1

Dustin Locke

argues that assuming that p is (rather than engage in probabilistic reasoning regarding p) is rationally permissible when it is practically certain. So understood, moral or practical certainty is epistemic in nature, though it seems to fall short of epistemic certainty.

According to Locke, when is something practically certain? Can something with a lot (your soul, mine) resting on the result ever be "practically certain"? If so, can we be practically certain in some ways (there is good evidence and something like a consensus) but not in others (you or the community may be unreliable). When is enough, enough, so that we no longer need to assess the alternative, let alone let it figure in our moral/practical decisions?

  • I agree with your perplexity... Id we agree that humans cannot have "absolute" certainty, also in scientific context we have to rely on "plausibility" (also in this case we fall short of "probabilistic reasoning"). What is "plausible"? what has been agreed upon by the scientific community. – Mauro ALLEGRANZA Mar 20 '23 at 10:19
  • we make few moral decisions grounded in scientific theories @MauroALLEGRANZA or at a minimum we don't routinely have to apply them (i shouldn't jump off here, gravity) –  Mar 20 '23 at 10:20
  • I've tried a parallel between the two context: if we follow it, what is "practical certainty"? What is agreed upon by the community... – Mauro ALLEGRANZA Mar 20 '23 at 10:22
  • i suppose an epistemic consensus coupled with the will to believe it may always suffice @MauroALLEGRANZA –  Mar 20 '23 at 10:34

2 Answers2

0

The answer of Locke is given in the abstract of the paper. We are practically certain if and only if

the actual epistemic probability that p differs from epistemic certainty that p only in ways that are irrelevant to the decision one currently face.

I.e. if our decision is based on knowledge/expectations/perceived certainty we may well be wrong in our assumptions and it is asked too much to base decisions on an "absolute" certainty that cannot be achieved anyways. As long as the points and extent of that wrongness would not make a difference for the outcome of our decision process, we can and should still call that practical certainty since for all practical considerations, we are certain enough, even if we are wrong, strictly speaking.

This does by no means address the epistemic problem that we cannot know the actual probabilities of p and whether the difference would change our decision and, by extension, we cannot know whether and at which point we maybe already are, indeed, practically certain, ie. this notion of practical certainty has no bearing on the practical problems faced in decision processes. Therefore, ironically, this notion is absolutely impractical because we are uncertain in our decisions only if our epistemic uncertainty is such that it could make a difference for our decision if actual probabilities differ.

In other words: Well, duh, if we are uncertain in our decisions we are practically uncertain because we are epistemically uncertain. But we also could be practically certain (in Locke's sense) without knowing it, so it actually boils down to an epistemic problem no matter from which angle we look at it. And that is exactly the point Locke wants to stress in that paper.

Philip Klöcking
  • 13,015
  • 3
  • 39
  • 69
  • did you read the paper? the abstract does not answer the question, because it doesn't say what 'relevance' is, and it's not clear from it that it is never rationally permissible to assume that p if doing so would make any difference to our actions (which i think is what you mean) –  Mar 22 '23 at 03:47
  • 1
    @zero But yeah, that is exactly the point that he makes: Every reasonable concept of *practical* certainty boils down to *epistemic* certainty. He establishes a reasonable concept of practical certainty and investigates the consequences to hammer that point home. – Philip Klöcking Mar 22 '23 at 05:41
  • ok, thanks for the answer then –  Mar 22 '23 at 07:18
0

Conceptually, at least, a degree of certainty can be expressed on a continuous scale from zero to one. There is no fixed point on that line that can be said to divide practical certainty from lesser degrees of certainty, as the difference depends on many factors, including personal judgement. As a general principle you might say that you are practically certain of something if you consider any related uncertainty to be purely academic, but that is just shifting the problem to being one of clarifying what one means by purely academic. One way to address that question is to ask whether the residual uncertainty makes a practical difference- if not, then you are practically certain. The possibility that I might be able to swim the English Channel is so remote that I would not waste time contemplating it, so I am practically certain I would fail if I tried. However, if I were offered a billion pounds to swim the Channel, I might want to re-examine the possibilities, so one's view of practical certainty can vary depending on the circumstances as well as on the question itself.

Marco Ocram
  • 8,686
  • 1
  • 8
  • 28