Studies in Romance linguistics Ed. by Michio Peter Hagiwara (review)

BOOK NOTICES 471 A preferable solution to the tense-labeling problem, inaugurated by Damourette-Pichon and used in a number of recent French grammars, is simply to refer to tenses (' tiroirs') by a surface form, e.g. 'tiroir a lavé, lavant'. The authors' enthusiasm for tables, lists, and 'combinatoires maximales' has further led them to postulate theoretical tense-forms unlikely to occur in actual discourse (have been being washed, ayant eu été regardé). The fact that they are not native speakers of English does not excuse an appalling number of questionable examples—e.g., They were used to her being joking; Having done, he took his leave; His being smoking again was remarked upon; John [sic] having disappeared was a common occurrence. Proper sequence of tenses also seems to be a problem (/ am painting the house before he has arrived / after he has left; Je repeins la maison avant qu'il ne soit arrivé), particularly in the case of anteriority relationships—which in English, as opposed to French, often need not be marked by a 'perfect' tense: I'm painting the house (= I'm going to paint the house) before he leaves. Judgments of acceptability in both French and English are at times questionable, as is the interpretation of certain sentences. Thus He had his car stolen is cited to illustrate causative had; but unless the subject is out to defraud his insurance company, he is more likely patient than agent. On balance, this book falls short of achieving its objectives. Notwithstanding the purportedly innovative approach, elaborate formalism, and numerous intriguing diagrams, the study has not furthered my understanding of the workings of aspect. [Suzanne Fleischman, Berkeley.] Studies in Romance linguistics. Ed. by Michio Peter Hagiwara. Rowley: Newbury House, 1977. Pp. viii, 345. This collection of nineteen papers, plus one panel discussion (on the role of applied linguistics in the teaching of Romance languages), represents the proceedings of the 5th Symposium on Romance Languages, held at the University of Michigan in 1975. Unlike the first in this series of gatherings, held at the University of Florida in 1971 and also published by Newbury House, the present volume has no central theme. The result is a perhaps random but not uninteresting selection of papers, treating various Romance languages from differing perspectives. Ten of the papers are on phonology and morphology, and only five on syntax and semantics—though there is a combined syntax and phonology paper on French Creole dialects. There are also two papers specifically on substratum theory, and a general paper on past and future trends in Romance linguistics by R. A. Hall Jr. This imbalance between phonology and syntax reflects a tradition within Romance studies which has only recently begun to be redressed, thanks to the advent of transformational approaches to language. This influence is not necessarily a source of satisfaction to all linguists: thus Hall's paper contains an attack on modern syntactic theory, claiming that, although the data collected are very useful, they will, 'eventually have to be recovered and restated, where necessary, in better terms' (313). Of the papers on syntax, two in particular (D. J. Napoli and M. Nespor on Italian comparative structures, and T. Terrell and M. E. Garcia on Mexican Spanish mood) reflect the concern in the 70's with the relevance of pragmatic presupposition to syntactic descriptions. The paper by J. Casagrande and L. Jackson on quantifiers in French and Italian seeks to justify the use of transderivational constraints, and thus likewise demonstrates an interest in areas of concern to generative semanticists. A further notable imbalance is in the range of languages covered. French, Italian, and Spanish are well represented, but there is only one paper on Portuguese, and none specifically on Rumanian. Two exceptions are the papers by F. Agard on Romance auxiliaries and by Y. Malkiel on gender; both give space to Rumanian, a language of particular interest with regard to gender because its treatment of the Latin declension system has given rise to the so-called ambigenerics (masculine forms and agreement in the singular and feminine in the plural). A. Valdman's paper on French Creole dialects also shows a healthy concern for more peripheral Romance, and furthermore...