The exotic syntax of spoken discourse: Challenges from spontaneous narration in Wan (Mande)

Tatiana Nikitina

Monday, January 28, 2019
10:30 a.m.–Noon

Humanities Center Conference Room D


Recent decades have seen an upsurge of interest in assessing applicability of generative syntactic frameworks to data from newly described languages. Fueled by the rapid development of tools that assist fieldworkers in building annotated corpora of spontaneous data, that interest has led to major advances in the description and analysis of unwritten languages, as well as prepared the ground for reevaluating some of the theoretical assumptions of traditional derivational frameworks. Newly available corpora of spontaneous discourse abound in structures that are difficult to account for using the syntactician’s readily available formal tools. When attested in languages with a literary tradition, some of these structures are easily dismissed as performance errors, yet speakers of languages with a rich oral tradition sometimes insist on treating them as an essential component of “the proper way stories are told” (Nikitina 2018a).

In this talk I illustrate some of the challenges of analyzing the syntax of spontaneous discourse in an unwritten language based on a corpus of traditional narratives in Wan (Mande, Ivory Coast). I describe several constructions that are very prominent in the corpus yet are not attested in elicitation, and which do not yield easily to a standard clause-level analysis. Some of these structures involve repetition and insertion of complex constituents that from a Eurocentric perspective look like disfluencies, yet are systematic and are treated as felicitous by speakers. Others encompass units of discourse larger than a sentence, suggesting the need for developing formal tools for treating phenomena at a parasyntactic level (Halford 1990, inter alia).

I focus on one particular feature that is characteristic of Wan spoken discourse: the use of a typologically unusualbidirectional case marker (Heath 2007, Nikitina 2018b). As in a number of other West African languages, a special marker is used in Wan to separate subjects from objects in clauses where they would have otherwise been adjacent. In terms of alignment typology, it serves to indicate unambiguously that the preceding constituent is a subject and the following one is an object; yet it is not associated with either of the constituents and cannot appear when either one of them is missing. Being a fundamentally linear order phenomenon, bidirectional case marking is difficult to account for within derivational syntactic frameworks (for example, it is not treated properly by accounts relying on restrictions on case assignment, such as Koopman 1992).

In conclusion I discuss some of the methodological and theoretical issues raised by the discrepancy between our largely graphocentric expectations and the reality of spoken discourse. In addition to providing a perfect testing site for formal syntactic frameworks, the “exotic” syntax of natural discourse opens new avenues for exploring how syntactic structures are built and interpreted, how clause-level syntax interacts with the structure of larger-than-sentence units, and how syntax is affected by cultural practices and technological developments.


Halford, B. K. 1990. The complexity of oral syntax. In H. Pilch & B. K. Halford (eds.) Syntax gesprochener Sprachen, pp. 33-88. Tuebingen: Narr.

Heath, J. 2007. Bidirectional case-marking and linear adjacency. NLLT 25: 83-101.

Koopman, H. 1992. On the absence of Case chains in Bambara. NLLT 10: 555-594.

Nikitina, T. 2018a. When linguists and speakers do not agree: The endangered grammar of verbal art in West Africa.Journal of Linguistic Anthropology 28(2): 197-220.

Nikitina, T. 2018b. Focus marking and differential argument marking: The emergence of bidirectional case marking in Wan. E. Adamou, K. Haude & M. Vanhove (eds.) Information Structure in Lesser-Described Languages, pp. 195-216. Amsterdam: John Benjamins.