L. and O. adduce examples to refute the claim that informative declaratives are characteristic of a 'normal' conversation or that declaratives are 'normally' used to inform. They claim that the previous attempts to formulate 'rules of conversation' clearly did not result from a serious examination of real conversations.