During the past few decades, a great progress has been made in American linguistics, especially in descriptive, or, more exactly, structural, linguistics. Historically, there is of course a connection between American and European linguistics. Descriptive linguistics, building on the Prague school to evolve a phonemic theory and taking hints from F. de Saussure for general orientation, has proved popular in the United States. It is natural that we should find a similar movement going on in Europe over the past same years. Among the chief founders of present American descriptive linguistics were W. D. Whitney, F. Boas, E. Sapir and L.. Bloomfield. The scientific descriptive linguistics intends to describe a language in terms of its own structure, not in terms of the Greek and medieval notions of linguistics. This American development can be explained in several ways. Science in general has recently progressed so much as to enable scholars to build up really scientific descriptive methods. Structuralism has developed very largely in connection with the modern development of scientific grammar as opposed to conventional one. It may well be said that American linguistics has also developed very largely as both anthropological and missionary approach. As attention widened to include languages other than the Indo-European, especially the native languages of America, the parochial and limited character of the traditional linguistic metalanguage has become obvious. Cultural anthropologists, more or less standing on the linguistic relativity principle, think of language as the symbolic guide to culture. Structural linguistics intends to describe `language' structure through applying the methods of linguistic analysis to the raw material of `speech.' It deals not with the whole of speech activities, but with the regularities in certain features of speech. These regularities are in the distributional relations among the constituent elements of language. The substitution procedure through what may be called the bipartite division is the funfamental method of linguistic analysis, It is primarily directed at the discovery of distribution as the only relevant criterion. Structural linguistics is based on the phonemic approach to linguistics. However, it sometimes emphasizes too much the phonemic criteria in determining constructional units. In dealing with a culture language, the influence of the literary tradition in written form cannot be neglected. The linguistic relativity principle is apt to fall into linguistic determinism. It should no be considered that there exist perfect correlations between cultural norms and linguistic patterns. In structural analysis, the knowledge of meanings is only a means that makes the substitution procedure possible. However, in order to study culture through language, we should return to E. Sapir's standpoint that took lexical as well as structural meanings into great considerarion.Descriptive linguistics deals not with the whole of speech activities, but with one phase. It is only a part of the language study in a wider sense. American linguistics has too much emphasized its one phase. We should not forget that the complete language study consists in the combination of descriptive, historical and comparative linguistics. When we think of human language in general, we shall also have to consider universa grammar, even in describing a language in terms of its own structure. Nevertheless, we should admit that structural linguistics is making a great contribution to the scientific study of language.