p27

As we have seen earlier, the most obvious kinds of counterexamples – e.g. mass, charge, spin, etc. – can be responded to by a structuralist stance towards physics. Physics only tells us about the causal role that these kinds of properties play, rather than any intrinsic qualities underlying this role. However, one might argue that geometrical properties, like being spherical, might constitute a different kind of counterexample. On the face of it, properties like being spherical seem to be intrinsic properties that have a non-dispositional nature. There are several things to say about such geometrical examples. First, being spherical does not seem to be a fundamental property, and by “quality” I mean to only be referring to (intrinsic) fundamental qualitative quiddities. More generally, a standard view is that spatiotemporal structure is grounded in fundamental spatiotemporal relations, such as distance relations, that ground shape 33 See Lee (2019) for further discussion of the possibility that our introspectable macrophenomenal properties are realized by non-introspectable microphenomenal properties. 34 Technically speaking, it might be that the space of phenomenal qualities is a disunified subspace of the (unified) space of all possible qualities. But this seems implausible: phenomenal qualities seem more unified to each other than they are to allegedly non-phenomenal qualities. 13 properties like being triangular or being spherical (e.g. a macroscopic object is only spherical because of the spatial relations that its parts stand in). Such a view is perfectly compatible with the claim that all possible qualities are phenomenal qualities, since such a view does not posit any fundamental spatiotemporal qualities. It should also be noted that the view that there are only fundamental spatiotemporal relations (rather than qualities) is compatible with a substantival view of space-time, according to which there are fundamental space-time points, which are themselves structureless entities that lack any spatiotemporal extension and dimensionality, that stand in various spatiotemporal relations to one another.35,36 A second point to make is that there are strong independent reasons for thinking that, at least in the context of science, spatiotemporal concepts pick out functional properties rather than intrinsic qualities. For example, one could have a phenomenological analysis of spatiotemporal concepts, where spatiotemporal concepts pick out whatever properties play a role in bringing about certain spatial/temporal experiences, or a non-phenomenological analysis of spatiotemporal concepts, where spatiotemporal concepts pick out whatever properties play a suitable “spatiotemporal role” in fundamental physical theories.37 Such a functionalist approach is needed to account for a variety of fundamental physical theories, such as certain philosophical interpretations of quantum mechanics and certain speculative theories of Quantum Gravity, where spacetime is “emergent” rather than fundamental.38 A third and last point to make is that one could just as easily modus tollens potential geometric counterexamples to Modal Idealism by means of the argument in the previous section. On the face of it, phenomenal qualities and wholly non-experiential geometric qualities seem radically different in kind. If the space of possible qualities consisted of a space of phenomenal qualities together with a totally disjoint space of geometric qualities, this would cry out for explanation. What could explain why there are only these two specific kinds of qualities, rather than some other third kind of quality? In the absence of any kind of unity between the space of phenomenal qualities 35 See Pooley (2013) for an overview of the debate between substantivalism and relationism. One might think that an ontology of space-time in terms of space-time points might still need to posit fundamental intrinsic geometrical properties to space-time points, such as the property of being point-sized. But being point-sized might be better understood as a “merely negative” property, such as lacking spatial extension and lacking dimensionality, and merely lacking a certain property should not be understood as a fundamental property itself (see note 43 for more discussion). Sider (2006) discusses this point further, writing that “a natural and economic theory of points of spacetime is that each one is a partless, truly bare particular that stands in a network of spatiotemporal relations” (393). 36 Substantivilist views of space-time need not be committed to a fundamental ontology of space-time points. For example, Schaffer (2009) argues for a monistic conception of space-time, according to which the spatiotemporal manifold as a whole is the only fundamental entity. 37 Both such analyses are explored in Chalmers (2021), focusing mostly on the spatial case. Also see Lam and Wüthrich (2018) for more on the relevant functional role of spacetime in physical theories. Chalmers also discusses whether we might have a primitive (non-functional) concept of “Edenic shape” properties that are represented in perception. I have two responses to this. First, perhaps Edenic shape properties are also better thought of as non-fundamental properties that are grounded in Edenic spatial relations, and second, I will be further discussing the general class of “Edenic qualities” later in this section. 38 For more on spacetime emergence, see Huggett (2021). 14 and geometric qualities, it’s unclear how there could be an answer to this question. But if there is no answer to this question, then there would be brute necessary facts about the space of possible qualities. Let us now turn to a second kind of counterexample, which is inspired by perception rather than physics. Within perception, it is natural to think that our perceptual states represent external objects as having certain primitive non-phenomenal qualities, which Chalmers (2006) calls “Edenic” qualities. For example, color experience seems to represent the surfaces of external objects as being primitively and qualitatively colored, where there is no a priori connection between such primitive Edenic colors and (say) the surface-reflectance properties studied by physics. Naïve Realists about perception believe that external physical objects really have such primitive (nonphenomenal) qualities, and they typically appeal to such Edenic qualities to explain the phenomenal character of our (veridical) perceptual experiences. In my own view, there are good reasons for thinking that external objects don’t have Edenic qualities. However, the mere fact that external objects don’t actually have Edenic qualities doesn’t let Modal Idealism off the hook. If it is even possible for there to be (non-phenomenal) Edenic qualities, then Modal Idealism would be false. Moreover, Edenic qualities have the interesting feature that they seem to be very closely related to phenomenal properties, so they might serve as a counterexample to Modal Idealism that also avoids our argument in the previous section. In particular, perhaps Edenic qualities are sufficiently closely related to phenomenal qualities that they would not render the space of possible qualities “disunified” in the way that other kinds of qualities might. The purpose of the rest of this section will be to address this important potential counterexample to Modal Idealism. However, before I begin to argue that Edenic qualities are impossible, I want to make a (less important) dialectical point. In section 3, I flagged that I would be making the assumption that experiential phenomena cannot be reduced to non-experiential phenomena, contrary to Physicalists and “Panprotopsychists”. However, it is natural to think that, if Edenic qualities are possible, then it would be possible to reduce facts about experience to facts about Edenic qualities (together with other (non-phenomenal) psychological facts). After all, one of the main motivations for positing Edenic qualities is precisely to account for our (perceptual) phenomenal experiences in terms of such Edenic qualities.39 So, technically speaking, the possibility of Edenic qualities might not be relevant to the claim that I wish to defend. The metaphysical possibility of Edenic qualities does pose a counterexample to the unconditional claim that “Modal Idealism is true”, but it’s unclear whether it poses a counterexample to the conditional claim that “if phenomenal qualities cannot 39 This is one of the main motivations behind “Panqualityism” (see Chalmers (2013) and Coleman (2017)). Although this is one of the main motivations for positing Edenic qualities, see Cutter (2022) and Mihalik (2022) for arguments that this motivation cannot succeed. 15 be reduced, then Modal Idealism is true”, because the possibility of Edenic qualities might make the antecedent of that conditional false. Having made this dialectical point, let us set it to the side and see whether a case can be made that Edenic qualities are impossible. In broad strokes, I think that Edenic qualities only seem to be possible when we narrowly focus on Edenic colors. When we broaden our focus to the general category of Edenic qualities, we find that the general concept of an Edenic quality starts to unravel, and we find that there are good reasons to reject the general concept as incoherent. Here is a formal way to state my argument against the possibility of Edenic qualities. If Edenic qualities were possible, there would have to be a precise and non-arbitrary answer to the following question: Question: What is the space of all possible Edenic qualities? However, there is no precise and non-arbitrary answer to this question. So, we should reject the possibility of Edenic qualities. This way to put the argument is rather formal and unintuitive, but I hope that, as I defend it, it will start to become more intuitive why Edenic qualities are not possible. Start with the first step: why think that there has to be a precise and non-arbitrary answer to this Question? Because different answers to this Question entail different views about the space of possible fundamental qualities. If the answer to this question was arbitrary, then it would imply that the structure of modal space was arbitrary. However, as we’ve seen before, one of the main assumptions of this first argument is that modal space is not arbitrary. If the answer to this question was vague rather than precise, then, because I do not think that the concept of a “possible fundamental quality” is semantically vague, such an answer would commit to metaphysical vagueness. However, following most philosophers, I think that vagueness should be understood only as a feature of how we represent reality, rather than as a feature of reality itself.40 The next step is the important one: why think that there is no precise and non-arbitrary answer to this Question? Well, to start, the concept of an “Edenic quality” is supposed to pick out a (nonphenomenal) quality that a possible phenomenal experience can represent an object as having. So, in order for there to be answer to our Question, there would have to be answers to the following two questions: Question-1: Which possible phenomenal experiences represent objects as having a corresponding Edenic quality? 40 For an overview of arguments against metaphysical vagueness, see Barnes (2010). 16 Question-2: Among the Edenic qualities that possible experiences can represent objects as having, which are possible? The purpose of the first question is to pick out which Edenic qualities are represented by possible experiences, and the purpose of the second question is to decide, of the Edenic qualities picked out by the first question, which are genuinely possible. Note that the second question is needed because it is consistent to hold that (say) visual experience represents external objects as having Edenic colors, but nevertheless such Edenic colors are not genuinely metaphysically possible. Once we have answers to these two questions, then we immediately have an answer to our main Question. Conversely, it seems like any answer to our main Question will have to presuppose corresponding answers to Question-1 and Question-2. What is left is to argue is that there are no precise and non-arbitrary answers to Question-1 and Question-2. If neither question has a precise and non-arbitrary answer, then our main Question will not have a precise and non-arbitrary answer either. In fact, I really only need to argue that one of these two questions lacks a precise and non-arbitrary answer, but in fact I believe that neither question has a precise and non-arbitrary answer. Start with Question-1. It is most common to think that paradigm cases of perceptual experiences represent objects as having primitive non-mental qualities, but we can also ask more broadly: which phenomenal experiences in general represent objects as having Edenic qualities? When I start reflecting on this question, I find that I quickly start losing my grip on the question. Consider the following experiences: feeling anxious, feeling happy, feeling sad, feeling dizzy, feeling nauseous, feeling sexually aroused, having cognitive phenomenology (associated with thought and understanding), feeling the need to urinate, feeling nostalgia, feeling anger, feeling regret, feeling hot, feeling cold, etc. All of these experiences have a distinctive phenomenology associated with them, but for which (if any) of these experiences is there a corresponding primitive non-mental Edenic quality that is represented as being instantiated by some object? I’m not at all sure. But when I reflect on experiences of these various kinds, it seems pretty clear to me that there is no (non-arbitrary) precise line to be drawn between those experiences that represent objects as having primitive Edenic qualities, and those that do not. Moreover, my source of puzzlement about which of these experiences have corresponding Edenic qualities is peculiar. For example, when I try attending to an experience of feeling dizzy, it seems like I know perfectly well what that experience is like.41 Still, even given complete knowledge of the phenomenal character of dizzy-experiences, it’s still unclear whether the idea of a corresponding “Edenic dizziness” makes any sense (or “Edenic nausea”). This makes me wonder 41 See Lee (forthcoming) for further clarification and discussion concerning the extent of our knowledge of our conscious experiences. 17 whether the source of my puzzlement is really puzzlement about what it is to be an “Edenic quality” in the first place. In response to Question-1, one might simply say the following: a phenomenal experience represents a corresponding Edenic quality in some object if and only if that phenomenal experience is a perceptual experience. My first response to this suggestion is that it’s not clear that it is extensionally adequate, and my second response to this suggestion is that I don’t think the distinction between a “perceptual” experience and a “non-perceptual” experience is a precise distinction anyway (and what is needed is a precise and non-arbitrary answer to Question-1). Consider, for example, experiences like being in pain, experiencing an orgasm, feeling an itch, or feeling the need to urinate. It is fairly plausible that such experiences represent certain Edenic qualities in different portions of our bodies. For example, a pain in my foot plausibly represents a certain kind of primitive quality in my foot. However, are such bodily experiences supposed to count as “perceptual”? If perceptual experiences are defined as those that are supposed to give us information about the “external world”, then this question boils down to the question of whether our bodies count as part of the “external world”. This question seems to be a merely verbal question, but the question of which Edenic qualities there could possibly be should not be a merely verbal question. It should be a question of fundamental metaphysics. Sometimes perceptual experiences are thought of as those that we do not have voluntary control over (e.g. visual or auditory experiences), as contrasted with (say) thought or imagination, where we do seem to have more voluntary control. However, feeling dizzy and feeling nauseous are experiences that we do not have voluntary control over, and it is not at all clear whether such experiences have corresponding Edenic qualities. Lastly, consider the experiences we have while dreaming, or while imagining. Perhaps it is most natural to categorize such experiences as “non-perceptual”, but it’s not clear whether such experiences lack any corresponding Edenic qualities, contrary to the suggestion that an experience represents a corresponding Edenic quality if and only if it is perceptual. Having argued that there is no precise and non-arbitrary answer to Question-1, let us now turn to Question-2. Of the possible Edenic qualities that phenomenal experiences can represent objects as having, which are genuinely possible? It seems to me that at least some of them are not genuinely possible. Consider an experience of intense pain in your foot (perhaps you just stubbed your toe). It seems like such an experience represents a certain Edenic painful quality as inhabiting your foot. However, Edenic qualities are supposed to be non-mental qualities, and such non-mental qualities should be able to be instantiated by physical objects in the absence of any experiencers (just like Edenic colors). But could there really be an intensely painful quality that is not experienced by anyone? For example, could a rock floating in space instantiate an Edenic quality of intense pain, 18 without being experienced by anyone? It seems like it couldn’t. Intense pain seems to have an intrinsically motivating force to it that it doesn’t make sense to ascribe to an unexperiencing floating rock. Perhaps in response one could say that, while Edenic pain is possible, it is nevertheless necessary that, whenever it is instantiated, there must be a corresponding experiencer that is “acquainted” with the relevant Edenic pain. There are two worries with this response. First, if it really is necessary that Edenic pain must be experienced whenever it is instantiated, then it is unclear why Edenic pain would still count as a “non-mental” quality in the first place. Second, insofar as we can make sense of a non-mental quality that is necessarily experienced, such a view would seem to be positing a brute necessary connection between wholly distinct existences, and it was an assumption of my original argument that there can’t be brute necessities of this kind. Suppose, then, that at least Edenic pain is impossible. Then so long as some Edenic qualities are impossible, the question immediately arises as to where is the precise line between those Edenic qualities that are possible and those that are not? Prima facie, it’s hard to see where such a precise line should be drawn. Let us think about some other experiences. Consider the taste experience of something that is really spicy. Is Edenic spiciness possible? Well, sufficiently spicy experiences can be, in a way, painful: they can be very unpleasant in a way that quickly makes us try to relieve the spiciness. So, if Edenic pain is impossible, then it seems like intense Edenic spiciness should be too. However, once one says that intense Edenic spiciness is impossible, what should one say for mild Edenic spiciness? Since there doesn’t seem to be any non-arbitrary precise line to draw, it seems like we should say that, no matter how mild it is, no Edenic spiciness quality is possible.42 Similarly, consider sourness. Intensely sour things have very different effects on us than intensely spicy things, but in a way, they are also unpleasant, and so for the very same reasons, it seems like Edenic sourness (no matter how mild) should be thought to be impossible. Similar things can be said for sweet and bitter taste experiences (although mildly sweet experiences can be pleasant rather than unpleasant). Once one sees this pattern, it is easy to apply all over the place. For example, consider a very loud auditory experience. Such an experience is unpleasant, and so for similar reasons one should think that a corresponding loud Edenic quality is impossible. But, because there is no nonarbitrary dividing line, one should similarly conclude that any auditory Edenic quality, no matter how soft, is impossible. Similar things can be said for (good-smelling or bad-smelling) olfactory Edenic qualities. One might think that these kinds of examples suggest an interesting response to Question-2: perhaps only phenomenal experiences with neutral valence (i.e. ones that are neither pleasant nor unpleasant) have corresponding Edenic qualities that are possible.43 This is an intriguing suggestion, but it’s not clear that it works. For one, consider a holistic taste experience that has 42 A structurally similar argument was given by Berkeley in his Three Dialogues between Hylas and Philonous. 43 The exact nature of valence is contested, but see Carruthers (2018) for two contrasting accounts. 19 both positive and negative aspects to it, but these positive and negative aspects exactly “cancel out” to result in a holistic taste experience that has overall neutral valence (e.g. perhaps something is sweet and spicy, and the pleasantness of the sweetness exactly cancels out the unpleasantness of the spiciness). Is it possible for there to be a corresponding “holistic” Edenic taste quality? Insofar as I have intuitions about this, the answer seems to be “no”: so long as there is an aspect to the experience that is not neutrally valenced, then there needs to be a corresponding experiencer who is affected by that valenced-aspect (just like the pain case). In response, one might suggest that only phenomenal experiences that don’t have any aspect to them that is positively or negatively valenced have corresponding Edenic qualities that are metaphysically possible. One worry about this view is that it’s not clear whether it picks out a precise class of phenomenal experiences, since it is not clear if there are precise facts about “all aspects” of an experience (or even whether there are precise facts about whether a particular aspect is exactly neutrally valenced). It is also unclear whether there are any phenomenal experiences that satisfy this description in the first place (partly because it’s unclear how to make sense of “every aspect” of an experience). Lastly, the project of stripping away all possible valence out of an experience doesn’t seem to be getting at the heart of the issue.