Knowledge and Time

Knowledge and Time

I shall make some remarks about a topic neglected by epistemologists—the relationship between knowledge and time, particularly future time. The relationship is not simple or easily grasped; there is a reason for the neglect. I will try to keep it as uncontroversial as possible; this is to be preliminary groundwork. Truisms not breakthroughs. The big question is this: Can we know, perceive, refer to, or have justified beliefs about the past, the present, and the future? I think it would be generally agreed that we can and that we have these relations to the past and the present—but to the future not so much. There might be some argument about whether we perceive the past, or even whether we perceive the present: with what sense do I perceive what happened yesterday, and what about the time lag between the event perceived and the perception of it? Don’t we infer the past from our present memories of it, and don’t we really perceive an event in the past given the time it takes for a perception to be formed? But in the case of the future there is really no doubt that we don’t perceive it; the question has been whether we can still know it. For perception requires causation and causation never runs from the future to past: what happens tomorrow cannot cause what happens today, in the mind or elsewhere. This familiar point is surely correct and scarcely disputable, but it needs to be fully absorbed: we cannot be acquainted with the future; we cannot directly apprehend it; we cannot be consciously aware of it. We cannot know it in the perceptual sense; we can only know it, if at all, by inference. It is necessarily imperceptible. In the case of the past, we can know it directly (by memory of past perceptual encounters) or by inference from these, but we can never know the future in this direct way—because we can never perceive it. You can’t now see what will happen tomorrow, no matter how much you strain your eyes—light doesn’t travel into them from the future. So, this basic source of knowledge is completely unavailable to us, in principle and forever. Given that perceptual knowledge is the basis of all knowledge, the question must then arise as to whether we can know anything about the future. Isn’t it just too cut off to be known? Aren’t we limited to mere guesswork, chance truth, accidental match? And these are not instances of knowledge. Shouldn’t we be absolute agnostics about the future? What you can never see you can never know. The case is even worse than other minds, because at least in that case we have causal relations between object and subject: the other’s mental states cause my mental states via his behavior, which I see with my own eyes. But the states of the future can never cause mental states in me, or any other present states. The past is not an effect of the future, as the future is an effect of the past. Thus, I cannot know the future by perceiving it, or any part or sliver of it, however indirectly or remotely. I am completely shut off from it. We are separated by an epistemological wall, based on a metaphysical necessity.

But it doesn’t follow that we can’t have true justified beliefs about the future: so, can we? Let me first note that if we can it won’t follow that we can have knowledge of the future (knowing-that). It would follow only if knowledge is, or can be, true justified belief; but the future provides a clear counterexample to that analysis. Suppose I have a true justified belief that war will break out tomorrow: do I thereby know it will? Intuitively, no: I don’t know this fact, I only believe it.[1] I think it will and for good reason, but I don’t really know it—not like I know that there’s a cat in front of me. I am not acquainted with that future war. We would be perfectly within our rights to deny that anything about the future is ever known, even if we allow that we can have reasonable true beliefs about the future. And indeed, I venture to suggest that this is the common opinion: the future is not knowable—though it is conjecturable. You can have beliefs about it, but these don’t amount to cases of knowledge. So, the JTB analysis of knowledge is insufficient (this is a kind of Gettier case, in effect). But we can still ask whether such beliefs are ever justified, discounting the knowledge question. Now this is extremely well-trodden territory, which I don’t intend to re-tread. I will make two points about it. First, this is not the “problem of induction”: that problem is not inherently about the future; it can apply to both the past and the present. Were all past swans white and are all present swans white? The problem of induction is about generalizing from a sample to a whole population, not about inferring the future from the past. The second point is that induction is the only way we can know about the future, since we are perceptually closed to the future. We can perceive the past (perhaps we always do) and we can perceive the present (pace the time-lag argument), but we cannot as a matter of necessity perceive the future. This puts belief about the future in a much worse position than belief about the past and present, since we don’t know even what it would be to perceive the future. What would it even feel like? If an alien had such a sense, would we be able to grasp its phenomenology? On top of that we have the problem of induction itself, which strikes even regular people as problematic. True, we reflexively form expectations for the future (as Hume famously observed), but this has nothing to do with reason; we would have such reflexes whether the future resembles the past or not. Induction is notoriously difficult to justify. At best future predictions are perilous and indemonstrable. You don’t have to be a skeptic to feel that we are deeply ignorant of the future (you have never been there); indeed, this is hardly worth calling skepticism, since ordinary folk are already queasy with talk of justification and knowledge regarding the future (this is not so for the past and present). In sum: we have neither perceptual knowledge of the future nor solid justification of beliefs about the future—just instinct, conditioning, and blind faith. This is why there is no history of the future—no narrative of what will happen. One might be forgiven for supposing that the future is not a fit object of human knowledge; we just talk as if it is for pragmatic reasons. Strictly, we shouldn’t even have beliefs about the future, since belief presupposes justification; we should only have attitudes of surmise and speculation (good Popperians about what will be). At any rate, that is a position with an intelligible rationale.

This problem has always haunted science, because science purports to be predictive—and yet empirically warranted. Empiricism bases knowledge on experience, but we don’t experience the future; it ought then to be out of epistemic bounds. I have no “impression” of the future events I predict, so how can I know about them? How then can an adequate philosophy of science be empirical? Popper took a radical line; others have suggested that science doesn’t make factual predictions but is only a useful tool for getting along in the world. But it is always future-oriented and hence open to criticism from a consistent empiricist. History has no such problem—or logic and mathematics and philosophy. The epistemology of science has therefore always been under a cloud, as exceeding what can be humanly known. Hume was well aware of this (Popper made a big deal of it). Proust wrote a long book called Remembrance of Things Past but not Expectations of Things Future, because there is so little to say under the latter head; there is no madeleine of the future. This is our human epistemic predicament and the source of much of our anxiety (it isn’t only death). We can describe the future and fear it, but we can’t know it—not really. We can know (sic) the future only by using the past in conjunction with induction, but induction is eminently questionable, so we are in perpetual doubt about the future. The future itself is terra incognita. At best we go on external signs of it.[2]

[1] See my “Perceptual Knowledge” and “Non-Perceptual Knowledge”.

[2] The idea of the crystal ball is instructive: the only way the future can be genuinely known is by being seen in the shape of a transparent sphere—a portal to the future. Time travel is similar (and equally fictional): we can know the future only by going there and clapping our eyes on it—up close, directly, under our nose. Pure fantasy, of course, but it feeds off our epistemic anxiety concerning the future: the future is the unknown in its purest form, outdistancing even the most remote galaxy or secretive mind. We can “know” it only by comparing it with the past—its very opposite. How could the past ever tell us about the future? Time has no patience with our intellectual limitations. The future is the twilight zone but without any light. That is the terrible truth.