Churchland and Adams, et al. at an Impasse: A Way Forward?

Churchland and Adams, et al. at an Impasse: A Way Forward? Patricia Churchland has established a reputation for her staunchly reductionist theory of consciousness. But unlike other notable physicalists such as Jaegwon Kim, 1 Churchland has remained unmoved by the surge of arguments for the irreducibility of qualia. 2 In this paper, I outline one such argument, what I ll call due to how it is commonly stated by Robert Adams, et al. the why-this-rather-than-that argument. Following the outline of the argument, I ll present and analyze Patricia Churchland s objection to it, which seems to leave Churchland and defenders of the argument at an impasse. I suggest, however, that the distinction between necessary and contingent identity statements provides a clear way forward. I. The Why-This-Rather-Than-That Argument Compared to other arguments for the irreducibility of qualia, such as the what-it-is-like argument, the why-this-rather-than-that argument has received scant attention. But it is no less an argument for that, and it has able defenders in Robert Adams, Richard Swinburne, David Chalmers, and Jaegwon Kim. Perhaps the clearest statement 3 can be found in Robert Adams interesting paper, Flavors, Colors, and God : For suppose that the experience of seeing red is caused by brain state R, and the experience of seeing yellow by brain state Y (both R and Y being patterns of electrical activity).... Why does R cause me to see red? Why doesn t it cause me to see yellow or smell a foul odor? We do not imagine that R is itself red, or Y yellow. It is hard to conceive of any reason why a particular pattern of electrical activity would be naturally connected with the peculiar kind of experience that I call the appearance of red, rather than with that which I call the appearance of yellow. Indeed, it is hard to conceive of any reason why a pattern of electrical activity would be naturally connected with either of these appearances, rather than with no phenomenal qualia at all. Let us be clear that I am not denying that R and Y are in fact constantly correlated with the experience of red and yellow respectively. I am also not denying that R and Y cause me to experience red and yellow, respectively. What I want to know is why these relationships between brain states and phenomenal qualia obtain rather than others and indeed why any such regular and constant relationships between things of these two types obtain at all. 4 The puzzle here is why some particular brain state causes some particular quale rather than some another, or why certain qualia are connected to certain brain states. Whatever the answer to this 1 Jaegwon Kim, Physicalism, Or Something Near Enough (Princeton, 2007). 2 Attributed largely to the influence of David Chalmers The Conscious Mind (Oxford, 1997). 3 Though Richard Swinburne s is very similar. See his The Evolution of the Soul (Oxford, 1986. Rev. ed. 1997), pp. 186ff. 4 Ch. 16 of Robert Adams, The Virtue of Faith and Other Essays in Philosophical Theology (Oxford, 1987), pp. 244-245. The passage quoted (see also those quoted below) should make clear why I have chosen to refer to the argument in the unwieldy fashion I have. It also nicely compliments the what-it-is-like argument (a.k.a. the knowledge argument). The why-this-rather-than-that argument is sometimes called the explanatory gap argument. See Joseph Levine, Purple Haze: The Puzzle of Consciousness (Oxford, 2001), pp. 76ff. Dimitris Platchias, Phenomenal Consciousness: Understanding the Relation Between Experience and Neural Processes in the Brain (McGill-Queen s, 2011), pp. 127-136; I have avoided referring to the argument as the explanatory gap argument because some authors refer to knowledge arguments generally as explanatory gap arguments. 1

puzzle is, call it the correlation explanation. Consider next David Chalmers statement of the problem. Given any account of the physical processes purported to underlie consciousness, says Chalmers, there will always be a further question: Why are these processes purported accompanied by conscious experience? 5 No doubt because of his unyielding fondness for his zombie twin, Chalmers illustrates the problem accordingly: where Chalmers zombie is a physical duplicate of him but with inverted spectra, Chalmers argues that any story about physical processes applies equally to my inverted twin, who sees blue where I see red: it follows that nothing in that story says why my experience is of one variety rather than another. 6 Chalmers presentation of the problem is a little short. 7 As a final statement of the argument, then, consider Jaegwon Kim s. Say some brain states N and M cause some phenomenal qualia i and t (say, having an itchy sensation and having a tickling sensation, respectively), or that i and t supervene 8 on their respective brain states N and M: (S) N i (S*) M t Kim explains the problem: If we had supervenience laws like (S) and (S*) as explanatory premises, we could of course explain why a particular instance of itchy sensation or ticklish sensation occurred in a person: She was itchy because she was in neural state N, and she was ticklish because she was in state M. But precisely what is in need of explanation is why supervenience relations like (S) and (S*) hold in the first place. 9 Kim continues, asking the characteristic why-this-rather-than-that question: Why do we experience itches just when we are in neural state N? Why don t we experience tickles instead? What is it about states N and M that account for the fact that itches emerge from M, and not the other way around? 10 The argument could be outlined as follows: (1) If reductive physicalism is true, then i is reducible to N. (2) If i is reducible to N then N will explain via only natural laws, physical facts, etc. why N is correlated with i rather than, say, t. (3) But N cannot explain via only natural laws, physical facts, etc. why N is correlated with i rather than, say, t. 5 Chalmers, The Conscious Mind, p. 106. Several defenders of the argument note this open question feature of it. Cf. Adams: For as long as the distinction [between phenomenal properties and physical properties of the brain] remains, we can still ask why brains that have those physical qualities also have these phenomenal qualia. Flavors, Colors, and God, p. 259; 255, 257. 6 Chalmers, The Conscious Mind, p. 107. 7 Churchland cites only Chalmers as a voice for this argument. 8 Where the supervenient relation is as weak as you like; e.g., nomological supervenience rather than metaphysical. 9 Jaegwon Kim, Philosophy of Mind (Westview, 1998), p. 174. Kim s emphasis. See also Physicalism, Or Something Near Enough, pp. 168-170. 10 Kim, Philosophy of Mind, p. 174. 2

(4) Therefore, i is not reducible to N. (2, 3 MT) (5) Therefore, reductive physicalism is not true. (4, 1 MT). Assuming canonical reductive physicalism, premise (1) is definitionally true. This seems equally true of (2). But consider (2) for a moment. Examples of correlative reduction in the sciences (e.g., optics to electromagnetic radiation, genes to DNA molecules, etc. 11 ) suggest that a successful reduction of one phenomenon r to another (say, R) depends crucially on explaining, in terms of more general laws, how R causes r. For example, to say temperature is reduced to molecular energy is to say, given the molecules and their motion governed by physical laws, there is an explanation for why the molecules produce or cause a certain temperature. With such an explanation at hand, we can then say whatever property it is that meets these causal powers is temperature, 12 thereby reducing temperature talk to talk about molecular motion and physical laws. So if there is to be a successful reduction of i to N, then we must expect an explanation [for why N causes i] in terms of these laws and the properties of the material stuff the laws govern. 13 Because (4) and (5) are conclusions, that leaves (3) as the argument s main premise, and the defenders of the argument spend most of their intellectual acumen to defending it. II. Churchland on the Why-This-Rather-Than-That Argument Patricia Churchland, a canonical reductive physicalist if there ever was one, rejects the whythis-rather-than-that argument. In this section, I will consider Churchland s objection to the argument as found in her paper A Neurophilosophical Slant on Consciousness Research, 14 and why I think this response lands Churchland and Adams, et al. in a deadlock. 2.1. Churchland s Objection As Churchland understands them, defenders of the why-this-rather-than-that argument maintain that even if neuroscience were to discover with what brain states being aware of a burning pain on one's left ear is identical, we would still not understand why just those brain states are identical with precisely that sensation, as opposed, say, to feeling a desire to void. Neuroscience, it will be averred, will never be able to explain why conscious states Y = brain states X, rather than say, brain state Z. 15 Notice that Churchland interprets the puzzle as being why certain phenomenal states are identical to certain brain states. In so doing she reveals that she doesn t even entertain the possibility that a correlation explanation could be being given weaker terms. As Churchland herself observes, X can be correlated with Y for a range of reasons: X causes Y, Y causes X, 11 Actually, there is some controversy as to whether these and other canonical cases of reduction are, in fact, successful. See Steven Horst, Beyond Reduction: Philosophy of Mind and Post-Reductionist Philosophy of Science (Oxford, 2007). That they are is unquestioned dogma exclusive to philosophy of mind, according to Horst. 12 Kim, Philosophy of Mind, p. 175. Kim s emphasis. 13 Ibid., p. 174. 14 In Progress in Brain Research 149 (2005), pp. 285-293. Churchland only cites Nagel and Chalmers as representatives of these arguments. 15 Ibid., p. 290. 3

they have a common cause, or X and Y are actually the same thing under different descriptions; i.e., X = Y. 16 But nothing short of identity will do, for [d]iscovery of identities (i.e., that X = Y) is typically needed to make the crucial step towards theoretical authority. Any correlation explanation given in weaker terms, then, does not have the requisite authority for theory-hood. The word correlation in the above outline of the why-this-rather-than-that argument, then, must be understood to mean identity. So understood, Chruchland thinks the argument rests on a misunderstanding. The misunderstanding is that identities ought to follow general laws i.e., that if S = Y, then we ought to know why S = Y in the sense that there should be an explanation only in terms of natural laws, physical facts, etc. for why S = Y. But, Churchland protests, by and large science does not offer explanations for fundamental identities. 17 Rather, science discovers fundamental entities, but the identities it discovers just are the way things are. 18 Churchland, then, rejects premise (2) of the argument. If i is reducible to N, it is not true that N will explain why N is correlated (read: identical) with i rather than, say, t. If phenomenal states are identical to brain states, then there is no puzzle about why some brain state N is not identical to some other phenomenal state t, because identities are just brute facts. The question is tantamount to asking why is Venus, Venus? So, the problem disappears, and does so without even a spitball being returned in response to Adams, et al. defense of premise (3). 2.2. An Impasse? General problems might be raised for Churchland s insistence that phenomenological states are identical to brain states. Swinburne, Chalmers, and Kim all think that saying phenomenological states are identical to brain states commits a category mistake of sorts, perhaps on par with saying the number 2 smells pink. But let s grant Churchland the intelligibility of the identity claim. In fact, below we ll indulge the fiction that Adams, et al. agree with Churchland that the correlation is identity. Defenders of the why-this-rather-than-that argument are likely to see Churchland s response as concessionary: to say the problem (in scare quotes because, according to Churchland, there really is no problem) admits of no explanation and so is simply a matter of brute fact is to concede the point. If an answer to the correlation explanation cannot be non-reductive, nor reductive way described above, then Adams et al. agree with Churchland that the only alternative will be to leave it [the correlation] a brute, inexplicable fact (which seems pretty implausible). 19 Chalmers: the accompaniment [i.e., correlation] must be taken as brute (Chalmers); 20 Kim: It seems that the N-itch, M-tickle, and other correlations must be accepted as brute facts not subject to further explanation. 21 Unfortunately, with the possible exception of Adams, defenders of the argument don t say why inexplicability and bruteness in this case are problematic. So, Churchland is happy with correlation explanation being brute and inexplicable, and Adams, et al. aren t. The road of explanation has to stop at some point, right? It just stops much sooner for Churchland than for Adams, et al. 16 Ibid., p. 289. 17 Ibid., p. 290. 18 Ibid. Notice the appeal to the agental powers of science. It is science, not scientists, who discovers things. Scientists are just the unwitting pawns through which the real causal buck, Science, works its miracles. 19 Adams, Flavors, Colors, and God, p. 252. 20 Chalmers, The Conscious Mind, p. 107. 21 Kim, Philosophy of Mind, p.176. 4

Perhaps Adams finds the claim that the correlation is identity, and hence the correlation explanation is brute and inexplicable, implausible for the following reason. Adams insists that even supposing N and i are identical, we can still ask why N (the physical brain state that is the itchiness), rather than M, seems to us the way it does. As long as there is a way in N seems to us, the question of why that is so can be asked. But how does this help? Churchland will not likely be impressed. She ll just say Adams is asking a meaningless question. How does one go forward? Is this just a dreaded intuition stalemate? Below I will suggest that by distinguishing necessary from contingent identity relations, the argument can move forward. III. A Way Forward? As Dimitris Platchias observes, what is being sought by proponents of the why-this-ratherthan-that argument is a contrastive explanation, which he calls the most influential notion of explanation in the philosophy of science. 22 Platchias broadly characterizes a contrastive explanation of some fact P as an explanation of P relative to a class of contrasting alternatives. It explains P rather than relevant alternatives to P, that is, why P rather than Q or R or S or T? 23 Interestingly, Platchias diagnoses the need for explanation when it is conceivable that things could have been different than they are. 24 For example, why is grass green rather than, say, blue, or why do women tend to live longer than men (rather than the other way around), or why does water boils at 212 F rather than, say, 100 F at sea level? As the facts come in, alternative explanations are ruled out. Grass s molecular composition is dominated by chlorophyll. Chlorophyll absorbs the red and orange wavelengths of light, leaving it to reflect green, not, say, blue. But it seems clear that things could have been different: had grass contained indigo rather than chlorophyll, it would have reflected blue. This helps clarify a reductive explanation will not, in general, be a contrastive explanation. This is because a reductive explanation amounts to an identity statement, and contrastive explanations tend not to be sought for identities. When it is asked why grass is green, the is is to be understood not as the is of identity, but of predication. That is, we are asking why green is predicated of grass, not why is green identical to grass. It is because green is predicated of grass that we can conceive of things being different than they are. But what of identity statements? Can we conceive of, say, Venus being different than Venus? It seems straightforwardly necessary that Venus is Venus, and that it could not have been otherwise. In short, identity statements seem to be necessarily true. I say a reductive explanation will not, in general, be a contrastive explanation. A reductive explanation will not be a contrastive explanation in cases where the explanandum is a necessary identity statement. However, it is arguable that not all identity statements are necessary. If there are contingent identity statements, then it s not at all clear that one can t offer a contrastive explanation of it. To motivate the thought, consider first some putative cases of contingent identity. We have the statue David and the lump of bronze that composes David (call it lumpy). 25 Plausibly, David and lumpy are identical, but they might not have been that same lump of bronze could have composed a different statue. So if David and lumpy are identical, their identity is a contingent 22 Dimitris Platchias, Phenomenal Consciousness, p. 128. 23 Ibid., p. 129. 24 Ibid. 25 To slightly modify Allan Gibbard s famous example of the statue Goliath and the lump of clay, Lumpl, of which it is composed. See his Contingent Identity, Journal of Philosophical Logic 4 (1975), pp. 187-221. 5

fact. Or consider Lewis example of a network of railway tracks that could have included an additional track. Lewis thinks that the current network of tracks is contingently identical to the network of tracks that includes the additional track. 26 It has also been argued that the relationship between dispositions and their causal bases is contingently identical. 27 Indeed, the same has been argued to hold between sensations (e.g., pain states) and brain states. 28 Necessary facts do not obviously cry out for explanation in general, much less contrastive explanations (compare: why does 1+1 = 2 rather than 3? Why is grass green rather than blue?). But contingent facts do cry out for explanation. If there are contingent identities, then, we should expect them to cry out for explanation. So, we might ask, if David is contingently identical to lumpy, does it make sense to ask why is David identical to lumpy? And that seems like an intelligible question one that can be intelligibly asked in several different contexts, even. For example, suppose Michelangelo had several different of lumps of bronze or other kinds of material from which to choose in creating David. Why is David lumpy, rather than, say, clumpy? Alternatively, why is lumpy David and not, say St. John the Baptist? These are genuine questions that, presumably, have legitimate answers. So the crucial question is what kind of identity relation holds between phenomenal states and brain states. Is it necessary, like that between Venus and Venus, or is it contingent, like that between David and lumpy? 29 We already know what Chruchland believes. So maybe a better question is: why think the identity relation between brain states and phenomenal states is necessary rather than contingent? Churchland could attempt to void the question by denying that there are contingent identities; i.e., deny that David and lumpy are identical. But to deny that is to accept that purely material entities, like statues and lumps of clay, of admittedly identical material constitution at all times, may nonetheless be distinct, though distinguished only by modal, dispositional or counterfactual properties 30 which of course is far too welcoming to well-known modal arguments for dualism. 31 So this is not a promising route for Churchland. It is true that most metaphysicians reject contingent identity. That identity relations are thought to be necessary has been established philosophical dogma at least since Naming and Necessity. There are powerful dissenting voices, however, and I think it is be safe to say that contingent identity will become a popular topic in the near future, and that the strength of Kripkean orthodoxy will subside. 32 But this is not a paper defending contingent identity. My only point is that even if Churchland is granted her claim that the relationship between mental states and brain states is identity, that claim is not enough to render the why-this-rather-than-that question inappropriate. It is inappropriate only if the identity relationship is necessary. But that is something Churchland hasn t argued for, much less against. Surely, Churchland will think a contingently identical relationship between mental states and brain states is too week, and 26 David Lewis, On the Plurality of Worlds (Blackwell, 1986), p. 248ff. 27 David Armstrong, Dispositions: A Debate (Routeledge, 1996). 28 J. J. Smart, Sensations and Brain Processes, The Philosophical Review 68 (1959), pp. 141-156. 29 Further complication: in which modal domain is necessary best understood (logical, metaphysical, nomological)? Orthodoxy treats the identity relation as logically necessary, so I m assuming narrowly logical necessity is at play here. But I suppose it need not be, assuming contingent identity is a live option. 30 See Harold Noonan, Identity, The Stanford Encyclopedia of Philosophy (Winter 2011 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/win2011/entries/identity/>. 31 Or at least the non-identity of body and mind. See Stephen Yablo, The Real Distinction Between Mind and Body, Canadian Journal of Philosophy; Supp. Vol. 16 (1990), pp.149-201 32 See Wolfgang Schwarz, Contingent Identity, Philosophy Compass (forthcoming 2012). 6

Adams, et al. will think it s too strong. But it seems to me that a sensible moderating voice in this dispute will advise Adams, et al. to grant for the sake of argument Churchland as much as possible and proceed from there. Contingent identity is the way to do that. Failing to distinguish necessary and contingent identities has left Churchland and Adams, et al. at loggerheads with each other. The accuracy of this diagnosis of a clear way forward is evidenced by the fact that it also explains why Adams, et al. think that treating the correlation explanation as brute and inexplicable is implausible: contingent facts require explanation, whereas necessary facts (ordinarily) do not. 33 Both Churchland and Adams, et al. would likely agree with Alexander Pruss that claiming something to be a brute fact should be a last resort. It would undercut the practice of science were things claimed to be brute facts where not implausible putative explanations, propositions that would be explanations were they true, can be formulated. 34 The difference between Churchland and Adams, et al. is that the latter believe plausible alternative explanations can be formulated (i.e., non-reductive ones). Furthermore, believing the correlation explanation to be a contingent fact, treating it as brute and inexplicable rightly smacks Adams, et al. as what Pruss has elsewhere called the taxi-cab fallacy : arbitrarily dismissing a request for explanation (like a cab) once one wants to go no further. 35 Understanding it to be a necessary fact, Churchland is bothered by none of this. Conclusion If this is an accurate diagnosis, then this seems to break the tie by putting the ball in Churchland s court. This is because Adams, et al. already have at their disposal arguments purporting to show that the identity relation (if it is an identity relation) must be contingent: e.g., inverted spectra, 36 bodily interchange, 37 multiple-realizability, 38 and Kim s pain box. 39 So, in addition to giving reasons for thinking brain states are necessarily identical to phenomenal states, Churchland must also rebut or undercut these arguments to the contrary. REFERENCES Adams, Robert. Flavord, Colors, and God, in The Virtue of Faith and Other Essays in Philosophical Theology (Oxford, 1987), pp. 244-245. Chalmers, David. The Conscious Mind: In Search of a Fundamental Theory (Oxford, 1997). Churchland, Patricia. A Neurophilosophical Slant on Consciousness Research, Progress in Brain Research 149 (2005), pp. 285-293. Gale, Richard and Pruss, Alexander. A New Cosmological Argument, Religious Studies 35 (1999), p. 461-476. 33 See Alexander Pruss, The Principle of Sufficient Reason (Cambridge, 2010). 34 Ibid., p. 255. 35 See Richard Gale and Alexander Pruss, A New Cosmological Argument, Religious Studies 35 (1999), p. 470. 36 Chalmers, The Conscious Mind, pp. 99-101. 37 David Lewis, Counterparts of Persons and their Bodies, in Philosophical Papers: Vol. 1 (Oxford, 1983), pp. 47-54. 38 E.g., Hilary Putnam, The Nature of Mental States in Capitan and Merrill (eds.), Art, Mind, and Religion (Pittsburgh, 1967), pp. 37-48. Jerry Fodor, Special Sciences: Or the Disunity of Science as a Working Hypothesis, Synthese 28 (1974), pp. 97-115. 39 See Physicalism, Or Something Near Enough, pp. 168-169. 7

Kim, Jaegwon. Philosophy of Mind (Westview, 1998). Lewis, David. On the Plurality of Worlds (Blackwell, 1986). Noonan, Harold. Identity, The Stanford Encyclopedia of Philosophy (Winter 2011 Edition), Edward N. Zalta (ed.). Platchias, Dimitris. Phenomenal Consciousness: Understanding the Relation Between Experience and Neural Processes in the Brain (McGill-Queen s, 2011). Pruss, Alexander. The Principle of Sufficient Reason: A Reassessment (Cambridge, 2010). 8