A SOCIOLINGUISTIC SURVEY OF THE MAMBAY LANGUAGE OF CHAD AND CAMEROON by Cameron Hamm SIL International 2002 2 CONTENTS 0ABSTRACT 1 INTRODUCTION 1.1 Fieldwork 1.2 Research Goals 1.3 Procedures 1.3.1 Lexicostatistical Analysis 1.3.2 Recorded Text Test (RTT) 1.3.3 Group Questionnaires 1.3.4 Individual Questionnaires 1.3.5 Itinerary 2 BACKGROUND INFORMATION 2.1 Linguistic Classification and Literature 2.2 Speech Area 2.3 Population 2.4 Village Research Sites 3 LANGUAGE AND DIALECT SITUATION 3.1 Dialect Situation 3.2 Lexical Similarity 4 INTELLIGIBILITY 5 LANGUAGE ATTITUDES AND LITERACY PREFERENCES 6 LANGUAGE USE IN THE RELIGIOUS SERVICES 7 BILINGUALISM AND LANGUAGE VITALITY 8 FINAL CONCLUSIONS AND RECOMMENDATIONS Appendix A. Group Questionnaire Appendix B. Church Leaders’ Questionnaire Appendix C. Individual Questionnaire Appendix D. Self-Evaluation Bilingualism Questionnaire (SEBQ) Appendix E. Word Lists Appendix F. Texts and Comprehension Questions Appendix G. Map of the Area Studied References 3 0 ABSTRACT This paper presents data from a sociolinguistic survey conducted in the Mambay speech community of Cameroon and Chad. The survey was designed to provide language planners with information about this community in order to determine how they can best encourage language development in the region. Mambay is a part of the Mbum group of languages (Niger-Congo, Adamawa) and its closest relation is to Mundang and Tupuri. The data presented includes a general overview of the previous linguistic literature and a description of the geographic area and population, as well as results of interviews, word list comparison, and individual comprehension testing. These results concern comprehension of Mundang, language use and vitality, bilingualism, and attitudes toward both written and oral forms of Mundang. In addition, information by local leaders on the literacy and religious situations is included. Word list results show a very distant relationship between Mambay and Mundang, leaving some question as to genetic affiliation. Additionally, the individual comprehension test results show that Mambay speakers have a very low level of comprehension of Mundang. Acceptance of Mundang is low, with a strong preference for Mambay in most areas. There are no indications of language shift now or in the near future. It is recommended that Mambay should be developed and literacy be encouraged in the region. 1 INTRODUCTION This paper reports on the results of a sociolinguistic survey conducted among the Mambay speech community of Western Chad and Northern Cameroon. The survey was designed to provide language planners with information about this speech community in order to determine how they can best encourage language development in the region. This section includes a description of the fieldwork, the research goals, and the methodology applied during this survey. The subsequent section presents pertinent background information, including a review of the previous literature on this speech variety and a description of the language area and population. The results sections then present the sociolinguistic situation of the area and are followed by a conclusions and recommendations section and appendices, including questionnaires used, maps, and a bibliography. 1.1 Fieldwork The fieldwork for this survey took place between April 26 and May 1, 2000 and was conducted by a research team consisting of: Cameron Hamm and Erik Anonby, members of the Association SIL Tchad, and Calvain Mbernodji, member of the Association Tchadienne pour l’Alphabétisation, la Linguistique et la Traduction de la Bible (ATALTRAB). We extend our sincere appreciation to the Sub-Prefect of Léré, the Sub-Prefect of the Mayo-Louti Department in the Arrondissement de Figuil in Cameroon, the canton chief in Biparé, and the chiefs of Kaboni and Kakala villages for their welcome and for the help they provided in enabling us to successfully carry out this survey. 1.2 Research Goals The goal of this survey was to get a more accurate picture of the linguistic and sociolinguistic situation in the Mambay region. Prior to the survey, we had information suggesting that the Mambay could possibly understand Mundang in order to use literature in this language. Specifically, the objectives of the survey were the following: 1. To determine the degree of lexical similarity between Mambay and Mundang. 2. To investigate and quantify the degree of inherent comprehension of Mundang by Mambay speakers. 4 3. To investigate the acceptability of literature in Mundang and the mother tongue in the Mambay area. 4. To determine the vitality of Mambay by looking at domains of usage and other indications of its continued use in the future. 1.3 Procedures For this survey, we used several methods designed to gather information needed for a global view of the linguistic and sociolinguistic situation of a language community in a relatively short period of time. The data gathering instruments used are briefly explained in the following sections. 1.3.1 Lexicostatistical Analysis A standard word list of 227 terms was taken in Mambay, as well as in the reference speech form of Mundang (see appendix E for a phonetic transcription of the word lists). These word lists were compared to determine the degree of lexical similarity. This was done on the basis of apparent phonetic similarity, and no investigation has been made to see if the words are true historical cognates. SIL has found that knowing the percentage of lexical similarity between languages aids in making judgments in regard to the level of intelligibility between these speech varieties. It has been recommended within SIL that when the percentage of apparent cognates between two speech forms is less than seventy percent (including a certain percentage to allow for a range of error), these speech forms should be considered as two separate languages for the purpose of language development. If, however, the level of lexical similarity between speech forms is seventy percent or greater, dialect intelligibility testing is normally recommended to determine the level of comprehension between the speech forms (Bergman 1989). 1.3.2 Recorded Text Test (RTT) To investigate the degree of comprehension between Mambay and Mundang, Recorded Text Testing (RTT) was used. Texts between two and four minutes in length were recorded in Mundang (at Léré) and Mambay (at Kaboni). Texts that had too many loan words were rejected. Texts that did not have adequate detail, which merely described lifestyle or which did not relate a specific incident in the storyteller’s life, were also rejected. A written French translation of each text was made; noting where phrases ended and pauses could be inserted. Twenty key pieces of information were identified for each text. The French translation and the comprehension questions for each story are listed in appendix F. Both the respondent and the researcher listened to the stories through headphones. The tape was stopped at each pause and the respondent was asked to retell the story. If the retelling included the key piece of information (whether correct or incorrect), the answer was recorded on the score sheet and the next section was played. If the response did not include information related to the question, the respondent was prompted more specifically, and his response was recorded, with a note that a prompt had been necessary. If a section was replayed because the respondent requested it, this was noted on the answer sheet. If a section was replayed because the researcher thought a person or a sound had disturbed the respondent, it was not noted on the answer sheet that the section was replayed. 5 After testing each story on speakers of that particular language the list of twenty key pieces of information was narrowed to ten, based on which ten all hometown subjects were able to readily identify. Speakers of the other language were evaluated, based on their ability to correctly understand these ten pieces of information.1 Subjects listened first to the test recorded in their own language to familiarize themselves with the testing method (and to screen out any subjects with hearing or other difficulties). The subject’s score on the test in their own language is then used to compare with their performance on the other language, e.g., if a subject missed one on their own language and only one on the test in another dialect, the subject will be considered to understand that other language satisfactorily. The numerical results of the RTT can not be interpreted as an exact measure of the degree of comprehension between the speech forms, but rather is an indicator of the approximate level of intelligibility. The percentage at which one can reasonably claim adequate comprehension between speech forms is not always clear. SIL guidelines indicate that seventy-five percent is the lowest acceptable indicator of comprehension, whereas eighty- five percent is usually considered high enough to guarantee good communication. Between these two scores is an area of marginal comprehension, where other sociolinguistic factors must come into play (J. Grimes 1995:22). The standard deviation of the RTT scores is important in evaluating whether the level of intelligibility that has been measured is inherent or acquired. Since inherent intelligibility is a feature of the language itself, it is reasonable to assume that speakers who have mastered their own language will understand the reference speech form equally as well (or poorly) as others who speak their language. Thus, if the intelligibility measured by our test is inherent intelligibility alone, the standard deviation of the scores should be small (i.e., <0.15)2 (J. Grimes 1995:30). A large standard deviation (>0.15) indicates that some subjects scored significantly higher than others did, presumably because their score reflects an acquired comprehension in addition to any inherent intelligibility. 1.3.3 Group Questionnaires Sociolinguistic questionnaires were used to gather information about the sociolinguistic situation, including the geographical extent of the various speech forms, language use, language vitality, and general attitudes towards the various speech forms and the implementation of literacy programs.