Contextualism (October 2003):
Many different theories fall under contextualism. What these theories have in common is that in one way or another they make the claim that what counts as knowledge -- or justification or warranted assertibility -- varies from context to context. (The opposite view is invariantism which holds that there are fixed conditions which must be satisfied for something to be knowledge.)
So how do knowledge conditions vary? According to one contextualist theory, there are two different standards for knowledge: a weak (ordinary) standard and a strong (philosophical) standard. This seems to me to at least adequately capture the way we use the word "know" (which is not to say that it necessarily captures what it actually is to know). Consider a dialogue. One person says "I know that X", and the other says "Do you really know X?". The first person then concedes that he doesn't really know X. What is going on here? I think there are two different meanings of "know" here, two different standards for attributing knowledge.
This is just one example of a contextualist theory. In some theories, there there are a broad range of standards for knowledge. In some circumstances we are more willing to attribute knowledge than in others. The requirements for accepting something as knowledge are different in casual conversation than they are in a court or in science. The level of justification required varies from context to context.
One important dichotomy that I'll mention briefly is that between subject and attributor contextualism. Consider the statement, uttered by person A, that person B knows Y. According to subject contextualist, knowledge standards are relative to facts about the person making a claim. According to attributor contextualism, knowledge standards are relative to the attributor of knowledge.
There are many more types of contextualism. One view which is somewhat popular among contextualists is the relative alternative account. This theory holds that knowledge is true belief where the believer is able to rule out all relevant alternatives to the held belief. This can be seen as a form of contextualism because the matter of determining which alternatives are relevant is context-sensitive.