I've written a few posts about that argument already and was planning to write a fourth. In fact, a long unfinished fourth post has been sitting in my drafts for two months now. But the reason I never finished it is that I discovered some points in my thinking about this argument that needed to be worked out in better detail. I'm still working it out, so my thoughts on this argument are still unfinished, but it might be helpful to write out some of what I've been thinking about.
First, just to recap a little: the argument from reason is a family of arguments that begins with the assumption that we are able to form rational beliefs, and then it's not so much an argument for theism as it is an argument that this ability is not compatible with a naturalistic account of the universe. Obviously, naturalism doesn't work unless it can be reasonably affirmed. But the argument is that, in some way or another, naturalism undermines our ability to trust our own rationality. In other words, naturalism is accepted on the basis of certain arguments, but naturalism also seems to imply that our beliefs are the inevitable product of non-rational, physical processes, which calls the rationality of those beliefs into question. In this way, naturalism seems to be self-defeating.
I initially said that there were four versions of the argument from reason that pushed me toward theism (or at least away from naturalism). What tripped me up when I started work on the fourth post was that the fourth version suddenly seemed to be a rehash of the third version, and the third version had already turned out to be two distinct (albeit closely related) arguments. Then I began to wonder if the first version (which was aimed at only a very extreme version of naturalism called eliminative materialism) was adding much to the discussion. So the first challenge for me was just in organizing my thoughts in a way that made sense.
As I reflected further on the different arguments, I also realized that I needed more clarity in what each of the arguments were targeting. Two issues turned out to be especially important. First, one criticism of the argument from reason that often comes up is that it's possible for us to hold rational beliefs without needing to justify them with arguments. For instance, I believe that I am looking at a computer screen. This belief is not based on arguments or other beliefs but on my direct perception, and even though my perceptions can be mistaken, my belief that I am looking at a computer screen is still rational. After all, it's true! So either the argument from reason needs to restrict itself to beliefs based on logical inferences (which in some cases I was already doing, without realizing why it mattered so much), or it needs to show that even the reliability of our basic beliefs (the ones not based on other beliefs) is called into question by naturalism. Now, there is a famous version of the argument from reason developed by Alvin Plantinga that is usually called the Evolutionary Argument Against Naturalism (EAAN). This was the argument I originally had in mind as the fourth version of the argument from reason that I wanted to defend. From what I can tell, the EAAN is targeted at beliefs in general, not just logical inferences. If that's true, then it really is a distinct argument from the others that I'd already discussed. It also seems to be the most influential of these sorts of arguments.
But this leads to the second issue, which is that some versions of the argument from reason depend on an internalist view of justification, whereas many people (including Plantinga and others who defend the EAAN) accept an alternative to internalism called naturalized epistemology (or at least an alternative called Reformed epistemology that is directly based on naturalized epistemology). Very briefly (and assuming that I've got this all straight in my head), internalism holds that beliefs are justified based on factors internal to the person who holds those beliefs. For instance, a belief is justified when the person has good reasons for accepting it, or when they have fulfilled their intellectual duty by examining the evidence for it. Internalism seems straightforward, but from what I understand, it becomes more complicated when dealing with basic beliefs (which, again, are not based on arguments or other beliefs), and then there's also the question of whether belief in internalism itself can be justified in a non-circular way. On the other hand, Plantinga's preferred approach to epistemology replaces the concept of justification with warrant, and it says that a belief is warranted when it is formed in a reliable way, or when our cognitive faculties are properly functioning. This, too, can lead to circular reasoning if not handled carefully, but if I am not mistaken, much of Plantinga's philosophical project has involved showing that (1) this approach to epistemology is probably correct, (2) it only seems to work if something like theism is true. It's possible that I've got some of the details of this wrong, but that's my understanding so far.
Personally, I've never known what to make of Plantinga's work or of his Reformed epistemology. It always struck me as a way of arguing with naturalism on its own terms, but not necessarily as a good approach to epistemology, especially because of the dubious (I would even say obnoxious) way in which it's often used to justify Christian theology (including by Plantinga himself). But naturalized epistemology, from what I understand, is the preferred approach for many naturalist philosophers. Upon realizing how big of a role these different views play in the argument from reason, I'm finding myself wanting to explore the debate over internalism and naturalized epistemology and see which approach makes the most sense to me. I still think I'm inclined toward internalism, since it makes perfect sense to me that beliefs that aren't based on direct experience would need to be justified through careful reflection on arguments and evidence in order to count as rational beliefs. But I'm genuinely curious to see if there is anything important that I'm overlooking here. At the very least, I want to know if my understanding of what makes a belief rational needs to be vindicated before the argument from reason can go through. I may finally need to get around to reading some of Plantinga's work on this topic.
POSTSCRIPT
I had one more thought lingering in my head after I posted this. It occurs to me that, regardless of whether we think that beliefs in general need to be justified in order to be regarded as rational, it still remains true (at least I think it does) that a logical inference can only be justified if it follows validly from other beliefs that are more plausibly true than false. The belief itself could perhaps be deemed rational on other grounds. But as an inference, it must follow validly from other truths.
So perhaps one does not need to come to a resolution about the internalism/externalism/naturalized epistemology debate in order to agree that conclusions in an argument are justified when they follow validly from premises whose truth is more plausible than their falsehood. In that case, I think that several versions of the argument from reason will still go through. EAAN should go through as well, especially if it is applied to logical inferences rather than to beliefs in general.
I will probably still take a look at Plantinga's work on this topic when I have the opportunity, but I don't think it will necessarily change anything in my thinking about naturalism. I still don't see how I could rationally affirm determinism if my belief concerning determinism is determined by non-rational physical forces, or even by an external agent (rational or otherwise). I also don't see how I can rationally deny the reality of mental causation if my beliefs about mental causation are caused by non-rational physical forces. If these arguments are sound (and I don't see how I could reasonably deny them), then evolutionary theory makes no difference, because it still does not justify either a belief in determinism or a rejection of mental causes. The real value of EAAN, then, would be in offering further criticism of a naturalistic account of human origins that is already known to be false on other grounds. It just so happens to be a very powerful criticism since it takes the naturalist view of human origins seriously on its own terms.