It's generally well-known that Kant was responding, amongst other influences, to Hume's critique of the empirical method on purely logical grounds. One could consider him as a modern-day Pyrrhonniste.

The problem of induction is sourced from a brief argument in Hume's Treatise, but the SEP shows a discussion in thoroughly analytic terms - Popper and Carnap are mentioned, for example.

What was Kant's response to the same question, if he in fact did respond to it? It seems likely that a response could be fashioned out of his philosophy on the basis of his categories as pure concepts of his understanding.

Question: How does Kant, or Kantians or neo-Kantians solve or understand the problem of induction?

  • 1
    Am I wrong to assume that the SEP answers this here: plato.stanford.edu/entries/kant-hume-causality?
    – user3164
    Jun 23, 2014 at 19:59
  • @Watson: I think it does:"It is in precisely this way that Kant thinks that he has an answer to Hume's skeptical problem of induction: the problem, in Kant's terms, of grounding the transition from merely “comparative” to “strict universality”. Jun 23, 2014 at 20:08
  • Well, there you go then. :)
    – user3164
    Jun 23, 2014 at 20:10
  • @Watson: its pretty dense, but luckily PVJ has done a precis below. Jun 23, 2014 at 20:11
  • I think Kant states that induction is unreliable, but not necessarily invalid and the scientific method holds despite the unreliability. Jun 15, 2019 at 17:25

2 Answers 2


In short, Kant's answer is that 'causality' isn't, contra Hume, merely constant perceived conjunction. If this is the case, then the problem of induction applies and it is not possible to infer that there is a necessary connection between a cause and its effect. Instead, Kant argues that causality is an a priori concept of the faculty of understanding. Because the concept of causality a priori mediates our experience of the world it isn't a purely subjective matter, as Hume claimed. The categories of understanding, among which 'Causality and Dependence', a priori structure our experience of the world and thus license the idea of necessary connection.

That said, it should be noted Kant doesn't deny that there are causal laws which lack the necessary character of 'pure' causal laws.

Since you mentioned the SEP, you might want to have a look at the entry on Kant and Hume on Causality. It discusses the problem you want to address in much more detail.

  • thanks, @Watson has already mentioned that particular reference. Jun 23, 2014 at 20:10
  • @MoziburUllah You're right, I hadn't seen the comment before posting!
    – PVJ
    Jun 23, 2014 at 20:41

The answer above (by @PVJ), although broadly correct, is too imprecise to understand the depth of Kant's answer. It, of course, involves causality, but not in a banal manner. For Kant it doesn't suffice to show that categories are applicable to experience (which the Transcendental Deduction of the Categories achieves). It also requires proving the principle of Uniformity of Nature (UN) which is the job of the Analogies of Experience.

Strictly speaking, we will only be discussing first two analogies. The third defines absolute simultaneity using bicausal connection which is irrelevant for the proof of UN. These are the first two analogies of experience:

  1. First Analogy of Experience: All objects of our spatio-temporal intuition are articulated into substance and state, where the quantity of substance is preserved in time (from which conservation of matter later follows).
  2. Second Analogy of Experience: All change of state has a cause, i.e. there always exists a (counterfactual-supporting) rule describing a necessary connection between subsequent states of substances. This is equivalent to Leibnizian principle of sufficient reason.

Most commentaries on Kant's proofs are absolutely horrendous. I'd reccommend reading the relevant sections of the Critique of Pure Reason yourself, if you're interested.

NOTE: These continue to be valid under Einsteinian physics but their status is more problematic in quantum mechanics.

These two principles are "sufficient" (actually, Kant's proof involves some unacknowledged assumptions beyond the two principles mentioned, which makes it fallacious... but Kant thinks it is sufficient) to establish the principle of inertia, i.e. Newton's first law, which Kant later does in his Metaphysical Foundations of Natural Science. This is because the state of substance doesn't depend on its spatio-temporal location (obviously, since they cannot ever be perceived, i.e. they have their empirical determinacy only due to objects in space and time), so velocity is a state, but position isn't. This principle is essentially a special case of the UN, as formulated, for example, by James Clerk Maxwell (the physicist) [cited after E. Cassirer, Determinism and indeterminism in modern physics, p. 42]:

The difference between one event and another does not depend on the mere difference of the times or the places at which they occur, but only on differences in the nature, configuration or motion of the bodies concerned.

...since velocity is only one state, among many, of a substance. The principle of sufficient reason together with the first analogy guarantee that there is always a sufficient reason for change of substance deriving solely from empirically perceivable circumstances. This, in principle, guarantees that it is possible to give a comprehensive, causal account of nature, which is what Hume doubted.

Most modern philosophers, beggining with the logical empiricists, including (perhaps, especially) Kant scholars, sadly, don't appreciate Kant's insights.

You must log in to answer this question.

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