Our thinking is often compartmentalized. We are able to walk and talk, in one compartment we attend to where we are and where we are going while another part of us is focused on a conversation.

David Lewis used this fact in an attempt to avoid an inflation of knowledge in his account of knowledge. A compartmentalized subject, he suggested, could know the truth of two propositions and yet fail to know their (trivial) logical consequence when they are known in different compartments.
 
In evaluating this suggestion I will show that Lewis’s epistemology is more anti-skeptical than previously supposed, but also, that compartmentalized knowledge has  interesting general features. A compartmentalized subject will have a harder time knowing that she does not know a proposition than knowing that she knows. A Lewisian subject will find it hard to even known that she doesn’t believe a proposition.

I will close with some general considerations regarding compartmentalized knowledge.