Linguistic realities: An autonomist metatheory for the generative enterprise By Philip Carr (review)

392 LANGUAGE, VOLUME 67, NUMBER 2 (1991) One of the main stated goals of this book is to explore the extent to which sociolinguistic method can be applied to the analysis of a dead language, in this case Ancient Greek. The book is in no sense a pilot study, however. Rather, it is based on analysis ofan extremely large body of data: the main corpus comprises a total of 5,480 documents from 33 different geographical regions. Bubenik's specific goals were to trace the decline of Classical Greek dialects and the concomitant rise of Hellenistic Koine, through analysis of scribal documents in the Hellenistic (300-150 B.C.) and Roman (150 B.C-300 A.D.) periods. The details ofthis analysis are provided in Chs. 3 and 4. In Ch. 3 B presents detailed discussions of the decline of 17 dialects (from eight dialect groups—'Strict' Doric, 'Middle' Doric, 'Mild' Doric, Elis, Achaea, Aeolic, Arcado -Cypriot. and Pamphylia). In Ch. 4 B provides detailed analyses of the rise of seven different varieties of koine (Attic-Ionic, Aegean Doric, Achaean Doric, North West Doric, Egyptian, Eastern [Syro-Palestinian], and Asia Minor). These chapters are preceded by three chapters presenting background and methodological information. The first chapter (not numbered and entitled 'Introduction') surveys the range of Greek dialects before the rise of Koine (in the 5th century B.C.), followed by a discussion of the definition of 'koine' with comparisons to Modern Arabic and Modern Greek. The second chapter, Ch. 1 , discusses methodological issues, such as the use of written materials to analyze phonological variation. This chapter also summarizes the inscriptional corpus used in the analysis. B includes as much stylistic variation as possible in this corpus, and he also discusses the classification of inscriptions into various types of public and private styles (e.g. decrees, dedications, and funeral inscriptions). The third background chapter (Ch. 2) surveys the Hellenistic social and linguistic context, dealing with the prevailing political, social, religious, and educational conditions and describing the typical patterns of bidialectalism and bilingualism. Following the two main analytical chapters, B presents a short survey of Hellenistic Koine contact situations (e.g. with Egyptian, Phoenician , Aramaic, and Arabic). Finally, in the Conclusion , B summarizes the main findings of the study and assesses the extent to which the methodological goals have been achieved. He notes (301) that the main value of the study is in the application of sociolinguistic methods to a dead language, rather than in developing new theoretical hypotheses. He further notes that it is not possible to investigate the typical social parameters of age, education, sex, and religion in a dead language, but that it has been possible to investigate the linguistic correlates of nonnative performance, bidialectalism, bilingualism , and to some extent social class. With regard to these latter features, B offers a detailed analysis of the sociolinguistic patterns among Ancient Greek dialects. [Douglas Biber, Northern Arizona University.) Linguistic realities: An autonomist metatheory for the generative enterprise . By Philip Carr. Cambridge : Cambridge University Press, 1990. Pp. ix, 157. $39.50. Some observers see a paradox in the work of linguists who view the theory oflanguage as part of the cognitive and biological theory of human nature; these investigators often seem committed to applying none but standard linguistic research methods in the actual development of linguistic theory. At least in methodological terms, psychologistic generative practice doesn't look much different from Platonist generative practice. Carr's book can be read as an attempt to rationalize this state of affairs. He offers an alternative metatheory that is meant to countenance the nonexperimental research practice he describes. In Part 1, 'The methodological question", C presents what he terms a realist philosophy of linguistics. In a briefreview of some basic issues in the philosophy of science he adopts and defends a perspective due to Karl Popper. He then asks how linguistic theories can be 'proposed and elaborated in the absence of experimentation , and independently of any knowledge of cognitive psychology or social theory' (30). He argues that such an enterprise is indeed scientific in Popper's terms, but that the domain of inquiry is non-nomic—that is, its generalizations should not be construed as...