Thomas Nagel: “one cannot really understand the scientific world view unless one assumes that the intelligibility of the world, as described by the laws that science has uncovered, is itself part of the deepest explanation of why things are as they are. So when we prefer one explanation of the same data to another because it is simpler and makes fewer arbitrary assumptions, that is not just an aesthetic preference: it is because we think the explanation that gives greater understanding is more likely to be true, just for that reason.” Mind and Cosmos (2012), p. 17.
I’m not sure what he means, of course, but here is a stab at it, and at what is wrong with what he says. To “really understand” would be, say, to understand why it is that the world is scientifically intelligible. If we assume that there is an answer to that question, then I suppose it must be that some background conditions are necessary for the explanans to do its work and entail what is explained. To know the explanation is correct, we have to know the background conditions to be satisfied. But where does our information that the background conditions are satisfied come from? I guess it would have to come either from scientific procedures, or some other way. If the first, we seem to get in regress trouble, unless we use our explanation to explain how we know the background conditions for its application are satisfied, and that looks suspiciously circular. I guess Nagel embraces the circularity. What would it be for the intelligibility of the world to be part of an explanation? Is it that it is part of the explanans, from which the explicatum is to be deduced, given the background conditions? Then, that the world is intelligible, with some other assumptions, given further assumptions, leads to the conclusion that the world is as described in science. On one reading that seems obvious and trivial—we can indeed argue that the world is intelligible, such and such observations have been made, it is possible to argue from observations to what is really so if we are careful enough (since the world is intelligible), and so, being careful (i.e. scientific in our procedures), we argue to the world fitting the description given by science. In effect, we have pointed out that application of “the scientific method” (taken in the widest and loosest possible sense), if it is to be justified, requires that the world be intelligible (i.e. through that method). But surely it does not follow from this that the world’s intelligibility implies, or causally contributes to, the world being the way that method shows us it is. For one thing, it implies this only given the background information that enables us to deduce, assuming the validity of our method, that the world is as science reports it is.
If I explain how it is something can be known, that does not amount to explaining why it is so! Say I explain that a given particle can be detected because it produces certain electromagnetic phenomena which are observable using the instruments we have available to us. It does not follow that the reason the particle produces such phenomena is in order that it be detected, or that its detectability causes it to produce those phenomena. Rather, the production of those phenomena causes it to be detectable. Evidence that it is so is one thing, the causes why it is so another, even if it follows both from the evidence and from the cause that it is so. Try it with Nagel’s point as he completes his remark—it turns out that seeking simplicity leads us to the truth. Why? I can imagine an explanation, built up around how it is that we (us, not just any old imagined being) are able to discover the truth (that assumes we can, of course, but assume that, and take “discovering the truth” to be whatever it needs to be so that we can be justified in our assumption). Now, to do this, we need to worry about how creatures built as we are are able to accomplish this business, given the way the world is. Everything hangs, one supposes, on the details of the machine that realizes the function of truth discovery here, and the environment necessary for the machine to function properly. It turns out that, given those things, the machine finds what it views as simpler explanations more likely true than more complex ones. In order to make this work, we need to do a lot of work figuring out what “simplicity” is here, and noting that it is an aesthetic notion is only a bare beginning to that. We need to understand it completely enough to evolve our explanation, based on our understanding of the machinery that realizes truth detection. I see nothing crazy about this. It looks like a fascinating question, actually, and we would learn a lot about science and scientific knowledge if we were able to work out an answer. And, of course, right now we can only speculate a bit on what that answer will turn out to be if we find it, since it is a very difficult scientific problem that is set us here. But say we do figure it out. Will we then have found that the intelligibility of things (through the search for simple explanations) is “part of the deepest explanation we have why things are as they are”? Not at all. Why the conditions are satisfied for the effective functionality of the machine that uses simplicity as a clue to realize truth-discovery is not anything we have even investigated yet. (It could be investigated, of course…)
This can all be pursued farther by trying to see how we are to understand “causes” versus “evidence” (“reasons to believe”). They both involve reasoning to the thing (the thing caused, the thing we have evidence for), but some further restriction on cause must clearly be recognized (and no doubt a different further restriction on evidence—evidence makes it evident to someone, and how that is done all depends on who the someone is). To get at that would probably clarify what we want to say when we speak of how things really are, since the cause must be real, not merely apparent. But its near lunch time and the dog needs to go out.