An Answer to the Skeptic

Skepticism gains traction from the true justified belief theory of knowledge, because it can be argued that our beliefs are seldom if ever justified. But that is just one theory, not a datum. What if we adopt another type of theory? I observe, to begin with, that other types of knowledge than knowing-that are not subject to skeptical argument: knowing-how and knowledge by acquaintance. You can’t use a brain-in-a-vat scenario to undermine the claim that we have knowledge-how and knowledge by acquaintance, because these are far removed from what is called propositional knowledge (Russell’s knowledge-by-description). I have knowledge-how if I have a certain ability, whether I can justify the claim that I have this ability or not, even while I am a brain in a vat and have never done the thing in question, e.g., throw a ball. I can also know by acquaintance what red is without knowing whether there is an external world; it just depends on what I have experienced not what beliefs I can justify. These are not evidence-based types of knowledge, so the quality of the evidence cannot be impugned. So, the concept of knowledge is not inherently susceptible to skeptical challenge. But what about knowledge-that?

Suppose we go for a perception-based theory of knowledge not a justified true belief theory: that is, we extend acquaintance to knowledge of facts.[1] For example, I might take myself to know that I am lying in bed, and suppose I am: do I really know that I am? That depends on whether I perceive that I am lying in bed, which doesn’t follow from taking myself to be and this actually being the case. Suppose I do perceive this (the fact of my being in bed causes me to have the experience); then we say I am acquainted with this fact—whether I can justify the belief or not. My knowledge depends only on the facts not on my ability to have true justified beliefs about them. Even if I can’t rule out the hypothesis that I am a brain in a vat, I am still causally connected to the fact in question, if indeed it is a fact. Thus, I can know this fact independently of my ability to justify my beliefs about it: for my knowledge is not based on any such justification; it just arises from my being in a perceptual relation to the fact. Knowing facts by acquaintance (perceiving them) isn’t susceptible to the standard skeptical argument. But if that’s what knowledge is, then we have defeated the skeptic about knowledge: the knowledge exists—whether we can know this or not (we are not discussing second-order knowledge). Perceptual knowledge does not depend on possessing justified beliefs. The skeptic has no argument against the possibility of first-order knowledge of facts based on direct acquaintance.

This point may be conceded, but what about all the putative knowledge that is not based on acquaintance but on inference? What about belief-based knowledge based on evidence and justification? Surely the skeptic can get his fangs into that! Here we might agree but insist that no sensible person has ever supposed otherwise: of course, we can’t know what goes beyond our direct apprehension of fact; we can only surmise. It is a misuse of the concept of knowledge to suppose otherwise.[2] Perhaps we can stretch a point and agree that we might call such belief “knowledge” in a relaxed frame of mind, but it was never really knowledge, as distinct from reasonable speculation, just loose talk for pragmatic purposes. I don’t know that atoms exist, though I might have reason to believe that they do; I only truly know what I myself directly perceive. If so, it was never part of (sensible) common sense to apply the concept of knowledge beyond its proper domain, so the skeptic is tilting at windmills and parading truisms. I know hugely many facts about the external world just by perceiving it, even though there are many things I reasonably believe that don’t count as knowledge. Isn’t that what we normally suppose?

Here the skeptic may retract his horns: okay, he says, but we still don’t have adequate justification for the beliefs we hold, despite the fact that we have a lot of knowledge by acquaintance. To that weakened form of skepticism we can reply as follows. We can simply agree with the skeptic but point out that he has said nothing to rule out knowledge in the areas proper to it; he is talking about something else entirely, i.e., justified belief. But second, we could recommend a comparative notion of justification combined with some absolute cases of it. Thus, I am fully justified in believing I am in pain and I am more justified in believing that eagles fly than that pigs fly. We need not claim that all justifications are created equal in order to rebut the justification skeptic—that would be absurd. Justification comes in degrees of cogency and not all measure up to the perfect case—whoever denied it? So, the justification skeptic has not raised a startling new epistemological challenge that undermines commonsense epistemology. We can all agree that our justifications are pretty shabby, judged objectively, but still maintain that the concept of justification is in good order with useful applications. What we are not going to agree on is that there is no such thing as knowledge of the external world; and the correct concept of knowledge does not invite any such conclusion. So, the skeptic has left commonsense epistemology more or less where it was, not counting those rash epistemological optimists who ought to know better. Sound minds have always known that human knowledge is a more limited affair than has sometimes been advertised; that is not skepticism but realism. Human knowledge: its scope and limits.

In case you think this kind of anti-skepticism is toothless, let me note its consequences for knowledge of other minds and the past. For we can now be said to know facts about other minds and the past: that is, such facts can act as the cause of our perceptual states. I know you are in pain because your pain has caused behavior that I perceive as pain-expressing—that is a fact. I can’t justify my belief that you have a mind to the skeptic’s satisfaction (and not unreasonably), but that doesn’t prevent me from being in a knowledge relation to the fact in question. Similarly, past facts cause current memories, so I know them by something akin to perception (it might even be perception)—even if I can’t justify my belief that there is a past. Thus, I can know facts about other minds and the past by something like perceptual acquaintance, though (arguably) I can’t justify my beliefs about these things. Knowing facts is one thing, justifying beliefs is another. To put it simply, if knowing is seeing, then I know a great many things; what beliefs I can justify is another matter, and may well be shakier than some people have supposed. Since no genuine knowledge is constituted by true justified belief—that is just an incorrect analysis—it is irrelevant to knowledge if adequate justifications for belief are unforthcoming.[3]

[1] See Michael Ayers, Knowing and Seeing (2019); also, my “Perceptual Knowledge”.

[2] See my “Non-Perceptual Knowledge”.

[3] A virtue of the account given here is that it concedes some territory to the skeptic—he isn’t just barking up the wrong tree—but it doesn’t concede his most radical claim, namely that nothing is known about the external world (or other minds and the past). Our alleged justifications don’t really warrant the kind of strong belief we are apt to derive from them, but that has nothing to do with our ability to have knowledge; and indeed, we have a lot of that. Knowledge proper was never about warranted belief. Human knowledge, like animal knowledge, is in good shape, though quite restricted; belief on the other hand cries out for justification and often falls short of it. That’s why some philosophers (e.g., Popper) dispense with it in serious contexts.

Space, Time, and Logic

Philosophy needs a metaphysical vision. Humbly (and pretentiously), I will provide one. As far as I know, it has no predecessor, though echoes of other theories will be apparent. Neither does it have a name: it might be called “Logical Spatio-Temporalism” (LST), or just “Spatialism” because space is central to it. It isn’t idealism or materialism, since mind and matter don’t figure in the foundations. The elements of it are space, time, and logic: this is what reality is fundamentally made of. Matter, mind, and mathematics are derivative from these three elements; I would even say they cause them. I take it that matter is implicit in space not just annexed to it or superadded. This is a modern viewpoint, though it admits of various elaborations. Matter is a modification of space, a version of it. I favor the idea of a space-matter continuum; there is an underlying unity here (as there is said to be a unity called space-time). We could say that space became matter as we know it; it was already matter in some other form (no particles of the familiar types). Space is matter before it has been cooked, so to speak. Matter, as it exists now, evolved from space. To space we must add time—the medium of change. Without time matter is static and unchanging, unrecognizable as matter. Time renders space dynamic. The modern physicist will insist that space and time are not separable, and if you are happy with that curious identification, you are welcome to adopt it; then the theory will have space-time as one element of it. Could time exist without space? Let’s not even go there; say what you like, you can still share my metaphysical vision. The third element is the novel one: logic. I don’t mean predicate calculus or modal logic or any other symbolic scheme; I mean necessary consequence of the kind we call logical (and let’s not go into that question either). Heuristically, think of Frege’s abstract realm of “Thoughts”—the subject matter of reasoning. We adjoin this to space and time in order to capture the realm beloved of the rationalists—logic, mathematics, philosophy itself. The picture, then, is that mathematics, as it now exists in human civilization, is an outgrowth of logic (possibly combined with space in the field of geometry). Together, these three elements constitute the foundations.

The vision is that space, time, and logic precede mind, matter, and mathematics. These are relative latecomers in the construction of the universe as it presently exists. Everything that now exists owes its being to these three things; we might even say that the present universe is supervenient on space, time, and logic—determined by them. They cause everything. After God created them, he could slope off and take a nap; his work was done. Alternatively, the ontological structure of the universe has them at the bottom holding up the superstructure. They are bedrock. Mind and matter are mere side-effects, not foundational at all. Any universe like ours will have them as its infrastructure. But this is not to say that every possible universe is so structured: for it may be that space and time can vary across logical space. Maybe in some possible worlds space has a different geometry—of 27 dimensions or infinitely many. Maybe time loops and curls somewhere in logical space. Logic, however, remains the same, being metaphysically necessary. These universes will look nothing like ours and may have nothing corresponding to our mind and matter. I am not legislating across all possible universes. But in our universe our space and time call the shots; they determine what will be or not be. These are the metaphysical elements that fix the reality of this world. The fundamental layer consists of space, time, and logic (or space-time and logic). Logic is the province of the a priori; space and time are the progenitors of the a posteriori.

Notice, however, that the scheme is metaphysical not epistemological; not a trace of the epistemological shows up in this metaphysical system. It is strictly by the book. This is reality as it exists completely independently of all or any knowledge. Our knowledge results from this reality; it doesn’t bring this reality into being. Not space, not time, not logic (shades of Frege). How much we know of this reality remains to be determined; it could all be completely unknowable by us. What we mean by space (our conception of it) might be nothing like space as it exists objectively, and similarly for time and logic. My own bet is that the gap is surprisingly large, but that is a separate question. We can at least responsibly surmise that reality is structured in the tripartite way described. Reality has the architecture (to use a fashionable term) I have conjectured: here space, there time, yonder logic. It has three basic ingredients (like bread—flour, water, and yeast). If you want to bake a universe, these are the ingredients you need—assuming you want a universe like ours (if you cut out the yeast, you end up with something pretty flat).

It will be observed that this is a minimalist theory; it tries to cut everything down to the minimum number of elements necessary. This is desirable because we don’t want to populate the universe with too many basic features, or else we won’t know how it exists. Yet it does contain one of two elements beyond what some systems envisage (the ones we call monistic—idealism, materialism). We must strike the right balance between profligacy and miserliness. Occam’s razor must not cut too deep. It really does seem to me that the three elements I have identified are genuinely distinct and individually necessary; whether they are sufficient is the mooter point. Some may urge that we need an extra ingredient if we are going to get all that we need—call that ingredient “God”. The itch that prompts this urge is certainly real, but we do better to live with the itch than succumb to superstition and quack cures. In any case, STL eschews such expedients and adventures. It carries a light backpack.[1]

[1] The thing with metaphysical visions is that they are best presented pithily and pitilessly, so they can penetrate the carapace of prejudice that seeks to repel them. Then the reader can contemplate them at his or her leisure and not be swamped by detail and qualification.

Bernard Williams and Me

One day, over twenty years ago, I ran into Bernard Williams in the corridor at NYU. He remarked: “The thing about you, Colin, is that you think you’ve either solved the problems of philosophy or they can’t be solved at all”. I paused for less than a second and replied, “I believe you’re right”. I assume his point was that this is a rather self-confident attitude, perhaps not entirely justified by the facts. But I think, on mature reflection, that it was perfectly reasonable, and not for “narcissistic” reasons. In the case of the mind-body problem, I had at that time been thinking about it for over thirty years and was well-versed in all the standard theories, as would be any half-way competent philosopher; and I had no idea what such a solution would look like. I also knew personally all the top philosophers of the time, and they had no idea either (though some may have thought they did). It was phenomenally unlikely that I, or any of them, would come up with the correct theory any time soon, and there were principled reasons for urging pessimism. It is perfectly rational to believe that no one living will come up with the solution. Is it rational to believe that someone not now living will come with it? But what will they have that we don’t? On the other hand, there are philosophical insights that have been gained in recent times, and I have as much access to them as anyone else; so, what I believe about the relevant questions is likely to be correct, or at least eminently defensible. Bernard was wrong if he thought that I mistakenly believed myself to have come up with these insights—that is demonstrably false. But I share them, like numerous others. The essential point is that no one I know (including myself) is anywhere near solving the mind-body problem, so it is not absurd for me to hold that the problem is not within sight of a solution. It is not that there is anyone X such that X can be counted on to solve the problem. Even the great Saul Kripke, who might be thought a plausible value of “X”, declared the problem “wide open and extremely confusing”. So, Bernard was quite right about my attitude, but it wasn’t all that silly. It isn’t as if Saul qualified his remark by saying, “But I hear Colin McGinn is working on the problem, so perhaps we will get a solution in a week or two”. That would be ridiculous. In this we see the true nature of philosophical problems. It isn’t like Watson and Crick and DNA or the Higgs boson or Darwin’s theory of evolution.

A Law of Subjectivity and Objectivity

Things can be subjective or objective, and so can conceptions of things. The former distinction pertains to types of objects and concerns their proximity to the mind: are they explicitly mental or at any rate mind-dependent? In this use, we say that color is subjective but shape is not. The latter distinction pertains to how we think of objects or otherwise mentally represent them: do we specify them conceptually in terms of their relation to us? Accordingly, objective things can have subjective modes of conception, and subjective things can have objective modes of conception. Here I will state a law that may seem paradoxical but is really truistic: the more subjective a conception is the more objective its object is apt to be, and the more objective a conception is the more subjective its object is apt to be. Objective things characteristically have subjective representations, and subjective things characteristically have objective representations. This sounds paradoxical because one would think the subjectivity or objectivity of an object would carry over to its mode of conception. But that is to confuse the two uses of “subjective” and “objective”, which have two meanings—objectual and conceptual. The reason the law is a truism is that the further an object is from the knowing mind the more we have to rely on its relations to our subjectivity (in the objectual sense) to get it in our sights; and the closer an object is to the knowing mind the less we have to rely on such indirect methods to get at the object conceptually.  For example, our thoughts about galaxies rely on our subjective modes of perceiving them, i.e., our visual experience; but our thoughts about our pains can skip all this and go straight to the pain, not relying on our senses. We know about pain directly, as it is in itself, but we know about galaxies indirectly, as they appear to us. The physical world is subjectively represented (initially at least, and maybe always), while the mental world is objectively represented in that we have immediate access to what it objectively and intrinsically is. Intuitively, we see the subjective appearance of galaxies from a certain position in space, but we see the real objective thing in the case of pain and other mental phenomena. We know galaxies from our particular point of view (from Earth, with our sense of sight), while we know pains from no point of view, but directly, and just as they are (by introspection). This is why the law I stated holds: the more out there an object is the more we rely on what is in here to conceive of it (the germ of truth in empiricism), but the more in here the object is the less we have to rely on anything indirect and peculiar to us, and can just go right to the object (hence infallibility doctrines of self-knowledge).

Granted that we have hold of a law here, we can ask how it applies elsewhere, in particular to mathematics and ethics. For if those subject matters are subjectively represented, then they must obey the law and be objective in themselves. But if they are conceived in a demonstrably objective manner, we can infer that they are subjective in the objectual sense: objective epistemology, therefore subjective ontology. If we have a murky humanly-relative conception of numbers, say, then numbers will be objective in the objectual sense—possibly Platonic. But if we conceive of numbers just as they are in themselves, with no subjective intrusion or mediation, then, according to the law, they should be closer to the mind, or even be mental entities. We can infer something like Platonism from conceptual subjectivity and something like intuitionism from conceptual objectivity. That is, we can do this unless the law breaks down in the case of mathematics. Similarly for ethics: if our ethical conceptions are subjective (subject-centered), we can infer objectivity in the objects; but if they have the marks of conceptual objectivity, then ethical subjectivism about the objects would appear to be indicated. If we think of the good relatively to our particular point of view, invoking our own psychology, then ethical truth is apt to be objectively determined; but if we think of it directly, and hence objectively, then ethical subjectivism would appear indicated. Suppose we think of the good as what produces the emotion of approval in us, but not so for all beings that think ethically; then we can argue that ethical fact must transcend our modes of mental representation. We must be thinking of the good in a subjective manner, perhaps peculiar to our species. The contingent mode of presentation of ethical facts that we employ must fall short of the reality in question. By contrast, if our mode of thinking is objectively true to the nature of the ethical facts, then those facts must be subjective in nature (in the objectual sense). If the good really is just what gives us pleasure, say, then we know its nature, because we know what pleasure is and what things produce it. But if the good is something epistemically remote from us, then it must transcend our subjective modes of apprehension (as Plato thought). Subjectivity in the conception implies objectivity in the object (assuming it has an object); and objectivity in the conception implies subjectivity in the object. So, it looks as if all we need to do to settle the question of the objectivity of ethics is to find out whether we think subjectively or objectively about it; and similarly for mathematics, mutatis mutandis. The law will do the rest.

So: do we think about moral value subjectively or objectively, and similarly for mathematics? Do we think about these things from our point of view or from the object’s point of view? The trouble is that it is hard to see how to answer this without having an independent grasp of the ontology in question, as we do in the case of physical objects (we call it “physics”). We can’t determine the epistemology without already knowing the ontology, i.e., whether the objects are subjective or objective in the objectual sense. We might find signs of subjectivity, such as variation in the manner of conception, but these can be interpreted as evidence of relativism; we can’t report that the conception comes apart from the facts, which are uniform. I myself believe that individual or cultural variation in ethical conceptions indicates that people have only a subjective grasp pf ethical values, which themselves exist universally and objectively; but it is hard to prove this (we don’t have an ethical analogue of physics). All we can say is that if ethical conceptions are subjective in the sense explained, and thus analogous to conceptions of the physical world, then ethics is objective on the objectual sense, i.e., not constituted by psychological facts. For if it were so constituted, there would not be this kind of variation; then we could affirm ethical subjectivism with regard to subject matter. (People don’t differ about what pain is because its nature is so evident.) Still, the law applies to the ethical case, though it is hard to use it to argue for one position or the other. If people rely on their own nature to anchor their moral thinking, perforce as it were, then moral values must be objective in the objectual sense, i.e., not “in the mind”. If people referred to the good as “the (non-mental) cause of these feelings”, then we would have reason to suppose that the good is more than the feelings thus referred to. But we don’t have any clear reason to suppose that they do. We must therefore rest content with the conditional claim. That is a non-trivial result.[1]

[1] Needless to say, these are very difficult and obscure matters, over which I have lightly skated.