A standard kind of design argument is one that takes a nomic generalization, and asks why the law in question holds. One can either ask about particular laws (e.g., fine-tuning arguments do that) or about why there are laws at all. I have nothing against such arguments, but I want to suggest that it’s worth thinking about a different family of design arguments: arguments from non-nomic generalizations in science.
Here are some generalizations that appear to be non-nomic:
- Every species on earth, with the exception of one root species, has an evolutionary explanation of its existence.
- Every particle is a P1 or a P2 or … or a Pn. (Here, n is a small finite number, many orders of magnitude smaller than the number of particles in existence. Physics isn’t yet capable of making this generalization, since we don’t yet have a complete list of all particles.)
- Every planet in the solar system is in approximately the same orbital plane.
Now, in some cases, one might seem to have a perfectly fine physical explanation. It seems we can explain (3) in terms of the way our solar system actually formed from a bulgy disc around the sun. There may, further, be facts that imply that this method of planetary formation is much more probable than other methods, given the initial conditions. But, while I am willing to concede this case, I want to make two points. The first point is that the explanation may well end up depending on non-nomic (and perhaps only statistical) generalizations about the initial conditions of our universe, so the problem may simply get pushed back. The second is that the explanation depends on the principle that nothing can come from nothing. After all, if planets could come into existence ex nihilo, then to explain why all planets in the solar system have coplanar orbits would require us to explain why no new planets came into existence ex nihilo with weird orbits. And it is not clear that one can, in the end, defend the principle that nothing comes from nothing without defending the most controversial premise of the cosmological argument.
But let’s put (3) aside. It is, nonetheless, plausible that there are some non-nomic generalizations that cannot be explained in this way. Certain generalizations about the initial conditions might be like that. And (1) and (2) may be like that. Now, the theist can give explanations of non-nomic generalizations by noting that the existence of such generalizations is a case of order, and order is a good thing, and hence the sort of thing we would expect a deity to produce. God could have some species evolve and others come into existence in some other way, but it is particularly elegant if all evolve except the first one, unless some further principled distinctions can be drawn between where evolution happens and where it doesn’t. (To see that (1) is unlikely to be a nomic generalization, note that (1) is false if Russell’s five minute hypothesis is true. But, plausibly, Russell’s five minute hypothesis is compatible with the laws of nature, as far as I can tell.) Likewise, there is an aesthetic value to a complex universe built out of a small set of natural kinds as building blocks.
It may well be that we need to wait for more scientific developments to be able to give better arguments along these lines. But I thought I’d share the idea and maybe commenters will have other examples of non-nomic generalizations that are of interest. There may also be other design arguments not based on generalizations. For instance, consider the puzzle of why Russell’s five minute hypothesis is false. We have a very good theistic explanation of that falsity, but it is not clear we have any good non-theistic explanation.
A related interesting observation is that the scientist who is not a theist will have less reason to believe in non-nomic generalizations like (1) and (2) than the scientist who is a theist.