Reconstructing Grammar: Comparative Linguistics and Grammaticalization (review)

From the vantage of nineteenth-century comparative linguistics, this title would seem something of an oxymoron: Traditionally, phonemes and words were reconstructed, not grammar. During the past two decades, however, grammaticalization theory has begun to alter such perceptions. This collection of eight articles, most dealing with the historical morphosyntax of indigenous languages of the Americas (two discuss African languages), combines successful insights from genetic and areal linguistics with advances in internal reconstruction, a generally underappreciated method of great potential value for deciphering the origins of complex polysynthetic structures. While traditional comparativists based their diachronic sound laws upon synchronic regularities observed across different languages, practitioners of internal reconstruction focus on exceptional features in a single language, seeking to explain them as the end product of typologically ordinary changes. Until recently, the strength of this method as a tool of reconstruction has only been realized sporadically, since those who study grammaticalization have more often been primarily concerned with demonstrating the synchronic relevance of typological universals.

One of the articles, ‘Internal reconstruction: As method, as theory’ (107–59), by Talmy Givón, analyzes the origin of Swahili verb morphology before attempting to unravel a thornier typological conundrum: the development of a complex series of prefixes in Tolowa (Oregon Athabaskan), an SOV language that should be expected to have suffixes instead. Givón makes a plausible case that Athabaskan, like Bantu, developed its polysynthetic prefixal morphology when serial verbs, some already containing their own inflectional suffixes, procliticized as aspectual markers on the rightmost verb in the series, which retained its TAM suffix as the verb’s only suffixal position.

Most of the remaining articles employ cross-language comparison alongside internal reconstruction, demonstrating the value of both methods when used together. Alexandra Aikhenvald’s ‘Areal typology and grammaticalization: The emergence of new verbal morphology in an obsolescent language’ (1–37) demonstrates how Tariana (Arawakan) recently acquired verb compounding through diffusion from the genetically unrelated Tukano. In ‘Florescence as a force in grammaticalization’ (39–64) Wallace Chafe investigates the development of noun incorporation in Northern Iroquoian which, though highly productive, creates new lexical items and thus does not constitute an actual case of grammaticalization. Spike Gildea’s ‘On the genesis of the verb phrase in Cariban languages: Diversity through reanalysis’ (65–105) examines the origins of five different patterns of constituency and uses comparative data as well as internal reconstruction to account for their origin and present distribution. Bernd Heine’s ‘Grammaticalization chains across languages: An example from Khoisan’ (177–99) explores the distribution of person-number-gender markers on different parts of speech in several Central Khoisan languages. Heine also concludes that internal reconstruction must be augmented by comparative data to justify any claims about the stages in the development of gender concord. Sérgio Meira’s ‘The accidental intransitive split in the Cariban family’ (201–30) uses internal reconstruction and family-internal comparison to explain how a semantically anomalous split- S case-marking system in Cariban developed when a detransitivizing affix acquired new functions. Marianne Mithun uses both internal and comparative Iroquoian evidence in ‘The reordering of morphemes’ (231–55) to explain positional shifts among verb suffixes in Cherokee.

Joseph Greenberg’s ‘The concept of proof in genetic linguistics’ (161–75) seems out of place here as it does not actually confront any specific instance of grammaticalization. Greenberg argues more generally for a ‘proto-comparative’ taxonomic approach to genetic classification that does not insist upon systematic sound correspondences or reconstruction as a prerequisite for accepting new language taxons, demonstrating that no such standard of proof was originally required for now universally accepted families such as Uralic or Indo-European.

Although most of the contributing authors employ internal reconstruction as a subsidiary approach, the awareness this book will bring to general problems of diachronic morphosyntax can only draw attention to it as a tool for investigating other typologically anomalous...