General

that children shift from reliance on semantic, pragmatic and frequency factors to a sense of structural transparency and saliency; and that several general children's deviations persist strongly among lower-class young teenagers and sometimes even into adulthood (for example, word-initial ' f' for ' p ' in verbs of the type pizer and past tense forms of the type niketi for standard nikiti). It is striking that there is no talk of gender skewing, of the type reported by Cheshire, in Trudgill (ed.), Sociolinguistic patterns in British English, 1978. Not surprisingly, other non-standard forms were typically juvenile, some almost vanishing by the age of five and others persisting for various lengths of time. Ravid devotes the majority of her book to a thoughtful discussion of her findings in terms of (a) interaction between structural opacity in the adult language and strategies of language acquisition, particularly rote memorization, linear simplicity (impeding stem alternations among very young children), formal consistency (thus generalizing three-letter roots across the entire root system), semantic transparency (thus reducing b/v, k/kh, p/f alternation) and saliency (tending, for example, to prefer the analytic to the synthetic); and (b) general factors arrayed in language change: opacity, consistency, salience, inertia and literacy. Here, Ravid articulates a notion of benefit weighed against the cost of offending such basic principles as markedness, of damage to the overall system or simple flouting of tradition. Her conclusion is that language change is due to the deviations not of young children but of older children, naive (i.e. less literate) speakers and naive (i.e. non-selfconscious) speech. One can agree with her analysis of the strategies involved, treating Hebrew acquisition like that of any other language; but I have strong reservations about the author's understanding of the creation of Modern Hebrew and of how this somehow makes for a very special case of language change. Quite simply, we have scant idea of what first-generation spoken Hebrew was like. It is clear that the first generation of Hebrew speaking Ashkenazim were adult males (of the first and particularly the second Aliyah), drawing on a passive proficiency in Hebrew, but we know little about the morphological features of the Hebrew they contrived to produce or indeed about the morphology or phonology of the Ashkenazi Hebrew they had read or heard in Europe—and the relationship between the two. (See for example Glinert, ' On the sources of modern colloquial Hebrew', Leshonenu, 55). The first generation of Hebrew-speaking children may in fact have acquired from their elders some of the persistent non-prescriptive forms that Ravid regards as juvenile formations. Indeed, the same may hold for some of the non-prescriptive forms prevalent among disadvantaged (sc. Sephardi) speakers; here, too, owing to the policy of settling Sephardi immigrants in largely immigrant Sephardi neighbourhoods, children may have been particularly exposed to a Diaspora-based Sephardi Hebrew of which we know little. All in all, Ravid is unwise to regard Modern Hebrew as issuing directly from Classical Hebrew; ancient norms have indeed been prescribed and even used by teachers, authors and their ilk, but this has been just one factor in the emergence of spoken usage. Thus, while some spirant/plosive deviations such as word-initial spirants (fizer) are undoubtedly new juvenile forms, others such as lid/ok may well have a long pedigree in the so-called 'corrupt' Hebrew of Yiddishspeaking Europe.