The critique by Lanin-Kettering and Harrow (1985) of Chaika's article (1982) overlooks the essential distinction between speech, an overt behavior, and thought, a cognitive process inferred on the basis of many different overt behaviors including speech. It is argued that the fundamental importance of this distinction requires the fullest possible analysis of psychotic speech—qua speech— since no inference, however desirable, can be more secure than its base in observation.