1. See Michael Behe Darwin's Black Box (New York NY: The Free Press, 1996). For other defences of intelligent design, see William A. Dembski Intelligent Design (Downer's Grove IL: InterVarsity Press, 1999), and Michael Behe's more recent ‘Darwin's breakdown: irreducible complexity and design at the foundations of life’, in William A. Dembski & James Kushiner (eds) Signs of Intelligence: Understanding Intelligent Design (Grand Rapids MI: Brazos Press, 2001). For a sampling of criticism of intelligent design, see Kenneth R. Miller Finding Darwin's God (New York NY: Harper Collins, 1999), 129–164, and Niall Shanks God, the Devil, and Darwin (Oxford: Oxford University Press, 2004), 160–190.

2. In this paper I will understand ‘traditional theism’ (which I will sometimes shorten to ‘theism’) in the standard way. Traditional theism is true (roughly) if and only if there exists a supernatural being which is omnipotent, omniscient, and omnibenevolent.

3. Behe Darwin's Black Box, 69–72.

4. Ibid., 78–97.

5. Ibid., 120–126.

6. Ibid., 5.

7. Both of these are situations where some contemporary philosophers and scientists have also wanted to introduce design inferences.

8. Allow me to say a bit about this direct intervention. I am assuming that we are working within a framework where laws specify how some entities will be involved in the production or generation of other entities. (This is sometimes informally called an ‘oomphy’ conception of laws.) Imagine that at least some of the fundamental laws are indeterministic (specifically, that they are probabilistic), and that the following is a sample law: 70 per cent of particles of type A will decay into particles of type B. I am supposing that this law specifies that there is an objective chance of .7 that any given particle of type A will decay into a particle of type B. In this situation, any external impetus that changes the objective chance of any type A particle's decay will count as a direct intervention. This is so even if the overall statistical pattern of decays conforms quite closely to what the law leads us to expect. Later, we will have occasion to discuss irregularity in worlds, and I take it that any interference in these sorts of objective chances will contribute to irregularity, regardless of the character of overall statistical patterns. (I also assume that laws of nature are not ‘gappy’ – there are no situations where the laws leave what should occur completely unspecified.) I am grateful to an anonymous reviewer for this journal for making me aware of the need to clarify my assumptions about laws.

9. I am understanding the probabilities in question here to be epistemic – as something like what are often called ‘degrees of belief’. Aside from assuming that these are and should be constrained in various straightforward ways by the probability calculus, I will not propose any more substantive characterization of them in the present context.

10. See Behe Darwin's Black Box, 225.

11. Here I will set aside issues about whether it is likely that a knowledgeable and powerful designer lacking in goodness might produce a product that was lacking in optimality or efficiency.

12. ID theorists who believe that God (rather than lesser supernatural beings or aliens) is the intelligent designer will typically appeal to background considerations (and hence prior probabilities) to draw this conclusion, not to confirmation advantages that the God-hypothesis has over competing alternatives with respect to the ID evidence.

13. Caveats and elaborations may be required to make sense of indeterministic quantum processes, but I will set these issues aside. I also use the terminology of ‘natural event’ rather than ‘physical event’ to leave open the possibility that the natural order involves primitive mental properties, which can enter into lawlike relationships with other properties without miraculous intervention. Since humans lie outside the mainstream of the present discussion, I am ignoring all issues related to free will for the time being.

14. I am indebted to an anonymous reviewer for this journal for suggesting that these clarifications be made.

15. It is also important to note that these conclusion about P(L/N) being low do not seem to depend on rejecting God as the designer of physical laws or the initial conditions of the universe. If the progression of the universe is indeterministic (as there is good reason to believe), and that indeterminism is of a sufficiently substantial nature (as there is also good reason to believe), then it will not be possible for God to rig things up at the start of the universe and produce the biochemical systems with any acceptable likelihood. (A possible complication, which I will not address here, is if God were to have middle knowledge of various indicative conditionals about quantum decay and so forth. For some speculations on this matter, see Michael J. Murray ‘Natural providence (or design trouble)’, Faith and Philosophy, 20 (2003), 325–326.) Even if the world is deterministic (when left to its own devices), there is still good reason to suppose that God could not produce the biochemical systems just by tinkering with the initial conditions, because of chaos considerations (and also assuming God does not create laws with various ad hoc caveats, as discussed above). For some further discussion, see Thomas Tracy ‘Evolution, divine action, and the problem of evil’, in Robert John Russell, William R. Stoeger, and Francisco J. Ayala (eds) Evolutionary and Molecular Biology (Vatican City: Vatican Observatory; Berkeley CA: Center for Theology and Natural Sciences, 1998), 529.

16. See Peter van Inwagen ‘The problem of evil, the problem of air, and the problem of silence’, in J. Tomberlin (ed.) Philosophical Perspectives, 5, Philosophy of Religion (Atascadero CA: Ridgeview Publishing, 1991). All references here will be to the reprinted (but unaltered) version in Daniel Howard-Snyder (ed.) The Evidential Argument from Evil (Bloomington IN: Indiana University Press, 1996), 151–174.

17. More on this ‘for all we know’ qualification below.

18. van Inwagen ‘The problem of evil, the problem of air, and the problem of silence’, 161.

19. Ibid., 161. van Inwagen is also worried about irregularity because it contributes to the systematic deception of higher-level sentient creatures (like humans). Because this is not an intrinsic problem with irregularity, I won't discuss it at this juncture. Suffice it to say, though, that there is considerably more to van Inwagen's distaste for massive irregularity than just its tendency to deceive.

20. Ibid., 154.

21. For a criticism of van Inwagen's general strategy of employing defences to counter evidential arguments from evil, see Almeida, M. J. ‘Refuting van Inwagen's “refutation”: evidentialism again', International Journal for Philosophy of Religion, 44 (1998), 23–29CrossRefGoogle Scholar.

22. There may be tricky issues generally in evaluating the probabilities of theodicies/defences that make claims about values (claims that if true, are necessarily true), but I will set this aside as irrelevant here. Also, for the remainder of the paper, I will not be addressing the sceptic about value commensurability. The general problem of evil and the specific manifestation of it we will be discussing are much less troublesome for such an individual.

23. Richard Swinburne Providence and the Problem of Evil (Oxford: Oxford University Press, 1998), 51.

24. See, for example, Tracy ‘Evolution, divine action, and the problem of evil’, 525; John Polkinghorne Faith of a Physicist (Princeton NJ: Princeton University Press, 1994), 85.

25. Here I ignore issues about God's ability and/or obligation to make the best of all possible worlds (if there even is such a thing). Whatever one's views on these general matters, it is not hard to generate the intuition that in this specific case, if the one option is better than the other, we would expect God to make the better world.

26. We are, of course, assuming that God does not actually do this, and then simply covers His tracks.

27. We are assuming here that ID does in fact bring about massive irregularity on a scale which is similar to that of these artificial measures taken to prevent or relieve suffering. Below, I will tackle the suggestion that there is a serious difference in quantity of irregularity. But that is a separate objection.

28. I am assuming a dualist picture here, since on a physicalist picture, it would be metaphysically impossible for an animal not to feel pain if it were in the ordinary pain-brain state, since the pain would be identical (in some fairly straightforward sense) to the brain state. Consequently, even God could not affect the mental state in this case without affecting the brain state.

29. Here I am forced to set aside consideration of options like creating the Earth five minutes ago complete with apparent memories and substantial history, since the difference between this and ID really would be a difference in quality – a giant one-time intervention vs a (perhaps large) number of smaller, less dramatic, interventions.

30. See Behe Darwin's Black Box, 227–228.

31. See, for example, Miller Finding Darwin's God, 162–163.

32. In order for this intervention by agents to constitute irregularity, we need not assume that the laws are deterministic. Even indeterministic laws of the sort standardly accepted by physicists do not permit the sorts of influences spoken of here. I am grateful to Bill Hasker for pointing out the need for such a clarification.

33. Incidentally, in the 1950 papal encyclical Humani Generis, the first document from the Catholic hierarchy endorsing the compatibility of evolution with Catholic doctrine, Pius XII warns that this compatibility does not extend to belief in the naturalistic production of the human soul. He says that God's intervention is required for the production of ‘the human soul’, presumably implying each human soul.

34. I would like to thank Todd Moody and Bill Hasker, both of whom read previous drafts of this article and suggested a number of helpful improvements. I am also grateful for the detailed comments of an anonymous reviewer for this journal, whose suggestions have helped me to clarify several points.