Auxiliation: An Enquiry into the Nature of Grammaticalization (review)

Auxiliation is a subfield of the more familiar field of grammaticalization. The process of auxiliation is similar to that of grammaticalization in that they share the same causal factors and mechanisms, and ‘auxiliation theory’ makes use of the same theoretical notions familiar from ‘grammaticalization theory’, such as the idea of ‘grammaticalization chains’, ‘grams’,¹ the concept of a ‘locus for change’, and the importance of ‘pragmatic inferencing’. Auxiliation differs in that it deals only with the way in which ‘complex lexical verb structures develop [ . . . ] over time into auxiliary grammatical structures’ (2), in other words, ignoring all nonverbal lexical structures that have also been shown to grammaticalize. Kuteva has restricted her investigation to auxiliation because she wishes to build on Bernd Heine’s 1993 theoretical work on the grammaticalization of auxiliaries, which concentrated on the source-concepts that form the start of the auxiliation process and on the cognitive forces underlying it. In order to build up a (fuller) ‘theory of auxiliation’, K concentrates on the discourse-pragmatic situation in which a new stage of auxiliation is reached, and she wishes to shed light on the interrelation between the discourse-pragmatic and the more general cognitive-semantic forces in the process. The emphasis, therefore, is on the synchronic situation, on language processing, and on the language users (speakers as well as hearers) involved in the situation. It seems to me, however, that her findings here are also generally applicable to grammaticalization. This latter point, indeed, raises the question of how much auxiliation needs a special theory. As a corollary of this, auxiliation can be said to share a problem noted for grammaticalization theory: Is grammaticalization an independent phenomenon or a conglomeration of changes that occur elsewhere independently (see the objections offered by quite a number of historical linguists in a special issue of Language Sciences 23 (2001))?

The book contains seven chapters followed by a bibliography and a subject and name index. In the first two chapters, K discusses the general framework, the assumptions, and the approach used. She makes a broad distinction between functional/cognitive and generative theory. The essential difference between the two approaches lies in the foundation for investigations, with generativists opting for the ‘reality’ of the grammar and cognitive-functionalists taking language output as their point of departure, with the result that functionalists place much more emphasis on the social and semantic-pragmatic forces involved in communication while generativists are more interested in the structure of the grammar itself. From the functional-cognitive theory of language adopted here follows the notion of gradience, the idea that there are no discrete categories; that is, ‘categorical distinction[s are] blurred’ (13), synchrony and diachrony are intertwined, changes are not abrupt, and layering and variation are the rule. In a way, both theories, in an attempt to be elegant and consistent and perhaps also in an attempt to give themselves a clear profile, overemphasize the formal, in casu quo the conceptual-pragmatic frame that structures language. In my view both form and function are important and heavily intertwined. K, like many grammaticalization theorists, emphasizes conceptual semantic motivations throughout and tends to lose sight of formal considerations. I can sympathize with K’s idea that ‘semantics is the common ground where the conceptual system and the language system meet’ (4), but I think it goes too far to state that ‘conceptual shift from lexical to grammatical content precedes all other shifts’ (11) and that ‘auxiliation as a process is conceptually-semantically motivated rather than arbitrary’ (20). Thus the causes and mechanisms in auxiliation, which K mentions in Ch. 2, are all pragmatic-semantic, such as metaphor and metonymy (27, 35), semantic generalization (32),² and pragmatic inferencing (35). Interestingly, K writes that ‘[a]ssuming that auxiliation is a conceptually motivated process, the student of auxiliation might well expect neat developments of the kind “one lexical source structure—one target grammatical structure”. This, however, is not what we find...