Alvin Plantinga and Thomas Nagel are well-known senior philosophers at Notre Dame and NYU, respectively. Plantinga, a Christian, is known for his contributions to philosophy of religion, while Nagel, an atheist, is known (nevertheless) for his resistance to purely materialist/naturalist/physicalist theories of the mind (e.g. in his famous article, "What is it like to be a bat?"). Now Nagel has reviewed Plantinga's most recent book in the NYRB, giving it a much more sympathetic reading than most naturalists would offer. (For what it's worth, Plantinga is a supporter of Intelligent Design, and Nagel has often spoken of it approvingly, while not quite buying the whole sales pitch.) Jerry Coyne offers a reasonable dissection of the review. I wanted to home in on just one particular aspect because it was instructive, at least for me. There is a long-standing claim that "faith" is a way of attaining knowledge that stands independently of other methods, such as "logic" or "empiricism." I've never quite understood this -- how do we decide what to have faith in, if not by the use of techniques such as logic and empiricism? Plantinga offers an answer, which I think is at least internally consistent -- but that's part of the problem.
So far we are in the territory of traditional epistemology; but what about faith? Faith, according to Plantinga, is another basic way of forming beliefs, distinct from but not in competition with reason, perception, memory, and the others. However, it is
God endows human beings with a sensus divinitatis that ordinarily leads them to believe in him. (In atheists the sensus divinitatis is either blocked or not functioning properly.)
Plantinga is clearly trying to separate "faith" from merely "things we would like to believe are true" -- faith is knowledge that is put directly into our minds by God. Points for at least trying to offer a reason why we should put credence in beliefs based on faith even if the logic and/or evidence aren't there. Here, as I see it, is the problem. Any time we have beliefs of any sort, we need to admit the possibility that they are incorrect. Even if we have think that some result has been reached by nothing but the application of pristine mathematical logic (e.g. the ABC conjecture), it's always possible that we simply made a mistake -- have you ever multiplied two numbers together and gotten the wrong answer? Certainly in an empirical endeavor like science, we recognize that our theoretical understanding is necessarily contingent, and are constantly trying to do better, via more precise and far-reaching experimental tests. These are methods of reaching knowledge that have built-in methods of self-correction. So what about faith? Even if your faith is extremely strong in some particular proposition, e.g. that God loves you, it's important to recognize that there's a chance you are mistaken. That should be an important part of any respectable road to knowledge. So you are faced with (at least) two alternative ideas: first, that God exists and really does love you and has put that belief into your mind via the road of faith, and second, that God doesn't exist and that you have just made a mistake. The problem is that you haven't given yourself any way to legitimately decide between these two alternatives. Once you say that you have faith, and that it comes directly from God, there is no self-correction mechanism. You can justify essentially any belief at all by claiming that God gave it to you directly, despite any logical or evidence-based arguments to the contrary. This isn't just nit-picking; it's precisely what you see in many religious believers. An evidence-based person might reason, "I am becoming skeptical that there exists an all-powerful and all-loving deity, given how much random suffering exists in the world." But a faith-based person can always think, "I have faith that God exists, so when I see suffering, I need to think of a reason why God would let it happen." Sometimes you will hear that "science requires faith," for example faith that our sense data are reliable or that nature really obeys laws. That's an abuse of language; science requires assumptions, just like anything else, but those assumptions are subject to testing and updating if necessary. If we built theories on the basis of our sense data, and those theories kept making predictions that turned out to be wrong, we would examine and possibly discard that assumption. If the universe exhibited a chaotic jumble of non-lawlike behavior, rather than falling into beautiful patterns, we would abandon that assumption as well. That's the most compelling thing about science: it always stands ready to improve by casting out an old idea when the evidence demands it. Okay, this is probably belaboring the obvious for atheists, and completely irrelevant for believers. But it's useful to have a specific definition of "faith" right there on the page, if only to understand what its dangers are.
a wholly different kettle of fish: according to the Christian tradition (including both Thomas Aquinas and John Calvin), faith is a special gift from God, not part of our ordinary epistemic equipment. Faith is a source of belief, a source that goes beyond the faculties included in reason.