¹ Hume, D., An Enquiry Concerning Human Understanding, Beauchamp, T. (ed.) (Oxford: Oxford University Press, 1748/1999), section IV.CrossRefGoogle Scholar

² Hume, D., An Enquiry Concerning Human Understanding, Beauchamp, T. (ed.) (Oxford: Oxford University Press, 1748/1999), sections IV, V.CrossRefGoogle Scholar

³ Goodman, N., Fact, Fiction and Forecast, 4th ed. (Indianapolis: Bobbs-Merrill, 1983), 68–69.Google Scholar

⁴ Hempel, C., Aspects of Scientific Explanation (New York: Free Press, 1965), 10–13.Google Scholar

⁵ Goodman, N., Fact, Fiction and Forecast, 4th ed. (Indianapolis: Bobbs-Merrill, 1983), chapter III.Google Scholar

⁶ Hempel, C., Aspects of Scientific Explanation (New York: Free Press, 1965), 14–20.Google Scholar

⁷ Goodman, N., Fact, Fiction and Forecast, 4th ed. (Indianapolis: Bobbs-Merrill, 1983), 71.Google Scholar

⁸ Cf. Howson, C. & Urbach, P., Scientific Reasoning: The Bayesian Approach (La Salle: Open Court, 1989).Google Scholar

⁹ Cf. Earman, J., Bayes or Bust? A Critical Examination of Bayesian Confirmation Theory (Cambridge, MA: MIT Press, 1992), chapter 5.Google Scholar

¹⁰ Howson, C., Hume's Problem: Induction and the Justification of Belief (Oxford: Oxford University Press, 2000), 239–240.CrossRefGoogle Scholar

¹¹ Howson, C., Hume's Problem: Induction and the Justification of Belief (Oxford: Oxford University Press, 2000), chapter 7.CrossRefGoogle Scholar

¹² Quine, W. V., ‘Natural Kinds’, in Ontological Relativity (New York: Columbia University Press, 1969), 114–116.Google Scholar

¹³ Mill, J. S., A System of Logic, 8th ed., London: Longmans, Green and Co., 1904) section III.VIII.2.Google Scholar

¹⁴ Dretske, F., ‘Conclusive Reasons’, Australasian Journal of Philosophy 67 (1971), 1–22Google Scholar; Nozick, R., Philosophical Explanations (Cambridge, MA: Harvard University Press, 1981), 172–196.Google Scholar

¹⁵ Nozick, R., Philosophical Explanations (Cambridge, MA: Harvard University Press, 1981), 248.Google Scholar

¹⁶ There is more work to be done to articulate how this probability is to be understood, but I would like to capture the idea that confirmation in its tracking element is a matter of degree.

¹⁷ This way of putting the point was given to me by Paul Dicken.

¹⁸ As Tim Williamson pointed out to me, this shows why it is not fully adequate to construe the tracking condition as saying simply that if the hypothesis had been false the evidence would have been different. For suppose that the hypothesis is an arbitrary conjunction of the observed and the unobserved, where the observed conjunct is more ‘fragile’ or contingent than the unobserved conjunct. In this case the simple tracking condition would be satisfied even though there is no confirmation. What is required I think is that the observed track the unobserved, which is not the case for this conjunction.

¹⁹ Cf. Lipton, P., Inference to the Best Explanation 2nd edition (London: Routledge, 2004).Google Scholar

²⁰ Cf. Chakravartty, A., ‘Semirealism’, Studies in History and Philosophy of Science 29 (1998), 391–408.CrossRefGoogle Scholar

²¹ Cf. Woodward, J., ‘Realism about Laws’, Erkenntnis 36 (1992), 200.CrossRefGoogle Scholar

²² Cf. Lewis, D., ‘Counterfactual Dependence and Time's Arrow’, in Philosophical Papers, Vol. II (New York: Oxford University Press, 1986), 32–66.Google Scholar

²³ Cf. Kripke, S., Naming and Necessity (Oxford: Blackwell, 1980), Lecture III.Google Scholar

²⁴ I am grateful to audiences at Oxford University, Cambridge University, University of East Anglia, Leeds University, University of Aarhus, and the Royal Institute of Philosophy, and for helpful reactions to my talks about ravens, and to Alex Broadbent and Paul Dicken for comments on a draft paper.