Writing the self through the Other’s language? An ethnographic analysis of notebooks kept in rural Mali∗

“…in between is as much a place to be at home as any other”1 This paper focuses on one empirical result of an ethnographic research on personal notebooks kept by villagers in Mali. This research was part of a wider study on literacy practices (Mbodj-Pouye 2007). While working as an ethnographer on literacy practices ranging from agricultural records to shopping-lists, a common practice caught my attention: keeping a personal notebook for various notations2. These notebooks are multilingual, mixing Bamanan, French and Arabic. When focusing on those writers whose repertoires include Bamanan and French, it appears that both languages are used in distinctive ways, and that French is the dominant language for the more personal writings. The way such a result can be obtained from the analysis of texts which are highly multilingual will be discussed at length, as well as the extent to which the writings are personal. This paper aims at exploring the meanings of this unexpected language choice. The use of French for personal writings immediately recalls broader discussions on languages in postcolonial contexts. The question of language is a prominent issue in debates over the language of education. The dominant discourse is that literacy, whether in adult literacy classes or at school, is more quickly achieved by using the first language of the population. An underlying assumption is that writing in one’s own language provides a better chance for effective use. Another common assumption, of a more nationalistic tone, is that African languages are more suited to transmitting the cultural lore and to convey the intimate feelings. Here we come across an argument which has also been raised by writers defending their choice in favor of the vernacular in creative writing. In contrast with this strong advocacy for the written uses of African languages, the actual sociolinguistic situation remains a diglossia, with French, the only official language, dominating as the written language in formal settings. Even if Mali stands as one of the “French-speaking” countries where their promotion has been more sustained, through rural Aïssatou Mbodj-Pouye ([email protected]), Centre d’études africaines, EHESS, Paris. Paper presented at the African Studies Association Annual Meeting, Chicago, November 13th-16th, 2008. 1 (Cooper 1994: 1539). 2 I rely here on research data collected mainly in a Bamanan-speaking village located near the town of Fana. I have conducted interviews and observations, during which I have collected a corpus of texts by photographing them. My final corpus consists of 424 pages from notebooks kept by 23 villagers. ∗

Draft version - please do not quote

1

literacy and bilingual schooling, the uses of the African languages (defined as “national languages”) remain limited. The investigation into grassroots literacy practices proves interesting because it is an instance in multilingual practices where the ideological debates hardly peep through, thus providing an insight into the series of language choices ordinary people make in their daily writing practices. In the set of studies on what Karin Barber has termed “tin-trunk” literacy, stemming from anthropological and historical research in Africa, the issue of language choice for writing is touched upon (Barber 2007). For instance, Stephan Miescher in his analysis of the autobiography of Ghanaian catechist points to the fact that this text is written in English, with the exception of a section devoted to a list on witchcraft activities in Twi (Miescher 2006: 34). Language alternation is thus limited within the text, though it appears central for the oral performance, which is in Twi (ibid.: 44). Though most contributions to the same edited volume deal with texts in English, some focus on writing practices in other languages (Zulu, Hausa). However, the kind of written products analyzed, mainly letters and diaries, are in the form of texts designed as such, with strong issues of self-presentation. The choice of a language for writing is thus inseparable from the decision to write. I will deal here with much less-structured writings, where switches from one genre to another within the same notebook and often within the same page offer more space in an investigation on language choice and code-switching. The question mark my title ends with thus encompasses several layers of interrogation. A preliminary point is to assert how far the notebooks can be treated as “writings of the self”? Secondly, as the writings are multilingual, ascertaining that French is the dominating language is not straightforward. How is this empirical result obtained? How do biliterate writers handle the several languages that compose their written repertoire? Finally, what does it mean to write in French? To what extent is French still the language of “the Other” in a postcolonial setting?

1. Languages, genres, and themes My first point will be to provide an overview of the notebooks: the diversity of languages and scripts is intertwined with a diversity of genres and themes of writing.

a) The heterogeneity of a common practice My initial research interest focused on literacy practices in the cotton-growing area where functional literacy campaigns in Bamanan have been organized since the mid 1970s by the Draft version - please do not quote

2

CMDT (Compagnie Malienne pour le Développement des Textiles). The Company had its own interests in developing literacy classes as it was involved in a major organizational change, which implied delegating many tasks to local villager organizations. This process required that some villagers would be literate in Bamanan, and able to keep records and perform a series of accounting, measurements, reading and writing tasks. My point of departure was in line with many other ethnographic approaches following the New Literacy Studies framework: my idea was to understand how a population “takes hold of literacy”, assuming that the villagers, confronted with this functional training would also make their skills serve other purposes (Street 1993). I did not intend to focus on one specific practice; rather I tried to approach a wide range of domains of use. But while carrying out interviews, early in my research, I was presented with notebooks disclosing personal writings. I became more and more interested in this practice of keeping a notebook for one’s personal needs. The domains of use present in the notebooks range from farming (noting dates for sowing or treating cotton fields, meteorological data), business (keeping credits records), family and village life (dates of marriages, births and deaths; keeping accounts of contributions at ceremonies), religion (copying prayers), medicine (recording recipes). Writing in the notebook involves keeping track of eye-witnessed or local events as well as collecting information from other sources, oral - notably the radio for records of news, sport results, song titles - and written - by copying books or documents. A wide array of school, professional and administrative models are borrowed to write in the notebooks. These various uses intermingle in the notebook in ways that greatly differ from one writer to another. Given that there is no generic label for the personal notebook, this heterogeneity calls into question the possibility of treating the practice as one common practice. However, all the notebooks draw from the same limited repertoire of type of texts, which attributes a common aspect to them3. They also handle common problems: as most of them mix several of these types of text they must find ways of graphically delineating these units. In fact, it is the very heterogeneity of the practice which gives it some of its shared features. The main common point is that writers draw on conspicuous uses of exercise-books in school as well as within professional and institutional settings, but they are not given any model of personal writing as such. It is striking that the practice does not follow a model of journal3

I use “type of text” as a more general descriptive category than “genre”. Some of the types of texts that we find in the notebooks are identified genres, either oral (“kilisi”, magical incantation, spell) or written (addresses, history, memories), but not all (such as lists, accounts, notes written down from radio-listening…).

Draft version - please do not quote

3

writing: even when events are recorded, the writing does not follow a daily routine. Nevertheless, the distinct moments of writing might coincide with distinct textual units, providing the notebook with an entry-form style. In line with this idea of a practice emerging without following a model4, an important point is that very often the notebooks are not initially dedicated to personal writing. They might have been received during an agricultural training course to which literate villagers are often convened. On such occasions, the organizer gives out exercise-books to the audience. At the end of the training, each participant will have conscientiously filled in a part of the notebook by copying down the statements written on the blackboard by the teacher. This notebook will then be taken home with them. Another way of getting a notebook is to rescue a school exercise-book in order to write in the margins or in other spare space. Broadly redefined to include such cases as keeping a notebook on which personal notes are written down, the practice appears to be quite common: in the village where I conducted my research, most of the literate villagers I interviewed keep such notebooks. As my argument runs on the demonstration that the more personal writings are in French, locating within the writings what is “personal” appears a necessary preliminary task. As the foregoing description of the practice suggests, this is not straightforward, so I will only provide a few distinctions here, which help disentangle this issue of the “personal”. Among the different meanings the notion of “personal” uncovers, the following three can be distinguished. Firstly, it can lead to a survey of the topics of writing in search of private issues. As we know from historical scholarship, writing played a prominent part in the way the colonial power took hold of populations and individuals (Hawkins 2002). As a “technology of the individuals”, to quote from Michel Foucault (Foucault 2001), writing makes its impact through identification procedures from birth-registers to individual papers. Personal writings designed in response to these official writings appear in the notebooks, more or less close to these assigned models whose initial language is mainly French. However, appropriating these models in personal notebooks opens a choice between keeping French and translating. Secondly, when focusing on the conditions of transmission, the information can be more or less secret versus public. In oral uses, public utterances are strictly controlled on some 4

In this respect, my work addresses questions quite similar to the ones Liz Gunner poses when asking “How does one begin to keep a diary when no one has ever suggested the possibility? What cultural models assist the keeping of a “diary” or “journal” especially if these concepts do not, as yet, exist in your mother tongue?” (Gunner 2006: 155).

Draft version - please do not quote

4

specific kinds of knowledge, labeled as “secret”, gundo, with restricted forms of transmission and circulation5. Writing opens a vast new terrain for more individualized relationships to such bodies of knowledge, even if the risk of dissemination is taken seriously. Collecting magical incantations and medicinal recipes and writing them down in one’s notebook is one of the villagers’ favorite uses of writing. This practice has distinctive implications on language choice, as privacy can be sought by choosing a language which is not known in the closest family environment, but at the same time the value of the traditional or religious lore is often tied to its original linguistic form. Lastly, the “personal” can be located in the expressive ways of creating a space of one’s own. This goes along with specific practices: marking materially a notebook with signatures and initials; giving a title to the whole notebook if it is dedicated to personal writing, etc. We will have to question the role of languages in this expressive work.

b) Language repertoires In the village I have been working in, different ways of acquiring literacy are present, involving three languages: adult literacy classes organized by the CMDT are in Bamanan; schooling combines two languages, Bamanan and French, as one of the first bilingual schools of the country was opened in this village in 1979; Islamic learning implies the use of Arabic, oral and written, as well as few written and oral uses of Bamanan and French. Outside these educational settings, other experiences, especially when migrating for work in urban places provide occasions for learning specific uses of literacy. Thus the paths to literacy are plural, and the writers’ repertoire often encompasses different languages. Multiscriptuality is common too: Bamanan and French share the Latin alphabet; Arabic is used both in Arabic script and in Latin transcription. I must specify that I focus here mostly on literacy practices by Bamanan-French biliterates for different reasons including my own language proficiency6. These Bamanan-French biliterates are often former pupils of the bilingual school. Before the opening of the bilingual school, others have been trained at school in French only. They have later acquired literacy skills in Bamanan in literacy classes or in more informal ways, as writing down Bamanan in the Latin alphabet can seem easy when literate in French. This precision show that there are several language varieties, so the written repertoire must be determined as specific in regard to the oral one, not only in terms of languages included, but 5

On the notion of secrecy, see the collection of essays in (Jansen and Roth 2000). For an analysis on the ambiguities of what is known and talked about privately but still should never be mentioned in public, see particularly Julianne Freeman’s contribution to this series (Freeman 2000). 6 French is my first language, and I have learned Bamanan before fieldwork in Paris.

Draft version - please do not quote

5

also in terms of their varieties (Blommaert 2008: 68-72). As Bamanan is almost always added to French to the repertoire of schooled villagers (who count for the main bulk of the literates), only those who went only to the adult literacy class are monoliterate (though often with notions of Arabic). Biliteracy is thus common. As an oral language, French is seldom used in the village (apart from school and interactions with strangers). The writers, who have often migrated to neighboring cities, might have been exposed to code-switching with this language in urban use of Bamanan, as well as through the media. Although in Mali, French has not developed in such a specific vernacular variety as for instance in Ivory Coast, some writers might have been in contact with the Ivorian variety through migration to this country. However, the French code in their written repertoire appears to be close to a standard variety. Bamanan is used in an orthographic code which varies from the official orthography to a less codified transcription, usually backed up by writing habits in French. Arabic, when it appears in the writings of the Bamanan-French biliterates I study, is prominently used in Latin script, more or less close to the scientific transliteration. As this description clearly shows, naming the writers whose repertoire includes French and Bamanan “biliterates” does not mean that they are balanced biliterates. On the contrary, the multilingual character of the practice arises from the discrepancies between skills in the different languages. Moreover, the writers must be situated on the outskirts of broader “economies of literacy”, to borrow from Jan Blommaert (Blommaert 2008). Thus, one should bear in mind that language choice is strongly constrained by the limited skills in the different languages. But these constraints still leave the writers with a space where biliteracy stands as one of their expressive resources.

2. Distinct languages for separate uses Given this complex repertoire, how do notebook-writers handle these different languages in their writings?7 It is worth noting that all biliterates I worked with make use, albeit in some cases in a limited way, of both Bamanan and French in their writing practices. Most of them do so in their personal notebooks. In this respect, this practice is different from other practices, such as 7

I would like to acknowledge here how much these analyses rely on an ongoing collaboration with Cécile Van den Avenne, undertaken while I was writing my doctoral thesis. She has brought her linguistic skills to the description of my corpus. Our first presentation of the analyses developed here was a talk given at the Ecole Normale Supérieure in Lyon on the 6/04/2006. See (Mbodj-Pouye and Van den Avenne 2007; Mbodj-Pouye and Van den Avenne in press).

Draft version - please do not quote

6

letter-writing, where the language choice is made once and for all at the beginning. Here language choices are continuously made throughout the writing of the notebook8. When trying to make sense of the switch from one language to another in this writing practice, the first difficulty is to determine a level of analysis: code-switching occurs within the different units and between them. Of course, a precise description requires that all those levels be handled at the same time, because the notebooks are often multilingual both at intraunit level and between them. However, provided that an overall identification of the language of a unit is possible, the latter level appears as the easiest site to look for motivations for codeswitching: how do writers choose from their written repertoire the language to write according to the genre or domain of writing? Some of the domains require the use of a specific language. Notably, Bamanan is the main language used in writing down magical incantations and medicinal recipes, even in notebooks whose writers try and conform to writing in French. For the spells, this language choice seems linked to the necessity to retain its original form, as the materiality of sound is essential to the efficiency of the utterance. French does appear in other kinds of notations, designed on the model of the recipe - which, as Jack Goody shows, is a highly graphic form (Goody 1977). However, even in those texts, the names of trees and herbs remain in Bamanan. Here, the ignorance of the French can account for the absence of translation: the writers simply do not know the French word, if it exists, nor the scientific name. In this specific domain, Bamanan also has a specific culturally-loaded value. “Bamanan” refers to a language, but also to cultural values and attitudes, whose core is a set of magical practices, which borrow from Islam, but are still thought of as specific9. Another domain closely linked to a language is Islamic knowledge, even if in this domain translation occurs quite frequently. Here too, a cultural association as well as a strong link with the original linguistic form can account for this language choice. French appears systematically for texts bearing an official character, such as addresses or copies of documents. In these cases, the models are originally in French. This overview makes it clear that biliterates gain from being able to write in two languages, and make use of their multilingual written repertoire. This result counters common views on literacy in national languages which assume that it is useless as soon as literacy in French is 8

Multilingual literacies have often been treated as implying the association of each language or script to one specific set of uses. See, for a classical approach of such a multiliterate setting (Scribner and Cole 1981), and more recently (Martin-Jones and Jones 2000). 9 On the continuum between “bamanan” and “mori”, see the classical article by Jean Bazin (Bazin 1985). Catherine Barrière study of magical practices in the region of Segu further explores this double reference (Barrière 1999).

Draft version - please do not quote

7

acquired. However, this approach explains some typical choices but not the constant alternation from one language to another that characterizes the notebooks. It does not account for the fact that very common genres, such as accounting, farming information, records of events, are written in both languages, but mainly in French, even when they do not follow any model of official writing. Furthermore, a methodological pitfall of this approach is that it opens up explanations for language choice that draw upon distinct criteria: a “cultural” association between one language and a domain of writing, an incidence of the context of writing, a tendency to keep the “original” language… The main trouble with this analysis is that it does not pay attention to the specificity of the practice, which lies in the association of genres and topics. A pragmatic approach seems the only way out, allowing us to understand how the multilingual character of the notebook is part of its fabric. Languages are not selectively drawn from a compartmentalized repertoire to meet specific needs: on the contrary, the writers make sense of this plurality in creative ways.

3. Making sense out of the plurality of languages Handling the plurality of languages gives way to distinctive acts, such as quoting and translating. A pragmatic approach helps to identify broader strategies of managing the plurality of languages. At this point, the level of analysis should not be the textual unit, but the notebook as a whole. In the notebook, the units are often framed and introduced, either by graphic devices (the spatial layout of the notebook or of the page) or discursive devices (titles, commentaries…). These are of great interest in exploring language choice within multilingual texts.

a) Quoting from: transcribing and copying down Quoting is a central dimension of oral communication (Penfield 1983). In written form, it appears through specific acts: transcribing and copying down. At a first glance, quoting can be seen as an act which does not imply any language choice, as it imposes keeping the original form. However, attention given to the ways the writer indicates or not this act reveals a more complex situation. Code-switching appears as one feature among others that highlight the quotation. Other devices include quotation marks, graphic layout, change of color of ink, and discursive markers. All these are, however, far from being consistently used throughout my corpus, and language alternation is the most constant marker in biliterates’ writings. Draft version - please do not quote

8

Let us survey the uses of quotation in Demba Camara’s notebook. Demba Camara is a retired employee from the CMDT, where he used to be an instructor. He went to school for 5 years, starting in 1944, and learnt to read and write in French. After being recruited by the CMDT, he was, as a teacher, involved in the literacy classes during the mid 1970s, where he acquired written skills in Bamanan. This notebook is his current personal notebook, dedicated to his own purposes, mainly religious notations. If we focus on the use of the three languages, we can identify the following uses, displayed for example in the following double page (doc. 1).

Doc. 1 Double page from Demba Camara’s notebook

Arabic is used only for reported speech, for excerpts from Koranic texts that he copied from several books. Interestingly enough, the print material is already multilingual and multiscriptual, including translations from Arabic into Bamanan and French, and transcriptions into the Latin alphabet10.

10

One of the most common reading material in the village are Islamic booklets, which have been studies by Francesco Zappa (Zappa 2004).

Draft version - please do not quote

9

Bamanan is used, in textual form, only for the incantations, which are also reported speech. In one case, for a list of contributions to a ceremony, Demba makes use of Bamanan for single words and syntagmas. Code-switching to Bamanan in French also occurs frequently. French is the language of titles and indications used both for both prayers in Arabic and incantations in Bamanan. It is also used extensively for an address (untitled as such) and a “contract” with a Fula herdsman (doc. 2) - this last piece is interesting as it also involves a specific form of translation to which we shall return. Other small notations are also in French. Though far from quantitatively dominant, French appears as the “linguistic frame reference” of this notebook (McLaughin 2001: 166). Arabic and Bamanan appear in quotations that are framed by titles and indications for use. The titles’ language ascription is not easy, given that borrowings are common. Arabic prayers’ titles draw back to a literate tradition: for instance, page 7 “Roudi” is a Bamanan borrowing from Arabic - the original form is “k’a wurudi”, to tell one’s beads. Incantations in Bamanan, collected through face-to-face interaction, are also generically untitled by the phrase borrowed from Arabic “bismillâhi”, the opening word of the utterance11. For incantations as well as for prayers, indications for use are always in French: “Récité dans l’huile”, “(à réciter 129) fois” etc. The only exception is for the second incantation, where the indication of the medicine to be used begins in Bamanan, but it is the name of a tree, which, as previously discussed, is a lexical domain where the use of Bamanan is common. The switch between Bamanan and French is visually marked, as Demba follow a convention of writing Bamanan in script, and in lower case, whereas French uses the standard handwritten cursive, with capitalization when needed. This pattern, where French assumes the role of the metalanguage (the language which introduces the text) and the language for prescriptions, vis-à-vis quotations that borrow from Arabic and Bamanan, is shared by the different notebook-writers12.

b) Translating as “creative writing” Translation appears in the notebooks in one very constrained style, the lexical list. Interestingly, whereas some writers only produce copies from lexicons, others borrow from this model to assist their informal learning of a new language. In this rigid form (the lists are 11

Demba’s orthography is not stable enough to provide solid cues on his language choice, even though the titles seem to rely more on French orthographic conventions than the inner text. 12 The way Bamanan and Arabic are respectively dealt with is not exactly the same though. Copying and transcribing are different acts, and involve distinct processes of “entextualization”. For Arabic, the texts are drawn from a well-established literate tradition, which provides meaningful titles, which is not the case for Bamanan incantations. Those are transcribed after face-to-face transmission, and the notebooks often traces the text to the person who delivered it to the writer, whereas the written sources of the Islamic texts are not indicated, though Demba orally specifies where he copied each text from.

Draft version - please do not quote

10

usually displayed in two columns, each in one language, with the opposite word being the translation of one other), the variations are limited. This appropriation of the lexicon model suggests that, in this highly multilingual society, the uses of written translation are extended outside school settings. As Derek Peterson remarks, “in central Kenya as elsewhere in the colonized world, translation was always more than a scholarly project. Translating was a popular activity (…)” (Peterson 2004: 9). Indeed, translation appears fully as a creative work in more complex sets of practices, where it is associated to ways of articulating not only languages but also cultural systems of reference13.

Doc. 2. A contract-like notation for the work of a herdsman

A good example is the common practice of translating dates and providing corresponding terms in another calendar, which appears in Demba Camara’s notebook in the “contract” quoted above. It consists in giving the date in two forms: in French, in the Gregorian calendar, which is the civil calendar in use in Mali, and, in Bamanan, in the moon-based calendar with Bamanan names, which follows nowadays the Islamic calendar14. The act of translating is here involved in a process of putting together distinct systems of reference. When inserted in the birth-certificate model, this practice gives to the local calendar a space that it does not have in official writings, even if the use of Bamanan maintains a distance.

c) Uses and avoidance of the mix Considering the notebooks as a whole, we can now investigate the writing strategies of the writers. Translation and quotations often reveal an unequal treatment of languages. In fact, as 13

Penelope Papailias addresses similar questions on the meaning of code-switching and transcription in (Papailias 2005) chap. 5 “America Translated in a Migrant’s Memoir”, p. 179-225. 14 The names of the month in this calendar are season-based, which testify to a distinct use in the past. According to Maurice Delafosse, the concordance took place in the 19th or early 20th century (Delafosse 1921).

Draft version - please do not quote

11

Jan Blommaert notes, code-mixing can paradoxically provide evidence for the monoglossic character of a text, understood as the “conscious attempt (…) to produce ‘pure’ and ‘standard’ language” (Blommaert 2008: 68). In this view, the graphic devices which are used to flag the code-switching, such as quotation marks, brackets, uses of the margins, appear when Bamanan is introduced in a monoglossic French text, and not in the reverse case (Mbodj-Pouye and Van den Avenne 2007). This indicates not only that French is perceived by the writer as the standard, but also that for French, such a requirement (writing in the standard) is meaningful, whereas it is not for Bamanan, where code-switching occurs frequently without being indicated at all. At the level of the notebook too, titles appear as a crucial site for understanding the practice. In the notebooks dedicated to personal writing, the few titles are in French, though other opening forms are used, such as the usual opening form “bismillâhi”. This determination of the meta-language of the notebooks as French needs to be complemented by a specific investigation on the enunciative features of the texts in the notebook. As the initial description of the practice readily suggests, searching for instances of the first-person proves rather deceptive. However, we find some texts using the pronoun “I” (or the personal pronoun “me”)15. In Bamanan, these are spells, where the space for the enunciator is provided, often with an apposed surname, for instance: “me, Hinda”. Interestingly however, we encounter the form “ne, karisa”, “me, so-and-so”, where the first person enunciation is somewhat depersonalized, rather providing an empty space for any possible enunciator. On the contrary, in French, the first-person texts refer to the individual writer, even if it does not disclose an intimate self, as the “history” by Moussa Camara shows. In this case, the writer is in a situation similar to the one Jan Blommaert describes as a struggle with established genres that characterizes the two texts written in Congo that he analyzes (Blommaert 2008). The monoglossic choice for French is patent when this notebook is compared to the other notebooks Moussa Camara keeps: he records the weight of cotton collected by each family member during harvest on a small hand-cut notebook, written in Bamanan; another notebook displays notes copied down during a training session on the organization of a parents’ association, written in Bamanan. The use of French sets his personal notebook apart. This kind of text is what appears as the most “personal” writing in the notebooks. It is personal in that it deals with private and individual matters, and also in that it involves a specific expressive work.

15

This analysis is restricted to texts in Bamanan and French.

Draft version - please do not quote

12

Thus, French appears not only as the language framing the writing practice, but also as the language in which self-writing emerges. Contrary to a quantitative survey, which leaves the determination of a “dominant” language quite uncertain, this pragmatic approach provides us with a result that we now need to account for.

d) Why French? I will draw here on the analyses of the notebooks as well as on excerpts from the interviews I conducted with some villagers16. These interviews confirm the overall privilege given to French. The main argument is that writing French is easier (“tubabukan ka nogo, tubabukan ka teli ne bolo”), as Baïné Coulibaly, schooled for 7 years (complete curriculum in bilingual primary school plus one year), states. AM Est-ce que tu as l’habitude d’écrire des lettres? BC L!tiri ? N y!r! ta ? N b’o s!b!n. AM A ka ca bamanankan na ni tubabukan na ye ? BC A ka ca tubabukan na, hali n’a ma &a, a ka ca o de la, bamanankan n t! se k’o s!b!n k’a b!! &!nab'. Translation (from French and Bamanan). AM Do you often write letters ? BC Letters? My own ones? Yes I do. AM Is it mostly in French or in Bamanan? BC Mostly in French, even if it is not well-written, most of them are [in French]. Bamanan, I can’t deal with everything when writing it. This statement must be taken seriously, even from villagers whose proficiency in French is very limited. Writing is as much about language proficiency as about writing habits in a given language. In this respect, a six year-long schooling, either in French or bilingual but with a strong orientation toward French, leaves a stronger imprint than a few months of literacy class in Bamanan17. Some specific writing habits, such as the essay form uses of “I”, are only developed in French. Therefore, genres such as “history” are more easily performed in French. Writing, to a large extent, still equates writing in French as long as high-valued forms are concerned. As we have seen, a care for writing in a standard variety applies prominently to French18.

16

I rely on the analysis of the transcript of 56 semi-structured interviews conducted in this village, mostly in Bamanan, complemented by a few others in other places in the CMDT region. 17 For an assessment of the results of the type of bilingual school writers have been confronted with, see MbodjPouye and Van den Avenne, in press. 18 In oral expression, a sense of “proper Bamanan” can be displayed in specific settings (for instance radio games testing auditors for their ability to avoid French borrowings). For written Bamanan, an attention to the orthographic code may appear, but far less strongly than for French.

Draft version - please do not quote

13

French also draws some of its value from the fact that it is less known, at least in rural settings19. This argument is two-folded: the use of French can be strategic, as a means not to be understood by the immediate family environment; but it is also distinctive in Bourdieu’s sense (Bourdieu 1979), setting the writer apart from the non-educated, even if literate in Bamanan, villagers. Tahirou Coulibaly expresses this idea quite clearly. He begins by stating that his own writings are in French whereas when he writes for others he may use Bamanan: Ne ka s!b!nni fanba, bon, ni m'g' ta t!, ne b’a s!b!n tubabukan na. Translation (from Bamanan). The bulk my writings, well, if it is not for others, I write it in French. Then he further elaborates on this point: N y!r! b! se ka bamanankan s!b!n n y!r! ; bamanankan, ne b! se ka l!tiri ci m'g' ma bamanankan na, mais ... tubabukan de ka di ne ye ka t!m! a kan quoi. Sisan n b! se ka l!tiri s!b!n, k’a ci m'g'w ma bamanankan na ; n b! se k’a kalan n y!r! ye bamanankan na, mais ne ka branche fanba b! tubabukan de la quoi. Translation (from Bamanan). I know how to write in Bamanan myself. I can send letters in Bamanan, but… I really prefer French. Of course I can write a letter in Bamanan, and send it to others, I can also read it myself, but my main domain [litt. “branch”] is French. Of course, the higher status of the official language endorses it with a specific value. However, we should be cautious not to reduce this kind of statement to a simple adhesion to a shared language hierarchy. Ndiamba “prefers” French, and expressions of “liking” and “inclination” for this language are a recurring motive in the interviews. I suggest that French has two specific values for the writers. Firstly, writing in French, especially in a personal notebook, does not mean fitting into a language order that in any case remains remote. It means taking hold of it, making it serve other purposes. In that respect, French is used in the same way as many other features that convey meanings of officialdom, in a conscious attempt to deviate them from its usual meanings. As Karin Barber notes, “the diary is (…) a site in which Boakye Yiadom (…) stiffens up the outlines of his self with the carapace of officialdom” (Barber 2007: 198). Thus, writing in French is a way to appropriate not only writing, but also French, in experimental

19

The opposite argument is made when dealing with letters: here too French appears to be preferred language, especially for letters sent in towns, on the ground that it is easier to find someone able to decipher French than Bamanan outside the village.

Draft version - please do not quote

14

and personal uses that are not offered by oral communication, where people are more reticent to expose the limits of their proficiency. Secondly, the fact that French is a distant language, in a sense a borrowed one, besides its cryptic value in a village setting where it is not the most common written language, gives it a special meaning. The set of cultural values and attitudes associated with French is especially important for the young villagers who longs to belong in the emerging urban and youth culture (this is a role which can be endorsed by English in urban settings). In a sense, the detour of this distant language is something that can account for the choice to write in French when experimenting self-writing.

Conclusion I would like to highlight the two main results of this investigation. Firstly, the fact that French appears as the main language for personal writing is not so surprising as long as we consider the models for writing that are available. In this respect, my work provides an assessment of the ambiguities of the current language policy in Mali. On the one hand it improves literacy in Bamanan, which is effectively used. On the other hand, these uses are limited. Even bilingual schooling still conveys an implicit hierarchy of languages where French dominates, thus fostering a tendency to write in this language. This reflects a wider sociolinguistic situation where French remains the only official language. Above all, written models in Bamanan remain scarce and limited, especially when it comes to creative works. However, taking hold of writing opens a new space for appropriating French, for villagers who seldom have any occasion to speak this language. This study thus testifies to the complexity of language choice in multilingual settings, in line with sociolinguistic works that question simplistic assumptions that expressing the self would be easier in the vernacular (Sebba and Wooton 1998). Secondly, even with scarce models and under constraints, the writers make expressive uses of language diversity. Handling several languages offer a space for practices of translation and other forms of association of languages where the distinct spaces of reference the writers belong to can be rearranged or at least put together. Gifted with an unbalanced written repertoire, the writers are often not fully equipped to make language choices on the basis of arguments. The practice shows, however, the agency of these grassroots actors, who are able to create for themselves unexpected ways of expression, sometimes disentangling assumed links, and thus displaying the plurality of “African modes of self-writing” to borrow from Achille Mbembe (Mbembe 2002). Draft version - please do not quote

15

References. BARRIÈRE Catherine (1999) “Techniques d'agression magique en pays bamana (région de Segu, Mali) : emprunts réciproques entre islam et religion traditionnelle.” Journal des africanistes 69(1) : 177-197. BARBER Karin ed. (2006) Africa's Hidden Histories: Everyday Literacy and Making the Self, Bloomington, Indiana University Press. BARBER Karin (2007) The Anthropology of Texts, Persons and Publics. Oral and written culture in Africa and beyond, Cambridge, Cambridge University Press. BAZIN Jean (1985) A chacun son Bambara. Au cœur de l’ethnie. Ethnie, tribalisme et Etat en Afrique. J.-L. AMSELLE and E. M’BOKOLO, eds., Paris, La Découverte. BLOMMAERT Jan (2008) Grassroots Literacy. Writing, Identity and Voice in Central Africa. London, Routledge. BOURDIEU Pierre (1979) La distinction. Critique sociale du jugement, Paris, Minuit. COOPER Frederick (1994) “Conflict and Connection: Rethinking Colonial African History”, American Historical Review 99-5: 1516-45. DELAFOSSE Maurice (1921) “L'année agricole et le calendrier soudanais” L'Anthropologie tome 31 : 105-113. FOUCAULT Michel (2001) “La technologie politique des individus” in Dits et écrits II, 1976-1988, Paris, Quarto Gallimard: 1632-1647. FREEMAN Julianne E. (2000) “Knowledge, Secrecy and the Practice of Senior Womanhood Among the Bamana of the Bèlèdugu (Mali)” Mande Studies 2: 115-127. GOODY Jack (1977) The Domestication of the Savage Mind Cambridge, Cambridge University Press. GUNNER Elizabeth (2006) Keeping a Diary of Visions: Lazarus Phelalasekhaya Maphumulo and the Edendale Congregation of AmaNazaretha. Africa's Hidden Histories: Everyday Literacy and Making the Self. K. BARBER, ed., Bloomington, Indiana University Press: 155-179. HAWKINS Sean (2002) Writing and colonialism in northern Ghana : the encounter between the LoDagaa and "The world on paper", 1892-1991 Toronto, University of Toronto Press. JANSEN Jan and Molly ROTH (2000) Secrets and Lies in the Mande World. Mande Studies. 2. MARTIN-JONES Marilyn and JONES Kathryn, eds (2000). Multilingual literacies: reading and writing different worlds, Amsterdam, John Benjamins. MBEMBE Achille (2002) “African Modes of Self-Writing” Public Culture 14 (1): 239-273. MBODJ-POUYE Aïssatou (2007) Des cahiers au village. Socialisations à l'écrit et pratiques d'écriture dans la région cotonnière du sud du Mali, Doctorat de sociologie et anthropologie, Université Lumière-Lyon 2. MBODJ-POUYE Aïssatou and Cécile VAN DEN AVENNE (2007) “"C'est bambara et français mélangés". Analyser des écrits plurilingues à partir du cas de cahiers villageois recueillis au Mali” Langage et société n°120 : 99-127.

Draft version - please do not quote

16

MBODJ-POUYE A. and C. VAN DEN AVENNE (in press) Comment les langues se mélangent-elles à l'écrit ? Pratiques actuelles de deux agriculteurs passés par une école bilingue (francobambara) au Mali, in Languages and Education in Africa. A Comparative and Transdisciplinary Analysis. B. BROCK-UTNE and I. SKATTUM, eds. Oxford, Symposium Books. McLAUGHLIN Fiona (2001) “Dakar Wolof and the Configuration of an Urban Identity” Journal of African Cultural Studies 14 (2): 153-172. MIESCHER Stephan (2006) "My Own Life": A. K. Boakye Yiadom's Autobiography. The Writing and Subjectivity of a Ghanaian Teacher-Catechist in Africa's Hidden Histories: Everyday Literacy and Making the Self. K. BARBER., ed. Bloomington, Indiana University Press: 2751. PAPAILIAS Penelope (2005) Genres of Recollection: Archival Poetics and Modern Greece New York, Palgrave Macmillan. PENFIELD Joyce (1983) Communicating with Quotes. The Igbo Case Westport, Greenwood Press. PETERSON Derek R. (2004) Creative writing: translation, bookkeeping, and the work of imagination in colonial Kenya. Portsmouth, Heinemann. SCRIBNER Sylvia and Michael COLE (1981) The Psychology of Literacy. Cambridge, Massachusetts, Harvard University Press. SEBBA M. and T. WOOTON (1998) “We, they and identity. Sequential versus identity-related explanation in code-switching” in Code-switching in conversation: language, interaction and identity. P. AUER, ed. London, Routledge: 262-286. STREET Brian ed. (1993) Cross-cultural approaches to literacy, Cambridge Studies in Oral and Literate Culture, Cambridge, Cambridge University Press. ZAPPA Francesco (2004) L’islamizzazione della lingua bambara in Mali. Tra pubblicistica scritta e letteratura orale Supplem. n° 2 Rivista degli Studi Orientali. Istituti Editoriali e Poligrafici Internazionali, Pisa/Roma, LXXVII: 119.

Draft version - please do not quote

17