There Is More Than One Thing

There Is More Than One Thing Philip Goff For those with a taste for desert landscapes, the existence monism of Horgan and Potr 1 is the ultimate thrill. No tables, no chairs, not even particles in the void. Just the One jelly-like blobject, wobbling away or to put it more precisely: instantiating various properties in spatiotemporally local manners. Unfortunately, this beautifully minimalist view cannot possibly be true, for it is inconsistent with certain thing we know to be true. At this moment, I am experiencing no pleasure, but I do feel a terrible pain in my knee. I know that there is something that feels pain but no pleasure. Even if it turns out that I am in The Matrix, this is something the evil computers cannot be deceiving me about. Sitting across from me is Dave, who is eating an ice cream. He tells me he is currently enjoying a pain free existence, and is in fact extracting great pleasure from his ice cream. If common sense is to have any sway in my metaphysical speculation, then I must set aside scepticism about other minds when compiling the data a theory must explain. I therefore take it that Dave, in his own case, has certainty that there is something that currently feels pleasure but no pain, akin to the certainty I have that there is currently something that feels pain but no pleasure. It follows that there are currently in existence two things: one that feels pain but no pleasure, and another that feels pleasure but no pain. Therefore, existence monism is false. Of course, this is too quick. Horgan and Potr accept the truth of all kinds of sentences which quantify over a plurality of objects, but deny that the truth of such sentences is a matter of a direct correspondence between the quantificational posits of the sentence and the items of the world. Rather, such sentences are made true by an indirect correspondence between the sentence and the one blobject instantiating properties in spatiotemporally local manners. The sentence T: There are two tables in regions 1 and 2 of my front room is true in virtue of an indirect correspondence between T and the state of affairs of the blobject instantiating certain physical properties R1-ly and R2-ly. Given their respect for common sense, I take it that Horgan and Potr will accept that sentence S: There currently exists something that feels pain but no pleasure, and something that feels pleasure but no pain is true, but will hold that the truth of S is a matter of an indirect correspondence between S and a state of affairs involving the blobject having properties in spatiotemporally local way, rather than a direct correspondence to a state of affairs involving two distinct things. There seem to be two ways in which this might be done. Firstly, a physicalist proposal, according to which S is made true by the blobject instantiating physical properties in spatiotemporally local ways. Secondly, a non-physicalist proposal, according to which S is made true by the universe instantiating irreducible phenomenal properties in spatiotemporally local ways. I shall consider and reject both of these proposals in turn. The physicalist proposal 1 Horgan and Potr 2008; this volume. S is made true by the fact that the blobject instantiates c-fibres firing R1-ly but not R2-ly, and instantiates p-fibres firing R2-ly but not R1-ly (where c-fibres firing is the physical property which underlies pain, p-fibres firing is the physical property that underlies pleasure, and R1 and R2 are spatiotemporally local manners of instantiation). Sentences make demands on reality, such that if those demands are satisfied the sentence is true. In order for the physicalist proposal to be true, S must make demands on the world such that those demands are satisfied by the universe instantiating physical properties in spatiotemporally local manners. In order to assess the physicalist proposal, we must first answer the following question: are the demands S makes on the world a priori accessible (for someone who understands S, and in virtue of understanding S)?2 Let us say that a sentence is transparent iff it is a priori for someone who understands the sentence and has sufficient rational powers what that sentence requires of the world for its truth; There is a Euclidean sphere , There is an omnipotent being , seem plausible examples of transparent sentences. The sentences, The Earth is getting hotter , Everything is made of Gold seem plausible examples of non-transparent sentences.3 I will assess the physicalist proposal first under the assumption that S is transparent, and secondly under the assumption that s is non-transparent. S as transparent It is not plausible that it is a priori that pain is c-fibres firing and pleasure p-fibres firing; one can competently use the terms pleasure and pain without knowing any brain science. But if (i) S is transparent, and (ii) the physicalist proposal is true, then there must be a priori accessible demands of S which are satisfied by facts about the blobject s instantiation of c-fibres and p-fibres. Therefore, for the physicalist-transparency combo to possess a modicum of plausibility it must involve a commitment to some kind of analytic functionalism, which will provide a conceptual middle man between the mental and the physical. The claim must be that we know a priori that to feel pleasure is to be in the state of having some state that is playing the pleasure role, and to feel pain is to be in the state of having some state that is playing the pain role, together with the claim that as it happens p-fibres firing and c-fibres firing respectively play these roles.4 It is only by taking this detour through functional facts that we can make it at all plausible that I can know a priori what S demands of reality, such that those demands are satisfied by physical facts about the blobject. Unfortunately, this reductionist proposal is demonstrably false, not in the sense that one can produce premises that are inconsistent with it, but in the sense that one can make apparent to oneself its falsity by taking Cartesian doubt to its natural end. I can doubt that I have causal powers, More precisely: sentence S makes P a priori accessible iff there is a possible world where someone comes to know that P a priori, in virtue of understanding S. 3 It may be that something of what a non-transparent sentence demands from the world is a priori accessible. For a sentence to be transparent it must reveal everything that it demands from the world. 4 I will charitably assume that Horgan and Potr can give an account of the blobject having a state that plays the pain role, in terms of its instantiation of physical properties in spatiotemporally local manners. 2 but I cannot doubt that I am currently in pain. By supposing that I have no causal powers, whilst at the same time supposing that I am in pain, I find I am conceiving of myself as something that feels pain but has no causal powers. If it were an analytic truth that to feel pain is to be in the state of having some state that is playing that pain role, this would not be conceivable. Therefore, it cannot be analytic that to feel pain is to be in the state of having some state that is playing the pain role.5 Note that this is not a repeat of Descartes argument for dualism, and hence is not subject to well known critiques. There is no dodgy move here from epistemic premises to a metaphysical conclusion. Rather the move is from an epistemic premise that I can conceive of myself as a thing that feels pain but has no causal powers to an epistemic conclusion it cannot be part of the concept of feeling pain that to feel pain is to be in the state of having some state that plays a certain causal role. We are currently trying to make sense of the view that that the demands of S are both a priori accessible, and satisfied by the physical facts. A commitment to a certain view about our mental concepts namely analytic functionalism is required to make sense of the demands of S having both these features, and that view about mental concepts is inconsistent with certain evident facts about what is conceivable.6 S as non-transparent I turn now to assessing the physicalist proposal under the thesis that S is non-transparent. In the last twenty years, many physicalists wrestling with the hard problem of consciousness have been drawn to semantic externalism about phenomenal concepts, by which I mean those concepts we employ when we think about our conscious states in terms of what it is like to have them.7 Such semantic externalism seems to offer hope of reconciling the conceptual gulf between the phenomenal and the physical with the metaphysical intimacy that these two realms are supposed to enjoy according to physicalism. If the content of the phenomenal concept <pain> is determined by facts outside of what is a priori accessible to the concept user, then this explains why the physical nature of pain is not a priori accessible to the concept user, and consequently why it is conceptually coherent to suppose that pain is not a physical state. Semantic externalism offers hope of making sense of physicalism without relying on analytic functionalism. If Horgan and Potr could make sense of S being non-transparent, then they could perhaps pursue a similar strategy for making sense of the physicalist proposal currently under 5 I give this argument from the conceivability of ghosts to the falsity of analytic functionalism in more detail in Goff forthcoming. The materialism of Lewis and Armstrong does not count as a view according to which S is transparent, although according to it we have a transparent understanding of the functional states we use to pick out mental states. The consideration I advance in favour of the transparency of S below, therefore, would count as reasons against the Lewis/Armstrong view. In Goff forthcoming, however, I also argue that the conceivability of ghosts is inconsistent with the Lewis/Armstrong view. 6 There are of course a number of other well known arguments against analytic functionalism, such as zombieconceivability argument and the knowledge argument (these arguments are intended as arguments against physicalism as such, but work best as arguments against forms of physicalism according to which the mental facts are a priori entailed by the physical facts). 7 Levine [1983]; Loar [1990; 2003]; Papineau [1993a; 2002]; Tye [1995]; Lycan [1996]; Hill [1997]; Hill and McLaughlin [1998]; Block and Stalnaker [1999]; Perry [2001]. consideration. They could simply claim that, as a matter of a posteriori fact, S demands that the there be two spatially local manners of instantiation, Rx and Ry, such that the blobject currently instantiates c-fibres firing Rx-ly but not Ry-ly, and instantiates p-fibres firing Ry-ly but not Rx-ly. This version of the physicalist proposal cannot be defeated by considerations of conceivability; if it is a posteriori that S demands that P, then the conjunction S but not P is conceivably true. Compare: it is a posteriori that There is water in the glass demands that there be H2O in the glass, and hence There is water in the glass, but there is no H2O in the glass is conceivably true. But is it plausible to hold that S is non-transparent? I will argue that it is not, by first arguing against the plausibility of semantic externalist accounts of phenomenal concept. It will be useful to sketch a bit more of a framework to help in thinking about the issues here. Let us distinguish between three kinds of property concept as follows: Transparent property concept: A property concept is transparent iff everything of what it is for the property denoted to be instantiated is a priori accessible (to someone possessing the concept, and in virtue of possessing the concept). Translucent property concept: A property concept is translucent iff something but not everything of what it is for the property denoted to be instantiated is a priori accessible (to someone possessing the concept, and in virtue of possessing the concept). Opaque property concept: A property concept is opaque iff nothing of what it is for the property to be denoted is a priori accessible (for someone possessing the concept, and in virtue of possessing the concept).8 A plausible example of a transparent concept is the concept <Euclidean sphericity>. For something to be spherical in Euclidean geometry is for it to have all points on its surface equidistant from its centre. For someone possessing this concept <Euclidean sphericity>, this information concerning what it is for something to possess Euclidean sphericity is a priori accessible. A plausible example of a translucent property concept is <being a Euclidean sphere roughly the same size as the Earth>. The property denoted here has two aspects: being a Euclidean sphere, and being roughly the same size as the Earth. The concept reveals what it is for the former aspect to be instantiated, but empirical investigation is required to know what it is for the latter aspect to be instantiated. The concept denoting the property formed of those two aspects, therefore, is translucent. If there are concepts which are such that reference is determined wholly by facts outside of what is a priori accessible to the concept user for example, by causal relations between concept and referent, by sub-personal recognitional capacities of the concept user, or by the evolutionary history of the concept then such concepts will be opaque. Clearly, then, semantic externalism about phenomenal concepts is committed to the opacity of phenomenal concepts. There are a couple of reasons why this view is implausible. 8 See footnote 1 for an analysis of a priori accessibility. Firstly, supposing that the phenomenal concept of pain is opaque makes it difficult to see how many of those normative judgements we generally take to be well-grounded could indeed be wellgrounded. I believe that the fact that someone is in severe, unnecessary pain, to be of great normative significance. I come to this conclusion not by doing experiments, but by reflecting on what it s like to feel pain. But for a judgement of the normative significance of a certain descriptive property to be well-grounded, I must surely have some understanding of the nature of that property, of what it is for it to be instantiated. Therefore, if my phenomenal concept of pain provides me with no understanding of what it is for something to feel pain, then the judgement of the normative significance of pain that I form when employing that concept must be ill-grounded. Secondly, supposing that phenomenal concepts in general are opaque makes it difficult to make sense of the fact that we have a priori knowledge concerning internal relations phenomenal qualities bear to each other. What it s like to see red is similar to what it s like to see orange. This is an internal relation these qualities bear to each other, that is, it s a relation which is entailed by the intrinsic nature of the relata. The mere existence of phenomenal red and phenomenal orange is sufficient for it to be the case that these two qualities resemble. If I have no a priori access to the nature of these qualities, it is hard to see how I could have a priori access to the fact that they have a nature such that they essentially resemble each other.9 These considerations make the thesis that phenomenal concepts are opaque overwhelmingly implausible. What about the thesis that phenomenal concepts are translucent? Making sense of this thesis requires dividing between the aspect of, say, how pain feels, that we transparently understand, and the aspect of how pain feels that we can denote but don t have a transparent grasp of. The trouble is that the conception of my pain I have when I reflect on it and think about it in terms of how it feels, seems too simple to be divided up in this way. To my knowledge, there has been only one proposal according to which phenomenal concepts are translucent, and that is Robert Schroer s proposal that phenomenal concepts have descriptive and demonstrative elements.10 According to Schroer, <phenomenal red> picks out phenomenal red under the description that property which is such and such a combination of phenomenal hue, phenomenal saturation and phenomenal lightness , where reference to these more basic elements of phenomenology is determined by facts outside of what is a priori accessible. Schroer s proposal is hopeless for a number of reasons which I have laid out in detail elsewhere.11 Here s a very simple but very serious difficulty with the view. In paradigmatic deployments of phenomenal concepts, we pick out phenomenal qualities in virtue of instantiating them in experience. I am looking at a red ball, and I turn my attention to how the red looks, and think about it as such. But this fact that I pick out phenomenal red in virtue of its presence in my current experience seems inconsistent with the thesis that I pick out phenomenal red descriptively as that property, whatever it is, that is composed of such and such a combination of certain elements. It is very hard to see how this fact and this thesis could be merged into a single account of the reference fixing of phenomenal concepts. I have defended these arguments against semantic externalism about phenomenal concepts in more details in Goff 2010. 10 Schroer 2010. 11 Goff MS. 9 If we know that phenomenal concepts reveal something of the nature of phenomenal qualities, and we cannot make sense of their being translucent, then we must suppose that they are transparent. But if we know a priori what it is for something to feel pain, and we know a priori what it is for something to feel pleasure, then we surely know a priori what the sentence There is currently something that feels pleasure, but not pain, and something that feels pain but not pleasure demands for its truth. All that has been added to the terms expressing transparent concepts to form this sentence is some quantifiers.12 In summary, 1. If S is transparent, then either analytic functionalism is true or the physicalist proposal is false. 2. S is transparent. 3. Analytic functionalism is false. 4. Therefore, the physicalist proposal is false. The non-physicalist proposal The demands of S are satisfied by the fact that the blobject feels pain R1-ly but not R2-ly, and feels pleasure R2-ly but not R1-ly. If a competent speaker of English inspected a world and found that there existed one thing that felt both pleasure and pain, she would not describe that world as one where There exists two things, one of which feels pleasure but no pain, and another of which feels pain but no pleasure , rather she would describe it as a world where There exists one thing that feels both pleasure and pain. Of course, the proposal under consideration is that the blobject experiences pain in one spatiotemporally local manner, and pleasure in another, quite distinct, spatiotemporally local manner. It s not easy to make sense of what it would be for a phenomenal quality to be instantiated in a spatiotemporally local manner, but let us suppose that this reflects our cognitive limitations rather than of an absence of possibility. Still, once we know about a world that it contains one thing that feels both pleasure and pain, we know enough about that world to know that it would not be appropriate to describe it as a world where S is true. I am not simply supposing here that truth must always consist in direct correspondence, that each of the objects quantified over in true sentence must exist in reality. I am inclined to agree with Horgan and Potr that there are numerous cases in which a state of affairs can make true a given sentence, despite the fact that the quantificational structure of the state of affairs does not mirror the quantificational structure of the sentence. Consider a world where there are no composite objects, This is not quite correct: we have also added reference to the present moment. Even if our temporal concepts are not transparent, it is difficult to see how this could help Horgan and Potr make sense of the physicalist proposal. 12 but in which there are particles arranged yellow-sofa-wise and brown-table-wise13 in my front room. I would be inclined to describe this world as one where the ordinary English sentence, There is a yellow table and a brown sofa in my front room is true. Ordinary English sentences are not used to convey information about the metaphysical structure of reality, and therefore an absence of direct correspondence should not be taken to entail falsity. Nor am I denying that the single blobject is capable of serving as truthmaker for many ordinary English sentences which quantify over numerous distinct objects. Consider a blobjectivist world where the blobject instantiates yellowness and solidity sofa-shaped-region-in-my-front-room-rightnow-ly but not table-shaped-region-in-my-front-room-right-now-ly, and browness and solidity tableshaped-region-in-my-front-room-ly but not sofa-shaped-region-of-my-front-room-ly. I would be inclined to describe this as another world where the ordinary sentence of English There is a sofa in my front room which is yellow but brown, and a table is my front room which is brown but not yellow. Appreciation of the everyday purposes to which such a sentence is put, e.g. getting round the room, achieving certain goals such as sitting down,14 makes it implausible that the sentence demands a direct correspondence with reality for its truth. But the purposes of mental ascriptions are different. When I m relaxing in front of the TV on a Sunday, I don t care whether my front room contains a sofa, or whether my front room contains particles arranged sofa-wise, or whether the blobject instantiates solidity sofa-shaped-region-of-myfront-room-ly. So long as it feels comfortable when I sit down, I m happy. This is why, metaphysically speaking, sentences about tables and chairs are so easily pleased. But my interest in mental properties of others is not so metaphysically neutral. When, whilst watching TV in my front room, I worry about how many are suffering in Syria right now, I have a concern that goes beyond practicalities. I am concerned with how the world is in and of itself. I am worried because I take it to be inherently bad that there are many things in the world thing in the sense of something that has properties instantiating the property of feeling pained. Interest in mentality is interest in the world. This is a crucial difference between talk of tables and chairs and talk of mental properties, which accounts for the fact that when assessing whether a given state of affairs is sufficient truthmaker for a given mental ascription, we are much less inclined to say Oh that ll do , than we are when assessing whether a given state of affairs is sufficient truthmaker for some sentence about furniture arrangement. But the ultimate reason the non-physicalist blobjectivist proposal fails is simply that that nobody who understands the sentence There is a something which feels pleasure but no pain and something which feels pain but no pleasure , upon considering an irreducible state of affairs of a single object feeling both pleasure and pain (albeit in distinct spatiotemporally local manners) would describe that world using that sentence. Rather they would describe it using the following sentence, There is one thing which feels both pleasure and pain . Conclusion I use yellow-sofa-wise as shorthand for whatever arrangements of particles would be sufficient for the existence of yellow sofas, if there were yellow sofas. This kind of phrasing originates in van Inwagan 1990. 14 Of course if blobjectivism is true there is not really a room to get around, nor a person which might sit down, but I will assume for the sake of discussion that Horgan and Potr could come up with some appropriate paraphrases of these goals. 13 I hope to have shown that a world where there is only one thing could not be truly described as a world where There are two things such that one feels pleasure and not pain, and another than feels pain but not pleasure. I am inclined to think that I can know with Cartesian certainty that the thing that feels my pain is not identical to the thing that feels Dave s pleasure; even the evil computers couldn t be deceiving me into thinking I m not feeling the pleasure as of eating ice cream right now when in fact I am. If I m right about this, then either blobjectivism is false, or Dave and his ice cream joy is a figment of my imagination. Of course, the same would go for all the other conscious states I ordinarily suppose to currently exist, but I am certain that I am not currently instantiating. Blobjectivism would entail solipsism, which would make the view very implausible indeed. Maybe you believe it yourself, but why bother telling me about it? But suppose I am wrong in thinking that I can rule out with certainty that the thing that instantiates my conscious states also instantiates Dave s and everyone else s. Still, the view of Horgan and Potr is not supposed to involve the denial of commonsense truths, and to deny truth to sentences such as S concerning the existence of distinct subject of experience would clearly be a mortal sin against common sense. I hope to have shown that Blobjectivism cannot account for the truth of such sentences, and hence cannot be reconciled with common sense. References Block, N. and R. Stalnaker, 1999. Conceptual Analysis, Dualism and the Explanatory Gap, Philosophical Review 108/1: 1-46. Goff, P. 2011. A posteriori physicalists get our phenomenal concepts wrong,' Australasian Journal of Philosophy, 89: 2,191-209. Goff, P. Forthcoming. A priori physicalism, lonely ghosts and Cartesian doubt, Consciousness and Cognition. Goff, P. MS. Do physicalists finally have a plausible theory of phenomenal concepts? Hill, C. 1997. Imaginability, Conceivability, Possibility and the Mind-Body Problem, Philosophical Studies, 87/1: 61-85. Hill, C. and B. McLaughlin 1999. There are Fewer Things in Reality than are Dreamt of in Chalmers s Philosophy, Philosophy and Phenomenological Research, 59/2: 445-54. Horgan, T. and Potr , M. 2008. Austere Realism: Contextual Semantic Meets Minimal Ontology, Cambridge, Massachusetts, London, England: MIT Press. Horgan, T. And Potr , M. This volume. Existence monism trumps priority monism. Levine, J. 1983. Materialism and Qualia: the Explanatory Gap, Pacific Philosophical Quarterly, 64/9: 354-61. Loar, B.1990 CHANGE THIS. Phenomenal States, in Philosophical Perspectives, 4: 81-108. Loar, B. 2003. Qualia, Properties, Modality, Philosophical Issues, 13/1: 113-129. Lycan, W.G. 1996. Consciousness and Experience, Cambridge, Mass: MIT Press. Papineau, D. 1993. Physicalism, Consciousness, and the Antipathetic Fallacy, Australasian Journal of Philosophy, 71/2: 169-83. Papineau, D. 2002. Thinking about Consciousness, Clarendon Press: Oxford. Perry, J. 2001. Knowledge, Possibility and Consciousness, Cambridge, Mass: MIT Press. Schroer, R.2010. Where's the Beef? Phenomenal Concepts as Both Demonstrative and Substantial, Australasian Journal of Philosophy, 88/3: 505-22. Tye, M. 1995. Ten Problems of Consciousness: A Representational Theory of the Phenomenal Mind , Cambridge, Mass: MIT Press. van Inwagan, P. Material Beings, Ithaca and London, Cornell University Press.