Certainty, Causality, and Knowledge

Certainty, Causality, and Knowledge

I am going to argue that our concept of knowledge is essentially contradictory. This fact is written into the history of epistemology. It lurks just below the surface. It is both surprising and predictable, given how we use the concept. Arguably, knowledge is logically impossible–conceptually, by definition. We can still usefully employ the concept, but don’t expect it to make sense. I am well aware that this will strike the reader as wildly paradoxical: how can the concept of knowledge (knowledge itself) be contradictory? It’s not that hard, as we shall see, though certainly contrary to long-held assumptions. A cursory analysis of the concept will reveal its contradictory nature.

Let’s get right to the point: knowledge is conceived as involving both certainty and causality, but these are in tension with each other. There are two components to the concept, but they cannot coexist—they pull in different directions. I am talking about ideal grade-one knowledge itself not what we are prepared to say for pragmatic reasons—the essence not the extraneous.[1] Consider what we normally take to be the paradigm of knowledge, before philosophy has got to work on it, viz. perceptual knowledge, particularly visual knowledge. I see a red bird perched on my car and I have the corresponding impression that there is a red bird there. I am quite sure there is a red bird on my car and the bird itself causes me to have this certain knowledge: I have inward certainty and external causality—what more could you ask? I am not in doubt (there is no reason to be) and my visual state is not caused by anything other than a bird on a car. The world causes certainty in my mind by impinging on my mind; it doesn’t come from my own malfunctioning brain, and it doesn’t consist in a state of doubt. Knowledge at its finest is certainty plus causation (by the object of knowledge). We can argue details and subtle points, but that’s the essence of the matter.

But these two conditions do not exist in harmony: if there is causality, there can’t be certainty; and if there is certainty, there can’t be causality. For consider: causality implies contingency, but contingency is not compatible with certainty. If there is a causal connection between world and mind, then it might break down—I might be hallucinating or be a brain in a vat. Then I cannot be certain. If I am (justifiably) certain, there cannot be causality; so, if there is, I can’t be certain. But then, I don’t know. On the other hand, in cases where there is no causality, I can be certain—as in knowledge of mathematics, logic, ethics, and analytic truth. But then, there is no causality, and knowledge requires causality; hence the nagging suspicion that these cases are not really cases of knowledge—for they lack an essential feature. We thus have doctrines of mere empty tautology, conventionalism, non-cognitivism, formalism, pragmatism, and the like. In this way we retain certainty, but at the cost of departing from the paradigm of perceptual knowledge; at best this “knowledge” is second-rate, dubiously so-called, just a manner of speaking. The reason there is certainty is precisely that there is no contingent causality; but if there is causality, there can’t be certainty. Hence, there is no knowledge proper. Causality lets in skepticism, which is incompatible with knowledge; while certainty undermines causality—how you be certain of something only contingently connected to your state of certainty? Thus, certainty precludes causality and causality precludes certainty; so, knowledge is impossible. It requires both, by definition, but they are not compatible. If knowledge is an effect, it can’t be certain; but if it is not an effect, it can’t be causally dependent on reality. There cannot, then, be knowledge of the external material world, and there cannot be knowledge of the abstract world of mathematics, logic, and ethics. Not in the classic ordinary sense anyway; we have to water it down at best.

It might be protested that there are exceptions to this sweeping conclusion, namely self-knowledge and knowledge of other minds. Don’t these harmoniously combine certainty and causality, thus producing knowledge in the fullest sense? No, they don’t, because we don’t have certainty about other minds, notoriously so; and self-knowledge is not based on any recognizable causal mechanism. Thus, it is doubtful that we have knowledge of other minds at all, or if we do it is substandard; and causal self-knowledge is problematic at best, and dubiously compatible with introspective certainty. How do mental states cause beliefs about themselves, or perceptions of themselves? How can we have certainty about our own mind if the causal relation is contingent? This is why some philosophers have denied that we have self-knowledge at all. And it is hardly a central case of knowledge, following the pattern of the perceptual case. The problem is that the paradigm cases of knowledge fail the test of combining certainty and causality in equal measure. The two attributes pull apart instead of happily coexisting together. The epistemic necessity of certainty is not compatible with the metaphysical contingency of causality. Something has to give: certainty in the case of the material world, and causality in the case of the abstract and mental worlds. Our ordinary concept of knowledge tells us that knowledge is both certain and causally dependent on the object of knowledge; but these are contradictory requirements, since causal contingency is incompatible with the necessity implied by certainty (epistemic necessity). Causal connections are never certain, so knowledge can’t be a causal connection that is simultaneously certain. Knowledge is therefore impossible au fond. We can manufacture weaker notions of knowledge (“knowledge”), but we can’t get the real thing as we naively think of it. The two concepts, certainty and causality, just don’t make compatible bedfellows—each wants to kick the other out of bed. You can have one or the other, but you can’t have both joined in a single concept. We should really have two concepts not one: certainty, on the one hand, and causal epistemic dependence, on the other. We are certain about some things (e.g., mathematics), but not causally dependent on them; and we are causally dependent on some, but not certain (e.g., the material world). These are not two types of a single category of knowledge, construed as an epistemic natural kind, but different concepts altogether; the concept we have is an unholy alliance. There is no such fact in the world, though the word has its uses. The concept of knowledge is really a mind-fuck (in the technical sense).[2]

[1] See my “Perceptual Knowledge” and “Non-Perceptual Knowledge” among other works. It is important not to be taken in by the promiscuous use we make of the word “know” in ordinary language. Look at the facts not the speech acts. Speech follows its own rules, which are not the rules of reality. Analytical epistemology has been far too obsessed with what we say.

[2] See my learned treatise Mindfucking (2008) for an elaboration. Yes, I know I will be accused of doing just that in this incendiary little essay, but I think mature reflection will incline the reader to see the point of my indictment of the concept of knowledge. The concept is trying to do two jobs at the same time, but they are not unified; they answer to different authorities—Cartesian and naturalistic, you might say. Knowledge is conceived as both a causal relation in nature and an inner state of the soul (somewhat spiritual). It has religious connotations as well as naturalistic affinities. Like a tree falling on a church: the inner sanctum in a world of impinging crashes and bangs. The two don’t happily combine, still less coalesce. Colorless green ideas and all that. (Meaning is not dissimilar.) Concepts evolve for practical purposes, and they are not always as logical as one might wish. To know is to be inwardly certain and subject to causal contingency: but these are inconsistent conditions.