Argument Realization (review)

This volume contains articles unified both by subject (surface realization of arguments in the clause) and framework (lexical-functional grammar, LFG). Within LFG, the focus is foremost on lexical mapping theory (LMT), which pays particular attention to valency alternations, describing them in terms of mapping from arguments to grammatical functions, mediated by the level of abstract features.

The book opens with an editors’ ‘Introduction’ (1–13) which offers a good survey of the recent developments in LFG and the contribution of LMT, placing this framework in a general perspective with some other modern syntactic theories.

The papers fall into three thematic sections. Section 1, ‘Morphology vs. syntax’, opens with an important article by Kersti Börjars and Nigel Vincent, ‘Multiple case and the “wimpiness” of morphology’ (15–40). The authors’ main claim is that the morphology is an active determinant. Their argumentation rests upon evidence from languages that exhibit multiple case marking, as in Kayardild (Australia), where the noun phrase this woman in with this woman’s good net receives both genitive and instrumental marking: maku-karra-nguni ‘woman-gen-instr’. Alongside such clear instances of ‘case stacking’, the authors see the traces of multiple case in overriding (‘case attraction’), in particular, in Classical Armenian: The dependent noun phrase borrows the case from the element on which it depends (by the king’s(gen) wife(abl) → by the king’s(abl) wife(abl)). The authors have overlooked an important theoretical issue, however: how to prove that case attraction results indeed from a multiple case marking, not from the secondary replacement of the initial case relation in the surface syntactic structure (gen→abl). The authors propose a useful classification of types of multiple case marking in terms of several parameters (agglutinative/fusional type of language, stacking/overriding, independent/agreement case). However, their (intuition-based?) claims regarding the impossibility of some combinations seem unfounded. In particular, case stacking in a fusional language (considered impossible by the authors) might be exemplified by possessive pronouns. At least in some languages, on the one hand, they are marked for genitive, and, on the other, take agreement case (cf. Russian ixnij ‘their’, prohibited in the literary language but common in colloquial ‘low’ speech; ix is the genitive form of oni ‘they’, which ‘illegally’ takes the adjectival agreement case: ixnij ‘they.gen+nom.sg.m’, etc.).

Rachel Nordlinger’s article (41–71) also deals with multiple case marking. Nordlinger offers an interesting LFG-based theoretical framework, the constructive case model. She argues that in many Australian languages case morphology not only marks the grammatical function of a nominal (subject, object, etc.) but also carries information about its larger syntactic context (tense/aspect/mood). Louisa Sadler discusses the properties and morphosyntax of noun phrases in Welsh (73–109).

Section 2 is dedicated to a favorite topic of LFG, complex predicates. George A. Broadwell (‘Choctaw directionals and the syntax of complex predication’, 111–33) claims that in some languages directionals (forms expressing the orientation of motion) are to be treated as forming complex predicates with several verbs. This leads the author to an important lexicological issue: Which verbs can be combined with directionals and thus can be conceived as oriented? Broadwell argues that these are verbs that contain the predicate go in their lexical structure. In spite of appealing to the Whorfian theory of linguistic relativity, this solution appears to be ad hoc for verbs of perception or emotion, which thus should be decomposed, for instance, as ‘gaze goes from X to Y’ (for to see) or ‘thought goes from X to Y’ (for to be said about).

Yo Matsumoto (135–69) discusses the Japanese morphological causatives in -(s)ase. He argues that, alongside causatives that are biclausal at both the level of argument structure and functional structure (permissives and persuasive causatives) or at least at a-structure (coercive causatives: make someone do . . .), there are causatives...