Epistemology is a curious discipline, which seeks to uncover the nature of knowledge so that we can come to know that we, in fact, know.
That has been how the discipline has been pursued since Descartes, which places scepticism at the centre of inquiry.
I believe, however, this view to be mistaken. It is a view, I think, most epistemological naturalists, among whom I count myself a number, would concur with. That said, most naturalists still stick with the traditional view of knowledge as justified true belief.
The scepticism of Descartes, whatever his intentions, should not be interpreted as outlining a sceptical view of knowledge in order to provide us with a steady and sure foundation for knowledge.
Rather, Descartes should be interpreted as providing a reductio against the traditional conception of knowledge. Since, assuming knowledge to be justified true belief, you cannot know that you are dreaming, subject to the whim of an evil demon, or a brain in a vat, but nonetheless we do know that we know ,it follows, I would argue, that we have a reductio against the traditional picture of knowledge.
The traditional picture of knowledge leads to scepticism, but akin to Hume, we know scepticism to be hopeless so therefore we abandon the traditional view, because it leads to absurd conclusions, not the concept knowledge.
Descartes, but also the ancient sceptics, effectively told us that our intuitive picture of knowledge, justified true belief, cannot be correct long before Gettier gave us his counter examples.
In this sense, then, analytical philosophy is barking up the wrong tree. The analytical style is to break up our concept of knowledge to reach clarity, but analysis of an absurd concept is foolhardy.
I had occasion to think of these matters upon reading a paper by Richard Kitchener  on the naturalism of Bertrand Russell. Kitchener argues that Russell needs to be viewed, especially regarding his epistemology, in a naturalist, rather than an analytical, vein.
I think Kitchener is on to something. Russell always had a passion for the sciences, which he followed closely.
However, Kitchener spends a good amount of time arguing that Russell was concerned with developing a naturalistic account of Peirce’s abduction understood as inference to the best explanation.
I don’t think that is how abduction should be viewed, whatever Russell may have thought of the matter. I think that Harry Frankfurt’s views on abduction are correct.
The interesting thing about Russell’s naturalism, according to Kitchener, is that it was, like Quine’s, behaviourist. Which is to say it’s wrong, but it’s wrong for interesting, and illuminating, reasons.
I would be on safe ground when I assert that Russell, Carnap and Quine were the three leading empiricist philosophers of the 20th century and each can be put into the company of behaviourism. I tend to think that their empiricism, their behaviourism as it were, flows not from an epistemological commitment, that is to empiricism, but rather from an ontological commitment namely physicalism.
To be a naturalist is to be a physicalist and what better naturalistic building block is there for an epistemology that is consistent, especially in their era, with physicalism but behaviourism? It is the commitment to physicalism that leads to an externalist, as opposed to internalist, epistemology.
But trying to develop a theory of knowledge based on a physicalist ontology is to leap frog ahead quite unjustifiably. The first step is to develop a mentalist, internalist, picture of knowledge, leaving physicalism aside until the sciences of the mind are unified with physics, and of physics of an unspecified type.
When you think of matters like so you are led to adopting an epistemology that is coterminous with the cognitive sciences. Most naturalists don’t think like so, alas. The dominant strain of naturalism is neurophilosophy but that too arises from a physicalist ontology, and that too is too premature.
Naturalists need to purge themselves of their proclivity for physicalism and learn to be more patient.
Chomsky always stated that for him the study of language was not necessarily motivated by a desire to understand language per se. Rather, it was hoped, the study of language would shed light on problems of mind and epistemology.
His account of our knowledge of language is the best theoretical tool a naturalist has to build a theory of knowledge. It is one that is branching out into the domain of moral epistemology, and I see no reason why it shouldn’t shed light on science itself.
But when you examine Chomsky’s theory of knowledge of language you begin to appreciate that the traditional justified true belief picture does not hold. Our knowledge of language is not based on externalised warranted assertability. Our knowledge of language is the most powerful Gettier counterexample that we know of.
There’s a good connection here, I believe, between the proper interpretation of scepticism, Gettier, and our knowledge of language.
Naturalists pay heed.
Why are naturalists impatient? Let me venture an hypothesis. The search for knowledge does not arise because we seek to sate the needs of reason. No. The search for knowledge is a passion and the passionate cannot wait. They must know, they will know, they will know now be damned!!
Reason dictates patience but passion demands gratification.
So thus demonstrating that nobody knew us like that master of us all, David Hume.
 Richard Kitchener, “Bertrand Russell’s Naturalistic Epistemology,” Philosophy, Vol 82 No 319 (Jan 2007), pp115-146.