There are certain illocutionary acts (such as hypothesizing, conjecturing, speculating, guessing, and the like) that, contrary to John Searle's (1969, 1975, 1979) speech act theory, cannot be correctly classified as assertives. Searle's sincerity and essential conditions on assertives require, plausibly, that we believe our assertions and that we are committed to their truth. Yet it is a commonly accepted scientific practice to propose and investigate a hypothesis without believing it or being at all committed to its truth. Searle's attempt to accommodate such conjectural acts by claiming that the degree of belief and of commitment expressed by some assertives “may approach or even reach zero“ (1979: 13) is unsuccessful, since it evacuates his thesis that these are substantive necessary conditions on assertives of any force. The illocutionary acts in question are central to scientific activity and so cannot be plausibly ignored by a theory of speech acts. The problem is not limited simply to Searle's theory, since even theories which depart markedly from Searle's in other respects are often committed to similar characterizations of assertion.