The Humean analysis of causation would hold that there is no actual relation between two events (being 'cause' and 'effect'), and that any sense of 'causal powers' can be reduced to talk about the presence or absence of brute facts of events. But if this is the case, why is it that we don't observe more causeless events, or events that occur without any accompanying event? Why does the overwhelming majority of our observations of change display regularities if there is nothing to such regularities like a relation between the events therein, that makes them be so? Why, in short, are there always causes to effects when we seek after them, if they don't need to be there?

  • This seems to be part of a series you're working through on Hume and induction (philosophy.stackexchange.com/questions/29097/… / philosophy.stackexchange.com/questions/28621/… / ).
    – virmaior
    Commented Nov 9, 2015 at 4:37
  • 3
    I think your first sentence is wrong as interpretation of Hume, specifically I think you're confusing his skepticism which denies our ability to infer cause on epistemic level with a blanket assertion that there is "no actual relation between two events". Do you take Hume to be denying cause on a metaphysical level or denying our ability to infer cause on an epistemic level? (If metaphysical, can you adduce some textual evidence to support that?)
    – virmaior
    Commented Nov 9, 2015 at 4:39
  • From an extreme point of view, if time is an aspect of animal minds, and not of reality, then we can say there is no necessity to the succession of cause and effect. We will observe cause and effect, because that is our viewpoint on the world, and not because it is necessary logically.
    – user9166
    Commented Nov 9, 2015 at 23:07
  • @virmaior That's a worthy question. I'm not necessarily taking Hume's work in a certain 'way' (be it metaphysical or epistemic). I'm only raising certain questions I have about his philosophy, which I admittedly believe to be assumed perhaps too often in modern philosophy.
    – Chosen One
    Commented Nov 13, 2015 at 23:00

1 Answer 1


Hume is agnostic about causality, only asserting that no conclusive evidence for it exists, but let's say that he would humor you and defend the negative position. First, by his analysis we do not know if observed causes are really causes, and even assuming we are right about that there is an obvious selection bias at play. Our observing evolved to aid our survival, we benefit from focus on repeatable patterns that can be relied upon. We do not focus on why the temperature is 74.53 degrees and not 74.49, or why dropped spoon landed where it did, or why somebody has a birthmark. That we could find out if we looked is merely an optimistic extrapolation. Eventually it runs into the issue of absolute precision for systems sensitive to perturbations, and there can be no precision beyond quantum limits. Motion of planets is an exception, not a rule, and in most cases we test causation in tightly controlled experiments, in very improbable environments. But even then we never observe the supposed cause determine its supposed effect in every detail, not even once. Sufficient cause is a pure idealization.

But even aside from that Hume could say that no explanation is owed here. You seem to be running a particular case of Putnam's no miracles argument: it would be a miracle if all this regularity occurred without the underlying causation. Well, let's assume that regularity is a very good test for causation, namely the probabilities of false negatives (observing no regularity conditioned on causation) and false positives (observing regularity conditioned on no causation) are very low. What can we say about probability of causation conditioned on observed regularity? Intuitively, it seems that it should be very high, and this intuition explains the appeal of no miracles arguments. But this intuition is flawed, in fact any value, including arbitrarily small one, is consistent with the assumptions. The reason is that we do not know the "base rate", the prior probability of causation before observation, nor do we have any way of making sense of it. For that matter, we do not even have a good reason for asserting low rate of false positives, because causation is certainly not the only mechanism that produces regularity.

Since you are into Hume's reasoning and counterarguments you may find Howson's Hume's Problem an interesting read. In particular, he discusses the base rate fallacy as it applies to causality in detail.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .