Principles of syntactic reconstruction (review)

For all the stunning accomplishments of nineteenth-and twentieth-century linguistics in the areas of phonological and morphological reconstruction, syntactic reconstruction remained a backwater throughout most of this period. It is of course easy to see why this should have been the case. The arbitrary nature of the linguistic sign together with the concrete, discrete nature of the linguistic items that feed into the comparative method enables us to reconstruct a starting point; this point may be interpreted as the latest construct that will allow for an unambiguous one-way mapping relationship back to the input items. The syntagmatic axis of this reconstruction is occupied by sounds that are generally subject to well-documented change based on a combination of auditory and articulatory factors. The nature of syntactic reconstruction is significantly different. The constructions we may wish to feed into the comparative method are less discrete, are more abstract, and may not involve the same material substance. They are patterns that are not subject to the same arbitrariness as the word. Consequently, any reconstruction that we arrive at will not possess the same determinable quality as phonological reconstruction.

With such considerations as background, the articles included in this volume were solicited and brought together. After the ‘Foreword’ by Giuseppi Longobardi (ix–xvii), the first article by the editors, ‘Syntactic reconstruction: Method and new insights’ (1–26), introduces a general background on the comparative method, offers a brief historical overview of syntactic reconstruction as well as its relationship to generative syntax, and presents synopses of the six articles that form the core of the volume. The articles are: ‘How much syntactic reconstruction is possible’ by Acrisio Pires and Sarah G. Thomason (27–72); ‘Reconstruction in syntax: Reconstruction of patterns’ by Alice C. Harris (73–96); ‘Reconstructing complex structures: A typological perspective’ by Ferdinand von Mengden (97–120); ‘Competitive Indo-European syntax’ by Rosemarie Lühr (121–60); ‘Principles of syntactic reconstruction and “morphology as paleosyntax”: The case of some secondary Indo-European verbal formations’ by Irene Balles (161–86); and ‘Syntactic change and syntactic borrowing in generative grammar’ by Claire Bowern (187–216).

The main issue that pervades every article in this collection, including the foreword and the introductory overview, is the status of syntactic reconstruction relative to generative grammar. A key distinction is drawn between what Noam Chomsky (1986) called internalized language (Ilanguage), an ‘element of the mind of the person’ (22), and externalized language (E-language), or language as an externalized entity manifested in linguistic data. In contemporary synchronic linguistic theory, it is the former that is universally deemed to be the proper object of linguistic research; and, unless synchronic and diachronic linguistics are different disciplines, each with separate objects of concern, synchronic notions of language should be applicable to the diachronic arena as well, because diachronic layers within language history presuppose synchronic states that are linked to each other via linguistic change. ¹ It therefore follows that I-language should in theory be the proper object of diachronic analysis as well. Although it is perhaps achievable for the diachronic linguistics of the future by those who will study linguistic change in a technologically rich environment from the late twentieth century forward, such a goal is generally impossible to meet for the diachronic linguistics focusing on linguistic states of the past for numerous reasons. First, we do not have access to randomly generated corpora of spoken language of all varieties; rather, only corpora of stylized texts are available to us. Second, any act of syntactic reconstruction recovers merely a piece of the system, not an entire system ‘où tout se tient’, which can then be evaluated from the point of view of I-language as a complete Gestalt. Third, many if not most of the insights and analyses attained by modern syntactic theory depend on grammaticality judgments of native speakers, as they frequently serve...