Knowledge, Memory, and Time



Knowledge, Memory, and Time


We have knowledge of the past, the present, and the future—or we think we do anyway. That makes three types of knowledge, classified according to time. But when you think about it you see that there is really only one type of knowledge: knowledge of the past.  [1] Knowledge of the future is based on knowledge of the past; it has no other source. We know what has happened so far and we project this knowledge into the future, relying on inductive inference. Knowledge of the future is simply knowledge of the past extrapolated forward: if there were no knowledge of the past, there would be no knowledge of the future. Knowledge of the present, so-called, is also really knowledge of the past: for it takes a finite time for a stimulus to cause a percept, and it takes a further finite time to form a belief based on a percept. By the time you have formed a belief about what is presently the case it may no longer be the case. This time lag is notoriously writ large in the case of distant astronomical objects: by the time you have perceived and formed a belief about such objects they may be long gone. Our knowledge is always of the past, though we may project it forward by making present-tense judgments, hoping that things haven’t changed too much in the interim. What we know is always strictly about what has been (this is true even for introspective knowledge, since it too takes time to formulate). In the most basic cases our knowledge is a trace of the past—an effect of a prior cause. In fact, there is always a complex sequence of events mediating the outside stimulus and the knowledge it eventually produces—events that are both computational and physiological, occurring in the nervous system. These function like memories of earlier events in that they involve holding information for later processing: the retinal image is a kind of memory of the light pattern that reaches the eye, and subsequent neural patterns are like memories of the retinal image. There is information storage over time, though the time might be measured in milliseconds.

Given that memory is the retention of information over time, we could say that all knowledge, including present-tense knowledge, is based on memory. Indeed, the light pattern that reaches the eye from a distant (and possibly extinct) galaxy might profitably be viewed as itself a type of memory of the galaxy in question: for the light preserves information about that galaxy over time, possibly for many years. The universe remembers! If we were to go externalist about memory, we could include this kind of information retention in the mechanisms of memory: the light preserves information about a remote object and then the nervous system of the observer preserves information contained in the light. In any case we have information preservation over time, the mark of memory. We can simplify matters theoretically by stipulating that the process of belief formation elicited by past events is a memory process: information retention is always a type of remembering. Even when an object is right in front of you now there is a complex memory process as information is gleaned, processed, and stored—eventually leading you to form a belief about the object (“This is red”). Subjectively it all seems to happen in an instant, but in fact there is always a time lag between the object having the property and the observer coming to know it has that property. If we were being very careful (the skeptic looking over our shoulder), we would more cautiously assert only that the object was red, albeit a fraction of a second ago. The sensory modules inside us, which have a much more fine-grained experience of time, might confine themselves to judgments along the lines of “I seem to recall that such and such an event happened in the recent past”. By the time the light stimulus makes itself felt in the visual cortex at the back of the head, after a long trip through the optic nerve and the brain, the retinal image might be a distant memory as far as our visual processors are concerned. The essential point is that information retention is the method whereby knowledge of the present is acquired—that is, knowledge of the past. In short, all knowledge is memory knowledge. When the positivists spoke of “observation statements” as forming the epistemic bedrock of science (and not just the positivists), they misspoke: it is memories of observation that form the basis of scientific knowledge. Observation itself is essentially a memory process, but in addition the observation must be retained over time, in memory and by means of external record (e.g. writing). The real basis of scientific knowledge is memory: no memory, no science. Everything we know about the world must be contained in what we remember about it. Our knowledge of the future and the present is really nothing but our knowledge of the past plus some inferential moves (more or less dubious according to the skeptic).

            The question must then arise as to how good memory is as a source of knowledge of the world. Does memory tell us how the world really is? Memory is, of course, notoriously unreliable, subject to distortion, fleeting and fallible. It needs some kind of external check if its deliverances are to be credited—and the checks too need somehow to be remembered. But that is not the point I want to emphasize: I want to focus on a deep structural feature of memory—what memory constitutively is. This feature is somewhat elusive to characterize, though we all recognize what it is; it tends to be described metaphorically or impressionistically. Hume spoke of memory images as “faint copies” of sensory impressions, which are vivid in comparison—as if memories are faded and worn, indistinct and blurry—and this description carries intuitive appeal. We might say that memories are impoverished or etiolated or drained of substance or sketchy or fragmentary or thin or insubstantial. Marcel Proust is the poet of memory as a feeble route to the past: it simply cannot recreate the psychological effect of past experience itself (this is why we are in search of lost time).  [2] Only the madeleine can bring back the past, enabling us to relive it, and even that cannot really recall the past experience in all its glory. Time past is time irrecoverable. Let me put the point this way: memory cannot provide us with a mode of apprehending the world that captures its full reality. Memories of the past are glancing, stripped down, and partial–ghostly remnants of what they purport to record. They are mere traces, like footprints in the sand: they don’t afford us the kind of full brimming knowledge of the past for which we yearn. They are not like seeing or touching or tasting something: remembering a taste is not like tasting a taste. It is true that some people reportedly experience the past in this quasi-perceptual way, and normal people can sometimes have remarkably vivid memories, but generally memory is a feeble conduit for knowledge of the past. It is normally pretty thin gruel. And once we appreciate the role of memory in perceptual knowledge we start to wonder how much is omitted even there: maybe a lot of information is lost between stimulus and response—even Humean impressions might be drastically etiolated compared to reality itself. If memory images are faint compared to sensory impressions, sensory impressions might be faint compared to the objective world; indeed they presumably are, given that enormous amounts of information don’t even get beyond the initial sensory receptors (the retina is by no means locally omniscient). We can conceive of creatures that have super vivid sense impressions, compared to which ours are pale and weak. In any case human memory is just not very saturated or all encompassing. And if memory is limited in this way, our knowledge must be too, proceeding as it does from such a bleached out basis. Our conception of reality is formed from our memories of it when you get right down to it, but memories are just not very rich or full or penetrating. Just think of your general conception of the past: it is shadowy, reduced, distant, full of blanks, and skeletal–even when operating at its most efficient. You really can’t think of the past as it is in itself (this was Proust’s lifelong lament). The past is gone, epistemically speaking; it is fundamentally irrecoverable, even when memory is accurate. The past is a shadow world (Plato should have used the parable of the cave to illustrate the feebleness of memory). But our conception of the world derives from memory and inherits its limitations. You might suppose that perceptual experience offers a way to something more vivid—more real—but don’t forget that experiences themselves become objects of remembering. Memories of experiences are themselves weak and faint, nothing like having the experience (and even that suffers from time lags and limited information retention). Fundamentally, our whole picture of the world is conditioned by our powers of memory, but memory is ineluctably Proustian—it can only offer us so much lost time.

Our epistemic relation to time is all out of whack: we can’t grasp time while it is happening, so to speak. Time is not obliging to human knowledge—not the past, not the present, not the future. We are always late to the party (or much too early in the case of the future), and poorly dressed and equipped. Time and human knowledge are just not cut out for each other. The result is that we view reality through a misty portal, a narrow and smoky window. This means that we never really grasp reality as it is, de re as it were; at best we form ideas that map its structure, or we feel its effects on our consciousness. God doesn’t remember the past, as if at a distance from it; he perceives the past—feels it in his marrow (or the divine equivalent). Think of an animal’s memory-mediated conception of reality, say a mongoose’s: that must be thin gruel indeed. Fortunately the animal is equipped with a bunch of instincts and doesn’t need to rely on abstract knowledge; we on the other hand fancy ourselves endowed with intellectual apprehension, while actually laboring under the constitutive limitations of memory. Memory is frustratingly flimsy, a perpetual source of existential angst, a Proustian nightmare. It all comes down to memory for us, but memory is very far from all there is. Reality itself has no truck with memory limitations and won’t agree to be represented by memory. Reality has all the talent and its agent memory is but a weak and impoverished echo of it. The past is always much more than memory can ever know.  [3]                             


  [1] I will put aside questions of innate knowledge; our concern here is with acquired empirical knowledge. A prioriknowledge is not temporally bound and is not founded in memory, but empirical knowledge of the world needs memory-assisted sense perception to get off the ground.

  [2] There is a good discussion of this philosophical theme in Proust (who was influenced by Bergson) in Nabokov’s Lectures on Literature (1980). Nabokov writes: “To recreate the past something other than the operation of memory must happen: there must be a combination of a present sensation (especially taste, smell, touch, sound) with a recollection, a remembrance, of the sensuous past… In other words, a nosegay of the senses in the present and the vision of an event or sensation in the past, this is when sense and memory come together and lost time is found again.” (p.249) I would add that such recreation can never be complete. Memory alone cannot do the job, so we need to supplement it with the more vital and full-blooded force of perception; but this combination is still impotent to produce the impact of direct past experience. Thus we search in vain for lost time: the past comes to us only in hints and morsels, despite that magical madeleine. The only way truly to recover the past is to travel back to it—and then we are still stuck with the memory-like nature of perception.  

  [3] Here we might speak, not so much of cognitive closure, as of cognitive claustrophobia—the feeling of being squeezed into a space too confined, cognitively speaking. This is the prison of memory: the way knowledge is held in check by the very nature of the memory faculty. And it is not just a matter of what is remembered but of the wayremembering works—that faintness, that etiolation, that fatal thinness. The world for us is a remembered world, but remembering is subject to extremes of deletion and dilution. The veil of remembering (as opposed to the veil of perception) is an apt metaphor: the past hides coyly behind this veil, showing itself only sketchily and dimly. This must condition the way we think of reality.


Fact and Value


Fact and Value


It seems to be commonly assumed that fact and value belong to separate categories: facts are not values and values are not facts. We might call this Humean dualism. The two categories are held to be exclusive and exhaustive. But is the dualism really so clear-cut? First, there is the question of how homogeneous each of the alleged categories is (the same question applies to Cartesian dualism): aren’t there many types of fact and many types of value? Maybe a more pluralist view would be better.  [1] But second, aren’t facts valuable and values factual? Facts can be good, bad, or indifferent—they are not value-neutral. Pleasure is good, pain is bad, and hedonic neutrality is a matter of indifference. Every fact can in principle bear on our desires, wishes, emotions, and judgments of desirability—not to mention our aesthetic values. The fact that there is oxygen in the earth’s atmosphere, for example, has value so far as terrestrial animals are concerned. Also, values are themselves matters of fact: it is a fact that pleasure is good, say, and genocide is bad. Don’t say that such judgments are subjective or relative or mere projections: that doesn’t render them non-factual, since there are subjective, relative, or projected facts (take colors). Facts don’t need to be objective in order to be facts (not that we should accept these subjectivist views of value). Even tastes and emotions are factual (it’s a fact that blue cheese tastes good to me).  [2] The truth is that statements about value, like other statements, are statements of fact—they are true or false, believable or not believable. What we really have is a dichotomy between non-value statements of fact and value statements of fact. So all (non-value) facts have a value aspect and all values are species of fact. The world isn’t divided into a set of value-free facts and a parallel set of non-factual values: value facts are among the facts of the world, and the non-value facts are value-imbued. It is true that reality contains two sorts of property—value properties and non-value properties—and we have vocabulary suited to both sorts of property; but this implies no grand metaphysical division between fact and value. Facts are not intrinsically opposed to values, and values are not prohibited from belonging to the fact club. It would be better to say that everything in the world is a combination of value properties and non-value properties (compare certain views of the mental and the physical). The case is analogous to primary and secondary qualities: there are indeed two types of property, but it would be wrong to express this by saying that primary qualities correspond to facts while secondary qualities don’t, or that primary qualities can exist without accompanying secondary qualities. Primary qualities come with associated secondary qualities, and both are perfectly factual (even if secondary qualities are subjective or relative or some such). The shape-color distinction is not a distinction between the factual and the non-factual. Likewise goodness is a property belonging to facts in the sense that ascriptions of goodness are factual, and insofar as non-moral facts have a moral aspect (or other evaluative aspect). The world is a combination of value properties and non-value properties, as it is a combination of primary and secondary properties. There are not two worlds, one of fact and the other of value. These distinctions should not be inflated beyond their proper dimensions.  [3]


  [1] Values fall into several subspecies: moral, aesthetic, practical, taste, whim, caprice, existential, etc. It would be wrong to assume that all these values form a homogeneous group; in particular, moral values are really nothing like food preferences or an appetite for fast living. It is the same for the mental side of traditional mind-body dualism: what we call “the mind” includes sensations, beliefs, moods, wishes, linguistic understanding, character traits, etc. We shouldn’t let a high-level dualism blind us to deep differences within a broad category.

  [2] I am not considering such doctrines as emotivism and prescriptivism, which assume a highly tendentious conception of the factual and have little basis in the actual semantics of moral sentences. Moral sentences have the grammar of truth-bearing propositions (you can easily slide them into a Tarski-style biconditional).

  [3] I am reminded of warnings on tires not to overinflate. Philosophy has a checkered history of over-inflation, forever pumping up distinctions beyond their proper dimensions.


Is Existence Possible?



Is Existence Possible?


I am going to consider an argument for the conclusion that existence is not possible. Not that we don’t know that anything exists, or that nothing in fact exists, but that existence is impossible: it could not be that anything exists. This is, to put it mildly, a startling conclusion—startling to the point of jaw dropping. Yet it follows from relatively simple assumptions and requires no fancy or revisionary analysis of what “exists” means. The argument is not apodictic, but it is certainly worrying; it isn’t easy to see where it goes wrong, if anywhere. It starts with a fairly innocuous observation, namely that when new things come to exist they do so in virtue of pre-existing things, generally by being made of these things. Thus a new house is made of pre-existing bricks, a new painting is composed of paint that existed before, and a new organism is made of bits of matter that were already around. Nothing is truly new; everything is derived from something else.  [1] Let us say that things generally have a derivative existence. We can usually explain the existence of a new entity by reference to old entities that pre-date it: it is typically a matter of re-combination. Sometimes this procedure doesn’t work because there is (apparently) too much novelty in the new entity: notoriously, sentient beings have an existence not easily explained by the existence of what preceded them (insentient matter); similarly for organic beings in relation to pre-organic material. But here we generally assume that we are missing something, and it is clear that some aspects of the new entity owe their existence to pre-existing entities (such as the constituents of the body). Existing things come from other existing things, often going back a long time. Nearly all of what we observe in the way of existence is derivative existence.

            But isn’t there also original existence—things that exist without reliance on other things? What about the parts of newly existing things—where do they come from? They may in turn come from other pre-existing things, ultimately going back to molecules and atoms; and these entities may owe their existence to even earlier realities, given what we hear about the state of the universe at the time of the big bang. Maybe atoms derive from superhot plasma as it cools. But at some point we reach things that don’t derive their existence from other things: what should we say about these? Now matters are apt to turn sticky; a certain intellectual panic sets in. Our usual paradigm of existence starts to break down, since these entities don’t have an existence that can be explained in terms of antecedent entities. An array of more or less unpalatable options presents itself. The first is that these entities come from nothing at all: they simply spring into existence de novo. Not even God plays a role, since he is an existent being (allegedly) for whom the same question arises: what explains God’s existence? But putting God aside, we have the idea that the basic things of reality pop into existence from pure nothingness. This seems utterly incomprehensible: as the old adage goes, nothing comes from nothing. And if such a miracle were possible, why isn’t it still happening—why don’t we see things popping into existence out of nothing on a regular basis? It might seem that a dose of modal metaphysics could get us out of this jam: what precedes existence and provides its foundation is possibility. We might even change the terms of the discussion and describe the antecedent possibilities as themselves existing, so that the question becomes how possibilities become actualities. This is certainly an intriguing metaphysical theory: pre-existing possibilities give rise to the existing actual universe that we observe—not something from nothing but something from a possible something. There was a possible world existing before the actual world and it is the basis for the existence of the actual world; we just need to pump up our conception of reality and then we can find the basis for actual existence, thus preserving our usual paradigm of derivative existence. There are (at least) two problems with this approach. The first is that the same question will arise for the antecedent realm of existence: where do these possibilities come from? If possible worlds really exist, where do they come from—do they pop into existence from nothing? But second, and more decisive, possibilities have no tendency to turn into actualities; so they cannot play the role of existence generators. We can’t say that actually existing things are made of possibilities that pre-date them—what would that even mean? The relation between existence and possibility is nothing like the relation between a thing and its parts. This would really be another version of the something-from-nothing approach: things come to exist of their own volition, so to speak, without any earlier preparation or precursor or preamble—they just burst onto the scene at some assigned or random time.

            The obvious next move is to declare eternality: the original existences are nothing like the derivative existences, which have a time of origin and a finite lifespan, but are eternal beings without beginning or end. Thus we might suppose that elementary particles are eternal beings for which the question of origin does not arise. This is no doubt a tempting move (a knight’s move), given the unfolding dialectic, but it is far from satisfactory. First, it is flagrantly ad hoc: we are forced to make a clean break from our usual conception of existence in order to solve a metaphysical puzzle. We have no other reason to postulate eternality for non-derivative existences than to solve the problem of how certain things manage to exist; and on the face of it the idea sounds preposterous. It isn’t that these entities are like numbers in being (arguably) necessary existences: they exist contingently but eternally. So it is not in their very nature as necessary beings to exist in all possible worlds at all times; no, they are contingent beings that just happen to exist for all time. Nor is it clear that this postulation would solve the basic problem: for isn’t there still the question of what explains the existence of these entities? Let’s assume they exist eternally: we still need to know why they exist at all, given that they don’t have to. Is their existence simply a brute fact with no explanation? But why these entities and not others—why the particular types of particles that populate our universe? Their existence remains a mystery even if they exist through all eternity. We can understand the existence of most of what we observe, since we see how the existence of one thing depends on the existence of another, but with the basic entities we are confronted with things that exist for no reason—unintelligibly, by brute stipulation. Surely it would be preferable if we could give some account of their existence as contingent beings: unlike numbers they exist in time and can change over time—yet we are told they have no origin, no means by which they came into existence. Is their allegedly eternal existence any more acceptable than saying that ordinary objects could exist eternally as a way of explaining their existence? Suppose we tried saying that sentient beings exist eternally, which is why we can’t explain their origin in terms of antecedent facts—wouldn’t that be totally unbelievable, a complete abnegation of intellectual responsibility?

            So we seem to have reached the conclusion that existence is impossible. It seemed as if we could understand the existence of ordinary objects by reference to pre-existing objects, but it turns out that we have no account of the existence of those objects—which means we don’t really understand the existence even of derivatively existing objects. In the case of non-derivative existence we have only an array of unpalatable metaphysical speculations, more or less desperate and ad hoc. We really have no understanding of existence at all when you get down to brass tacks. And it looks as if any type of existence will face this critique—what to say about the existence of the basic realities. In particular, are we really committed to the eternal existence of concrete contingent beings just by the very notion of existence? Was that part of the bargain when we agreed to operate with the concept of existence? Or are we forced to accept that existence can be conferred by nothing at all, or that it springs miraculously from mere possibility? None of this is remotely attractive; it all seems like a steep and whistling descent into metaphysical nonsense. Thus the proponent of the argument encourages us to draw the obvious conclusion: existence is impossible. Nothing exists in any possible world. The idea of existence is incoherent, puzzling to the point of paradox. It isn’t just that nothing does exist; nothing could exist. For existence leads us to the problem of how the basic existences come to be—for which we have no adequate account. Maybe there is an option we haven’t thought of, or maybe existence is an impenetrable mystery; but as things stand, it looks as if we have a kind of proof that existence is impossible. We are familiar with arguments that purport to show that meaning is impossible, there being no workable account of what meaning is  [2]; the present argument purports to show that existence itself is impossible—of meanings, of material objects, of selves, of anything (except perhaps numbers). Ultimately, we have no conception of how the existence of the most ordinary things is possible.


  [1] It is an interesting fact about our universe that it permits the upsurge of new things: many things exist now that did not exist in the past. You would think that given the means at the universe’s disposal nothing new could be generated, just re-combinations of old stuff. This is why one can sympathize with the adage “There is nothing new under the sun”.  In some possible universes presumably this is the case—no new entities ever come to exist. In a sense existence is quite close to non-existence—that is, new entities are ontologically close to the entities they come from and can easily revert to them. It is surprising that they exist at all as separate things in view of this closeness. Why not just make do with the old things?

  [2] I am thinking of Kripke’s discussion of meaning in Wittgenstein on Rules and Private Language (1982). This is Zeno-level super-charged military-grade skepticism: the complete impossibility of both meaning and existence.


Knowledge of Things



Knowledge of Things


In chapter 5 of The Problems of Philosophy (“Knowledge by Acquaintance and Knowledge by Description”) Russell writes: “Knowledge of things, when it is of the kind we call knowledge by acquaintance, is essentially simpler than any knowledge of truths, and logically independent of knowledge of truths, though it would be rash to assume that human beings ever, in fact, have acquaintance with things without at the same time knowing some truth about them. Knowledge of things by description, on the contrary, always involves, as we shall find in the course of the present chapter, some knowledge of truths as its source and ground.”  [1] He goes on: “We shall say that we have acquaintance with anything of which we are directly aware, without the intermediary of any process of inference or any knowledge of truths.” Thus Russell proposes a sharp duality in our knowledge of things: knowledge we have by bringing things under descriptions and knowledge we have by means of direct unmediated experience. As illustrations he cites knowledge of sense-data and knowledge of the external object that causes sense-data: we are “immediately conscious” of such sense-data as color and smoothness, he says, whereas we have no such direct awareness of the table itself, which we know only by means of the description “the physical object which causes such-and-such sense-data”. So our knowledge of things divides into two subspecies that act quite differently and are made possible by quite distinct mental faculties. The phrase “knowing X” is capable of two types of analysis depending on whether the knowledge relies on acquaintance or description. We might have thought we have a unitary concept here but in fact an irrefragable dualism confronts us: there are two completely different ways a thing may be known (or known of or known about).

            The distinction appears robust when we follow Russell’s way of articulating it. We do know about sense-data by the faculty of introspection, which is different from the faculties of external perception and associated inference; and description is a linguistic matter while immediate awareness is not. There is also certainty concerning our acquaintance with our own minds, whereas we have no such certainty with respect to the external world.  Further, it sounds intuitively correct to say that we have “direct” knowledge of some things and “indirect” knowledge of others. So Russell seems to be barking up the right tree when he announces a clear-cut distinction between two types of knowledge of things—as clear as the distinction between a priori and a posteriori knowledge, say. We are accustomed to such dualisms in philosophy and we can add this one to our stock. Certainly the distinction has been warmly received and smoothly transmitted down the philosophical generations. But is it as sharp and principled as it is easy to suppose that it is? Are there really two completely different ways of knowing about things with nothing to unify them? First, let us note two features of so-called knowledge by description as characterized by Russell: it is linguistically mediated, and it is dependent on knowledge of truths (propositions). These would certainly mark a key difference, but both are questionable, and indeed clearly wrong. It is not necessary to have and deploy a language in order to have so-called knowledge by description, or else animals and small children would have no such knowledge; what is necessary is the possession of concepts—ways of thinking about things. I can think of an external object as the cause of my sense-data without putting this thought into words—I just need to deploy the appropriate concepts. We can call these concepts “descriptive” if we like but that does not imply that they take a linguistic form (not in a public natural language anyway). Knowledge by description is therefore no more language-dependent than knowledge by acquaintance. Second, it is not part of the idea of knowledge by description that any knowledge of truths is essential: for a description is not a sentence; it does not express a proposition. Of course, if we accept Russell’s theory of descriptions, we get that result; but it is no part of the idea of knowledge by description itself. If we thought that some knowledge necessarily depends on names (“knowledge by naming”) that would not imply that such knowledge requires knowledge of truths (propositions). So again, that mark of distinction lapses. We are left with the claim that some knowledge of things is based on concepts—ways of conceiving of things, constituents of thought, elements of the reasoning faculty.

            Attention now shifts to knowledge by acquaintance: is it not dependent on concepts? Russell appears to think so, but the question is debatable. He seems to think that our knowledge of sense-data has no conceptual component—that sense-data just loom up and slap us in the face, so to speak. But surely introspection involves bringing sense-data under concepts: I think of my sensation of red as a sensation of red when I introspect it. This means that I conceptualize sense-data according to their intrinsic character: I classify them, relate it to other sense-data, and have truth-bearing thoughts about them. I don’t just gaze at them blankly in uncomprehending wonder, as I might gaze at some strange animal quite unknown to me (imagine seeing an octopus for the first time). So when I am acquainted by introspection with my sense-data I apply concepts to them—whatever exactly concepts turn out to be. I am not cognitively void with respect to them. So isn’t knowledge by acquaintance a species of knowledge by description, i.e. concept-mediated knowledge? Consider your knowledge of the color red: you are acquainted with red and know what it is—is this knowledge completely independent of any means of mental representation? No, because you have to perceive the color in order to know what it is—it has to come before your mind. There must be some sort of intentionality involved: no cognition without mental representation. It isn’t that red just grabs your mind; your mind has to grab red, i.e. mentally represent it. There must be a mental act of grasping or perceiving or apprehending—some sort of intentionality. But then knowledge by acquaintance is like knowledge by description in that both involve a faculty of mental representation. It isn’t that knowledge of external objects requires intentionality while knowledge of internal objects doesn’t (which appears to be Russell’s view); both require it. The dualism totters; the distinction collapses; the division disappears. There is really only one type of knowledge of things—the type that builds in a system of mental representation. To be sure, we can distinguish different objects of knowledge, and record differences of epistemic status, and observe variations in the type of mental representation involved: but there is no deep distinction of the kind Russell claims—no difference of fundamental structure.

            Interestingly, Russell himself essentially gives up on the contrast when he notes that all knowledge by description depends on knowledge by acquaintance. As he famously states: “Every proposition which we can understand must be composed wholly of constituents with which we are acquainted” (his italics). This means that all knowledge by description is reducible to knowledge by acquaintance—just not always acquaintance with the thing known. He should really have distinguished two types of knowledge by acquaintance: acquaintance with the thing itself and acquaintance with other things connected to that thing in various ways (as sense-data are connected causally to external objects). He could have called these “knowledge by direct acquaintance” and “knowledge by indirect acquaintance” or “knowledge by acquaintance with connected things”. If he had done that he wouldn’t have overdrawn the contrast with simple acquaintance by supposing that somehow language is essential to indirect knowledge. What he didn’t see is that acquaintance itself is also a matter of mental representation, and so is always a mode of “description”. All knowledge by description is knowledge by acquaintance and all knowledge by acquaintance is knowledge by description (in the wide sense). To put the point differently, some knowledge is by denotation (as applied to concepts as well as words) and some is by perception or introspection, but both of these are semantic relations in a broad sense. For example, I may know of a person merely by being told about him and I may know of a color by seeing it, but both of these relations involve a structure of intentionality: the distinction between methods of knowing is real, but it doesn’t entail the kind of cognitive duality postulated by Russell. Denoting and perceiving are indeed different relations, but both fall under a broader concept—and that concept unites the different instances of knowledge of things.

            The same applies to knowledge of meanings: there are not two types of semantic knowledge, description-based and acquaintance-based. It isn’t that names are grasped “by description” and demonstratives for sense-data are grasped “by acquaintance”. In both cases we have mixture of direct acquaintance (possibly with universals) and indirect description (i.e. conceptualization); even when I grasp the meaning of “this pain” I invoke my introspective conception of pain, not just the pain itself. Nor are there two types of meaning corresponding to the different ways meanings are known—descriptive meanings and acquaintance meanings. A description theory of names will require a bedrock of acquaintance for the elements of the description; and a direct reference theory of mental demonstratives (“this pain”) will require a mode of conceptualization. Linguistic understanding is fundamentally uniform not divided into two completely different forms—the aloof inferential verbal form and the slap-in-the-face confrontational nonverbal form. This bears on the question of whether there are two types of definition, often labeled “verbal” and “ostensive”. I can give you some words to define a given word (e.g. “unmarried male” to define “bachelor”) or I can direct your attention to something in the world by pointing at it (e.g. I can point to something red in order to define “red”). So it is supposed that something fundamentally different is going on in the two cases: the word “definition” denotes two completely different procedures. But this distinction blurs when we examine things more closely: verbal definition depends on a prior grasp of the defining words, ultimately tracing back to something non-verbal; and so-called ostensive definition (as Wittgenstein taught us) is enmeshed in a sophisticated matrix of pre-existing conceptualization. The former is not so purely verbal as it might appear, and the latter is not so purely confrontational as we are inclined to suppose. We must already categorize the world in order to benefit from acts of ostensive definition, as we must already know the meaning of words non-verbally in order to benefit from acts of verbal definition. And doesn’t a verbal definition tacitly point to words in order to convey a meaning (“bachelor” means this: “unmarried male”)–as an ostensive definition converts the world into samples in order to convey a meaning, where samples act like elements of language? There is pointing in verbal definition (the “paratactic theory of definition”) and there is symbolism in ostensive definition (the “semantic theory of samples”).  [2] These activities are not as distinct as we might casually suppose. There is no deep dualism of definition of the kind we tend to think. Distinctions can doubtless be drawn but they don’t add up to the sharp division by which philosophers are prone to be enchanted. Definitions are not cleanly divided into “definitions by description” and “definitions by acquaintance”, though words and perceptions can play different roles in different ways of getting meaning across in different contexts. It isn’t that some definitions involve simply pairing words with other words while others involve pairing words with chunks of reality, as if neither activity presupposes anything of the other. That would be an untenable dualism, a typically philosophical piece of distinction mongering. Acquaintance and description (so-called) bleed into each other. The terminology itself should be abandoned as misleading.  [3]


Colin McGinn                        



  [1] The book was first published in 1912 and has, I believe, exercised an enormous subliminal influence on the course of subsequent analytical philosophy, so much so that Russell’s distinction is taken for granted (I myself took it for granted until very recently). It is beautifully, and deceptively, seductive, largely owing to its confident and limpid style.

  [2] On samples as parts of language see Wittgenstein, Philosophical Investigations, sections 16 and 50. The paratactic theory of definition would mirror Davidson’s paratactic theory of indirect discourse in “On Saying That”: “Oscar said that: We are all in the gutter but some of us are looking at the stars”. That is, that-clauses are analyzable as involving a demonstrative directed at a subsequent piece of discourse, as in “’Bachelor’ means that: ‘Unmarried male’”. In general, indexicality and description are closely intertwined.

  [3] The terminology applies quite reasonably to another distinction, not Russell’s distinction–that between knowing about a thing by testimony and knowing about it by sense perception. Of course, it is true to say that testimony knowledge is acquired by means of “description”, i.e. verbal report, and also true that you can be “acquainted” with a person or place by actually seeing him, her or it, thereby acquiring knowledge: but that has nothing to do with Russell’s distinction. This is one of those cases in which a perfectly legitimate distinction drawn in ordinary language is used to confer spurious legitimacy on a quite different philosophical distinction. One always needs to study the fine print carefully and not be carried away by putative vernacular precursors.   


Wavelength and Color


Wavelength and Color


We are told that wavelength determines color. But what does “determine” mean here? It is not in doubt that the wavelength of incoming light causes the perception of a specific color, though we should note that it only does so in conjunction with non-trivial facts about the perceiver’s eyes and nervous system. This is compatible with allowing that other types of stimulus could causally determine color vision: for example, the wavelength of sound waves might conceivably cause colors to be seen, or indeed a knock on the head. The causal relation is quite contingent and extrinsic with respect to the nature of color. What we know is that wavelengths are causally correlated with the perception of particular colors, but this is a far cry from supposing that they constitute what colors are. Kicks to the shins are causally correlated with bruises of various hues, but no one would suppose that bruises are kicks.

            Is there any closer relation? It might be supposed that an identity theory is possible: that way we would have a nice materialist account of what color properties are. It is worth being very clear about what this would mean, which requires us to be explicit about what wavelength is. According to the wave theory of light, light has a wavelike structure analogous to the waves you can see in the ocean: waves have both frequency (how many cycles reach a certain point in space per unit of time) measured in Hertz units, and they have length (the distance between successive crests or troughs) measured in meters. Wavelengths are very short for light waves and hence are measured in nanometers. The visible spectrum (between 400 and 700 nanometers) is just one part of the much wider electromagnetic spectrum.  Thus the identity claim is that a given color can be identified with waves of light with a specific distance between incoming peaks—this being nothing other than the phenomenon you can observe on many a beach (though involving a light medium not a water medium).  So the contention is that colors have a wavelike structure, travel through space, and have specific distances between their wave crests: red, say, is shaped like a wave, traverses space, and has a certain wavelength (longer than other colors). These are attributes of light, so by Leibniz’s law they must be attributes of color, if colors are identical to light waves. That is the nature of the color red, its true character, its inner essence: to be red is to have a particular wavelength, i.e. a certain distance between successive crests.

            But is this the way red looks? If you stand on a beach and look out to sea, you will see waves rolling into shore, and these waves will look to have a specific wavelength, generally of several meters: the shape and height of the waves is apparent to you and the distance between them is also apparent. These waves look precisely like the waves they are. But do colors look like waves? Evidently not: when you look at a red apple you don’t have an impression of waves of light approaching your eyes—you don’t see anything wavelike at all. You don’t see crests and troughs, or wave heights and the distances that separate them. The phenomenology of seeing colors does not include any perceiving of wavelike structure—you don’t, say, see red as having a longer wavelength than blue. Why not if colors just are wavelengths of light? You might say these waves are too small to see, that’s why you don’t see them—but they are still what visible color actually is. But if red were made of much bigger waves would we see them when we looked at a red object? Would red look like the waves it is if the waves were big enough to see? Clearly not: it would presumably look something like colorless ocean waves.  And if we looked through a microscope and could see our actual light waves, would we still be seeing red? No, the redness would disappear, to be replaced by magnified waves of light. So colors don’t appear to be wavelengths, as we ordinarily perceive colors—yet we are seeing some sort of property. The property we see cannot then be the property of being a wave with a specific wavelength.  [1] Some philosophers will respond by saying that these color appearances are actually appearances of light waves with their corresponding wavelengths, with no further property interposed between them and the perceiver. But this is a bad theory of perception: we clearly do perceive color properties when we see colored objects–we see things as red, blue, green, etc. But the property that we see can’t be identified with a wavelength property, because it looks nothing like a wavelength property. At the very least we would need to be given an explanation for why colors don’t look the way they intrinsically are, according to the identity theory; but no such explanation is forthcoming. So the phenomenological unreality of the alleged wavelike nature of color is a count against the theory that colors are identical with such properties.  [2] And it isn’t as if the way colors look gives us a hint of their real wavelike nature: there is nothing wavelike about the way color looks to us. If anything, colors look non-wavelike. We could, of course, go eliminative about colors, claiming that colors don’t really exist as properties, though light waves do—then there would be nothing that is not covered by the physics of color. But if we insist instead that colors are real and also reducible to wavelengths, then we have to face the question of why they don’t seem that way. On the face of it, the identity theory of color is refuted by the appearances. To be sure, wavelengths trigger the perception of color properties—they are the external cause of color perception—but they are not what color properties are. These properties are ontologically separate, a distinct realm of existence.

            There isn’t even a supervenience relation here, since the same wavelengths could correlate with distinct colors in different possible worlds, depending on the eyes that respond to these wavelengths. The correlation in the actual world is just that—a contingent correlation. Waves in the ocean often correlate with gleaming peaks and a crashing sound, but it would be quite wrong to think that there is some necessary connection here; a trip to a nearby possible world would quickly disabuse you of that misconception (in that world ocean waves form only at night and merely murmur). We must not mistake correlation for identity. The relationship between wavelengths and colors is an external contingent relationship not a relationship of identity or constitution. The reason colors don’t look like waves of light, equipped with frequency and wavelength, is that this is not what they are. The same can be said of heard sounds and sound waves in the atmosphere: pitches don’t sound like waves with frequency and wavelength because they are not such waves. We can hear waves in the ocean and gauge their frequency and wavelength, but nothing analogous is true of heard sounds themselves—they are not phenomenologically wavelike. Sound waves cause sound perceptions, certainly, but the sounds heard are not identical to sound waves in the atmosphere: one doesn’t hear middle C as a particular spatial separation of wave crests. Sensible qualities like these are not reducible to wave patterns in a physical medium.


  [1] A further point: if colors are identical to wavelength formations, which are invisible to the human eye, how come we can distinguish colors from each other? How can we tell that red is not the same as blue if we can’t see the wavelength difference that constitutes the difference of color? The obvious answer is that the property we see when we see red is not the same as the property of having a certain wavelength but is a property that is entirely visible to the human eye.

  [2] We may note, too, that wavelength is a continuous quantitative property, being simply distance between two points, but colors vary qualitatively and are discontinuous: red is not simply “longer” than blue. So wavelength is unable to capture the way the visible spectrum is divided up. There ought to be just one color, varying along a single dimension, if colors were definable in terms of wavelength.


Puzzles of Color


Puzzles of Color


The mantis shrimp challenges philosophical reflection. This little crustacean is reputed to have the most complex eyes in the animal kingdom. Mounted on stalks, these eyes can move independently and rotate freely; it has between 12 and 16 photoreceptors compared to the human 3. It can see into what is called deep ultra violet. It uses these eyes in the close capture of prey where great precision is needed (it also has a powerful club with which it stuns its prey).  There is every reason to believe that this creature’s eyes are superior to human eyes; in particular, it has superior color vision. Presumably we don’t know what it’s like to be a mantis shrimp, since its color phenomenology outstrips ours (compare the bat’s perception of sound). Its visual phenomenology is like Technicolor compared to our dull monochrome. It sees more colors and it sees them better.

            This raises puzzling questions. First, it suggests that colors exist independently of vision: just as some colors exist independently of human vision, so there may be colors that exist independently of all (terrestrial) vision. For there could be a species superior even to the mantis shrimp in the perception of color, so that color does not consist in actually causing color experiences. The wavelengths of light clearly pre-exist perceivers and it seems right to say that things were red (say) before any animal saw them as red. We can try saying that they had a disposition to appear red, but then there is no categorical property that being red consists in. For various reasons it is preferable to accept that things have colors before being perceived as such. The mantis shrimp thus has access to mind-independent colors that many other animals have no access to. But it is a puzzle what exactly these colors are: not wavelengths, and not dispositions to appear—but then what? They seem neither objective nor subjective but somewhere between the two; they challenge the usual binary opposition of objective and subjective. They seem to have phenomenology written into them and yet they are not identical to anything in the mind. They are, we might say, objective-cum-subjective. Perhaps they are a primitive type of property that belongs neither to the world of physics nor to the world of psychology. Yet they belong to physical objects and are perceived by the mind.

            But that isn’t the puzzling question I most want to talk about. Suppose we say that colors are secondary qualities in the classical sense—dispositions or powers to elicit perceptual experiences. Then we must ask whoseperceptual experiences: is it our experiences or those of the mantis shrimp? Since its color vision is markedly superior to ours we can’t identify the two, which means that the same objects can elicit different perceptions of a given color: red looks one way to them and another to us. Are there then two reds? No, there are two appearances of the same red. But the shrimp sees red better than we do, so it can’t be that our experiences are determinative: better to say that the shrimp’s visual system gets to decide. But then couldn’t there be perception of red that outclasses even the mantis shrimp’s? The problem is that colors admit of more or less accurate perceptions of their nature, but that is not compatible with claiming that colors are such dispositions. No one’s visual system gets to decide what red is, because the appearance of red to different perceivers varies: being red must transcend how red looks to different visual systems. Suppose we try saying that the best visual system on earth determines the identity of the colors, say the mantis shrimp’s: then it follows that if this shrimp goes extinct we should turn to the second best perceiver of color on earth; but this move shifts the identity of red from one subjective appearance to another. There must be more to the color red than its subjective appearance to sundry perceivers of red. Maybe our visually acute shrimp gets the closest to the true nature of the color red, with us running a distant second, but neither of us gets to determine the intrinsic nature of the color red. But then the traditional secondary quality view of color has to be mistaken.  [1] That would be fine if we could see our way clear to identifying colors with physical properties like wavelengths, but that way is blocked for familiar reasons.  [2] The upshot is that colors again emerge as sui generis basic properties belonging neither to physics nor psychology; they occupy a curious no-man’s land of the objective-cum-subjective. Perhaps the mantis shrimp sees red exactly as it intrinsically is (that is certainly an appealing thought), but this judgment of exact correctness presupposes that something outside the shrimp’s visual system fixes the nature of the property—the shrimp gets it right according to an objective measure. We, on the other hand, see it through a glass darkly, or through a haze laughably. We certainly can’t claim that our visual impressions constitute the very nature of the color red. No one’s do. Colors are like shapes in having a nature of their own that is logically independent of how they appear to this or that perceiver, simply because different visual systems represent them differently. We might be said to be color blind not only with respect to colors we can’t see but with respect to colors we can, since our visual system might be, as it were, legally blind with respect to the real nature of color. But legally blind perceptions of color are hardly capable of constituting the essence of color. The quality we are seeing, then, is not identical to a disposition to elicit imperfect perceptions of color in us; there is such a disposition, to be sure, but it cannot be the color. Our perceptive shrimp has a better claim to fixing the true nature of the color, but even she is not capable of constituting color properties (save per accidens).

            The same point can be made about other sensible qualities: smells, tastes, and sounds might not reveal themselves to our specific sensory systems.  [3] Animals with superior senses to ours might experience these qualities quite differently from us, and in ways closer to their true nature. Yet none may quite get to the heart of the quality in question—perhaps no animal on earth has ever experienced what sugar really tastes like. So we cannot assimilate such qualities to dispositions to appear in certain ways to existing perceivers, since these appearances may be more or less inaccurate or imperfect. So again, the traditional secondary quality account cannot be correct. This leaves us with an implausible physical reductionism or an acceptance of a range of puzzling qualities that are neither one thing nor another. The world of sensible qualities lies tantalizingly out of reach, perceived only as veridically as the perceiver’s limited senses allow. Human sensation, in particular, might be feeble and misleading compared to other “humbler” creatures. To be sure, we have big penetrating brains, but our sense organs might well be rather superficial and misleading even with respect to qualities with which we fancy ourselves well acquainted. We might have a rather poor idea of what red is—not as poor as a blind man’s, certainly, but pretty dismal compared to the mantis shrimp’s. If we had a hundred color receptors instead of three, the world of colors might have a startlingly different phenomenology for us. We might then see colors as they really are.        


  [1] The traditional secondary quality theory of color was proposed at a time when human biological superiority was presupposed—it was assumed that the human perception of color must be authoritative. In our post-Darwinian age we are far more ready to accept that the human species might not set the standard for veridical perception. 

  [2] These reasons are discussed in my The Subjective View (1982).

  [3] The sense of touch is less clear at least for the perception of shape (but perhaps human sensations of heat inadequately reveal the true nature of being hot).


The Ballad of Dolores Haze


I have been writing songs lately, so I thought I’d share this one with you: think of it as sung in Country and Western style. 


The Ballad of Dolores Haze


He broke my heart

But you broke my life

He was a creature of art

You were a monster equipped with a knife


A pentapod monster, I’d agree with that

His pen and photography were never so bad

I loved him so, I don’t know why

It’s not as if he was much of a guy


You killed my mother

So you could get to me

He saved me from another

Who would never let me be


Clare Quilty was his name

He was impotent, or so he claimed

You were anything but

But I’d rather be trapped in a desolate hut


No, I never vibrated to your touch

How could I when I hated it so much?

You loved me, you say

You’d take me far away


But I’d rather stay in this town of hicks

With my ironically named Dick

Than hitch myself to your terrible beam

And live again in that despicable dream


You broke my life, you see

And that’s not nothing, that’s all of me

Oh, you are hurting, I can tell

And I do pity you in your self-made hell


Goodbye honey, please dry your tears

They’re nothing compared to my childhood years

And the motels and the cars

And the bars and the barmen and the cold gray stars


Partial Skepticism


Partial Skepticism


The solipsist is a partial skeptic: he doubts the existence of other beings, sentient or otherwise, but he doesn’t doubt his own existence. He might go on positively to affirm that nothing exists but himself. He favors one object over all others. What is the analogue of this for the material world? Suppose I come to the conclusion that no material object exists except my coffee cup: it is the sole object that is real, all the rest being illusory. I am a kind of coffee cup solipsist—I favor one material object above all others. Perhaps I feel that I know my coffee cup (hereafter “the Cup”) better than any other material object, given my close daily acquaintance with it; in any case I select this object as the sole possessor of reality (apart from myself). I may think I am a brain in a vat of the classic type, but with one exception—the Cup. Just as the traditional solipsist picks out one individual as real (himself), so this “solipsist” picks out one object as real, namely a certain cup. Evidently this is a coherent position: it is an epistemic possibility that only the Cup exists among all the other apparent material objects. Perhaps the mad vat scientist has set things up so that the only veridical perceptions are of the Cup, the rest being hallucinations; or the architects of the Matrix have a peculiar cup fixation. Everything in my visual field is an illusion save for that solitary beverage container. Hey, it’s logically possible. It is a question whether this form of partial skepticism is a new type of skepticism: is it another skeptical scenario that needs to be considered? Apparently it is. And once we have this type of skepticism on the table we can construct varieties of it: for example, are oak trees the only real trees? We just select a class of material objects as exempt from general skepticism. Similarly, a partial solipsist skeptic could, logically, claim that only persons of a certain type exist—say, only people under six feet tall. That again is an epistemic possibility.

            Why do these partial skeptical positions seem completely arbitrary and unmotivated? Formally they resemble classic solipsism, but they seem entirely without intuitive appeal. The reason is obvious: they have no epistemological basis. There is no reason to favor a single coffee cup or a particular species of tree or a certain height of person. But there is a reason to favor myself over all other objects, namely my immediate infallible knowledge of myself: I know for certain that I exist but not that other people exist. By contrast, I have no greater certainty about the Cup than about other objects: so it seems completely arbitrary to favor that object above all others. I wish to make two points about this observation. The first is that there could be epistemological reasons to favor the Cup: that is, there are possible worlds in which there is more reason to believe in the Cup than in other material objects (irrespective of whether either really exists). Suppose I have regular, clear, and vivid perceptions of Cup (it has a name now) but irregular, unclear, hazy perceptions of other objects—so much so that the possibility of hallucination becomes highly plausible for them. Then I would have reason to think that Cup exists but other objects probably don’t—whether or not this supposition is actually correct. If so, I might be tempted to accept a form of partial skepticism about the material world, or even to deny that any material thing exists apart from my beloved Cup. That is, in this world a position analogous to traditional solipsism would be rational (assuming that solipsism is rational): there is more reason to believe in one material object than in any others, as there is more reason to believe in myself than in other selves.

            The second point is that solipsism as a positive doctrine starts to look distinctly irrational in the light the analogy. For why should we entertain the move from an epistemological point to an ontological one? It is true enough that Cup might be the only existent material object—this is an epistemic possibility—but why should this incline one in the slightest to suppose that it is the only existent object? Likewise, it is an epistemic possibility that only I exist—this cannot be ruled out with absolute certainty—but it is a total non sequitur to infer that it is actually true. Such a conclusion is as arbitrary as supposing that only my coffee cup exists, even when I have more reason to believe that it exists than to believe in the existence of other material objects. The belief in solipsism is as irrational as the belief that only precious Cup exists: both are logical possibilities, to be sure, but there is not even a smidgen of evidence to indicate that either supposition is true. What is true is that people care a lot more about themselves than about other people, whereas they don’t tend to have such strong feelings about their coffee cups; but that is not a reason positively to believe that you exist and other people don’t. Certainty is one thing, truth another. Solipsism is no more likely to be true than the corresponding belief about a particular cup. Neither position can be refuted as an epistemic possibility, but neither position has anything solid in its favor. Yet people have been more strongly drawn to solipsism than to the analogous position with respect to material objects. Why?  [1]


  [1] I have not discussed here alleged reasons for supposing that only the concept of one’s own mind makes sensegiven the nature of mental concepts (only first-person uses are properly intelligible), limiting myself to the more popular claim that epistemological reasons favor solipsism. The idea that one’s own mind is uniquely favored ontologically should strike us as remarkably self-centered, a reflection perhaps of our natural selfishness.