This is highly related to my Linguistic theory of user-centered design.
Kathryn Bock's From Conceptual Roles to Structural Relations Bridging the Syntactic Cleft mentiones a 1963 Humphry article which suggests that "the stuff of thinking and messages is not the stuff of language." (Kathryn's words)
This contends with the idea that thought is cast in linguistic form, the Sapir-Whorf Hypothesis. But I wonder if there is research out there in support of said hypothesis. Or perhaps only in partial support--isolated situations where language seems to be the basis of thought. How does this relate to the problem of mapping human thought to computer instructions?