Language Documentation and Description ISSN 1740-6234 ___________________________________________ This article appears in: Language Documentation and Description, vol 6. Editor: Peter K. Austin Managing linguistic diversity in the church ANICKA FAST Cite this article: Anicka Fast (2009). Managing linguistic diversity in the church. In Peter K. Austin (ed.) Language Documentation and Description, vol 6. London: SOAS. pp. 161-212 Link to this article: http://www.elpublishing.org/PID/075 This electronic version first published: July 2014 __________________________________________________ This article is published under a Creative Commons License CC-BY-NC (Attribution-NonCommercial). The licence permits users to use, reproduce, disseminate or display the article provided that the author is attributed as the original creator and that the reuse is restricted to non-commercial purposes i.e. research or educational use. See http://creativecommons.org/licenses/by-nc/4.0/ ______________________________________________________ EL Publishing For more EL Publishing articles and services: Website: Terms of use: Submissions: http://www.elpublishing.org http://www.elpublishing.org/terms http://www.elpublishing.org/submissions Managing linguistic diversity in the church: language ideological contestation within a shared moral framework in south-western Burkina Faso Anicka Fast Donc, il faut quand même comprendre que quand on élabore les projets de traduction, qu'il faut les limiter à un certain domaine. Je suis pour la traduction. Mais, que cette traduction soit accompagnée aussi d'une possibilité pour les gens d'avoir une langue de communication plus élargie. So, one needs to understand that when translation projects are introduced, they need to be limited to a particular domain. I’m in favour of translation. But this translation must also be accompanied by the possibility for people to have a language of wider communication. -- ÉÉMBF pastor Donc il y a tous ces cas de figure par rapport aux langues maternelles... Cela peut être une promotion de la langue, susciter l’intérêt, mais cela peut être aussi, “Bon, ça c’est vous qui trouvez que c’est important, mais nous-mêmes, on se contenterait soit du français, du dioula, ou du mooré... So there are all these scenarios with respect to mother tongues... It can be a promotion of the language, exciting interest, but it can also be, “Well, you’re the ones who think that’s important, but as for us, we’d be content with French, Jula or Mooré...” -- ÉÉMBF pastor **** A culture that for the first time possessed a dictionary and a grammar was a culture endowed for renewal and empowerment, whether or not it adopted Christianity. -- Sanneh 2003: 99 Giving people pride in their mother tongue is God's vision for them.... We can’t relate to the poor and oppressed without a mother-tongue connection. -- AIMM missionary Anicka Fast 2009. Managing linguistic diversity in the church. In Peter K. Austin (ed.) Language Documentation and Description, Vol 6, 161-212 London: SOAS. 162 Anicka Fast **** Est-ce le moment de nous rappeler de la valeur énorme pour un groupe ethnique d’avoir la Parole de Dieu dans leur propre langue? ... [ou] Est-ce le moment de nous rappeler que le Saint Esprit est pleinement capable de toucher le cœur d’un être humain par un témoignage au nom du Seigneur dans une langue qu’il comprend même si ce n’est pas la sienne? Is this the moment to remind ourselves of the enormous value for an ethnic group of having the Word of God in their own language? ... [or] Is this the moment to remind ourselves that the Holy Spirit is fully capable of touching the heart of human beings by a testimony in the name of the Lord in a language that they understand, even if it is not theirs? -- Bertsche 2006: 2 1. Introduction Having grown up in a missionary community where Bible translation was often seen as an unquestioned good and a source of empowerment for local communities, I was intrigued by reports that in the Mennonite1 churches of Burkina Faso, some aspects of the historical missionary focus on translation into vernacular languages were being overtly contested by church leaders. Working primarily in the province of Kénédougou in Burkina Faso, Mennonite missionaries have tended to focus on Bible translation and literacy in local languages, while the national Mennonite church leadership has generally focused on attempting to expand use of the regional lingua franca, Jula. With increased power-sharing and mutual accountability between these two groups, different ideas about the relative importance of Jula and the various local vernaculars have come into conflict. While critique of the missionary emphasis on the vernacular is not new (e.g., Adejunmobi 2004, Meeuwis 1999), this paper constitutes the first detailed examination to my knowledge of an open discussion of this question between ‘brothers and 1 The Mennonites are a denomination of Christians who split off from the Protestants during the Reformation, inspired by the teachings of Menno Simons (1496-1561). They are historically characterised by pacifism, non-violence, and a commitment to adult baptism. A 2006 census of Mennonite church membership shows 1.5 million church members world-wide, with the largest proportion (17.2%) living in Africa (Burkhardt 2006). Managing linguistic diversity in the church 163 sisters in Christ’ who continue to work within a very similar moral framework. In this paper, I investigate perceptions of language utility held by church leaders, expatriate missionaries, and church members in Burkina Faso, in order to shed light on the complex use of language ideologies in a situation of ongoing intra-church conflict. I also relate the language attitudes expressed both directly and indirectly by research participants to other discourses and ideologies of endangerment and authenticity circulating in both academic and ecclesiastical circles. By moving beyond the simple description of language attitudes to an explanation of their relation to structures of power and interest, I hope to exemplify a “demythologise[ed] sociolinguistics” which sees the use of language ideology and the expression of language attitudes as “a social practice in its own right” (Cameron 1997:64). I also intentionally situate myself as a ‘critically engaged’ researcher (Speed 2006:67) who shares a moral framework with research participants (Dobrin 2005, Fast 2007a). Methodologically, this research builds on a growing trend to tailor methods of attitude study to largely rural and non-literate contexts with nonindividualist modes of interaction (cf. Showalter 2001, Robinson 1996). I introduce or adapt various methods of language attitude study to allow both for comparison between literate and non-literate participants, and for the collection of meaningful data even in the absence of a shared language between myself and participants. I suggest that the discourses used by missionaries and church leaders reflect differing language ideologies that lead to competing definitions of church. The doctrine of translatability, which has historically provided strong theological impetus for Bible translation into vernacular languages, is tightly bound up for many Western missionaries with an essentialising ideology that connects the vernacular with deep identity and spiritual authenticity, thus functioning to contain diversity within the church through the idealisation of ethnically homogenous, monolingual congregations. In the Burkina Faso context, I suggest that this ideology functions to obscure participants’ unequal access to resources for ideological legitimisation by re-casting missionaries as agents of cultural revitalisation and dismissing alternative conceptions of church as insufficiently indigenous. Working from ‘within’ seems essential both to uncover the potency of this ideology and to suggest moral resources for reconciliation and power balancing, and I therefore conclude by proposing some theological perspectives that, from a shared Christian perspective, can lead to a greater balance in power. Despite being primarily committed to a moral framework shared with research participants, I strongly believe that an examination of this kind of ideological conflict is also relevant to all linguists. Dobrin (2009:1) has perceptively noted that documentary linguists do not seem to be very good at 164 Anicka Fast absorbing the implications of “language program failure”, since linguists’ “activist discourse clearly prepares us to respect and support certain choices more than others” (ibid.:9). The results of this study suggest that linguists would do well to examine both the kinds of ideological resources that they may be using to justify new forms of intervention in endangered language communities, and their tendency to underwrite the kind of ideologies on which missionaries draw in this conflict. 2. Background In this section, I briefly describe the sociolinguistic situation in south-western Burkina Faso, and outline the history of relations between Mennonite missionaries and church leaders in this part of the country. 2.1 The sociolinguistic situation of south-western Burkina Faso The southwest region of Burkina Faso is not only the most linguistically diverse in the country, as Figure 1 illustrates, but is also characterised by a high degree of individual multilingualism (Tiendrebeogo and Yago 1983:25). In addition to the languages from the Kru, Mande and Gur families that are spoken in Kénédougou province, the Mande language Jula, which spread to Burkina Faso as early as 1200 C.E. via Muslim merchants of the nearby Manding empire (Mansour 1993: 36), plays an important role as lingua franca for an estimated 1.2 million people in the region surrounding Bobo-Dioulasso (Showalter 2006:3). Research by Showalter suggests that current patterns of multilingualism in this region show much continuity with the pre-colonial situation (Showalter 2001:10). His detailed study of Jula proficiency among speakers of 36 indigenous languages spoken mostly in this region demonstrates that, while in the average community Jula competence is not as high as competence in the mother tongue, many communities are characterised by global good competence with pockets of excellent competence (Showalter 2004:21). However, he shows that levels of competence in Jula are not directly related to the degree of endangerment of indigenous languages (Showalter 2004:17). Despite high levels of multilingualism in western Burkina Faso, the attitude of the population seems to include both a desire to maintain their languages and a generally low level of national loyalty (Showalter 2001:25), such that the situation can best be described as one of relatively stable multilingualism rather than rampant language shift (Showalter 2006:41). Managing linguistic diversity in the church 165 Figure 1. Language map of Burkina Faso (SIL 2004) 2.2 Mission and Mennonite church in Burkina Faso The Africa Inter-Mennonite Mission (AIMM) was formed through the collaboration of several North American Mennonite denominations. In 1978, its work expanded to the province of Kénédougou in Burkina Faso (Traoré 1992:3). By 1992, seventeen AIMM missionaries were working in Burkina Faso, churches were meeting in three locations, and the denomination had received legal recognition and taken its current name of ÉÉMBF (Église évangélique Mennonite du Burkina Faso) (Traoré 1992:4). For AIMM, Bible translation into local languages was part of a larger church-planting strategy, rather than an end in itself. AIMM missionaries thus saw their strategy as distinct from that of Wycliffe Bible Translators, the much larger Bible translation organisation that was also working in the country (Bertsche 2006:1). Jim Bertsche, a former AIMM-BF2 missionary, describes the place of Bible translation in AIMM strategy as follows (Bertsche 2006:2, translation mine): 2 I will follow the practice of referring to AIMM missionaries in Burkina Faso as AIMM-BF missionaries, to differentiate them both from the AIMM central office and from AIMM workers in other African countries. 166 Anicka Fast For AIMM ... the analysis of languages and the translation of the Bible were a strategy to attain another basic goal, namely the evangelisation of the people of the Kénédougou region in order to plant, later, the Church of Jesus Christ of Mennonite orientation. It was our hope that by choosing to live in rural settings, taking their own languages seriously and at the same time valorising the positive qualities of their own cultures, we could create relationships of friendship and of trust that would encourage the villagers to respond to the invitation to accept Jesus as their Saviour. While AIMM-BF and ÉÉMBF had always collaborated to some extent (Kampen Entz 2001:38), the degree of formal collaboration between the two organisations has recently increased significantly. Between 2003 and 2006, AIMM initiated a major restructuring of the formal relationship between the Mennonite organisations that made up it up, the missionaries working in various African countries, and the national churches (Partenariat l’Afrique 2006:1). Following the restructuring, missionary representatives of each ongoing project meet several times a year with church leaders for collaborative decision-making in a forum known as the Table Ronde. At the same time, AIMM council meetings have been replaced by Partnership Councils with donor agency, missionary and national church representation. The new structure has been affirmed by both missionaries and church leaders, and appears to constitute a step closer to the vision articulated years ago by the current ÉÉMBF president, when he stated that “the moment has come where the church and the mission must find a common strategy conforming to the context in order to combine their force to accomplish the order of the Master” (Traore 1992:4). Moreover, since the new structure leaves less room for missionaries to act independently, while considerably augmenting the decision-making power of church leaders, it has allowed certain areas of long-standing conflict between church leaders and missionaries to be addressed more directly than in the past. The present research was motivated by the impression that differing views of the role of language in evangelism and church-planting form one such conflict area. Managing linguistic diversity in the church 167 3. Methodology 3.1 Sample selection and variables During seven weeks spent in Burkina Faso from 2nd June to 21st July 2007, I conducted 28 interviews with individuals and groups, in a sample stratified by position vis-à-vis the church (Table 1). In villages, interviews were conducted with 12 groups, typically made up of 2-5 people and homogeneous in age and gender (Table 2).3 Table 1. Stratification of sample by position vis-à-vis church Village groups 12 National church leaders 8 Missionaries 8 Total 28 3 Since this study was intentionally small in scope, it was inevitable that certain variables which were not included in sample stratification have clear effects on the results. One such variable was education. The groups were not homogenous with regards to education level (and hence knowledge of French) and despite the interpreter’s best efforts to draw out others’ opinions, this occasionally resulted in the most educated person dominating the discussion. Among Mennonite Toussian speakers, however, I could not introduce education as an additional variable, since there were not enough educated speakers to make this possible. 168 Anicka Fast Table 2. Stratification of village groups Younger (15-39) Village 1-Kourinion (5) Village 2-Djigouéra (5) Male Female Male Female 1 1 1 1 1 (mixed-gender)4 2 2 Village 3-Orodara (2) Total Older (40+) 3 3 1 1 1 2 3 While the sample is stratified, it also has aspects of a network approach (Milroy 1987) since it represents a high proportion of Mennonite missionaries, church leaders, and Toussian Christians in the area. I interviewed at least one member of all the Mennonite missionary families who worked in Burkina Faso, an estimated 90% or more of Toussian church members in the two designated churches, and a high proportion of Burkinabè Mennonite Bible translators and pastors working in Kénédougou, including the president of the Burkina Faso Partnership Council and the president of the ÉÉMBF. The selection of church leader and village group participants was made jointly by missionaries and church leaders before my arrival. 3.2 Interpretation and interview language Mr. Goarè Terri, a member of the Djigouéra Mennonite church who identifies with the Samo ethnic group, was selected by church leaders and missionaries to work as an interpreter for the group interviews.5 As a long-time schoolteacher, a well-respected church member and an experienced interpreter of church services to and from French and Jula, Mr. Terri was skilled in drawing out less talkative members of a group and in fostering a collaborative group dynamic. 4 5 This group was excluded from gender-based analyses. I had judged it more important to work with a Mennonite church member than with a Toussian speaker (there were no highly educated Toussian speakers in the Mennonite churches). Managing linguistic diversity in the church 169 Interviews were conducted, as far as possible, in a language in which participants were comfortable. I conducted the English and French language interviews with missionaries and most church leaders myself, with Mr. Terri serving as an interpreter in the two cases where church leaders stated a preference for Jula. In the case of group interviews, I trained Mr. Terri to conduct the interviews himself in Jula while I monitored the recording equipment and recorded responses.6 3.3 Ethics and oral consent Both in the initial request for permission to do interviews in the churches (Appendix 1) and in the contact with individual interviewee(s), I tried to adhere to the SOAS Statement on Ethics.7 Before all interviews, either I or Mr. Terri presented the interviewee(s) with information about the goal of the research, clarified my position as a fellow Mennonite and a student who had obtained permission from the ÉÉMBF to conduct these interviews, and clarified participants’ right to withdraw from the interview, to not answer specific questions, and to remain anonymous. I also presented all interviewee(s) with a one-page document (written in French) outlining the above points. 3.4 Moral dilemmas in my position as researcher I chose to present myself to participants primarily as a fellow Mennonite and a potential future missionary. My research was presented as a contribution to the ongoing conflict about language, and thus as a service to the church. However, I clarified my academic affiliations to all interviewees, and tried not to hide the personal advantage that could accrue to me through academic advancement related to this work. My presentation of myself in this way had particular repercussions, which may be considered as advantages or disadvantages. For example, questions of payment for the interpreter and of recompense for group interviewees’ time were determined by church leaders, relieving me of any need to negotiate 6 This training included interviewing Goarè Terri myself to give him a feeling for how participants might react to questions, and having him conduct a practice interview under my supervision of three people selected for this purpose. 7 Available online at:-----------------------------------------------------------------------------mercury.soas.ac.uk/research_and_galleries/ethics/StatementonEthics.pdf. 170 Anicka Fast payments or gifts.8 The very high degree of cooperation with every aspect of my research shown by missionaries, church leaders and church members alike, was almost certainly due to my positioning myself as a fellow believer. At the request of my first interviewee (who is also the president of the ÉÉMBF), I began interviews with a prayer when this seemed appropriate. It was often difficult to separate my research activities from discussions with missionaries and church leaders about my future work in the province. However, I felt strongly that discussing my ongoing discoveries or observations with others as I proceeded would have been unethical, even if this would have been advantageous to everyone in helping to clarify a potential future position for me as a linguist within the evolving collaborative structure. I communicated this concern to church leaders and missionaries, and remained purposely vague when people asked me if my research was yielding any interesting results. Overall, I sensed respect from most participants about the need to maintain confidentiality. 4. Methodology and results Various methods were used in order to address the following main research questions: 8 1. what languages do participants find useful in different domains (especially church)? 2. to what ideological resources do participants appeal to back up these perceptions? and 3. how do institutional linguistic practices map perceptions of the sociolinguistic situation (i.e., embody particular language ideologies [Spitulnik 1998:163])? I did not pay individual interviewees for their time, so as not to undermine the relationship of trust and the idea that this research was contributing to a mutual goal of helping to deal with a sensitive situation within the church. However, I did throw a party at the end of my stay with the express goal of thanking all the individual interviewees as well as those who had provided invaluable logistical help. Managing linguistic diversity in the church 171 Table 3. Overview of types of data collected Village groups Church leaders Missionaries 1. Demographic data X X X 2. Domain model activity X X X 3. Attitude statements X X X 4. Open questions No data collected X X 5. Participant observation 7 churches, representing the churches of all but 2 individuals and 2 groups In order to minimise the need for interview transcription, keep literate and non-literate participants on the same footing, and allow the groups freedom to discuss the questions in a language of their choice, I designed the domains activity and the attitude statements to yield numerical data that were easy to tabulate. I took the mean of divergent responses within groups,9 because the potential group effect made it impossible to assume that all responses represented individual opinions. I broadly transcribed responses by missionaries and church leaders to the open questions, and tabulated them according to major themes and sub-themes,10 following procedures used by Robinson (1996) and Reh (2004) as an aid to consistency and objectivity. An analysis of the data illuminates conflicting definitions of church that centre on the identity of church as ideally mono-ethnic or multi-ethnic. Moreover, the data suggest that missionaries’ definition of church as monoethnic is underwritten by their tendency to equate mother tongues with ethnic identity. 4.1 Demographic trends Besides the social variables of gender, age and position vis-à-vis the church by which the sample was stratified, I also collected demographic data regarding other social variables, in order to check possible correlations 9 Divergent responses occurred only in the attitude statements and the matched-guise activity and were relatively rare. 10 Given the relatively informal nature of the interviews and my encouragement of participants to tell stories and go off on tangents, I occasionally included in these tables relevant statements or topics that had been shared with me during a different part of the interview. Also, in order to protect the identity of participants, I will use the masculine pronoun to refer to respondents of either sex. 172 Anicka Fast between these variables and attitudinal data. In groups, these data were collected by asking each participant for an individual response. According to Goarè Terri and to another Burkinabè with extensive interviewing experience in the area, data collected in this way should be relatively free from a “group effect” (Robinson 1996:80) since participants will be unlikely to bend the truth about their possessions, age or other such information in front of others.11 Figure 2 and Table 4 show the demographic characteristics of church leaders, missionaries, and village groups. These data demonstrate that a clear understanding of the social correlates of membership in categories such as ‘missionary’, ‘church leader’ or ‘village group’ is necessary for a balanced interpretation of the attitudinal data. Position vis-à-vis the church in Burkina Faso clearly coincides with other demographic characteristics, so that missionaries, for example, also happen to be more educated, wealthy, and fluent in more languages, to see themselves as more powerful, and to have spent the greatest proportion of their lives in church. It is therefore difficult to know whether any given attitudinal trend is characteristic of missionaries, church leaders, or villagers, or simply of anyone with particular levels of education, church participation, wealth, self-ascribed power, or fluency in various languages. Multivariate analysis on a larger sample would be necessary to elucidate the relationships between these social factors. Figure 2. Self-reported fluency in various languages 11 The only question to which a major group effect seemed to apply involved participants rating their own power on a scale of 1 to 5. Participantss seemed reluctant to rate themselves differently from others, even when they were clearly much more educated or wealthy. Managing linguistic diversity in the church 173 Table 4. Demographic characteristics of participants Age in 2007 Years of schooling Year of adhesion to church Proportion of life in church Index of power (self-ascribed, scale of 1 to 5) Missionaries 48.4 22.5 1961 95.2% 4.0 Church leaders 41.3 16.5 1985 53.6% 3.0 Village groups 35.4 4.3 1996 31.6% 3.0 T-test (df = 7); p-values less than 0.1 shown; p-values italicised when p < 0.05 Church leaders vs. missionaries 0.14 0.01 0.001 1.9E-04 0.06 0.04 5.1E-11 2.6E-06 2.1E-06 0.01 0.30 4.9E-06 0.03 0.04 0.99 Self-reported fluency* in... All** languages Jula French Local*** languages English Mooré Missionaries 8.6 1.0 2.0 1.1 2.0 0.3 Church leaders 7.9 2.0 1.8 2.4 0.8 0.5 Village groups 5.3 1.9 0.9 2.4 0.1 0.0 Missionaries vs. groups Church leaders vs. groups T-test (df = 7); p-values less than 0.1 shown; p-values italicised when p < 0.05 Church leaders vs. missionaries Missionaries vs. groups Church leaders vs. groups 0.57 0.002 0.15 0.06 1.9E04 1.2E-03 0.001 2.9E-04 3.1E-04 6.5E13 5.1E-03 0.19 6.1E-03 0.99 0.01 0.44 4.4E02 * 2 indicates reported fluency; 1 indicates some knowledge of the language; 0 indicates no knowledge ** Sum of fluency scores in all languages including those not shown (e.g., German) *** Fluency in more than one local language may result in a score higher than 2 174 Anicka Fast 4.2 Domains model activity: perceived utility of different languages in different domains Photographs representing language domains and coloured rings representing languages were presented to participants. They were told to place zero, one or two rings on each photograph according to whether they thought competence in the relevant language would not be helpful, would be slightly helpful, or would be very helpful in that domain. I thus obtained an overview of the perceived utility of different languages in different domains. Counting the proportion of rings of each different colour provided a rough measure of overall perceived utility of a given language in an individual’s repertoire. By providing rings in five different colours, participants were implicitly encouraged to think in terms of their entire linguistic repertoire rather than being required to identify a single useful language per domain. Overall, the relative utility of the vernacular, Jula and French is perceived remarkably similarly across all three groups, as Figure 3 and Table 5 demonstrate. Figure 3. Perceived utility of mother tongue, Jula, French, Mooré and other languages in nine domains % of total rings used in domains model activity 40% 35% Church leaders 30% 25% Missionaries 20% Village groups 15% 10% 5% 0% % Mother tongue % Jula % French Language % Mooré % Other Managing linguistic diversity in the church 175 Table 5. Overall perceived utility for mother tongue, Jula, French, Mooré and other languages Total number of rings % Mother tongue % Jula % French % Mooré % Other Church leaders 29.9 28.9% 36.2% 28.4% 2.7% 0.8% Missionaries 33.5 32.0% 36.0% 24.6% 1.2% 2.8% Village groups 38.1 24.3% 33.4% 25.6% 7.0% 6.0% t-test (df = 7); p-values less than 0.1 shown; p-values italicised when p < 0.05 Missionaries vs. church leaders 0.17 0.61 0.97 0.37 0.29 0.46 Missionaries vs. group 0.08 0.02 0.50 0.75 3.8E-06 0.32 Church leaders vs. group 0.01 0.32 0.48 0.19 0.003 0.05 In most of the nine domains, all participants agree on which language is the most useful (e.g., mother tongue at home and in the field, French at school and in the government office). However, there are two main ways in which missionaries differ from other respondents. First, missionaries see more overall utility for the mother tongue than villagers do, holding a stronger preference for the mother tongue in six of the nine domains. Second, the church domain is contested. Missionaries see the mother tongue as the most useful, while the church leaders and groups see Jula as the most useful (Figure 4).12 12 Cross-tabulations showed strong relationships between perceived language utility and the factors of age and gender among the village groups. Within the village groups, the younger respondents see higher overall utility for French 27 % of total rings for younger participants vs. 23.5% for participants over 40; p < 0.05, df = 4). In addition, women see less utility for French in the mosque, more utility for the mother tongue in the market, and less utility for Jula at home or in the field, compared to men. It seems that women have an overall preference for the mother tongue in domains where Jula is also present. Other social factors had little or no explanatory value. 176 Anicka Fast Figure 4. Perceived utility of mother tongue, Jula and French in church Perceived utility (2 = most useful) 2.5 Church leaders 2.0 Missionaries 1.5 Village groups 1.0 0.5 0.0 Mother tongue Jula French Language 4.3 Domains activity: Linguistic repertoires and communication problems Participants were presented with four figurines that represented a young woman and man (ie. of marriageable age), and an old woman and man (of grandparent or post-childbearing age). They were asked which languages these people would be likely to speak if they lived in the same village/town as the participants. Again, zero, one or two coloured rings could be placed on each figurine to represent lack of fluency, partial or complete fluency in each language. Figure 5 shows the perceived linguistic repertoires of the young man, young woman, old man and old woman. The degree of agreement between church leaders, missionaries and village groups is striking for the repertoires of the old man and woman. However, the village groups have a significantly better opinion of young men’s and women’s levels of both Jula and French. It is also worth noting that the village groups perceive young people to have equal competence in the mother tongue and Jula, while both missionaries and church leaders perceive the average young man or woman to know their mother tongue better than Jula. Managing linguistic diversity in the church 177 Figure 5. Perceived linguistic repertoires of an average young man, young woman, old man and old woman Young man Young woman 2.5 2.5 Church leaders 2.0 Church leaders 2.0 Missionaries Missionaries 1.5 1.5 Village groups 1.0 0.5 Village groups 1.0 0.5 0.0 0.0 Mother tongue Jula French Mother tongue Jula French Old woman Old man 2.5 2.5 Church leaders Church leaders 2.0 2.0 Missionaries Missionaries Village groups 1.5 Village groups 1.5 1.0 1.0 0.5 0.5 0.0 0.0 Mother tongue Jula French Mother tongue Jula French Figure 6 shows the perceived difference in linguistic competence between old and young speakers. The missionaries differ from the other participants in that they perceive no change in mother tongue competence and only a relatively small increase in Jula and French competence in the younger generation. This differentiates them sharply from the village groups, who see a huge increase in both French and Jula competence. The church leaders, on the other hand, are unique in their perception that mother tongue competence has decreased significantly over time. 178 Anicka Fast Difference in reported competence between older and younger generations (positive value indicates a perceived increase in competence) Figure 6. Perceived difference in mother tongue, Jula and French competence between old and young 1.4 1.2 1.0 0.8 Mother tongue 0.6 Jula 0.4 French 0.2 0.0 -0.2 Church leaders Missionaries Village groups -0.4 -0.6 4.4 Domains activity: Modified ‘matched-guise’ activity Participants were next introduced to six figurines representing three young men and three young women, with differing linguistic repertoires as shown in Table 6. Table 6. Linguistic repertoires of characters in modified matched-guise activity Mother tongue Young male and female #1 Young male and female #2 X Young male and female #3 X Jula French X X X Managing linguistic diversity in the church 179 I asked participants to choose the most and least ideal church teacher, school teacher, farmer, and spouse and friend for their child. This activity represents a modification of the classic matched-guise test (Lambert et al. 1960), using visual rather than auditory stimuli and having participants choose between entire linguistic repertoires rather than single varieties. The indirect aspect is also removed since there is no mediating element between participants’ judgments about a particular speaker and their judgments about his/her language. These modifications reflect both the difficulty of representing an entire linguistic repertoire in an auditory stimulus, and the trend in language attitude studies in Africa to avoid highly indirect methods such as the classic matched-guise because of its implicit assumption of a one-to-one correlation between a language variety and a speech community (Fast 2007b:14-15), which the present research attempts to deconstruct. The data collected through this activity provide a view of the connections made by participants between language, and economic and educational aspirations. Previous sociolinguistic studies in West Africa (eg. Djité 1988, Woods 1995) have shown exoglossic languages of wider communication (henceforth LWCs), regional LWCs and local vernaculars to be associated with descending levels of prestige. Following this assumption, I assigned the mother-tongue-only character in this activity a score of zero, the Jula and mother-tongue speaker a score of 1, and the French and Jula speaker a score of 2. Due to the limited scope of this paper, only the data on the ideal church leader will be discussed here. The village groups most strongly associate the worst church teacher with the character who only speaks the mother tongue. This differs significantly from both missionaries and church leaders (p < 0.05). However, the general agreement between missionaries and church leaders masks significant differences seemingly related to urbanity and rurality. As Figure 6 shows, for the eleven church leaders and missionaries associated with rural churches, the ideal church leader is someone who speaks the mother tongue in addition to Jula; knowledge of French at the expense of the mother tongue is seen as a liability. There is thus a clear opposition between the attitudes of rural villagers and urban leaders (both church leaders and missionaries) on the one hand, and those of rural leaders on the other. Since eight of these eleven rural leaders are involved in translation work, there is a possible connection between language attitude and the presence or absence of vernacular Bible translation and literacy work. This would account for the different attitudes of the village groups, since no translation work is going on in their villages. However, data from members of urban and translation churches would be required to demonstrate this clearly. 180 Anicka Fast M e a n la n g u a g e s c o r e (2 = F R + J ; 1 = J + M T ; 0 = M T o n ly ) Figure 6. Preferences for best and worst church leaders among village groups, and urban and rural church leaders and missionaries 2.0 1.8 Urban church leaders and missionaries (n=5) 1.6 1.2 Rural church leaders and missionaries (n=11) 1.0 Village groups (n=12) 1.4 0.8 0.6 0.4 0.2 0.0 Best church teacher Worst church teacher 4.5 Attitude statements Following the domains activity, I presented participants with eighteen ideologically charged statements regarding language. Attitude statements were presented in French to all missionaries and most church leaders, and in Jula to all other participants. Participants could indicate their agreement with the statement on a five-point Likert scale.13 Table 7 shows the mean responses to the ten attitude statements that had a high discrimination value (the difference between at least two of the groups attained statistical significance at p < 0.05). 13 In all group interviews and most individual interviews, I avoided the need for literacy and attempted to prevent confusion by using a diagram with a sad-looking character, a happy-looking character, and three intervening spaces, to which participants could point. Managing linguistic diversity in the church 181 Table 7. Mean agreement with 10 attitude statements Key: 5 = strong agreement; 1 = strong disagreement Church leaders Missionaries Village groups 1. Speaking Jula has made us abandon our customs and traditions. 1.6 3.9 3.8 2. Jula is the language of Islam. 2.3 3.8 2.4 3.5 2.8 4.7 3.5 1.4 1.4 4.4 4.6 5.0 4.0 4.9 4.8 4.0 2.0 4.6 3.3 2.1 4.0 3.8 3.1 5.0 3.4 3.4 1.4 3. If we use Jula in church, we build unity in the Mennonite church of Burkina Faso 4. Using the mother tongue in church is a threat to unity among the Mennonite churches. 5. If we use the mother tongue in church, we build unity within our ethnic group. 6. Children will learn to read better when they are taught in their mother tongue. 7. There are enough resources to teach children to read in their mother tongue. 8. We will always be real [name of ethnic group], even if we no longer speak our mother tongue. 9. The people of my ethnic group will always speak their mother tongue. 10. Islam is a threat to our mother tongue. The examination of these responses suggests a distinct attitudinal profile for church leaders, missionaries and villagers. The village groups may be characterised as relatively carefree. They do not appear overly concerned about mother tongue preservation, but hold strong positive attitudes to both mother tongue and Jula, believing that both these languages contribute to unity of some sort within the church. In contrast, the missionaries are characterised by concern about the potential negative effects of Jula and the potential loss of the mother tongue, and by a strong belief in the superior unifying potential of the mother tongue over Jula. They also draw the strongest link between language and ethnic identity (statement 8). Finally, church leaders seem to be most concerned about the divisive potential of mother tongue use in Mennonite churches, and the least open to perspectives that connect Jula with a threat to customs or traditions. Total present (myself excluded) Hymnbooks 21 Jula, French Secondlargest city (population 434,000 in 2007) 9 Jula, French Rural translation churches Orodara Largest town in province; population 20,000 (2007 estimate) 110 No church copies Kotoura Sicité village 74 Songs copied into notebooks Samoghohiri Dzuun village with significant proportion of Sénoufo Other rural churches Kourinion Toussian village 28 30 Jula, Samogho (personal copies) No church copies Djigouéra Toussian village 32 No church copies 80% 55% 35% 0% 0% 10% 25% Songs: % Jula 20% 45% 65% 88% 57% 80% 69% Songs: % Mother tongue 0% 0% 0% 13% 43% 10% 6% Number of ethnic groups among church members (source of information) Sermon Summary: Main languages used in service 9 (pastor's wife) French French; Jula used only in songs 3 (pastor) 7 (pastor, missionary attendee) 1 (pastor, translator, former missionary) 4-5, but mostly Dzuun (translator, missionary) no information available 7 (pastor) French Jula, translated into French from pulpit (not for our benefit) French, translated into Sicité from pulpit Dzuungoo, translated into Jula from pulpit French, translated into Jula from pulpit Jula, translated into French from pulpit (for our benefit) Jula translated into French; individual prayers and Bible readings in Jula only French translated into Sicité for our benefit; normally, Sicité translated into Jula or French; individual prayers in Sicité Dzuungoo translated into Jula; individual prayers in Dzuungoo only; French used only by visitors Jula; French translation for visitors; announcements, personal testimonies and Bible readings in Jula only; Toussian used only for one song Jula; French translation for visitors; announcements, personal testimonies, prayers and Bible readings in Jula only; Toussian used only for one song French; Jula used only in songs Anicka Fast Songs: % French 182 Description of locality Capital city in Mossi-speaking area (population 1,181,000 in 2007) Bobo Table 8. Summary of observations on language use in seven Mennonite churches Urban churches Ouagadougou Managing linguistic diversity in the church 183 4.6 Participant observation While participating in church services in seven of the nine Mennonite churches in the region, I observed language choice for (1) major functions such as teaching and announcements; (2) functions where comprehension is not essential, such as singing; and (3) the inclusion and exclusion of visitors (Table 8). In this way, I hoped to gain insight into how the regular linguistic practices of institutions could embody particular ideologies of language (Spitulnik 1998:163), and how language choices could function as a map of the sociolinguistic situation, with valuations of different languages made implicit and “unattributable to any interests in particular” (Spitulnik 1998:180). An examination of these data suggests a classification of churches into three types: urban churches, rural churches with ongoing vernacular translation work in progress, and rural churches with no history of involvement with mother-tongue work. In urban and rural non-translation churches, French and/or Jula are the main languages, the choice of which seems to depend on comprehension levels within the congregation. Local languages are either absent or relegated to singing. Only in translation churches was a local language used either as the main language or as the language into which the entire service was translated. The proportion of songs in local languages was also highest in translation churches. Finally, a particular language is used for the inclusion of visitors in all churches: this is usually Jula in translation churches, and French elsewhere. 4.7 Open questions Through a set of open questions, asked only in individual interviews, I attempted to elicit discourse about language ideologies held by participants. The previous activities had usually served to bring to mind topics regarding language use in church that participants wanted to share. After asking the interviewee to indicate the language(s) he/she saw as appropriate for use in church, and the perceived advantages and disadvantages of each language, I asked questions about the perceived motivations of those working on vernacular translation and literacy, as well as of those who encourage the use of LWCs rather than vernacular languages in church. I also elicited theological/religious discourse that supported the participant’s policy stance, and gauged their commitment to this stance by asking for examples of times when they had personally attempted to influence the behaviour or attitudes of others in this matter. The responses to the open questions were a rich source of ideological meta-language that were often crucial in revealing the beliefs and convictions 184 Anicka Fast underlying the trends of the rest of the data. This section reviews the main themes of the responses and points out similarities and differences between missionaries and church leaders. 4.7.1 Preferred church languages A first question asked participants to state the languages that they felt should be used during the church services that they regularly attended.14 Table 9 and Figure 7 demonstrate a clear division in the results along urban-rural lines, with Jula and/or French seen as most the appropriate languages for use in urban church services, and the mother tongue in combination with an LWC seen as ideal in rural churches. This is one area where missionaries and church leaders who worked in similar contexts were in substantial agreement, and the disagreements that surface in the responses to the following questions should be interpreted with this in mind. Table 9. Preferred languages for church use among urban and rural missionaries and church leaders Urban (n=6) Rural (n=11) Missionaries (n=3) Church leaders (n=3) Mean Missionaries (n=6) Church leaders (n=5) Mean Mainly French, with Jula songs 1 1 33% 0 0 0% Jula & French 2 2 67% 0 1 9% Mother tongue & LWC 0 0 0% 6 3 82% Mother tongue only 0 0 0% 0 1 9% 14 In one case, a missionary expressed her opinion about the kind of church that she would like to see start among the people with whom she worked, since no church (yet) exists in that village. Managing linguistic diversity in the church 185 Figure 7. Preferred languages for church use among urban and rural missionaries and church leaders 90% 80% % of respondents 70% 60% Urban (n=6) 50% Rural (n=11) 40% 30% 20% 10% 0% Mainly French, with Jula songs Jula & French Mother tongue & LWC Mother tongue only 4.7.2 Advantages and disadvantages of various languages A subsequent set of questions asked participants to comment on the advantages and disadvantages of the mother tongue, Jula and French in church settings. Overall, missionaries and church leaders had quite similar views of the sociolinguistic situation with regard to French, generally agreeing that its utility was limited to the inclusion of non-Julaphones. They also cited very similar disadvantages to mother tongue use, centering on its potential to exclude those of other ethnic groups. However, views about the advantages of mother tongue use and about the advantages and disadvantages of Jula differed substantially. When interpreting these results, it is worth remembering that missionaries report significantly lower levels of personal fluency in both Jula and local languages (see Figure 2). While both missionaries and church leaders saw the wider communication possibilities as an advantage of using Jula, church leaders put more stress on communication than missionaries did, and also evoked issues of solidarity. One church leader emphasised that pastors who were educated in French should nevertheless make the effort to preach and pray in Jula for the sake of the uneducated people in the congregation, since a sermon in Jula would have more “power”. Church leaders from two rural churches also suggested that 186 Anicka Fast using Jula, rather than only the mother tongue, was a useful way for their congregation to learn Jula words. For missionaries, a similar advantage of Jula use was the possibility of connections with the larger Mennonite denomination that this brought. For missionaries, disadvantages of Jula use revolved around the superficiality of communication and its contribution to giving the church a “foreign” character, leading to a loss of credibility in evangelism in the local area. One missionary also stated that Jula use was a threat to the mother tongue. In contrast, church leaders saw fewer disadvantages to Jula use in church, and never described Jula as being in any way foreign. Finally, statements about the advantages of mother tongue use seemed to draw on very different sets of ideas. The reasons cited by church leaders involved practicality (especially in some rural churches where the mother tongue seems to be the language best understood) and ease of communication. They also referred to solidarity-related reasons such as the desire to reach people who spoke no other languages (specifically, the very young and the very old). A few pointed out reasons for mother tongue use that went beyond mere practicality or inclusiveness: a sense that worship is more meaningful in this language, that the mother tongue is a source of joy and that it makes church feel like their own. The missionaries, on the other hand, while also emphasising comprehension, added the idea that comprehension in the mother tongue is somehow inherently more profound or deep. Several cited the term “heart language” to express this idea. Themes of locality, participation and ownership were also strong, especially the idea that the mother tongue is associated with a particular area and therefore should be used in that area, and the belief that mother tongue use will make the gospel seem less foreign and will encourage greater local participation in church as a result. In sum, while there is much common ground in the views of church leaders and missionaries, there is also significant divergence. The choice of languages for church leaders centers around questions of inclusion, solidarity and practicality. While missionaries share these concerns to some extent, their views are also coloured by a complex of ideas about authenticity, identity and place that are almost absent from the discourse of their colleagues. 4.7.3 Perceived motivations underlying major policy stances Tables 10 and 11 present participants’ responses about the perceived motivation behind mother tongue translation and literacy work, as well as that underlying the preference for Jula or French use in church at the expense of the mother tongue. Where church leaders and missionaries gave very similar responses, these are italicised. Managing linguistic diversity in the church 187 Table 10. Perceived motivations for mother tongue translation and literacy work Theme Reasons given by missionaries (n=8) Reasons given by church leaders (n=8) Preservation of language (2 missionaries, 2 church leaders) Language preservation Desire to raise perceived value of language (1) Belief that the Bible as a written medium will contribute to valorising the language (1) Desire for promotion of the language (1) The pride of being able to read in one's mother tongue (1) Desire to encourage people to speak their mother tongues (1) Desire for people to have the word of God in their languages (1 missionary, 2 church leaders) Desire for people to become Christian (1 missionary, 1 church leader) Evangelism Desire to share the good news with Belief that evangelism in the mother tongue people who would not otherwise be is more effective (2) able to know it (1) Desire for [ethnic group name] to come to Christ (1) Desire for people to understand the gospel better (5 missionaries, 4 church leaders) Belief that Jula or French will not clearly communicate gospel or speak to People can understand who God is better people's hearts (2) in their language, which is the heart Better language; discovery of God through a Belief that gospel will be best compreunderstood in mother tongue translation different language may not be complete (1) hension of (2) gospel Deeper, more authentic understanding of gospel (2) The Bible in the mother tongue has more Desire to share personal experience of power and is understood better (1) understanding God's word in one's language (1) Desire to avoid syncretism arising from use of an unfamiliar church language Naturalness, (1) authenticity, indigeneity Desire to bring God closer by demonstrating that he speaks people's language (2) Desire to encourage indigenous theology (1) Belief that use of mother tongue in church service is more natural (1) Desire to preserve traditional cultures during shift to Westernisation/modernisation (1) Globalization Ethnicity Personal Desire to go against the irreversible trend of ethnic mixing and globalisation (1) Belief that a stronger identity (through mother tongue) will help people cope with Westernisation and materialism (1) Desire for [ethnic group] to come to Christ (1) Desire to create ethnic churches (2) Love of one's people (1) Personal fascination with languages (1 missionary, 1 church leader) Theological Desire to participate in fulfilment of imperative biblical prophecies (1) 188 Anicka Fast Table 10 (cont’d). Perceived motivations for mother tongue translation and literacy work Theme Reasons given by missionaries (n=8) Reasons given by church leaders (n=8) Desire to keep people backward and limited (as shown by the lack of schooling Concern about injustice of stronger languages running over weaker ones (1) and books other than the Bible in the mother tongue) (1) Desire to help children learn to read better (1) Pride of being able to read in mother tongue (2) Desire to valorise small groups and languages (1) Justice, empowerment Desire to help people read and resolve problems in their own language (1) Desire to empower people (1) Desire to give people documents to read in their own language (1) People who know only their mother tongue should not be neglected or prevented from discovering God (1) Desire to valorise people by valorising their culture and language (1) Desire to preserve traditional cultures during shift to Westernisation/modernisation (1) Desire to preserve cultural and linguistic heritage (1) Connection with identity Desire to institutionally strengthen the or culture local community in the local language (1) Desire to reinforce dignity, culture and customs (1) Desire to help people keep their identity (2) Concern that mother tongue preservation is crucial to identity preservation (1) Managing linguistic diversity in the church 189 An examination of Table 10 suggests that the main areas of agreement with regard to the motivation for Bible translation centre on language preservation, evangelism, and improved comprehension of the gospel message. In other areas, missionaries and church leaders differed significantly. For example, two church leaders perceived primarily negative motivations for mother tongue work, such as the neo-colonialist motive to keep people limited by their mother tongue, and the desire to create ethnic churches. In contrast, some missionaries interpreted language work as a way to counteract the forces of globalisation, while the church leader who also saw this motivation interpreted it in a highly negative way. Although missionaries and church leaders shared the desire to contribute to biblical comprehension, missionaries emphasised the importance of authenticity and profundity of understanding, with some expressing the belief that spiritual experience would be deeper in the mother tongue. Finally, one of the most significant motivations evoked by missionaries which is entirely absent from the responses of church leaders involves seeing language as essentially connected to culture or identity, such that mother tongue promotion work becomes a form of identity or culture preservation. The following statement by a missionary clearly exemplifies this implicit connection. [A]s a Mennonite... [I have] the strong sense that to cope with Westernisation and materialism ... people need to be strongly rooted in their own identity, and you can’t move backwards, but somehow mother tongue needs to be preserved, because without mother tongue, identity’s lost. 190 Anicka Fast Table 11. Perceived motivations for the preference of Jula or French over mothe tongue use in church Theme Lack of fluency in mother tongue Communication Reasons given by missionaries (n=8) Reasons given by church leaders(n=8) The person does not speak any languages other than Jula or French/does not speak the mother tongue in question (2 church leaders, 1 missionary) The person is not able to learn relevant languages (1) More possibilities for communication/can reach more people (2 church leaders, 1 missionary) Communication in a wider domain/realm (1) It is easier, more efficient, practical (1 church leader, 3 missionaries) Ease, efficiency, practicality Belief that people know Jula well enough, making mother tongue translation unnecessary (1) Taking the easy route, laziness (2) Using mother tongue is too much work when another language will unify everyone (1) Desire for unity; avoid problems of unity (1 church leader, 3 missionaries) Fear of disunity, ethnic groups or missionaries "doing their own thing" (2) Unity Fear of encouraging people to be who they are; forcing them into a melting pot (1) A belief that Jula produces unity (3) Individualistic evangelicalism, resulting in use of Jula to achieve a superficial unity (1) Belief that unity is the most important thing (1) Assumption that everyone knows Jula and French, since they do (1) Lack of personal capacity or awareness Personality issues (not related to language) (1) Past negative experiences (1) Lack of awareness of the threat to language and culture (1) True connection with people is not a priority (1) Fear that missionaries are out to keep Christians isolated (1) Resources Desire to make use of existing French and Jula resources (2) Better comprehension in urban areas (1) Comprehension Citizenship The language of a minority ethnic group will not be understood by many (1) Desire to experience citizenship (1) Managing linguistic diversity in the church 191 When it comes to explaining the reasons for preferring Jula and/or French over the mother tongue in church (Table 11), most missionaries impute one or more negative motivations to church leaders, such as fear, personal lack of capacity, laziness, or the influence of individualistic ideologies. While church leaders perceive motivation to centre on a personal lack of knowledge of the relevant mother tongue, and on a desire to communicate with more people and to encourage comprehension between ethnic groups, missionaries are almost unanimous in assuming that a (misguided) desire for unity is a main motivating factor. 4.7.4 Relevant theological resources I asked participants to list the biblical and theological concepts that they found relevant to the question of language choice in church. Overall, the missionaries referred to many more biblical texts and doctrines than church leaders, indicating either a greater familiarity with these texts or greater perception of pertinence to the situation. All participants seemed to agree that the Bible depicts linguistic diversity as a basically good thing. Many point out that the Bible provides no support for a linguistic or ethnic hierarchy. Several church leaders and missionaries emphasise the inherent suitability of all languages as vehicles for the gospel, refusing to accept any language as being somehow superior. This indicates a shared familiarity with the doctrine of translatability, which has historically justified Bible translation (Sanneh 1992b:1). However, the missionary interpretation of the biblical texts moves beyond the basic translatability doctrine or the acceptance of linguistic diversity by incorporating the concept of ‘mother tongue’ in a way that is bound up with concepts of authenticity and identity. Thus, while two church leaders explicitly extend the goodness of linguistic diversity to cover LWCs, several missionaries posit a contrast between ‘mother tongues’ and other kinds of languages. For example, one missionary interpreted the Babel story as God’s way of avoiding a “melting pot” and evidence that God “is not looking for a Jula church”. The link between mother tongue and identity is exemplified by another participant’s statement that true identification with others is not possible “without a mother-tongue connection”. 4.7.5 Experiences of exclusion Both missionaries and church leaders indicated that they had experienced exclusion based on language in church contexts, as Table 12 and Figure 8 show. However, they differed both in their degree of acceptance of such exclusion and in what languages they had experienced as communication 192 Anicka Fast barriers. Church leaders were more likely to mention exclusion based on lack of English or French fluency, while missionaries tended to experience Jula and mother tongues as communication barriers. In addition, missionaries tended to downplay the effects of a lack of Jula or mother tongue fluency, with five missionaries commenting that their lack of ability to participate fully did not bother them. One stated that exclusion did not bother him because he was an “outsider”. Another stated that being “out of the loop” was not a problem because “[t]his is the [ethnic group name] church. It’s not to be meeting my needs; it’s to be meeting [ethnic group name] people[’s]”. In contrast, church leaders saw exclusion on the basis of lack of mother tongue knowledge to be a much more serious matter. The two church leaders who mentioned experiencing such exclusion both stated that they had intervened to reprimand the relevant people. Table 12. Participants’ experiences of exclusion in church contexts Missionaries (n=8) Church leaders (n=8) Participants who indicate feeling excluded from church activities 4 4 Participants who cite language barriers that prevent full participation in church activities 7 4 (n=8) (n=8) English 0 2 French 2 2 Jula 6 0 Mother tongue 4 2 Languages perceived as barriers by participants Managing linguistic diversity in the church 193 Figure 8. Languages cited by missionaries and church leaders as barriers to full participation in church activities 7 # of participants 6 Missionaries (n=8) 5 4 Church leaders (n=8) 3 2 1 0 English French Jula Mother tongue Language 4.7.6 Personal involvement in language policy promotion All participants cited cases when they had promoted a particular view of languages, suggesting that all see themselves as implicated in local language policy (Table 13). As the table demonstrates, both missionaries and church leaders engage in mother tongue promotion in church, but church leaders are more likely to limit their promotion to the domain of singing. Outside church services, missionaries are much more involved in mother tongue promotion than church leaders. In contrast, no missionaries mentioned having encouraged others to learn more languages as a way to minimise communication barriers, while five church leaders indicated having done so. Finally, although both church leaders and missionaries stated that they had intervened to encourage use of a more widely understood language instead of English or French, only church leaders said they had intervened to encourage the use of Jula instead of or in addition to the mother tongue at church meetings 194 Anicka Fast Table 13. Personal involvement in language policy promotion (responses not relevant to Kénédougou excluded) Missionaries (n = 7) Church leaders (n = 8) Encourage more mother tongue use at church events Encourage people to sing in their mother tongues in church (2) In songs Encourage people to express themselves in their language in church, both in songs and testimonies (1) When a preacher asks the congregation what language he should preach in, encourage the choice of mother tongue In all parts of (1) church service In personal conversations, speak of importance of using the mother tongue in village churches (1) Intervene to encourage mother tongue use in a church meeting with translation At church meetings into Jula for benefit of three non-mother tongue speakers present (1) In church meetings, advocate the use of mother tongue in all domains of life, including children's Sunday school (with Jula translation only on demand) (1) Promote mother tongue use outside church services or meetings In general Decline offers of translation Convince church leaders of need for Bible translation Distribute or raise awareness of mother tongue materials Attempt to convince church leaders to use the mother tongue in church settings In personal discussions, try to convince Speak the mother tongue whenever attending a neighboring town wedding of people of the need to take mother tongues into consideration and to see the someone within ethnic group area (1) importance of culture in church life (1) When visiting other churches, decline offer of translation into Jula or French (1) Decline offer of a personal Jula-toFrench translator at a wedding (1) Invite a church leader to supervise ongoing translation project while missionary is in home country for a year (1) Run a conference for church leaders Convince pastors of one's own ethnic about the importance of Bible translation group of the importance of translation into (1) the mother tongue by appealing to the need for documentation of the language Participate in inauguration of a bible before it disappears (1) translation project in another church, including discussions where missionary attempts to convince others of the importance of mother tongue translation (1) Distribute biblical and traditional audio materials in order to save the language and evangelise at the same time (1) In church, intervene to remind people of available mother tongue material so that Bible readings can be in the mother tongue when possible (2) Speak privately with a church leader after his choice to have no translation into mother tongue at a church event (1) Managing linguistic diversity in the church 195 Table 13 (cont’d). Personal involvement in language policy promotion (responses not relevant to Kénédougou excluded) Missionaries (n = 7) Church leaders (n = 8) Initiate relationships with civil servants Initiate belonging in ethnic group; giving or selling relationships audio material (1) with civil Initiate relationship with mayor with goal of servants encourage leaders of ethnic group to work in literacy and cultural preservation (1) Connect Attempts to connect ethnic group mayor and ethnic group civil servants with resources for cultural with outside preservation (e.g. a US-based grant) (1) resources Encourage language learning Encourage members of the Partnership Council to learn English or French (whichever language they don't already know) (1) Encourage people to learn more languages (3) Encourage people to learn more languages Try to assure education possibilities for church members in French or Jula (1) Encourage the use of a common language Make an effort to speak Jula among a group of church leaders instead of French, so that non-French speakers will be included (2) Speak privately to a church member encouraging him to use Jula instead of French to share testimonies and prayer Instead of requests in church, so that everyone can French only When preaching in French, ensure that there is understand (1) an interpreter (1) When French is required because of presence of missionaries who do not speak Jula, ensure inclusion of non-francophones through Jula translation (1) Encourage a local church leader indirectly (through a third person) to have Scripture readings in Jula rather than French (1) Try not to encourage church members to speak English with them when this will exclude others Make an effort not to speak English to other Instead of (1) Partnership Council members who know English only Require other missionaries to submit both French, so that none will be excluded (1) English and French versions of their reports for Table Ronde meetings (1) Intervene in a church meeting to rebuke church members for speaking the mother tongue without translation and thereby excluding others and creating division (1) Instead of mother tongue only Encourage the use of a common language in church meetings (2) Instead of mother tongue In general Encourage use of a common language in any situation where multiple languages are spoken (e.g. village development meetings) (1) Try to convince people to use lower-down language on the common scale for comprehension reasons (1) 196 Anicka Fast 5. Analysis and discussion In this section, I draw on the research results to illuminate participants’ adherence to two conflicting language ideologies, both of which centre on definitions of church as ideally mono-ethnic or multi-ethnic. I then explore the power imbalance that characterises this ideological conflict, suggesting that it is largely due to the historical allure for powerful Western actors of an essentialising ideology that links language to cultural identity. After summarising recent research that demonstrates the historical power of this ideology and listing some of the potential distortions with which it has recently been associated, I trace its integration into Bible translation discourse and demonstrate how it functions to disempower church leaders in the current debate. I conclude by offering suggestions for power balancing, relevant both within and outside the Anabaptist Christian moral framework, that may allow ideological contestation to continue in a more fruitful manner. 5.1 The role of ideology in enacting the conflict 5.1.1 Conflicting definitions of church I suggest that the various disagreements between participants center on competing definitions of church identity: specifically, on whether the ideal church is multi-ethnic or ethnically homogeneous. An examination of the data demonstrates the constant recurrence of questions of church identity at the heart of participants’ efforts to distinguish insiders from outsiders, acceptable from unacceptable exclusion, and diversity from division. 5.1.1.1 Insiders vs. outsiders Both in the interviews and through church linguistic practices, competing definitions of who constitutes an ‘outsider’ are expressed that function to enact particular definitions of church. In translation churches, vehicular languages such as French and Jula are subordinated to the local language, thus defining ‘outsiders’ as those who do not speak the local language, and insiders as mono-ethnic. Church linguistic practice thus functions to underwrite a definition of church whose boundaries coincide with those of a particular ethnic group. Views expressed in the domains activity and the matched-guise activity, such as the belief that the mother tongue is the most useful or ideal language in church settings, and the tendency to define an ideal church leader as someone who speaks his or her mother tongue in addition to Jula, further reinforce this definition. Another Managing linguistic diversity in the church 197 example of this implicit distinction between insider and outsider is one missionary’s statement that in an ideal church the mother tongue would be the main language, with simultaneous Jula translation “in the back” for the benefit of the “many” civil servants posted in the area and of wives from different ethnic groups who had married into the community years ago but had not yet learned their husbands’ languages. Long-time residents of a community are thus assigned a different status based on a definition of the ideal church as ethnically homogenous, with Jula functioning as a language of accommodation to outsiders. This contrasts sharply with the linguistic practices of other churches, where the choice of Jula and/or French as main languages functions to include attendees from all over the Julaphone region as insiders, thus implicitly defining the church as a multi-ethnic body. 5.1.1.2 Acceptable vs. unacceptable exclusion Both church leaders and missionaries share a deep concern about questions of inclusion and exclusion through language in church settings. However, conflicting definitions of church are reflected both in their alternative understandings of what constitutes unacceptable exclusion and in their interventions to encourage the use of common languages in certain situations. Missionaries’ tendency to downplay the importance of their own comprehension of church services, even to the point of refusing efforts at accommodation, seems to be due to an ethnically-based definition of church from which they are necessarily excluded. This view leads them to express surprise at the extent to which non-mother-tongue speakers are accommodated in church services. One missionary explicitly connected such efforts at inclusion with language loss, saying that [ethnic group] people are “so hospitable” that they “bend over backwards to make Jula speakers at ease.” Another encouraged a switch to the mother tongue for part of a church meeting at which a small group of non-mother-tongue speakers was present, in order to encourage greater participation by those less fluent in Jula. In short, missionaries sometimes see a degree of exclusion of non-mother-tongue speakers to be necessary to ensure that all mother-tongue speakers can fully participate in the language they understand best. In contrast, several church leaders find it unacceptable for a mother tongue to be used that is not understood by all present. Moreover, a direct link between exclusion on the basis of mother tongue use and church identity is expressed in comments by two church leaders. One stated that the neglect of some church members through the use of a local language does not demonstrate “a Christian spirit”. Another expressed the view that village 198 Anicka Fast churches must transcend the tendency to ethnic exclusiveness exhibited by their non-Christian neighbours in the face of growing immigration into the Kénédougou area. In his view, “when the autochthonous people speak between themselves and those who are not of that ethnic group are not taken into account, this does not reflect the true identity of the church” (emphasis added). For this church leader at least, the identity of the church is not compatible with the continued existence of a category of Christians who feel like “strangers”. As a result, although some church leaders noted the importance of mother tongue use for the inclusion of the few non-fluent Jula speakers in their congregations (eg. old women and children), the exclusive potential of Jula is generally de-emphasised. Jula is seen as functioning not only as a language of accommodation but also as one of solidarity and inclusion: a way to prevent anyone from feeling like a stranger. 5.1.1.3 Homogeneity vs. unity and diversity vs. division While missionaries and church leaders share a view of the essential goodness of linguistic diversity, they both enact strategies to contain this diversity so that it does not pose a threat to church unity. However, these strategies are determined by differing views of the relationship between linguistic homogeneity and ethnicity. In urban and rural non-translation churches, the limitation of one language primarily to singing functions to include linguistic diversity in a “safe” way (Spitulnik 1998:180), and allows ethnic diversity in membership to be retained by the use of a common language for all main functions. This is consistent with church leaders’ greater tendency to see the mother tongue as a threat to church unity (Table 7). In their view, the use of the mother tongue can only contribute to unity when it is held in common by all church members, and such a situation is seen as unsustainable since, as one church leader stated, “globalisation is there at all levels. You won’t find ethnic groups all by themselves anymore.” For missionaries, linguistic diversity is contained through the idealisation of mono-ethnic churches in which mother tongues can realistically be used as main languages.15 Ethnic homogeneity is thus perceived to bolster church unity. 15 Since translation work is taking place in largely mono-ethnic communities, it is important to consider the extent to which the ethnic composition of a church determines its linguistic practices and the extent to which it mirrors them. Participants from various types of churches made comments indicating their view that the use of mother tongues“worked” in rural, largely mono-ethnic churches but “didn’t work” elsewhere. However, the selection of communities in which to begin translation work Managing linguistic diversity in the church 199 5.1.2 An essentialising ideology Having demonstrated the existence of alternative and competing definitions of church, it becomes important to understand why missionaries and church leaders define the church in opposing ways. I suggest that missionaries’ definition of church as mono-ethnic is underwritten by their tendency to equate mother tongues with ethnic identity. A strong link between language and identity is expressed by missionaries in various ways throughout the interview. It is present in missionaries’ strong rejection of the idea that members of a given ethnic group will still be “authentic” if they no longer speak their mother tongue, and motivates the view of mother tongue preservation as essential to identity preservation. It inspires the idea that mother tongue use is associated with more profound spiritual experience, and that it will help people to see church as truly “theirs”. It crucially underlies a rhetorical distinction between mother tongues and other kinds of languages that allows the goodness of linguistic diversity in the Bible to apply only to mother tongues and renders paramount the right of mother tongue speakers to primary ownership of church. Even in areas of otherwise substantial agreement between missionaries and church leaders, the missionaries’ point of view is often intertwined with this additional ideological strand tightly linking the mother tongue to identity, authenticity and profound spiritual experience. This clearly explains some participants’ preference for mother tongues as main church languages: in their view, this is the most logical way to ensure authentic spiritual experience for the majority, even when it leads to the exclusion of a minority. 5.1.3 Summary: two ideological profiles When the role of this essentialising language-identity link in the definition of church has been clarified, two main ideological profiles emerge which can be summarised as follows. First, the ideology held by many missionaries and some church leaders involves an implicit definition of the church as ethnically homogenous, with the healthiest churches being those whose members are confident of their ethnic identity, and use their cultural resources, including language, to access may also have been influenced by the existing ethnic composition of communities. Thus, Spitulnik’s conclusion that the embodiment of language ideologies in institutional practices can be both a source and a reinforcement of particular language valuations (Spitulnik 1998:181) seems to apply to the Mennonite churches in Burkina Faso as well. 200 Anicka Fast authentic and profound spiritual experience. LWCs are helpful for communicative purposes, yet embody a potential threat to this definition of church that must be carefully controlled. Linguistic diversity is celebrated, yet excluded from church services. Individual identities and linguistic repertoires are downplayed in favour of a group-based ethnic identity associated with a single language, and a certain degree of exclusion of those who do not speak this language is considered acceptable. The ideology held by many church leaders on the other hand, defines church as multi-ethnic yet crucially inclusive. The healthiest churches are those whose members’ ethnic identity does not cause a barrier to fellowship, and languages that index particular ethnic identities therefore pose a potential threat. LWCs can function as expressions of solidarity, equality and inclusion. Positive attitudes toward the mother tongue co-exist with a high awareness of the potential divisiveness of language and a preference to define church as made up of Jula speakers even when this might mean the exclusion of certain segments of the congregation who are not fluent in Jula. 5.2 Essentialising ideologies and their disempowering effects Given the strong emphasis in the Christian tradition on the trans-ethnic nature of church, it seems particularly crucial to clarify how an ideology that idealises mono-ethnic churches could take hold and why it would affect missionaries more than church leaders. In this section, I suggest a response to both these questions. After reviewing recent scholarship demonstrating the historical power of an ideology linking language and identity in other contexts, I show how this ideology has gained legitimacy in Kénédougou due to its pervasiveness in Bible translation discourse. I then trace the specific effects of this ideology in the current situation, showing how it functions to powerfully delegitimise the conception of the church held by Burkinabè leaders on grounds that it is not sufficiently indigenous, while erasing from view the wealth differential between missionaries and national Christians. 5.2.1 The power of essentialising language ideologies Scholars of language ideology emphasise that linguistics is not an ideologically neutral science (Schieffelin et al. 1998). It has been associated with the production of ideologies sometimes called “vernacularist” which insist “on the authenticity and moral significance of ‘mother tongue’ as the first and therefore real language of a speaker, transparent to the true self” (Woolard 1998:18, emphasis original). For example, Cameron has shown how the assumption of an “organic connection between a people” and a “mother tongue”, influenced by 19th century European nationalist ideologies, underlay Managing linguistic diversity in the church 201 the development of the comparative method in linguistics and the theorising of concepts such as “native speaker” or “speech community” (Cameron 2007:277-278, see also Foley 2005). History shows us that such ideologies can have potent political effects, both positive and negative. For example, the choice of the Cherokee Nation in the 19th century to adopt a syllabic writing system for their language, accompanied by the rejection of missionary efforts to develop a phonetic orthography that would facilitate comparison between Cherokee and other languages, is perceived by Owens and other Cherokee scholars as an “act of linguistic, cultural, political, and possibly even religious resistance” that was overwhelmingly empowering for the Cherokee people (Owens 2006:10). Owens connects the rapid spread of literacy using the new syllabary with the fact that “for the first time in history an evangelised people were reading the New Testament translated by their own kinsmen, into their own language, using a writing system developed, refined and popularised entirely as their own” (Owens 2006:8). On the other hand, scholars have demonstrated how this ideology may have a primarily disempowering function. During Africa’s colonial period, for example, the analysis of African languages by both colonial agents and missionaries contributed powerfully to colonial social control by allowing precolonial political structures to be re-drawn along linguistic lines (Irvine and Gal 2000:50; cf. Fabian 1986, Meeuwis 1999). Another example of an extremely negative political outcome is presented by Hutton, who shows how an implicit equation between the vernacular and national identity underlay the “mother tongue fascism” of Nazi race science and its efforts to limit the use of German to “true” Germans (Hutton 1999 in Cameron 2007:278). Clearly, essentialising ideologies of language and identity are not new, and they are historically associated with concrete political consequences. Moreover, these consequences are highly specific to particular cultural and historical contexts. The examination of any ideology thus involves clarifying how, in a particular context, a given understanding of language is linked to the “defense of interest and power” (Woolard 1998:7) while alternative viewpoints are subtly downplayed. While the conception of an organic link between language and identity has undoubtedly been empowering in certain contexts, an examination of more recent incarnations of this ideology causes me to sound two notes of caution before proceeding with an analysis of its role in the Kénédougou context. First, in today’s context, this ideology is being given legitimacy through its association with “objective” linguistic theory and through its seductive promise to remedy Western angst about the homogenising tendencies of globalisation. In recent years, theorists of language ideology have uncovered the familiar ideological linkage between language and identity in current 202 Anicka Fast language endangerment discourses. Muehlmann points out that such discourses tend to define injustice or disempowerment in a limiting way, framing language endangerment in terms of a global threat to “biolinguistic diversity” which must “be ameliorated in a way that implicates and constrains the involvement of indigenous people by essentialising their relationship to both language and nature” (Muehlmann 2007:18). In this way, the persuasive force of the concept of “indigeneity” can legitimise new forms of external intervention (Muehlmann 2007:17) while limiting the possible options for response by requiring “indigenous” people to act indigenous in order for their rights to be protected (Muehlmann 2007:27). Second, recent advances in the understanding of language attitudes in West Africa suggest that such ideologies may be particularly inappropriate in this context of complex multilingualism. Owens notes that the linguistic practices of complex linguistic societies tend to keep language and ethnicity somewhat separate (Owens 2004:41), with language less likely to be politicised as an identity marker (Owens 2004:42). And as DombrowskyHahn and Slezak point out, theories that are insufficiently informed by the West African context generally fail to explain the continued vitality of local languages subject to increasingly “global influences” (Dombrowsky-Hahn and Slezak 2004:49). Ideologies that tightly link language with ethnic identity seem to inherently de-emphasise the shifting and constructed nature of linguistic identity, and to gloss over the alternative sources of legitimacy available both through LWCs (Swigart 2001:90) and through multilingual practices that function as a strategy to claim locally relevant power (Stroud 2004:90). 5.2.2 ‘Mother tongues’ and ‘heart languages’: The power of language ideologies in Kénédougou 5.2.2.1 Essentialising ideologies in Bible translation discourse While missionaries presumably have some access to essentialising language ideologies via historical and ongoing connections with the discipline of linguistics, and through the ubiquity of popularised discourses of language endangerment, I suggest that a main source of this ideology for many missionaries is the discourse of contemporary Bible translation organisations. Sanneh attempts to explain the recent massive expansion of the church in Africa as deriving at least partly from the Bible translation enterprise. He insists that despite their undeniable ethnocentrism, missionaries’ choice to translate into the vernacular resulted in the ultimate empowerment of African recipients of Bible translation to resist their colonial masters and to develop a Managing linguistic diversity in the church 203 theology on their own terms (Sanneh 1992b:3-5). According to Sanneh, the historical motivation for Bible translation was the belief that the gospel message is inherently translatable into any language. This belief results in the relativisation of all cultures and languages, since none is inherently more fit to convey the gospel message, while simultaneously destigmatising any culture in which the gospel is expressed, since that culture too is a sufficient channel for God’s message (Sanneh 1992b:1). However, further examination of Sanneh’s work reveals a strong tendency to supplement the basic idea of translatability with an essentialising ideology that draws a natural link between the vernacular and ethnic identity. The vernacular is described as having received an “autonomous, consecrated status” through translation (Sanneh 1992b:208), and is associated with indigeneity via reference to Africans’ “unique vernacular voice” (Sanneh 1992a:97). Sanneh’s appeal to this vernacularist ideology makes it all too easy to for him to downplay the adequacy of “non-native” languages as carriers of the gospel in an ironic contradiction of translatability. Moreover, since it is due to the missionary’s activity that African languages are “organised”, “stripped of foreign or gnostic conceits” and “made available” to all (Sanneh 1992a:101-102), the translator is made into the source of cultural renewal, judging what is “authentic” and disseminating it to the indigenous recipient. Promotional literature on Bible translation is replete with similar connections between language and identity, in popularised form.16 One of the organising concepts used here is that of the “heart language”, which I suggest functions as a “keyword” (Muehlmann 2007:15), simultaneously evoking three separate concepts: the first language, the language of superior comprehension, and the language that indexes authentic identity and spiritual experience. In this way, comprehension, spiritual authenticity and identity are assumed to proceed organically from the mother tongue. In an article featured on the website of Wycliffe International, an anonymous author uses this discourse to argue directly for the importance of mother tongue translation over translation into a regional lingua franca in the multilingual Nigerian context. The author draws a qualitative distinction between LWCs and the vernacular by emphasising that the “sound of the mother tongue in the ear and its meaning in the heart are trustworthy because 16 Although it is not possible to state with certitude the extent to which Sanneh introduced this ideology into Bible translation discourse and the extent to which it was already present, Sanneh’s work is clearly a major source of inspiration for Western Bible translators and missionaries; evidence of this is the frequency with which his work is quoted in The Bible Translator, the journal of the United Bible Societies (e.g., Wendland 2006: 215), and on Wycliffe Bible Translators country websites (e.g., WBT UK 2005, WBT 2006, Wycliffe International 2007). 204 Anicka Fast they are one with the person” (Wycliffe International 2007:n.p.), while “the second language doesn’t reach down into the understanding of their hearts.” This special role for the mother tongue and its equation, not just with superior comprehension but with authentic or true understanding, is possible because of the direct connection assumed to exist between the mother tongue and one’s “essential identity”. 5.2.2.2 Disempowerment Muehlmann (2007) suggests that one of the effects of essentialising ideologies of language endangerment is to obscure real processes of disempowerment by making “endangerment...the necessary cause of social marginalisation” (Muehlmann 2007:30). I suggest that language ideology functions in a similar way in this conflict. When the conflict is framed in terms of indigeneity and vernacular authenticity, missionaries become agents of cultural revitalisation, while church leaders’ role in theorising the identity of the church is either rhetorically erased (Irvine and Gal 2000:38) or re-cast as an obstacle to the development of the true indigenous church. In my view, this ideology thereby functions to mask at least two injustices that characterise the missionarychurch leader relationship. First, the injustice of the wealth differential between missionaries and church leaders is obscured. Stumpf (1977) argues that missionary wealth was operative in undermining local support for vernacular work in Cameroon, where the missionaries’ language was interpreted by local people as a “verbe puissant”: a source of access to material success in the Western sense (Stumpf 1977:203). Similarly, even while missionaries in Burkina Faso vigorously promote the value of mother tongues, their relative wealth, higher level of education, and knowledge of many languages communicates a commitment to pursue the resources available through LWCs such as English and French. Crucially, the assignment of a semi-consecrated status to the mother tongue helps to erase this internal contradiction from view by re-casting missionaries’ work as an aid to the self-actualisation and authentic spiritual experience of others. A second injustice that is obscured is the way that missionaries’ greater education and knowledge of English allows them to more easily legitimise their actions through appeal to scholarly research, theology, biblical texts, and Bible translation rhetoric. For example, one missionary referred to Dye’s (1985) academic research in the Central African Republic to support the view that “people who go to church but use a language that they’re not really that familiar with are more likely to be syncretistic in their faith.” Another referred to missiologist Donald MacGavran’s “homogenous unit” principle (MacGavran 1970) to emphasise the importance of keeping converts rooted in Managing linguistic diversity in the church 205 their cultural identity rather than encouraging them to abandon this cultural and family structure in order to become part of the church. Others emphasised Anabaptist distinctives, such as a concern for social justice, to imply that church leaders’ concerns were tainted by evangelical individualism. To state, as did one missionary, that the view of language loss as a justice issue is supported by Anabaptist theology makes it difficult for church leaders to contest this way of framing the linguistic situation. By appealing to discourses that have wide currency in academic and Christian circles, missionaries move the debate into an international sphere to which they have privileged access, and where Burkinabè church leaders are at a disadvantage. At least one church leader perceived the contradiction between missionary lifestyle and rhetoric as strongly disempowering. Accusing missionaries of enacting a “new colonialism” by keeping rural people backwards in their villages, he stated that: on the one hand, they say that they want to liberate people by helping them to speak their language, but at the same time they confine them in their environment, they shut them in … So, I say to myself that when this kind of project comes from elsewhere, to be imposed, this is a kind of colonialism. You’re keeping them shut up in their language, while you, you’re developing your knowledge by learning other languages … And when you come and confine people in their languages, you don’t have the possibility of teaching them about everything that’s happening in the world. You just limit yourselves to teaching them literacy, and all they need to do is read the Bible, that’s all, they don’t need anything else. No, I think that that, that’s colonialism … And it’s wrong. It’s wrong with respect to the way in which our world is evolving, such that in a decade, many languages won’t hold out. So it’s necessary to look at the trends and work with them in order to help people better. In sum, the results of my research illustrate the way that an essentialising language ideology can function to obscure the ways that languages other than the mother tongue are actually contributing to empowerment and inclusion, while subtly legitimising an ongoing power imbalance between missionaries and church leaders that prevents this ideology from being contested. While the Partnership Council structure is an honest attempt to ensure a more equitable power balance between church leaders and missionaries, it seems clear that access to resources of ideological legitimisation is still sufficiently 206 Anicka Fast differentiated to have a strong dampening effect on the voice of church leaders in the current debate. 5.3 Resources for power balancing Although many linguistic practices in church services are contested, areas of commonality between missionaries and church leaders certainly exist, including shared motivations for Bible translation, an agreement that mother tongue music should be encouraged in church services and that the comprehension of preaching in the mother tongue is superior, a shared concern for inclusion of the less powerful through language, and a shared excitement about the potential of the Partnership Council to ensure more collaboration in decision-making. However, I believe that more drastic power balancing is needed before participants can discuss their opposing language ideologies in a fruitful way. In this section, I propose resources from within a Christian and Anabaptist tradition shared with research participants, that may serve as a catalyst to power balancing between missionaries and church leaders as they continue to navigate this ongoing situation of tension. Despite the dependence of these recommendations on a moral framework which not all linguists share, I believe that they nevertheless belong in an academic document which purports to demonstrate the impossibility of ideological neutrality as much within the field of linguistics as within the church. First, drawing on the Anabaptist view that the church is a transformed, alternative society characterised by justice, reconciliation, forgiveness and economic sharing (eg. Yoder 1997:49) could support a determination not to use supposedly neutral “facts” about language as a cover for the misuse of power. It could also encourage a determination to make membership in this new polity more important than ethnic identity, so that language ceases to function as a fixed marker of identity. In addition, it could allow celebration of the many different ways that languages, including LWCs, can be used. Finally, it could encourage people to show concern for the languages of others, including their ‘ethnic’ languages, in a subversion of the logic that one is incapable of caring for any language but one’s ‘own’. Second, the historical Anabaptist view of the church as an interpretive community encourages access to biblical interpretation for all believers, not only those whose interpretation is underwritten by advanced study (Yoder 1997:92). Kampen Entz affirms this point as one of the historic strengths of Anabaptism relevant to cross-cultural collaboration (Kampen Entz 2001:5859). Emphasising this tradition could lead to more fruitful discussions between church leaders and missionaries about how biblical texts about language should be interpreted. Managing linguistic diversity in the church 207 Third, missionaries should make every effort to end practices of selfexclusion from local churches, which occur through de-prioritisation of Jula learning, the refusal of offers of inclusion, and the encouragement of linguistic practices that underwrite a definition of church of which they are ultimately not a part. As long as missionaries continue to position themselves as outsiders, they will not encourage the development of a shared definition of church. A desire not to have too much influence on the ‘indigenous’ church is understandable in many ways, yet I have shown that it tends to legitimise a view of church as ethnic. Finally, ‘expert knowledge’ should be disseminated more carefully. Academics and theorists of language must recognise their complicity in the propagation of ideologies that disempower speakers of ‘indigenous’ languages. As Cameron 2007 states, the real political consequences that can result from these ideologies are a matter for “critical reflection” (Cameron 2007:279). An admission that ‘expert discourse’ is inseparable from questions of power and interest (Cameron et al. 1997:155) and an invitation for its evaluation within a shared tradition of inquiry (MacIntyre 1990:60) is necessary to help relativise it as one voice among many. 6. Conclusions Through this research, I have attempted to demonstrate the relevance of quantitative methods of language attitude study for an understanding of conflicting language ideologies. My analysis of language attitude data did not assume a mirror-like association between language and social structure (Cameron 1997:60), but directly examined not only how participants were acting to “construe the intersection of language and human beings in the world” (Woolard 1998:3), but also who had the power to make such representations, and what the specific consequences were (Woolard 1998:27). The clarity of the data seems to confirm that it is possible, through relatively simple exercises, to access participants’ perceptions of language utility without resorting to deceptive techniques. In addition, clarifying my political and moral allegiances to this church community not only encouraged greater openness and trust among research participants, but also made my research rationally intelligible within a moral tradition of inquiry that is shared with participants. Moreover, if MacIntyre (1990) is correct, this adherence to a particular tradition is a prerequisite to the intelligibility of my conclusions outside that tradition as well (MacIntyre 1990:60). In order to gain a clearer picture of the sociolinguistic situation in southwestern Burkina Faso, several areas would benefit from further research. First, further studies of mother tongue fluency would help to clarify the relationship between Jula and mother tongue knowledge. Second, additional attitude 208 Anicka Fast statements relating to French would help in gaining a better understanding of the role of this language in the current sociolinguistic setting. Third, the role of Jula as a language of solidarity requires further clarification through both qualitative and quantitative methods. Fourth, asking more open-ended questions in village settings would yield data that could help to further illuminate the language ideologies held by villagers, which at this point remain somewhat unclear. Fifth, a more direct exploration of the “heart language” discourse and other ideologies that surfaced in the interviews could occur through further interviews. Sixth, participant observation in additional domains would be crucial in further relating language ideologies to actual practices. During my fieldwork, language ideologies I had heard all my life, especially regarding the need to translate into a people’s “heart language”, took on new resonance through their contrast with alternative emphases on inclusion and empowerment through the use of non-vernacular linguistic resources. Contestation over the political and moral significance of missionary support for the vernacular lies at the heart of an ongoing conflict to define what constitutes church. An essentialising ideology connecting language and identity contributes to a definition of church on ethnic terms, while an ideology of inclusion and solidarity insists on separating church from ethnic identity. However, the conflict between these ideologies does not occur on a level playing field. The missionary ideology is supported by Western linguistic scholarship and Bible translation discourse, and functions to maintain unequal access to resources for legitimisation. As an Christian working from within to uncover the potency and power of this ideology, I have to conclude that this state of affairs does not reflect the true character of the church. References “Partenariat l’Afrique” 2006. AIMM organisational document, February 2006 version. 13 pp. Adejunmobi, Moradewun. 2004. Vernacular palaver: Imaginations of the local and non-native languages in West Africa. Clevedon: Multilingual Matters. Bertsche, Jim. 2006. Un petit aperçu historique sur l’oeuvre de l’AIMM au Burkina Faso. Unpublished manuscript, 2 pp. Burkhardt, Fern. 2006. Church census shows 1.5 million members. Mennonite Weekly Review, 22 November 2006. Retrieved 14 May 2007 from http://mennoweekly.org/NOVEMBER/11-27-06/MWC11-27.html Cameron, Deborah. 1997. Demythologizing sociolinguistics. In Nicholas Coupland and Adam Jaworski (eds.) Sociolinguistics: A reader, 55-68. New York: St. Martin’s Press. Managing linguistic diversity in the church 209 Cameron, Deborah. 2007. Language endangerment and verbal hygiene: History, morality and politics. In Alexandre Duchêne and Monica Heller (eds.) Discourses of endangerment: Ideology and interest in the defence of languages, 268-285. London: Continuum. Cameron, Deborah, Elizabeth Frazer, Penelope Harvey, Ben Rampton and Kay Richardson. 1997. Ethics, advocacy and empowerment in researching language. In Nicholas Coupland and Adam Jaworski (eds.) Sociolinguistics: A reader, 145-162. New York: St. Martin’s Press. Djité, Paulin G. 1988. Language attitudes in Abidjan: Implications for language planning in the Ivory Coast. Ph.D. dissertation. Georgetown University. Dobrin, Lise. 2005. When our values conflict with theirs: Linguists and community empowerment in Melanesia. In Peter K. Austin (ed.) Language Documentation and Description, Vol. 3, 42-52. London: SOAS. Dobrin, Lise. 2009. From linguistic elicitation to eliciting the linguist: Lessons in community empowerment from Melanesia. Language. Dombrowsky-Hahn, Klaudia and Gabriele Slezak. 2004. Local vitality: On a complex linguistic situation in the West African town of Banfora (BF). In Peter Probst and Gert Spittler (eds.). Between resistance and expansion: explorations of local vitality in Africa, 49-68. Münster: Lit. Dye, T. Wayne. 1985. Bible translation strategy: An analysis of its spiritual impact. Dallas: SIL. Fabian, J. 1986. Language and colonial power: The appropriation of Swahili in the former Belgian Congo 1880-1938. Cambridge: Cambridge University Press. Fast, Anicka. 2007a. Moral incoherence in documentary linguistics: Theorising the interventionist aspect of the field. In Naomi Hilton (ed.) Proceedings of the 5th Cambridge Postgraduate Conference in Linguistics, Cambridge, March 20-21, 2007. Fast, Anicka. 2007b. Reclaiming the politics of language attitudes: Directions and prospects for language attitude activism in Africa. Unpublished manuscript, 40 pp. William A. Foley. 2005. Personhood and linguistic identity, purism and variation In Peter K. Austin (ed.) Language Documentation and , Vol 3, 157-180. London: SOAS. Hutton, C. 1999. Linguistics and the third reich: Mother-tongue fascism, race, and the science of language. London: Routledge. Irvine, Judith T. and Susan Gal. 2000. Language ideology and linguistic differentiation. In Paul V. Kroskrity (ed.) Regimes of language: Ideologies, polities, and identities, 35-84. Oxford: James Currey. Kampen Entz, Donna. 2001. From Kansas to Kénédougou... and back again. Missio Dei 9: 33-61. Retrieved 16 August 2007 from http://www.mennonitemission.net/ resources/Publications/MissioDei/Downloads/V3.pdf. 210 Anicka Fast Lambert, Wallace E., R. Hodgson, R. Gardner, and S. Fillenbaum. 1960. Evaluational reactions to spoken language. Journal of Abnormal and Social Psychology 60: 44-51. McGavran, Donald. 1970. Understanding church growth. Grand Rapids, Eerdmans. MacIntyre, Alasdair. 1990. Three rival versions of moral inquiry: Encyclopaedia, genealogy, and tradition. Being Gifford Lectures delivered in the University of Edinburgh in 1988. Notre Dame: University of Notre Dame Press. Mansour, Gerda. 1993. Multilingualism and nation building. Philadelphia: Multilingual Matters. Meeuwis, M. 1999. Flemish nationalism in the Belgian Congo versus Zairean anti-imperialism: Continuity and discontinuity in language ideological debates. In Jan Blommaert (ed.) Language Ideological Debates, 381424. Berlin: Mouton de Gruyter. Milroy, Lesley. 1987. Language and social networks. 2nd ed. Oxford: Basil Blackwell. Muehlmann, Shaylih. 2007. Defending diversity: Staking out a common global interest? In Alexandre Duchêne and Monica Heller (eds.) Discourses of endangerment: Ideology and interest in the defence of languages, 14-34. London, Continuum. Owens, Jonathan. 2004. Ethnolinguistic vitality theory and complex linguistic societies. In Peter Probst and Gert Spittler (eds.). Between resistance and expansion: explorations of local vitality in Africa, 35-48. Münster: Lit. Owens, Pamela Jean. 2006. Bible translation and language preservation: The politics of the nineteenth century Cherokee Bible translation projects. The Bible Translator 57 1): 1-10. Reh, Mechtild. 2004. Writing and reading in English and L1: Attitudes among pupils in Lira and Mpigi, Uganda. In Katrin Bromber and Birgit Smieja (eds.). Globalisation and African languages: Risks and benefits, 163180. Berlin: Mouton de Gruyter. Robinson, Clinton D. W. 1996. Language use in rural development: An African perspective. Berlin: Mouton de Gruyter. Sanneh, Lamin. 1992a. “They stooped to conquer”: Vernacular translation and the sociocultural factor. Research in African Literatures 23(1): 95-106. Sanneh, Lamin. 1992b. Translating the message: The missionary impact on culture. Orbis: Maryknoll. Sanneh, Lamin 2003. Whose religion is Christianity? The gospel beyond the West. Michigan: Eerdmans. Schieffelin, Bambi B., Kathryn A. Woolard, and Paul V. Kroskrity. (eds.) 1998. Language Ideologies: Practice and theory. Oxford: Oxford University Press. Showalter, Stuart. 2001. The same but different: Language use and attitudes in four communities of Burkina Faso. Dallas: SIL International. Managing linguistic diversity in the church 211 Showalter, Stuart. 2004. Un profil du bilinguisme en dioula au sud-ouest du Burkina Faso. Communication présentée au Vème Colloque interuniversitaire sur la co-existence des langues en Afrique de l'Ouest. Université de Ouagadougou, 27-29 septembre 2004. Retrieved 17 August 2007 from http://www.showalterfamily.net/downloads.htm. Showalter, Stuart. 2006. The Endangered Languages of Burkina Faso. Paper presented at the Georgetown University Roundtable on Endangered and Minority Languages and Language Varieties. Washington, DC, March 3-5, 2006. SIL 2004. Languages of Burkina Faso [map]. In Raymond G. Gordon, Jr. (ed.) Ethnologue: Languages of the world. 15th ed. Dallas: SIL International [online edition]. Retrieved 6 November 2006 from http://www.ethnologue.com/show_map.asp?name=BF&seq=10. Speed, Shannon. 2006. At the crossroads of human rights and anthropology: Toward a critically engaged activist research. American Anthropologist 108(1): 66-76. Spitulnik, Deborah. 1998. Mediating unity and diversity: The production of language ideologies in Zambian broadcasting. In Bambi B. Schieffelin, Kathryn A. Woolard, and Paul V. Kroskrity (eds.) Language ideologies: Practice and theory, 163-188. Oxford: Oxford University Press. Stroud, Christopher. 2004. Revisiting reversing language shift: African languages in high modernity. In Katrin Bromber and Birgit Smieja (eds.). Globalisation and African languages: Risks and benefits, 85102. Berlin: Mouton de Gruyter. Stumpf, Rudolf P. K. 1977. La politique linguistique au Cameroun de 1884 à 1960 (comparaison entre les administration coloniales allemande, française et britannique et du rôle joué par le sociétés missionnaires). PhD thesis, Paris. Swigart, Leigh. 2001. The limits of legitimacy: Language ideology and shift in contemporary Senegal. Journal of Linguistic Anthropology 10(1): 90-130. Tiendrebeogo, Gérard and Zakaria Yago. 1983. Situation des langues parlées en Haute-Volta : perspectives de leur utilisation pour l'enseignement et l'alphabétisation. Abidjan: Institut de Linguistique Appliquée. Traoré, Siaka. 1992. Planting the Mennonite church in Burkina Faso. AIMM Messenger 59: 3-4. Wendland, Ernst R. 2006. Theology and ministry in Africa through Bible translation: “How firm a foundation”? The Bible Translator 57(4): 206-216. Woods, David R. 1995. Attitudes toward French, national languages, and mother tongues across age and sex in Congo. In Akinbiyi Akinlabi (ed.) Theoretical approaches to African linguistics, 403-415. Trenton: Africa World Press. Woolard, Kathryn A. 1998. Introduction: Language ideology as a field of inquiry. In Bambi B. Schieffelin, Kathryn A. Woolard, and Paul V. Kroskrity (eds.) Language ideologies: Practice and theory, 3-50. Oxford: Oxford University Press. 212 Anicka Fast Wycliffe Bible Translators 2006. Why Bible Translation is Important. Retrieved 4 September 2007 from http://www.wycliffe.org/Explore/WhyBibleTranslationisImportant.aspx. Wycliffe Bible Translators UK 2005. Expect explosive growth in faith movements to Christ: Quotes from Christian leaders about Bible translation into the heart language. Retrieved 4 September 2007 from http://www.wycliffe.org.uk/News_06_QuotesaboutBibletranslation.ht ml. Wycliffe International 2007. Explore Bible translation in Africa. Retrieved 4 September 2007 from http://www.wycliffe.net/home/Africa/tabid/444/Default.aspx. Yoder, John Howard. 1997. For the nations: Essays evangelical and public. Eugene: Wipf and Stock.