Discovering Dyula: the reach of a lingua franca in Burkina Faso

ABSTRACT

National identities must interact with the linguistic identities of groups within state borders. Individuals are typically categorized by their first-language capabilities, and theories about ethnic fragmentation and sub-national conflict are largely based on these categorizations. Though we are aware that most people in Africa are multilingual, this has not shifted our theorizing. If people are indeed multilingual, particularly in a lingua franca, then their economic and political behavior may be different from what we expect to emerge from their first-language identities. This paper seeks to push theorizing to include the linking role of lingua francas. It looks at a particular language, Dyula, and shows how it spread historically in West Africa and contemporarily within Burkina Faso. Using original survey research, it shows the geographic regions in which it has spread and the individuals who adopt it. This can help to predict its future role, and that of other lingua francas across the continent of Africa, suggesting an alternate end-point to the bifurcated debate about national identities that advocates either monolingual or linguistically demarcated states.

KEYWORDS:

• Language
• lingua franca
• Dyula
• Burkina Faso
• West Africa
• gold mining

Acknowledgments

The author would like to thank Bowdoin College for funding this field research. The deepest thanks go to research assistants in Burkina Faso: Eve Diarra, Philippe Sinoe, Francis Somé and Samuel Kambou.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes on contributor

Ericka A. Albaugh is Associate Professor of Government at Bowdoin College and teaches on Africa, ethnic conflict, development, state-building and language politics at Bowdoin. She has conducted field research in Cameroon, Senegal, Ghana, Cote d’Ivoire and Burkina Faso and has written articles on language politics, education, and elections in Africa that have appeared in International Studies Quarterly, Journal of Modern African Studies, and Democratization. Her first book is entitled State-Building and Multilingual Education in Africa (Cambridge University Press, 2014), and she recently co-edited an interdisciplinary volume, Tracing Language Movement in Africa (Oxford University Press, 2018).

Notes

1 This language is alternatively spelled Dioula, Jula, and Juula. It is different from Diola in Senegal.

2 I do not argue that language always corresponds to an ethnic category. Indeed, language exhibits more potential for change within an individual’s lifetime than an ethnic category, which is the point of this article. But as one of the most common markers of ethnic identity (Fishman, 1997), language is often conflated with ethnicity, and scholars studying one invariably overlap with those studying the other.

3 Mande is a major branch of the Niger Congo family of languages encompassing a wide variety of related but not necessarily mutually intelligible languages in West Africa. Present Dyula is part of the Manding variety within the Mande branch, which includes mutually intelligible Bambara (Mali), Maninka (Guinea) and Dyula/Jula of Cote d’Ivoire and Burkina Faso (https://www.ethnologue.com/subgroups/bamana). See publications by Valentin Vydrin (http://llacan.vjf.cnrs.fr/p_vydrin.php#publis) and Coleman Donaldson (https://www.ankataa.com/resources/) for more detail.

4 Şaul (1998) identifies the Watara war-houses as sonanki, while Person (1976) refers to ‘Sonhondyi’ princes.

5 The lack of control may have served the warrior needs as well. As Wilks (1976, p. 414) notes of the Gonja: ‘Centralizing states did not try to incorporate all of the surrounding areas. Unincorporated, they constituted pools of manpower from which the demands of raiding parties in search of slaves could be met.’

6 Army numbers from Person, 1971, p. 122; figures for population under his control in 1887–1888 from Person, 1968, pp. 2089–2090.

7 He noted the Dyula were present in urban centers, but he admitted that of all the languages in the AOF, Dyula was the only one he was not able to precisely situate on a map (de la Vergne de Tressan, 1953, p. 184).

8 Some cantons were split between Dyula and other languages, so these are rough estimates.

9 At this time, Tiendrebeogo and Yago (1983, p. 25) claimed that Dyula was spread across the entire Southwest as a vehicular language. The surveyors, however, only went to five towns in Hauts Bassins and three in Sud-Ouest. Of the 48 concessions (extended households) surveyed in the Sud-Ouest, 42 were in Diebougou town, four were in a Dioula village outside of Diebougou, and only two were in a relatively rural area in Gaoua sub-prefecture, near Kampti. Therefore, there was no real assessment of the wider region of the southwest, and it is doubtful that Dyula was as widespread as they claimed.

10 Bilingual education might have helped to spread Dyula in these regions as well, as the language has been used sporadically as a medium of instruction in some schools. It is currently used in 38 schools, all in the Boucle de Mouhoun, Cascades, Centre and Hauts Bassins regions, though not in the Poni and Noumbiel regions comprising the ‘pocket’ (Bagre, 2016).

11 The Ethnologue is aware of the critique that it tends to separate related language groups, and its database now names ‘macro-languages’ that encompass sub-groups and variants. SIL has also become increasingly pragmatic about using second language literature, seeing languages of wider communication not as threats, but as potential alternative opportunities for literacy. It has recently begun an effort to better represent the second-language capabilities of speakers (See Olson & Lewis, 2018).

12 From Berthelette et al 1995. In each village, males and females from three different age groups were selected. Individuals could score from 0 to 45, but in practice, the upper bound was about 30, with self-identified native Dyula speakers scoring an average of 31 (Showalter, 2008, pp. 7–12).

13 I first went to 21 villages and one mining site with research assistants, and then had the trained assistants conduct shorter questionnaires in 9 additional villages and three more mining sites.

14 Diosso had 19.15 SRT score in the 1990s and 93% proficiency in Dyula in 2016; Kovio had 17.34 and 83%; Djefoula had 15.71 and 99%.

15 It is used in 15 schools: 12 in Ioba, 1 in Bougouriba and 2 in Poni. Neither Lobiri nor Birifor have been used (Bagre, 2016).