Against the Proliferation of Senses

The official topic of Chapter 9 of Sense and Sensibilia is Ayer's discussion of the question: "Are we to say that the objects we directly perceive are sense data?," given Ayer's conclusion that the argument from illusion provides us with no reason to say so. Ayer argues that we should say so when discussing perception in order to avoid ambiguities in the way in which verbs of perception are ordinarily used. To employ that argument, Ayer attempts to establish that verbs of perception have many different senses. Austin in fact devotes the chapter to undermining the claim that verbs of perception have different senses. His method is far more general: he attacks the standard philosophical move of distinguishing various senses of words in a quite general way, merely using Ayer's discussion of perceptual verbs as a stalking horse. The whole setting is rife with ironies: Ayer, who has argued that words do not have meanings, or senses, in any general sense of the terms argues that perceptual verbs have many senses. Austin, who extended Ayer's argument that there are no meanings in a general sense of the term to an argument that there is no single kind of object of which our world is composed, whether that is material objects or sense data, replies that perceptual verbs typically have a "single, ordinary 'sense'" (98). Now, to the Chapter.

Can you see something that isn't there?

As Ayer would put it, is there a meaning, a use, of the perceptual verb "to see," on which one can see something that doesn't exist at all?

Austin flatly denies that there is any such use. If I truly believe that there are no ghosts, I cannot seriously claim to have seen one, or even to take any else to have seen one. I might say, "I thought I saw a ghost" ("Cousin Josephine thought she saw a ghost"), "I saw something that looked like a ghost," "I saw something as a ghost," but I cannot say, except perhaps as a kind of sloppy abbreviation, "I see a ghost." Similarly, if I am having double vision, and I know it, I can't literally and correctly say, "I see two pieces of paper"—there aren't two pieces of paper ,and I know that full well. I can say, "it looks like I am seeing two pieces of paper," and so forth.

What about, "I am seeing a silvery speck" and "I am seeing a lump of hot gas larger than the earth"? Those seem to Ayer to require that I am using the verb "to see" in two different senses. He even says that the two statements seem to contradict each other. But, as Austin points out, they don't: the silvery speck is the star, the lump …. Even when we are seeing one and the same thing, it can be described in many ways: "I am seeing the image from the telescope"; "I am seeing the rings of Saturn." "I saw NBC"; "I saw West Wing"; "I saw President Bartlett"; "I saw Martin Sheen"; "I saw Charlie Sheen's father." Those are all perfectly normal and correct ways of describing things that may occur in a single perceptual setting. What is seen need not differ in any of those cases.

We describe one and the same object in many different ways depending on our interests, commitments, and, as Austin puts it, how far we may be willing to stick our necks out. Thus, in 2003 I might have said all of the following: "I saw the President of the United States"; "I saw the man the Supreme Court held to be the President of the United States"; "I saw the unelected man who acts in the place of a President of the United States with the connivance of his father's Supreme Court." Those are quite plausibly not different in perceptual ways, but Austin argues that these cases are entirely analogous, and no more symptomatic of different uses of "to see" (99): "I saw a star"; "I saw Sirius"; "I saw a small twinkling speck"; "I saw the image on the 14th mirror of the telescope."

There are not, if Austin's argument is correct, multiple senses of "to see," and what is seen must exist. Austin can't handle two of Ayer's examples in that way: "I saw two pieces of paper," said by someone with double vision, and I saw "pink rats," said by someone with the DTs. He argues that those are just not normal uses of "to see." Unusual situations force us to adapt language in various ways, since we don't have ordinary resources to handle them comfortably. Surely it would be more careful to say, "It looked like there were two pieces of paper" and we would say of the person with the DTs, "he thinks he is seeing pink rats," or "he is hallucinating pink rats," and if even if we do say, "he is seeing pink rats," we know perfectly well that we don't really mean it, it is a shorthand for "he believes he is seeing pink rats," or the like.

Aside. I don't think Austin can legitimately just dismiss certain uses of "to see" as abnormal, as he does in the previous paragraph. He claims (in the next chapter) that the uses of words are given by characteristic examples, not by definitions, and that we often stretch them in various ways, to argue that there is no fact of the matter about what can be inferred from a use of a word. The uses of the verb "to see" that he takes to be abnormal stretch the word in perfectly ordinary ways, which is something he seems to be concerned to allow. Thus, he needs an argument that there is a real dichotomy between normal and abnormal uses. Such dialogs as the following suggest that it may not be so easy to do that.

I have been focused on Austin's discussion of Ayer's claim that there is an "existentially delusive" use of to see, in which one can say that one sees something when it just isn't there. If we discuss qualitative delusions, Ayer says something different: there is a use of "see" such that an object must have the properties it appears to have: "I see a twinkling speck"; and a different use in which it need not: "I see a huge object, a star, that appears to be, but is not, a twinkling speck." Austin denies that there are two different uses here, and he denies, quite rightly, that the speck is anything other than the star. We can just describe what we are seeing in different ways, depending upon what features are of interest and what beliefs we are prepared to take for granted.

People sometimes say things like, we don't really see objects, only a portion of their surfaces. No, Austin says, when we see an object, it may sometimes be appropriate to describe what we see as a portion of the surface of some other object, sometimes appropriate to describe what we see as the object. Moreover, we do sometimes see a whole object, not a part ("I saw the whole piece of paper, and it was completely blank." That does not imply that I saw the other side of the piece of paper. "Whole" is not so simple as that. In addition, we don't always see a surface at all: where is the surface of a cat?)

After all that discussion, Austin admits that he doesn't really think it has anything to do with Ayer's argument after all. Ayer says we need a single use for "see," not subject to the vagaries of ordinary use, and he elects to use the word so that what we see must always exist and have the properties it appears to have. Well, whatever the ordinary uses, he can just ignore them and adopt a new use, since he is clear that he is doing so. However, notice that the new use is not any of the ordinary ones he claimed to give examples of. However, his motivation is not "improving ordinary use," as he says. It is introducing incorrigible propositions, as will become clear next time.

If you adopt Ayer's use of "to see," what can you be said to see? Certainly none of the things we ordinarily see or think we see, and so the inevitable answer is sense data.

-- ShaughanLavine - 18 Apr 2003 - 10 Apr 2007