History repeats itself, with variations. Two decades ago American structuralists were trying, with indifferent success, to apply to morphology the same analytical techniques that had proved successful in the analysis of sound. For a number of reasons—including the lack of a suitable theory of meaning—the attempt made no headway at a time when phonology was still scoring advances with help from both acoustics and information theory. Morphemics still remains, in current texts on linguistics, a kind of relic of the 1940's. Now we witness a revived attempt from a different direction, but with essentially the same desire: to try out in a new field the techniques that have been developed in an older one. The new field is meaning, the old one is syntax, and the techniques are those of generative grammar. For the moment, morphemics is only slightly involved, but the signs are clear.

0.1. The aim of this paper is to investigate the relations between the layering in structural trees and the scale of rank. My standpoint will be that of scale-and-category grammatical theory as I understand it.

This paper examines the processes of question formation in French. My purpose is to propose a generative analysis to account for these processes for which one can make a reasonable claim of descriptive adequacy. Attention will first be focused on simple yes-no and ‘wh’ questions. Then, the analysis being expanded to accommodate the est-ce que series, a close relationship will be demonstrated between these and certain embedding mechanisms. Finally, I shall indicate how the analysis presented may be extended in a natural way to account for various other phenomena of French syntax.

In 1934, when Yuen-Ren Chao discussed the nonuniqueness of phonemic solutions, he demonstrated that different phonological analyses presuppose different underlying concepts. The thirty years since the publication of that paper have not produced any one phonological theory to which even most linguists will subscribe. The situation is not unusual in the history of science. Scientific theories are constantly submitted to change and revision; in an empirical science, the structure is never completed. Experiments become more precise, new phenomena are revealed, new concepts are developed. At all times we must be prepared to abandon our theory, remodel the foundations, and erect a new structure. But the history of science shows also that theories which turn out to be inaccurate can yet lead to the discovery of new facts; and each new theory retains some elements of the one it has supplanted. Acceptance of one theory does not require one to reject a rival theory in toto.

The dichotomy between exocentric and endocentric constructions seems well established in linguistics. Its importance in syntactic analysis is obvious. In this connection it is enlightening to refer back to Bloomfield's view, in which English syntax, for example, emerges as strongly exocentric, since the more common of his two favorite sentence types calls for the exocentric actor-action construction. The characterization is further supported by the frequency of his other exocentric constructions (relation-axis, clause-subordination, phrase-subordination).

1. Introduction. In the course of building an automatic essay-paraphrasing computer program it became apparent to me that certain quantitative features of style could be controlled in the paraphrasing process. This paper presents the results of several experiments in the control of two quantitative measures of style, frequency of occurrence of syntactic structures and frequency of occurrence of pronouns. The results suggest that any well-defined quantitative measure of style is subject to machine control.