SWIF
Philosophy of Mind,
10 October 2000. http://www.swif.uniba.it/lei/mind/texts/allen.htm
Texts
A Space Oddity: McGinn on Consciousness and Space Sophie Allen
Colin McGinn famously contends that the explanation of consciousness is cognitively closed to us: we lack the requisite cognitive abilities to grasp the essential nature of the mind, at least at this stage of our evolution. Meanwhile, he maintains that naturalism about the mind is true, in that there is nothing ontologically spooky about consciousness; it is objectively related to the other non-mental features of the world -- space, time, matter, natural or physical properties (if such entities exist) -- in what he describes as 'a naturalistic fashion' (1). The mystery of consciousness is only a mystery from our epistemic perspective, a hypothetical God's eye-view would reveal that the mental and physical realms fit together neatly into some unitary natural whole. Armadillos can't do maths, and we can't grasp the nature of our own consciousness; that much is, McGinn claims, is an unavoidable fact about our cognitive make-up. McGinn (1991, 1993b) gives several reasons for maintaining this viewpoint, but in the course of this paper I will investigate just one of them, which arises, he claims, out of the relationship between consciousness and space (1995, 97–108). I think McGinn's argument about the relationship between consciousness and space is suspect, and the pessimistic implications which McGinn draws from his conclusion concerning the limitations on our knowledge about consciousness even more so. Then, to add insult to injury, I think that there is reason to question the very basis of his whole approach, since he may be unable to characterise what he means by 'physical' without begging the question about the mind's essential properties against his more optimistic opponents. In the course of this paper, I will concentrate my criticism upon the former of these three issues, omitting discussion of the cognitive closure of consciousness which has been extensively challenged elsewhere (2) and McGinn's characterisation of the physical. If I am correct about what I say here, McGinn does not frame a significant and coherent problem regarding the spatiality of consciousness, so the question of whether his conclusion threatens the explicability of consciousness does not arise. Whatever the strength of his other arguments for consciousness being cognitively closed, I do not think that this one should be any cause for sleepless nights among those who remain convinced that the nature and workings of the conscious mind will eventually be intelligible to us. McGinn's 'Space Problem' The core of McGinn's argument in 'Consciousness and Space' centres on the Cartesian intuition that while matter is essentially spatial, mind is essentially not. The conscious experience of a yellow flash, for instance, is 'not the kind of thing that falls under spatial predicates', unlike a neural state with which it may be associated which, as a 'complex of neural structures and events' occupies a region of the brain; it has extension, contour, shape, size, dimensionality and so on, the kind of spatial properties we associate with ordinary physical objects (3). Conscious experiences fall into the temporal order, but they do not fit into the spatial one, and so they are essentially disanalogous to the essentially spatial physical entities with which we explain the remainder of the natural world. To apply spatial predicates to mental states is to be guilty of committing a category mistake. In order to counter any intuitions conflicting with this observation, McGinn denies that an analogy can be drawn between conscious mental states and the unobservable or theoretical entities of physical theories, such as sub-atomic particles, which we think of existing in unperceived space. On the contrary, he asserts, mental states are more analogous to numbers; even God would not perceive mental entities as being spatially defined, although he would perceive electrons as spatially arrayed. Nor can the anomaly between the conscious and the spatial world of material entities be explained away as the mistake of treating mental entities as objects, rather than event-types, states or processes. According to McGinn (1995, 98), 'Descartes was not just committing the simple howler of failing to notice that conscious phenomena are not objects and hence not spatial objects'; physical processes present themselves as occurring in space, physical states are states of space, but mental states and processes are dissimilar to physical ones in this respect. To the extent that a mental state appears to be located, its location is derivative, based upon the physical entity which we take to be its cause. To strengthen this position, McGinn attempts to explain away the intuition that mental states do have location, and hence are spatial, in virtue of their being associated with one particular human body or brain. This spatial feature of mental states, he argues, is no more than 'a sort of courtesy location', since locating mental states in this way trades on 'certain causal considerations': 'Events in particular physical objects are directly causally involved in changes of mental state, and we locate the mental changes roughly where those causally proximate physical objects are.' (1995, 99) I am where my brain is; but that does not impinge upon the intrinsic non-spatiality of my conscious states, since the contingent causal connection between my brain and my consciousness is not sufficient to make spatiality a feature of the mental, rather than just a feature of whatever neural structures constitute the causal basis of my thought. The metaphysical possibility of a world where there are no psychophysical causal connections -- a world of Leibnizian pre-established harmony, perhaps -- entails that we are not entitled to infer the spatiality of the mental from the spatial characteristics of its causal basis (4). A Leibnizian world would appear to be identical with the actual one, but the mental entities within it would not admit to being associated with any spatial location whatsoever, however vague. The Implications of the Space Problem This, in outline, is McGinn's defence of the intrinsic non-spatiality of mental entities and these considerations have, he claims, radical consequences for the philosophy of mind. There are two traditional ways, McGinn thinks, to alleviate ourselves of the burden of the explanatory gap which acceptance of the space problem generates: we can be materialists and deny the intuitions about the non-spatiality of the mental; or we can revert to a form of dualism which denies that there is a connection between the mental and the physical to be explained. Both of these he deems to be unacceptable (1995, 102-3), for familiar reasons which I shall join him in omitting to mention. The third way is to accept that there is a space problem with regard to consciousness and work out the implications that this has for the philosophy of mind. Firstly, McGinn suggests, the intelligibility of mental causation becomes questionable, since causes and effects are usually construed as standing in spatial relations to each other. On McGinn's picture, mental states do not do this, once the analogy between mental entities and the unobservables of the physical realm has been rejected: 'we understand mental causation only if we deny the intuition of non-spatiality. ...The standard analogy with physical unobservables simply lulls us into a false sense of intelligibility.' (1995, 100) This is not to deny that there is such a phenomenon as mental causation, ontologically speaking; in fact, McGinn's ontological naturalism about the mind gains much of its support from the strength of our intuitions that what we think directly causes what we do. However, we cannot fully comprehend the nature of causal links between mind and matter. Secondly, the origin of consciousness becomes inherently mysterious, since received cosmological opinion asserts that not too long ago in the history of the universe conscious beings did not exist. An intrinsically non-spatial realm has emerged, at some time in the history of the natural world, from the intrinsically spatial one of matter obeying the laws of nature: exactly how did this occur? This is the 'Space Problem' which afflicts the study of the mental realm and one of the questions that McGinn thinks we cannot answer (5). It is parallel to the more commonly discussed 'Hard Problem' of consciousness, due to Chalmers, Block, Jackson and countless others, concerning the explication of the generation of irreducibly phenomenal entities or qualia by the non-phenomenal realm (6), and is, McGinn claims, at least as serious. Responses to the Space Problem I will discuss two key areas in which I think McGinn's argument for the space problem is deficient and suggest that his reasoning is insecure and his conclusion most probably incoherent. In this case, we should not take what McGinn says seriously. This is not to announce that there are no conceptual obstacles in the way of our understanding consciousness; nor even to say that there is nothing unusual about the nature of its spatiality in particular. That there is something of a conceptual gap between our current theories of matter and mind is something upon which I and Colin McGinn agree. However, because I agree that there are features of the conscious mind which are difficult to fit in to our schema of physical explanation, I think it is extremely important to clarify exactly which features these are. The Supposed Disanalogy between Mental States and Unobservable Physical Entities Firstly, McGinn's argument for the space problem does not satisfactorily deal with the responses he attempts to pre-empt which draw an analogy between the spatial features of physical unobservables and mental states, since he appears to conflate the behaviour of distinct categories of metaphysical entities such as objects and properties, in the very way in which he is concerned to counsel against (1995, 98). Thus, the disanalogy he points to between the essentially spatial unobservables of physical theory and our conscious mental states or properties is not nearly as plausible as he makes it seem. As I mentioned above, McGinn is concerned to point out that we cannot construe mental entities as being analogous to unobservable physical entities which are located in unperceived space. The latter, he argues, exist in spatial relation to entities which we do perceive, whereas mental properties and states do not, being more similar to numbers in this respect. By any account of the nature of abstract objects, this second comparison strikes me as extremely odd: in what way is my occurrent belief that it is raining comparable to the number three? On one hand, we do seem to have 'no conception of what it would even be to perceive' the number three as a spatial entity, when it is considered platonistically as an abstract object, but McGinn is concerned to point out that the view he is presenting does not mistake occurrent conscious episodes for objects. On the other hand, the number three is instantiated all over the place (by the 't's in 'instantiated', for example) and, presumably, these instances of threeness are located. I do not find the analogy between mental states and numbers convincing -- if this is not merely a careless, throwaway remark about numbers, I must concede that McGinn's mental life is significantly different from mine -- but, even if it were, the case for saying that numerical properties are never located seems rather slim. It does not seem that mental states are abstract in the sense in which numbers are, rather in another sense which (confusingly) crops up in the metaphysical literature to describe the fine-grained ontology of particular entities which are incapable of independent existence such as property-instances or tropes. A particular instance of red, for example, cannot exist in the absence of the instantiation of other properties at the same location; but it has location nevertheless, unlike the number three on McGinn's conception of it. However, it should be noted that the spatial features of physical property-instances or tropes (which could be called 'abstract particulars', were it not going to confuse the issue) are themselves dissimilar to those of physical objects, whether observable or unobservable. By definition, many property-instances may be instantiated in the same spatio-temporal region: it makes no sense to ask where the mass of a particular stone is located in relation to the stone's other properties, whether it is to the right, or the left of its density and chemical structure for instance. The particular stone, of course, bears spatial relations to other entities; but the collection of properties it manifests do not (in general (7)) bear spatial relations to each other. They can be described as 'non-spatio-temporal parts' of the particular, and are definitely situated within space, but their respective locations are severally dependent upon the location of the particular in which they are instantiated. Moreover, when it comes to locating the physical properties instantiated by the unobservable physical constituents of the stone, rather than the unobservable objects themselves, it appears that we must trade upon the self-same causal considerations which McGinn suggests make the rough location of our mental states in the vicinity of our brain problematic. We say that the unobservable micro-structure of the stone, for example, has the physical micro-properties that it does in virtue of the stone's causal potential; that is, what those properties can cause (8). Just as in the mental case, this contingent causal story is consistent with a variety of metaphysical possibilities about the way the world is ontologically speaking, not all of which would have the micro-physical properties in causal contact with any in the observable realm. The quantum level, for instance, may merely be in pre-established harmony with chemical changes, despite the empirical story of quantum chemistry to the contrary. McGinn may be correct in thinking that the analogy with unobservable objects is inadequate (9), but it seems plausible to maintain that the location of quark colour (say) is no different from mental properties when these are roughly located in the brain, in its being accorded 'a sort of courtesy location' in virtue of its causal relations and the location of the objects which instantiate it. McGinn does not do enough to preclude an analogy being drawn between conscious mental properties and the physical properties of unobservable physical entities. The (Probable) Incoherence of Time without Space. Secondly, the argument for the essential non-spatiality of the mental is threatened by brute incoherence: it is extremely plausible to maintain that there is no notion of entities being temporally ordered without being spatially located also, but McGinn maintains that this is true of conscious states. Of course the nature of the space in which some such entities are located may outrun the three-dimensional folk theory of space required for the successful functioning of our visuo-motor systems, but that is not so unusual in the post-Einsteinian conception of space (10). In opposition to McGinn, it might be argued that the notion of temporality, or of entities falling into some temporal order, requires the notion of simultaneity; that is, what it is for an event to occur, or a property be instantiated at the same time or at different times. However, to make sense of this notion, the relative states of motion of the entities involved is important and, since motion is a partially spatial notion, simultaneity and temporality must be also. Assuming Einstein's special relativity to be somewhere near correct, any two events which are temporally separated with respect to one frame of reference, are spatially separated with respect to another frame. To give a space age example of why this might be the case, imagine two people with synchronised and extremely accurate watches who agree that neither of them holds a certain belief (call it P, say). They are each given an envelope, which contains incontrovertible evidence for P, and told to open the envelope when their respective watches say 9:09 (on a certain date: 9/9/99, say). In such a case, it would seem reasonable to say that both parties gain the belief that P when their watches say that it is 9:09 on 9/9/99. However, whether the two people gained the belief that P simultaneously is dependent on their states of motion relative to one another: if one is in a spaceship moving at extremely high velocity (approaching c, the velocity of light) and the other remains on earth, then the earthbound person's watch will show 9:09, 9/9/99, before that of the space traveller; so, according to our earthbound frame of reference the earthbound person will gain the belief that P first. Motion matters to temporal ordering, and we cannot make sense of motion without space and location: if something has a when, it also has where (11). If this objection against McGinn holds, then his characterisation of mental entities as being temporal and not spatial is ultimately incoherent. However, the force of the objection is not as clear as this, since it relies on Einstein's special theory of relativity which, however widely accepted and plausible, is an empirical claim about the nature of space and time. In response, it remains open to McGinn to provide an account of temporality which is independent of spatiality, but this would involve rejecting the Einsteinian paradigm which occupies such a fruitful and fundamental place in modern physics. It is not inconceivable that this could be achieved, but it is certainly something of a tall order. Moreover, since McGinn's materialist opponents have a reasonably unproblematic story explaining why conscious states do have spatial location, despite appearances to the contrary, the provision of an independent account of temporality becomes a matter of some urgency. The burden of proof has been shifted onto McGinn to explain how mental entities can be temporally ordered without being spatially located also. Furthermore, McGinn weakens his own position on this issue by upholding a form of ontological naturalism about the mind: it is still true that the non-spatial realm somehow emerged from the spatial, there is some objective connection between the two. However, in order to maintain this position, he is forced to concede that mental entities must be, in some sense, spatial; 'they bear an opaque and anomalous relation to space as currently conceived' but 'presumably... this is merely an epistemological fact rather than an ontological one' (1995, 105). The real problem, it seems, is with our conception of space, and not with the non-spatiality of conscious states at all. Although mental properties may present themselves as lacking spatiality, that is not how they ontologically can be, although they may be like some physical entities in being hyper-spatial; that is, (as I will stipulatively define it if necessary) having spatial features which do not fit into our folk, three-dimensional spatial scheme. Like quanta and superstrings, we might have a problem explicating the spatial features of mental entities in terms of the Euclidean space of ordinary middle-sized material objects, but they are not non-spatial in virtue of this. The Final Frontier I have pointed to two difficulties with McGinn's argument, the analogy between conscious properties and unobservable physical properties is still up for grabs, and by insisting on the temporality of consciousness without its spatiality, the position he is defending is threatened with incoherence due to the interdependence of space and time. If either or both of these objections are granted, it appears that McGinn does not succeed in setting up the space problem in the first place. In addition to this, in defending his ontological naturalism, McGinn is forced to admit that consciousness has a place within some conception of space unlike our that of our folk theory and thus that mental states are hyper-spatial, rather than devoid of spatial features entirely. In this sense mental entities are not dissimilar to entities within some fundamental physical theories which are presumed to exist within a counterintuitive conception of space. I suggest that McGinn accords too much weight to the Cartesian intuitions from which he generates the space problem, especially since, as an ontological naturalist, he must regard them as literally false. In fact, he coincidentally provides two plausible reasons to disregard these intuitions, by noting that their acceptance leaves both mental causation and the origin of consciousness shrouded in mystery. Rejecting the intuitions about the non-spatiality of consciousness would go some way to relieving the conceptual difficulties associated with explaining the origins and causal efficacy of conscious mental states. I propose that it is time to do just that. This is not to say that the explanation of consciousness will be easy, it is still likely that some paradigm shift will be required for the explanation of consciousness, and perhaps some of these changes will specifically concern our conception of space. However, this conclusion is not nearly so pessimistic as that of McGinn, which I have gone to some lengths to refute (12). ______________________________________ Notes: 1. 1995. Reprinted in Shear (ed.) 1997, 105. All page references to McGinn (1995) will be to Shear (ed.) (1997). 2. See, for example, Garvey (1997) and Hanson (1993). 3. 1995, 97. Following McGinn, I shall call these intuitions Cartesian; although it seems that McGinn takes the 'Cartesian' view rather further than Descartes himself. Descartes maintained the non-spatiality of pure thought, but allowed that both perception and imagination require the brain. Thus, Descartes would not have acceded to the example of the non-spatiality of a yellow flash given by McGinn, since he recognised that colour involves extension and shape which are both spatial concepts. (See Cottingham, Stoothoff, Murdoch, eds.: vol. 2, 265.) 4. McGinn also rejects the view that some mental states, such as pains, do present themselves to us as being located (1995, 99). Insofar as we do treat pains as located, we are treating them as bodily, rather than mental, states. 5. McGinn makes his own speculations about the origins of consciousness, but he relies on a controversial cosmological story which I will not discuss (1995, Section II). 6. See, for example, Shear (ed.) (1997, passim); Jackson (1982); McGinn (1991). 7. There might be exceptions here: depending on the account of colours, these properties may be located only on the surface of the stone but the other properties which the stone has, qua concrete particular, are not in spatial relation to each other. 8. This conception of physical or natural properties as essentially causal is explicitly proposed by Shoemaker (1980) but it also finds implicit acceptance in the work of Kim (1993), Armstrong (1978) and Lewis (1983, 213 - 8). 9. McGinn may not correct in thinking this, however. The analogy with unobservable physical objects may hold, if quanta are treated as a species of unobservable physical objects, since there are problems associated with their spatial properties also. 10. For example, superstring theory requires that the entities it postulates exist in ten dimensions, rather than the four commonly recognised in physical theory, such that each point of space-time is treated as a six-dimensional 'hypersphere'. It is suggested that the properties of these hyperspheres determine the behaviour of the entities of the subatomic domain. 11. For more discussion on this issue, see Lockwood (passim), Gibbins (1985), Gordon (1984). 12. I would like to thank Naomi Eilan, Olga Markova, Greg McCulloch, Nenad Miscevic, David Spurrett, James Tartaglia and Jonathan Webber for helpful comments about the arguments in this paper. ______________________________________ References: Armstrong, D M. Chalmers, David. Cottingham, J., Stoothoff, R., Murdoch, D., eds. Garvey, James. Gibbins, P F. Gordon, David. Hanson, P. Jackson, Frank. Kim, Jaegwon. Lockwood, Michael. 1984. Reply to Gordon. Analysis 44: 127-8. 1985. Einstein, Gibbins and the Unity of Time. Analysis 45: 148-50. McGinn, Colin. Shear, Jonathan, ed. Shoemaker, S. Spurrett, David. Whens without Wheres? (In preparation.) Van Inwagen, P, ed.
Department of Philosophy
University College London
sophie.allen@ucl.ac.uk
1978. Universals and Scientific Realism. Cambridge: Cambridge University Press.
1995. Facing up to the Problem of Consciousness. Journal of Consciousness Studies 2: 200 - 219.
1985. The Philosophical Writings of Descartes. Cambridge: Cambridge University Press.
1997. What does McGinn think we cannot know? Analysis 57: 196-201.
1985. Are Mental Events in Space-time? Analysis 45: 145-7.
1984. Special Relativity and the Location of Mental Events. Analysis 44: 126-7.
1993. McGinn's Cognitive Closure. Dialogue 32: 579-85.
1982. Epiphenomenal Qualia. Philosophical Quarterly: 32: 127-136.
1993. Supervenience and Mind. Cambridge: Cambridge University Press.
1984. Einstein & Identity Theory. Analysis 44: 22-5.
1995. Consciousness and Space. Journal of Consciousness Studies 2: 220-30. Reprinted in Shear (1997).
1997. Explaining Consciousness. Cambridge, Mass: MIT Press.1
1980. Causality and Properties. In Van Inwagen (1980): 109-35.
1980. Time and Cause. Dordrecht: D. Reidel.
© 2000 Sophie Allen