Derived intransitivity: a contrastive analysis of certain reflexive verbs in German, Russian and English By David J. Cranmer (review)

704LANGUAGE, VOLUME 54, NUMBER 3 (1978) deletion (for an explicit derivation and further justification, cf. Haiman 1974:130-35). An alternative theory, that predicate adjectives occur in the neuter unless they follow the subject nouns with which they are supposed to agree, cannot be maintained : (11)a. Staunchel-s (m) ei igl urn iu-s (m) alia radunonza. 'Tired, the man went to the meeting.' b. Bial-s (m) eis il cuolm (m). 'Beautiful is the meadow.' Thus there is nothing irregular about 9 and 10 if an abstract syntactic analysis for these sentences is permitted. Sentences like 6 continue to pose a problem. S has made an important contribution in discovering such sentences, but his analysis should be pushed further. Do these sentences differ in meaning from the corresponding sentences like 7, which alone are accepted by prescriptive grammarians? If so, do they differ in a way that would allow us to propose a more abstract underlying structure for 6, e.g. one in which the underlying subject was sentential? Perhaps the difference in meaning between 6 and 7 is parallel to the difference in meaning between (12)a. Too many cooks spoil the broth, (proverb reading) b. Too many cooks spoil the broth, (like 'Too many people borrow my security blanket.') On the whole, the most disconcerting thing about the whole book is the way its authors persist in treating Romantsch and Ladin as dead languages. This is not a cause for complaint in historical studies, like that of Kuen; but is frustrating in synchronic analyses such as that of Stimm, which have much to interest the nonspecialist . REFERENCES Boehmer, E. 1871. Prädicatscasus im Rätoromanischen. Romanische Studien 7.210-26. Gartner, T. 1910. Handbuch der rätoromanischen Sprache und Literatur. Halle: Niemeyer. Haiman, J. 1974. Targets and syntactic change. The Hague: Mouton. [Received 25 June 1977.] Derived intransitivity : a contrastive analysis of certain reflexive verbs in German, Russian and English. By David J. Cranmer. (Linguistische Arbeiten, 38.) Tübingen: Niemeyer, 1976. Pp. viii, 117. DM 26.00. Reviewed by Bernard Comrie, University of Cambridge This book is concerned with the contrastive analysis of Russian verbs in -sja (after vowels, -s'), German verbs with sich, and English transitive/intransitive verb pairs like boil (John boils the water vs. The water boils). Although 'reflexive verb' is used as an informal term, Cranmer argues that this is misleading, since so-called reflexive verbs do not necessarily have reflexive meaning, i.e. do not necessarily indicate an action performed by the agent on himself—e.g. Ger. Die Tür öffnete sich 'The door opened', Russ. Sobaka kusaetsja 'The dog bites.' Moreover, other constructions that may express reflexive meaning behave differently syntactically from the above-mentioned constructions, e.g. Russian verbs with the reflexive pronoun sebja, English verbs with the reflexive pronoun in -self—with a few nonproductive exceptions, e.g. (History) repeats itself (82-5). REVIEWS705 The theoretical basis of the analysis is provided by Bowers (ms), dealing mainly with English data, and Babby 1975, dealing with Russian data.1 In their model, a given verb is specified lexically as taking a certain number of NP arguments. It is possible that, for a given verb, some of these arguments may be optionally or obligatorily empty. In the course of a derivation, NP's may be moved to an empty node (thereby leaving behind an empty node); new NP's may be inserted by attaching them under an empty node (e.g. expletive es in German); and NP's may be deleted (again leaving behind an empty node). Cranmer's main argument, here following Babby, is that the basic function of Russ. -sja (and, likewise, Ger. sich) is to provide surface indication of certain changes in valency of these types. The main function of -sja and sich is thus syntactic, rather than semantic. Earlier analyses of Russ. -sja verbs posit a number of different functions of -sja, providing at best a taxonomy of uses and not attempting to give a general characterization of the particle's function. These specific uses include (i) passivization, e.g. èta dver' zakryvaetsja nami 'This door is being closed by us'; (ii) anti-causative, e...