Language classification: History and method (review)

Historical linguistics and language classification questions have recently been prominent in the field (as well as in the wider media) because of work demonstrating connections between Athabaskan and Yenisei languages (Vajda 2000, 2001) and the use of computational biological methods to study change, subgrouping, and population spread (Gray et al. 2007). Campbell and Poser have written a complex book that covers many aspects of language classification. They make contributions to the history of the field, to the development of ideas about particular language relationships, and to long-standing arguments about the validity of methods of long-range comparison. The authors’ aims are to demonstrate how methods have been used in historical linguistics, to explain which methods have worked and which have not, to determine how particular language families have been established (that is, under what criteria and how reliably), to evaluate some controversial proposals of distant genetic relationship, and finally to make recommendations about future research.

The first chapter (1–12) introduces the history of the evolution in Western thought regarding language classification and genetic relationship. Ch. 2, ‘The beginning of comparative linguistics’ (13–31), describes early contributions, while Ch. 3, ‘ “Asiatic Jones, Oriental Jones”: Sir William Jones’ role in the raise of comparative linguistics’ (32–47), concerns the role Sir William Jones and his famous ‘philologer’ passage have played and the accuracy of interpretation of the passage. The authors then move to method: how to show that languages are related, and applications of the comparative method to Indo-European (Ch. 5, ‘How some languages were shown to belong to Indo-European’, 74–86) and to other languages and regions (Ch. 6, ‘Comparative linguistics of other language families and regions’, 87–161).

Ch. 7, ‘How to show languages are related: The methods’ (162–223), concerns the type of evidence that can be used to demonstrate that languages are related, while Ch. 8, ‘The philosophical-psychological-typological-evolutionary approach to language relationships’ (224–33), is a brief discussion of glottogonic thought. Ch. 9, ‘Assessment of proposed distant genetic relationships’ (234–96), and Ch. 10, ‘Beyond the comparative method?’ (297–329), evaluate distant genetic relationships. Ch. 11, ‘Why and how do languages diversify and spread?’ (330–63), is about the mechanics of and reasons for the diversification of languages and the spread of their speakers. Ch. 12, ‘What can we learn about the earliest human language by comparing languages known today?’ (364–93), discusses the relationship between modern languages and how they can be used to make inferences about the earliest forms of human language, while the final chapter, ‘Conclusions: Anticipating the future’ (394–403), summarizes the book. There is an appendix about hypothesized distant genetic relationships, which provides a useful reference list and some comments about how controversial each proposal is.

The book touches on many areas related to language classification within historical linguistics: how questions of classification are bound to reconstruction methodology, how evidence is evaluated, and how much evidence (and what type) constitutes an appropriate demonstration of relationships. For the most part, however, it is concerned more specifically with a certain intellectual tradition and a particular argument over methods. The intellectual tradition is the role of the comparative method in historical linguistics, particularly as practiced in European and American universities. The argument concerns which methods are most accurate and reliable when we wish to use language to find out about past human history, and specifically how we should evaluate the contributions of Joseph Greenberg (see, for example, Greenberg 1966, 1987, 2002, and other publications). The methods and results of Greenberg’s multilateral comparisons have been comprehensively discredited. Critics have pointed to numerous problems in methodology, data, and assumptions (Matisoff 1990, Campbell 1997, Ringe 1999), that is, at all levels of the enterprise. In this book, the focus on tying arguments back to Greenberg’s errors (42–43, 57–58, 62–63, 66, 79–80, 127ff., etc.) at times draws the reader’s attention from points of more general history and developments of individual families. [End Page 706...