Austin’s Critique

Philosophy of Perception: Another Conception
    While scientists are interested in constructing theories about the mechanisms and processes underlying perception, philosophers should be primarily interested in our perceptual concepts. In philosophy of perception those who put forward philosophical theories have often ceased to make contact with our ordinary concepts of perception. The task of the philosophy of perception can be thought of as being to bring us back to our own concepts, to remind us how we use them. This is the non-theoretical conception of philosophy.
    How do we identify our perceptual concepts? We ought to look at the language in which we speak about perception. This is largely unchanged (in its basic features) over hundreds of years. Our language shows us what our concepts are, and these are what we’re trying to get straight. Only after we have a description of our perceptual concepts can we decide whether there is a theoretical job left to be done: we may not even need a philosophical theory of perception.
    Our perceptual concepts tell us what it makes sense to say about perception. They tell us that it makes sense to speak of perceiving an enormous variety of things like tables, chairs, people, sounds, smells, tastes, etc. They tell us that it makes sense to say that statements about objects like these can be known to be true on the basis of perception. They tell us that it makes sense to speak of the objects we perceive, not only when they are being perceived but also when they are not being perceived, and that these objects are also able to retain properties of the types we perceive them as having, even when they are not being perceived.
    Some philosophers have denied this. They believe that we never 'directly' perceive the sorts of things we think we perceive, but only our own sense-data (ideas, impressions, etc.). And they then tell us that although this sounds shocking, it’s really just what we’ve believed all along!

J.L.Austin: Sense and Sensibilia
    Austin tries to show that the theorists of perception haven’t even got their phenomena, the data for their theories, straight. does not see himself as defending ‘naive realism’, or indeed any kind of realism (Sense and Sensibilia, (Oxford: Oxford University Press, 1962), p.3). In fact he doesn’t put forward any theory of perception. His project is the more modest one of dismantling the arguments for sense-data, of “dissolving philosophical worries” (p.5), and of allowing us to learn something about the meanings of certain words. In other words, he is committed to the non-theoretical conception.

'Material Things'
    Austin emphasises that we have a chance to dig our heels in well before the end of the 'argument from illusion'. For one thing, naive realism systematically misrepresents the layperson’s perceptual concepts. The first place it does so is as soon as it suggests that commonsense involves the idea that we normally think that when perception goes right, what we perceive directly is always a material thing. There are two quite distinct mistakes here, one concerning ‘material things’ and one concerning ‘direct’ perception.
    The first mistake is the suggestion that the ordinary person thinks he or she normally perceives material things. The terms ‘material thing’ and ‘material object’ are philosopher’s terms of art, not common sense concepts. To explain these terms, adherents of the argument from illusion fixate on examples of ‘familiar objects’ (chairs, tables, pictures, books, flowers, pens, cigarettes), which Austin memorably calls “moderate-sized specimens of dry goods” (p.8). But ordinary people talk as if they perceive many things, by no means all of which are anything like these. As well as ‘familiar objects’ like tables and chairs, ordinary people are perfectly happy to talk of perceiving (or rather, seeing, hearing, tasting, smelling or feeling) all of the following: people, people’s voices, rivers, mountains, the earth, flames, rainbows, shadows, the sky, the beam of a searchlight, pictures on the screen at the cinema or on television, pictures in books or hung on walls, vapours, gases, reflections, flashes of light, tastes, and smells. Whilst all these things have a physical aspect, very few of them fit comfortably into the category of ‘material thing’.
    Austin perceptively remarks that “The trouble is that the expression ‘material thing’ is functioning already, from the very beginning, simply as a foil for ‘sense-datum’; it is not here given, and is never given, any other role to play” (p.8). The two terms, ‘sense-data’ and ‘material things’, he says “live by taking in each other’s washing - what is spurious is not one term of the pair, but the antithesis itself. There is no one kind of thing that we ‘perceive’ but many different kinds” (p.4). So much for the daft philosophical question ‘What is it that we perceive?’.

Non-Standard Cases of Perception?
    Equally wrong are certain philosophers’ suggestions about what ordinary people think is happening when perception is not standard or normal. They suggest that ordinary people believe that when they are not perceiving a material thing, they think they are being deceived by their senses, and, correlatively, that when they’re being deceived by their senses, they’re not perceiving material things. But perception of many of those things we can be said to perceive but which are not ‘material things’, like rainbows, mirror-images, etc., does not tempt one to think that one’s senses are deceiving one. There is no one kind of thing which we think we perceive when things are going wrong with perception, any more than there is a single kind of thing we think we perceive when things are going right.
    When we say that people are sometimes ‘deceived by their senses’, we are using a metaphor. Our senses, unlike other people, do not literally tell us anything (p.11). Gilbert Ryle says of this that:

The notion that our eyes, ears and noses are foreign correspondents who send us messages, which, on examination, turn out often and perhaps always to be fabrications, does enjoy a wide vogue... People make mistakes, are confused, fail to make things out, overlook things, and so on, in looking about them as they do in calculating, translating, demonstrating and playing games. But only misleadingly can these troubles be described as the outcomes of false or ambiguous messages from reporters. For reporters are themselves good or bad observers, and the critical or uncritical recipients of information from others. So to liken our eyes to reporters is simply to push back the question of the sources of error by one stage. (‘Perception’, in his Dilemmas, (Cambridge: Cambridge University Press, 1954), p.99, my emphasis).
Ryle’s complaint contains an important and devastating objection to the attempt to take the metaphor of senses as messengers too seriously. The objection is that to do so is to be implicated in the homunculus fallacy. This is the fallacy of postulating a ‘little man’ in the brain or mind in order to explain the capacities of the person. To commit the fallacy is to “extend predicates which can only intelligibly be applied to a human being (or sentient creature) as a whole to parts of the creature” (P.M.S.Hacker, Appearance and Reality, (Oxford: Blackwell, 1987), p.19). Homuncular ‘explanation’ is useless because the postulation of an homunculus reproduces whatever problem it was invoked to explain.
    Further, Austin insists that the layperson’s view is not correctly represented by the thought that ‘our sense-perceptions can be trusted’. This not only involves the misleading metaphor of communication, but also introduces some new entities, sense-perceptions, which don’t figure in our ordinary perceptual vocabulary. The idea is that “whenever we ‘perceive’ there is an intermediate entity always present and informing us about something else - the question is, can we or can’t we trust what it says?” (p.11). But none of this is any part of the layperson’s battery of perceptual concepts; in fact, Austin suggests, it is the philosopher’s view, wrongly foisted on the layman! ‘Sense-perceptions’ or ‘perceptions’ are nothing but obvious stooges for the forthcoming attraction: ‘sense-data’.
    The correct use of the term ‘deception’ is also far more restricted than the adherent of the argument from illusion supposes. There are just far fewer cases of ‘deception’ than the philosopher likes to think. We are not ‘deceived by our senses’ when we see perspective, ordinary mirror-images, after-images, or when we dream. And even the cases we do accept as cases of deception are of importantly different kinds.

Doubt and Scepticism
    Adherents of the argument from illusion chide us for not being sensitive to the fact that our belief in the existence of material things needs justifying, for not doubting whether we perceive tables and chairs, for being all too easily satisfied. Austin complains that there is often no room for doubt or suspicion. Doubt is not sophisticated and philosophical, but senseless in such cases.
    Ayer retorts that “the reasons which have led sense-datum theorists to speak of uncertainty in this connexion... remain untouched” by these considerations (Ayer, p.286). He insists that their point, although perhaps expressed in a misleading way, was the ‘purely logical’ one that “in any such situation as that described by Austin the occurrence of the experience which gives rise to the perceptual judgement is logically consistent with the judgement’s being false” (Ayer, p.286). In other words, there is no logical connection between the experience and the ‘material thing’ that is causing it. We could quite well have the former without the latter.
    This is untrue. For a start, one cannot have the experience of looking at or seeing a chair without there being a chair there that is seen. That’s because no situation in which no chair is in the offing can correctly be described as involving an experience of looking at or seeing a chair. To be looking at or seeing a chair, there must be a chair one is looking at. In general, to have the experience of looking at or seeing an X, (where an X is something which it makes sense to say is physically located) one must be in the vicinity of an X, awake, attentive, facing the right direction, and with the relevant sense organs in good fettle! Ayer complains that this description of the situation begs the question, but that

the point on which the sense-datum theorist takes his stand is that the situation does not have to be described in this way;... The kind of description which is needed for this purpose is one that will uncover rather than conceal the fact that the observer could be having the experience in question even though the physical object which he takes himself to be perceiving did not exist: that the occurrence of the experience is consistent with his having been hypnotized or otherwise deluded. (Ayer, p.286, my emphasis).
Well, is it a fact that “the observer could be having the experience in question even though the physical object which he takes himself to be perceiving did not exist”? That is, is our conception of experience such that this could be literally true? Could you correctly be said to be having that experience in the absence of the physical object which is allegedly causing it? Unfortunately for Ayer, the answer to this is ‘no’.

    As for alleged sceptical possibilities, Austin famously points out that “Talk of deception only makes sense against a background of general non-deception... It must be possible to recognize a case of deception by checking the odd case against more normal ones” (p.11). ‘Deception’ that is in principle unrecognisable is no deception at all. Ryle agrees:

In a country where there is a coinage, false coins can be manufactured and passed; and the counterfeiting might be so efficient that an ordinary citizen, unable to tell which were false and which were genuine coins, might become suspicious of the genuineness of any particular coin he has received. But however general his suspicions might be, there remains one proposition which he cannot entertain, the proposition, namely, that it is possible that all coins are counterfeits. For there must be an answer to the question ‘Counterfeits of what?’ (‘Perception’, in Dilemmas, pp.94-5, emphasis added).
Sometimes we make mistakes, and we may even do so at the very same time as we are reminding ourselves of this liability. We may, for example, miscount the number of chairs in a room. So it looks as if we ought to concede that we can never find out the number of chairs in a room by counting. But we do not concede this, and this is because we already have all the tests and checks we need. We can, that is, count again, carefully. And this carefulness will not be “merely a useless, anxious watchfulness against nothing in particular. It will be vigilance for just those specific slips which we and our associates have made before and detected and corrected before” (Ryle, p.95). We know by experience what it is like to miscount, and what it is like to avoid, detect, and correct those miscountings.
    Deception is something that happens, and that can be detected and corrected for, within our experience. The idea of a ‘logically unrecognisable or undiscoverable deception’, Ryle seems to be saying, is a nonsense, a contradiction in terms.

‘Direct’ and ‘Indirect’ Perception
    The argument from illusion introduces (that is, invents) a distinction between direct (or ‘immediate’) and indirect perception. It characterises the layperson’s view as being that we directly perceive material things. Against this, Austin insists that ordinary people should not be represented as believing that the things we say we normally see, hear, smell, taste, etc., are perceived directly. We would do well to resist the introduction of this distinction. Introducing the notion of directness already tilts the grounds of the argument in favour of the ‘indirect realist’, or some other advocate of ‘indirect’ perception. We ought, instead, to maintain that the distinction just doesn’t apply. Two objections to the distinction:
 (1) The idea of perception being indirect has no application, makes no sense, except where some instrument (like television, microscope, telescope, etc.) intervenes.
 (2) Transferring the distinction from cases where it is sensible to cases of ordinary unaided perception is unwarranted, because cases of ordinary unaided perception are paradigm examples of cases of direct perception, and they can’t viably be pushed into the class of cases of unbeknownst-to-us indirect perception.

It makes sense to talk of perceiving an object indirectly only if it makes sense to talk of perceiving it directly. One may say of a certain person that one could not, on a given occasion, see him directly but only in a mirror. But then, of course, it makes sense to speak of seeing him face to face... The idiom of direct/indirect perception applies paradigmatically to vision, and there principally to cases in which the direct line of vision is (or is not) broken. To hear something indirectly is to come to know of it by hearsay; and there is little if any use of the notion of smelling or tasting indirectly. (Hacker ibid, p.47).
It makes sense to talk of perceiving an object indirectly only if it makes sense to talk of perceiving it directly. So someone who complains that we never perceive material objects directly must be able to coherently specify some kind of possible but (by his account) non-actual situation which would count as a human being’s perceiving a material object directly. Typically, opponents of commonsense never do any such thing. But what if we now asked them to do so? What could they say? If they concede that there is no possible situation in which a person could directly perceive a material object, then they have failed to introduce a ground for their direct/indirect distinction: they are merely changing the concept of perception. On the other hand, they may say that there is a possible but non-actual situation which would count as directly perceiving a material object. For example, they may say that if the perceiver’s mind actually enfolded or was identical with the material object in question, that would count as direct perception. But here we can just make Malcolm’s response: the ‘possibilities’ envisaged simply don’t make sense, in fact they are each a contradiction in terms (since minds and material objects are in different categories).

Illusion and Delusion
    The argument from illusion also conflates illusion with delusion. Ayer cites phenomena like refraction, mirages, and reflections, and implies both that these are illusions, and that illusions and delusions are the same thing. But, as with the concept of deception, the philosopher has here stretched things too far. Both the implications of Ayer’s argument are wrong.
    What about the alleged fact that a coin ‘looks elliptical’ when seen from some angles? Here there appears to be some disagreement even among careful thinkers. Austin says that this is “exactly what we would expect and what we normally find” (p.26), and therefore not an illusion. “Familiarity”, he says, “takes the edge off illusion” (p.26). The fact that some setup might deceive a small child or an ignoramus does not mean that we can correctly describe it as an illusion. Adherents of the ‘argument from illusion’ persistently misuse the concept of an illusion. But he says that they look elliptical “in one sense” (p.26), so perhaps his verdict is not in conflict with that of Peter Hacker, who says that “pennies... do not, save in very poor light and from a distance, look the slightest bit elliptical; nor do they look as if they are elliptical. They look just as any round flat object does when viewed from such an angle” (p.218 note 6).
    Delusions are something entirely different from illusions, but again are misrepresented by hasty philosophers. They are a matter of having “grossly disordered beliefs” (p.23), and they don’t necessarily concern perception. The term ‘delusion’, but not ‘illusion’, suggests “something totally unreal, not really there at all” (p.23). And this implication is used by the ‘argument from illusion’ to suggest that in all of the cases cited something unreal or immaterial is conjured up. This implication is totally false. Nothing immaterial or unreal is conjured up when I look at a straight stick half-immersed in water and take it to be bent. I’m seeing a straight stick, and nothing bent is in the offing at all.

    Austin spends two chapters arguing that the uses of the word ‘real’ are multifarious, and that this variety is simply ignored by the sense-datum theorist who plays fast and loose with the term. Austin points out that the word ‘real’ is a normal word with an established meaning, not a philosopher’s technical term, but nevertheless that it has no single meaning which is constant across all contexts of its use. This does not mean, however, that it is ambiguous. The function of the word ‘real’, he says, “is not to contribute positively to the characterization of anything but to exclude possible ways of being not real - and these ways are both numerous for particular kinds of things, and liable to be quite different for things of different kinds” (Austin, p.70). Possible ways of not being real really are numerous. We can contrast real limbs with artificial limbs, real teeth with false teeth, real cream with synthetic cream, real diamonds with paste diamonds, real bullets with dummy bullets, real ducks with decoy ducks, real cars with toy cars, real horses with pictures of horses, etc., almost ad infinitum. Because there is no single contrast between the real and the non-real, Austin says,

there are no criteria to be laid down in general for distinguishing the real from the not real. How this is to be done must depend on what it is with respect to which the problem arises in particular cases. Furthermore, even for particular kinds of things, there may be many different ways in which the distinction may be made. (Austin, p.76).