Myth of the Given

There must be some firm foundation on which the rest of our knowledge is built by various inferential methods. This was the traditional view, foundationalism, and is still to some extent the common-sense view of how knowledge works. But it has been increasingly losing favour in philosophy, especially since Sellars' famous argument, "the myth of the given."

Sellars' argument works like this: 

Broken down foundationalism makes two claims.

(1) Epistemic Independence Requirement (EIR). There must be cognitive states that are basic in the sense that they possess some positive epistemic status independently of their epistemic relations to any other cognitive states. 

(2) Epistemic Efficacy Requirement (EER). Every none-basic cognitive state can possess positive epistemic status only because of the epistemic relations it bears, directly or indirectly, to basic cognitive states. Thus the basic states must provide the ultimate support for the rest of our knowledge. Such basic, independent and efficacious, cognitive states would be the given. 

Sellar argues that you cannot have both. The standard candidates for basic empirical knowledge either fail EER (e.g., knowledge of sense-data), or presuppose other knowledge on the part of the knower, and thus fail EIR (e.g., knowledge of appearances).