Explicable Knowledge

Explicable Knowledge

Suppose I don’t know where my phone is; it could be in a number of places in the house. I look around and find it on top of my desk, thereby coming to know that my phone is on my desk. This is a paradigm case of perceptual knowledge—knowledge by means of the senses. It is characteristic of such knowledge that I also know how I know—the means, the method. I don’t just suddenly know, inexplicably. I know that I know the fact in question by using my eyes to search the house and eventually see the phone sitting on my desk. I know that I moved my body through space and let the external world act on my senses so as to produce experiences of kinds that I also know. I know that I saw the phone from a particular angle, in a certain light, at a specific time. I know what led up to the knowledge, its preconditions and procedures. I have first-person explicable knowledge—knowledge of how such knowledge is produced, in general and in particular. It is no mystery to me. It gives me confidence in the belief I formed about the whereabouts of my phone.  So it is with what we are pleased (and proud) to call empirical knowledge in general—a posteriori knowledge, if we prefer the Latin (and its associated italics).[1]

But the same is not true of all knowledge. Suppose I know that 2+2=4 or that everything is self-identical or that bachelors are not married: do I know how I know these things? I do not. I could say “By reason alone”, but that is like saying I know where my phone is “by perception”: there is no detail, no explanation, just a general formula. There is nothing corresponding to walking around the house, casting my eyes this way and that, letting the world act on my senses (I could also feel my way around if it’s dark). I don’t know how I know—in fact, there is nothing like the procedures that led to my phone knowledge. I am inclined to say I just know, inexplicably (it is “self-evident”). My knowledge of mathematics, logic, and analytic truth operates in the absence of an accompanying knowledge-how. Movements of the body are not involved; nor is environmental causality. What we are pleased (and proud) to call Pure Reason has need of none of this: it proceeds in a vacuum, as it were. There is no knowledge of how the knowledge is produced. We have a first-person blank here, while in the case of perceptual knowledge we have an embarrassment of riches.

This enables us to make the following pronouncement: a posteriori knowledge is explicable knowledge and a priori knowledge is inexplicable knowledge—that is, to the agent or subject. The first kind of knowledge is knowledgeably known, but not the second kind; it is unknowledgeably known. Thus, the latter is a mystery-to-the-knower. We don’t understand how we come to have such knowledge. It is a kind of ignorant knowledge, though this doesn’t detract from its status as knowledge. The epistemic lack is higher-order: it concerns knowledge of knowledge. Conceivably, it could be the other way round: inexplicable perceptual knowledge (perhaps a bit like blindsight) and explicable rational knowledge (but what would this be like?); in any case, as things are, this is the way things line up. Predictably, the situation provokes suspicion of the a priori: it smacks too much of unsupported intuition, mere stabs in the dark. It would be nice if we could fit it under the sensory umbrella, as covertly perception-based; then it would enjoy the procedural transparency of perceptual knowledge. Some may say there can be no such thing, or that it amounts to vacuous tautology. And one can appreciate the motivation; there is a real asymmetry here, to the seeming detriment of the a priori (but see below). The a priori already seemed suspect from an objective explanatory point of view (mysterious in its operations); now we see that its first-person epistemology leaves much to be desired. The subject doesn’t know how he knows what he knows, yet he is convinced that he knows. This kind of suspicion will spill over into neighboring areas like introspective and ethical knowledge. For here too we find no clear analogue of the perceptual case: how do I know that I am in pain or have a particular belief, and how do I know ethical propositions? Not by moving my body around and deploying my senses, not by causal interactions and sensory impressions. Rather, I just know, inexplicably. Thus, some people question whether there is really knowledge in these cases—hence expressivist theories. If the subject can’t say how he knows, he doesn’t really know; and if he doesn’t know, there is no fact to be known. A lot of philosophy hangs on the asymmetry in question.

But an irony intrudes: if perceptual knowledge comes with knowledge of how it is acquired and justified, the question must arise of whether the means we use are up to the task. Hence skepticism. Brains in vats could use the same methods, or appear to. But not so for a priori knowledge: if no means are used, then there is nothing to criticize. Thus, a priori knowledge has been traditionally regarded as exempt from skepticism—not brain-in-vattable. It allows for (justified) certainty. Since it is first-person inexplicable, we can’t ask how good the explanation really is: what is not attempted cannot be faulted. A priori knowledge may be mysterious and inexplicable, but it cannot be accused of using shoddy methods, as a posteriori knowledge can be. We have a kind of Mexican stand-off: guns drawn but no victory, because what the a priori gains in certainty it loses in explicability, while the a posteriori can claim explicability only at the price of skepticism. This has been the basic layout of epistemology since Plato’s time: empirical knowledge makes sense but is open to skepticism, while rational knowledge is hard to make sense of but is not open to skepticism. The senses are fallible, so empiricism faces skepticism; reason is not similarly fallible, so it is (relatively) immune to skepticism. But rationalism is up to its neck in mystery (first-person and third-person), while having an easier time with skepticism. There is no obvious way out of this maze, as the last two thousand years testify.[2]

[1] This view of perceptual knowledge is defended by Michael Ayers (my old supervisor) in Knowing and Seeing: Groundwork for a New Empiricism (2019). I have not had the opportunity to read this book, but I have gathered the basic thesis from the book’s blurb. It is pretty straightforward.

[2] I first wrote about this topic in my MA thesis in psychology in 1972 and have returned to it intermittently.