Knowledge attribution revisited: a deflationary account