Hume pp 38-60 | Cite as

Causation and Induction

  • Terence Penelhum
Part of the Philosophers in Perspective book series (PHPE)


The problem of induction (the problem of the epistemological status of inferences from experience) and the problem of causation (the problem of the meaning of individual causal judgements) seem to be distinct from one another. Hume, however, discusses them together at all times. This is because he holds that inference from experience is most commonly inference from effects to causes or from causes to effects; and, far more fundamentally, because he holds that the nature of causal judgements cannot be understood until we recognise the way in which our inferences from experience are actually performed and what leads us to make them in the way we do. General questions about causation set the framework for his investigation of induction; but his answers to those questions presuppose theses about the justification and origin of inductive inference. The intricate relationship between these two topics, both explored by Hume in unprecedented depth, have led to his being celebrated, and attacked, for a variety of alleged discoveries or errors. It is universally agreed that Hume has told us something of permanent and unchallengeable importance about each theme, but there is little common agreement as to what this is in each case. In this brief treatment I shall merely attempt to indicate the exact nature of Hume’s argument and conclusions, as we find it in the Treatise, the first Enquiry, and the Abstract.


Causal Inference Inductive Inference Causal Judgement Causal Principle Billiard Ball 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Notes and References

  1. 1.
    On Hume’s views about volitions, see Chapter 6.Google Scholar
  2. 2.
    These contentions are of some importance in Hume’s strategy. His major target is the view that we have to ascribe unobservable powers to natural objects in order to account for their causal properties. But he is also seeking to refute the Occasionalists, who argued that connections between mental and physical phenomena could only be understood as miraculous interventions of God; and, I think, Berkeley, who held that physical causation could only be understood on the model of mental causation, so that the perceptions that constitute natural sequences in the physical world are controlled by the divine mind in the way in which the ideas of our imagination are controlled by human minds. Hume tries to undermine all these theories together by arguing that both physical and mental causation are wholly familiar, and that any mystery attached to one should be attached equally to the other.Google Scholar
  3. 3.
    See Part Two of Probability and Hume’s Inductive Scepticism. I am much indebted to this remarkable book, which no serious student of this topic can afford not to read.Google Scholar
  4. 4.
    Ibid., p. 91.Google Scholar
  5. 5.
    Ibid., p. 97.Google Scholar
  6. 6.
    This is the burden of arguments like those of Norman Malcolm in ‘The Verification Argument’ in Philosophical Analysis, ed. Max Black (Cornell University Press, Ithaca, New York, 1950).Google Scholar
  7. 7.
    See P.F. Strawson, Introduction to Logical Theory, chapter 9 (Methuen, 1952).Google Scholar
  8. 8.
    See e.g. F.L. Will, ‘Will the Future be like the Past?’ in Logic and Language, ed. Antony Flew, Second Series (Blackwell, Oxford, 1953).Google Scholar
  9. 9.
    This is the point at which to mention that in both accounts Hume proceeds on the assumption of the existence of external objects — and also on that of a continuing self which observes them. The sceptical doubts about these topics play a wholly trivial part in the Enquiry, where it is striking that the apparatus of impressions and ideas is only brought into play when Hume turns to the idea of necessary connection. In the Treatise the perplexing discussions of these topics come after the discussions of causation. Price has pointed out that if it is true that our ideas of cause and effect are the product of our observation of constant conjunctions, it is far from clear that such conjunctions are to be met with among our impressions, whatever we may say about physical objects. (‘It is very doubtful whether there are any constant conjunctions of sense-impressions. If we try to formulate any one drowsy nod or blink will refute us.’ (H.H. Price, Hume’s Theory of the External World, Clarendon Press, Oxford, 1940, p. 7). There seem to be three possible alternatives open to us if we attempt to supply the connections that Hume does not make. (1) We could say that the conjunctions are indeed between impressions, and are thought to be constant when they are not. (2) We could say that the conjunctions are between physical objects, so that the imaginative construction of the external world that Hume describes in Part IV, Section II, is completed before the mechanisms of causal thought can begin. (3) We could say that the development of one set of habits proceeds alongside the development of the other, so that we cannot really separate the belief in constant conjunction from the belief in the distinct and continued existence of objects. Hume avoids this issue by his piecemeal treatment; but it remains. Although it is speculative as well as difficult, J would suggest that only (3) is even moderately plausible, and that it could be worked out without formal inconsistency. Had Hume himself attempted it, his epistemological psychology would have been even closer to that of Kant than it is.Google Scholar
  10. 10.
    This is why causal judgements can be contested. If Hume thought the idea of causation were wholly derived from an internal impression, he would be committed to the incorrigibility of causal statements (see Price, ‘The Permanent Significance of Hume’s Philosophy’). The reason that the subjectivity of our idea of power or necessity does not prevent genuine empirical disagreement about causal judgements is that they commit those who make them to the objective characteristics of contiguity, succession, and constant conjunction. There is a close and deliberate analogy here with Hume’s analysis of moral evaluations. Although he adopts an ‘emotive’ interpretation of them, he also holds them to be occasioned by a supposedly objective survey of the actions to which we apply them, so that the judgements in which we express them can be genuinely contradicted. See Chapter 7.Google Scholar
  11. 11.
    See Flew, Hume’s Philosophy of Belief p. 127.Google Scholar
  12. 12.
    This seems especially evident in considering historical causes. See William Dray, Laws and Explanations in History (Oxford University Press, 1957).Google Scholar
  13. 13.
    There is a short passage in the Section ‘Of the Probability of Causes’ in the Treatise, reproduced in Section VIII of the Enquiry, ‘Of Liberty and Necessity’, in which Hume discusses the phenomenon of ‘contrariety’ in our observations. He is concerned with cases where a cause is apparently not followed by its usual effect, and wishes to explain them in a manner that preserves our commitment to the universality of the connection between causes and effects. He says that in such cases the expert will detect an additional ‘contrary’ cause which the layman overlooks — such as a speck of dust which prevents the force of the spring from moving the hands of a clock. Such discoveries preserve the principle that causes are regularly followed by their effects by showing that there are other causes present when they seem not to do so. MacNabb suggests (David Hume, pp. 56–7) that this passage is Hume’s account of why we believe in the Causal Principle. If it is, it is a failure, for reasons additional to those MacNabb offers. An explanation of why we believe that if a cause appears to fail there is another cause interfering is not an explanation of why we believe that every event has some cause or other. The former principle entails that every cause, left to itself, will always have the same effect. The Causal Principle does not entail this, nor is it entailed by it.Google Scholar

Copyright information

© Terence Penelhum 1975

Authors and Affiliations

  • Terence Penelhum
    • 1
  1. 1.University of CalgaryCanada

Personalised recommendations