No CrossRef data available.
Published online by Cambridge University Press: 12 April 2004
In this monograph, Glenn Levine explores what sociolinguistic variables may account for “incomplete acquisition” of a language even though a child speaker is exposed to it within the “critical (or sensitive) period” (Lenneberg 1967, Newport 1990). The participants in Levine's study are ten adult English-Yiddish bilinguals (ranging in age from 63–83) who acquired both languages in the United States simultaneously. The acquisition context that Levine examines is a classic example of a dominant/nondominant language situation, with Yiddish being socially subordinate to American English, as its use was restricted to the participants' immediate families. The main question that Levine addresses is the extent to which Yiddish acquired by his consultants in childhood diverges from the accepted norms (that is, from Standard Yiddish). The data used to answer this question come from the open-ended interviews between the author and the Yiddish-speaking adult bilinguals.
Levine's hypothesis is that some divergences are to be expected, given that all but one of the participants stopped using Yiddish actively at the age of ten and their grammars were therefore likely to have undergone some attrition. In addition, Levine suggests that his consultants may have even failed to acquire some aspects of Yiddish grammar due to the extradeficient nature of linguistic input. The challenge of an analysis, then, is to differentiate between those divergences resulting from language attrition and those that are a consequence of incomplete acquisition (the linguistic product of the latter is referred to as “fossilized child grammar”).
Levine observes that the language data from the ten consultants (“primary informants,” in his study) are insufficient to shed light on the issue, therefore he supplements his investigation with data from three other adult speakers of Yiddish (“secondary informants”), ranging in age from 20 to 55, who acquired Yiddish outside of the United States (Lithuania and Russia) and who maintained it into adulthood. These speakers serve as a control group in providing a baseline for comparison in the analysis. Levine views a more protracted period of use and exposure to Yiddish as the social conditions that bring about a more complete linguistic knowledge (or at least, knowledge that is qualitatively different from that of his primary consultants). The assumption throughout the study is that the divergences in the secondary consultants' speech are likely to result from language attrition, therefore whatever differences are to be found between the two data sets can be attributed to incomplete acquisition.
The structure of Levine's monograph is as follows. Chapter 1 introduces the reader to the goals and theoretical framework (a sociolinguistic/ethnographic approach). Chapter 2 describes the datacollection procedures and profiles of the Yiddish-English bilingual consultants whose speech is analyzed in the study. Chapter 3 provides a description of the data from primary and secondary consultants and argues that the divergences common to the two data sets result from language attrition. In chapter 4, Levine presents his sociolinguistic model of incomplete acquisition and language loss and discusses how the two processes bear on the issues of language death. Chapter 5 centers on the discussion of the “h/z rule” (that is, the use of auxiliaries hobn ‘have’ and zayn ‘be’) in the data of adult bilinguals. The argument is that the primary consultants failed to acquire the usage rules for the two auxiliaries (in contrast to the secondary consultants) and therefore the h/z rule is a feature of the fossilized child grammar.
As Levine points out himself, his study is not “sociolinguistic in the traditional sense” (17). He does not focus on investigating variation in the Yiddish grammars of his primary and secondary consultants, but rather treats the two data sets as two separate grammars of sorts (one is a fossilized grammar and the other is a grammar that has undergone some attrition). Furthermore, an analysis of linguistic features of the two grammar types is based on descriptive statistics and, at times, also lapses into somewhat general and sketchy observations. Chapter 3 begins with the argument that the primary Yiddish-English bilinguals are not native speakers, a conclusion that would have been more credible had it actually been based on a linguistic analysis. The author claims, however, that his primary consultants' speech “cannot be considered as fluent or spontaneous as that of secondary informants” (33), although it is more “fluent” than the speech of L2 learners of Yiddish (like himself). We are also told that the author's primary consultants do not have a substantial memory stock of lexical items and also have a rather limited communicative competence. While the question of what kind of linguistic knowledge defines a native speaker is interesting in itself, the study is not designed to address this question and so the reader is asked to take the author's conclusions on faith.
The second half of chapter 3 is more scientific because Levine discusses the “divergent features” in terms of statistical tendencies. Levine's error analysis is based on 100 consecutive utterances from the transcribed conversations with his primary and secondary consultants. The focus is on verbal and nominal morphosyntax, in particular, features such as subject-verb agreement, past participle formation, use of personal pronouns and determiners, and plural morphology. (The author does not motivate his choice of these linguistic features; however, one is immediately reminded of the study by Johnson and Newport [1989]). Levine argues that his two groups of consultants show quite comparable error rates on this set of features, with the bulk of errors concentrated in the domain of determiner use. All of Levine's consultants demonstrate a tendency toward neutralizing gender and case distinctions in the definite article system, reducing the various determiner forms to one gender- and case-neutral variant, de ‘the’. Due to the similarities of errors in his consultants' speech, Levine proposes treating the analyzed features as resulting from language attrition.
Chapter 5 is an inquiry into the acquisition of the present perfect tense by the two groups of speakers (referred to as the “h/z rule”). The present perfect in (Standard) Yiddish is formed by means of the auxiliaries hobn ‘have’ or zayn ‘be’ and a past participle. The selection of an auxiliary is governed by the aspectual and syntactic properties of the lexical verb that the past participle derives from. As in the case of the determiner system, there is an obvious tendency in the primary consultants' speech toward levelling out the distinction between the two auxiliaries. The primary consultants overuse hobn and extend it to the environments where a speaker of Standard Yiddish would use zayn (namely, to change of location and change of state predicates). The secondary consultants, in contrast, appear to have maintained the h/z rule, a conclusion supported by infrequent occurrences of hobn in zayn-contexts. Levine states that the errors are so sporadic among secondary consultants that no statistical analysis is necessary. (Note that the same argument was not invoked when the minimal error rates were reported for the set of features in chapter 3.)
No study is ever without shortcomings. The main problem with Levine's investigation lies in the assumption that his secondary consultants acquired Yiddish fully in childhood. There are a few reasons to question this assumption. First, the families of his two secondary consultants came from Lithuania and Russia, where Yiddish was certainly not a socially dominant language, and it is most likely that the consultants' parents were bilingual speakers before they arrived in the United States (naturally, there is no way of knowing whether they were balanced bilinguals). Furthermore, one secondary consultant was a U.S.- born speaker of Yiddish and was included in the control group based on her length of language use. (This consultant reported that she learned Yiddish primarily from her grandparents.) Thus it is quite possible that the secondary consultants were also acquiring Yiddish under the same conditions of extra-impoverished input to which the primary consultants were also exposed. If so, the divergences discussed in chapter 3 may well have resulted from “fossilized” competence as opposed to language attrition. Second, one is left to wonder whether the obliteration of the h/z rule in the primary consultants' speech might be simply a characteristic of the Yiddish vernacular dialects that these participants were exposed to as children. Levine mentions that some dialects of Yiddish do not show the h/z distinction, unlike Standard Yiddish, and therefore the consultants may have acquired the vernacular forms of the present perfect. Future studies of acquisition of Yiddish should address this issue.
Levine's study makes an important point by linking fossilized child grammars and language attrition to the process of language loss. The dominant/nondominant language settings in immigrant families certainly create a kind of linguistic discontinuity between younger and older generations of speakers. Children of immigrants (at least, in the United States) in many cases do not appear to view their parents' language as a social asset and therefore do not often make an effort to continue using it throughout adulthood. Levine's study also brings into focus the issue of incomplete child acquisition and opens up new directions for the studies of bilingual development. One obvious line of potentially exciting inquiry is a comparison of fossilized child grammars with fully developed child grammars. This is why Levine's discussion of his consultants' speech will be of interest to language acquisition researchers of various theoretical orientations.