One of the annoying habits of philosophers is to substitute without warning a normative for a descriptive theory of the topic they are discussing—that is, in what purports to be a statement of how the subject actually presents itself they tell us instead how it ought to present itself. Current treatments of the elusive topic “meaning” seem to me to supply capital instances of this vice. Defenders of a positivist or an operational theory of meaning (I pick these two because of their wide contemporary influence) often give us no hint that their conclusions are conclusions about how the meaning of words ought to be interpreted, not about how, by and large, it is interpreted. Without such a hint their readers naturally suppose that it is the latter problem with which they are concerned, but this is rarely the case; were it so, emphasis could hardly fail to fall on determinants of meaning that are too often neglected by philosophical analysts.