Styles of Rationality was written 20 years ago, well prior to any thoughts of (unicepts of) unicepts. The question of Hurley and Nudds’ book Rational Animals? was whether non-human animals were rational in any sense. I would have to think through that again now. It would be important not to allow it to become a verbal issue.
I am not sure now of any argument requiring language to be either prior to or coincident with declarative thought. Clearly it is enormously helpful, however, given the new projects of declarative thought.
Sellars’s “space of reasons” supplied his concepts with content, replacing “the given.” Talking with Sellars I am quite sure that he came to believe what later became Robert Brandom’s “Inferentialism.” He faltered, I think, at the end of his life by trying to understand animal thought on something akin to the same model. (Unfortunately he didn’t know Gibson’s work.)
Inferential connections among unicepts would differ pretty sharply from Sellars/Brandom inferential connections among concepts. They would not be determined by one’s language; they would differ from person to person.