Now there are lots of what purport to be reasoned arguments for the existence of god. The argument from design, the ontological argument, arguments from fine-tuning, and on and on. But two things about those arguments strike me. I don't think any one of them is at all rationally compelling. At the very least, an atheist can, I think, argue the theist to a stand-still with counterarguments. If you start out neutral with respect to god and try to reason your way to his existence by appeal to any of the traditional philosophical arguments, you just aren't going to get all the way to positive belief, in my humble opinion. And that I think is the very best that can be said for traditional arguments for the existence of god.
The very worst that can be said for them is that they are all demonstrably invalid and incapable of compelling rational belief in the existence of god. And if the worst that can be said is true, then that seems to suggest that belief in god is a form of unreason.
First for the very worst: As Macht notes, it is false that all the arguments are demonstrably invalid. What's more, it is demonstrably false, since you can construct a valid argument for any conclusion. Similarly it's not really a question whether it's 'rationally compelling'; 'rationally compelling' arguments are, if taken literally, probably a myth (no matter what the argument is, you aren't actually compelled by it, as Macht notes -- for one thing, and this is something we all do a lot, you can take a step back from it and decide that because you don't like the conclusion, there must be something wrong with the argument even if you don't see it), and if taken more loosely, as it usually is, to mean a really good argument that a lot of reasonable think supports the conclusion, it's pretty clear that a lot of people find some of the arguments rationally compelling, or they wouldn't keep popping up. Nor is it at all relevant whether atheists are ever won over by them in a dispute; what is relevant is whether a person on their own, thinking through the matter rationally, might find the argument convincing. There's a famous story about Bertrand Russell, from Russell himself, in which he had gone out for a tin of tobbacco, and on his way was thinking about various things; all of a sudden he stopped short, thunderstruck at something that had come to him, threw the tobacco tin up into the air, and shouted, "Great Scott, the ontological argument is sound!" Then, if I recall correctly, he went home and couldn't remember very clearly what had come to him to make him think that. The point is, when we are dealing with rather sophisticated arguments, as many arguments for the existence of God are, anyone, however dead-set against it, can have a mood where he thinks that maybe, just maybe, there might be something to it, and that maybe, just maybe, the objections to it are really quibbles after all; and other people, given their rational background and line of reasoning, will have much more than just a mood. The problem, by the way, with putting it in terms of a dispute rather than a single person reasoning is that it puts the atheist in very bad company, since there are all sorts of skeptics that, with sufficient ingenuity, can argue you to a standstill on many things that you have excellent reason to believe. The human intellect, unlike an argument, is wild and living; if it wants to go a direction different from where the argument leads, it can find a way to do so. And if you are in a situation of direct opposition, a sufficiently determined and ingenious opponent can keep up the argument indefinitely, however right or justified you may be.
So it's not a matter of validity, nor a matter of rational compulsion. It's not even quite a question of soundness, because not all of the arguments are strictly deductive arguments. The only real question here is whether they are legitimate inferences from true claims. And even setting aside my own very extreme view -- that there are a lot of excellent arguments for the existence of God, some of which can be considered demonstrative (that last clause is especially rare these days) -- it's clear that there are a lot of these inferences where a reasonable person could think that they are good enough, supposing that no adequate defeating arguments came along. If nothing else, there is almost certainly a gray area where reasonable people could reasonably disagree about the quality of the argument. A lot of inferences we think are good are actually quite tricky; we are not rigorous creatures, and even in our relatively good reasoning sometimes slop around a bit more than we should, and sometimes there are issues that are just hard on our given level of analysis. (This, incidentally, is why I still recognize that there are reasonable and rational atheists even though I have the strong view that God's existence is rationally certain. There is room for them, given that not every argument, even assuming that it is good, can be adequately pursued by everyone, and given that I may be wrong about some of them, and given that, even setting that aside, many of them might reasonably be rejected by people with different views of inference, or of some other relevant thing, than I. There's lots of room for being reasonable on lots of things -- it's just that some things have more room for being reasonable than others.)
[I didn't read it until after posting this, but as it happens, Alejandro Satz has a post that independently makes a similar point on this general issue from a very different, and atheistic, perspective.]
But here's the thing. I don't think the real basis of most believers' belief even purports to be anything like reasoned argument. I mean I don't think I've ever met a single person who's been talked out of belief by the failure of any of the traditional philosophical arguments or who's been talked into belief by the success of those arguments. Does that mean that most believers are unreasoning? Well, some surely are. But I'm not prepared to say that most or all are.
I don't know if I've met anyone, either; but, then, I don't go around asking people about such things. It certainly does seem that people can be pretty seriously disrupted by coming to think an argument for God's existence to be a bad argument; for instance, coming to think that Aquinas's arguments for the existence of God were failures played a bit of a role in Anthony Kenny's drifting away from Catholicism. Likewise, it does seem that people can be talked into belief by the success of an argument; witness Flew's conversion by design argument to a weak form of theism. In any case, it is no more surprising that people don't usually get talked into or out of theism by a particular argument than that they don't get talked into or out of any position by particular arguments -- it can happen, but as I think Newman points out somewhere (and in any case, Locke says similar things), mostly it is a matter of a whole bunch of beliefs slowly shifting over time in response to this testimony and that argument, and this experience and that sudden insight, until what seemed bad now seems good and what seemed good now seems bad. Rational belief change, when you think about it, is often a stunningly complicated thing. And note that nothing about this tells us whether that change is supported by reasoned argument. Arguments ramify like crazy; and it could very well be that a reasoned argument to conclusion A may, by defeasible but plausible inference B, lead us to hold C, which by analogy seems very suggestive of D, which in any case may be supported by E and is certainly useful for making sense of F (which I very much want to make sense of), and, hey, wait, if D is true then presumably G. If someone is converted to G by something like this, even if the steps are spread out over time so that the person doesn't remember (as it must be confessed even the most rational people usually don't) exactly what lead to G, it is unreasonable to say that its basis isn't reasoned argument; it's just that it's not all premise-premise-conclusion at one time. Not all reasoned argument occurs in regimented form under surveyable conditions, thankfully.
Finally he says:
The problem with this approach, as I see it, is that if you take yourself to be positing god merely in order to endow one's life with meaning and you do so with no rational basis for really and truly believing that god exists, then you seem to be engaging in a kind of pretense. But I wonder whether mere pretense is really enough to endow our lives with meanings that they don't already have. If mere pretense is enough, why can't we just decide to see our lives as meaningful in the first place, and skip the positing of god in whom we don't really believe.
The problem with this is that if you are positing God wholly in order to endow your life with meaning, it's too quick to assume that you are doing so with no rational basis. One comes across this problem with Kant's moral arguments for God's existence; yes, in fact, God's existence is given no stronger justification than being a postulate for making sense of our moral lives -- but since it's a matter of moral reason you can't actually say that the postulation is without rational basis. If you are positing something because it gives your life meaning, the question that arises is whether there's anything about giving your life meaning that gives you reason to think that what you are positing is not merely a fiction of convenience. But if there is, there is a rational basis for your belief. In other words, positing something's existence is perfectly fine; there are lots of circumstances in which it is not only reasonable, it is unreasonable not to do so. And if positing that something's existence really does what it's supposed to, that can often be, in and of itself, a good reason to believe that it exists. It varies from case to case. My point is not that 'giving one's life meaning' is a good reason to believe that God exists, since I very much doubt it is, at least without more specifics; but that you can't assume that positing X's existence for purpose P, and for no other reason than P, is a mere pretense that X exists.
Positing and pretense are in any case always very different rational functions even where we have no particular commitment to what we are positing. Someone who posits a center of gravity for the purposes of calculation, and only for the purposes of calculation, isn't making a pretense that there's a center of gravity; he's positing one, and it can be a perfectly rational thing to do whether centers of gravity actually exist or not. If a philosopher of mind posits for the purpose of a thought experiment a mind-reading psychic, he's not pretending that there are mind-reading psychics -- or if he is, he needs to go back and rethink his whole understanding of thought experiments. When for a given purpose we posit something to exist (like some astrophysical phenomenon that would make our mathematics come out right), rather than positing it to exist for a purpose (like positing that the earth is a perfect geometrical sphere for the purposes of roughing out the answer to a particular problem), things can become considerably more tricky. I suspect that what's really happening here is that Taylor is smuggling all sorts of things into his 'really and truly believe'; I doubt he just means 'actually believe' rather than 'believe in some special sense'. Without knowing quite what special sense he would mean, it's a little difficult to see where the argument is going. It could be that he means by 'really and truly believe' something like 'believe with conviction', or 'believe with passion', or 'believe in such a way that you could live your life in accordance with it'; but you can in fact do all these things based simply on a posit, if the purpose of the positing is sufficiently important. The problem with just skipping the positing step is that (in a case like this) the reason it obviously would exist in the first place is that the person doesn't see how he could make sense of the claim of a meaningful life unless there were something like what he was positing. The reason we posit X is always that X seems to be what's called for by our purposes; so long as those purposes remain, and that appearance remains, X is what you posit.
