The Structure of Dagaare (review)

Dagaare, a language of the Oti-Volta group of Gur (a branch of Niger-Congo), is spoken in Northern Ghana and adjacent Burkina Faso. The book is the most extensive description of Dagaare currently available in English. However, as with other volumes of comparable size in this Stanford series, exhaustiveness or depth of analysis is not a goal. Instead, Bodomo informs about the language in its social context and samples various grammatical processes, sometimes bringing in constructions from related languages for comparison, e.g. Mampruli, Moore. Some processes are treated with much greater detail than others. A cooperative field-method-like approach in selection of issues is indicated by references to unpublished student papers in conjunction with the author.

The book is divided into an introduction, 20 chapters, an appendix with a short vocabulary list, a list of references, and an index which includes grammatical terms, authors, and languages referred to in the text. Interestingly, word order is first discussed immediately after the chapters on phonology (and orthography)—Ch. 6, ‘Word order parameter and Dagaare syntax’ (40–46). This reflects the general strategy of presentation, such that larger units are discussed before smaller constituents. Thus, Ch. 7, ‘Syntax of the Dagaare nominal phrase’ (46–51), precedes the longer Ch. 8, ‘Morphology of the Dagaare noun’ (52–70); like most Gur languages, Dagaare nouns are marked for class. Ch. 11, ‘Structure of the verb phrase’ (80–83), precedes Ch. 12, ‘The preverbal particles’ (84–87), and Ch. 13, ‘The main verb’ (88–93). In such cases, successive chapters often continue discussion of an issue initiated in a preceding chapter.

Not surprisingly for its typological interest, as noted by B (109), much effort is put into the analysis of serial verb constructions (SVCs)—Ch. 17, ‘Complex constructions: SVCs’ (107–21), actually starting with Ch. 16, ‘Double object constructions (103–6). The transition is interesting because of the overlap between double object and SVC constructions in Dagaare and languages of the wider area, in contrast to more distant languages which only have SVCs. Thus, in preparation for the chapter on SVCs, B notes that some verbs like ko ‘give’ and tere ‘offer/pass’ can function either as bitransitive or serial verbs. The severe constraints on double object constructions in Dagaare suggest how they may have been lost in favor of SVCs in other languages. The discussion of SVCs continues into Ch. 18, ‘Complex constructions: SVNs’ (122–27), serial verb nominalizations, and Ch. 19, ‘Complex structures: Coordination’ (128–32).

Inevitably there are some oversights in description. For example, B discusses the verb aspectual suffixes minimally (90). There are two perfective markers. He glosses one as ‘transitive use’ and one as ‘intransitive use’. From this I gathered that when a verb like nyu ‘drink’ is used without an object, it is marked as intransitive. However, he also illustrated both for verbs which seem to be intransitive, e.g. zo ‘run’, where he provided a paradigm with zo-0 ‘perf’ and zo-e ‘perfIntr’. I was unable to understand the distinction between the two markings with intransitive verbs. In much later examples I noticed zo, i.e. zo-0, used in an SVC, e.g. ba zo gaa di la bondiri (they run-perf go-perf eat fact food) ‘They ran there and ate food’ (112). That suggested to me that the terms ‘transitive’ and ‘intransitive’ had been used in an unconventional way in glossing the perfective markers, but it remained unclear just what way that is. Similarly, much later it is revealed that the suffix glossed ‘perf-Intr’ is used in medio-passive constructions, e.g. a noɔ de-ε la (def hen take-perfIntr fact) ‘The sacrifice [the hen] has been accepted’ (141), but when the logical subject is expressed, even post-verbally, the ‘perfIntr’ suffix is not exemplified, e.g. a noɔ de la a bɔɔr (def hen take-perf fact the god) ‘The sacrifice is not accepted by the gods’. In this case, the absence of an example of the ‘perfIntr...