Documenting Vute Ethnobiological Inventories
|SOAS University of London
|Landing Page Handle
Summary of the collection
The main outcome of this project consists of a characterisation of ethnobiological lexical variation using several methodologies. The major outputs include a sketch grammar, thematic dictionary and a dense multimedia corpus of ethnobiological data allowing for systematic analysis of linguistic variables. The project provides a linguistic profile signifying the multilingual and language contact context of Mbenguédje, one of the Wawa villages. It uses social network analysis (Milroy 1980) to describe Wawa social patterns and to help explain the impact of social behaviour on multilingual lexical repertoires. Plants and animals feature prominently in the daily lives of the Wawa and the ethnobiological vocabulary within their linguistic repertoires is extensive, although this of course is not evenly distributed across the population or across languages. The corpus is organized using Microsoft Access, which allows efficient analysis, tagging, querying, linking and summarising data, and ethnobiological lexicons and practices are interlinked with photos, audio and video. Primary data is archived with ELAR and analyzed data is archived as it is prepared, within one year of fieldwork. Analyzed data includes a sketch grammar; translated, annotated and interlinearised audio and video recordings using ELAN; thematic dictionary (Mosel 2011); GPS data, sociolinguistic data using SPSS and social network data using UCInet. The sketch grammar outlines major differences between Mbenguédje and the Oumiari dialect, on which a grammar of Wawa is based (Martin 2011). The thematic dictionary covers the domain of ethnobiology and follows an ethnographic approach (Pawley2001) to include culturally relevant features.
The village lies within the Cameroon-Nigeria borderland, an area of high linguistic diversity (Connell 2007), where multilingualism is the norm and language contact occurs on a daily basis. The village population numbers around 1000, about half of which are Wawa. It is multi-ethnic, with cultural and language boundaries not clearly delineated. Mbenguédje Wawa are highly multilingual, competent in at least three languages, including Vute, Fulfulde, French and Mambila. Other languages comprise the linguistic landscape, including English, Pidgin English, Yamba, Hausa and Arabic.
Ethnobiological terms in Wawa retain prefixes of a former noun class system attested for Mambiloid languages that informs their affiliation with Bantu and Benue-Congo (Blench 1993; Thwing 1987). Martin’s (2011) documentation of the Oumiari dialect revealed that ethnobiology retains these prefixes more than any other domain and the use of these prefixes varies considerably. Documenting a broad range of ethnobiological terms and targeting variation will help to inform which prefixes have been retained and can be related to the noun class study conducted by Thwing (1987) on Vute.
The documentation of Mbenguédje ethnobiological inventories will provide baseline data for the development of further research questions, such as how language change affects ethnobiological knowledge, the loss of which is often ideologised in the literature (Batibo 2005; Harrison 2007). The use of social network analysis will be the first of its kind in the region and will complement other projects in the African context (Lüpke ongoing; Beyer 2010; Beyer and Schreiber 2013; Schreiber 2009). Within this region, language endangerment is a part of a larger topic concerning change and linguistic, cultural and social impacts. Concerns for language endangerment could be extended to multilingualism itself (Di Carlo and Good 2011) and linguistic diversity (Woodbury 2003). Language endangerment in the region is high, with at least seven languages moribund (Connell 1998:4). The multilingual situation of the past is unknown; my research forms an empirical foundation considering the dynamics of multilingualism from which we can better understand social factors and language change in this region. At this point in time, ethnobiological knowledge is highly endangered as the influence of technology and market goods steadily increase, easing the lives of busy agriculturalists. It is important to have a record of current knowledge and to understand the relationship of language change and ethnobiological knowledge. Changing habitation patterns also have linguistic, cultural and social impacts. Inhabitants of this region, who have historically been dispersed (Hurault 1998), are being urged to further centralise. It seems imperative to conduct more extensive sociolinguistic research with attention to multilingualism and the dynamics of contact before we can understand individual language endangerment in this region and accurately assess the rate and scale of loss (Hill 2001).
The project does not solely focus on documenting lexical knowledge. Ethnobiological knowledge goes beyond the lexicon, embedded in stories and cultural practice surrounding plants, animals and Wawa agricultural way of life. Wawa historically followed a 10-day week and used a 10-grain species of grass to track time and agricultural practices. Each day a grain was removed until all grains were removed, then this was repeated twice more, making a month (Griffiths and Robson 2010). Today this practice has fallen out of use, remembered only by older Wawa, while the 10-day week only survives through the practice of holding market every ten days. Additionally, Wawa practice a noteworthy method of botanical pesticides, which is little documented worldwide and is falling out of use as chemical pesticides become readily available. This project documents a variety of stories and practices reflecting these endangered or lost aspects of the Mbenguédje Wawa, including beekeeping, which during previous fieldwork I was told is in decline. I have used audio, video and photos to document these practices and prioritize specialized lexicon as it is often the most endangered (Mosel 2011).
This project builds on my previous research in Mbenguédje and draws data collection methods from Linguistics and Ethnobiology, with a complementary balance of qualitative and quantitative methodologies. I spent ten months in the field overall, with the first few spent learning Wawa better and observing basic social patterns. Wawa inhabit family compounds, commonly comprised of a man, his four wives and their children. With the assistance of male and female research assistants, we documented variation within Wawa and across other languages of a speaker’s repertoire, eliciting in one language at a time (Milroy and Gordon 2003). The compound was the basic unit of analysis, where we investigated lexical variation within compounds and expand to communities of practice (Eckert 2000) such as agricultural, market, religious, and other relevant ties. We conducted focus groups based on these communities of practice to elicit qualitative data on variation and social networks. Questionnaires elicited sociolinguistic and demographic information. I used judgment sampling to ensure stratification across social categories and communities of practice. The sample size depended on the size of social networks and distribution of lexical variation. Audio recordings were made using my Olympus LS-5, which was easily learned by assistants. Video was used in focus groups and to document non-verbal aspects of ethnobiological practices. I collected plant specimens as necessary, to avoid confusion of lexical referents. I continue my previous collaborations with botanists at Cameroon’s National Herbarium and the Royal Botanic Gardens, Kew, where I can deposit plant collections as a permanent record and obtain assistance identifying plants. I used field guides to collect names of animals and take photos for species not represented.
I built on previous data management practices and used Microsoft Access to organise and represent the complexity and multi-dimensionality of the relationships between variation and social categories by interlinking database tables such as demographics, social networks, ethnobiological data, GIS and metadata. I consulted with Kakia Chatsiou in ELAR on data infrastructure. The Access database results have been exported to an archivable format and allows for full accountability (Bird and Simons 2003), linking primary and analysed data. I transcribed at least ten percent of recordings, the rest of which has been annotated in detail and time-aligned for later use as needed. I have used about 20 hours of recordings (Thieberger 2009) and the time spent in the field and annotating the corpus has equalled more than one-third of the entire project. ELAN was used to transcribe, annotate, translate, gloss and code for languages. All character encoding has used Unicode. We recorded descriptive metadata in a format based on ELAR’s spreadsheet template, manipulating it to include important ethnobiological information and maintain administrative data to track workflow and access permissions. All data has been backed up with hard-drives and memory cards throughout the project.
I structured the Access database to facilitate exportation of relevant fields into FLex for creating the ethnobiology dictionary. Fields indicate language(s) and cross-reference lexical relations as it may prove difficult to directly translate across languages (Haviland 2006). A print version targets community members and we conducted focus groups to develop community-driven form and content, with the hope that it caters towards use in the resource-limited school. We maintain an electronic version for revision and further research projects. I am aiming for a minimum of 500 headwords, consistent with the common number of named species (Berlin 1992). A monolingual dictionary would not represent a speaker’s repertoire; therefore the entries include inter- and intra-language variation, targeting multilingualism. The dictionary is grouped by semantic fields of culturally salient subdomains, including partonyms, along with referent images and culturally relevant symbols to represent culturally important categories (habitat, use, etc.) and folk taxonomies (Cablitz 2011). An orthography has not been developed for Wawa, therefore we used broad transcription.
SPSS was used to analyse sociolinguistic data and UCInet was used for social network data. Social network analysis was used as one of the tools to explain the impact of socio-cultural factors. I determined networks and categories of importance through participant observation and informal interviews. The approach revealed how factors such as cohesion, density and position in a network impact lexical variation. The qualitative outputs of this approach are sociograms, which are visual representations of networks, allowing for correlating lexical variation at different layers including individual, compound and community. Quantitative outputs have included network scores based on relational data. The outputs were used to determine important and relevant aspects of the network and to identify the nexus of lexical variation, which were then further investigated qualitatively.
Another analytical tool was the use of a GPS (Global Positioning System), which is essential in documenting ethnobiological knowledge, especially of plants, not only to record the context of the plant, but also as a permanent record of the location at a specific time. No detailed maps exist for this area; therefore it was essential to map the village and outlying compounds. Geography for the most part is an unchanging variable and georeferencing locations of compounds, distribution of kin groups and social networks aided in the visualisation of linguistic (Di Carlo 2013) and social network data. Using GIS (Geographic Information System) analysis in conjunction with a social network approach will contextualise socio-cultural factors and handles large data sets, allowing data layering to efficiently investigate correlations.
As I further refined my research methods, I worked closely with my supervisors and my university’s Research Office to ensure that my research complies with the university’s Research Ethics Policy and the Data Protection Act. I attained approval through my university’s ethics review process before my project began. Throughout my project I maintained openness, honesty, tolerance, fairness and responsibility. As part of my moral obligations to academia, I archived primary data to allow verification of any analyses.
Collection of ethnobiological data must consider ethical and legal implications such as bio-piracy and the commercialisation of traditional knowledge (McClatchey 2012). My collection of ethnobiological information follows Economic Botany Collection Standards (Cook 1995) and the International Society of Ethnobiology’s Code of Ethics. Before arriving in the field I obtained a research permit from Cameroon’s Department of Scientific Research and Innovation through the assistance of a newly established International Research and Training Centre in Cameroon’s capital and operated by the University of California, Los Angeles. I have maintained contact with the head of the village and have verbal consent to continue research. Once in the village, I maintained close communication about my research with the head of the village. I continue to maintain a good working relationship with the community, respecting their goals in regards to my research and maintaining openness about my research goals and content. I carefully handled data and place protections where necessary to avoid misuse of information and respect community rights. Mbenguédje Wawa are secretive about certain aspects of plant knowledge and I discussed and was fully transparent with how information will be used and with whom it will be shared, ensuring that sensitive data will be handled accordingly. Asking information about social networks can be quite personal and the use of sorcery can complicate the methods of collecting network information as well as the elicitation of certain plant uses. I informed each individual that information they give would be anonymous and confidential. The accessibility of the data is determined by community members and will kept accessible after fieldwork. I elicited data from children as they are an important age group that should not be omitted in this project. No extensive studies have been conducted on horizontal linguistic transmission in the African context and especially not with regards to multilingualism, making it imperative in this project to document the lexical variation of children, especially across languages. I obtained prior consent from the child’s parent or guardian and fully discuss how data will be used. Before making any audio recordings, I clearly explained how the person’s data would be used and that data may be used in my thesis and other publications. I explained how the data will be stored in an archive in Yaoundé and with ELAR, including access issues and explaining ELAR’s open access policy. I obtained prior informed consent and obtain permission for any photos or video documentation. All consultants and research assistants were fairly compensated accordingly with a scale appropriate for the local economy. I ensured that all contributors are fully acknowledged. Throughout the project I maintained ongoing consent.
Overviews of language endangerment in West and Central Africa exist (Blench 2007; Connell 2007; Dimmendaal & Voeltz 2007), yet few studies examine the issue comprehensively and consider the multilingual, intense contact contexts. Both variation and multilingualism are often viewed as inconveniences (Milroy and Gordon 2003). This project sheds light on the patterns and social functions of variation and multilingualism, contributing to the betterment of methods and discourses for assessing endangered language situations in the African context where endangerment is not well understood (Lüpke and Storch 2013; Lüpke 2009). At this point it remains an empirical question as to what is actually occurring within the linguistic ecologies of small African communities that have constantly been in flux and it is unclear whether a lingua franca existed pre-colonialism or whether multilingualism was the norm for communication (Di Carlo and Good 2011). Documenting the use of the current lingua franca, Fulfulde, will be of interest and will give insight into the contact between pastoral and agriculturalist ethnobiological knowledge. Analyses of social networks and communities of practice helps clarify the function of the Wawa language and will give insight into the relationship of network strength and language maintenance (Milroy and Gordon 2003).
Researching these topics with the social network approach is pioneering within the field of Linguistics. Few African studies use social network analysis (with the notable exception of Beyer 2010; Beyer and Schreiber 2013; Schreiber 2009) and the integration of GIS is just now coming into practice (Di Carlo & Pizziolo 2013; Veselinova & Booza 2009). This project is part of a movement away from the monolingual perspective, on which most lexical variation research is based. Most social network approaches concern western, urban and monolingual communities, with a few bilingual studies. The few ethnobiological studies concerning bilingualism involve a former colonizer language (Schaeffer et al 1999; Ohmagari and Berkes; 1999; Lizzaralde 2001; Müller-Schwarze 2006). The domain of ethnobiology typically involves a significant amount of variation, which allows for a broad typology of Wawa lexical variation. The documentation of lexical variation quantifies the language ratio in a speaker’s repertoire and reveals the dominant language(s) within the domain of ethnobiology. Documentation of linguistic repertoires will inform language management and pedagogy. The project will help to inform our knowledge on how social networks and other social factors play into multilingual and language contact situations and will provide a visualisation of the patterning of multilingualism within and between communities of practice. The multilingual focus complements other multilingual studies of the region (Di Carlo 2012; Connell 2009). Lastly, this project acts as a foundation for development of further research questions such as how the loss of highly specific ethnobiological terms affects conceptual, sociocultural and practical knowledge.
An extensive analysis of Wawa multilingualism has not been conducted. Other than my ethnobotanical research, the only other research on the Mbenguédje dialect is a word list collected by Connell (unpublished). Two students recently conducted doctoral research on the Oumiari dialect and one produced a grammar based on that dialect (Martin 2011). Other literature on Wawa includes a short description (Guarisma 1987), a sociolinguistic survey (Starr 1989), anthropological research (Gaussett 1999), linguistic research (Connell 1996, 1998, 2000, 2001) and an article on Wawa language ecology (Griffiths and Robson 2010).
During previous fieldwork, I assessed the endangerment of the Mbenguédje dialect based on observation and interviews covering a range of sociolinguistic factors. I characterised Wawa as seriously endangered (Edwards 1992), although further detailed, domain-specific analysis is necessary.
In order to understand the linguistic landscape of this region, socio-historical factors need to be considered. The village is situated in an area that has experienced social upheaval prompted by the arrival of the Fulbe 150-200 years ago (Gausset 1998). They enslaved some groups and caused others previously dispersed to form alliances against raids (Hurault 1998). The Fulbe were the first in the region to adopt Islam and their political dominance has facilitated the spread of the religion (Njeuma 1990).
It became clear during my previous research that Wawa cannot be studied in isolation. As I elicited plant names, I was often given names in other languages, revealing the complexity of inter- and intra-language variation. Several people attributed their extensive plant knowledge to their social habits, such as where and with whom they spend time. It will be essential for the aims of this project to consider the language ecology of Mbenguédje, as well as extralinguistic and sociolinguistic factors. My master’s degree in ethnobotany and personal interests, coursework and teaching in biology, along with a BA and varied postgraduate coursework in Linguistics prepared me with the knowledge to carry out such a multidisciplinary project.
Blench, R. M. (1993). An outline classification of the Mambiloid languages. Journal of West African Languages. 105-118.
Connell, B. (2010). Language Ecology and Language Endangerment: An Instance from the Nigeria-Cameroon Borderland. Journal of West African Languages. 1-11.
Connell, B. (2009). Language Diversity and Language Choice: A View form a Cameroon Market. Anthropological Linguistics. 130-150.
Connell, B. (2007). Endangered Languages in Central Africa. In M. Brenzinger, Language Diversity Endangered. 163-178. Berlin: Mouton de Gruyter.
Connell, B. (2001). An Introduction to the Mambiloid Languages. In N. M. Mutaka, & S. B. Chumbow, Research Mate in African Linguistics: Focus on Cameroon: A Fieldworker’s Tool for Deciphering the Stories Cameroonian Langaguaes Have to Tell. 79-92. Cologne: Rüdiger Köppe Verlag.
Connell, B. (2000). The Integrity of Mambiloid. Proceedings of the Second World Congress of African Linguistics.
197-213. Leipzig, Cologne: Rüdiger Köppe Verlag.
Connell, B. (1998). Moribund Languages of the Nigeria-Cameroon Borderland. In M. Brenzinger, Endangered Languages in Africa. 207-225. Cologne: Rüdiger Köppe Verlag.
Connell, B. (1996). The Roots of Mambila: Convergence and Divergence in the Development of Mambila. 26th Colloquium on African Languages and Linguistics. Leiden.
Gausset, Q. (1999). Islam or Christianity? The Choices of the Wawa and the Kwanja of Cameroon. Africa. 257-78.
Griffiths, S. S., & Robson, L. (2010). Cultural Ecologies of Endangered Languages: The Case of Wawa and Njanga. Anthropological Linguistics. 217-238.
Guarisma, G. (1987). Dialectométrie Lexicale de Quelques parlers Bantôides Non Bantous du Cameroun. In G. Guarisma, & W. J. Möhlig, La Methode Dialectometrique Appliquee aux Langues Africaines. 281-299. Berlin: Verlag von Dietrich Reimer.
Komáromi, R. (2009). Sharing Knowledge: Intra-cultural Variation of Ethnobotanical Knowledge and the Factors that Pattern it in a Mambila Community in the Cameroon-Nigeria Borderland. Unpublished Master’s Dissertation, University of Kent.
Martin, M. (2011). A Grammar of Wawa: An Endangered Language of Cameroon. Unpublished PhD thesis. University of Kent, Canterbury, UK.
Njeuma, M. Z. (1990). The Lamidates of Northern Cameroon. In M. Z. Njeuma, Introduction to the History of Cameroon: Nineteenth and Twentieth Centuries. London: Macmillan.
Starr, A. (1989). Sociolinguistic Survey of Wawa A Mambiloid Language of Cameroon. SIL Publication. Yaoundé, Cameroon.
Todd, H. (2011). Language Decline and Ethnobotanical Knowledge in a Multilingual Setting in Cameroon. Unpublished master’s dissertation. University of Kent, Canterbury, UK.
Acknowledgement and citation
To refer to any data from the collection, please cite as follows:
Todd, Heather. 2014. Documenting Vute Ethnobiological Inventories. Endangered Languages Archive. Handle: http://hdl.handle.net/2196/00-0000-0000-000E-D157-4. Accessed on [insert date here].