Different ways for belief to be independent of evidence
Necessary truth: has to be true and could not be false.
For example, Red is not blue.
But not all apriori truths are necessary.
For example, "I am here" is an apriori belief (doesn't need evidence) but not necessary.
On the other hand, not all necessary truths are apriori.
For example, "Mark Twain is Samuel Clements".
Some analytic truths are not necessary.
For example, "Henry II's foot is one-foot long".
Kant and Quine
All analytic beliefs are apriori.
But Kant said: not all apriori beliefs are analytic (synthetic apriori belief).
For example, geometric axiom.
Before we describe and understand the world, we have to make some assumptions.
No belief is pure analytic and sage from rejection.
All beliefs are located somewhere in the vast network called "web of belief".
People in principle have the choice of either keeping or abandoning any belief whatsoever.
Kant said there are certain beliefs must always be the center of the web.
Quine said there must always be a center, but different beliefs can occupy there depending on
certain world situation.
Two-way linkage of concepts
The web of belief is also a web of concepts.
Apparently, every belief depends on its component concepts for making sense.
Experimental versus theoretical concepts
For the former, specifically connected to classifying sense data or gathering evidence for them.
For the latter, explain complicated phenomena by constructing theories.
The degree of abstractness of concept and its muti-relatedness with other concepts give it some
from its perceptual aspect or link to sense data.