Having been reconciliatory to this extent, I still want to insist that there is a clear way in which an internalist approach, in addition to being intellectually legitimate on its own, has a fundamental kind of priority for epistemology as a whole, so that externalist views, whatever their other merits, do not constitute satisfactory responses to the general issue with which this essay is concerned: that of whether we have any good reasons to think that any of our beliefs about the world are true (and what form these reasons might take).
This is so because externalist justification simply does not speak to this global and essentially first-person issue. One way to see this is to note that if an epistemologist claims that a certain belief or set of beliefs, whether his own or someone else’s, has been arrived at in a reliable way, but says this on the basis of cognitive processes of his own whose reliability is for him merely an external fact to which he has no first-person, internalist access, then the proper conclusion is merely that the belief or beliefs originally in question are reliably arrived at (and perhaps thereby are justified or constitute knowledge in externalist senses) if the epistemologist’s own cognitive processes are reliable in the way that he believes them to be. Of course there might be a whole series of hypothetical results of this sort: cognitive process A is reliable if cognitive process B is reliable, cognitive process B is reliable if cognitive process C is reliable, and so forth. But the only apparent way to arrive at a result that is not ultimately hypothetical in this way is for the reliability of at least some processes to be establishable on the basis of what the epistemologist can know directly or immediately from his first-person, internalist epistemic perspective…
The basic question (which each person must in the end ask for himself or herself) is whether I have good reasons for thinking that my beliefs are true (and, if so, what form those reasons take). And the reason that this leads to an internalist view is that the reasons in question are supposed to be reasons that I have, not in the impossible sense of having them explicitly in mind at every moment, but in the sense of their being more or less immediately available or accessible. (Epistemic Justification, pgs. 39, 174)
[One might instead ask why we need to regard skepticism as a main concern in epistemology. An extended reply would take me to far afield of the present post, but in short, the question suggests its own ironic answer: "why do I need an answer to this question in order for this to be true or known to be true?"]
With the skeptical concern in mind, I turn to this excellent passage by Michael Williams, to which I have nothing further to add at this time:
The essential feature of Agrippan skepticism is that it is universal. The Agrippan Argument applies to any arbitrary belief or claim. This is why it is available for indefinite reiteration. The skeptic’s challenge is to explain how it is possible for us to know (or be justified in believing) anything whatsoever (Stroud, 1989).
The skeptic’s question is peculiar. If I tell a child about dinosaurs, she may ask me how it is possible to know anything whatsoever about them. After all, they went extinct thousands of years ago, so no one has ever seen a dinosaur. I will reply by telling her about the fossil record and how it gives us clues to what different kinds of dinosaur there were, their different structures in turn giving clues to how they lived. But of course, in giving an explanation like this, I am only explaining how it is possible to know some things on the basis of others. What the skeptic wants – and what the traditional epistemologist means to provide – is an explanation of the possibility of knowledge in general. An explanation of the possibility of knowledge that takes certain facts for granted – treats them as if they were known – will lack the requisite generality. The skeptic imposes – and the traditional epistemologist accepts – a Totality Condition on a properly philosophical understanding of knowledge and justification (Williams, 1992).
The Totality Condition creates pressure to accept a further constraint: internalism, or full epistemic self-awareness. Internalism is the view that, to be justiﬁed in holding a belief, we must have “cognitive access” to its “justiﬁcation-makers.” So-called “externalist” theories of knowledge and justiﬁcation, by contrast, allow epistemically appropriate believing to result from factors of which we are not aware. For example, an externalist may say that a belief of mine is epistemically appropriate if it is formed by a method that is in fact highly reliable, whether or not I know about the reliability of the method I used. For externalists, such reliability-knowledge is relevant to the quite different question of whether I know, or am justiﬁed in believing, that my original belief is justiﬁed. One can have beliefs that are epistemically appropriate without understanding why. Presumably, the “knowledge” we attribute to animals is like this. According to externalists, human knowledge is not essentially different (Kornblith, 2002).
Prima facie, internalism is not particularly plausible, at least if it is taken as a fully general view of ordinary justiﬁcation (Goldman, 2001). Everyday justiﬁcation often seems to work as externalists say it does, as, for example, when it ﬂows from the unselfconscious exercise of dependable recognitional abilities (Fogelin, 1994, chapter 3). We do not always require people to have reﬂected systematically on their abilities at large, or even on their performance in the situation at hand. However, in the peculiar context of the skeptical challenge, it is easy to persuade oneself that externalism is not an option.
An explanation of how knowledge or justiﬁcation is possible has to do more than show that knowledge or justiﬁcation is logically possible: that there is a way of thinking about knowledge that does not involve a contradiction. Externalists can surely manage this. They can sketch a consistent picture of the world in which we credit ourselves with reliable faculties and so, by externalist lights, with epistemically appropriate beliefs. But are we justiﬁed in believing that our faculties are reliable? Is that belief epistemically appropriate? If not, then for all we know, we have no justiﬁed beliefs. This is a signiﬁcant concession to skepticism. We want a reply not just to the claim that we know nothing, but also to the meta-skeptical claim that for all we know we know nothing. We want to know that we know. This too pushes us towards internalism.
Admittedly, committed externalists can resist this line of thought. They can say that our epistemic beliefs (meta-beliefs about the reliability of our faculties) may indeed be epistemically appropriate – by externalist standards! But how do they know that? Well, they believe it: appropriately, too, if this meta-epistemic belief is formed in some suitably reliable way. And so on. But that is the problem. More traditionally minded theorists will surely feel that externalists who go down this route are either accepting inﬁnitism or simply turning their backs on the sceptical problem. The traditionalist thought is that the epistemic self-understanding we seek can be attained only if the epistemic appropriateness of our beliefs at large can be made in some way evident. This is an internalist demand.
This pressure to adopt internalism gets further reinforcement from the thought that the context of philosophical inquiry is inherently reﬂective. In doing philosophy, we step back from all particular practical engagements in order to make explicit presuppositions that we normally take for granted without formulating in a precise way. But once such presuppositions are made explicit, the question of their epistemic appropriateness cannot be avoided. In philosophy, we want to get clear about what, at the deepest level, we are committed to and whether we are entitled to those commitments.
To understand philosophical reﬂection this way is to link such reﬂection with a particular ideal of self-understanding. It is because it adopts this ideal that traditional epistemological reﬂection is conducted from a ﬁrst-person standpoint. By contrast, externalist approaches to knowledge and justiﬁcation are elaborated from a third-person point of view. Accordingly, to the traditionalist, externalist epistemologies embody an attitude that we might take to someone else: we see that he is a reliable informant about this or that, and so we take his beliefs in that area to be epistemically appropriate, not worrying about what, if anything, he knows about his own reliability. But it is not clear what would even be meant by proposing to take such an attitude towards oneself. Again, the very character of traditional epistemology, as a quest for total epistemic self-understanding, pushes us to adopt internalism.
Externalist epistemologies seem plausible, the skeptic will suggest, because they accord with what we already believe. Our common-sense–scientiﬁc picture of the world suggests that, within limits, our basic cognitive faculties – perception, memory, and so on – are fairly reliable, so that beliefs formed with the aid of those faculties tend to be epistemically appropriate. But in taking the common-sense picture of the world for granted, we are not explaining how it is possible to know anything whatsoever. (Contemporary Debates in Epistemology, pgs. 207-208)