Let's take our intuition, our faculty of memory, our powers of perception and introspection, our powers of deduction and inference, and put them all in a box. This box can be seen, maybe a bit too
simplistically, as a machine that produces belief. Somewhere within
us, no doubt in our brains, this box receives input from our senses and outputs a representation of the external world. To keep things simple, let's call this
amalgamation of neurology our faculties. And, without worrying about how it works, let's worry about whether it works at all. There are different senses in which our faculties can 'work'. We are able to navigate our environment without dying, and that's a sort of success. But it's possible to navigate effectively while being completely wrong about what one is navigating. Our faculties could very reliable for keeping us alive, but be very unreliable for delivering truth. The psychologist might be very interested in the first sense that our faculties are reliable, and want to study it more (maybe with the hopes of replicating them; AI technology is much sought after). But the philosopher is interested in the second, he wants to know whether there is a serious threat of skepticism. It is this latter sense we will be concerned with.
Many think there is an argument against Naturalism and for Theism from the reliability of our faculties. After all, whether or not God was directly responsible for human origins, we can expect he would intervene in some respects. Maybe God doesn't care if we're bipedal or quadrupedal, or whether or not we have a tail or are covered in fur. But God does care about our intellect, and our ability to discover truth. God cares, then, about our brains. And so, on Theism, there is an available explanation for how our faculties might be reliable, in that they were designed to be reliable by God. But on Naturalism, how can this issue be addressed?
The standard scientific explanation is that our brains, along with the rest of us, evolved by natural selection over millions of years from a long extinct ancestral species. And that species likewise evolved from something even older, and so on and so on. Our faculties, then, were produced somewhere along this line of animal evolution. But herein lies the problem. There doesn't seem to be any selective pressure for having reliable rather than unreliable faculties. One might think an animal must know where the water is, know who the predators are, and know what to mate with, to successfully pass on its genes. But this isn't quite right. What matters is that the animal goes towards water when it's dehydrated, hides from predators when threatened, and pursues a mate of its own species. In other words, what matters is the animals behaviour, and not his beliefs. If the animal moves in all the right ways, then his mind can be on vacation and he will survive just as well. As a result, it seems the truth of an animal's belief does not factor into his evolutionary fitness.
Given an alternate evolutionary history, humans might have had a very different understanding of the world. Imagine, in this alternate reality, someone sees a tiger and, thinking it is friendly, tries to befriend it. But, they think, playing tag with it and running away as fast as they can is the best way to befriend it. In their mind, tigers respect fast runners. When they run away, it doesn't matter what was going on inside their head. Their belief about the tiger still motivated the right action, even though it was false. And, if you think about it, there are many ways a network of beliefs could be beneficial but false, but only one way in which they could be true. Maybe instead of thinking the tiger is friendly, this fellow thinks the tiger is lucky. And that seeing a tiger means you will be blessed with finding riches if you run away immediately. I'm sure there are an endless many silly little stories we could tell about why a false belief would cause someone to perform the right action.
But if there are many more ways in which our faculties could get it wrong than right, and if natural selection has no way to differentiate between the two, then we would expect our faculties to become very finely tuned to keep us alive, but not at all reliable for discerning truth. And so the probability of our faculties being reliable (R) on evolutionary naturalism (E&N) must be very low.
Many think there is an argument against Naturalism and for Theism from the reliability of our faculties. After all, whether or not God was directly responsible for human origins, we can expect he would intervene in some respects. Maybe God doesn't care if we're bipedal or quadrupedal, or whether or not we have a tail or are covered in fur. But God does care about our intellect, and our ability to discover truth. God cares, then, about our brains. And so, on Theism, there is an available explanation for how our faculties might be reliable, in that they were designed to be reliable by God. But on Naturalism, how can this issue be addressed?
The standard scientific explanation is that our brains, along with the rest of us, evolved by natural selection over millions of years from a long extinct ancestral species. And that species likewise evolved from something even older, and so on and so on. Our faculties, then, were produced somewhere along this line of animal evolution. But herein lies the problem. There doesn't seem to be any selective pressure for having reliable rather than unreliable faculties. One might think an animal must know where the water is, know who the predators are, and know what to mate with, to successfully pass on its genes. But this isn't quite right. What matters is that the animal goes towards water when it's dehydrated, hides from predators when threatened, and pursues a mate of its own species. In other words, what matters is the animals behaviour, and not his beliefs. If the animal moves in all the right ways, then his mind can be on vacation and he will survive just as well. As a result, it seems the truth of an animal's belief does not factor into his evolutionary fitness.
Given an alternate evolutionary history, humans might have had a very different understanding of the world. Imagine, in this alternate reality, someone sees a tiger and, thinking it is friendly, tries to befriend it. But, they think, playing tag with it and running away as fast as they can is the best way to befriend it. In their mind, tigers respect fast runners. When they run away, it doesn't matter what was going on inside their head. Their belief about the tiger still motivated the right action, even though it was false. And, if you think about it, there are many ways a network of beliefs could be beneficial but false, but only one way in which they could be true. Maybe instead of thinking the tiger is friendly, this fellow thinks the tiger is lucky. And that seeing a tiger means you will be blessed with finding riches if you run away immediately. I'm sure there are an endless many silly little stories we could tell about why a false belief would cause someone to perform the right action.
But if there are many more ways in which our faculties could get it wrong than right, and if natural selection has no way to differentiate between the two, then we would expect our faculties to become very finely tuned to keep us alive, but not at all reliable for discerning truth. And so the probability of our faculties being reliable (R) on evolutionary naturalism (E&N) must be very low.
This is where
the argument gets interesting, for if we must accept that
Pr(R|E&N) is very low, then we must accept that E&N is a
defeater for R. But if E&N is a defeater for R, then it is
likewise a defeater for all the beliefs we know were produced by our
faculties. In a strange twist, we recognize that E&N is beliefs
of ours—having been produced by our faculties—and, as such, is
self defeating. Since the evolution of man is an
established fact, Naturalism must be rejected. Ironically,
one of the perceived pillars of naturalistic thought defeats
our warrant for believing Naturalism.
Or so some think, but there's more to the matter. However beneficial it
is to have the right beliefs, it's even more beneficial to be able
to draw the right conclusions. An animal who is able to infer truths
about his environment can acquire knowledge of things he has never before experienced, which allows for planning and
forethought to guide the animal's behavior. Suppose a species is plucked from their
home environment and dropped into one they've never before
encountered. Evolution couldn't have possibly equipped them for
this new environment, since they only ever evolved to suit their old
one. And yet, they might already have crucial knowledge of this new
environment simply by making inferences. Clearly this species will
survive at a greater frequency than one who doesn't have any such inferential power.
But inference is a garbage in, garbage out process; only true premises guarantee true conclusions. This provides selective pressure for evolution to maximize the number of truths an animal possess, offering him a large pool of premises to safely draw conclusions from. On the other hand, a false belief allows the animal to draw false conclusions (even if the reasoning is flawless). And so to minimize the chance of error, evolution is pressured to weed out falsehoods. Furthermore, false beliefs can have a sort of domino effect: one might allow many in, which in turn might allow many more. And so the pressure to not have false belief is fairly strong.
Looking back, the stories of the fellow and the tiger all involved him having more beliefs than necessary to motivate the right behaviour. If he merely understood that the tiger was dangerous, that would surely be enough. But in thinking the tiger was friendly, he had to have other false ideas of how to befriend it. In thinking it was lucky, he had to have other false ideas about how to pursue riches. Since there is evolutionary pressure to minimize an animal's false beliefs, these over-complicated, false belief structures are actually less likely to be beneficial, and thus less likely to evolve, than the simple truth. And so, given an evolutionary origin, we would actually expect our faculties to be more or less reliable.
But inference is a garbage in, garbage out process; only true premises guarantee true conclusions. This provides selective pressure for evolution to maximize the number of truths an animal possess, offering him a large pool of premises to safely draw conclusions from. On the other hand, a false belief allows the animal to draw false conclusions (even if the reasoning is flawless). And so to minimize the chance of error, evolution is pressured to weed out falsehoods. Furthermore, false beliefs can have a sort of domino effect: one might allow many in, which in turn might allow many more. And so the pressure to not have false belief is fairly strong.
Looking back, the stories of the fellow and the tiger all involved him having more beliefs than necessary to motivate the right behaviour. If he merely understood that the tiger was dangerous, that would surely be enough. But in thinking the tiger was friendly, he had to have other false ideas of how to befriend it. In thinking it was lucky, he had to have other false ideas about how to pursue riches. Since there is evolutionary pressure to minimize an animal's false beliefs, these over-complicated, false belief structures are actually less likely to be beneficial, and thus less likely to evolve, than the simple truth. And so, given an evolutionary origin, we would actually expect our faculties to be more or less reliable.
That being said, many parts of our faculties (particular
intuitions, or other belief-forming mechanisms) are actually quite unreliable. On evolutionary naturalism, scattered cases of unreliability here and there are not at all surprising; evolution by natural selection is a messy, imperfect process. But this isn't a problem because over time we've developed
methodologies to help us along. We don’t just blindly follow our gut anymore, but instead put more weight behind the
scientific method than behind any single metaphysical intuition. For example, the results of quantum physics are mindbogglingly odd, but we accept them anyway. And even for fields like ethics where intuition still plays a significant role, we have learned to filter out intuitions through thought experiments, and thus aiming at generalized theories.
Over time we have refined the process by which we acquire belief. But if our faculties were designed by someone who cared about our epistemic situation, why would this refinement be necessary? If God exists, why would he leave so many people in the dark for so long? Why would he have left us to pursue false religions, false ethical theories, false medical practices and so on, especially since these areas actually matter? And so, if anything, it seems our epistemic situation strongly supports Naturalism over Theism. Not only can Naturalism explain the reliability of our faculties but, unlike Theism, it can explain the extent to which our faculties are reliable.
Over time we have refined the process by which we acquire belief. But if our faculties were designed by someone who cared about our epistemic situation, why would this refinement be necessary? If God exists, why would he leave so many people in the dark for so long? Why would he have left us to pursue false religions, false ethical theories, false medical practices and so on, especially since these areas actually matter? And so, if anything, it seems our epistemic situation strongly supports Naturalism over Theism. Not only can Naturalism explain the reliability of our faculties but, unlike Theism, it can explain the extent to which our faculties are reliable.