- Functional categories
The distinction between lexical and functional categories (FCs) is well established in linguistic theorizing and is couched in different theoretical frameworks, although under different labels (content words, referential categories, and so on versus function words, grammatical categories, empty categories, and the like). It gained particular importance in generative grammar when parameterized principles of universal grammar (UG) were argued to relate exclusively to functional categories, more specifically to the featural composition of functional heads (Borer 1984, Chomsky 1991). Under this assumption, FCs define the locus of parametric variation and play a crucial role in all types of grammatical variation and development, including crosslinguistic variation, language acquisition, and diachronic change. The particular importance of FCs for numerous [End Page 975] areas of linguistic research stands in stark contrast to the fact that there exists no broadly accepted consensus on their theoretical status. In fact, even the questions of which elements may justly be classified as functional, whether an identical set is instantiated in all human grammars, or whether a specific FC will consistently occupy the same position in clause structures across grammars, as well as related issues, are a matter of controversy. Interestingly, these disagreements reflect not merely theoretical dividing lines, as was to be expected, but they also persist among researchers following the same or similar theoretical approaches. The debate on the set of FCs specified by UG, for instance, illustrates the controversy: whereas Chomsky (2000) and others aim to drastically reduce their number in the minimalist framework, the ‘cartographic’ school (e.g. Cinque 2006) advocates finely grained structures that contain a considerable number of functional heads. A study of the status of FCs and their manifestations in typologically and genetically distinct languages, in language contact, and in historical change is therefore a desideratum in linguistic research.
This is what Pieter Muysken sets out to do, and I cannot think of anyone more appropriate to tackle the challenges that such a task presents. On the one hand, he is highly respected by linguists who take different theoretical approaches (both functional and formal); on the other hand, he has made significant contributions to grammatical theory and more applied areas of linguistic research, most notably to the study of language contact phenomena. M has also carried out extensive empirical work on a variety of languages, including non-Indo-European ones like Quechua. He has precisely the profile required if one attempts to ‘provide an analytic survey of this topic’ from an ‘integrated cross-linguistic perspective’ (2).
The volume comprises eighteen chapters, organized in four parts, preceded by an introduction and followed by conclusions. In the ‘Introduction’ (1–10), M starts with an intuitive definition of FCs. He argues that FCs do not represent a homogeneous group of categories and also mentions an ‘intermediate status’ between functional and lexical elements. This leads him to favor a gradient treatment of the functional/lexical distinction and to advocate a multidimensional approach. The concept of FCs is claimed to ‘lie at the interface of different representations—morpholexical, syntactic, phonological, and semantic’ (5), with its ‘roots’ in morphosyntax (7) since they account for the structural skeleton of clauses. Surprisingly, M refrains from giving an explicit definition of FCs and does not state explicitly which properties an element must have in order to be classified as an FC.
Part 1,‘Grammar’, consists of four chapters that review treatments of FCs in various theoretical approaches to language and grammar, mainly in search of defining criteria for FCs. The survey begins in Ch. 2, ‘Functional categories and language typology’ (13–25), with a short presentation of the views held by Edward Sapir and Roman Jakobson, focusing subsequently on linguistic typology, before it finally assesses the contribution of ‘Theoretical syntax: The generative tradition’ in Ch. 5 (53–70). This discussion produces several sets of defining criteria and a growing list of FCs. The resulting picture becomes more complex when ‘Lexical, morphological, and phonological dimensions of functional categories’ are taken into account in Ch. 3 (26–41) and when aspects of ‘Semantics and pragmatics...