Language loss and the crisis of cognition: Between socio- and psycholinguistics By K. Dallas Kenny (review)

440 LANGUAGE, VOLUME 74, NUMBER 2 (1998) (Terrence Kaufman, co-director of the Mixe-Zoquean linguistic project that has been underway since 1993, has postulated that there are no vowel-initial lexical morphemes in any of the Mixe-Zoquean languages and that such 'vowel-initial' forms as C has listed should be represented as beginning with a glottal stop. The persistence of such glottal stops in the middle of compounds testifies to the correctness of this analysis.) In the earlier version of the dictionary, Popoluca verbs were listed as bare stems without inflectional affixes. In most cases in the new version, however, the main entry of verbs is the first-person-singular present-tense form. This practice, coupled with the fact that person is marked prefixally in Sayula Popoluca , results in almost all verb-entries appearing in the same twelve-page section. Transitive verbs also include a direct-object prefix in their main entries. (A number of verbs are also listed a second time by their third-person-singular present-tense forms, even though in most cases these forms are predictable on the basis of regular morphophonological processes.) In spite of such oddities, this new version of Cs dictionary is in most regards clearly a valuable improvement over the earlier one and. in accordance with the hope that C expresses in his introduction, it should provide both speakers and scholars with an important tool for using and understanding the Sayula Popoluca language. [Dennis Holt, Southern Connecticut State University.] Language loss and the crisis of cognition : Between socio- and psycholinguistics . By K. Dallas Kenny. Berlin & New York: Mouton de Gruyter, 1996. Pp. xiv, 306. The study of language attrition and semispeaker status has usually been undertaken by comparing grammatical and phonological patterns used by less fluent speakers with fluent monolingual speakers of the language in question. Here Kenny takes a different approach, using quantitative and qualitative analyses of hesitation phenomena and pauses to trace the stages of language attrition. This 'macro-fluency model' is instantiated by the study of Arabic-speaking Palestinian immigrants in the United States, classified according to length of stay in the U. S., degree of immersion in anglophone society and culture, and other sociolinguistic variables. Following a short introduction , the book contains ten substantive chapters , a concluding chapter, and a number of appendices. After briefly reviewing various methods of measuring L1 fluency, K presents his 'non-textural, temporal model' as a psycholinguistic alternative capable of producing quantifiable results. Recorded and transcribed interviews were rated by a panel to detect hesitation phenomena, including pause-filling sounds, vowel and consonant lengthening, silent pauses, repair and repetition. These in turn were calibrated against the calculated rate of speech for each informant and against 'verboseness', i.e. the number of words used to respond to a given question. Code switching (mostly from English to Arabic) was also taken to reflect L1 attrition. Interspeaker quantification was achieved by comparing the ratios of 'disfluencies ' per 100 words. Linear regression techniques provide the quantitative correlations between putative L1 fluency and hesitation phenomena. The author departs from the hypotheses that loss of fluency in L1, leading to eventual attrition and loss, will show a linear correlation with time spent in the L2 setting, in this case in the United States. Each of the types of hesitation phenomenon receives a chapter-length treatment, including numerous graphs and charts which break down the data by speaker category. Overall, the data from filled pauses and other verbalized hesitation phenomena do show a positive correlation with length of residence abroad, but silent pauses decline gradually until about the fourteenth year of residence in the U. S., after which point the frequency rises again, thus providing a V-shaped pattern. Verboseness tends to rise with length of stay abroad while rate of speech shows a downward trend. K postulates a crisis period of bilingual shock during the first year or so in the new setting followed by a 'cross-competence' phase lasting until about the fourteenth year, during which bilingual accommodation is gradually achieved. Finally, the 'parallel attrition' phase takes over, leading speakers in the direction of L1 erosion. The results of the quantitative analysis of 'disfluencies ' seem paradoxical. According to...