Knowledge of language (and probably much else) can properly be regarded as a system of mental representations and operations to form and modify them. In these terms, many problems concerning language and knowledge can be reasonably approached, sometimes solved, though many mysteries remain, and prospects are opened to incorporate this inquiry within the natural sciences.
Syracuse Scholar (1979-1991):
2, Article 2.
Available at: http://surface.syr.edu/suscholar/vol4/iss2/2