In lieu of an abstract, here is a brief excerpt of the content:

BOOK NOTICES 473 dialectology, while Humberto López Morales presents an excellent summary of dialect research in the Spanish Caribbean to date. The latter paper will be of interest to those researchers working in variability theory, particularly phonological variability. Two papers are based directly on speech data, both from the Caracas speech project: Paola Bentivoglio discusses the use of de que and que as relators in sentences like Comentó con alguien de que ... 'He commented with someone (of) that ...' This paper suggests future studies of this use of de as a syntactic variable for Spanish. Rosalba Iuliano presents a frequency study of the functions of ir a plus infinitive in speech. Mark Goldin's paper, although not based on speech data, presents an interesting discussion of the 'personal' a as a syntactic variable. There is a paper by Deborah Rekart on sibilant development in Catalan, and one by John C. Birmingham Jr. on Papiamentu and its relationships with West African creóles. Gary Scavnicky shows that the active comprehension of indigenous terms in Guatemalan society is diminishing. Two articles deal with bilingualism and bilingual programs. The first is a not very interesting suggestion to use William Bull's model of the communication process in bilingual programs. The second is a nice study of gender acquisition based on tests with 315 bilingual and monolingual students in Los Angeles County. There are two articles on Spanish syntax—one, by Donald Greco, on indirect questions, and the other, by Theodore Higgs, on the epenthetic -n often added in sentences like Estaban golpeándosen 'They were hitting each other.' Robert Hammond discusses the phonological analysis of /b d g/ and their allophones in Spanish: the question of which variant—the fricative or the stop— should be used in the underlying form is discussed, with arguments for and against both analyses. Joan Hooper's archisegment proposal is not included as a possibility. [Tracy D. Terrell, University of California, Irvine.] In defense of structuralism: Transformational and structural morphology . About two rival approaches to the Rumanian verb system. By Alphonse Juilland. (Stanford French and Italian studies, 5.) Saratoga , CA: Anma Libri, 1978. Pp. 77. During the 50's and 60's, there was fierce discussion in several Rumanian journals (and an occasional Western one) as to the status of a set of Rumanian nouns with forms and agreement that match masculine nouns in the singular and feminine nouns in the plural. This is not the place to discuss the relative merits of dubbing such nouns 'neuter' or 'ambigeneric', and to some extent the argument is merely a question of terminology; but it does serve to illustrate how hotly debated a rather trivial issue can become in a language which—although far from ignored in linguistic studies—is, in comparison with other Romance languages, relatively obscure. Juilland's book is the third round in another battle over what is ostensibly a similarly minor issue—whether a set of Rumanian verbs should be regarded as belonging to two separate conjugations, or to two subclasses of the same conjugation. In the earlier controversy, the proponents were generally structuralists of one kind or another; but here J is defending a structuralist approach to morphology against the transformationalist approach followed in Merritt Ruhlen's critical review (in Romance Philology, 1974) of A. Juilland & P. M. Edwards, The Rumanian verb system (The Hague, 1971). J's arguments are quite respectable, and include the following. Ruhlen's criticism is piecemeal: 'To be sure, a solution which accounts for part of the data may prove incapable of accounting for all. The relationship between -ea and -e verbs may be such that a framework devised for their exclusive benefit may be unable to accommodate the more complex relationships between other groups in the system' (8). Ruhlen creates rules that are so complex in terms of their environments that they amount to mere listing, which is what he accuses Juilland & Edwards of. The generative phonology approach supported by Ruhlen has the advantage of hindsight, and uses known historical facts as the basis for underlying forms, even though the child has access only to synchronic data during acquisition. Ruhlen questions the value of a statistical approach, but resorts to it...

pdf

Share