commensurability.

A problem with response theory, however, is that it appears to be the solution to the wrong problem: It appears to be the solution to the problem of whether or not attraction and repulsion (or whatever response one chooses) are unified and commensurable, not to the problem of whether pleasures and pains are unified and commensurable. The only way in which response theory could be an argument for the unity and commensurability of pleasures and pains is if pleasures and pains were just responses: if a pleasure were a pleasure by virtue of its attractive force and a pain were a pain by virtue of its repulsive force.

This position, which I shall label strong response theory, is held by some. Richard Brandt, for example, argues that “for an experience to be pleasurable is for it to make the person want its continuation.”[16] Similarly, Richard Hall argues that “The unpleasantness of pain sensations consists in their being disliked,”[17] and Chris Heathwood suggests that “a sensation S, occurring at time t, is a sensory pleasure at t iff the subject of S desires, intrinsically and de re, at t, of S that it be occurring at t).”[18] Christine Korsgaard also defends a species of strong response theory. Korsgaard writes:

The painfulness of pain consists in the fact that these are sensations which we are inclined to fight … If the painfulness of pain rested in the character of the sensations . . . our belief that physical pain has something in common with grief, rage and disappointment would be inexplicable. For that matter, what physical pains have in common with each other would be inexplicable, for the sensations are of many different kinds. What do nausea, migraine, menstrual cramps, pinpricks and pinches have in common, that makes us call them all pains?[19]

Strong response theory, as suggested by Brandt, Hall, Heathwood, and Korsgaard, does offer a possible solution to the problem of the unity and commensurability of pleasures and pains. It does so, however, at a high cost, and I shall now argue that the view is almost certainly false.

A first problem is that strong response theory appears to get things backwards. To see why, we can approach the relationship between pleasure and desire with a Euthyphro question: Do we desire things because they are pleasurable, or are things pleasurable because we desire them? Think of pancakes. When you desire pancakes, do you desire them because they are pleasurable or are they pleasurable because you desire them? Introspection favors the former over the latter: You desire pancakes because of their pleasurable taste. Indeed, accepting the alternative view seems to have an awkward implication: If things are pleasurable by virtue of being desired, then we can never use “because it is pleasurable” as an explanation of why we desire something, and the statement “I desire pancakes because they are pleasurable” would be empty, amounting to “I desire pancakes because I desire them.” The only explanation that could be given granted strong response theory is an explanation of this form: “I desire pancakes because of their sweetness.” This is an explanation, but it leads to a regress, for why does one like sweetness? At every point, the strong desire theorist must answer “because I desire it.” In criticizing this view, Andrew Moore argues that it is “hard to see how merely directing one joyless entity at another might constitute a joyful whole,”[20] and in T. L. S. Sprigge’s view, strong response theory ends up with “a strikingly joyless picture of pleasure.”[21] The picture is joyless since, if it is correct, the reason why we desire something is never that it gives us a good feeling, but always merely the fact that we are drawn towards it.[22]

A second problem is that strong response theory makes it a necessary truth that we desire all pleasures and are averse to all pains. Though pleasure and desire, and pain and aversion, are intimately related, it seems that we can both fail to desire a pleasure and fail to be averse to a pain. Think, for example, of masochism. It also seems that we can experience (mild) pain without desiring to end or weaken it and (mild) pleasure without desiring that it continues. If this is right, then pleasure and pain are conceptually independent of desire and repulsion.

A third problem is that response theory makes desire-satisfactionism and hedonism identical theories. Though this might not be a fatal implication, it is an odd one, since having one’s desires satisfied seems to be different from experiencing pleasure. I can imagine feeling pleasure without having my desires satisfied (if I am pleasantly surprised, say, and I do not have time to form any desires) and having my desires satisfied without feeling pleasure (if I have long desired something, yet find that having my desire satisfied does not give me the pleasure I expected). In conjunction with the three earlier worries, this mounts a considerable weight against strong response theory.[23]

We should ask, however, what could plausibly motivate strong response theory. I see three sources of motivation. The first source of motivation is that desires often correlate with, and sometimes contribute to elevating, the hedonic level of an experience. If one strongly desires a bottle of special French wine—say, one has been waiting a month to have it delivered and has dreamed about it at night—one’s desiring is likely to influence how good one finds that the wine tastes. If one pays attention to such cases, however, I believe one will realize that what goes on is that the desiring works causally as a factor that raises the hedonic level of the experience. It is by virtue of altering the way one experiences the wine that the desire becomes significant. In and by itself, the desire would not be pleasurable. As Aaron Smuts has pointed out, desiring by itself is often more painful than pleasurable.[24]

A second source of motivation is that strong response theory helps solve cases such as the coffee paradox. The coffee paradox is the curious fact that coffee tends to taste bad when you are a child but good when you are an adult, even though qualitatively, coffee seems to taste the same at both stages. Coffee, it appears, has the same taste when you are a child and when you are an adult—it is just that when you are an adult, you find its taste pleasurable. This paradox might lend support to the view that the pleasurability of an experience is not intrinsic to the experience. If pleasurability were intrinsic to our experiences, then presumably the pleasure and pain element in the experience could not change without the quality of the experience changing. If response theory is correct, however, the coffee paradox is not a paradox at all: As adults, we simply come to desire a new gustatory quality. This might count in favor of response theory, but as I shall argue below, response theory is not the only theory that has resources to resolve the coffee paradox.

Apart from the fact that desire satisfaction tends to be pleasurable and its solution to the coffee paradox, it therefore seems that the only motivation for holding strong response theory is that it helps solve the heterogeneity problem. That, however, is not a motivation that lends epistemic support to the theory. Thus, if neither the fact that desires correlate with pleasures nor the coffee paradox provides us with a strong reason to favor response theory over competing theories, it seems that strong response theory should be rejected in light of its oddities. If so, we might retreat to weak response theory, but that theory is not relevant in this context, since weak response theory is a theory about desire and repulsion, not about pleasure and pain.


Split Experience Theory

Let us now turn to a theory that seeks to reconcile homogeneity with unity and commensurability without locating unity and commensurability in our responses to our experiences: split experience theory. According to split experience theory, our experiences have two components: One qualitative component (which is heterogeneous, disunified, and incommensurable) and one hedonic component (which is homogeneous, unified, and commensurable). The most famous advocate of split experience theory is Jeremy Bentham. In Principles of Morals and Legislation, Bentham concedes that in one respect, our pleasure and pain experiences are heterogeneous. There are, he writes, “pleasures of sense, pleasures of wealth, pleasures of skill, pleasures of power, pleasures of piety,” and these all have different qualitative feels.[25] The crux, however, is that although they all have a qualitatively different feel, they differ only quantitatively with respect to their pleasurability. The “pleasure” part of an experience, Bentham argues, is something separate from the qualitative experience; it “accompanies,” is “derived from,” “results from,” or is “produced by” our qualitative experiences. If Bentham is right, then if you have a headache, you do not just have one experience, but two experiences: A certain qualitative feeling in your head in conjunction with a certain hedonic level attached to that qualitative feeling.

To visualize Bentham’s theory, imagine that you have an inner hedonometer. Depending on what sensory inputs you have, the marker on the hedonometer goes either up or down, or it stands still, giving you an experience at a certain hedonic level. The hedonometer can be bombarded with all kinds of heterogeneous information from your senses, but it still makes a sum of these experiences, so that the hedonometer, at any given time, marks a certain hedonic level. In this respect, the hedonometer is just like a thermometer. A thermometer can also receive a lot of heterogeneous information—from, say, sunbeams, boiling water, and ice cubes—but irrespectively of the heterogeneity of the input, condense all the information into a certain point on a quantitative scale.

If our experiences of pleasure and pain are like Bentham suggests, then we can have non-hedonic experiences that vary qualitatively and hedonic experiences—produced by and attached to these—that vary quantitatively. Bentham’s theory can thus save heterogeneity since it makes room for heterogeneity on the qualitative side of our experiences. It can save unity, moreover, since it makes room for unity on the quantitative side. Pleasures and pains, on this view, are unified since there is something that all pleasures and all pains have in common that accounts for why these and only these are pleasures and pains, namely being either high or low on the hedonic scale. As Rem B. Edwards explains Bentham’s view, this is how Bentham can claim that “the quality of pleasure is always the same no matter how it is obtained.”[26] Being high or low on the hedonic scale, moreover, is a property that exists in terms of more and less and thus it allows for commensuration. For this reason, Bentham can claim that pleasures and pains are unified and commensurable without rejecting heterogeneity.[27]

Though I think we should concede that this view, if true, would account for unity and commensurability, it is doubtful if it is true. Human nature could perhaps have been like Bentham describes it, but as it happens to be, it probably is not.

Stuart Rachels has presented an introspective argument against Bentham, using the example of jamming one’s finger. When you jam your finger, Rachels argues, you experience just one thing, not two things. You do not experience a certain feeling in the finger, which by itself is hedonically neutral, and in addition to that, feel a general shift in hedonic level. Rather, Rachels claims, the pain you feel is just as located and immediately present in the finger as is the qualitative sensation. Indeed, Rachels argues, you feel just one thing: pain in the finger.[28] This seems right, and if it is right, it creates an explanatory problem for split experience theory: If we have two experiences, why does it seem as if we have just one?

Karl Duncker has presented a similar argument, appealing to the phenomenology of wine drinking.[29] Duncker seeks to clarify what counts as a cause, and what does not count as a cause, of the pleasures we get from drinking wine. To do this, Duncker asks and answers a series of questions. First he asks: Is the wine a cause of the pleasure we get? His answer is yes. Second: Is the drinking of the wine a cause of the pleasure we get? Yet, again he argues that yes, the drinking of the wine is also a cause of the pleasure. Third: Is theexperience of drinking the wine a cause of the pleasure we get? Here Duncker’s answer is no. The experience of drinking the wine is not a cause of the pleasure of wine drinking. Rather, the experience is the pleasure of wine drinking; it is the very taste of the wine thatconstitutes the pleasure of drinking the wine. The pleasure, Duncker claims, is in the experience. If he is right, then split experience theory introduces one step too many.

A third objection has been raised by William Alston. Alston argues that if our experiences were split the way Bentham suggests, then feelings of pleasure would distract us from the particular things that we find pleasurable. Granted that our attention is generally drawn toward pleasures, it would seem, on Bentham’s view, that intensely pleasurable experiences, such as having sex, would draw our attention away from what we are doing and over to the hedonic level itself, which is supposedly an experience separate from the qualitative experience of having sex. This, however, seems not to be the case. Rather, the opposite seems to be the case: The more pleasure we get from a certain activity, the more our attention tends to be drawn towards that activity.[30]

A fourth problem for split experience theory is to explain how different parts of our experiential field can simultaneously have different hedonic tones. Try eating a chocolate bar while pinching your finger. If you do, you will (if you are like me) feel pleasure and pain simultaneously in different areas of your experiential field. Though a sufficiently strong pain grabs one’s attention and overrides other experiences, it seems that if both the pleasure and the pain in question is fairly mild, we can simultaneously feel both. This is mysterious if our pleasure and pain level is determined by an inner hedonometer. It seems that the only way Bentham could account for a plurality of hedonic tones is by positing several hedonometers or by holding that one hedonometer can record several different hedonic levels simultaneously. Such a move—although perhaps not impossible—would deprive the theory of the explanatory simplicity that makes it appealing in the first place. For these reasons, split experience theory seems unappealing.

Again however, we should ask what counts in this theory’s favor. Although I do not believe split experience theory is as fundamentally mistaken as strong response theory is, I see few reasons to positively believe in it—except for the fact that split experience theory allows for heterogeneity while saving unity and commensurability. That, however, does not lend the theory epistemic support. The only additional reason could be that split experience theory also neatly explains the coffee paradox, and does so without resorting to response theory. If split experience theory is correct, the coffee paradox is explained by certain qualitative feels changing causal connections to our inner hedonometer. Split experience theory, however, is not the only non-response theory that can explain the coffee paradox. In lack of further supporting reasons, the theory should be rejected.[31]

If we reject both response theory and split experience theory, however, it seems difficult to account for unity and commensurability in face of the heterogeneity objection, for it seems that, in some sense, that which is unified and commensurable must be separate from that which is heterogeneous. Thus, to account for the unity and commensurability of pleasures and pains, it seems that pleasures and pains must be either extrinsic to our experiences (response theory), or at least, extrinsic to the qualitative part of our experiences (split experience theory). After all, it seems impossible that the very same phenomenon can be both heterogeneous and unified at the same time.

One theory, or quasi-theory, that might be seen as countering this, and that should be addressed parenthetically, has recently been suggested by Aaron Smuts. In Smuts’ view, “pleasurable experiences are those that feel good.” This is a refreshingly plain and obvious answer to the question of what pleasure (and conversely, pain) consists in, but as Smuts himself admits, “This is not an illuminating suggestion.”[32] The reason why is that it is closer to a restatement than to an explanation or an analysis. Smuts argues, however, that we cannot take for granted that it is possible to give an explanation or an analysis of what pleasure is. At a certain point, our explanatory and analytic regress must come to an end, and pleasure might be the natural place to stop. Phenomenally, pleasure seems to be a sui generis experience, and qua sui generis experience, it might well not permit further analysis. Perhaps James Mill was thus right in claiming that all we can really say about pleasure is that: “A man knows it, by feeling it; and this is the whole account of the phenomenon.”[33] This might be right, and if so, Smuts might have given the most thorough explanation that can be given. Facing the heterogeneity problem, however, saying that “all pleasures feel good” is not an answer that is likely to move those critical of unity and commensurability. Smuts’ theory amounts only to “look!” or “feel!,” but we can neither see nor feel that pleasures and pains are unified and commensurable. If the “feels good” theory is the best we can hope for, therefore, the evidence for unity and commensurability is scant.

I believe, however, that a better account of the nature of pleasure and pain—and of their unity and commensurability—can be given, and I shall now present and briefly defend this view. This view lies close to both split experience theory and Smuts’ feels good theory, but avoids the central problems that these theories face.



Dimensionalism is the theory that pleasure and pain have the ontological status of opposite ends of a hedonic dimension along which experiences vary. Several philosophers have earlier hinted to this view, but none have worked it out in detail. I will now work it out in some detail, defend it, and explain how it offers a solution to the problem of the unity and commensurability of pleasures and pains.

An early hint towards dimensionalism is found in C. D. Broad’s Five Types of Ethical Theory. Broad writes:

[T]here is a quality, which we cannot define but are perfectly acquainted with, which may be called ‘Hedonic Tone.’ It has two determinate forms of Pleasantness and Unpleasantness. And, so far as I can see, it can belong both to Feelings and to those Cognitions which are also Emotions or Connotations. … ‘A pleasure’ then is simply any mental event which has the pleasant form of hedonic tone, and ‘a pain’ is simply any kind of mental event which has the unpleasant form of hedonic tone. There is not a special kind of mental events, called ‘pleasures and pains;’ and to think that there is is as if one should solemnly divide human beings into men, women, and blondes. It is of course true that the commonest, and some of the most intense, pleasures and pains are feelings, in my sense of the word. But remorse, which is memory of certain events, having a certain emotional tone, is plainly a pain as much as a toothache. And hope, which is expectation of certain events, having a certain emotional tone, is plainly as much a pleasure as the sensation of smell which we get from a rose or a violet.[34]

What Broad suggests in this paragraph is that pleasures and pains, rather than being separate kinds of experiences, are “tones” or “qualities” of other experiences. This is emphasized by his further claim that “any mental event which has hedonic quality will always have other qualities as well.”[35] Pleasure and pain, on Broad’s view, do not ontologically belong on level with experiences such as experiential sweetness, greenness, and warmness. Rather, pleasure and pain are tones with which all experiences—including sweetness, greenness, and warmness—are imbued.

A similar view is proposed by Karl Duncker, who argues that every pleasure and every pain is a “side,” a “property,” an “abstract part,” or a “hedonic tone pervading an experience,” and that in and by themselves, pleasure and pain are “essentially incomplete experience[s]” that cannot exist in the absence of any particular experience being pleasurable or painful.[36]

Clarifying the dimensionalist position by means of analogy, Shelly Kagan suggests that pleasure and pain are related to qualitative experiences the same way auditory volume is related to sounds. Auditory volume, Kagan observes, is neither a component nor an object of auditory experience, but rather, an “aspect of sounds.” Applying the analogy to pleasures, Kagan suggests that we should “identify pleasantness not as a component of experiences, but rather as a dimension along which experiences can vary.” The fact that pleasure is not a kind of experience, then—returning to the analogy—is just as “obvious” as the fact that “loudness is not a kind of sound.” Rather than being a “kind of sound,” loudness is a dimension along which sounds vary.[37] Thus dimensionalism.

Dimensionalism, as I defend it here, is the claim that:

Pleasure and pain are opposite sides of a dimension along which experiences can vary.

It is possible and perhaps natural to pair (1) with the further claim that:

All experiences belong at a certain point on a hedonic dimension.

This further claim has been proposed by John Searle, who argues in The Rediscovery of the Mind that a “general feature of each modality [of consciousness] is that it can occur under the aspect of pleasant or unpleasant.” One can always, Searle claims, ask about an experience: “Was it fun or not?,” “Did you enjoy it or not?,” “Were you in pain, exasperated, annoyed, amused, bothered, ecstatic, nauseous, disgusted, enthusiastic, terrified, irritated, enchanted, happy, unhappy, etc.?”[38] On this view, whenever you experience something—spotting a friend, tasting honey, feeling an itch, reading a paper, coughing, or seeing a blue dot—one of the dimensions along which that experience varies is a hedonic dimension.

In this paper I commit to (1), but not to (2). There are two reasons why. First, (2) is rendered less certain than (1) by the fact that the hedonic dimension, unlike most other dimensions, is a dimension with axes stretching out on both sides of the zero point. It is unclear, moreover, what the difference is between an experience being at the zero point on the scale and an experience not being on the scale—the latter of which would seemingly be incompatible with (2). Second, the problem of reconciling heterogeneity with unity and commensurability does not depend on the truth or falsity of (2). If (2) is false, this restricts the range of experiences that are pleasurable or painful. It does not, however, alter the fact that those experiences which are pleasurable or painful are also unified and commensurable.

What reasons do we have to believe in (type 1) dimensionalism? Providing a comprehensive defense would require work beyond the scope of this paper. I shall, however, indicate my reasons for favoring it over competing theories. Let me start by explaining how dimensionalism solves the challenges raised against response theory and split experience theory.

Dimensionalism faces none of the problems faced by response theory. Since dimensionalism holds that pleasurability and painfulness are intrinsic to our experiences, it comes out on the intuitive side of the Euthyphro problem: It allows for explanations of liking in terms of pleasure and pain. For this reason, it has no problem accounting for hedonic surprises, and dimensionalism leaves open the question of whether there is a necessary connection between pleasure and liking.

At the same time, dimensionalism does not face the problems faced by split experience theory. First, dimensionalism has no problem explaining why pleasurable experiences are not distracting. If hedonic tones relate to experiences the same way auditory volume relates to sounds, then pleasure should distract no more from pleasurable experiences than volume distracts from sounds. On the contrary, it should attract attention, and this is what it does. For a similar reason, dimensionalism does not have a problem explaining why, when we jam a finger, we feel pain right there in the finger, since according to dimensionalism, it is the very feeling in the finger that is imbued with a negative hedonic tone. Moreover, dimensionalism has no problem explaining how we can simultaneously experience different hedonic tones in different parts of our experiential field, since there is nothing in dimensionalism that forbids different experiences from simultaneously having different hedonic tone.[39] This becomes clear if we formulate dimensionalism in terms of qualia. Formulated in terms of qualia, dimensionalism holds that rather than pleasure being a quale and pain being a quale, pleasure and pain are the opposite sides of a dimension along which qualia vary (or perhaps: exist). To the extent that we can simultaneously experience several qualia, we can also experience several hedonic tones.

Dimensionalism also offers a solution to the coffee paradox. It can solve the paradox since it does not hold that hedonic tone is part of the object or the content of an experience. Holding that hedonic tone is a dimension along which experiences vary, dimensionalism allows for a compatibility range between qualitative experiences and hedonic tones. Broad discusses the issue of compatibility range inFive Kinds of Ethical Theory. He asks:

[Is the] connexion between such and such non-hedonic quality merely causal and logically contingent, or is it intrinsically necessary? It is, e.g., logically possible that there should have been minds which had experiences exactly like our experiences of acute toothache in all their sensible qualities, but in whom these sensations were pleasantly toned?[40]

Broad does not answer the question. Duncker does, however, and writes that “A feeling-tone of pleasantness may reside in any kind of experience.”[41] I am not convinced that Duncker is right in holding that pleasantness may reside in any kind of experience. It is not certain that the qualitative feeling of being burned could ever have a positive hedonic tone. How wide the compatibility range happens to be, however, is not something that must be defined in order to defend dimensionalism; the crucial point is that dimensionalism allows for a compatibility range. To the extent that it does, it allows for an experience to be imbued with different hedonic tones at different points in time.

Dimensionalism can also explain how we are able to experience pleasure and pain, even though, as a puzzled James Mill noted inAnalysis of the Human Mind, pleasure and pain have “neither organ, nor object.” We have no designated pleasure and pain organ, and pleasures and pains are not objects in our environment that we occasionally stumble upon. Rather, Mill notes, “We have pleasures and pains of the eye, the ear, of the touch, the taste, the smell…”[42] Dimensionalism makes sense of why this is so. If pleasure and pain are dimensions of experiences as such, we need neither hedonic objects nor a designated hedonic sense. Rather, we should expect pleasurability and painfulness to be distributed along all different sensory modalities—and this seems to be how it is, since both sound, sight, taste, smell, and touch can be hedonically valenced. It is unclear what other ontological status than dimensions of experiences could account for this.

For these reasons, dimensionalism appears introspectively plausible. I also believe it makes sense biologically, however, and here is a speculative explanation of why: Evolution operates by the rule that a trait is selected for if and only if it promotes survival and reproduction. If we take for granted that consciousness evolved, consciousness would somehow have to promote survival and reproduction in order to be selected for. If consciousness did not promote survival and preproduction, it would not be selected for, and to the extent that it were biologically costly, it would be selected against. The only way consciousness could promote survival and reproduction, moreover, is by virtue of guiding an organism’s actions, prompting it to perform survival and reproduction enhancing actions – and the only way in which consciousness could prompt an organism towards survival and reproduction seems to be by imbuing experiences with a certain valence or a pro/con attitude. Without a valence or a pro/con attitude, it is unclear how an experience would be able to guide an organism’s actions. Evolution, moreover, cares for action, not for experiences as an end in itself. It therefore seems that if consciousness were to ever get going, valence would have to be present from the very start. Otherwise, consciousness would disappear as fast as it occurred. This suggests that hedonic valence phylogentically is as old as consciousness itself, which in turn lends support to the view that hedonic valence lies at the heart of consciousness. This supports dimensionalism, moreover, since according to dimensionalism, pleasure and pain—rather than being two things out of the many things we can experience—imbues all (or, if (2) is false, almost all) our experiences. Indeed, one might, from a dimensionalist approach to consciousness, argue that the first experience any organism ever had was an experience of either pleasure or pain, and that consciousness of the kind our species has today is a more fine-grained version of something that is most fundamentally a pleasure/pain mechanism. This, if true, gives supports the dimensionalist view.

This speculation concludes my argument in support of dimensionalism. Let me now turn to the question of how dimensionalism can help reconcile heterogeneity with unity and commensurability.[43]

Dimensionalism has no problem allowing for heterogeneity, since it places no restrictions on how heterogeneous our experiences may be. It places no more restrictions on how heterogeneous pleasures may be than our common assumptions about loudness place restrictions on how heterogeneous sounds may be. At the same time, dimensionalism allows for unity. Shelly Kagan touches on this point when discussing pleasure’s parallel to auditory volume. Kagan writes that a “recognition of the qualitative differences between the sounds of a symphony, rain falling, and a bird chirping, does nothing at all to call into question our ability to identify a single dimension—volume… .”[44] What unites all pleasures, according to dimensionalism, is the fact that they belong within a certain range on the hedonic dimension. What unites all pains is that they belong on the opposite range on the hedonic dimension. These are both instances of genuine unity, moreover, since it is by virtue of being on a certain range of the hedonic dimension that an experience is either a pleasure or a pain.[45]

For a unified group to be commensurable, the property by virtue of which the group is unified must be a property that exists in terms of more and less. This is the case with pleasures and pains, according to dimensionalism, since these mark different points on a hedonic dimension, and dimensions—by their nature—exist in terms of more and less and thus allow for commensuration.

Dimensionalism, therefore, is not only an introspectively and biologically plausible theory of pleasure and pain. It also helps reconcile our two opposing intuitions: It explains why, in spite of phenomenal heterogeneity, pleasures and pains are unified and commensurable.


Alston, W. (1967). Pleasure. In P. Edwards (Ed.), The encyclopedia of philosophy (pp. 341–347). New York: Macmillan Publishing Co. & The Free Press.
Aydede, M. (2000). An analysis of pleasure Vis-à-vis pain. Philosophy and Phenomenological Research, 61(3), 537–570.
Bentham, J. (1781/1996). Principles of morals and legislation. In J. H. Burns, H. L. A. Hart (Eds.). Oxford: Cleardon Press.
Brandt, R. (1979). A theory of the good and the right. Oxford: Cleardon.
Broad, C. D. (1930). Five types of ethical theory. London: Routledge.
Crisp, R. (2006). Reasons and the good. New York: Oxford University Press.
Duncker, K. (1941). On pleasure, emotion, and striving. Philosophy and Phenomenological Research, 1(4), 398–99.
Edwards, R. B. (1979). Pleasures and pains: A theory of qualitative hedonism. Ithaca: Cornell University Press.
Feldman, F. (2006). Pleasure and the good life. New York: Oxford University Press.
Hall, R. J. (1989). Are pains necessarily unpleasant? Philosophy and Phenomenological Research, 49(4), 643–659.
Heathwood, C. (2007). The reduction of sensory pleasure to desire. Philosophical Studies, 133, 23–44.
Kagan, S. (1992). The limits of well-being. Social Philosophy and Policy, 9(2), 169–189.
Katz, L. (2009). Pleasure. The Stanford encyclopedia of philosophy (Fall 2009 Edition). Edward N. Zalta (Ed.), URL = <>.
Korsgaard, C. (1996). The sources of normativity. Cambridge: Cambridge University Press.
Locke, J. (1690/1975). An essay concerning human understanding. Oxford: Oxford University Press.
Mason, E. (2007). The nature of pleasure: a critique of Feldman. Utilitas, 19, 379–87.
Melzack, R. (2005). The McGill pain questionnaire: from description to measurement. Anesthesiology, 103(1), 199–202.
Mill, J. (1869). Analysis of the human mind. London: Longmans Green Reader and Dyer.
Moore, A. (2008). Hedonism. Stanford Encyclopedia of Philosophy (Fall 2008 Edition). Edward N. Zalta (Ed.). URL =<>.
Parfit, D. (1984). Reasons and persons. Oxford: Oxford University Press.
Plato (1997). Philebus. Translated by Dorothea Frede. In Plato: Complete works. J. Cooper (Ed.). Indianapolis: Hackett Publishing Company.
Plochmann, G. (1950). Some neglected considerations on pleasure and pain. Ethics, 61(1), 51–55.
Rachels, S. (2000). Is unpleasantness intrinsic to unpleasant experiences? Philosophical Studies, 99(2), 187–210.
Rachels, S. (2004). Six theses about pleasure. Philosophical Perspectives, 18(1), 247–67.
Searle, J. (1992). The rediscovery of the mind. Cambridge: MIT Press.
Sidgwick, H. (1907/1981). The methods of ethics, 7th ed. Indianapolis (Ind): Hackett Publishing.
Smuts, A. (2010). The feels good theory of pleasure. Philosophical Studies, 155(2), 241–265.
Sprigge, T. L. S. (1988). The rational foundations of ethics. London: Routledge.
Sumner, L. W. (1999). Welfare, happiness, and ethics. Oxford: Oxford University Press.


1 The McGill Pain Questionnaire is a standard questionnaire for measuring pain. See Melzack (2005, pp. 199–202).
2 Aydede (2000, p. 540).
3 Locke (1690/1975, II, XX, §15).
4 Sidgwick (1907/1981, pp. 93, p. 127)
5 Katz (2009). Some philosophers oppose this wide usage of the terms “pleasure” and “pain.” Roger Crisp (2006, pp. 103–109) suggests that rather than speaking of “pleasure” and “pain” in this wide sense, we should speak of “enjoyment” and “suffering.” Stuart Rachels (2004, pp. 247–48) suggests that we can keep “pleasure,” but that we should not use “pain” as its antonym. “Pain,” Rachels suggests, should more narrowly be reserved for the negative experiences brought about by nociception, and he argues that the proper antonym for pleasure is “unpleasure.” I have no principled reason to oppose such word usage, but for the sake of simplicity I keep to the wide usage of “pleasure” and “pain” in this paper.
6 Plato (1997, 12 c-d).
7 Parfit (1984, p. 493).
8 Feldman (2006, p. 79).
9 Edwards (1979, p. 40).
10 Edwards (1979, pp 34–35, 73).
11 This view is sometimes referred to as “externalism.” See Sumner (1999, pp. 87–91).
12 Sidgwick (1907/1981, p. 127).
13 Alston (1967, p. 345).
14 Sumner (1999, p. 90). Sumner labels this view the “attitude model.”
15 Edwards (1979 p. 35).
16 Brandt (1979, p. 38).
17 Hall (1989, p. 649).
18 Heathwood (2007, p. 32).
19 Korsgaard (1996, p. 147–8).
20 Moore (2008).
21 Sprigge (1988, pp. 131–2).
22 Similar criticisms have been raised by Rachels (2000, pp. 187–210) and Mason (2007, pp. 388–97).
23 For a more in-depth discussion of these points, see Smuts (2010).
24 Smuts (2010).
25 Bentham (1781/1996, pp. 43-46).
26 Edwards (1979, p. 34).
27 A first reading of Bentham might give the impression that he holds that pleasures and pains, qua pleasures and pains, vary qualitatively. Bentham lists seven axes along which pleasure and pain can vary: intensity, duration, certainty, propinquity, fecundity, purity, and extent. With the exception of intensity and duration, however, none of these concern matters intrinsic to the nature of pleasures and pains. The other axes refer to different causal roles that pleasures and pain can play, and the different ways in which they can be distributed. “Purity,” in Bentham’s words, refers not to the phenomenological purity of a pleasure or pain, but to “the chance it has of not being followed by sensations of the opposite kind: that is, pains, if it be pleasure: pleasures, if it be pain.” “Extent” concerns the number of individuals who experience pleasure or pain. Bentham, (1781/1996, pp. 38-40).
28 Rachels (2000, p. 196).
29 Duncker (1941, pp. 398–9).
30 Alston (1967, p. 345).
31 It should be said in Bentham’s defense, however, that his theory fares somewhat better when it comes to pleasures than when it comes to pains. Pleasures have more of a holistic feel to them, and are not located in the same way as pains. While you can have a pain in your index finger, you can’t have a pleasure in your index finger; pleasures seem to be much more “inside” and “everywhere,” as if the qualitative feel caused a higher hedonic level in us. I do not, however, think that this is sufficient to support split experience theory, and as such that the theory—though not obviously false—should be rejected.
32 Smuts (2010, p. 254).
33 Mill (1869, p. 2:184).
34 Broad (1930, pp. 229–30).
35 Broad (1930, pp. 229–30).
36 Duncker (1941, p. 400).
37 Kagan (1992, pp. 170–72). Aaron Smuts might also be interpreted in this direction when writing that pleasure is “a tone that cannot be cleanly extracted or focused on apart from the experience itself,” and that “pleasure is not a distinct form of experience.” Smuts (2010, p. 16).
38 Searle (1992, pp. 38, 129).
39 For an interesting discussion of this, favoring the same conclusion, see Plochmann (1950, pp. 54-55).
40 Broad (1930, p. 231).
41 Duncker (1941, p. 412).
42 Mill (1869, pp. 1:37, 2:185)
43 A more thorough defense of dimensionalism would require addressing several other issues. The most central of these, I think, is the problem of explaining what mechanism determines what experiences are imbued with what hedonic tone. That, however, must be the topic of a different paper. My aim in this paper is merely to argue that dimensionalism is a very plausible theory, and that—if true—it solves the problem of the unity and commensurability of pleasures and pains.
44 Kagan (1992, p. 172).
45 I also believe that a dimensionalist can agree with the traditionally arch-heterogeneous claim that there is probably no such thing as pure “pleasure” or pure “pain,” and that all we ever experience is particular pleasures and particular pains. This is so because on the dimensionalist view, “pleasure” and “pain” are abstractions: They are concepts by which we isolate the property of being on either the positive or the negative side of the hedonic dimension, while omitting the particular distance from the zero point as well as the particular content of the experience. The fact that there are only particular pleasures and particular pains, therefore, need not be a threat to the unity and commensurability of pleasures and pains any more than the fact that there are only particular heats and particular cools is a threat to the unity and commensurability of heats and cools.