Introduction
When I quench my thirst with a cool glass of water on a hot day, I have a paradigmatically pleasant sort of experience. It seems obvious that this is a good experience for me to have. It seems to be non-derivatively good for me, in the sense that it contributes directly to making my day go better for me.
Consider the following three claims regarding this and other paradigmatically pleasant experiences. First: the experience itself feels good; thus, any experience that feels the same way would also feel good, and would be non-derivatively good for the person who has it. This is the objectivist claim. Second: if a person is indifferent to an experience that feels just like my pleasant experience, then that experience is not non-derivatively good for them. This is the subjectivist claim. Third: it is possible for there to be a subject who is indifferent to an experience that feels just like my pleasant experience. Many philosophers are prepared to endorse this possibility claim. So they face an inconsistent triad:
Objectivist claim. Some kinds of experience (namely, pleasant and unpleasant experiences) are non-derivatively good for all possible subjects who experience them.
Subjectivist claim. Necessarily, an experience is non-derivatively good for a subject only if that subject is not indifferent to experiences of that kind.
Possibility claim. For every kind of experience a subject can have, possibly a subject is indifferent to experiences of that kind.
This is what I call the pleasure problem. There is also a pain problem, in which every instance of good is replaced with bad.
Naturally, there are three basic strategies for resolving these two problems. Subjectivists take the first strategy: in each case, they resolve the problem by rejecting the objectivist claim (Sobel Reference Sobel2005; Heathwood Reference Heathwood2011). Objectivists take the second strategy: in each case, they resolve the problem by rejecting the subjectivist claim (Goldstein Reference Goldstein1989; Rachels Reference Rachels2000; Bramble Reference Bramble2013). The third strategy—rejecting the possibility claim—is unpopular. No one seems to have adopted it in response to the pleasure problem, with the exception of Timothy Sprigge (Reference Sprigge1987). Accordingly, I call it the Spriggean strategy. Sprigge's arguments have not gained much traction—objectivists and subjectivists are both quick to dismiss the idea that there might be necessary connections between our attitudes and our experiences.
In this essay I make the case for the Spriggean strategy. First, I show how and why the possibility claim is relevant to the debate between objectivists and subjectivists. Then I rehearse Sprigge's original argument against the possibility claim, and develop my own version of it. Along the way, I show that many philosophers of mind have independent grounds for rejecting the possibility claim. I conclude that we ought to follow Sprigge in claiming that ‘pleasures and pains are of their nature liable to affect behaviour in certain directions’ (Sprigge Reference Sprigge1987: 142). In this way, we can resolve the pleasure and pain problems, while accommodating both the objectivist and subjectivist intuitions.
1. Terms and Conditions
The pleasure and pain problems involve claims about ‘kinds of experience’. In this context, the kinds are individuated phenomenologically, or by ‘what it is like’ to experience them. Experiences e1 and e2 differ in kind just in case ‘what it is like’ to experience e1 differs from ‘what it is like’ to experience e2.
It is useful to have at hand some paradigmatically pleasant and unpleasant kinds of experiences. Imagine, then, that you take a cool sip of water on a dry summer day, quenching your thirst and causing yourself to have a pleasant experience. I call this and all other experiences of the same kind a quench-experience. Like every kind of experience, quench-experiences involve everything about ‘what it is like’ to be a particular subject at a particular time: they involve everything about ‘what it is like’ to be you on the particular occasion on which you have a cool sip of water. If your vision is blurry on that occasion, then you must have blurry vision to have a quench-experience. If you feel a pebble between your toes, then you must feel a pebble between your toes to have a quench-experience. The upshot is that quench-experiences are extremely specific sorts of experiences, and it is unlikely that anyone but you will ever have one. Even so, it is helpful to talk about the quench-experiences of other subjects. So I make a simplifying assumption: in discussing quench-experiences, I assume that the subjects having those experiences are similar to you in all relevant respects. Like you, they are parched. Like you—maybe—they have blurry vision and pebbles between their toes. Thus, there is no bar to claiming that they have quench-experiences—that is, experiences that are exactly like your total experience as you sip water on a specific occasion. Nothing of philosophical significance turns on this assumption, but it makes the discussion go more smoothly.
Quench-experiences, I claim, are paradigmatically pleasant experiences. Now imagine a different scenario: your unprotected hand is thrust into an open flame, thereby causing you to have an extremely unpleasant experience. I call this and all other experiences of the same kind a burn-experience. Again, burn-experiences all feel exactly alike. They all feel just like the total experience that you get on a particular occasion from having your hand thrust into an open flame. I invoke my simplifying assumption again: In discussing burn-experiences, I assume that the subjects having those experiences are similar to you in all relevant respects. Thus, there is no bar to claiming that they have burn-experiences—that is, experiences that are exactly like your total experience as your hand is burned on a specific occasion.
I assume that quench- and burn-experiences are among the kinds of experiences that figure in the pleasure and pain problems. According to the objectivist, then, quench-experiences are non-derivatively good for all possible subjects who have them, and burn-experiences are non-derivatively bad for all possible subjects who have them. As an aside, this is consistent with the claim that some quench-experiences and burn-experiences are bad simpliciter and good simpliciter, respectively—perhaps on account of those experiences being undeserved for some subjects (Goldstein Reference Goldstein1989). Objectivists are committed only to claims about those experiences’ values for subjects: quench-experiences are necessarily non-derivatively good for subjects; burn-experiences are necessarily non-derivatively bad for subjects. Subjectivists, in contrast, claim that those experiences are only non-derivatively good or bad for subjects who are not indifferent to them.
I understand indifference in terms of caring: we are indifferent to something if and only if we do not care about it. The term caring, in turn, is intended to cover the various attitudes that subjectivists employ in their various preferred theories of experiential value. For Derek Parfit, the relevant attitude is ‘hedonic (dis)liking’ (Reference Parfit2001: 53) So, for Parfit, an experience is good (bad) for a subject just in case they (dis)like it. For Fred Feldman, the relevant attitude is ‘attitudinal (dis)pleasure’ (Reference Feldman2004: 55–66); for Chris Heathwood, it is ‘genuine attraction/aversion’ (Reference Heathwood2019: 671–75). My term caring is meant to be neutral between these and other proposals.
Suffice it to say that caring involves some combination of behavioral dispositions, mental dispositions, and phenomenology. For example, I care about eating ice cream: I am disposed to be attracted to eating it, I view the prospect of eating it with gusto, and I feel good about eating it—perhaps not in a strictly phenomenological sense. I also care about drinking battery acid. I am disposed to be averse to drinking it, I view the prospect of drinking it with horror, and I feel bad about drinking it—again, perhaps not in a strictly phenomenological sense.
I understand non-derivative goodness and non-derivative badness in the usual way. Burn-experiences, I assume, are non-derivatively bad for me. They are bad for me and not in virtue of being related to the goodness or badness of anything else. In contrast, the act of thrusting my hand into an open flame is merely derivatively bad for me. It is bad for me because it causes me to have burn-experiences, and burn-experiences are non-derivatively bad for me. Similarly, it is derivatively good for me to drink water when I am thirsty, in part because it causes me to have quench-experiences, which are non-derivatively good for me. Other forms of derivative value correspond to other relations that states with non-derivative value can bear to other states without such value.
2. The State of the Debate
The possibility claim is relevant to the debate between objectivists and subjectivists, where objectivists are those who accept the objectivist claim, and subjectivists are those who accept the subjectivist claim.
The first thing to do is to distinguish this debate from another, closely related debate. As I have stated, the objectivist and subjectivist claims both tell us something about which possible experiences have the properties of non-derivative goodness and badness. So the debate, as I am understanding it, is a debate about which possible experiences have those normative properties. The debate is not concerned with why those experiences have those properties. That is the subject of a distinct, but closely related debate:
Objectivist explanatory claim. Whenever an experience is non-derivatively good or bad for a subject, it is non-derivatively good or bad for that subject in virtue of being an experience of the kind that it is.
Subjectivist explanatory claim. Whenever an experience is non-derivatively good or bad for a subject, it is non-derivatively good or bad for that subject in virtue of their caring about it.
I am mainly interested in the former debate, rather than the latter. But much of what I say regarding the former debate has straightforward implications for the latter. For example, in this section I consider various arguments put forward in the former debate, and I argue that Spriggeans can avoid all of them. Those same arguments are also put forward in the latter, explanatory debate, and the Spriggean can avoid them in that context too. Thus, my discussion of the arguments bears on the explanatory debate in a straightforward way.
The arguments I consider turn on the possibility claim: for every kind of experience a subject can have, possibly a subject is indifferent to experiences of that kind. Thus, the possibility claim is highly relevant to the debate between objectivists and subjectivists. It is implicated in objectivist arguments against subjectivism, and subjectivist arguments against objectivism. Rejecting it would deflate all those arguments.
2.1. Objectivist Arguments
Consider the following kind of standard objectivist argument. The argument begins with a description of something like the following case:
Ultra-Spartans: The Ultra-Spartans are a race of aliens. They are much like human beings in many respects. In particular, they have hands, and when those hands are burned by open flames, they sometimes have burn-experiences. That is, they have experiences that feel just like an experience that an ordinary human being might get from having their hand burned by an open flame. But Ultra-Spartans, unlike human beings, are indifferent to burn-experiences: they are not at all disposed to avoid them, nor are they distracted by them. They simply do not care about those experiences at all.
Cases like Ultra-Spartans have a famous history in the philosophy of mind. In that context, the purpose of invoking Spartan-style cases is to try to show that our inner experiences can come apart from the dispositions with which they are associated (see, for example, Putnam 1965: 102–4; Lewis 1980: 122). Notably, however, this is not how Spartan-style cases are used by objectivists in their debate with subjectivists, since objectivists and subjectivists both typically agree that there are no necessary connections between experiences and dispositions. Rather, the objectivist uses Spartan-style cases in order to motivate a value claim: paradigmatically unpleasant experiences are bad for us, even if we are indifferent to them. After all, the objectivist argues, the Ultra-Spartans’ experiences feel exactly like our experiences. Just think about what it would be like to have a burn-experience—that is, to thrust your hand into an open flame. Could the Ultra-Spartans have experiences just like that without being worse off for having them? When the question is framed in this way, the objectivist intuition is rather forceful. It seems difficult to imagine that Ultra-Spartans are not made worse off by their burn-experiences. And if they are made worse off by those experiences, then subjectivism is false:
Anti-subjectivist Spartan argument
P1: There are possible creatures who are indifferent to burn-experiences, but whose burn-experiences are bad for them.
P2: If P1, subjectivism is false.
Conclusion: Subjectivism is false.
When objectivists make this anti-subjectivist argument, they do not appeal to the particular alien creatures I called Ultra-Spartans. But they describe cases with a similar structure. For example, Irwin Goldstein argues as follows: ‘In principle, emotional reaction can be severed from any pain sensation without the sensation changing qualitatively. If all of pain's unpleasantness and badness were contingent on concurrent aversion to pain, any pain, however intense, could in principle shed all unpleasantness while remaining qualitatively unchanged. In some people intense pain might have no trace of unpleasantness or badness. This seems impossible. Concurrent aversion is not necessary for unpleasantness and badness’ (1989: 261).
Guy Kahane pursues the same line of thought to the same conclusion. If subjectivism is true, he tells us, then we must embrace an absurd possibility: ‘that I could be in the same total experiential state I am in when suffering from excruciating pain, yet that this state may not be bad at all, or may even be intensely enjoyable and thus good. This, I believe, is not a suggestion we can make sense of. Perhaps there will be those who will deny this. But it is not by accident that, although subjectivism about pain's badness is widespread, we are never told that this is one of its implications’ (Reference Kahane2009: 334). In both cases, the basic point is the same. It would be bad for one to have paradigmatically unpleasant experiences, even if one is indifferent to those experiences.
Subjectivists reject this value claim. They claim that creatures like Ultra-Spartans would not be made worse off by their burn-experiences. More generally, they claim that any subject that is indifferent to its burn-experiences is not made worse off by those experiences. However, subjectivists could offer a different response to the anti-subjectivist argument. They could reject the metaphysical assumption that creatures like Ultra-Spartans are possible. Indeed, this is exactly how some philosophers of mind react to these sorts of cases. This would amount to rejecting the possibility claim in the pleasure and pain problems: if no possible subjects are indifferent to burn-experiences, then it is not the case that for every kind of experience a subject can have, possibly a subject is indifferent to experiences of that kind. In this way, the subjectivist could undermine the objectivist's argument.
2.2. Subjectivist Arguments
This dialectical situation is exactly mirrored in arguments against objectivism. Subjectivists describe cases in which subjects’ experiences come apart from the dispositions with which they are associated. The only difference is that whereas objectivists tend to describe cases in which subjects are indifferent to paradigmatically unpleasant experiences, subjectivists tend to describe cases in which subjects are indifferent to paradigmatically pleasant experiences. Consider the following case, for example:
Ultra-Ascetics. The Ultra-Ascetics are a race of aliens. They are much like human beings in many respects. In particular, when they quench their thirst with a cool sip of water, they sometimes have quench-experiences. That is, they have experiences that feel just like an experience an ordinary human might get from drinking cool water on a hot day. But Ultra-Ascetics, unlike human beings, are indifferent to quench-experiences: they are not at all disposed to be attracted to them. Those experiences have no appeal for the Ultra-Ascetics. They simply do not care about them at all.
Like objectivists, subjectivists are not trying to show that experiences can come apart from the dispositions with which they are associated. Rather, they are trying to motivate a value claim. The subjectivist claims that paradigmatically pleasant experiences are not good for subjects who are indifferent to them. After all, the subjectivist will argue, the Ultra-Ascetics are psychologically unlike human beings. They, unlike us, are in no way engaged by their quench-experiences. Why, then, should we think that those experiences make them better off?
When the question is framed in this way, the subjectivist intuition is rather forceful. It does seem difficult to imagine that the Ultra-Ascetics are made better off by their quench-experiences. And yet, quench-experiences are the sorts of experiences that, according to objectivists, are necessarily good for anyone who has them. So if they are not good for Ultra-Ascetics, then objectivism is false:
Anti-objectivist ascetic argument
P3: There are possible creatures whose quench-experiences are not good for them.
P4: If P3, objectivism is false.
Conclusion: Objectivism is false.
When subjectivists have made this kind of argument, they have typically had in mind a particular theory of pleasure: namely, the theory that pleasurable experiences share some sort of phenomenological commonality. This is sometimes called the felt-quality theory of pleasure. For example, Chris Heathwood writes,
On the felt-quality theory, it must be just a contingent fact about us humans that we tend to like and want this feeling of pleasure . . . Realizing this invites us to imagine creatures indifferent to this feeling (in the same way that you are probably indifferent to, say, the white color sensation you are experiencing while looking at this page). Of course, we can suppose that the sensation of pleasure is good in itself for anyone who experiences it, and if we do, it is plausible to maintain that it provides reasons. But given this conception of pleasure, it is hard to see why we would want to say that pleasure is good in itself for us in the first place (in the same way that it would be hard to see why we would want to say that the white color sensation you are experiencing while looking at this page is an intrinsically good sensation for us to experience). (2011: 94)
Although Heathwood's argument targets the felt-quality theory, the argument applies equally to any kind of objectivism about pleasure and pain. Rather than talking about ‘the feeling of pleasure,’ Heathwood might have talked about quench-experiences. The same goes for a similar objection from David Sobel: ‘It must be metaphysically possible, on this [felt-quality] conception of pleasure, that someone not like it. . . . So let it be that we finally find someone who really does not like the flavor of sensation of pleasure. Should we think that this person is necessarily making some sort of mistake? Well what mistake would it be? I myself do not understand what sort of mistake could be thought to be necessarily involved in a failure to like this or that phenomenological state’ (2005: 445).
Objectivists respond to these arguments by rejecting the value claim. They contend that even subjects who are indifferent to their quench-experiences are in fact made better off by them. However, objectivists could instead reject the metaphysical assumption that creatures like Ultra-Ascetics are possible. Again, this would amount to rejecting the possibility claim in the pleasure and pain problems: if no possible subjects are indifferent to quench-experiences, then it is not the case that for every kind of experience a subject can have, possibly a subject is indifferent to experiences of that kind. By rejecting the possibility claim, the objectivist could undermine the subjectivist's argument against objectivism.
2.3. The Possibility Claim
My contention is that we should reject the possibility claim. But one might worry that if we reject it, then we will commit ourselves to some sort of extravagant metaphysical worldview. So I want to pause briefly to address this worry, before moving on to my positive argument against the possibility claim.
Here is the worry. If the possibility claim is false, then some experiences are such that, necessarily, we care about them whenever we have them. And this might seem to entail that which is forbidden by Hume's dictum: the existence of necessary connections between wholly distinct things (Wilson 2010). There is room to debate how exactly Hume's dictum should be understood, but the general idea is clear enough. Consider the two apples on my desk: they are wholly distinct from one another—they are entirely different chunks of reality—so Hume's dictum tells us that the state of one apple does not necessarily have any consequences for the state of the other. And this does seem at least prima facie plausible. In contrast, consider the properties of being an apple and being a fruit. It is prima facie plausible that there are necessary connections between these properties—it seems clear that being an apple is necessarily co-instantiated with being a fruit—and Hume's dictum does not forbid our saying so, because being an apple and being a fruit are not wholly distinct things. Part of what it is to be an apple is to be a fruit.
The present worry is that our experiences and attitudes are wholly distinct things—they are more like the pair of apples on my desk, and less like the properties of being an apple and being a fruit. Thus, the claim that they are necessarily connected is a violation of Hume's dictum, and ought to be regarded as metaphysically extravagant. Putting this together, we arrive at a simple argument:
Extravagance argument
P5: It is metaphysically extravagant to reject Hume's dictum: the claim that there are no necessary connections between wholly distinct things.
P6: Our experiences are wholly distinct from our attitudes towards our experiences (including, for example, attitudes of caring or indifference).
P7: If P6 is true and we reject the possibility claim, then we must accept that there are necessary connections between wholly distinct things.
Conclusion: If we reject the possibility claim, then we must accept something metaphysically extravagant.
The argument, thus understood, is far from airtight. One might deny P6, by embracing a theory of mental states on which our attitudes towards our experiences are not always wholly distinct from the experiences themselves. One such theory is role functionalism. According to the role functionalist, all it is to be an experience of a certain kind is to play a certain causal role—that is, to stand in certain causal relations to other things. With this theory at hand, we might say that part of what it is to be a burn-experience is to be something that causes the experiencer to care about it. On this theory, a given mental state simply does not count as a burn-experience unless it causes the experiencer to care about it. So it turns out that your burn-experience is not wholly distinct from your caring about it; they are not entirely different chunks of reality. It simply would not count as a burn-experience if it did not cause you to care about it. P6 turns out to be false. So, in rejecting the possibility claim, we need not reject Hume's dictum.
Of course, some may prefer to reject P5. Not everyone thinks that it is a metaphysical extravagance to reject Hume's dictum. And if one rejects it, then further options become available. For example:
Anti-Humean experientialism. There are some kinds of experience that bear necessary connections to (distinct) dispositions that are characteristic of caring, such as attraction and aversion responses. One's having those dispositions necessarily causes one to have an experience of one of those kinds, or one's having an experience of one of those kinds necessarily causes one to have those dispositions.
This kind of non-Humean theory of caring and indifference can be supported by more general metaphysical views. One might adopt a causal or dispositional theory of properties, along the lines of Sidney Shoemaker (Reference Shoemaker1984), John Heil and David Robb (Reference Heil and Robb2003), and Henry Taylor (Reference Taylor2018). On this view, experiential properties are first-order physical properties that bear necessary connections to other, distinct physical properties. A related view posits that there are phenomenal powers: phenomenal properties that produce certain effects in virtue of what it is like to have them. This view is defended by Hedda Mørch (Reference Mørch2014, Reference Mørch2017, Reference Mørch and Coleman2019a, Reference Mørch and Seager2019b, Reference Mørch2020), Harold Langsam (Reference Langsam2011), and David Builes (2020). The important point is that there is a range of views—both Humean and non-Humean—that are consistent with rejecting the possibility claim. So, rejecting it does not commit us to controversial metaphysical claims in any straightforward way.
3. The Spriggean Argument
Sprigge responds to the pleasure and pain problems by rejecting the possibility claim. He does so in the context of his particular view about pleasure and pain: he maintains that pleasure and pain are necessarily such that when we experience them, they dispose us to do certain things. In particular: ‘The pleasurableness of an experience tends of its very nature to promote activity within the stream of consciousness which tends to sustain and repeat it, while the painfulness of an experience tends of its very nature to promote activity which will remove it’ (Sprigge Reference Sprigge1987: 142). Sprigge also affirms that certain kinds of experiences are essentially pleasant, while others are essentially unpleasant (Reference Sprigge1987: 140). So if I am having a quench-experience, then necessarily I am disposed to ‘promote activity which tends to sustain and repeat it’ (Sprigge Reference Sprigge1987: 142). And if I am having a burn-experience, then necessarily I am disposed to ‘promote activity which will remove it’ (Sprigge Reference Sprigge1987: 142). Sprigge strongly suggests that having these dispositions is sufficient for caring about those experiences (Reference Sprigge1987: 142–43). On the resulting Spriggean view, we cannot possibly be indifferent to those experiences.
Sprigge offers several different arguments for his view. Most of those arguments are negative: he argues against various competing theories of pleasure and pain, and he concludes that his own view is the only one left standing. I am more interested in Sprigge's positive argument, which he summarizes as follows: ‘My own view is that there are [nonanalytic] necessities, at least of tendency, and that the reinforcing powers of pleasure and pain are conspicuous examples of such. If we deny that these, in virtue simply of being the specific qualities they are, have an intrinsic tendency to influence behavior as positive and negative reinforcers in the way we have roughly characterised, we must either analyse them behaviouristically or pretend that there would be nothing intrinsically odd to counter-hedonically guided behaviour’ (Sprigge Reference Sprigge1987: 148).
While this passage is open to multiple possible interpretations, I think the following reconstruction is plausible. According to Sprigge, it is necessarily fitting to respond with aversion and attraction to unpleasant and pleasant experiences, respectively. Thus, it is necessarily odd to respond with aversion to pleasant experiences, or to respond with attraction to unpleasant experiences. Call this thesis Hedonic Fittingness. Sprigge goes on to say that we can adequately explain Hedonic Fittingness only if we accept that there are necessary connections between certain kinds of experience and dispositions. In particular, if one has a paradigmatically pleasant or unpleasant experience, one must be disposed to be attracted to it or averse to it, respectively (1987: 148). So Sprigge's argument is simple:
Spriggean argument
P8: Hedonic Fittingness
P9: We cannot adequately account for Hedonic Fittingness unless we accept that there are certain kinds of experience such that necessarily we are not indifferent to those experiences.
P10: If Hedonic Fittingness and P9, then there are certain kinds of experience such that necessarily we are not indifferent to those experiences.
Conclusion: There are certain kinds of experience such that necessarily we are not indifferent to those experiences. (The possibility claim is false.)
Although there is room to doubt P8 and P10, I believe that P9 is the most doubtful premise here. Stuart Rachels, one of the few philosophers to engage with the argument, points out that Hedonic Fittingness can be explained without appealing to any necessary connections (Goldstein [Reference Goldstein1980, Reference Goldstein1983] makes the same point, but not as a response to the Spriggean argument). Rachels's explanation is simple: unpleasant experiences are bad for us, and it is fitting to be averse to things that are bad for us (Reference Rachels2000: 201). This explains why the Ultra-Spartans’ attitudes are odd. Their burn-experiences are bad for them, and yet they are not averse to those experiences. I think that Rachels's objection is successful. However, I also believe that Sprigge's argument is on the right track. We can modify it to reach the same conclusion while avoiding Rachels's objection.
First, we can leave out any appeal to a relation of fittingness. The important point is not that certain dispositions fit certain kinds of experiences, but that certain kinds of experience coincide with certain kinds of dispositions. They coincide in the perfectly prosaic sense that for each of those kinds of experiences, whenever one has an experience of that kind, one also has a corresponding disposition. Second, we can expand the argument beyond paradigmatically pleasant and unpleasant kinds of experience. As I will argue, there are many kinds of experience that coincide with specific kinds of dispositions. Indeed, our phenomenology systematically coincides with our dispositions. We cannot explain this systematic coincidence in the way that Rachels explains hedonic fittingness. The best explanation is that there are necessary connections between kinds of experience and kinds of dispositions. So we ought to accept that there are such necessary connections, and we ought to reject the possibility claim as a consequence of this general commitment.
5. The Neo-Spriggean Argument
The neo-Spriggean argument begins with the following thesis:
Systematic Coincidence. Many kinds of experience are such that, whenever one has an experience of that kind, then one has a corresponding disposition.
To get a feel for this thesis, we can start with some obvious cases in which our experiences coincide with our dispositions. For example, if I feel an itch, I am disposed to try to scratch. This is not to say that, on any occasion on which I feel itchy, I will try to scratch. I might be distracted, or I might have some reason to want to feel itchy. These are most naturally understood as cases in which my disposition to try to scratch is masked by countervailing circumstances. So these cases are consistent with the thought that everyone who feels itchy is disposed to try to scratch. And this thought, although couched in philosophical jargon, is a piece of common sense. Similarly, it is common sense that thirsty people are disposed to try to drink, and tired people are disposed to try to rest.
Some kinds of experience are not obviously associated with outward behavior, in the way that itchiness is associated with scratching, and thirst is associated with drinking. Consider color experiences, for example—there is no particular kind of behavior that is obviously associated with experiences of redness. Color experiences do not threaten Systematic Coincidence even if they do not coincide with certain dispositions. Systematic Coincidence merely tells us that many—not all—kinds of experience coincide with corresponding dispositions. But it is worth noting that, in point of fact, there is a plausible case to be made that color experiences do coincide with dispositions. As John Hawthorne explains:
Consider the trio: phenomenal red, phenomenal orange, phenomenal blue. It is certainly true that when a subject enjoys all three phenomenal states simultaneously and is invited to judge which pair is most similar, she will judge that phenomenal red and orange are most similar. Phenomenal colors are thus disposed to produce certain similarity verdicts. These dispositions are causal powers of the phenomenal colors. . . . Of course, various familiar puzzles attending ascriptions of dispositions arise here too: We say that a certain poison is disposed to kill you when ingested even though it will not do so when ingested with an accompanying antidote. I say that phenomenal colors dispose certain similarity verdicts even though, doubtless, there are some extraordinary situations in which the characteristic manifestation of the disposition will not be forthcoming. That all our ordinary disposition claims may be false approximations to the truth is not a matter I need worry about here. What is crucial is that there are causal powers essential to phenomenal colors, not that I have succeeded in characterizing one of them with full exactitude. (2006: 354)
Hawthorne focuses on our dispositions to discriminate between experiences, but we could just as well focus on our dispositions to discriminate between things in the world. Suppose I have a mixture of red and green candies, and suppose I know that the reds taste much better than the greens. Naturally, I am motivated to eat the best-tasting candies I can, so I will tend to go for the reds and not the greens. If I had the same motivation but were completely color blind, I would act differently—I would be somewhat at a loss. My color experiences thus dispose me to act differently than I would if I had no such experiences. Compared to my experiences of itchiness—for example—color experiences bear a less overt or direct connection to behavior. But it does seem that they make a difference to what we tend to do in various circumstances. So even in the case of color experience, there is a case to be made that there are systematic connections between experiences and dispositions. The same considerations apply, mutatis mutandis, to sensory experiences pertaining to other modalities. And notice also that we can discriminate across sensory modalities. We can discriminate between—for example—redness and loudness. (See Hawthorne Reference Hawthorne2006: 254n18.)
Whatever one says about sensory experiences, it is clear enough that many kinds of experience—such as experiences of itchiness, thirst, tiredness—systematically coincide with kinds of dispositions. Systematic Coincidence is thus a striking fact about our experiences and dispositions. According to the Spriggean argument I am pursuing, the best explanation of this striking fact entails that there are necessary connections between experiences and dispositions. It will entails the thesis I call Necessary Connections:
Necessary Connections. Many kinds of experience are such that, necessarily, if one has an experience of that kind, then one has a corresponding disposition.
Necessary Connections is intended to cover whichever experiences and dispositions are covered by Systematic Coincidence. Whereas Systematic Coincidence merely tells us that those experiences and dispositions coincide, Necessary Connections tells us that they cannot fail to coincide, as a matter of metaphysical necessity. Take the coincidence of feeling itchy and being disposed to try to scratch, for example. If this coincidence is covered by Necessary Connections, then it is not a contingent quirk of human psychology, nor is it the product of any contingent laws of nature. Even if our psychology and our laws of nature were very different, the relevant conditional would still be true: if a subject feels itchy, then they are disposed to try to scratch. And the same goes for all other pairs of experience and disposition that are covered by Systematic Coincidence.
From Necessary Connections, it is only a short leap to the falsity of the possibility claim. If many kinds of experience bear necessary connections to dispositions, then presumably quench- and burn-experiences are among them. They certainly seem to coincide with certain sorts of dispositions. People who have burn-experiences tend to respond aversely: they tend to try to get rid of those experiences. Those who have quench-experiences tend to welcome those experiences: they tend to try to savor them. The details do not matter, so long as having the relevant dispositions is sufficient for caring about—and thus, not being indifferent to—the relevant experiences.
Putting this all together, the neo-Spriggean argument runs as follows:
Neo-Spriggean argument
P11: Systematic Coincidence
P12: The best explanation of Systematic Coincidence entails Necessary Connections.
P13: If Systematic Coincidence and P12, then Necessary Connections.
P14: If Necessary Connections, then necessarily it is impossible to be indifferent to paradigmatically pleasant and unpleasant kinds of experiences.
Conclusion: The possibility claim is false.
I believe that this argument succeeds where the original Spriggean argument fails.
I have made the preliminary case for P11, or Systematic Coincidence. And I have already considered how Systematic Coincidence might be challenged by appealing to kinds of sensory experience—like color experiences—which do not obviously coincide with kinds of dispositions. I responded by arguing that even in these cases, the relevant kinds of experience do plausibly coincide with kinds of dispositions. More to the point, Systematic Coincidence is consistent with the claim that some experiences do not coincide with dispositions, since it is a claim about many—not all—kinds of experiences. There are doubtless other cases in which it is not obvious how our experiences coincide with our dispositions, but my responses will be the same. First, even if the connection is not obvious, upon reflection there may be a nonobvious connection. Second, even if we find a case in which a kind of experience genuinely does not coincide with any disposition, this would not threaten Systematic Coincidence.
The premise doing most of the heavy lifting in the neo-Spriggean argument is P12. The idea is that Systematic Coincidence cries out for explanation, and the best explanation entails that there is some sort of connections between kinds of experiences and kinds of dispositions. Furthermore, the best explanation of Systematic Coincidence will have it that the connections are necessary. Thus, the best explanation of Systematic Coincidence is Necessary Connections.
One might object, along the lines of Rachels's original objection to Sprigge, that there are better explanations to be had. The most obvious alternative explanation is purely psychological:
Psychological Connections. Human beings are psychologically constituted such that, for many kinds of experience, if one has an experience of that kind, then one has a corresponding disposition.
According to the proponent of Psychological Connections, the connection between itching and scratching is merely a fact about human psychology. In principle, other sorts of creatures might be constituted such that itches do not at all dispose them to try to scratch. Indeed, for all we know, the universe contains many creatures who feel intense itches all day and night but never feel the least bit inclined to scratch. There may also be creatures whose feelings of thirst do not dispose them to drink, or whose color experiences do not dispose them to discriminate between things in their environment. On the proposal we are currently considering, these odd creatures cannot be dismissed as mere metaphysically possible oddities. Rather, we must be open to the idea that they exist in the actual world.
I predict that few will be willing to go this far. Among those who claim that itchiness can come apart from the disposition to try to scratch, most will claim that this is a mere metaphysical possibility. It is not something we think may actually happen. Those who adopt this line of thought should deny that the connection between our experiences and dispositions is merely psychological. They would be better served by claiming that the connection is nomological:
Nomological Connections. The actual laws of nature are such that, for many kinds of experience, if one has an experience of that kind, then one has a corresponding disposition.
In contrast with Necessary Connections, Nomological Connections tells us that Systematic Coincidence is a product of the actual laws of nature. Thus, if the world had different laws of nature, then our experiences could come apart from our dispositions in any number of ways—our feelings of itchiness might not dispose us to try to scratch, for example. Necessary Connections tells us that this is impossible, even with very different laws of nature.
I think that Nomological Connections is the most promising alternative to Necessary Connections. But even so, we ought to prefer the latter thesis. We should not be afraid of the claim that, as a matter of metaphysical necessity, feelings of itchiness dispose us to try to scratch.
To begin with, there are some intuitive grounds for preferring Necessary Connections. Phenomenal-dispositional coincidences do not appear to be metaphysically contingent in the same way that other coincidences appear metaphysically contingent. For example, oak trees lose their leaves in the winter, but this appears to be a metaphysically contingent sort of coincidence. It is easy to imagine that (if the laws of nature were different) oak trees might hold onto their leaves throughout the winter. It is similarly easy to imagine that (if the laws of nature were different) there might be lightning without thunder, or fire without smoke. In contrast, even permitting ourselves to imagine worlds in which the laws of nature are very different, it is hard to imagine cases in which our phenomenology comes apart from our dispositions. Imagine feeling extremely itchy, but having no tendency to try to scratch. Of course it is easy to imagine feeling itchy, but lacking any disposition to scratch. One can imagine being a blob-like creature with no functional limbs, like the creature from ‘I Have No Mouth, and I Must Scream’ (Ellison Reference Ellison1967). In that case, one cannot be disposed to scratch, since one is simply incapable of scratching. However, one can still try, in the sense of making a futile mental effort. The relevant imagined scenario is one in which we are not even disposed to try. And that is simply very hard to imagine.
Now returning to the Spartan- and Ascetic-style cases: imagine cases in which subjects have no disposition to respond with attraction or aversion to their quench- and burn-experiences, respectively. These cases strain our powers of imagination in a way that the preceding cases—such as fire without smoke, lightning without thunder—do not. And this provides us with some grounds for thinking that phenomenal-dispositional connections are different. On reflection, the connection between feeling itchy and trying to scratch, or between pain and aversion, seems tighter than the connection between lightning and thunder or fire and smoke. Whereas the latter connections seem to be merely nomologically necessary, this does not seem adequate for capturing the connections between experiences and dispositions.
Relatedly, our experiences seem to explain our dispositions. If you know that I feel itchy, you can reasonably infer that I am disposed to try to scratch. You can make this inference just in virtue of knowing ‘what it is like’ to feel itchy. In contrast, I cannot make this sort of inference in ordinary cases of nomological connections. I am familiar with the phenomenon of lightning, but this familiarity alone does not license the inference that lightning is followed by thunder. I know that a given flash of lightning will be followed by thunder only because I know that, as a matter of fact, lightning is reliably followed by thunder. No amount of thinking about lightning as such will reveal its connection to thunder. Similarly, no amount of thinking about fire as such will reveal its connection to smoke, and no amount of thinking about oak trees as such will reveal when they lose their leaves. These connections are opaque. We learn of them—in the first instance—by repeated observation, as is typical of merely nomological connections.
If we accept Necessary Connections, we will be well positioned to explain why repeated observation is not necessary for knowing that itches dispose us to try to scratch. We can claim that, in making this inference, we are latching onto the necessary connection between the experience and the disposition. For example, we might say that the property of having an itch is identical with the property of being disposed to try to scratch (or to some other, more complicated disposition). Alternatively, we might say that feeling itchy analytically entails having the disposition to try to scratch. Or we might say that the experience grounds the fact the disposition to try to scratch. The important point is that, to the extent that we are at least dimly aware of this necessary connection, this awareness can explain how we know that itches tend to cause us to try to scratch. In particular, we need not posit laws of nature that connect our experiences with our dispositions. But this is what we must do if we accept nomological connections. We must say that the connection between itching and scratching is a nomological connection, just like the connection between lightning and thunder. Just as there is nothing in the nature of lightning that connects it with thunder, so too is there nothing in the nature of itchiness which connects it with trying to scratch. Just as the connection between lightning and thunder is the sort of thing we learn on the basis of induction, so, too, must we rely on induction to learn that itches tend to make us try to scratch. But all this seems false—in these respects, the relationship between itching and scratching is not like the relationship between lightning and thunder. All this suggests that there is a more than merely nomological connection between feeling itchy and being disposed to try to scratch. On balance, then, we ought to think that the connection is necessary. We ought to accept necessary connections.
The reasons I have given in support of P12 are closely related to reasons that Hedda Mørch has given in support of the phenomenal powers view (Reference Mørch2014, Reference Mørch2017, Reference Mørch and Coleman2019a). Like Mørch, I contend that there is a peculiarly tight epistemic connection between our experiences and our dispositions. Also like Mørch, this contention leads me to think that there are metaphysically necessary connections between experiences and dispositions. But unlike Mørch, I remain neutral about what ultimately explains these metaphysically necessary connections. For all I have said, the necessary connections may be explained by phenomenal powers. But they may be explained by something else, such as analytic connections between experiences and their functional roles, or by phenomenal-dispositional identities. I only contend that the metaphysical connection between experience and disposition must be sufficient for a necessary connection, so that our awareness of this connection can explain the peculiarly tight epistemic connection between our experiences and our dispositions.
In comparison with P11 and P12, P13 should be fairly uncontroversial. That premise tells us that if the best explanation of Systematic Coincidence requires that we posit necessary connections between certain kinds of experiences and dispositions, then that is the explanation we should accept. That much seems hard to deny.
In contrast, P14 is perhaps somewhat more open to controversy. That premise tells us that if many kinds of experience are necessarily connected with corresponding dispositions, then, in particular, paradigmatically pleasant and unpleasant experiences are connected with corresponding dispositions. Furthermore, those dispositions are such that having them is sufficient for not being indifferent to them. This gets us the result that necessarily, it is not possible to be indifferent to paradigmatically pleasant and unpleasant experiences.
One might challenge P14 by looking for an actual case in which a subject is indifferent to their pleasant or unpleasant experiences. For example, these are certain anomalous cases in which a subject reports that they feel pain, but reports that this pain is not unpleasant. (Though some philosophers have doubted it, I take for granted that these abnormal pains can be properly described as pains; for discussion see Park [Reference Park2019].) Abnormal pains can occur in subjects who are on strong painkillers. They can also occur in subjects with a rare brain condition called pain asymbolia. According to one interpretation of these cases, the anomalous subject's pain feels exactly the same as any normal subject's pain. That is, abnormal pains are the same kind of experience as ordinary pains. The only difference is that, whereas ordinary subjects tend to dislike their pains, the anomalous subjects are indifferent to them. This subjectivist interpretation is endorsed by Richard Brandt (Reference Brandt1979: 37–38), Parfit (Reference Parfit1984: 501), and Richard Hall (Reference Hall1989), all of whom leverage the relevant cases in support of subjectivism. If the subjectivist interpretation is correct, then it would constitute a dramatic refutation of the Spriggean view.
It is clear enough that Spriggeans need to reject the subjectivist interpretation. They need to say that the overall experience of abnormal pain differs phenomenologically from the overall experience of ordinary pain. Notice that this is a very general interpretation—there are any number of different ways in which the overall experience of abnormal pain might differ phenomenologically from the experience of ordinary pain. Different proposals have been suggested by Rachels (Reference Rachels2000: 199–200), Nikola Grahek (Reference Grahek2007: 2; 111–12), Colin Klein (Reference Klein2015: 158–60), Hedda Mørch (Reference Mørch2014: 121–23), and David Bain (Reference Bain2014: 306–7, 319). Each proposal is a variation on the phenomenological interpretation (or can be charitably interpreted as such). If any one of these proposals is correct, then abnormal pains do not threaten the Spriggean.
It would be a mistake to try to sort through all of the empirical data which bears on the subjectivist and phenomenal interpretations of abnormal pains. For present purposes, it is enough to note that interpreting the empirical data is no easy task. There is a dearth of reliable reports from subjects who experience abnormal pains. Subjects with pain asymbolia are exceedingly rare, and they often have severe language deficits (Klein Reference Klein2015: 512n17). Subjects on strong painkillers are not overly articulate because they are high on drugs (Rachels Reference Rachels2000: 199). Furthermore, subjects’ reports are not easily interpreted. Subjects tend to report that their pains are not unpleasant—but that is not the same as reporting that their pains feel exactly the same as ordinary pains, nor does it rule out that the difference in unpleasantness amounts to a difference in how their overall experience feels. There are some empirical studies that suggest that the unpleasantness of pain can be experienced independently of pain itself (Ploner, Freund, and Schnitzler Reference Ploner, Freund and Schnitzler1999). If this suggestion is correct, then perhaps abnormal pains differ from ordinary pains in virtue of lacking this component of experiential unpleasantness. There is much more to be said about the asymbolia and painkiller cases, but suffice it to say that the subjectivist interpretation is far from mandatory. Spriggeans can and should reject it without embarrassment.
Opponents of P14 might instead turn to a more prosaic sort of case: differences in taste. I love the fizzy feeling of drinking seltzer water—for me, that experience is paradigmatically pleasant. But my friend Paul is indifferent to the fizzy feeling. On the assumption that Paul is getting the same kind of experience as me, it follows that Paul is indifferent to a paradigmatically pleasant experience. From consideration of this and other differences in taste, one might conclude that even if we accept Necessary Connections, we ought to deny that we necessarily care about paradigmatically pleasant experiences.
It is clear how a Spriggean should respond to this line of thought: drop the assumption, and claim that Paul—as well as others who are indifferent to the fizzy feeling—are in fact getting a different kind of experience than me. To be sure, the Spirggean need not say that Paul and I are having entirely different sorts of experiences. It is not as if he does not feel the bubbles or the coolness of the water. His experience, like my experience, is an experience of drinking cool seltzer water. But Spriggeans should say that Paul's overall phenomenology differs from mine—’what it's like’ for Paul to drink seltzer is not quite the same as ‘what it's like’ for me to drink seltzer. Elsewhere, I defend this way of thinking about differences in taste in greater detail (Reference PalliesPallies forthcoming). For present purposes, it is enough to note that this is how Spriggeans should respond to differences in taste: they should say that what it is like to drink seltzer and like it is different from what it is like to drink seltzer and be indifferent to it. (For a defense of this response, I defer to my arguments elsewhere [Reference PalliesPallies forthcoming].)
In general, we ought to think that there are necessary connections between our phenomenology and our dispositions. This is the best way to account for the systematic harmony between how we feel and what we do. As a consequence of this general commitment, we ought to accept Sprigge's thesis. We ought to think that ‘pleasures and pains are of their nature liable to affect behaviour in certain directions’ (Sprigge Reference Sprigge1987: 142). In particular, they affect us in such a way that we are not indifferent to them.
6. Conclusion
I have argued for a conciliatory solution to the pain and pleasure problems. Objectivists are right to endorse the objectivist claim, and subjectivists are right to endorse the subjectivist claim. Both camps of ethicists are correct in their ethical claims, but both go wrong in endorsing the metaphysical possibility claim.
In closing, I would like briefly to consider how this Spriggean view bears on what I have called the explanation debate. Recall that this debate concerns the following two claims:
Objectivist explanation. Whenever an experience is non-derivatively good or bad for a subject, it is non-derivatively good or bad for that subject in virtue of being an experience of the kind that it is.
Subjectivist explanation. Whenever an experience is non-derivatively good or bad for a subject, it is non-derivatively good or bad for that subject in virtue of their caring about it.
The Spriggean view makes some negative progress toward resolving this debate. It tells us that we cannot make progress by appealing to the Ultra-Spartan and Ultra-Ascetic arguments. Those arguments, if successful, would disprove subjectivist and objectivist explanations, respectively. But they are not successful, so they do not disprove those explanations. Different arguments are needed.
More ambitiously, the Spriggean view suggests a conciliatory solution to the explanatory debate. Consider the debate as it pertains to quench-experiences. Objectivists will claim that those experiences are good for us because of their particular phenomenology, and subjectivists will claim that they are good for us because we care about them. But given that we necessarily care about experiences with the total phenomenology of a quench-experience, we should ask: To what extent is the attitude of caring metaphysically distinct from the quench-experience? If they are wholly distinct, then there is still an important debate to be had between objectivists and subjectivists. But suppose they are not wholly distinct. Suppose, for example, that caring about one's quench-experience is fully grounded in—is nothing over and above—having a quench experience. Then the explanatory debate looks rather more fiddly and less pressing. (Compare: is it good for us to have relationships of mutual admiration, affection, and respect—or is it good for us to have friendships?) And in the limit, if one's caring about one's quench-experience is simply identical with one's having a quench experience, then it is unclear that there is any room for an explanatory debate at all. Settling this issue is a topic for future work. For present purposes, it is enough to note that the Spriggean view suggests a path forward for resolving the debate.
Whatever we conclude about the explanatory debate, the Spriggean view entails that there is much about which objectivists and subjectivists agree. For all possible experiences, the objectivist and subjectivist can agree about whether or not those experiences are good or bad for us. Consider, for example, the claim that all possible burn-experiences are bad for us, and all possible quench-experiences are good for us. This is clearly a significant ethical claim. And contrary to what is regularly assumed, it is a claim about which objectivists and subjectivists can agree.
