In the concluding section of the Two Dogmas of Empiricism, Quine describes our entire knowledge as a “man made fabric,” (Two Dogmas of Empiricism, p. 39) an epistemological system that’s essentially a human construct consisting of statements, but one that’s constrained or limited by the vast accumulation of experience. When new experience that’s not confirmed or compatible with the existing construct occurs, there are repercussions that alters statements within the construct, and in turn, causes a ripple effect that changes other “logically connected” (p. 39) statements.
For statements in the center of the construct that aren’t directly linked to (or represented by) any experiences, they can still be affected indirectly through new experiences that occur on the fringes and consequently the linked statement is changed. Such new experience has long range repercussions for the entire construct. This is how human knowledge progresses, by a self-correcting mechanism that incorporates new information and changes, grows, evolving accordingly. Some new experiences bring only minimal changes, which add incrementally to the body of knowledge, and others spark a daisy chain reaction that eventuates into a full blown shift that reorients the entire structure.
Given this epistemological structure, Quine claims it’s misleading to limit empirical content to individual statements alone (p. 40). Statements are not isolated islands of information that float in a sea of inchoate reality, for they are all interconnected in a system of knowledge – a system that is always open to revision.
In addition, he claims that there’s no such divide between analytic and synthetic statements (p. 40). The statements closer to the boundary of the structure are more “synthetic,” because they’re directly linked to the empirical content of experience. As for those deeper in the interior, they are more “analytic,” because they are indirectly linked to experience, but there is no hard and fast divide between the two.
The rejection of the analytic-synthetic distinction has a major consequence: Quine says the entire construct is “underdetermined” (p. 39) by experience, i.e., there is more than one theory that’s consistent with the facts of experience. In other words, there is enough flexibility in the system that allows for adjusting a particular statement in order to incorporate new experience that expands or alters the system.
Moreover, every statement, including those at dead center of the construct, is open for revision (p. 40). That includes even the supposedly timeless laws of logic: Quine cites the strange properties of quantum mechanics that do violate certain fundamental laws of logic such as the law of the excluded middle (p. 40).
Quine offers a horizontal picture of epistemology, a counter-example to the traditional vertically oriented picture of epistemology, and a pragmatic solution that escapes the quagmire that has trapped previous systems that depended on the analytic and synthetic distinction.
Two Dogmas of Empiricism, by W. V. Quine, in The Philosophical Review, Vol. 60, No. 1 (Jan., 1951), 20 – 53