A review of a work in which a systematic and general theory of the nature of the conventions governing the semantics of a natural language is developed, with the object of offering a conceptual framework within which semantic phenomena can be understood in relation to syntax and to the communicative and social aspects of language. The empiricist theory of language is criticized for not supplying an adequate framework for the explanation of language learning. Taxonomy is a solution to the problems (...) caused by the limited application of general names. General concepts are not mastered individually, but rather as systems of the arbitrary division of ranges. Linguistic devices are systems of rules showing how human beings communicate with each other in order to forward specific human purposes. One such device can incorporate another and a natural language is merely a system of such systems. An attempt is made to put Wittgenstein's idea of "meaning as use" to work in the context of a general theory of language. A discussion follows on the relationship of Harrison's views to Chomsky's work in grammar and to recent studies in child language learning. R. Gaskill. (shrink)