>The problem is that these things require the
>existence of a rational, logical universe to work.
>Assume that the theists are right. Then there is
>no guarentee that a = mv, because God could
>screw with it at any time.
It seems to me that you have the notion in reverse. Theists believe that the universe possesses this sort of order because there is a rationally orderly God who created the universe. Indeed, it was such theists who founded modern science. What justification would an atheist have for the uniformity of nature? There really is no logical principle to tell us that physical laws will hold true in places where we haven’t tested them (even if that place is the future). Classical atheism held the universe holds to the metaphysical view of a universe dominated by chance events and hardly implied an orderly universe. Perhaps some justification could be made that’s compatible with atheism (I think it can be done), but theism proved to be more conducive to the order in the universe when it came to the birth of modern science. Thus, order in the universe and theism can clearly be quite compatible and perhaps even complementary.
More to the point though, what are the ultimate foundations of science? First, I think we need to recognize that science has more philosophy integrated into it than some might think. For example, the basic presuppositions. One must believe that knowledge is possible, that our senses are at least sometimes reliable, and so forth. Justifying such beliefs falls into the realm of philosophy (and more specifically epistemology). Another example: one myth of science is that a theory can be chosen solely on the basis of empirical data. However, one well-known fact in the philosophy of science is that empirical data underdetermine theories. Any given body of data (no matter how large) will always be agreeable with an unlimited number of alternative theories. Invariably there are many theories that explain the exact same data, and at least some of the theories will contradict each other. Some of these empirically indistinguishable theories may be elegantly simple and others may be outrageously complex, but multiple alternatives exist for any set of data.
Here’s a quick example of an empirically identical but utterly wrong theory. Suppose I proposed the theory that the moon is made of cheese. To refute this theory, many people would point out that astronauts have gone up there and found out that it is more like a rock than a huge piece of cheese. I could counter that argument by saying something like, “the moon with its great age would naturally accumulate massive quantities of rocks and other particles from space. Under that layer of space debris, however, is the cheese.” Of course, it is possible to rationally discard this absurd theory, but the point is one cannot do this merely by pointing to the data. When the right ad hoc hypotheses are made, the theory of the moon being made of cheese becomes empirically identical to the moon being rock-like. This sort of thing is not limited to ridiculous theories about the moon’s composition. It’s possible to modify virtually any theory so that it’s consistent with whatever data that might come up.
Since these competing theories are empirically indistinguishable from each other, if science is to pick out a theory from among these numerous competitors and claim that it is correct, then such a selection must be based on nonempirical principles (whether they be philosophical, personal, societal, or whatever). Some such values are needed. What are some of these values? Ockham’s razor, fruitfulness (rise to other understandings and having stimulated pioneering investigations and advancements), self-consistency, etc. I think at least some of these are quite reasonable, and the situation isn’t as dire as it might appear. We can use these values to discard, for example, the moon made of cheese theory, and can do so with confidence. But philosophy still plays a significant role here. When using values to select a theory, we must have some philosophical basis for believing that nature’s preferences are similar to ours. And for many of these principles there is no logical rule to imply their reliability. For example, in picking out a theory from among it’s empirically indistinguishable competitors (and when all other factors are held constant), the notion that reality favors simple theories over complex ones is nevertheless a philosophical principle.
Ultimately, then, we must have some basis for these philosophical beliefs and presuppositions for science’s ultimate foundation. Therein lies the answer.