Full Idea
A way of knowing there are cookies in the jar - visual perception - is not a way of knowing what one knows to be implied by this - that visual appearances are not misleading.
Gist of Idea
Knowing by visual perception is not the same as knowing by implication
Source
Fred Dretske (The Case against Closure (and reply) [2005], p.29)
Book Reference
'Contemporary Debates in Epistemology (2nd ed)', ed/tr. Steup/Turri/Sosa [Wiley Blackwell 2014], p.29
A Reaction
Why is the 'way of knowing' relevant? Isn't the only question that of whether implication of a truth is in infallible route to a truth (modus ponens)? If you know THAT it is true, then you must believe it, and implication is top quality justification. No?