Williamson’s argument is that the idea of naturalism – that we can somehow reduce all properties and things to physical things through the scientific method – is restricted and therefore not able to make the general claims that it does, and that it has become dogma. This strikes me as extreme straw manning, because I haven’t seen this dogmatic attitude in actual philosophy in general, although I am sure we can find some examples here and there.
In general, naturalisms are quite modest – attempts to make epistemology, truth or ethics a natural fact; which often fall afoul of things like the is-ought distinction: if we can say some act is right, we cannot get that from a natural description. This leads to moral antirealism or “error theory”, in which it is simply a mistake to say that ethical values are objectively true. Similarly, a naturalism of epistemology tells us only that some method or relation between knower and the known obtained successfully in the past. It will not tell us what will work in future.
Rosenberg takes an obvious line of defence: naturalisms have worked in the past, and we have every confidence to think that they will work in the future. While he does not mention it, the entirety of modern and western philosophy and science is grounded on the presumption that the natures of things (what they are made of and how) determines what they can or will do. Naturalism underlies all science and really, all non-magical thinking. He goes somewhat further, though: there is no other kind of knowledge. Jerry Coyne pipes up to defend scientism: we should not accept anything other than science.
I have a more agnostic approach. A philosopher cannot, unlike an evolutionary biologist, presume the conclusion to a philosophical investigation. In philosophy if not biology, question begging is a fallacy. Yes, we can agree that science has worked (but not always). It doesn’t follow that we cannot find knowledge any other way [those who think we do not find knowledge through science are beyond arguing with]. So I want to investigate this a little.
Naturalism is a project: It aims to try to explain all things, including mental activities and experiences, origins of living things, ethics, and so forth, which traditionally have been explained in terms of transcendental entities like God, Mind, and Form, by natural processes which are nomologically constrained.
Now when it does this (say, perception, which used to be an activity of soul), we can say “Science has shown perception to be natural.” When there are reasons to think we will shortly do this (say, heredity before 1944, when Schrödinger wrote his book), we can say “Science will show us how this happens naturally.”
And there are things that science can, and has, shown to be nonexistent, like crystal spheres, the firmament, aether, phlogiston and caloric. These are eliminated because we show they are empirically disproven (things like comets pass through the spheres, the moon has blemishes and is therefore not a pure sphere, etc.), or because they are fundamentally contradictory to things we know to be true (the action of “mind” on other minds in ways that are not physical contradicts everything we know about energy and causation).
But there are things we cannot show not to be real, because they are neither empirically disproven nor contradictory to our best knowledge. God (at least the kind of God I call “empirically inoculated“) is one of them. So the question now is: could God exist in a non-natural (which basically means a non-physical causal) fashion?
And of course that kind of God could exist. So the sole remaining question is: why should we believe in that God?
We cannot believe in that God because natural science has demonstrated he exists, because he is, by definition, non-natural and hence we cannot use natural means (scientific protocols, etc.) to investigate him (or her, or it, or them) So what warrant is there for believing in him/her/it/them (HHIT)?
There are only three:
1. Revelatory: We have had revealed to us that HHIT exists. This leads to questions of the reliability of revelation and whether we can test or rely upon it.
2. Logical: The existence of such deities is mandated by our other commitments (such as the existence of a universe). This leads to problems of implicature. Since the pre-Socratics, we have thought that a universe might be self-existent.
3. Tradition. We believe in God because we were raised as Christians/Jews/Muslims/Mormons… This leads to a pure cultural relativism about gods. However, it is, I think, the main reason why we think Gods exist. Most Christians do not wish to be excluded from their community and family by denying something so vital to the cultural traditions. God is a kind of tribal marker (arguably, that’s why belief in deities evolved in the first place).
Suppose I lack the tradition, have no revelation worth the name, and see no reason for adopting belief in HHIT from logic. I therefore have no motivation or warrant for that belief, and will not move from nonbelief to belief. But if I already believe, I might adopt the empirically inoculated kind of God to avoid losing my conceptual and community investments. So I would not adopt naturalism as a project that extends beyond my scientific commitments.
The question of why people change their conceptual commitments is fraught with difficulty, but basically I think it is based on trying to find a Pareto optimum between competing epistemic and social utilities. People convert when they reach a saddle point such that a move in one direction is going to increasingly approach that optimum (i.e., when they slide down the hill to a basin). If you are already in a given basin (in my case, the physicalist naturalist position) then you will not find the motivations and reasons given by others compelling to you. But if you are in an unstable conceptual region, you may.
Naturalism is a kind of default position: it is what you believe when nothing else intervenes. But this is not the case for all rational agents. Some might begin with conceptual investments that mean the shape of the hills they are climbing or descending are different from some other agent who has not been inculcated with the values of the proto-theist, -dualist or -moral realist.
For example, we often read Christians who assert that they were at some time atheists, but that they converted when they found the emptiness of that conceptual scheme (e.g., C. S. Lewis). Now I think people reconcile competing value commitments and conceptual commitments through their life. They do not start, as the myth goes, from a kind of level and rational playing field. Lewis was raised in the Christian tradition, and went to schools that held daily services. He was hardly more than a superficial atheist, one whose primary purpose was to disagree with the surrounding culture (as many adolescents do). But his underlying value system was theist, and specifically Christian, so it is hardly surprising that he converted thus. It would have been more surprising had he converted to Hinduism or some African animism.
We are inculcated with ideas and conceptual stances as we grow and learn. We are told that it is intuitive that we are made of souls (and yet until Pythagoras and Plato this was not a widely held view in the ancient western world; possibly it was based on Vedic missionaries’ teachings), and that we have irreducible inner lives (a view not held widely until the modern era of Descartes and Pascal), and that moral values are objectively true (contrary to every theodicy prior to the middle ages). Naturalism is a project to eliminate these cultural and traditional biases.
We should only accept what we can demonstrate to be the case where investigation is possible. Where it is not, then I submit that it really doesn’t matter what you believe. As I like to say, I don’t care if you believe that we should have sex with trees, although I would suggest you mind the splinters, so long as you do not allow that belief to affect the teaching and doing of science, which is, as Rosenberg says, the best way, if not the sole way, we have of knowing things.
A final point: in American discourse, since Philip Johnson so labeled it, “naturalism” means the exclusion of all religion from science. This is historical and philosophical nonsense. Science is exclusive of religion (and there is no such thing as “Augustinian Science” either, contrary to Plantinga’s bit of special pleading). A rather useless distinction is made between “methodological naturalism” and “metaphysical” or “philosophical naturalism”. In philosophy, we have terms for this: science and physicalism (or, if you read pre-Einsteinian philosophy, materialism), and the use of “naturalism” here is simply obfuscatory.
Naturalism is at the least the observation that science studies, and can only study, the natures of things. At the broadest it is the claim that there is no other way to learn things. Even if you reject the latter (and I do not), you cannot take issue with the former unless you are a Republican candidate for office in the United States.