Space and Sensibilia

Michael Bevan 19 Space and Sensibilia Michael Bevan T here is a reasonably common view, amongst proponents of causal-realist theories of perception, which says that objects of immediate perception inhabit a spatial realm that is totally cut-off from the one inhabited by material beings.I The reasons for this are two: (i) Visual percepts are often described by causal realists as ‘colour patches,’ implying that they have shapes and sizes, and are extended. Also, non -visual percepts seem to be positioned; sounds can be to the left or to the right, for example. Since any being with extension must have a location, it follows that the objects of our immediate perception are spatial entities. (ii) Since the causal realist posits that the objects of our direct perception to be non-material, it would be odd if percepts enjoyed some spatial relation to material things. Is the brown colour-patch caused by the table in the same location as the table? If so, then why does one directly perceive the former but not the latter? Such percepts cannot all be in the locations of their material causes since some (i.e. hallucinations) have no material causes. But if my table-shaped sense-datum is not at the table itself, then where is it? Is it swimming behind the cornea, or perhaps floating about the brain, haunting the synapses? How big is the colour patch (in meters squared/cubed?) All of these questions, which appear non-sensical, would be Michael Bevan is a undergrduate at the University of Reading, UK, where he is currently in his final year studying for a BA in philosophy. 20 Sense and Sensibilia legitimate if we were to admit that immaterial percepts inhabit the same space as material beings. These considerations say that if causal realism is true, there is a ‘phenomenal space’ inhabited by percepts, one which is distinct from the space inhabited by material beings. There are dissenters, of course: not very long ago, O’Shaughnessy took sense-data to inhabit body-relative physical space, that is, the space inhabited by material beings.II There are others however, Smythies being to my knowledge the most recent example,III who explicitly postulate a second space, and it is against the latter sort of causal realist that I shall aim my objections. In what follows, we shall be looking at the consequences of the ‘two-space’ view, and arguing that such consequences are absurd. 1. Space Oddities 1.1. Preliminary No matter one’s view of the ontology of ‘spaces,’ one can say that, where x and y are spatially located entities and where xDy reads as ‘x is some distance from y,’ x and y inhabit the same space if and only if xDy.IV To define the class of inhabitants of a space A then, one need only identify some inhabitant x A of A , and then define the class of A ’s inhabitants as I(A ) =df {y|yDxA}. D is an equivalence relation, meaning that I(A) is identical to the equivalence class [a]D where a ∈ I(A ). Thus, for any spaces A and B, either I(A) and I(B) are disjoint, or I(A ) = I(B). Since if there were some z in both I(A ) and I(B), it would follow that I(A ) = [z] D = I(B). Therefore, spaces are either completely disjoint, sharing no inhabitants, or completely overlapping, sharing all inhabitants. There is one other thing which we should prove before moving on. Call any part of x which possesses a location a spatial part of x. What we shall prove is that it is impossible for a being to only partially inhabit some space, by which we mean that for any being x and space A , either every spatial part of x inhabits A , or no spatial part of x inhabits A . This follows from the fact that a being must be located at least partially wherever any of its spatial parts are, and so will bear the symmetric and transitive D relation to each. For example, if x has spatial parts a and b with a inhabit- Michael Bevan 21 ing A , then we have xDa, so aDx by symmetry, and xDb, from which aDb follows by transitivity, so that b also inhabits A . 1.2. Causation If material beings are not spatially related to percepts, then the space inhabited by material beings (’material space,’ or ‘M’ for short) is totally disjoint from the space inhabited by percepts (‘phenomenal space,’ or ‘P’ for short). But a causal realist must say that the inhabitants of P and M enjoy some sort of causal relation. But what kind of causal relation could they possibly enjoy? Smythies suggests that the causal relations enjoyed by the inhabitants of P and M are ‘Humean,’ where a Humean causal relation is one of regularity: the X s are Humean causes of the Y s only if there is a constant conjunction of Y s following X s.V For example, Smythies might say that the table is a Humean cause of the brown colour-patch since the presence of the table in M is always followed by the presence of an appropriately shaped brown colourpatch in P. But to say that y follows x suggests that the occurrence of y comes after the occurrence of x. So if the inhabitants of M and P are related by Humean causal relations, then they must be temporally related. The trouble with saying that inhabitants of P and M are temporally and not spatially related is that this implies that spatiality and temporality in the material realm are separable in a way that conflicts with the broadly accepted scientific view on the matter which says that the two are bound together quite inextricably. The physicist will say that it is, strictly speaking, wrong to think of material things as entering into spatial and/or temporal relations, rather they enter into spatiotemporal relations. Further, in order for two entities to be a defined spatiotemporal ‘distance’ from each other, they must be a defined spatial distance and a defined temporal distance from each other, since the total spatiotemporal distance will be a function of these values. If the twospace causal realist wants to take current science seriously, they will be pressured to say that the events of P and M cannot temporally relate. Material beings and sense-data, therefore, must inhabit disjoint spacetimes. This means that Smythies suggestion of Humean causation is something of a non-starter. 22 Sense and Sensibilia As a possible answer, Smythies might suggest something like the following: while it may be true that events in P and M share no temporal relations, the order of events in P matches up with the order of events in M. Thus, the sequence of events in M: A window is present before my eyes A desk is present before my eyes Birds sing in proximity to my ears Matches up appropriately with the sequence of events in P: A blue colour-patch appears A brown colour-patch appears A twittering sound occurs Perhaps what makes each relevant event in M the cause of an event in P is that each P-event matches up appropriately with some M-event in terms of its place in the ordering of the two sequences. The first problem with this picture is its symmetry. While it is true that the order of events in P appropriately corres (...truncated)