This volume presents a series of corpus-based studies analyzing the various grammatical devices used to express discourse functions and to structure narratives in some ten African languages. Thanks to the editors’ efforts in setting up comparative concepts, the scope of this book doubtless goes far beyond African linguistics and should interest descriptive linguists and typologists, as well as specialists of discourse studies. The book opens with an introductory and synthetic chapter by the editors. The nine following articles, each devoted to studies on particular languages, are then ordered along language families (or phyla) of Africa.
In their substantial introductory article ‘Discourse structuring and typology: How strong is the link with aspect?’ (1–22), Shahar Shirtz and Doris Payne very clearly define the scientific context and the general approach used here. As a common ground, the authors propose to rely on Labov and Waletzky’s (1967) definitions of (a) ‘narrative’ as a sequence of nonoverlapping (thus bounded and perfective) events, and (b) the main event line (henceforth MEL, identified here with ‘foreground’) of a narrative as including propositions expressed in an isomorphic (or iconic) order to the story events. Based on previous studies and the studies in this book, Shirtz and Payne have identified seven major types of verbs or constructions used crosslinguistically to communicate foreground or MEL, in opposition to devices for background or non-MEL: (1) verbal constructions coding past-perfect(ive) or at least bounded or completive tense-aspect semantics, (2) verbal constructions coding ‘situational dependency’ (Robert 2010) of that proposition on some frame of reference, (3) syntactically independent clauses, (4) syntactically dependent clause chaining, (5) clause conjunction, (6) Austronesian-type voice, and (7) word order. Elaborating on these various points, they have discovered that there is not always a correlation between the MEL and grammatical aspect: the putative correlation between past-perfective tense-aspect forms and expression of the narrative MEL is not universally valid. After summarizing the main contributions of the various articles, Shirtz and Payne briefly present three specific models, all grounded in cognitive linguistics, which have been fruitfully used by several authors in this book for analyzing discourse structuring: those of Fauconnier (1994 ), Dinsmore (1991), and Botne and Kerschner (2008). They rightfully conclude with a broader invitation for typology and cognitive sciences to collaborate on investigating the grammar-discourse interface in a crosslinguistic perspective.
The Nilo-Saharan family is illustrated first by Doris Payne’s ‘Aspect and thematic clause combining in Maa (Nilotic)’ (23–52). This article is remarkable both in the thorough analyses conducted on a corpus of narratives and in the use made of Fauconnier’s (1994 ) model for this analysis of discourse structuring. Exploring whether this language has a dedicated morphosyntax for coding the temporally sequenced MEL in discourse, Payne first demonstrates that, though it can occur on sequential and semantically perfective main events, the uses of the so-called narrative [End Page 243] form (commonly attested in African languages) are not limited to marking narrative events, and she argues for a new characterization as marking ‘high thematic continuity between equally-ranked propositions’ (23). As for the form previously treated as a perfective aspect or a past tense, its various uses for indicating anteriority or relevance for the present (or current) space lead her to recategorize it convincingly as a perfect. Payne shows that this verb form is not used to demarcate foregrounded MEL material from nonsequential background but rather to initiate a mental space that elaborates or justifies the claim about a previously reported fact.
The next article, by Helen Eaton, ‘Main event line structure and aspect in Sandawe narratives’ (53–80), focuses on an isolated language spoken in Tanzania. Eaton argues that Sandawe does not have a specific verb form that fulfills the function of advancing the MEL in a narrative, nor does it have a set of dedicated aspect morphemes. However, the language makes use of an aspect distinction in...