Copulas: Universals in the categorization of the lexicon (review)

This is an important book on a fascinating topic. It marshals a great deal of fresh data from languages around the world and introduces new techniques and analytical perspectives. Like Stassen 1997, it is a must-read for anyone seriously interested in typological and functional approaches to copula constructions and intransitive predication generally. I first give an overall summary, then expand and comment on some specific findings and issues of interest.

Pustet’s book consists of five chapters, followed by detailed appendices. Ch. 1, ‘Copulas in current research’ (1–26), sets the scene, considers definitional questions, and relates copula constructions to issues of parts of speech and markedness. P’s key criterion is that copula ‘does not add any semantic content to the predicate phrase it is contained in’ (5). Hence the ‘copula paradox’ and one of the central questions of this study: if copulas are meaningless, why should they exist at all? Ch. 2, ‘Copulas in cross-linguistic perspective’ (27–82), reports on a grammar-based survey of 131 languages, followed by personal follow-up with native speakers on twenty-two languages. It confirms previous findings that the hierarchy nominals > adjectives > verbals is strongly correlated with the distribution of lexical copulas, and raises the issue of to what extent markedness, in the sense of Croft 1991 (i.e. low discourse frequency coupled with high structural complexity), can be held accountable.

Ch. 3, ‘Copularization and lexical semantics’ (81–152), is a more fine-grained semantic study based on original data from twenty-two languages. Using a lexical sampling technique, it identifies and compares language-specific ‘lexical minimal pairs’ which differ in whether they occur with a lexical copula, for example, English sleep vs. asleep, smell vs. smelly. The main finding is that four semantic parameters—transience, dynamicity, (semantic) transitivity, and dependence—account for most of the distributional differences. Each of these parameters is discussed at some length and their connection to the parts-of-speech issue is explored. Ch. 4, ‘The multi-factor model’ (153–84), sets out to combine information from several parameters to increase the language-internal predictiveness. It proposes that the lexicon can be partitioned into seventeen lexical classes, each with a distinctive ‘semantic profile’ defined in terms of three semantic parameters: transience, dynamicity, and valence (a conflation of dependence and transitivity). It presents a series of diagrams showing how the distribution of copulas in ten different languages is arrayed in three-dimensional ‘semantic similarity space’. The main claim is that ‘copula usage follows patterns which are imposed by a coherent semantic superstructure that organises the lexicon of any one language’ (179).

In Ch. 5, ‘Synopsis’ (186–95), P relates these findings to principles of markedness and ‘cognitive economy’, drawingon prototype theory, and she makes the case that her theoretical and methodological approach can be fruitfully applied to broader questions. She also revisits the issue of why so many languages have copulas in the first place, since many other languages apparently get by without them. There are six appendices (196–247) giving, among other things, details of the languages, further statistical breakdowns on copula distribution by lexical category, the questionnaire used to sample copula usage in the minimal pairs study, and an analysis of the semantic structure of the lexical samples. Aside from Leon Stassen, the other scholars most frequently cited are Bill Croft, Talmy Givón, Ronald Langacker, and Anna Wierzbicka. A valuable feature of the book is the frequent focused discussions of facts of particular languages, including Lakhota, Jacaltec, Basque, Turkish, Burmese, and Indonesian, often from the author’s original field data.

Despite its many useful features, there are certain gaps and weak areas in P’s study. The first concerns its scope. By confining her attention to semantically empty copulas, P may be preventing herself from developing a full explanatory account. Stassen (1997), in contrast, considered a broader field, which also included identificational ‘be’ and locational ‘be’, in uses such...