Wednesday 23 March 2016

Travis on seeing aspects

Two months of not being able to think or write (and making very slow progress on a chapter about the differences between McDowell and Wittgenstein on the harmony of thought and reality) may be coming to an end in some conversations with Peter Lucas initially inspired by a paper on aspect perception by Charles Travis called ‘Suffering intentionally?’.

Here’s a na├»ve way of thinking why Travis’ take on aspect perception might be interesting. The duck-rabbit is an ambiguous figure and one can see it as a duck or as a rabbit (and as a duck-rabbit, a Jastrow figure etc etc). One can see that it is all these things too. But it is also possible to experience changes in the aspect seen, perhaps best expressed by the spontaneous exclamation “Now it’s a duck!”. Given that, obviously (in two senses of that word), the figure does not itself change, it seems that the perceptual experience must change in some other way. Whilst Wittgenstein criticises the explanatory resources of any shifting inner image with differing ‘organisation’, it is tempting to think that there is an innocent account of the organisation of experience in which different elements of one’s conceptual repertoire are drawn on in the structure of experience itself when one takes in the differing aspects.

In the philosophy of science, NR Hanson concludes that all seeing is conceptually-articulated seeing-as. Stephen Mulhall claims that the point of the discussion of cases of changes in aspect is to illustrate the general nature of continuous aspect perception where the latter characterises our normal immediate response to words and to the world. I’ve never liked this because it seems to me that Wittgenstein makes it clear that the experience of aspect perception is contingent. It would be possible to be aspect-blind – ie not see shifting aspects even whilst able to judge that there is a duck and rabbit image in the picture – and hence changing aspects do not highlight the general nature of thought’s bearing on the world. Still, it is tempting to think that whatever is going on in aspect changes is a matter of the nature of the content of the visual experience. And hence it seems necessary to think that visual experience has some representational content to accommodate aspect changes. Travis famously rejects that claim, however, so what can he say about this?

Having sketched Wittgenstein’s rejection of an inner image to explain things, with the thought: ‘The inner image offers no better candidate for what a shift in looks consists in than the Necker does itself. It makes no progress.’ He gives part of his own positive view.

On my understanding of the Necker (perhaps not Wittgenstein’s) what changes in a shift is what is seen, ‘see’ here a straightforward perceptual verb, its object thus something there for one to see. For what I see to change here would then be for there to be a change in uptake – in what, there anyway to be seen, impressed me, was registered – under those special conditions the Necker poses, in which taking in one thing precludes taking in another. If this view of the Necker is correct, then I may say, ‘I see the A-cube’, if that is what I am doing, using ‘see’ as a straightforward perceptual verb. [ibid 51]

The first element of this is that Travis offloads the two aspects onto a feature of the world: the ambiguity of the picture. This stems from something I have often forgotten:

The thing to remember is that the rabbit, the duck, the tiger or the yacht difficult to pick out among the mass of dots, are, first of all, images in pictures, and, second, things which are there to be seen. [ibid: 48]

We are not talking about seeing actual ducks and rabbits. In this context it may seem illicit to appeal to the intentional properties of pictures but I don’t think it is. Pictures only depict on the basis of background practices – as Wittgenstein stresses using examples of signposts – but assuming those practices, then there are looks or appearances there to be noticed (for those with eyes to see).

So the first clue to an account of aspect perception which denies that perceptual experiences have content is that there is ambiguity in what is there in the world to be seen. The pattern of lines making up the Necker cube or the duck rabbit is there to be seen but so is the image / picture of a cube, duck or rabbit. The duck-rabbit really does look like a duck and look like a rabbit. Both of those looks or appearances are there in the cunningly constructed figure and so both are there to be seen.

Two (related) things are distinctive about the Necker (and the duck–rabbit, the black/white cross, etc.). First, while it provides us two different things to see, it does this while occupying a given location, and providing us the same thing at that location to be seen (or missed) no matter which of the two options we take up. At that location is a pattern of lines. It remains the same where we see the A-cube and where we see the B-cube. And (unless we miss it, hence see neither cube) that occupant of the location is something we do see throughout. [ibid: 52]

But that is only part of the solution because it has yet to accommodate the fact that, at any particular instant, the visual experience of looking at such figures is not ambiguous. The figure is experienced as a duck (not that Travis deploys the word ‘as’ to mark this). The passage continues:

Second, in this case, seeing the one thing (the A-cube, say) excludes seeing the other. The exclusion lies in the seeing: what is there to be seen leaves either option open. So the question is how seeing can impose a choice between these two. Here, clearly, uptake is all. Our responses to what we are presented with visually must choose what kind of visual experience we have, or what it is of. Without varying what we are presented with, our responses, or uptake, might make what we see either of two things. [ibid: 52]

Given that the worldly object does not change, the change has to lie on the side of the subject or agent in their ‘uptake’, or in what they ‘register’. But this idea is also put under tension in Travis’ account because he stresses here, as in a number of other papers, a distinction between seeing construed as a perceptual verb and seeing-that as an expression of thought or judgement.

But, as Frege noted (1897: 149; 1918: 61), ‘see’ has non-perceptual uses. When we say, for example, ‘He sees that that flower has five petals’ (one of Frege’s examples), ‘see’ is not used to report perceptual awareness. One way to see this is to note that, while the petals are on the flower in the garden, and from thence form images on retinas – they are that sort of thing – that the flower has five petals is not in the garden. Nor is it on the kitchen table. It is not the sort of thing to be located, a fortiori, to form images on retinas. Rather, it is the sort of thing one recognises by exercising capacities of thought. In thought, one can represent the flower as falling under a certain generality; as being a certain way there is for a flower to be – five-petalled. [ibid: 46]

Elsewhere, he suggests that ‘recognition’ is ambiguous between two senses. One is to re-cognise the same individual. The other is to make a judgement that an item is sufficiently alike others to count as an instance of a generality. Again, these divide between perceptual and non-perceptual seeing.

The distinction between perceptual seeing and thought or judgement seems threatened by cases of aspect perception.

As I read Wittgenstein, he is developing reasons for caution in how we appeal to responses here, or in just what in responses might do that selecting which, patently, is done. Where ‘see as’, read non-perceptually, adds something to what would be said here in the simple ‘see’, prima facie, at least, that something is not something appeal to which would do the needed selecting. So, for example, I may see how it could be a diagram of an A-cube, or see it to be, or see it as one (as Sid may see Vic as a rival). But that sort of uptake, if it belongs to what is added in seeing-as – an addition purely in thought, on a par with recognising what I see as the cube on the exam yesterday – cannot be what distinguishes seeing an A-cube from seeing a B-cube in looking at a Necker. If uptake does the choosing, then I see the A-cube just where I am responsive in a given way to what I see. For me, things are, visually, just as they would be in seeing the A-cube. Such is an attitude on my part. But it is not the same as seeing that such-and-such, nor seeing the Necker as such-and-such, where ‘see-as’ does not speak of perceptual awareness. [ibid: 52]

The most obvious ideas for an uptake, for what is made of a visual experience, is a form of judgement or thought (and hence the second form of recognition). But that cannot be the distinction between the two aspects as one can judge that the duck-rabbit can be seen as a duck even whilst having the rabbit aspect. Subsequent judgement of that sort cannot explain how the aspect is fixed at any given moment. In aspect perception it seems that the experience is a matter of object perception not of judgement.

In fact, contrary to popular usage, Travis doesn’t use ‘seeing-as’ to signal aspect perception. He suggests instead (aside from some cases of illusion) that seeing-as denotes the judgement or thought made on the basis of a properly visual experience.

Obviously, though, Wittgenstein is not thinking of such uses [ie. cases of illusion] of ‘see as’ (whether he allows them or not). The particular case he mentions is seeing the rabbit–duck as a rabbit. This is not to be a matter of seeing the rabbit-likeness. It is not a matter of visual awareness. So it is not to be classed with seeing convex as concave; nor with the sort of Necker-like switch one experiences when he finally gets what he knows ought to be a rabbit–duck to look like the rabbit, and not the duck. So, presumably, it is something which may occur or not no matter which way the picture looks to the viewer at the moment. Someone might say, for example, ‘The funny thing about this rabbit picture is that, sometimes, if you screw your eyes up right, it looks just like a duck-picture.’ Someone who said that would be one who took the picture for, classified it as, a rabbit picture. This, of course, is a way of thinking of it, and not a form of visual awareness. [ibid: 57]

Given that neither seeing that nor seeing as will do to mark the nature of the uptake that transforms a properly visual experience, it is hard to know how that notion can be characterised. But if I follow the line of thought, this is part of the point.

It is thus, I think, that the Necker and its kin serve as a stepping stone to a range of further cases. Of some of these, at least, Wittgenstein makes the following remark: 
Here one must guard against thinking in traditional psychological categories – say, decomposing experience into seeing and thought; or things of that sort. (LW I 542) 
This ‘here’ refers to a particular kind of case (or several). It is not as though, for Wittgenstein, Frege was wrong to distinguish between seeing and thinking as he did. An ordinary case of seeing is just that, seeing. It is not thinking. Its objects are fundamentally different: in one case, a pig chewing turnips (an episode taking place at a particular location in history); in the other that the pig is chewing turnips, something without location, either temporal or spatial (though what it represents as some given way is the way things are at a time, and, perhaps, a place). But there are special cases of the use of ‘see’ such that something speaks in favour of regarding them as perceptual uses, something in favour of regarding them as not perceptual; and in such cases, Wittgenstein suggests, the urge to choose should sometimes be resisted. [ibid: 52-3]

One such example is seeing a similarity between two faces which Travis suggests is a matter of seeing a look which a face can have, something there to be seen. Is the difference between seeing and not seeing that look a difference in visual experience? Travis suggests that there are reasons for answering both yes and no. He suggests that to represent what is seen may involve a kind of distortion, not just sketching again the two faces in question. It will involve a representation that has to be understood a particular way: a mark of judgement or thought. Such seeing not only registers what is visually there but also brings conceptual generalities to bear. And yet it seems as though the look itself is directly visually present as an object. It distorts Frege’s distinction.

Still I am confused by the nature of the uptake needed since any attempt to fill it out seems destined to tip it into non-perceptual seeing that. There is one further example: staring at a kitchen counter and not seeing some keys there and then suddenly noticing them. Although this thus allows subsequent judgements or thoughts, Travis comments that ‘my new awareness is just of what it is that history furnishes to do the instancing’. This makes this a case of perceptually seeing. But what is the difference between not seeing and then seeing the keys in this case? What is the uptake captured in the ‘new awareness’ given that it is not framing judgements?

It now seems to me that Mulhall and Hanson offer an explanation of aspect-seeing which overly domesticates the phenomenon. Indeed, a feature of their accounts is that it is ubiquitous. Travis, correctly in my view, suggests that it is a special case and also offers a kind of explanation of why it is, pre-philosophically, an odd experience. The sense that there has been both a change and no change. But it is less clear what his account of the phenomenon itself is.

Travis, C. (2015) 'Suffering intentionally?' in Campbell, M. & O'Sullivan, M. (eds.) Wittgenstein and Perception, London: Routledge