Oliver Stegen

Despite increased efforts by linguists and educationalists to facilitate literacy and literature development in minority languages, there are still many languages worldwide which do not have a written form. One area that needs attention in literature production for a newly written language is the question of writing style. As the features of good style are language-specific, writing style guidelines have to be developed for each language anew. It has been assumed that such vernacular writing style develops predominantly by mother tongue speaker intuition. However, very few studies have been carried out to verify this.
This research is set within the confines of the literacy project in the Rangi language of Northern Tanzania. As a contribution to the development of a natural writing style in Rangi, this research investigates what evidence for stylistic preferences can be found in texts that were produced by Rangi authors writing in their mother tongue for the first time. The main data of this study are 112 texts which were collected during a one-day writers workshop conducted between May 2005 and January 2006 in four different locations.
One way of observing stylistic preferences is through analysing the changes which authors make in successive versions of their text. Of the 112 texts in the database, 71 display stylistic changes between draft and revised versions. These texts are then investigated in more detail, e.g. with regard to text length, lexical density and story components. The subsequent comparative analysis of draft version versus revised version of each text operates at three levels: narrative elements at the text level, lexical choice at the word level, and word order, tense-aspect verb forms and participant reference at the clause level. At all three levels, stylistic conventions could be identified, e.g. formulaic introductions and codas, elimination of Swahili loanwords, or certain tense-aspect usages.
Despite such commonalities, this research suggests that, far from developing intuitively, vernacular writing style is influenced by a variety of factors, not least by previously available literature in languages of wider communication or in the target language itself. Among the concluding recommendations of this study for future vernacular writers workshops is the advice to employ guided editing which encourages multiple drafting and treats the different levels of editing separately, i.e. story structure, lexical choice and grammatical features.

Abstract: The present article describes three forms to express the verb 'to be'in Rangi, a Tanzanian Bantu language. Particular attention is paid to their functions in sentences and texts. In summary, it is shown how each form has developed in a distinct direction, either semantically or syntactically.

The research described in this paper is embedded in the Rangi language development project in Tanzania, and specifically in Rangi literature production. From a literature review, the areas of text and sentence length, clausal organization, vocabulary elaboration, and participant reference are identified as worthwhile to be investigated concerning the difference between oral and written style. Two traditional Rangi stories, recorded both in an oral and a written version and then edited into a joined story, are then ...

The research described in this paper is embedded in the Rangi language development project in Tanzania, and specifically in Rangi literature production. From a literature review, the areas of text and sentence length, clausal organization, vocabulary elaboration, and participant reference are identified as worthwhile to be investigated concerning the difference between oral and written style. Two traditional Rangi stories, recorded both in an oral and a written version and then edited into a joined story, are then analysed. For length and participant reference, the previous findings are confirmed. With regard to clausal organization and vocabulary elaboration, the Rangi written texts do not yet exhibit specifically written stylistic features as Rangi literacy is only recently emerging. In editing both versions into a publishable form, the editors have employed both written and oral stylistic features, thus enriching the simple original writing, which had been influenced by L2 writing in Swahili. Further investigation is suggested in the areas of audience feedback and identification of genre-specific stylistic features in Rangi.

The main object of research described in this paper is Rangi, a scarcely investigated Bantu language of Northern Central Tanzania. Rangi phonology and morphonology are briefly sketched, including a classification with regard to both Vowel Height Harmony and Advanced Tongue Root activity. The main body of the paper consists of a detailed description of Rangi derivational processes, which follows the pattern established in Maganga & Schadeberg’s 1992 description of Nyamwezi, a closely related language. Both verbal derivation which exclusively uses the extensional slot of the verb structure, and nominal derivation which employs noun class prefixes and a few suffixes are covered.

In this paper, the results of an investigation into the status quo of writing in Rangi society, a vernacular language group in Northern Central Tanzania, are reported. This investigation includes a sociolinguistic questionnaire in which respondents comment on their use of different genres, and a study of both official and personal letters. The main findings emphasize the importance of personal letters in Rangi society, mainly written in Swahili, Tanzania’s national language, but also a surprising number reported as written in Rangi. Writing styles and conventions are primarily shaped by those of Swahili, given that this is the language in which all Rangi writers have learned to write.
As literature production both in Swahili and Rangi is a desired outcome of the Rangi language development project, further research is suggested into what the Rangi consider ‘good style’ for each relevant genre as well as whether and how vernacular writing skills should be taught.

While the United Nations guarantee the right to use minority languages freely in private and in public, this right is in reality restricted by several circumstances, one of them being the fact that many minority languages do not have a body of written literature, or even an orthography, thus restricting language use to the oral medium. In this paper, the situation of one particular minority language is presented: Rangi of Northern Central Tanzania. After sketching the brief history of the Rangi language development project which started in 1996, the focus rests on the key factors of language vitality, language attitude, and project ownership. For the first two, a positive development is observed, and consequently, Kutsch Lojenga’s (1996) participatory research approach is recommended as suitable project approach. The paper ends with an outlook on avenues for further research and possible future action, suggesting an increase in sociolinguistic investigation as well as an in-depth discourse analysis.

Abstract This report is organised as follows. Section 2 is devoted to a description of the field methods which were employed. Section 3 presents our primary findings relating to tone in the associative construction. The final two sections are more speculative. Here, we give some pointers to future analysis work. In particular, we argue that the four-way tone distinction claimed for Bamileke Dschang is not present on all syllable types, and that there is a syllable length distinction which must play a central role in future empirical and ...

This conference presentation (46th CALL at Leiden, 29-31 August 2016) traces the representation of Zanaki vowels from 1948 to 2014 as well as relating a phonetic study based on recordings done in December 2014. Speaker variation points to the Zanaki vowel system being between a 7-vowel system with +ATR harmony, and a 10-vowel system with symmetric ATR harmony.

The study described in this paper grew out of a particular concern in a project of translation and literature development for the Rangi language. It had been observed that the project’s Rangi translators, when encountering rhetorical questions (RQs) in the source text, often translated these as RQs into Rangi, claiming that RQs were both natural in Rangi and common in all contexts. To verify that claim, an investigation of RQs in the project corpus was undertaken which includes a primer, a story booklet, 71 texts collected during writer’s workshops in 2005/2006, and 15 texts collected for a discourse workshop in 2010.
A survey of the literature on RQs reveals a gap that should be filled: RQs are often defined only negatively as “question[s] for which the speaker does not request an answer from the addressee” (Hackstein 2004: 167), a definition whereby various functionally quite dissimilar phenomena are joined into a single category. Most existent typologies of RQs do not seem to comprehensively differentiate RQ functions. Correspondingly, studies of RQs in Bantu languages (e.g. Zerbian 2006) also do not sufficiently distinguish the functions of RQs. This paper aims to contribute both to a functional typology of RQs and to the application of such a typology to Bantu languages.
In addition to a primary distinction between RQs with discourse functions and RQs with social functions (Hackstein 2004), all RQs in the Rangi corpus have been categorised with regard to medium, length and genre of the text in which they occur, speaker and addressee(s), and question form and question words used.
RQs are indeed not an infrequent feature in Rangi; 37 of 106 texts contain at least one RQ. However, the functional range of RQs found in the corpus is not as broad as claimed by the Rangi translators in general. The absolute majority of RQs occurring in dialogue expresses rebuke (cf. example 1) whereas most RQs occurring in monologue express uncertainty or doubt (cf. example 2).
1. sà tɕɛ́ ʊ̀lʊ́ːgʊ́ɾî:ɾʲɛ̀ íbǎːⁿdɛ̀ ráːnɪ́
for what 2sg:burn:PRF:CAUS 5:grasshopper 5:1sg:POSS
‘Why have you burned my grasshopper?’ implying: You should not have burned it.
2. kɔ̀ːnɪ̀ sìːbà jʊ̌ːdʑìɾɛ̀ ⁿdʊ̀ːsɛ́ tɕɛ́
when 9:lion 9:come:PRF 1sg:say:SBJV what
‘When the lion comes what should I say?’ implying: I don’t know what to say.
Apart from showing this typologically relevant distinction between RQs in dialogue and RQs in monologue, the paper also discusses the role of the speaker (e.g. narrator versus participant) and the role of genre (narrative versus hortatory). The paper concludes with an outline of distinguishing factors for a functional typology of RQs and with a summary which of the established categories occur specifically in Rangi.

References
Hackstein, O. 2004. Rhetorical questions and the grammaticalization of interrogative pronouns as conjunctions in Indo-European. In A.Hyllested, A.R. Jørgensen, J.H. Larsson & T. Olander (eds.). Per Aspera Ad Asteriscos. Innsbrucker Beiträge zur Sprachwissenschaft, 167-186.
Zerbian, S. 2006. Questions in Northern Sotho. ZAS Papers in Linguistics 43: 257-280.

Even though grassroots level literacy projects in minority languages are by definition localised, they often depend on outside support from translocal organisations. One of these global players in mother tongue education is SIL International. A central strategy for localised literature production in SIL projects is the running of writer’s workshops in which incipient writers are trained to write in their mother tongue.
In this paper, I investigate the influences which literature on writer’s workshops has on individual SIL literacy projects. The Rangi language project in Northern Tanzania serves as a case study. A first one-day writer’s workshop was conducted for Rangi between May 2005 and January 2006 in four different locations. I give an overview of the preparation, implementation and results of these four workshop occasions.
The study pays particular attention to the workshop leaders’ assumptions, which are at least partly based on anecdotal evidence in SIL publications from the 1970s and 1980s that a natural writing style evolves “intuitively” in newly written minority languages. I then compare these assumptions with later instructions on writer’s workshops as published by SIL literacy consultants with international experience (summarised in Weber et al. 2007).
The results of my investigation suggest that available literature on best practices was not implemented to its fullest potential in the first Rangi writer’s workshop. The paper concludes with suggestions for better distribution and implementation of localisation instructions.

With rising international support (e.g. the 1993 UN "Declaration on the Rights of Persons Belonging to National or Ethnic, Religious or Linguistic Minorities"), the last few decades have seen a steep increase in vernacular literacy among the almost 7000 languages worldwide. Most of these languages are spoken by comparatively small communities of less than one million speakers, like the Rangi of Northern Central Tanzania, one of around 120 languages in that country. Literacy efforts in Rangi have been ongoing since 1998 under the auspices of SIL International (http://www.sil.org).

This talk will investigate aspects of language awareness raising, especially the question which features of their mother tongue aspiring Rangi writers are becoming aware of as they develop their own writing styles. This is considered to be an important step in training Rangi writers as most if not all of them have learned to read and write first in Swahili, the national language of Tanzania. Consequently, when now writing their own language for the first time in their lives, they map their Swahili literacy skills onto Rangi which impedes a natural and full-fledged use of their mother tongue in writing.

The data for this investigation consists of a subset of a larger corpus: 23 of 107 texts collected during the first four writer's workshops conducted in the Rangi literacy programme. Each of these texts exists in two versions: a draft version which each author produced at the start of the workshop, and a revised version which was produced at the end of the workshop. As both versions of each text are compared clause by clause, a caleidoscope emerges of those stylistic features which the authors deem worthy or necessary to be changed. Future training for Rangi writers will have to pay attention both to those areas where such changes across authors intersect thus giving evidence of potential consensus, and to those where uncertainty about the direction of change seems to prevail. Interviews with literacy supervisors and editors, all of them Rangi mother tongue speakers, complement the findings.

This talk will report particularly on three areas: the discourse marker 'maa' (which has been considered problematic by Rangi writers from the start of the literacy programme), the interaction of tense-aspect forms, and participant reference. In conclusion, implications for training vernacular authors will be summarized.

In recent years, so called “cluster projects” have become en vogue within the organization of SIL International: All languages in an area are researched simultaneously by a group of researchers rather than individually. The underlying approach is strictly participatory (Kutsch Lojenga 1996), i.e. giving mother tongue speakers the primary role in the linguistic analysis. The first three workshops in such a project aim at phonological analysis and orthography development: 1. data elicitation, 2. phonology at the word level, 3. phonology and morphology at the phrase and clause levels.

This talk presents some procedures and results from the phonology workshops for five Bantu languages in the Northern Tanzanian Mara region: Jita (E.25), Kuria, Ikoma, Kabwa and Zanaki (all E.40). On the one hand, I will talk about how the phonemes (or rather graphemes) were arrived at for vowels, consonants and tones without employing phonetics. On the other hand, I will report the discovery of vowel harmony systems of ever increasing complexity. For this second part, two models are relevant as background: 1. Hyman’s 1999 article on historical development of vowel systems in Bantu, and 2. Casali’s ATR typology. In particular, I will present a harmony conflict in Ikoma and conclude with implications for phonological analysis by mother tongue speakers themselves.

In hundreds of languages worldwide, minority language speakers are these days developing their own written literature. As they do so, questions of written style arise, particularly whether vernacular writing style should be “kept pure” or whether it should copy the writing style of other languages. As most minority language speakers receive their education in a regional or national language, influences from these mediums of instruction on their mother tongue use seem to be inevitable. In this talk, I am going to look at the narrative style of emerging authors in the Rangi language of Northern Central Tanzania, and to investigate any evidence of stylistic influences from Swahili, Tanzania’s national language. In particular, I am going to compare two fables which have been written twice by the same Rangi author, first in Swahili, and then, years later, in Rangi. In the comparison, I will pay special attention to opening and closing formulae, temporal adverbials and the use and flow of tense-aspect forms. In conclusion, it will be shown how stylistic influences actually go both ways and not just from Swahili to Rangi.

Kell (2000) describes a discrepancy between how letter-writing is taught in a particular South African adult literacy class, and how the learners actually do write personal letters. Reasons given for those differences are, that:
1. classes are based on an understanding of literacy as an individual activity, whereas literacy for many is a collective and shared event,
2. teachers lack expertise in catering for the actual needs of learners, and
3. the outcome of these literacy classes is increasingly "used for the purposes of certification rather than for social purposes" (p.212).
Other factors mentioned are the means of letter transmission (postal service versus hand-carried), and traditional constraints on addressing between in-laws.
After giving an overview of Kell's study, I will relate her findings to my own observations in a language development project among the Rangi of Tanzania. Striking similarities will be revealed between both settings, including teaching material and methods on the one hand, and letter writing practices and means of transmission on the other. In conclusion, I will discuss the implications of my comparative findings for the planning of adult literacy, and particularly the teaching of writing skills, within the confines of the Rangi language development project.

Minority languages in Tanzania continue to survive despite of their absence from national forums and despite of decades of pressure from Swahili. Neither the government’s language policy nor the educational situation, both of which are briefly sketched, actively support the vernaculars. In a case study, the Rangi language of Northern Central Tanzania is investigated, and especially some of the survival strategies which Rangi speakers have developed in the use of their mother tongue. Of particular interest are those domains of life where Rangi was successful in acquiring higher status than the national language Swahili, and how Rangi exerts pressure of its own onto smaller neighbouring languages. The paper ends with an outlook of how successful language survival can be continued into successful language development.