Received conceptions of the meanings of scientific terms assign to meanings an essentially benign methodological role: the meaning of a term consists of principles or inference rules which are, always or for the most part, (approximately) true or reliable. In fact, many scientific terms have meanings which are malignant: which are mainly false or misleading and which detract from, rather than contribute to, scientific progress. Kuhn's conception of incommensurability can be fruitfully extended to take account of malignant meanings. Malignant meanings are especially implicated in cases, like that of human sociobiology, in which the influence of social ideology on scientific practice is especially profound.