ON THE NONEXISTENCE OF PERCEPTIONS

Abstract: Perceptions as philosophers understand them are absolutely crucial for grounding the problem of knowledge, the problem of appearances and reality, the dreaming argument and the argument from illusion, the existence of epistemology. The problem is that no arguments for their existence are sound, and investigation into their existence by way of nonphilosophical examples betrays them. What do exist are quite different entities with different logics in different and specific examples, not all the examples of seeing, etc. as the problems require.

I argue here for the claim that perceptions as philosophers have understood them (and as many others guided by philosophers have understood them) do not exist. There are hosts of interesting linked political and historical and psychological issues around these questions--since perceptions don't exist, why in the world have so many apparently fierce and tough-minded thinkers rolled over to have their tummies rubbed by these mythological beasts?--but I'm going to pass all those by.

This essay is a response to the traditional issues of appearance and reality, to the dreaming argument and the argument from illusion. If my argument is correct, then those problems and those arguments have something wrong with the thinking which underlies them. The argument from illusion and the dreaming argument require the ubiquitous existence of some things which in fact do not exist except in particular cases, namely perceptions/appearances/the way things seem to me now/the way things look to me now, and without the ubiquitous existence of those things not only is there no answer to the problem but there's no problem. This essay, then, results not in an answer to the problem but instead in the problem's dissolving or evaporating. All this occurs partly because, again, we can clarify how, when, and in which examples perceptions do exist--in brief, they exist only in particular kinds of examples of our seeing things, and not in all examples (as the problem requires).

Russell begins his The Problems of Philosophy with not one but a nest of problems. Appearance and reality, after all, is a problem about what exists besides appearances, but it also is a problem about how appearances may be a source of knowledge, a problem about the nature of perceptions or appearances themselves, a problem about human beings' sensory apparatus and psychology. It is a problem about the nature of reality but also is about the knowability of reality; it is a problem in part about the relation of body to mind (since appearances are taken to be mental objects), a problem about the relation between subjectivity and objectivity, about privacy within the mind and the public reality outside, and perhaps a problem about human alienation from the world.  We are, in other words, at the headwaters of a large and complex philosophical watershed.

Here's a reminder what we mean by appearances when we are doing philosophy.  We mean a family of things, designated by vocabulary largely interchangeable with each other.  We mean how things look to us, how they seem to us, how things are perceived by us--where perceived is to be distinguished from the purely mechanical and biological workings of our senses--and we mean these things to be in contrast with the reality or the possible reality of the matter when we are being careful to designate our immediate experiences of things from our judgments based on those experiences.  One aspect of this meaning which has found its way into the talk of educated persons so much that it is no longer controversial is our talk of perceptions.  Psychology texts now regularly tell high school and college students that the psychology of sensation and perception involves a distinction between, on the one side, the causal workings of the world and our biological and neurological apparatus of eyes and ears (the working of which is called sensing) and, on the other, the seemings, looks, aspects, qualia, experiences which result within us from those causal workings.  And both these are to be distinguished from human belief systems about the world which are based at least in part on our sensations and perceptions. [Footnote to Zimbardo and other texts in Psych]  Perceptions talk, in other words, is a contemporary way to get at what is meant by appearances talk in philosophy.  Indeed, one suspects it is the way.

Consider the argument from illusion, in the following form

The way that things look/seem/appear/are perceived right now could be the same whether the result of illusion/dream/hypnosis/psychedelic-drugs/hallucination/delusion or the result of reality.  Therefore I cannot tell the difference.   Therefore I cannot tell that I am not now suffering from some illusion/etc.  Therefore I do not know that the world is as I perceive it to be.

This argument has been thought by many philosophers to establish the need for epistemology or philosophy of knowledge. The argument from illusion has also been taken to establish the problematic nature of human knowledge of the external world.  In its more narrow version as the dreaming argument it has proven itself remarkably invulnerable to attack--no suggestion of certain signs by which I may know I am not now dreaming will stand against the logical possibility that the suggested sign, even if is is applicable in the present moment, could be dreamed.

First, there is a problem in that the argument is framed in such a way as to take Cartesian dualism for granted. Look at the wording of the main premise: "The way that things look/seem/appear/are perceived right now could be the same whether the result of illusion/dream/hypnosis/psychedelic-drugs/hallucination/delusion or the result of reality." Whatever else is being done by the subject of the sentence, it clearly refers to mental phenomena, and then the phrase "whether the result of illusion/dream/hypnosis/psychedelic-drugs/hallucination/delusion or the result of reality" divides the possible causes of the mental things referred to in the subject into those causes which are mental on one hand (the illusions) or physical on the other (reality). Another way to say this objection is that the argument from illusion begs the question of whether there are the private internal perceptions upon which we allegedly build our inferences about the public external world. That is, those internal perceptions are taken to be mental results not because it is argued that they are and not because we have gone to check any perceptions to find out whether they are private or mental, but because we take it for granted that they are. In other words, though the argument looks like an argument for the difficulty of bridging a gap between appearances and reality, it establishes that gap not by argument but by building it in at the beginning, using terms which import the gap. The charge then is that the argument from illusion requires but does not establish the ubiquitous existence of perceptions, which come trailing clouds of problems.

There is only one alternative can save the argument from this charge.  That alternative, that rescue, requires we independently establish that there are clearly all the time the appearances/perceptions/seemings/looks which the argument from illusion requires.  Here is how that would go.  Take seeing things first. Whenever I see something, there is a way that that something looks to me. Similarly, supposedly, whenever I hear something, there is a way that that something sounds to me, and generally my perceiving something involves there being a way that something seems to me and is perceived by me and appears to me.  This is not an unknown thought. The story that Berkeley tells along the way to claiming that material objects do not exist, or the story in Hindu philosophy of the veil of Maya, or the story about perception generally told by contemporary psychology textbooks, if true, would save the argument from illusion.  Between my seeing and the world, if this alternative is true, lie perceptions (or, to invoke some technical terms, qualia or sense data), such that my seeing is really seeing of an intermediary, and my understanding of what I see is in part inferential, based on processing the appearances/perceptions/looks/seemings in order to make judgements about the existence of the real things which lie underneath and which cause those seemings.

We might attempt to establish that this alternative is true, and thus rescue the existence of perceptions and rescue the problems, by sketching out how we have to perceive things, and finding in that sketch a requirement that perceptions are always present.  This too is not an unknown thought. That sketching out would look like this.  I see out my office window the redwoods into which the great white egrets come to roost about now each evening.  That, my seeing, works as follows.  There is the sun, which shines light on all, including the redwoods; most of the light is absorbed but a little gets reflected off in all directions.  A tiny fraction of what gets reflected gets reflected toward me and strikes the front of my eye.  It is there gathered and focused by the lens and humors of the eye onto the retina, is there converted by nerve cells including the rods and cones into neurochemical impulses which travel up the optic nerve to the visual centers of the brain, which gather and process them in order to construct a representation of what I see, such that finally there crawls over the boundary between body and mind a tiny thing, a perception, which in this case is the way the redwoods look to me.

You recognize from the snotty tone that I find this problematic.  You recognize that the story is not a story of what is found in the brain but instead is a story of what we project into the brain, as if we were shining our junior batlight into the example looking for bats and triumphantly seeing the shadow of the silhouette we have on the lens.  The account of physiology of the eye and neurophysiology of the visual system has nothing wrong with it but does not include anything about the results being the way things look to us.  If the argument of this paper is right, one consequence is that those brain researchers who are hoping some day to provide the neurological account of how I get a perception of the redwoods when I see them--they can give up that task and go do something less nonsensical.

Now, sometimes there are perceptions.  They occur in particular kinds of cases, cases in which the reliability of what we see is at issue, cases in which there is something problematic about our senses, cases in which the thing seen is peculiar or improbable so we doubt our senses enough that we want to hedge our bets, and some other kinds of cases.  But that it is in those cases that perceptions show up supports the idea that they don't show up in the cases in which, well, they don't show up. We can think that they show up in all cases only by forgetting the kinds of circumstances in which they do show up and by forgetting those details which make sense of their showing up. And careful attention to the other cases in which our rescue attempt needs us to come up with seemings, looks, perceptions or appearances, reveals their conspicuous absence.

Russell, for instance, tells us in Chapter Three that a coin looks oval unless we are looking at it straight on.  In Chapter One, talking about the table which looks oblong in different proportions to those whose points of view around the table are different, he brings up the art student as someone who needs to be aware of these matters because they will endeavour to draw the table, not as it is, but as it looks.

With these remarks as guides and as temptations, we can check against cases.

We can imagine teaching a class of eight- and nine-year-olds how to draw, setting for them various tasks, including a still life with some coins in it.  When one of the students draws circles for the coins though he is off to one side we could tell her, "Now think how the coins appear as you move from directly above them off to one side" [drawing on the blackboard a circle, then a couple or three ellipses of increasing eccentricity, then a double line with milling marks to portray the edge only of the coin] "until you get to a position where you can see only the edge of the coin. From here at your easel, which of these ovals matches best how the coins look?"

When teaching drawing, then, coins look oval when seen from an angle. Here's a case in which coins look oval.

Imagine another case.  You are a police detective for a local community, come back to the University to finish your degree so you can get a promotion.  Coming back to your first meeting of the Intro to Philosophy course after the Spring Break week off, you enter the classroom to find a group of shocked and horrified students standing around the body of the professor, who is lying with his head in a pool of blood, a piece of chalk in his hand resting on the last stroke of the word written in his characteristic block capitals, "COIN," and near his pocket scattered on the concrete floor is his pocketknife and half a dozen coins, some clearly not U.S. in origin.  You call in on your belt transceiver, and the campus cops ask you to do a quick inventory of the crime scene and to get people away to keep it from being disturbed.  As part of this, without getting too close as to obliterate any faint footmarks, you make a note regarding each coin, whether it is resting heads up or tails, what kind it is, and so on. Remember that you will carefully acknowledge any limitations in your point of view--for instance if a couple of the coins are gold in color you will say that rather than that they are gold--or you may say they appear to be gold.  (Though if they are Sacajawea dollars you will probably only note that fact and perhaps leave the color off.)  If one is an old-style British twenty-pence piece you will remark that it is seven-sided, and if one is a penny that has been through the F.A.O. Schwartz souvenir milling machine in San Francisco so that it comes out oval with the name of the store and the city embossed into it you will say so, remarking  that it is oval.  Of a fire-blackened and warped coin whose value or nationality you cannot tell you may say only that, and that it is resting heads up, with a head in profile, indistinctly.

If the professor has told the class that his ex-wife works at a toy store in San Francisco, the oval coin, you think, may turn out to be a clue.  But you will not say, kneeling off to the side a bit and carefully noting every trace of blood, every detail of the coins, his grip on the chalk, that the coins appear oval to you or that they look oval.  You will not say it of the round coins or of the seven-sided 20P or of the oval coin, the penny from F.A.O. Schwartz.

Like the dog that doesn't bark, this result, that you will not say of any of these coins that they look oval, even though one is oval--this result is important to our inquiry.

One more example: Suppose the Denver mint, stamping out the new Sacajawea gold-colored dollars, has a malfunction for a few minutes and suppose that about three thousand of the coins get squeezed side to side in stacks while they are still hot enough to be malleable.  The result is that they are slightly oval.  Very slightly--holding one up and knowing there is a chance it might be oval, it is still difficult to tell, though if you put one into a stack of standard round Sacajawea dollars, you will be able to tell where the sides are indented and where they bulge out, about the thickness of a fingernail.  A majority of the oval dollars, you find out, have by chance come to your town and most have made their way out the change machines at your local post office before the word gets out and collectors get interested.  You regularly visit the post office to mail packages in the evening, and have been tossing the gold-colored dollars into a drawer for the last couple of months. You go check the drawer to find you have six (out of thirty) of the oval coins, worth about twenty dollars apiece now but perhaps more soon. Few locals seem to know, and you begin asking for them in change at the bank and local stores.  You examine each one, but unless you have a round one or an oval one in your pocket you cannot be sure what you have.  "This one looks oval," you say.  By saying so you are saying there is a chance but you are not sure.  After you check, whether it is oval or not, you will not say it looks oval but that it is oval.

We have a case, the art class, in which coins look oval, a case, the murder scene, in which coins do not look oval, and a case, the mint malfunction case, in which coins sometimes look oval but don't after we check even if they are oval.

We can abbreviate other examples.  Perhaps one sign of a certain kind of astigmatism will be that circles, including coins, look oval to you on a ten-o'clock-to-four-o'clock axis.  Some ancient coins were produced in rolling mills rather than produced with die stamps, and so you might as a tourist in Ionia be advised not to automatically pass by the oval ones (not, "Don't automatically pass by the ones that look oval," with its puzzling implication  that you have astigmatism or haven't learned your kindergarten shapes). These add to our store of examples in which coins look oval, and in each case there are particular details of the case which make sense of saying they look oval. Conspicuously, those details have to do with particular cases and not with all cases. Do coins look oval in general, separate from these kinds of details? No they do not. Russell is wrong.

At this point some persons in my audience may wish to defend the view that the coins nevertheless do look oval to me in all the cases in which I see them other than straight on, whether I would say so or not. When asked for argument to support this view, especially in the face of the issue raised by the possibility that they are seeing the result of having dualism taped to the fronts of their flashlights--that is, that they are projecting their dualism and their causal theory of perception into the case, the arguments tend to be of the form spelled out by John Searle under the heading of the "assertion fallacy" in his book Speech Acts or a closely related form spelled out by Paul Grice in his essay "The Causal Theory of Perception" and then later under the rubric of "logical implicature" in essays such as "Logic and Conversation."  These arguments both roughly support the claim that the pronouncements at issue (several examples from ordinary language philosophers are given) need not be said because they are implied by what has been said, and that it is one of the burdens (and turfs) of philosophers that they have the duty to say things about examples which would not be said by those in the examples.  Searle, perhaps because his essay has been shown to be profoundly question-begging or perhaps only out of uneasiness, appears to have backed off his position regarding the assertion fallacy, and Grice's argument fares no better when subjected to critical assessment. (Footnote-- see Frank Ebersole's "Does It Look the Color It Is?" in Things We Know, Fred Mosedale's "On Saying What is Obvious" in Metaphilosophy v. 9 no. 1, and my "Telling the Truth vs. Being Frank." Mosedale points to when we would say that something is implied in the example as a way to distinguish intelligible cases of implying from the cases Searle and Grice deploy.  The Sacajawea dollars which are all gold color and are well known as being gold color and so that you could leave off in your report that they are gold color might be a relevant case for assessing these claims about implying.) The oddity (Searle calls it fishiness) of saying the coin looks oval is not a proof but it is not nothing either. Once the question has been raised whether perceptions exist in exactly those cases in which we ordinarily do not remark on them, it begins to look more likely that their existence is something we conjured up or took for granted without thinking it through, based on our having bought into a Cartesian sketch of human beings, including a sketch of human perception. At the very least, the arguments for such a claim (that coins look oval when we see them other than straight on) need to be articulated and examined.  Arguments based on accepting a dualistic psychology of perception will not work here because they are circular.  Arguments based on Gricean or Searlean accounts of implication and obviousness are question-begging. Others are hard to find.

Another objection to the view that there are no such things as the coin looking oval in most of our seeing coins might go as follows: it is hopeless to use ordinary language as a guide to whether the coin looks oval, because ordinary language is notoriously sloppy and imprecise, and people will say anything.  The proper response to this is to ask for support.  It is, in fact, false that ordinary language is sloppy and imprecise.  If we stick to one particular issue at a time, it will be easy to check.  And, since we already have a particular issue at hand, try it out.  Construct examples, with enough detail that one can tell what sounds nonsensical, in which someone says within the example that a coin looks or coins look oval. [Footnote: O. K. Bouwsma addresses this alleged sloppiness in his essay, "The terms of ordinary language are . . . ." in his Philosophical Essays.]

**cut this? There is a weak biographical argument against the view that coins look oval in all cases of seeing coins obliquely. Some who have looked at the cases and examined their own temptations have realized that their impulse to defend the claim that the coins nevertheless do look oval rests on a conviction that the causal theory of perception and thus the ubiquity of appearances must be right.  For them, taking the issue seriously leads to a feeling of confusion and a realization that they may have to take back an unquestioned philosophical view because their support is circular.

Please note that this is not an argument for the claim that the coins do not look oval. It is just as unintelligible to claim the coins do not look oval as it is to claim that they do.  Unintelligibility, though, has become a fighting term among philosophers to such an extent that the discussants sometimes stop listening to each other.  Here, I've spelled out any claims about unintelligibility in terms which refer to existence instead. That is sloppy and perhaps not adequately excused by a desire to avoid treading on inflamed toes.  My claim that perceptions as philosophers conceive of them do not exist might be better put in a more careful way. The intelligibility of standard philosophical  claims about perceptions rest on an elaborately circular argument, in which arguments for the urgency of the problem of knowledge and the problem of appearances and reality rest on accepting the view for which those arguments are allegedly support.  Let's say that over again: the legitimacy of the problem of appearances and reality depends on our accepting the terms of the problem without argument-which involves taking the conclusion as our support. My objection is that they, perceptions, have to be thought of in such a way that they will exist, not only in those examples in which people would remark on how things look, seem, appear, are perceived, but in precisely those examples wherein we would say no such thing.  Because the problems of appearance and reality, problems of knowledge and problems of how to appraise the dreaming argument and argument from illusion require the existence of perceptions in all cases in which anyone sees anything, it is crucially important that our arguments for such existence be careful and non-circular.

Some more comments toward clarifying the position in this paper.

How does one get from what is said in particular examples to claims about the usually more abstract issues regarding the philosophical problem? The answer is that the problem gives us our temptations and makes relevant the kinds of examples against which that problem may be checked. The conversations in the examples are not directed toward the philosophical problem, of course, and so there may seem to be a kind of built-in prejudice against the problems. But our interest is given by the fact that we are philosophers, in the business of trying to clarify a problem and investigating the first moves toward supposing the problem is justified and supposing the problem rests on clear and unambiguous lines of thought. It is possible for problems to be bogus or to rest on mistakes, and so a basic excavation of the thinking on which the problem rests will be a good idea if we have any suspicion that the problem rests on any confusions or begged questions. Examples not guided by our philosophical thinking may call into question the way we got the question going. That's not a proof the question is a bad question, but articulating how the question can be made sense of without importing other philosophically questionable thinking becomes an urgent need. If the only way to make the problems of appearance and reality into problems involves taking some few examples (the word contexts, usually a mistake, might do here-see below) to be the same as all examples--or if it requires importng a view of human beings which is shaped by Cartesian dualism, then that will necessarily change the issue into one of how to take some few examples for all or how to justify the Cartesian dualism into which we so easily fall.

Context. It might be supposed that this paper is shaped by contextualist theories of meaning. There may be some small part of this idea which is correct, but it is seriously misleading. For one thing,the philosophical notion of context (which may show up in pragmatics in linguistics) is very much at odds with those examples in which we would remark about things said in context or out of context. For another, what counts as a context is a matter of context, and so general claims about meaning depending on context are problematic. Most of the therapeutic value of talk about context is better put in other ways, partly because the whole notion of meaning depending on context imports serious mistakes about how language works, in particular the division between utterances or other signs and what is supposedly conveyed by their means. Those who say things about something (?) meaning or being different in different contexts suffer a genetic predisposition to think they know what the thing is that crosses contexts. Look at the example of teaching children how to draw, in which it makes sense to speak of (that is, within the example we would say such a thing) coins looking oval. From outside the example, we will be strongly tempted to say things like, "The meaning of 'the coins look oval' depends on the context of teaching the kids about drawing." But it will be a difficult matter indeed to ascertain the identity of the thing we think can be carried from one context to another, without imposing that identity onto the examples in a way which would never be remarked in the examples. Try having the teacher suffer from the kind of astimagtism which has the effect of making circles elongated on a ten-o'clock-to-four-o'clock axis, and set up an example in which one might say the thing in quotes is the same in that context as in the other of teaching. What we will find is that all the fierce temptations regarding identity, types and tokens and substance and accidents, ride into the examples upon the backs of the philosophers rather than emerging in the examples. To speak of examples as contexts is to invite the idea that there is a thing which remains the same across contexts. Perhaps saying that what makes sense of a great deal that we say is given us by the details of the examples in which we say it--perhaps that too invites us to think that the "it" is a migratory animal which somehow maintains an identity. Against that is the reminder that it is our temptations which provide that identity--we were investigating whether there is such a thing as coins looking oval when we see them other than straight on, and found that often there is not and that where there is we find no encouragement to think the cases all the same.

One might prophesy where this investigation could end up, given perseverance and luck.  It seems to be pointing toward something like the following: Internal mental states are only sometime things, making the sense they do by occurring in examples which provide in their details the ways to make sense of them and sense of what we would say about them. Some of the cognitive states to which human beings allegedly have constant access might be only occasional--memory, for instance, has nothing to do with my ability to finish a sentence in a way consistent with its beginning, nothing to do with my continuing ability to walk to my classes, nothing to do with my ability to speak and listen, except in the particular cases in which we would claim to remember, or deny remembering, or ask about memory.  Consciousness, which many philosophers think of as ubiquitous among non-sleeping (well, among sleeping but dreaming too) human beings, will prove to be a much less generally prevalent condition when tied to examples in which awareness or consciousness would make sense as part of what we would say.  Studies of our psychology are likely to be more careful and yield much less arrogant pronouncements regarding how the mind works, and mind itself will turn out to be a smaller deal than the mythological beast toward which the Chomskys and Pinkers have been sending their graduate student beaters.

Addition:

The Barn Looks Red

Here's an abbreviation of an attack against several things: against the argument from illusion, against (indirectly) the claim that we represent when we see or see perceptions when we see, and against any of that host of claims associated with the representation theory of mind. (For the non-abbreviated version, see J.L. Austin's Sense and Sensabilia.) It's also meant as an illustration of a therapeutic method in philosophy, namely checking to see if the terms in which we set up our problems or set up the tempting lines of philosophical thought, checking to see whether those terms require us to beg questions. It is meant as an illustration of checking on philosophical temptations by comparing how we talk while tempted with other ways we talk, other ways we have of making sense of things, looking then to examples of nonphilosophically-loaded conversations.

Though of course we have philosophical concerns which lead us to look at these examples or to remind us of their existence, a fair number of people (five or six) have taken to calling these examples nonphilosophical examples or ordinary language examples. A better phrase perhaps would be therapeutic examples. --or just examples.

First, here's the philosophical temptation, in the form of an argument.

The ways things {seem to us/appear to us/are experienced by us/look to us/are perceived by us} could be the same whether they are based on reality or on {illusion/hallucination/afterimages/delusion/dreaming/mirages, etc.};

Therefore, we don't know anything.

This is one of the standard versions of the argument from illusion, maybe the only argument in philosophy. In trying to figure out whether it is a good argument, the procedure I'm recommending here would lead you to back up and worry about the terms, and whether they beg philosophical questions. In this case, they do. The person who gives this argument is already a dualist, or the first line would not make sense. That is, she thinks that whenever we see something there is a difference between the thing we are looking at and the way that it seems-looks-appears etc. Is that true outside of philosophy?

Second, then, is an attempt to think about the philosophical temptation by comparing the way we think while tempted to how we think or talk when we are not.

Consider a pair of contrasting examples (the two are the same up to a line where they branch off from each other):

You are having a garage sale at your house out on the Arcata Bottoms. I'm trying on Saturday morning to find the sale, but I'm being unsuccessful and get lost. I call you up on my cell phone: "I can't find your goddamned garage sale." You: "Well, where are you?" I: "I don't know. There was a sign for some Lanphere-Christensen Dunes, but that was several left turns ago." Y: "Can you see any landmarks?" I: "There's a big, hip-roofed barn across the road, by itself." Y: "What color is it?"

[That's where they branch. So, ]

Example A.

I: "It's red." Y: "Okay. The only painted barn out here is Mr. MacGregor's. You need to do a u-turn, go about a mile to an intersection with a big cluster of mailboxes and a little shelter for kids to wait for the school bus, turn right, . . . ." etc.

Example B.

I: "It looks red." Y: [long pause] "John, have you been dropping acid again? [or] Are you wearing those purple John Lennon sunglasses? [or] Oh, yeah, you're in that lowrider pickup with the dark green-tinted windows--roll down the window. [or] John, are you colorblind and never told me?"

[and then, we hope, Example B rejoins Example A.]

Here come the morals: Our attempt to examine our tempting line of thought suggests the following: The way it looks to me does not arise or make sense in nonphilosophical examples unless there is some oddity about the case that makes it unlike a great many (maybe most of) the cases in which I see barns. In order to think that in Example A there has to be a way the barn looks to me, I have to already be a dualist. The support for the argument works only if the conclusion is true, not otherwise. That's the fallacy known as begging the question.

In other words, it is crucial to the argument from illusion that there be a gap (which means a possible difference) between how things look to us (or how they seem, or how they are perceived, or their appearance, and so on) and how they are. This argument of mine on this page is not for a claim that the gap does not exist, and it's not an argument that the claim that there is such a gap is false. Instead, it is an argument for the claim that the gap does not make sense in cases in which it should make sense, unless we stipulate or take the gap for granted as we begin, rather than argue for it. Again, begging the question. In examples in which we should be able to see and comment on how things look (and we could do this for the other vocabulary--seems, appears or appearance, is perceived) we find that to make sense of that talk the example has to be a particular kind of example and not every example. One basic requirement of the argument from illusion thus evaporates under investigation.

Now, what's the contrast here with, say, philosophical examples? The answer is that there is a philosophical example here, which is example B but without all the ways to make sense of it that you offer over the phone, and without any other ways to make sense of it either. That is, the philosophical example is the one in which the required line does not make sense. This is, of course, an overly combative way for me to put it, but it is at least this possibility that makes philosophers add intelligibility to their list of crucial virtues.

Further addition: What's the alternative to -well, it's called by various names: a causal theory of perception, the existence of appearances/perceptions/seemings/looks/immediate experiences/sense data?

The two examples contrasted, the red barn with it looking red when I see it and the red barn that's red, suggest an alternative that goes like this:

When there's something weird in the examples, then there is room for talk of appearances and perceptions and looks, and so in those cases there is a way to make sense of, let's just call them perceptions. When there's not something that makes sense of that talk, then there are also no perceptions. The desperate move adopted by Paul Grice of trying to legislate that in those cases the talk that goes, "It's red." is actually the same as or implies or takes as obvious that "it looks red" still has the contrast between the two kinds of examples as an argument against it. That was too long a sentence; taking the looks talk as implied or obvious or the same as the talk without looks flies in the face of and contradicts the differences between the two kinds of examples.

To the philosophically inclined, this won't seem an alternative at all. Is the alternative being argued for, against the alternative that when we see we do that seeing through a veil of perceptions or appearances or sense data, that really we can see directly, penetrate right though the veil? NO. There is no veil, either to obscure or to penetrate, except in particular kinds of cases and most emphatically not in all cases. Is the alternative being argued for that the barn's being red is more than how the barn looks? maybe that the barn's being red actually inheres in the object instead of being something in us? NO. If anything, the barn's being red is less than the barn looking red, because the barn's looking red raises or acknowledges the color of the barn as something like an issue, and so something that requires more details or background to make sense of it. And talk of inhering in objects or in us projects into the example the same old categories of material and mental, while we are trying to get a grip on whether that distinction applies at all.

It still won't seem as though we've been offered an alternative to the story about perceptions and the reality over on the other side of those perceptions. Perhaps we should say, partly this is because the alternative is hideously various and complex, (though this is a lie, it is a therapeutic lie that allows philosophers to continue thinking of themselves as doing subtle and difficult work). The alternative involves taking the example of the garage sale and saying the color of the barn, without the perceptions talk, taking that example to be perfectly fine as it stands. We could multiply examples, of both kinds, those in which we see things with no issue about looks or perceptions or appearances and those with those issues, and the message will remain that looks and perceptions etc. do exist in some examples but not in all, not, even, in many. The alternative is that perceptions exist in cases in which there are ways to make sense of them in the cases and do not exist in cases in which all we have to make sense of them is the projected image of our Junior Batlights.

The alternative to the existence of perceptions is their nonexistence. In the face of that alternative, the arguments for their existence look more and more like a house of tissue cards, maintained by carefully directed currents of hot air. Indeed, there's hardly argument at all, only a conviction that there must be such things as perceptions for otherwise we have no account of how we can get by the dreaming argument or the argument from illusion, no account of how we inside here up behind our eyes can have any knowledge of the outside world. And now the jig is up. The argument for the existence of perceptions is an insistence that they must be there in order for these other enormously appealing issues and lines of thought to proceed. But that's only a way of talking that begs to be turned on its head-without perceptions, an edifice with millenia of history erected by brilliant and earnest philosophers at great cost of effort does not so much come crashing down as it just evaporates.

The conviction that we have not offered, really, an alternative, may be because we cannot shake a conviction that the problem is a legitimate problem-the problem being the problem of appearances and reality, and the problem being that to which the story about perceptions was meant as answer or part of an answer. But the existence of that problem depends on accepting the account that was supposed to be a solution to it, namely perceptions. We are in a labyrinth and we went by this intersection before. How did we get here? Let's retrace our steps.

Notes and Biblography

1. We could find formulations of the argument from illusion by heaving rocks in philosophers' offices and then picking up whatever's under them. Descartes' First Meditation will do very well, and the dreaming argument is a powerful opening wedge for the more general argument discussed here. In our own time, A.J. Ayer provided a much-discussed version in his Foundations of Empirical Knowledge, (and later in The Problem of Knowledge) and J.L. Austin's Sense and Sensabilia is a relentless ripping of Ayer limb from limb. Ted Honderich's "Seeing Qualia and Positing the World" makes clear that the debates about qualia are new wine in old bottles, and that the argument from illusion is the key to disguised new versions of debates about subjective or private objects involved in perception (Honderich's essay is in A. J. Ayer: Memorial Essays, ed. P. Griffiths [Cambridge U.P.: 1991]). Current talk about the closure principle might seem promising, but when applied to skepticism always begins after the first premise of the argument from illusion is stated as obvious: here we are with these experiences, and which hypothesis shall we accept as better explanation, an evil genius or brain in a vat story, or that there is a world?

2. While this paper is mainly a record of my own thinking, battles against the existence of percepts or sense-data or perceptions or qualia are not at all new. Sellars has famously written about the "myth of the given." J.L. Austin's work is mentioned above, as well as Honderich's essay, which has a summary outline of the issue. Wittgenstein's On Certainty, which I regard as a first unfinished draft record of W's working toward a goal he could not himself see clearly, shows a struggle with methods which would lead to the results I give. Frank Ebersole takes up issues about perceptions in several essays-"Feeling Eggs and Pains," "The Objects of Perceptions and Dreams," "And Then I See," and "Does It Look the Color It Is?" I'm not sure, when I remind myself of these other thinkers, that I am saying anything new except that it's still an issue and that philosophers have not properly listened.