Perceptual Knowledge

The thesis to be defended here is that perception is knowledge and the most basic kind of knowledge.[1] This knowledge has nothing essentially to do with belief, except by way of repudiation. Perceiving is knowing and knowing is perceiving (more or less). In practice this comes down to the thesis that knowing and seeing are interchangeable—two names for the same thing. Knowing is seeing a fact not believing a proposition. This seeing is foundational and constitutive. I begin with an anecdote: I was sitting in a restaurant in Barbados (“La Luna”) and a bird flew over and perched on a nearby ledge; it seemed interested in something, but it was hard to know what. It sagely kept its distance, never getting too close; if I moved towards it, it moved away. Eventually it flew onto the table, obviously knowing the risk. At an opportune moment it did something remarkable: it flew quickly to the sugar tray and snatched a packet of sugar, instantly flying away with its haul. About this clever bird I would say that two things are obvious: one, it had knowledge, lots of it; and two, it had no beliefs. It knew but it didn’t believe. It knew because it could see what was going on; its seeing was a case of knowing. It didn’t need any additional belief state—it could simply see that there was a packet of sugar on the table. I doubt it had ever had a belief in its life (more on this later). The seeing was not just a matter of passively receiving a sense-datum; its visual system was tightly hooked into its motor system—the executive branch. Clearly, it had learned this trick from other birds, presumably by imitation; it had a learning history, a background of memory. The whole performance was a demonstration of avian knowledge. The bird was consciously aware of my presence and of the desired sugar: it had a kind of direct acquaintance with the objects and facts in question. This awareness constituted its knowledge, suitably embedded in the animal’s sensorimotor system and learning history. It had no language, no concepts, no opinions, no beliefs—but it did have knowledge. It knew in the plainest and least metaphorical sense. Its perceptual-executive system gave it knowledge—useful, actionable knowledge. No doubt this capacity had a long evolutionary history: perceptual knowledge is a very useful adaptation. This bird trusted its eyes implicitly to give it the necessary information; it didn’t just think there was sugar there, or that I was potentially dangerous. As we say, it knew, without a doubt. Knowledge of this kind is a primitive automatic response to sensory contact, though obviously sophisticated. It is not a matter of considered belief, evaluative justification, or careful deliberation. Nor does it arise by testimony. It comes straight from the senses. There is no such thing as perception without knowledge; an identity theory would seem to be appropriate. People like to say “Seeing is believing”; that is obviously false, but it is not false to say “Seeing is knowing”. We often say “I saw it with my own eyes” to indicate well-founded knowledge, and we are not wrong to do so. Yet recent epistemology has been obsessed with knowledge as true justified belief (“propositional knowledge”), neglecting the kind of sensory knowledge I am drawing attention to. In fact, there would be no knowledge unless this kind of primitive belief-independent knowledge existed. We must not intellectualize knowledge; it would be completely wrong to say that the perception that leads to it is a species of opinion, judgment, surmise. Perceptual awareness is knowledge.

It is a kind of empiricism to say that all knowledge rests on perception, where perception is construed in the pre-conceptual way I am recommending—direct consciousness of facts. But we should not take this to exclude rationalist epistemology: for there is room for the idea of direct rational (in)sight, say of logical connections. Similarly for mathematical and ethical knowledge. You can see that certain things have to be so, rationally. This is a perfectly natural way to talk, and quite unexceptionable. It might even be true that we use this intellectual notion of seeing in understanding the nature of vision by means of the eyes: we often see by seeing. Someone draws a diagram on the board and looking on we exclaim, “Oh, I see”. I myself often see (understand) by seeing (directing my eyes)—say, in watching a tennis demonstration. The eyes and the intellect both see and often work together to do so. So, rationalist epistemology can be perfectly empiricist in the sense that it accepts the foundational role of episodes of seeing. The important point is that perceptual knowing is basic epistemologically—and is not a type of belief. In fact, it wants as little to do with belief as possible, as we shall see. What this means is that what is often called “knowledge by acquaintance” is epistemically basic; but we should note that the notion is broader than is often supposed. For it covers all kinds of entities: mental entities, physical objects, qualities, facts, events, theories, necessities—anything that we can be said to perceive. It is not what we are acquainted with that counts but the acquaintance relation itself. Nor should we be deterred by the difficulty of understanding the nature of this kind of knowledge, involving consciousness and intentionality as it does. William James says at one point, “Now the relation of knowing is the most mysterious thing in the world. If we ask how one thing can know another we are led into the heart of Erkenntnisstheorie and metaphysics”. (The Principles of Psychology, p.216). But mystery is no reason for rejection, so we shouldn’t be suspicious of perceptual knowledge for this reason. Okay, all our knowledge rests on mysterious foundations—so what! Facts are facts, even if the facts are difficult to comprehend. Neither should we be concerned about the complexity of sensory processes, or their occasional fallibility (nothing is infallible). What matters is that knowledge by acquaintance is a superior type of knowledge, providing immediate contact with the thing known; it can’t be acquired in any other way (say, by testimony). Acquaintance is as good as it gets, the gold standard (even gold varies in value). If you could know everything by acquaintance, you would leap at the chance—so much better than mere description! God knows everything this way—he sees everything (no conjecture or inference). We feel frustrated that not everything can be perceived; life would be a lot simpler that way (consider other minds). We dream we could see into every nook and cranny of the universe, especially the future. Perceptual knowledge is good. If you see, you don’t need to infer. Seeing is your first resort. Carnal knowledge is sense-based not inferred, and generally thought desirable. Seeing is the epistemic ideal.

We must also note some further special features of perceptual knowledge: informational richness and selective attention, particularly. A lot comes in visually, far more than in a verbal report. So, we derive a great deal of acquaintance knowledge from a momentary perceptual encounter. We can also track an object with our eyes, learning more about it as time passes. Seeing is a very economical way to gain knowledge, much faster than reading about something. In addition, we can attend to certain portions of the perceived environment, thereby acquiring fine-grained knowledge; this makes our knowledge useful and relevant to our concerns. Seeing is a mighty faculty epistemically. We often notice things we didn’t expect, or see things we weren’t supposed to (where would spies be without vision?). Sight is really a lot better than blindsight—or telepathy, intuition, and gut feelings. It makes you a superior class of knower. Moreover, perceptual knowledge possesses a certain kind of generality: not only do you gain knowledge of a token state of affairs; you also gain knowledge of its type, or many types. Properties are exemplified in the token that apply to other tokens, actual and possible. If you’ve seen one cat, you’ve seen them all—you know what a cat is. So, your knowledge goes well beyond the particular token you are now perceiving; it generalizes. You see the general features that characterize a cat. You are not confined to knowledge of the token in question; you become acquainted with much more. But it must be admitted that perceptual knowledge is limited—we only perceive certain parts of the world. We need to go beyond this; the question is whether we can knowingly do so. I will consider that question in the next paper. For now, we have a tolerably clear idea of the nature of perceptual knowledge and why it deserves to be so described. It is indubitably knowledge, if anything is.[2]

[1] Again, I am indebted to Michel Ayers’ bracingly unorthodox (but commonsensical) work in what follows, particularly his Knowing and Seeing (2019).

[2] Ayers calls perceptual knowledge “primary knowledge” as distinct from the “secondary knowledge” that comprises knowledge by inference from primary perceptual knowledge. The question I will be concerned with is whether so-called secondary knowledge is really knowledge—really and truly.

Language, Self, and Substance

I will offer some sketchy remarks on meaning and the self in the light of the anti-substantialist view of the mind. First, there has to be something wrong with the Cogito as traditionally conceived, since the self (reference of “I”) is not a substance. We can’t say, “I think, therefore I exist as a substance”: the meaning of “I exist” cannot be that a certain substance exists, because nothing mental is a substance. The word “exists” here can’t be functioning as a predicate of a substance, and “I” can’t be a singular term denoting a substance; for then the sentence would be meaningless for lack of reference. What it does mean is obscure. Second, physical substances contribute to the meaning of sentences about physical objects, but so do non-substantial states of consciousness, since grasp of meaning implicates the ontology of consciousness. Meaning must be a combination of substance and non-substance, a peculiar hybrid. We are familiar with sense and reference; well, this is absence of substance and presence of substance. Meaning is going to be something rather special given its ontological underpinnings—a juxtaposition of external substance and internal lack of substance. It has an intelligible ontology and an unintelligible (to us) ontology at the same time. Third, sentences about the mind will find themselves in an awkward position: they are not about anything substantial, so their meaning cannot be fixed by the substance denoted. How then can they have meaning? Not in the way physical sentences do. They are about non-substance and also grasped by non-substance; so, they must be semantically quite different from sentences about substantial things. How can they have meaning? Must they have a pure use type of meaning, while sentences about physical things have a denotational substantial meaning? Fourth, the self cannot be a substance, so its nature and persistence through time cannot be like the nature and persistence of substances proper. Perhaps we need to look with more favor on etiolated psychological continuity theories, or admit complete bafflement. Nor can the word “I” function as a substance-denoting word: there is no such substance to be denoted; and it is unclear what else might be its denotation, if any. In sum, the elusive ontological status of the mind poses problems for standard theories. A semantics based on substance, such as our ordinary physical sentences demand, is inapplicable to psychological sentences, but nothing else suggests itself. Yet the sentences look very much the same. There is a real threat that psychological sentences can have no genuine truth-conditional meaning. A psychology without substances looks like a psychology that cannot be talked about. How can there be a science of such a thing? There can be no doubt that external substances play a formative role in the creation of meaning, but if the mind has no substantial ontology, it cannot play the same kind of role—so mental language ought not to be meaningful at all. This is a lot worse than indeterminacy of meaning, because now there are no (mental) rabbits to talk about, i.e., substance-like mental entities. Mental talk has no articulable subject-matter.[1]

[1] I feel paradigms shifting beneath my feet. Have we been complacently assuming a substance ontology for the mind in our theorizing about language and thought? What if we gave that up?

What Makes Consciousness Mysterious?

Today it would be widely agreed that consciousness is mysterious, rather mysterious or extremely mysterious. It would not, however, be widely agreed what makes it mysterious—what precise characteristic confers the mystery. Some would say there is no mystery at all: consciousness is nothing but higher-order thought or self-ascription or not being asleep or activity in the reticular formation. I won’t discuss these views, as they strike most of us as non-starters. More promising, we have the ideas of intentionality, privacy, incorrigibility, and non-spatiality. These don’t seem inherently mysterious, however, though it may be that it is mysterious how the brain contrives to produce them. I don’t think many people look within, find these characteristics, and think, “Wow, that’s so mysterious!” So, granted that consciousness is mysterious, it is a bit of a mystery what makes it so—a mysterious mystery, as we might say. We can say what makes gravity a mystery, or dark matter, or the workings of black holes; but we find it difficult to identify the source of the felt mystery in the case of consciousness. It is certainly close at hand and not afraid to present itself, but it is obscure what makes it stand out as a mystery—how it differs from other natural phenomena in the mystery sweepstakes, particularly the brain.

Here is one popular answer: consciousness is subjective while other non-mysterious things are objective.[1] To be more specific, it has a peculiar epistemic property, viz. that it can only be known by beings that share its particular character. It is mysterious because it has this property (the brain doesn’t). We tacitly recognize that it is epistemically restricted in this way, so we deem it mysterious. It isn’t simply that it has the what-it’s-like property—why exactly is that a mystery? It’s that this property gives rise to the peculiar epistemic situation in which we find ourselves—knowing what it’s like to be human and not knowing what it’s like to be a bat. However, I don’t think this is plausible as an explanation of our sense of mystery with respect to consciousness. First, is this really how consciousness immediately strikes us when we sense its mystery? Isn’t it rather an ingenious (though correct) point about the epistemology of consciousness? The point might never have occurred to you (it takes some arguing for) even though you have a primitive sense of mystery about consciousness. It seems too surprising to be the explanation we are looking for. Second, what if we suffered from no such epistemic limitation—would we feel no mystery?  Suppose we happened to have a mechanism in our brain that reliably produced the knowledge in question (bats and all): would consciousness then seem devoid of mystery? Doubtful. Third, what if we developed an objective phenomenology that enabled us to comprehend any type of experience no matter how remote from our own? Again, would the sense of mystery then disappear? Would consciousness no longer seem like a thing set apart, a metaphysical oddity, a natural wonder? No, we don’t seem to have put our finger on what exactly gives rise to the feeling in question. Isn’t there something more intrinsic and irremediable about the mystery of consciousness? The epistemological point seems too extrinsic and contingent. So, the mystery of the mystery remains—we haven’t been able to specify what it is about consciousness that makes it so mysterious. And this is a problem, because then we are defenseless against the claim that there is really nothing mysterious going on—we have a false sense of mystery. We don’t understand what makes consciousness an especially intractable problem if we can’t say whence the impression of mystery arises. We might expect that answering this question will reveal something deep about consciousness and our conception of it. I will attempt to answer the question in the sequel.

[1] See Thomas Nagel, “What is it Like to be a Bat?”

An Argument Against Idealism

Imagine we lived in a world in which idealism was the dominant philosophy (in fact, it is the actual world, but that’s another story). The prevailing doctrine is that everything that exists is mental, i.e., a state of consciousness. Not just the mind itself but also the so-called physical world—mountains, molecules, the brain. All these things consist of ideas in the mind, specifically sense experiences of certain kinds—for example, episodes of seeing the thing in question. To be a brain, say, is to be a sense experience as of a brain. Physical objects are really collections of sense-data, as the terminology has it. These sense-data may exist in the human mind or God’s mind or just in my mind (solipsistic idealism]. Physical objects are reducible to such mental entities; this is their essence, their mode of being. Everything has a mental nature.

But suppose there is opposition to such a doctrine: some people firmly believe, eccentrically, that some things are inherently not conscious states; their essence is to be not-conscious, mind-independent. Their battle-cry is, “The not-conscious world is what makes idealism metaphysically impossible”. They reject the claim that physical objects are reducible to sense experiences. The idealists wonder what else these objects could be: it would have to be something unknown, of an unfamiliar nature, a mysterious substance of some sort, subject to skepticism. It would have to be something there is nothing it is like to be, but what would that be? Everything we really know is like something—visual experiences, pains, etc. They have trouble getting their minds around this supposed non-mental reality: how can we even think of it? However, the anti-idealists have an argument that they believe can break this stalemate of intuitions and settle the question once and for all. It is quite an ingenious argument, centering on a conceptual claim. They point out that everything conscious is such that there is something it is like to be it: therefore, it is only possible to grasp a given type of mind by sharing it. You only know what it is like to be a human by having human experiences—it can’t be done from no “point of view”. It takes one to know one; you can’t (fully) know what is like to be a bat because you aren’t one. This property of consciousness they call subjectivity: you can only form a concept of another’s mental state if you are a subject similar to the other subjectively—if you are mental birds of a feather, so to speak. But this is not true of physical states: these you can grasp even if you don’t share them—you don’t need to be a mountain to know what a mountain is; you don’t need to share a bat’s brain to know what a bat’s brain is. In short, no point of view is built into physical concepts: they can be grasped from any point of view. You can do physics if you are human, Martian, or Vulcan—or any species with the requisite intelligence. You can do geometry even if you are not yourself similarly geometrical; you don’t need to be Euclidian in order to understand Euclidian geometry, say. The physical may be defined as whatever there is nothing it is like to be, so there is no like that you must share in order to grasp it. But then, it is not possible to reduce the physical to the mental, because these are concepts of a different order, denoting properties of different kinds. It would not be possible, say, to reduce the physical brain to sensations as of brains, because such sensations would embody a distinctive subjective quality graspable only by beings that have similar sensations; but that is not true of the brain itself, because that can be grasped by beings with arbitrarily different types of sensations. Thus, there must be more to reality than is contained in sensations of reality—the physical cannot be the mental. Physical things are essentially objective: they can be grasped from many points of view and therefore cannot be subjective in nature. Things there is nothing it’s like to be cannot be reduced to things there is something it’s like to be. Therefore, idealism must be false: it tries to explain the objective in terms of the subjective. It cannot be an accurate account of things that lack consciousness, precisely because it explains everything in terms ofconsciousness. The body cannot be explained in terms of the mind, on pain of subjectivizing the body.

Clearly, this argument parallels a familiar argument against reducing the mind to the body; it simply reverses that argument. If the familiar argument is valid, then so is this one. Thus, that argument defeats bothmaterialism and idealism in one fell swoop. It can be used against Hobbes but also against Berkeley. It can be used against physical anti-realism as much as against reductive physicalism. Basically, the argument is that a color-blind man cannot understand color vision but he can understand the brain science of color vision (and the rest of physics). For if he could not, the brain would not be a physical object, which it is. The objective cannot be reduced to the subjective, as the subjective cannot be reduced to the objective. The view from somewhere cannot be reduced to the view from nowhere, and the view from nowhere cannot be reduced to the view from somewhere.[1]

[1] Is this argument perhaps just a little bit too powerful?