The application-conditions for design inferences: Why the design arguments need the help of other arguments for God's existence

International Journal for Philosophy of Religion

Volumn 57, Issue 1, 2005, Pages 1-33

  Himma, Kenneth Einar ^a  

^a SEATTLE PACIFIC UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 34247259600     PISSN: 00207047     EISSN: 15728684     Source Type: Journal    
DOI: 10.1007/s11153-004-5894-7     Document Type: Article

References (38)

1. Undecided voters are, other things being equal, more likely to vote for the first person on a ballot; thus, getting the top line in a ballot affords the candidate with an advantage.
2. Though the court declined to convict Caputo, it was primarily because the court needed more than (1) through (4) to satisfy a criminal burden of proof, which requires proof beyond a reasonable doubt; in particular, it needed a causal story as to how Caputo could have brought this about.
3. David Hume, Dialogues Concerning Natural Religion (edited with an introduction by Norman Kemp Smith), New York: Social Sciences Publishers, 1948, p. 143.
4. In contrast to the argument from analogy, William Paley's famous watchmaker argument explicitly identifies a sort of irreducible functional complexity as a reliable indicator of design: "[S]uppose I found a watch upon the ground, and it should be inquired how the watch happened to be in that place, I should hardly think . . . that, for anything I knew, the watch might have always been there. Yet why should not this answer serve for the watch as well as for [a] stone [that happened to be lying on the ground]?. . . For this reason, and for no other; viz., that, if the different parts had been differently shaped from what they are, if a different size from what they are, or placed after any other manner, or in any order than that in which they are placed, either no motion at all would have been carried on in the machine, or none which would have answered the use that is now served by it." Though frequently interpreted as an argument from analogy (see Louis P. Pojman, Philosophy of Religion (Mayfield Publishing Company, 2001), 32;
5. and Ed. L. Miller, God and Reason: An Invitation to Philosophical Theology, 2nd Edition (Prentice Hall, 1995), 72), Paley's argument is a clear precursor to the mathematically more sophisticated contemporary versions.
6. Robin Collins, "A Scientific Argument for the Existence of God," in Michael J. Murray (ed.), Reason for the Hope Within (Grand Rapids, MI: William B. Eerdmans Publishing Co., 1999), 47-75, 49-50. Hereinafter referred to as SAG.
7. Collins's argument presupposes that this is the only actually extant material universe. If the hypothesis that every nomologically possible material universe actually exists in some superspace is true, then the probability that a fine-tuned universe exist is 1.
8. It is worth noting that, strictly speaking, this version of the fine-tuning argument provides a very weak reason to think God exists. To see why, let O be the observation that John Doe wins a 7,000,000-to-1 lottery. Let H1 be the hypothesis that an omnipotent, omniscient God wanted John Doe to win the lottery and intervened for the purpose of ensuring that John's numbers are drawn. Let H2 be the hypothesis that the lottery numbers are randomly selected. Notice that under H1, the probability that John wins the lottery is 1; for it is not possible for an omnipotent, omniscient God who intervenes to ensure John's numbers are drawn to make a mistake in those circumstances. Notice further that under H2, the probability that John wins the lottery is very small; for, as I described the example, the odds are 1 in 7,000,000. Thus, O is considerably more probable under H1 than it is under H2. Accordingly, it follows from PC that we have reason for preferring H1 over H2. It is true, of course, that PC provides some reason to prefer H1 over H2; after all, if O is all you know about the world, then the fact that O is more probable under H1 than under H2 gives you some epistemic reason to prefer H1 over H2. However, it should be readily evident that such a reason is easily defeated by other observations in the world: the fact that we know that winning numbers are sometimes selected by random mechanisms is sufficient to defeat H1, which theists would readily concede. As Collins formulates the argument, the occurrence of fine-tuning provides a pretty easily defeated reason to prefer the design hypothesis. For more detailed discussion of this point, see Kenneth Einar Himma, "Prior Proabilities and Confirmation Theory: A Problem with the Fine-Tuning Argument," International Journal for Philosophy of Religion, Vol.51, no.4 (June 2002), 175-194.
9. George N. Schlesinger, New Perspectives on Old-time Religion (Oxford: Clarendon Press, 1988), 124-148.
10. There is a second major problem: it is not even clear that the event has the right structure. Assuming that it makes sense to think of the creation of the universe as involving some sort of lottery-like selection event, it lacks one of the elements that justify suspicion in the other examples. In the other examples, each selection results in something of independent value, like a preferred ballot position or a cash prize. The selection of one fine-tuned property, however, does not result in something of independent value. It is valuable only because it leaves open the possibility of winning something that is of independent value namely, a universe hospitable to life; as will be recalled, had any one of these properties been significantly different, carbon-based life at least would not be possible. In this respect, the selection event resembles a lottery game that proceeds by selecting several numbers that must all be matched in order to win a cash prize. This latter event is not, as is readily evident, the kind of improbable event that naturally leads us to suspect that someone was cheating.
11. William A. Dembski, Intelligent Design: The Bridge Between Science and Theology (Downers Grove, IL: InterVarsity Press, 1999), 147. Hereinafter ID.
12. Michael J. Behe, Darwin's Black Box: The Biochemical Challenge to Evolution (New York: Touchstone Books, 1996), 204. Hereinafter DBB.
13. I believe that H. Allen Orr was the first to make this argument. See H. Allen Orr, "Darwin v. Intelligent Design (Again)," Boston Review (December 1996/January 1997), 28-31. I am grateful to an anonymous reviewer for the citation.
14. As Behe concedes "evolution might be far chancier than we thought, but still possible" (DBB 40).
15. Dembski defines "specified complexity" as follows: "Whenever we infer design, we must establish three things: contingency, complexity and specification. Contingency ensures that the object in question is not the result of an automatic and therefore unintelligent process that had no choice in its production. Complexity ensures that the object is not so simple that it can readily be explained by chance. Finally, specification ensures that the object exhibits the type of pattern characteristic of intelligence" (ID 128). Irreducible biochemical complexity is one type of specified complexity - as is the complex specified information that Stephen C. Meyer believes is contained in DNA. See, e.g., Stephen C. Meyer, "Evidence for Design in Physics and Biology: From the Origin of the Universe to the Origin of Life," in Behe, Dembski, and Meyer (eds.), Science and Evidence for Design in the Universe (San Francisco: Ignatius Press, 2002).
16. See William Dembski, The Design Inference (Cambridge: Cambridge University Press, 1998), Sections 2.2 and 6.5. Hereinafter DI.
17. Unless, of course, the probability of that hypothesis is 1. Even if one hypothesis has probability 0, we cannot conclude that another is more probable unless we have some reason to think its probability is greater than 0.
18. One might be tempted to think that this is true only if we have information about the probability of the hypothesis that is being rejected. Thus, on this line of analysis, if we know that A has a non-zero probability and we have no information about the probability of B, we could be justified in rejecting B on the strength of such information, but we could never be justified in rejecting A on the strength of such information. It seems to me, however, that only agnosticism about B is epistemically justified at this point. Should we achieve information about B that bears on its probability down the line, we would have to assess it relative to that of A. Though I will continue to assume as much in expressing the relevant epistemic principle, it is important to note that nothing turns on this assumption.
19. 150) happen all the time, we need some explanation as to why the occurrence of highly improbable specified complexity, as a general rule, justifies an inference of design.
20. 12. Even if every one of these human beings was kidnapped by mad kidnappers, placed inside exploding cardshuffling machines, and subjected to the same cardshuffling experiment as S, even with this many kidnapping victims, it would still be highly unlikely to survive" (DI 186-7). It is clear that this passage, which is representative of his efforts in Chapter 6, tells us nothing about the probability of the design hypothesis; its point is to show that, even if there we assume that there are a trillion kidnappings, the probability of any favorable outcome being explained by chance is extremely low.
21. See Dembski, The Design Inference, Chapter 1, for a discussion of these examples.
22. Dembski, No Free Lunch: Why Specified Complexity Cannot Be Purchased without Intelligence (Rowman & Littlefield, 2002), 5.
23. Michael Behe makes the point as follows: "A few years ago it was reported that an image of Elvis was formed by mold growing on the refrigerator of a lady from Tennessee. Again, the resemblance could be seen, but it was slight. Suppose, however, that the resemblance was actually very good. Suppose that the image was made up not only of black mold. Suppose that there was also Serratia marcescens - a bacterium that grows in red sheets. And suppose there were colonies of the yeast Saccharmyces cerevisiae, which are shiny white. And there was also Psuedomonas aeruginosa, which is green, and Chromobacterium violaceum, which is purple, and Staphylococcus aureus, which is yellow . . . In fact, suppose the bacteria and mold on the refrigerator formed an image of Elvis that was well nigh identical to one of those velvet posters of him that you see in variety stores. Can we then conclude that the image was designed? Yes we can - with the same confidence that we conclude that the dimestore posters were designed" (DBB 198-199). Of course, part of what contributes to our confidence here is that we already know there exist human beings capable of doing similar sorts of things (some people, e.g., can paint detailed pictures on very small surfaces).
24. There are a couple of other ways to set up this thought experiment if the reader has doubts about whether it is coherent. First, one could assume that the truth of the hypothesis that all possible material universes exist in some actually extant superspace and then assume that we have been transported across this superspace to another material universe. Alternatively, one could assume that we are traveling in this one and mistakenly come to believe that we are in some other logically possible world and have no other information about the probability of intelligent life in this universe (apart, of course, from knowing we exist).
25. This assumption is harmless since, again, proponents of design arguments do not want to rely on the other arguments for God's existence.
26. At the very least, they are not principally design inferences. If it is true that the observation of some sort of complexity plays some role in even this second inference (the grill, after all, is fairly intricate), the similarity to other objects nonetheless bears most of the weight. By itself, the observation of complexity does comparatively little to raise the probability of the design hypothesis; it is the similarity to other artifacts that, I think, raises the probability to such an extent that we are justified in accepting the design hypothesis.
27. Of course, ordinary and uncontroversial background knowledge about the world is assumed here.
28. Computers that are programmed, for example, to randomly select a set of numbers operate by means of nomological processes. To say that an event is random is not to deny that it has a cause; it is rather to assert that the causal processes that produced it were, as the matter is sometimes put, blind.
29. One might, of course, doubt that we can rule out such events as irrelevant, but the point here is not to show that these sorts of arguments work. Rather, the point is to show that the inferential strategy is different from the one that grounds the design argument.
30. Because there is such evidence, this is not an instance of the post hoc ergo propter hoc fallacy.
31. Which of these possibilities will presumably be worked out on the basis of further evidence - such as, for example, what J.C. has to say about what happened.
32. William Dembski, "Reinstating Design within Science," Rhetoric and Public Affairs, vol.1, number 4 (Winter 1998), 506. Since Dembski says very little about the particulars of this very intriguing example or its analysis, I will try to flesh it out in a bit more detail here.
33. id.
34. See ID, Chapter 5, for Dembski's discussion of this sequence.
35. We are, for example, sending out radio transmissions in the hope that some extraterrestrial intelligent life receives it.
36. As will be recalled, we cannot be justified in rejecting one hypothesis on the strength of its probability unless we know there is an alternative hypothesis with a higher probability.
37. There is, of course, another epistemically legitimate use. Design inferences enable us to tell when a being we already know to exist is capable of intelligent behavior.
38. Insofar as one believes that the arguments for God's existence provide some reason to think that God exists, the design arguments would seem to provide additional reasons - even if one thinks that the arguments against God's existence provide more reason to think that God does not exist. As long as one believes that she is justified in assigning some non-zero probability to God's existence based on some argument for God's existence, the design argument can add to that. Thus, I suppose that the design argument can be used to tip the scales in favor of theism for an agnostic who believes that the epistemic force of the arguments against God's existence is exactly counterbalanced by the epistemic force of the arguments for God's existence.