Home / Archive / 2015 / Why is Chakali still spoken?
Document Actions


1. Introduction


The speakers of Chakali (iso 639-3: cli), a Southwestern Grusi language of Ghana, recognise that their language is slowly disappearing. Using published materials, interviews, and observations, the article reports on the vitality of their language. It should assist in the documentation of a previously little-known language by helping to identify key sociolinguistic factors involved in endangerment dynamics. In doing so, it suggests some factors which are causing the language to slowly vanish and others which are responsible for the language still being spoken today.


Every year since 2007 I have carried out a field trip to the Wa East district of Ghana. The majority of the time was spent in a Chakali speaking village, mainly Ducie, with overnight stays in Motigu and Gurumbele, and several day trips to Katua, Sogla, Tiisa, and Tuosa. The remaining time was spent in Wa, the regional capital of the Upper West Region. These locations appear on the map in Figure 1. The majority of Chakali speakers who have received an education above Junior Secondary School (JSS) lives in and around Wa, or in other major towns and cities of the country. With many of these individuals I fraternise, share my observations, and debate ideas. At the village level, elders especially were the most important source of information. The dynamics of a language’s vitality can be substantiated with information on various time frames, that is why elders’ views are invaluable. The results I present are at most qualitative at present, while the quantitative data which this paper uses was gathered mainly in 2008 when a population survey was carried out in Katua, Motigu, Sogla, Ducie, and Gurumbele.

The article starts in §1.1 by introducing the ethnographic context. Some introducing remarks on geographical locations and historical background will be given. In § 2 I will briefly refer to previous statements and studies of Chakali sociolinguistics before presenting a vitality assessment based on the nine factors proposed in UNESCO (2003). Complementary factors and hypotheses dealing with why Chakali is declining, but also with how the speakers appear to keep Chakali alive, are proposed in § 3 .

1.1. Chakali in context


Figure 1 displays some of the villages of the area. Geographically described, Chakali is bordered to the east and south-east by areas inhabited by Waali and Bulengi speakers, two undocumented languages that can provisionally be classified as Western Oti-Volta based on folk linguistic impressions. Waali is the language spoken in Wa and surrounding villages and can be considered to be the lingua franca of the Upper West Region of Ghana. Bulengi is the language of Bulenga (and surrounding villages like Gilan, Chagu, and Dupari), a fast-growing town in terms of population and development whose language is yet to be compared with Chakali and Waali. To the north, Chakali is bordered by Pasaali speaking villages, i.e. a Sisaalɪ dialect, and Kpalewagu, whose inhabitants maintain a Mande language known as Kantosi. Tampulma speakers are mainly found in some villages of the Northern Region, but a few villages to the east of Chakali are within the Upper West Region's border (i.e. Holumuni and Belezing). The south and south-west areas comprise Vagla speaking villages and the uninhabited Mole National Park.


Since its first appearance in Goody (1954), Chakali has been listed in linguistic classifications as a member of the Grusi sub-group of the Gur language family. Today the language is classified as a Southwestern Grusi language alongside with Dɛg, Vagala, Tampulma, Kyitu (Siti), Phuie, Winyé, and varieties of Sɪsaalɪ (Naden 1989Lewis et al. 2014), all minority languages spoken in Northwest Ghana, Southwest Burkina Faso, and Northeast Ivory Coast. Grusi as a language cluster has been defined and confirmed in several publications (Delafosse 1912Köhler 1958Bendor-Samuel 1965Manessy 1969a,bKleinewillinghöfer 1997Brindle 2011) but, while the term ‘Grusi’ is associated with a language group, the term and its spelling variants (i.e. Grussi,GurunsiGrunshieGourounsi, etc.) there has never been unanimity on their designations in the French and English colonial vocabulary (Tauxier 19211924Rattray 1932a,bNicolas 1952Duperray 1984).

The history of the Grusi people before the middle of the nineteenth century, especially of the groups found today in Ghana and Burkina Faso, is for the most part unknown. Manoukian (1951:13) claims that “what are now the Northern Territories was peopled some 500 years ago by ancestors of the present-day Tampolense [Tampulma speakers JAB], Vagala and certain Isala groups, and by some Konkomba groups to the east”  [2]. Based on the maps of the divisions of the nineteenth century Gonja state (Goody 1967Wilks et al. (1986), Chakali villages were apparently within Katua, Daboya, or Wasipe districts. No mention of Chakali (or Chakalle or Cakali) is found in Goody (1967); Wilks et al. (1986), nor in Duperray (1984).

Figure 1 : Map of Chakali villages and beyond


With Chakali, three concepts are identified. The term may be used to name a land (tʃàkálɪ́ hàglɪ́ɪ̀), an ethnic group (tʃàkálɪ́ɛ́), or a language (tʃàkálɪ́ɪ́). However it would be wrong to think that a member of the Chakali ethnic group or someone living on Chakali land necessarily speaks the language. This is what Goody (1954:2) describes:

[t]he use of one self-applied name may cut across the linguistic frontiers which have already been defined. The Chakalle who inhabit the eastern part of the Wa district are split into those speaking a language of the Mossi group and those speaking a Grusi language. “Speaking a language” refers to the tongue which dominates in the child’s play group; the eastern Chakalle who use a Grusi language in this context are in fact mostly bilingual. The common name for the group derives from recognition of uniformity in other social activities.


Since the work focuses on the language, all these concepts are captured under Chakali, the term being accompanied with specification when necessary. However it is crucial to keep in mind that these three concepts (i.e. land, ethnicity, and language) are interwoven. For instance, according to Daannaa (1994), Chakali land and people consist of thirteen communities and their inhabitants: Bulenga, Tiisa, Sogla, Tuosa, Chagu, Motigu, Ducie, Katua, Bisikan, Kandia, Dupari, Gilan, and Gurumbele. Yet the sociolinguistic research I carried out indicates that Chakali can be considered the language of the inhabitants and forefathers of Tiisa, Sogla, Tuosa, Motigu, Ducie, Katua, and Gurumbele exclusively. [3]


The collective demonym for the people of the latter seven villages literally translates to m̩̀ ŋmá kàà, lit. ‘I say that’, whereas the one for the people of Bulenga and surrounding villages translate to ŋmɪ́nɪ̄ŋ dʒɔ́ŋ̀ ‘What is this?’ in the Bulenga lect (Búléŋíí). In this folk-sociolinguistic categorisation, the Waala are the ǹ̩ jɛ́ jàà lit. ‘I say that’ in Waali. [4] This way of making a distinction between groups corresponds more or less to what Goody (1954:2-3) calls Eastern Chakali and Western Chakali: the Eastern block is known locally as the m̩̀ ŋmá kàà while Western block as the ŋmɪ́nɪ̄ŋ dʒɔ́ŋ̀.

Table 1: Collective Demonyms and associated villages

Demonym 1

m̩̀ ŋmá kàà

ŋmɪ́nɪ̄ŋ dʒɔ́ŋ̀

ǹ̩ jɛ́ jàà

Demonym 2




Demonym 3 (Goody, 1954:2-3)

Eastern Chakali

Western Chakali





















Another popular distinction is that of ‘black’ and ‘white’ Chakali: respectively, tʃàkàlbúmmò ‘Black Chakali’ is a notion which connotes with secretive individuals and possessors of powerful medicine. To the best of my knowledge, this is equivalent to whatm̩̀ ŋmá kàà represents. The notion of tʃàkàlpʊ̀mmá ‘White Chakali’ corresponds, according to my ‘Black Chakali’ consultants, to talkative people who cannot hold back. They comprise the inhabitants of Bulenga, Dupari, Bisikan, Chagu and Gilan, that is, those villages included in what Daannaa (1994:2-3) identifies as Chakali people, minus the villages where the language is said to be indigenous. Obviously, if one asks the same question in Bulenga and surrounding villages one may get a different interpretation of the distinction between ‘black’ and ‘white’. [5Table 1 organizes the information for convenience. It also constitutes a hypothesis to be tested since the denominations do not necessarily map one-to-one and discussions I had about those concepts were often confusing. For instance, some men interviewed in Tuosa in 2014 told me that Tiisa, Tuosa, and Katua are not m̩̀ ŋmá kàà, but are tʃàkàlbúmmò.

Within the ‘Black Chakali’ or the m̩̀ ŋmá kàà, there exist lexical, phonological, and prosodic variations which validate folk dialectology. People's account of lectal variations partially parallel the proximity of geographical locations. The lect of Katua, Tiisa, and Tuosa, on one hand, and the lect of Ducie and Gurumbele, on the other hand, are considered the farthest apart on a continuum. [6]. Examples are presented in Table 2 .

Table 2: Distinctions between Katua and Ducie lect

Katua lect

Ducie lect




morning greeting

pɪ́ɪ́ / pɪ́á

kpa͂́a͂́ / kpa͂́a͂́má


bʊ́ɪ̀ / bʊ́sɪ̀

bɪ́ɪ́ / bɪ́á













kola nut (ultm. Waali and Gonja resp.)



type of staple food


















Awie (proper noun)


ʊ̀ bájʊ́ʊ́

ʊ̀ wáwʊ́ʊ́

he/she/it came

ʊ̀ bɪ́jʊ́ʊ́

ʊ̀ bɪ́wʊ́ʊ́

he/she/it is cooked





A recurrent example of folk dialectology is how each village would express ‘to eat yam’. Motigu, Gurumbele, Tuosa, Tiisa, and Katua ‘chew’ yam (tie), whereas Ducie ‘eat’ yam (di). And while ‘yam’ is pronounced kpã́ã́ŋ in Motigu, Gurumbele, and Ducie, it is pronounced pɪ́ɪ́ in Tuosa, Tiisa, and Katua. Thus, if someone says tìè kpã́ã́ŋ, he/she is identified as being from either Gurumbele or Motigu. The expression dí kpã́ã́ŋ is typically uttered by someone from Ducie, and tìè pɪ́ɪ́ by someone from Tuosa, Tiisa, and Katua.

Still, each village is recognised by the speakers themselves to have a set of unique features. For instance, within the Ducie-Gurumbele lect, Ducie people consider that Bɛ̀lɪ́lɪ́ɪ́, the Gurumbele language, is a lect resembling a mixture of Dùsèlɪ́ɪ́, the language of Ducie, and the lect of Gbanwale (gbʊ́ŋwálɛ́ɛ́) i.e. a variety of Tampulma. They also say that Bɛ̀lɪ́lɪ́ɪ́ is ‘stretched’, and that the speakers ‘pull their words’. The people of Gurumbele believe the opposite, saying that Dùsèlɪ́ɪ́ is ‘short’. These statements are reported in (1)  [7].















‘The language of Gurumbele; it pulls/stretches.’















‘The language of Ducie; it is short.’

Whereas people from Ducie say that those from Gurumbele “waste their time when they speak”, they themselves acknowledge that they “swallow some of the sounds”.

Table 3: Lectal variations in Ducie and Gurumbele

Gurumbele lect

Ducie lect






ʊ̀ kpágá lùò rē

ʊ̀ kpágá tɪ́ɛ́ rā

she bears a child (foetus)

dʒɛ̀bálàŋ (or) ámìõ̀

músì (or) díèbíè




small hoe

ɪ̀ mʊ́gáʊ́

ɪ̀ dáráʊ́

you are lying (false statement)



young girl



type of ant



ego's mother's brother



ego's senior brother



clear away grass in dry season






mature girl



type of cooking pot



raised bed (gardening technique)




small yam mound






young boy



type of Genet






pouched rat



old woman


kùò rē

kùò rō

farm FOC

sùkúù rē

sùkúù rō

school FOC


Table 3 presents some lectal variations recorded in Ducie and Gurumbele. The lexical variations express the same entity or proposition in each case and are the default expressions that correspond to the English translation in the last column in both villages [8]. The prosodic variations are essentially the perceived pitch associated with a word. The first three involve the stem {b|w}ie - {b|w}ise (sg. pl.) ‘child’ or ‘small’ in stem final position whose tone is on the surface LH (rising) in the Gurumbele lect but HL (falling) in the Ducie lect. An example of phonological variation is the surface form of the focus marker. In short, of the two phonological rules operating in this domain in the Ducie lect, i.e. ATR-harmony and rounding-hamony, the Gurumbele lect only has the ATR-harmony. Despite the above linguistic and anecdotal facts showing the presence of variations, all the Chakali lects are mutually intelligible.

This section introduced a geographic view of Chakali and provided a brief historical background, which are essential to understanding the sociolinguistic situation. Speakers are aware of Chakali varieties as each community is known to have its own lectal variant and community glossonym (i.e. kàtʊ́ɔ́lɪ́ɪ́bɛ̀lɪ́lɪ́ɪ́mòtígíítíísálí, etc); each lectal variant having more affinities with some than some others. Some examples of folk-linguistic classification and dialectal variations were given in order to show that Chakali, like all natural languages, is not bound and homogeneous.

2. Chakali linguistic vitality

2.1. Previous sociolinguistic surveys of the Chakali language area


In the last 40 years, two sociological surveys have been conducted in the Chakali language area. In 1974, the Ghana Institute for Linguistics (GIL) – changed in 1982 to the Ghana Institute of Linguistics, Literacy and Bible Translation (GILLBT) – investigated the need for Chakali language development (Reimer & Blass 1975). The authors of the report attested high Waali comprehension in the two Chakali villages visited, i.e. Motigu and Ducie. [9] Since written material already existed in Waali, the report concluded that future development should be linked to Waali literacy efforts because Waali written material was assessed to be comprehensible in their survey.


A follow-up survey was carried out in 1995 and published in Tompkins et al. (2002). Commenting on the 1974 survey, Tompkins et al. concluded that (i) twenty subjects in two villages is not enough to demonstrate L2 comprehension; (ii) understanding Waali material does not imply speaking Waali, therefore the degree of bilingualism of the Chakali people cannot be demonstrated; and (iii) in word lists, the presence of Waali terms does not indicate bilingualism; borrowing is a more probable explanation.Tompkins et al. reiterate that “the purpose of the survey was to help administrators [of GILLBT] decide whether there is a need for Chakali language development or whether these communities should be linked to Waali literacy efforts”. Chakali language development needs were to be determined by: (i) assessing Waali comprehension, first from the reported data and later from tested data if necessary; (ii) assessing the use of Chakali and Waali in various domains, to determine whether there were indications of language shift toward Waali; (iii) assessing language attitudes related to potential community involvement in Chakali language development; (iv) gathering information on population, education levels, literacy, religious environment and language contact with Waali speakers; (v) assessing Dagaare comprehension and Dagaare use, since Dagaare is the regional language used for educational purpose in the Upper West Region; and (vi) assessing Vagala comprehension, as Vagala is closely related to Chakali.Tompkins et al. carried out a survey in Katua, Motigu, and Ducie to address these questions.


Amongst many other revealing figures, their results show that all 37 individuals interviewed in the three villages report the ability to speak and understand Waali. While 74% can understand Dagaare, only 8% can speak it; and while 44% claim to understand Vagala, only 3% report the ability to speak it. The village of Ducie reports the least use of Waali in public domains. Further, Katua reports the most use of Waali at home; Katua and Motigu reports more daily use of Waali than Ducie, and overall more females than males report Waali in their daily use.

The information provided in Tompkins et al. (2002) is very similar to what I observed during my field trips. Although I did not use comparable questionnaires, their conclusion clearly reveals that inhabitants of Ducie, Motigu, and Katua do speak and understand Waali. The GILLBT administration concluded that “Chakali language development would not be pursued” because “Waali comprehension and use was high enough and its use was extensive enough for the use of Waali written materials” (Tompkins et al.2002:23). Although GIL and GILLBT carried out two surveys in the Chakali area, I have been told that no contact with any community members have been established since 1995.

2.2. UNESCO’s ‘nine factors’


To the best of my knowledge, statements qualifying Chakali as ‘endangered’ only appear in Dakubu (2005) where the language “can indeed be classed as an endangered language in the medium term”, and in Awedoba (2006), which asserts that “the language is almost dying out”. Lewis et al. (2014) treat the language as ‘threatened’, perhaps based on the evidence provided in Tompkins et al. (2002).

To determine the vitality of Chakali, I use the questionnaire developed in UNESCO (2003). This document is “designed to assist language communities, linguists, educators, and administrators (including local and national governments and international organizations) in finding ways to enhance the vitality of threatened languages.” It provides nine factors that both identify imperative needs and gives researchers a broad idea of the state of endangerment of a language. The authors of the questionnaire acknowledge that no single factor can be used to assess a language’s vitality or its need for documentation. The integration of the nine factors is ranked on a continuum from stability to extinction. In this section, I evaluate each factor defined in UNESCO(2003).


These factors are (i) intergenerational language transmission, (ii) absolute number of speakers, (iii) proportion of speakers within the total population, (iv) trends in existing language domains, (v) response to new domains and media, (vi) materials for language education and literacy, (vii) governmental and institutional language attitudes and policies, (viii) community members’ attitudes towards their own language, and (ix) amount and quality of documentation. These factors are evaluated either on a scale or provided with absolute numbers in § 2.3 .  [10]


It is important to note that information is extensive for certain villages and limited for others. Only the villages Tiisa, Tuosa, Sogla, Motigu, Ducie, Katua and Gurumbele are considered (see Figure 1): the villages Bulenga, Chagu, Chasia, Bisikan, Kandia, Dupari and Gilan are not included. As mentioned in the discussion above, the reason is that there was general agreement among consultants who advised me not to look for the Chakali language in these villages. The majority seemed to agree that Ducie and Gurumbele are the places where one can hear Chakali in its ‘uncorrupted’ form, even though Sogla is claimed to be the first Chakali settlement. Needless to say, the outcome is mostly qualitative and must be regarded as provisional since I did not carry out a systematic survey for each factor in all villages. [11]

2.2.1. Factor 1: Intergenerational language transmission


It is fair to state that all the children in Ducie and Gurumbele speak Chakali, and that in Motigu and Katua the great majority of children speak the language. In Katua, there are more women from Pasaale speaking communities who are married to men of the village than anywhere else, and family members of the chief who speak Waali also live in the village. For Sogla, Tiisa, and Tuosa, children may be said not to speak Chakali but a minority may be able to understand their parents or grandparents when they speak. In Motigu and Katua, Chakali is used by the majority, whereas in Sogla, Tiisa, and Tuosa the language is used by the parental and grandparental generation only. The language is therefore definitively endangered in the latter three villages as it is no longer being learned as mother tongue by children in the home. I was told that Waali is the language which has taken over. With respect to inter-ethnic alliances, it is possible and very likely that in any village, non-Chakali speaking mothers speak their own native language (e.g. Dagaare, Waali, Vagala, Tampulma, or Pasaale), but a child’s usage of the language would then be restricted to specific social domains (i.e. such as at home with the mother and when maternal family members come to visit).

2.2.2. Factor 2: Absolute number of speakers


Wilks (1989:17) reports that the population of Ducie was 350 in 1948, and 529 in 1960. Naden (1989:144) classifies Chakali as a medium-small language (i.e. 5,000-50,000 speakers) and Tompkins et al. (2002) estimates that there are 5,000 speakers. [12]Dakubu (2005) writes that “[t]he 2000 census (…) suggests far fewer [than 5,000]”. Lewis et al. (2014) estimates the population at 6,000, an estimation probably based on the number offered in Tompkins et al. (2002) and a population growth rate. The question then is what was meant by ‘Chakali’ when these censuses and surveys were carried out. For instance, if one adopts the ‘Chakali’ of Daannaa (1994) mentioned in § 1.1, the populations of thirteen communities need to be added up.

My estimate of the absolute number of Chakali speakers is based on a population survey I carried out in 2008 in the four villages Motigu, Ducie, Gurumbele, and Katua. [13Table 4 gives the village and section names, the number of individuals, households and houses, and finally, in the last column, the percentage of households per house. [14] From the grand total in the second column of Table 4 (i.e. 3,484), the number associated with Fulani, Dagaaba, and Waala settlers (i.e. 500) is subtracted. [15]

Table 4: Populations of Ducie, Gurumbele, Motigu, and Katua (author’s field survey, January 2008)





















































Belekani (Dagaaba)








































































Chief palace















Grand Total






However, a liberal figure of 500 is added to the population of these four villages to represent a projected number of speakers from the parental and grandparental generations in Sogla, Tiisa, and Tuosa, in addition to speakers of the diaspora (e.g. Chakali-speaking women marrying in a non-Chakali speaking village). Seasonal workers and Kayaayos – women porters in urban centres – were not considered as diaspora since they were included by the repondents of the survey among the household's individuals.Table 5 presents the results of a population survey I carried out in Sogla in April 2010. About a fifth of the population (i.e. 96), mostly individuals of Tindambanɪɪ, said they can understand Chakali and that some of them speak it at home. The rest of the population speaks Waali or Dagaare. I inferred the percentage of Chakali speakers of Tiisa and Tuosa by extending the information I gathered in Sogla: Tiisa and Tuosa joint population is about 1,500.

Table 5: Populations of Sogla, April 2010

































Therefore, after adjustment based on projected population of (non-) speakers, the estimate of the absolute number of Chakali speakers is 3,484. Future researchers may want to (i) compare the numbers proposed with current and future national censuses and population data, (ii) confirm/oppose the extrapolation, and (iii) look closer at Chakali as a second language and its level of proficiency in the villages of the area and in the diaspora, especially among children living outside the Chakali-speaking villages. [16]

2.2.3. Factor 3: Proportion of speakers within the total population


The estimate of the population of Chakali speakers (i.e. 3,484) in § 2.2.2 constitutes 0.5% of the Upper West Region’s total population, and 0.01% of the Ghanaian population. [17] With these numbers, Chakali is among the smallest languages in Ghana in terms of L1 speakers (Lewis et al. 2014). These numbers are at odds with those provided by the Wa East District Assembly, which suggests that Chakali make up 19% of the Wa East district [18], which is estimated at 66,358 inhabitants (GSS 2002), i.e. approximately 12,000 speakers. Here again, the interwoven concepts of land, ethnicity, and language seem to have affected the result.

UNESCO (2003:9) maintains that the number of speakers of an ancestral language in relation to the total population of an ethno-linguistic group is a significant indicator of language vitality. If one believes the elders I talked to and accepts that the villages Sogla, Tiisa, Tuosa, Katua, and Motigu were at a certain point in time entirely Chakali-speaking villages, i.e. the majority would speak Chakali in daily activities, then only a minority of the Chakali people, the majority of whom live in Ducie, can be said to speak the language today.

2.2.4. Factor 4: Trends in existing language domains


The languages of formal education are common to all villages. Currently in Ghana, English is the main language of instruction, and each region has a local language with an established orthography recognized for education purposes by the Ministry of Education. [19] In Chakali schools, the indigenous Ghanaian language being taught is Dagaare. Teachers in Ducie and Gurumbele, the majority of whom are from Wa or Dagaare-speaking communities, told me they have a hard time teaching this subject, and added that it is often dropped altogether towards the end of the school year in order to focus on the core subjects, and those considered to be more important for the final exams. One reason given to me by both the schoolchildren and the teachers is that they do not understand ‘book Dagaare’ as it does not resemble the South Dagaare dialect they are used to, of which Waali may be considered a variety. In point of fact, it is no surprise that teachers and schoolchildren struggle with this imposed language; Dagaare has important dialect divisions (Bodomo 1994) and “there has been some dispute over the viability of the current standard” (Anyidoho & Dakubu 2008:156).


Compared to other places in rural northern Ghana, the schools in the villages of the Chakali area have in all probability an average success rate in terms of government education. The literacy rate in English is low, the passing rate of the Basic Education Certificate Examination is under 50%, and the drop-out rate is reported high (Akyeampong et al. 2007). In each Chakali village, there is infrastructure and teachers, and sometimes also the required material. The schools, the teachers, and the pupils can be described as relatively normal in a report or after an afternoon visit. When it comes to the goal, i.e. education, it is hard to tell whether the situation can be considered successful. In 2008 I often took over for absent teachers in Ducie and Gurumbele schools, teaching mathematics in a Primary 6 class using English as the medium of instruction. I witnessed caning, some Junior Secondary School students (12-16 years old) who could simply not read, full-day labour work (farming, food making, digging, etc.), among other disturbing scenes. Another observation was that classrooms are devoid of dialogue. Pupils seemed afraid and confused, and whenever a child spoke, it was rote repetition usually disconnected from his/her understanding.


The local language policy is probably one of the most disconcerting problems. The former headmaster of Ducie primary school had revised the policy in 2008; instead of being caned for speaking Chakali in school (and in the school yard), a fine had to be paid, an amount about the same as the wage of a woman’s one day work of collecting and selling firewood. Usually the teachers collected the money. I was told that similar enforcement takes place in other schools of the area. Such harsh practices are easily recalled by educated Chakali men. These are difficult to document today because one-day visitors, e.g. NGO or government representatives, will never witness such abuse. Nevertheless, I witnessed the reverse strategy in Gurumbele, where some local teachers use Chakali to interpret and summarise readings, and to explain concepts to pupils. Obviously, there is active repression of Chakali, both by the school management and the imposed national curriculum. The school is thus the first formal social and educational institution a Chakali child meets in life which depreciates and devalues his/her language. [20]


Another domain where Chakali was and still is suffering from active repression is the domain of external relations. Leaving aside the historical details of the original Chakali-Waala alliance, the imposition and enskinment of Waala chiefs by the colonial administration in Chakali villages upset the pre-colonial cooperation between the Chakali and the Waala. According to Daannaa (1994: 62), strangers making decisions in a village is unprecedented in Chakali history. In colonial times, chiefs were known for their abuse of power. In consequence, Chakali inhabitants tried their best to avoid a case leaving the village and being reported to the chiefs and native courts set up by the colonial administration in Bulenga, Busa, and Wa. It is certainly around that time (i.e. approximately sixty to eighty years ago) that Chakali speakers started to suppress their language in favour of Waali. This may have been out of cautiousness, in order to avoid misunderstandings and conflicts with chiefs. However such a language shift may have been first limited to administrative meetings and further extended to general social contexts. I will address this point in § 2.5.1.


Nowadays, the government of Ghana speaks through an assembly man at the community level. He/she (predominantly he) is elected at the community level and represents one community, or sometimes two, at the district level. Even though that person is indigenous, he/she is generally educated and does not live on-site. The fact that the assembly men are rarely in the communities forges dissatisfaction among people. This is why, recently, local teachers, who are literate and reside in the village, are preferred candidates. Too often I have witnessed governmental officials visiting Ducie while the assembly man was absent. [21] Since English, Waali, and Akan are frequently used on such occasions, the solution is to select someone who speaks a common language with the visitors in order to have their message transmitted. When the common language is other than Waali, the selected person provides the interpretation for the audience. Because the area is still very remote, governmental services are not found in the area, except for a clinic in Ducie established in 2001. The clinic staff speaks English and Waali.


In summary, the situation in Ducie and Gurumbele is one of multilingual parity since English and Waali are used as the primary languages in most official domains. Speakers do experience active repression of Chakali in domains like education and modern administration. Chakali is integral to a number of public domains, especially traditional religious institutions, local stores and markets, and where members of the community socialize. In Ducie, Gurumbele, Motigu, and Katua, the coexistence of Waali and Chakali is a good example of diglossia, which was already established in Tompkins et al. (2002), whereby Chakali is used in informal and home contexts, and Waali is generally used in official and public contexts. For Sogla, Tiisa, and Tuosa, diglossia is representative of only a minority of the overall population. The majority speaks Waali in all contexts. Nevertheless, Chakali in Motigu and Katua may be categorized as dwindling; Waali spoken at home and in religious contexts is reported to be increasing in both villages. In these two villages there are many signs which suggest that traditional religious and administrative institutions are fading away. However, while Waali is spoken and understood by a majority in all the villages, there are at most 1 to 4 individuals per village who can speak and understand English, apart from the local school staff. At this point in time, it is difficult to determine i) how ‘prestige’ plays a role in linguistic choices, and ii) the level of production and comprehension of English amongst schoolchildren. These present interesting subjects for studies in themselves.

2.2.5. Factor 5: Response to new domains and media


Since the villages lack electricity  [ 22 ], most modern media devices are useless there. Battery operated radio is the only reception device found in the village, and talking drums are used to transmit messages to neighbouring villages. The situation should change since the number of mobile phone owners is increasing significantly, and small generators and solar installations are considerably less expensive than they were five years ago. The first radios – known as àkàsànʊ̀má, an Akan word – were brought to Ducie by Waala teachers around the 1960s. Today many youths and adults own a radio. They are mostly listened to at night or brought along to the farm, especially when one plans to stay overnight. The broadcast languages are Waali, English, and Dagaare by Radio Upper West and Radio Progress, both transmitting from Wa. News in Tumulung Sissala is presented as well by both stations, but few Chakali can understand the language.

Despite the lack of electricity and a stable network, mobile phone owners in Ducie and Gurumbele abound. There are one or two specific geographical points in these villages where one can access the network. English expressions incorporated in a Chakali matrix (e.g. tʃáádʒì ‘charge’, tɛ́gɛ́sɪ̀ ‘text’, kɔ́ɔ́lɪ̀ ‘call’, sɛ́ɛ́tɪ̀ ‘set’) are used for mobile phone activities. Although illiterate users often ask schoolchildren to read received texts (mostly advertisement), text messaging is not used.


The literacy rate is not known, but apart from schoolchildren, teachers, and a handful of adults who attended primary school, no one else can read or write in the villages. Books and old newspapers are available in school but it would appear that students rarely use them. Speaking Chakali is forbidden in many schools, as mentioned in § 2.2.4 . Waali is used in the mosque for discussion, and some Arabic lines are learnt through rote memorization. Chakali is used to preach by a small number of Christians in Motigu, and Ducie, but this is the only case I found where the language is being used in a relatively new domain. Therefore, based on these details, a typical response to new domains is to use (or be exposed to) a language other than Chakali or a non-Chakali vocabulary.

2.2.6. Factor 6: Materials for language education and literacy


The language can be said to have neither an orthography nor material. However, a Brazilian pastor known as Ronaldo designed in 2001 two illustrated booklets with the intention of starting a vernacular literacy project. According to one of the language consultants involved, 5 to 10 individuals could read with good fluency after just a few lessons. The first booklet introduces the alphabet designed. I was unable to locate any copies of the booklet. The second 20-page booklet consists of syllables and short sentences thematically organized. A photocopy of the single remaining copy is now in my possession. The program was never widely adopted, and the quality of the work, especially the orthography design, was far from what has been currently achieved by local organizations working with and on literacy in Ghana.

In 2008, with the agreement of the headmasters, I introduced a pilot orthography to some school children in Ducie and Gurumbele. They all learned it after a few classes. Daniel Kanganu and I wrote and printed two short books which were distributed to the schools in Motigu, Katua, Ducie, and Gurumbele (Kanganu and Brindle 2008a,b). In 2009, I gave one copy of a Chakali-English dictionary to 5 individuals in Ducie and Gurumbele to corroborate its content. The following year one copy of an edited version was distributed in four schools and given to native teachers. There were no follow-ups with those individuals.

On the whole, I cannot claim that “a practical orthography is known to the community” and that “some material is being written” (UNESCO 2003:12), since the orthography is not ‘known to the community’ as yet, only to a handful of schoolchildren and consultants of past projects, and, while it is true that material has been written, the majority of Chakali speakers are not ‘aware’ of it. Some people I talked to are upset about the fact that most of their neighbours’ language has been written or had a literacy program  [23]. Education in the native language is presumably essential for language vitality, especially in a country where education is a political priority. Several studies in recent years have shown the necessity of teaching non-language subjects in the child’s native language: Hinton and Hale (2001) offers many case studies. Chakali has not been implemented in either formal or non-formal education. Today it is hard to say whether the speakers wish their language to be written. On the one hand, ‘yes’, they do, perhaps only for the status. But the fact that decision makers in the communities are illiterate makes their motivation puzzling. On the other hand, the sense of achievement and the benefit one gets from having an educated and working-class relative in an urban center are reasons and motivations to keep the system as it is today, i.e. English as a medium in school. A conceivable solution would not choose one or the other language but design a pedagogy where the child’s native language is used when necessary. If one accepts the claim that literacy is directly linked with social and economic development, then materials on all topics for various ages and language abilities and people to teach them are needed, i.e. a language program (see Grenoble and Whaley (2006:7). That being said, such a language program should expect the challenge of lect/variety choices and orthography decisions.

2.2.7. Factor 7: Governmental and institutional language attitudes and policies, including official status and use


In the government’s records, the Chakali people are either ethnically Waala or Grussiother Grusi (GSS 20022008), two ill-defined concepts  [24]. Like the majority of Ghanaian languages, there are no efforts being made by the government to recognize Chakali and promote it. The Ministry of Education favours English and Dagaare in the Upper West Region. The Adult and Non-Formal Education Division, targeting people within the working age group 15-45, especially women, has made some efforts to teach communities other languages than the official ones, but the Chakali area has not been considered (Fuseini & Abudu 2014). In previous sections I mentioned that the government has explicit policies in education and that communication with the government (or through any of its institutions) at the community level is conducted in Waali, Dagaare, Akan, or English. Chakali can arguably be said to face active assimilation since “the government encourages minority groups to abandon their own languages by providing education for the minority group members in the dominant language” and that “speaking and/or writing in non-dominant languages is not encouraged” (UNESCO 2003:13). Hence Chakali is neither recognized nor protected.

2.2.8. Factor 8: Community members’ attitudes toward their own language


I do not claim to have explicitly evaluated community members’ attitudes toward their own language, but even without having surveyed the question with proper sampling it is possible to reflect on this factor. Generally, the people in Ducie and Gurumbele praise the attention I have shown to their language and would support language maintenance initiatives. I had the same impression in 2011 when a consultant from Gurumbele and myself visited Tiisa, Tuosa, Motigu, Sogla, and Katua on a language awareness campaign. In these villages people we met were conscious of the language shift. The elders at those meetings generally put the blame on the youth – who chose to use Waali at home for reasons the elders cannot understand – and/or on themselves, who unconsciously communicate with their children in Waali.


Since the majority of Chakali people also speak Waali, they are not hindered from economic mobility or integration into 'mainstream' Wala society. It is not a problem, and is even a source of pride, for two men to speak Chakali to one another when they find themselves in Wa. This attitude is relatively recent and surely due to a different kind of top-down regulation, that is, a decrease of chiefs’ domination and increase in justice controlled by national courts and police apparatus. Adaptation and conversion to Waala language and lifestyle is easily evitable; the average Chakali person comes to Wa to trade and goes back to his or her village. There is no fear of retaliation by dominant Waala.


Nevertheless, a social group which may be highly influenced in terms of attitude toward their language is the group of Senior Secondary School (SSS) students leaving the villages every year for education. When they move to Wa (or towns in the region where SSS education is provided), they normally become Muslim (or Christian) and adopt Waali (or Dagaare). I suspect that this group alone perceives Chakali to be characterized by qualities such as backwardness and behind in ‘progress’ or ‘development’. This attitude develops from the prospect of Chakali failing to meet the challenges of modernity, i.e. being politically, culturally, and economically different from what they see in towns, or what media exposure tells them. Rarely do such pupils move back to settle in the villages upon completion of their education.

2.2.9. Factor 9: Amount and quality of documentation


In pursuance of “assessing the urgency for documenting a language”, (...), “the type and quality of existing language materials must be identified” (UNESCO 2003:16). This criterion makes reference to language description work such as grammars and dictionaries, but specifically to “written texts, including transcribed, translated, and annotated audiovisual recordings of natural speech”. The only work of that nature is Brindle ( 2011), an unpublished dissertation which contains a grammatical sketch, a word list, and annotated texts. A reworked and expanded version has been proposed to Language Science Press, African Language Grammars and Dictionaries Series. Audio and video recordings exist, some of which have been annotated, but are not yet accessible for research. Given the current situation, this factor may be graded as undocumented-inadequate.

2.3. Score


The discussion regarding the nine factors suggests a division of the Chakali villages into three groups (see Table 6 ). Sogla, Tiisa, and Tuosa correspond to the villages where intergenerational transmission is ineffective and where Waali is used in formal and informal domains. They are the E(ndangered)1 villages. Motigu and Katua correspond to E2 villages. In both villages, Waali is encroaching on Chakali in formal and informal domains. The situation is not alarming since Chakali is spoken by the majority and intergenerational transmission is effective, but given the average population size of the villages and the recent conversion to Islam of their youth, it is worth considering that a language shift to Waali may take place within a short time.


A. B. Sakara and H. S. Daannaa, both born in Tuosa, told me that Chakali was spoken by everyone in their village when they were children, i.e. 1950-1960. There are no signs indicating that the same language replacement which took place in Tuosa cannot take place in Motigu and Katua. Finally, the E3 villages, Gurumbele and Ducie, show the most effective intergenerational transmission of the Chakali language. Both villages also establish local alliances (i.e. marriage, common shrines, one assembly man for both villages, etc.). Waali is spoken and understood, yet it is usually spoken in specific domains, for instance in official visits from the district or regional capital conducted by governmental bodies, and to Waali-speaking visitors, traders or migrant farmers.


UNESCO (2003:9) proposes that the nine factors should be assigned a score. In the assessment, the score is either graded on a 0 to 5 scale, or absolute, as for factor 2 . This can be illustrated with an example taken from Table 6 , which summarizes the vitality of Chakali. For factor 1 , grade 5 is assigned to Ducie and Gurumbele, grade 4 to Motigu and Katua, and grade 2 to Tuosa, Tiisa and Sogla. Grade 5 represents a safe language which is used “by all ages, from children up”, stage 4 represents an unsafe language which is used “by some children in all domains and all children in limited domains”, and stage 2 represents a severely endangered language, as it is mostly used “by the grandparental generation and up” (UNESCO 2003:8).

Based on observations, interviews and a population survey, my estimated degree of endangerment of Chakali is severely endangered in E1 villages, unsafe or vulnerable in E2 villages and safe for at least another generation in E3 villages. Overall the language may be said to be highly vulnerable.

Table 6: Estimated degree of endangerment for the E1 (Tousa, Tiisa, Sogla), E2 (Katua, Motigu) and E3 (Gurumbele, Ducie)






intergenerational language transmission

severely endangered (2)

unsafe (4)

safe (5)


absolute number of speakers

[3,484] [25]


proportion of speakers within the total population


trends in existing language domains

highly limited domains (2)

dwindling domains (3)

multilingual parity (4)


response to new domains and media

[inactive-minimal (0-1)]


materials for language education and literacy

[no orthography available (0)]


governmental and institutional language attitudes and policies, including official status and use

[active assimilation (2)]


Community Members’ Attitudes toward Their Own Language

all members value their language and wish to see it promoted (5)


Amount and Quality of Documentation

[undocumented-inadequate (0-1)}

2.4. Discussion


The factors on which the result in § 2.3 is based were originally meant to be “useful for characterizing a language’s overall sociolinguistic situation” and are thus insufficient in understanding a specific sociolinguistic situation. Lüpke & Storch (2013:270-278) go as far as to say that UNESCO (2003) (i) has a “monolingual bias”, i.e. the ideology which assumes one community = one ethnic group = one language; (ii) has a “graphocentric ideology” (Blommaert 2004:645), the ideology which assumes that a language needs an orthography and reading material to compete; (iii) sees multilingualism as negative and threatening, meanwhile multilingualism is the standard sociolinguistic situation in most areas of Africa; and (iv) ignores important factors responsible for cases of language attrition in Africa (political, economic, geographical, etc).

While I certainly agree with the points raised in Lüpke & Storch (2013:270-278), predefined questionnaires, like many elicitation tools in (field) linguistics, are usually designed for comparative purposes and for their usefulness when background knowledge is at its lowest. It is always the case that the ‘eco-specifics’ need to be identified and provided by the researchers and the communities, thus ad-hoc solutions are very often needed. Hence next in the exploratory section, I propose some sociological and historical factors that may be seen as linked to the language’s vitality and responsible for language change. The discussion suggests possible avenues for future investigation.

2.5. Punctuation events as blows to linguistic vitality

2.5.1. Colonial Administration


The first blow to linguistic vitality came from the imposition of Waala chiefs by the Colonial Administration, which was mentioned in § 2.2.4 . As for other groups in northern Ghana, the institution of chieftaincy was certainly known by the acephalous Chakali, but it was imposed on them gradually starting with the British colonial rule of 1901. Wilks (1989); Daannaa (1994); Salih (2008) all seem to agree that the position of war-chief (bòŋ nã̀a͂̀) in Chakali, a role similar to the leader of soldiers in Gonja organisation (mboŋwura), dates back to the time of, and was likely due to, the slave raids of the nineteenth century  [26]. At that time the Chakali and Waala cooperated for defence, in what is known as the koŋboŋi alliance (Daannaa 1994:63). At the beginning of the colonial era, the British favoured a centralized administration over the acephalous mode of administration found in Chakali. The district commissioners saw that these independent micro-states and their distance from Wa were unmanageable, and that trusted individuals were therefore needed. In short, the British colonial system first gave the war-chief of Bulenga authority over all other war-chiefs of Chakali land, and later on, Chakali as a whole was annexed to the Busa division of the Wa Native State, so the Chakali people became subject to the authority of the Busa chief. By 1943 “the whole of Chakali had come under the effective control of the Busa or Yijihi Gate of the Wa royal house” (Daannaa 1994:75). By then the chiefs were regarded as the authority by the colonial administration, and eventually, the government.


Preceding the introduction of chieftaincy, the institutions of landlordships (tɔ́ɔ́tɪ̀ɪ̀nà), the administrative body of elders representing the clans (nɪ́hɪ̃̀ɛ̃̀sā), the soothsayers (vʊ̀vʊ́tá), and the shrine representatives (vʊ́gtɪ́ɪ́ná) collaborated in the administration of communities. The day-to-day consequences of the colonial system on a Chakali-speaking village must have been considerable. Not only was authority no longer in the hands of a customary council, but strangers were ruling each Chakali village. The decisions surrounding enskinment were taken in Wa or Busa and the appointees were non-indigenous. As a consequence of that, the newly established chiefs must have spoken a Waali-like language and expected people to address them in this language. Daannaa (1994:75-80) writes that chiefs had unprecedented power and oral accounts tell us of much abusive, if not abhorrent, behaviour. Under such conditions, it would seem natural that villagers contemplated language shift as a defence mechanism.


When language shift as a defense mechanism is considered, one can partly interpret the linguistic vitality of Tuosa and Sogla. On my visits to these villages, I was told that Chakali was spoken by all in the 1940s and 1950s. Today Tuosa and Sogla are villages where: i) chiefs themselves and their extended families, not regents of absentee chiefs, have always lived, and ii) Chakali is severely endangered. Although no evidence is yet available supporting a correlation, it is worth asking whether the presence of a chief and the attitude of the community towards him have or have had impact on linguistic practice.

An understanding of the usage of Waali in Chakali communities cannot be based solely on today’s observations because Waali may not previously have been as widespread as it is today. One recent encounter supports this possibility. In 2013, I met by coincidence a witness who claimed that there was a time where Waali was not a language of wider communication in Ducie and that being understood using Waali those days was almost impossible. Mr. Nuhu was a Primary school teacher in Ducie between 1964 to 1966 with no family members in the Chakali area. A Waala himself, he learned Chakali in Ducie to be able to communicate. When we met he could retrieve a good number of words and short phrases. Today, he says, the teachers from outside Chakali do not have to learn the language, or do not bother trying, as almost everyone can speak Waali in the village.

2.5.2. Mole Game Reserve


I would like to suggest that another blow to Chakali linguistic vitality is the creation of what is known today as the Mole National Park. In 1958, the Mole Game Reserve was created, caging the traditional hunting grounds of many minority groups of the area (e.g. Hanga, Vagala, Tampulma, and Chakali). First intended as a game clearance area for tsetse fly control in the early 1950s, it was officially made a National Park in 1971 under the Wildlife Reserve Regulations (Clifford 2003:2). It is situated a few kilometers south of Chasia, Ducie, Gurumbele, and Holumuni (see Figure 1).


Wildlife Division (2005:17) reports that in the 1960s the inhabitants of five villages in the southern part of the park were resettled elsewhere. No villages of the northern part of the park are mentioned, and no mention is made of the names of the villages, who the people were, and where the people went. In Figure 2, the map which appears in Goody (1954) reveals that what is known today as the West Gonja district, in which the Mole National Park is situated, was formerly an area populated mainly by Tampulmas. In fact, since there are no records on the people inhabiting this area, Goody’s map must be interpreted with care. [27]

Figure 2: Map showing the tribes of the Northern Territories of the Gold Coast west of the White Volta Source: Goody (1954: 2) (Adapted JAB)


What I was told about this specific area is that the people living in the villages of Nyanga (ɲã́ŋɛ̃̀), Buge (bùgè), Gbantala (gbàntàlà), Kau (káʊ́), Saala (sààlá), Gbanwale (gbúŋwálɛ́), and Konkori (kɔ́nkɔ́rɛ̀) were removed by force from the area some time in the period 1963-1964  [28]. These were either Tampulma or Vagala settlements whose inhabitants in many cases had relations with the Chakali population of the north, through family ties and/or common shrines. Today, most of the vestiges in the park are of villages whose inhabitants were forced to move by the government. The population dispersed in all directions; some went southwest to villages like Jentilpe, Tuna or Sawla, some went south to Damongo, some others went east to Daboya or Bawena. For instance, the inhabitants of the Tampulma village Gbanwale were expelled in 1964  [29]; some went to Ducie, some to Gurumbele, while some others went to Jentilpe and other villages in Vagala land. Today Gbanwale can refer to three locations in a confined area; the name of a section in Ducie inhabited by these Tampulma migrants, the original deserted village (approx. 25 km south of Ducie), or a wildlife ranger camp belonging to the Mole National Park named after the original village.


Without going into detail about the history of Mole National Park, the point being made is that the creation of the park restructured the livelihood and the social network of these people. Losing their main hunting ground had a significant impact on their hunting activities and, consequently, their diet. As hunting and farming within the boundaries of the park became illegal, what many people could identify with disappeared. And since Mole National Park cannot in practice be accessed by local communities as crossing its boundary is treated as encroachment and offenders are tried in court, established social networks in the area have been disrupted. Thus, the Chakali, Tampulma, Vagala, and Hanga gradually fostered contacts in other directions with people who are mainly Muslim, have Gonja, Waali, and Dagbani as their first languages, and have a centralized mode of political administration. The former contact areas were more or less similar to each other: local religions based on shrine worship and masquerades as religious pillars, Grusi as language type, and an acephalous mode of administration. The consequences of building relations with their new neighbours are apparent: the introduction and fast growth of Islam, the new (and unequal) socio-economic relationships entertained with the Gonja, Waala, and Dagomba, the changes in gender roles, the gradual loss of traditional ecological knowledge, and the view held by modern authorities and the local youth that acephalous administration is backward and inefficient. I believe that the cost of adapting to new social circumstances and changing lifestyle is reflected in both their linguistic and cultural identity.

2.6. Why is Chakali still spoken?

2.6.1. Isolation and development


One factor favouring language maintenance is the combination of isolation and lack of development (read ‘westernisation’). The estimation in § 2.2 shows hints of a correlation whereby the further east one goes in the district, the safer the language is. Correspondingly, basic social and economic infrastructure decreases as one travels eastwards. One example is the road infrastructure, which in that part of the world is intimately connected to commercial relations and language contact. At the present time the road reaches Gurumbele, but until 2009 it stopped in Ducie, and until the 1970s it stopped at Tuosa. Before the 1980s, Katua and Ducie were the main markets, and around in 1980s a single lorry per week could reach Motigu and Ducie, and only in the dry season. The 6-day interval markets of Katua and Ducie attracted people from all Chakali villages and beyond, so people of Motigu and Gurumbele would go and trade at these markets. However, nowadays, large quantities of goods can be transported by lorry from anywhere in Chakali, so the much larger markets in Bulenga and Wa are preferred. The same ‘east-gets-least’ argument goes for the positive attitude of posted teachers and their willingness to work, the actions and presence of (non-) governmental organisations, mobile networks, electricity, etc. The fact that Ducie and Gurumbele are identified by inhabitants of the district as the locations where one can hear the language in its purest form indicates that the extreme east of Chakali land is typically seen as home of the places where Chakali is best preserved in folk linguistic terms, and the places which in turn are the most isolated and least developed of the district.

2.6.2. Indigenous religious beliefs and practices


A second factor that is favorable to linguistic vitality is religious beliefs and practices. It is very significant that Chakali indigenous relics, beliefs, and practices exceed Chakali land, ethnicity, and language. The Chakali are known to practice Sigmaa, an African traditional religion in which masks, worn by men, are at the centre of a ritual system comprising dance, music, songs, etc. (Dougah 1966Roy 1987Poppi 1993). Although Sigmaa is found in other ethnic groups of the area, many people I talked to claim that its source is the Chakali area  [30]. Whatever the truth concerning the origin of transmission, which would be a daunting task to prove, Sigmaa is probably more active in Chakali than anywhere else in northern Ghana; many Vagala and Tampulma communities report having stopped using masks, and just a few Waala communities are still performing. The logical reason is that Sigmaa dancers only come out for funerals of society members, and because of mass conversion to Islam and Christianity, which typically do not tolerate participation of converts in such activities, Sigmaa members are simply becoming scarce in many areas.


Also, the indigenous medicinal shrines in Sogla and Gurumbele and other ancestral shrines are specific systems of religious worship which appeal to people outside Chakali. Together Sigmaa and the two indigenous shrines are sources of pride for the locals. Whereas most of the region has converted to modern religions, Chakali indigenous religious beliefs and practices have survived rapid socio-cultural transformations. Since it is unthinkable to communicate with the ancestors in another language than Chakali, and since it is practically impossible to separate the religious from the social, the language must be used to engage in almost every aspect of everyday life. This factor is responsible for the apparent correlation between the Islamization and ‘Waalization’ on the one hand, and the vitality of the language on the other. Although data is needed to quantify the claim, the more one goes east, the fewer individuals would call themselves Muslims. Despite that, while indigenous religious beliefs and practices are still very strong in Gurumbele and Ducie, they are quickly fading away in other villages. Again, many young Muslims entertain the idea that Sigmaa performances and Islam are incompatible. Therefore, when the elders of Motigu solicit a neighbouring village to provide dancers and drummers for funeral ceremonies, it is a sign associated with a dying culture. Today the borrowing of dancers and musicians is commonplace, and the sources of those men are mainly Ducie and Gurumbele.


However, indigenous religious beliefs and practices cannot be the only reason for Chakali’s vitality because the language was estimated in §2.3 to be severely endangered in Sogla, but safe in Gurumbele. One can safely say that the two most popular shrines in Chakali land are Kʊ̀ɔ̀lɪ̀ɪ̀ in Sogla and Dààbàŋtólúgú in Gurumbele. The shrine in Sogla receives two to three times more visitors than the shrine in Gurumbele, and Sogla is located closer to Bulenga than Gurumbele, but these are probably minor factors for the vitality of Chakali in the two communities. I believe that there are two factors that have a significant influence on the vitality of the language in both communities: (i) the large number of strangers settled in Sogla, which is not the case in Gurumbele, and (ii) their respective attitudes to the village’s chief. Let us look at these factors in that order.


Strangers come to Sogla because Kʊ̀ɔ̀lɪ̀ɪ̀ cures people and can be replicated and exported; Dààbàŋtólúgú also cures people but cannot be replicated. At any time, there are on average 5 to10 foreign individuals being cured at Sogla. Only rarely is more than one person being cured in Gurumbele. The Busalabanɪɪ section of Sogla (see Table 5), established by an ex-patient (from Busa), is entirely Waali speaking, something which is not found in any E2 or E3 villages, except in a chief’s compound when the chief and family members are Waala. The only foreigners in Gurumbele are the migrant farmers, who stay at a certain distance from the village, and the Fulani, who do not mingle much. These facts suggest more intense contact with non-Chakali speaker in Sogla than in Gurumbele.


What do the attitude towards the chieftaincy and the language’s stability have in common? In Gurumbele and Ducie there are absentee chiefs, that is, a chief has a regent in the village to represent him. This has not always been the case, but ever since the annexation to the Busa gate around 1943 (Daannaa 1992:324), the Waala chiefs of Ducie and Gurumbele do not live in their respective villages. The distance to Wa and the life conditions were, and still are for many, simply uninviting for a ‘big man’ from Wa. As a chief never settles alone, Ducie and Gurumbele are thus free from members of royal families, unlike Katua and Sogla, for instance. The chief’s representative in Ducie is an ‘uncle’ – the son of a Chakali woman and Waala man –, a simple man who has his own farm, mingles with the others, and speaks Chakali. He is well-respected and, crucially, is not feared. By contrast, in Gurumbele, the chief’s representative is a Waala who is not recognized by the landlord and the elders. Locals do not even go to greet him in the morning. Hence, in Ducie and Gurumbele villages, chieftancy has in recent years no effect whatsoever on social and administrative affairs, unlike in other villages, like Sogla, where Waala chiefs settled and are now an integral part of village life and decision making.

2.6.3. Marriage trends


Marriage trends is a double-edged factor: depending on the location and the generation, marriage trends have had positive and negative effects on the vitality of Chakali. Hence different consequences on the linguistic landscape have emerged. This section brings together some observations.

By setting up a boundary between the Chakali people and their southern neighbours and by scattering the people with whom they have had a firm socio-historical and cultural network, the creation of the game reserve and the evictions of 1963-1964 are believed to have had impact on marriage trends over time. The Ducie elders with whom I discussed the topic told me that when they were children, around the 1950s and 1960s, they knew no more than three Waali-speaking women living in Ducie. One was the wife of an indigenous man, and the two others were related to the absentee chief of the time and thus lived in the chief’s house. They told me that before the 1963-1964 eviction, the men from Ducie preferred to choose their wives from within the village or from Gbanwale (evicted Tampulma village), Nyanga (evicted Vagala village), Buge (evicted Vagala village), Motigu (Chakali), Katua (Chakali), Tuosa (Chakali)  [31]. In general Ducie men did not marry women from Gurumbele; a prescribed purpose of their bilateral descent system disallows the union of a man to any of his mother’s sister’s daughters (mààwìé) and most women in Gurembele were related to potential husbands in Ducie in that very way. The elders could also remember women who originated from the Hanga villages of Jumfraanu (dʒɪ́mfʊ́ráánʊ́) and Murugu, and the Tampulma villages Saala (evicted Tampulma village) and Kau (evicted Tampulma village). When such information is taken into account, Ducie men ultimately married speakers of Chakali, Vagala, and Tampulma. Factors like spatial, cultural, linguistic, and social proximity must have been significant. Because of the linguistic proximity among the three Southwestern Grusi languages, there is less effort involved in learning Chakali for a Vagala or Tampulma woman than for a Waala woman, so it is entirely possible that new Vagala and Tampulma-speaking wives learned Chakali quite rapidly by vocabulary replacement and other strategies to make the idiolect resemble the language spoken by their husband, in-laws, and neighbours.

The marriage trend in Sogla was explained to me by the late landlord of Sogla Ŋmɛnɪ Bikpoŋ Maafo in 2008. He said that when he was young, in the 1950s, there were few women available. Ducie, Motigu, and nearby Vagala and Tampulma villages were the sources of the wives who came to Sogla. When it was difficult to find a wife, a man would try to get a woman from Mangwe, Busa, Guropie, etc., i.e. Waali-speaking villages. However, Sogla, like all villages in Chakali, did not appeal to foreign women because the more a village grew demographically, the less water a woman could fetch in a day. Besides, he said, not just any woman who could walk and carry water in those conditions like a Chakali woman could. A reason for the interest in Waala women he gave was that, previously and still today, Chakali women are free to divorce whenever they want; they just pack their things and leave. Such behaviour is not common amongst non-Chakali women of the area, so a Waala woman is a safer choice from the male perspective.


While Waali-speaking women may have refused to live in remote Chakali villages in the not so distant past, today the situation is different. Waali-speaking women see the Chakali villages, even if remote, as a proper place to stay: all villages are equipped with boreholes, so access to water is not such a critical factor in a woman’s decision as it was. The road is in relatively good condition and almost daily transport to and from Wa is available. Mobile networks and electricity are slowly entering Chakali from the west, and the vast majority can speak and understand Waali. Women from places like Chasia, Jayiri, Tuosa, Chagu, Bulenga, Gurupie, Mangwe, and Jirapa do marry men from Chakali-speaking villages, settle there, and speak Waali, Bulengii, and Dagaare to their children.


However, I observed that the matrimonial trend in Sogla is in sharp contrast with the one in Ducie. A survey intended to investigate the origin of wives was carried out in Ducie in 2010. I asked 112 households, selected at random, the origin of the married women of the household. The results reveal that 88% of the wives in Ducie are from Ducie. Similar surveys were not conducted anywhere else, but I believe that such a practice is not prevalent in smaller villages. The reason is that, an individual cannot marry from his/her own clan, and clans often match sectional divisions. Thus, the more sections there are in a village, the greater the chances for intra-village marriages.


The stability of Chakali can thus be viewed as partly due to the origins of the new wives entering a village. Therefore, the people of small villages like Sogla or Gurumbele tend to marry from outside the village. Most of the wives in Gurumbele are from Ducie, some others from Motigu or from Tampulma villages such as Holumuni, Gbantala, or Bawena, all of them majoritarily Southwestern Grusi-speaking villages. I was told that the recent trend in Sogla is to marry from within, something which was unthinkable 30-40 years ago as Sogla consisted of 3 to 5 houses, or to marry from the Waali-speaking villages to the west. Finally, another reason why Waala women are more in demand nowadays is that young men who convert to Islam prefer Muslim women, and these women are concentrated in Waali-speaking villages.


Based on GSS (2012: 103), it is estimated that approximately 47% of the Chakali speaking population are under 18 years of age. This means that the next 20 years will be decisive since it will be a period where Chakali men and women choose partners and establish families, the primary sociolinguistic unit to which a child is exposed. In terms of exposure, I already argued in § 2.2.4 that school management consciously suppresses Chakali. I cannot imagine that a family would suppress speaking Chakali at home, but the linguistic background of each partner and the linguistic dynamic at home is extremely significant for a child’s exposure to Chakali. More work is needed to provide support for the claims being made, but marriage trends in Ducie and Gurumbele are encouraging for the vitality of Chakali, while they are much less encouraging in other Chakali speaking villages.

3. Conclusion


Based on the result of UNESCO (2003)’s ‘nine factors’, an assessment of Chakali’s vitality has been presented in § 2.3 . I have considered factors which work against a thriving use of the language. Despite all of these factors, Chakali is still spoken and I have suggested some reasons for that. Grenoble & Whaley (1998:42), inspired by Brenzinger et al. (1991), state that the “most immediate threat to indigenous languages in Africa is not the language of European conquerors but other indigenous languages”. Being able to communicate in two or more languages is common among inhabitant of Ghana’s Upper West Region and Africa generally, both in urban and rural populations. Thus acquiring a regional lingua franca like Waali, or a national language like English, does not pose an imminent threat to the ‘health’ of the Chakali language. Still, when Waali is used in the village in the absence of Waali speakers, one wonders why. All development-related domains of communication have been shown to be domains where languages other than Chakali were used, favoured, and sometimes imposed.


Wendel & Heinrich (2012:158) write that the changing socio-economic basis of language ecologies and the relation of domination and inequality in contact between communities are at the heart of change in language ecologies. This is the picture I depicted in the previous sections: I have argued that the creation of a game reserve and the forced removal of the Chakali’s neighbours caused parts of the inter-village social fabric to disintegrate, and consequently caused substantial social restructuring, at the same time depriving inhabitants of their hunting territory; that formal education, through the school management, depreciates and devalues Chakali, and also imposes languages foreign to the linguistic ecology; and that the imposition of Waala chiefs has given an unprecedented chance for the Waali language, as well as for a different political economy to enter Chakali villages. All of these are circumstances beyond the Chakali communities’ control. The stability and safeness of Chakali in Ducie and Gurumbele is encouraging, yet more puzzling than the general prediction that the language may one day not be spoken anymore. I have proposed factors which are believed to be responsible for the stability of the language, factors which are nevertheless very fragile in my view.


Even though the 1992 Constitution of the Republic of Ghana should protect and promote basic linguistic rights of Ghanaians  [ 32 ], Ghana has not ratified the Convention on the Protection and Promotion of the Diversity of Cultural Expressions  [ 33 ], so policies which could improve the situation are not in place. One may wonder how long the language will survive under the current conditions. A correlation between development and language shift was hinted in several places in the article. Fortunately for the language, but not for the people, there are deficiencies in the provision of basic infrastructure and services in the area. The case of Chakali is no different than that of other small African languages: the ‘threat’ to Chakali may not necessarily be another language, but a general problem of its speakers coping with a slow and ostensibly irreversible social, economic, religious, and administrative shift. The Chakali language and culture face a process of slow attrition, and the individuals and organizations who could lobby in their defence have yet to speak out.


Akyeampong, Kwame, Jerome Djangmah, Abena Oduro, Alhassan Seidu, and Frances Hunt 2007

Access to Basic Education in Ghana: The Evidence and the Issues. Project Report ISBN: 0-901881-12-0, University of Sussex, Falmer, UK and Winneba, Ghana. Country Analytic Review (CAR), Ghana

Anyidoho, Akosua and M.E. Kropp Dakubu 2008

‘Ghana: Indigeneous Languages, English and an Emerging National Identity’. In: Simpson, Andrew (ed.) Language and National Identity in Africa. Oxford University Press

Awedoba, Albert K. 2006

‘The Peoples of Northern Ghana’. Tech. rep., National Commission on Culture. http://www.ghanaculture.gov.gh/modules/mod_pdf.php?sectionid=593 (Accessed 25.10.2013)

Bendor-Samuel, John T. 1965

‘The Grusi sub-group of the Gur languages’. Journal of African Languages 2,1:47–55

Blommaert, Jan 2004

‘Writing as a problem: African grassroots writing, economies of literacy, and globalization’. Language in Society 33,5:643-671

Boateng, E. O. 1984

The 1984 population census of Ghana: Special report on localities by local authorities. Upper West Region. Tech. rep., Statistical Service

Bodomo, Adams 1994

‘Language, History and Culture in Northern Ghana: An Introduction to Mabia Linguistic Group’. Nordic Journal of African Studies 3,2:25-43

Brenzinger, Matthias, Bernd Heine, and Gabriele Sommer 1991

‘Language Death in Africa’. In: Robins, Robert H. and Eugenius M. Uhlenbeck (eds.) Endangered Languages, pp.19–44. Oxford: Berg

Brindle, Jonathan Allen 2011

Aspects of the Chakali language. Ph.D. thesis, Norges Teknisk- Naturvitenskapelige Universitet (NTNU)

Clifford, Bugli 2003

Prospects and challenges of game and wildlife in Ghana: A case of Mole game reserve in the Damongo district. BA Dissertation in Integrated Development. University of Development Studies, Wa

Daannaa, Henry Seidu 1992

The Impact of State Law on Custom and Leadership in a Post-Colonial State: A Legal Historical Case Study of Centralised Wa and Acephalous Chakali in Northern Ghana. Ph.D. thesis, University of London, http://etheses.lse.ac.uk/2086/1/U553897.pdf(Accessed 15.06.2015)

Daannaa, Henry Seidu 1994

The Acephalous Society and Indirect Rule System in Africa. Journal of Legal Pluralism 34:61–85

Dakubu, Mary Esther Kropp 1976

On the linguistic geography of the area of Ancient Begho. In Trutenau, H. Max J., ed., Languages of the Akan area: papers in western Kwa linguistics and on the linguistic geography of the area of Ancient Begho = Transactions of the Linguistic Circle of Accra, III, Mitteilungen der Basler Afrika Bibliographien vol. 14, pp.63–91. Basel: Basler Afrika Bibliographien

Dakubu, Mary Esther Kropp 2005

Report of an exploratory trip to the Chakale-speaking area. 25-26 July 2005. (Unpublished manuscript)

Delafosse, Maurice 1912

Haut-Sénégal-Niger (Soudan français). Paris: Larose

Dougah, J. C. 1966

Wa and its people. Legon: Institute of African Studies, University of Ghana

Duperray, Anne-Marie 1984

Les Gourounsi de Haute-Volta. Conquête et colonisation, 1896-1933. Stuttgart: Franz Steiner Verlag

Fuseini, Moses Naiim and Amadu Musah Abudu 2014

Participation in Literacy Programmes of Non-Formal Education in the Wa Municipality: The Role of Inducement Factors. Global Educational Research Journal, 2(1)

Goody, Esther N. 2005

Contexts of Kinship: An Essay in the Family Sociology of the Gonja of Northern Ghana. Cambridge University Press

Goody, Jack 1954

The Ethnography of the Northern Territories of the Gold Coast, West of the White Volta. London: Colonial Office

Goody, Jack 1967

‘The over-kingdom of the Gonja’. In: Forde, Daryll and P. M. Kaberry (eds.) West African Kingdoms in the Nineteenth Century, pp.179–205. London: Oxford University Press

Grenoble, Lenore A. and Lindsay J. Whaley 1998

‘Toward a typology of language endangerment’. In: Endangered languages. Current issues and future prospects, pp.22-54. Cambridge: Cambridge University Press

Grenoble, Lenore A. and Lindsay J. Whaley. 2005

Saving Languages: An introduction to language revitalization. Cambridge

GSS, Ghana Statistical Service 2002

Population and Housing Census 2000: Summary report of final results. Tech. rep., Ghana Statistical Service, Accra

GSS, Ghana Statistical Service 2008

Ghana Demographic and Health Survey 2008: Preliminary Report, Tech. rep., Ghana Statistical Service and Ghana Health Service, Accra, Ghana

GSS, Ghana Statistical Service 2012

Population and Housing Census 2010: Summary report of final results. Tech. rep., Ghana Statistical Service

Hinton, Leanne and Ken Hale (eds.) 2001

The Green Book of Language Revitalization in Practice. New York: Academic Press

Holden, J. J. 1965

‘The Zabarima of North-west Ghana’. Transactions of the Historical Society of Ghana 8:60–86

Kanganu, Daniel and Jonathan A. Brindle 2008a

beŋ tʃelle (The law breaker). Trondheim: Institutt for språk- og kommunikasjonsstudier (ISK), NTNU

Kanganu, Daniel and Jonathan A. Brindle 2008b

bɪpɔlɪɪ naŋzɪmatɪɪna (The Clever Man) . Trondheim: Institutt for språk- og kommunikasjonsstudier (ISK), NTNU

Kleinewillinghöfer, Ulrich 1997

The Gurunsi languages: A summary of the state of art. In: Von Roncador, Manfred and Kerstin Winkelmann (eds.) Gur Papers/ Cahiers Voltaïques2:43–52. Bayreuth: Universität Bayreuth, Afrikanistik I

Köhler, Oswin 1958

'Zur Territorialgeschichte des östlichen Nigerbogens'. Baessler-Archiv, Neue Folge 6:229–261

Levtzion, Nehemia 1968

Muslims and chiefs in West Africa: a study of Islam in the Middle Volta Basin in the pre-colonial period. Oxford: Clarendon Press

Lewis, M. Paul, Gary F. Simons, and Charles D. Fennig 2014

Ethnologue: Languages of the World. SIL International, 17 edition

Lüpke, Friederike. and Anne Storch. 2013

Repertoires and Choices in African Languages. Language Contact and Bilingualism [LCB]. Boston, Berlin: De Gruyter

Manessy, Gabriel. 1969a

Les Langues Gurunsi: essai d'application de la méthode comparative à un groupe de langues voltaïques, vol. 1. Paris: Klincksieck (SELAF)

Manessy, Gabriel 1969b

Les Langues Gurunsi: essai d'application de la méthode comparative à un groupe de langues voltaïques, vol. 2. Paris: Klincksieck (SELAF)

Manoukian, Madeline 1951

Tribes of the Northern Territories of the Gold Coast. London: International African Institute

Naden, Tony 1989

‘Gur’. In Bendor-Samuel, J. (ed.) The Niger-Congo Languages, pp.140–168. University Press of America

Nicolas, François 1952

‘La question de l’ethnique ’Gourounsi’ en Haute-Volta’. Africa 22:170–172

Poppi, Cesare. 1993

‘Sigma! The Pilgrim’s Progress and the Logic of Secrecy’. In: Nooter, M. H. (ed.) Secrecy: African Art that Reveals and Conceals. New York: Museum for African Art

Rattray, Robert Sutherland 1932a

The Tribes of the Ashanti Hinterland, vol. I. Oxford: Clarendon Press

Rattray, Robert Sutherland 1932b

The Tribes of the Ashanti Hinterland, vol. II. Oxford: Clarendon Press

Reimer, Jean and Regina Blass. 1975

Chakali survey report. GIL

Roy, Christopher D. 1987

‘The spread of masks styles in the Black Volta Basin’. African Arts 20,4:40–90

Salih, Mohammed Bin 2008

The Kingdom of Wa: Elucidation of our origins and settlements. Accra: Salihsons

Tauxier, Louis 1921

Le noir de Bondoukou. Paris: Ernest Leroux

Tauxier, Louis 1924

Nouvelles notes sur le Mossi et le Gourounsi. Paris: Larose

Tompkins, Barbara, Deborah Hatfield, and Angela Kluge 2002

Sociolinguistic survey of the Chakali Lan- guage Area. Tech. rep., SIL International

UNESCO, Ad Hoc Expert Group on endangered languages 2003

Language Vitality and Endangerment. Tech. rep., International Expert Meeting on UNESCO Programme Safeguarding of Endangered Languages, Paris, http://www.unesco.org/culture/ich/doc/src/00120-EN.pdf (Accessed 25.10.2013)

Wendel, John and Patrick Heinrich 2012

‘A framework for language endangerment dynamics: The effects of contact and social change on language ecologies and language diversity’. International Journal of the Sociology of Language 218:145–166

Wildlife Division 2005

Mole National Park Management Plan. Tech. Rep. WDSP Report No 54, Forestry Commission

Wilks, Ivor 1961

‘The Northern Factor in Ashanti History: Begho and the Mande’. The Journal of African History 2,1:25–34

Wilks, Ivor 1989

Wa and the Wala: Islam and Polity in Northwestern Ghana. Cambridge University Press

Wilks, Ivor, Nehemia Levtzion and Bruce M. Haight 1986

Chronicles from Gonja: A tradition of West African Muslim historiography. Cambridge: Cambridge University Press

[1] I thank Mary Esther Kropp Dakubu, Tyson Farrell, Albert Awedoba, Rachel Selbach, and Felix Ameka for their comments on an earlier version. Special thanks to Per Wirehn for his help in the design and preparation of the map in Figure 1. I would also like to thank two anonymous referees for their valuable advice. This paper is a revised and expanded version of Chapter 2 of my Ph.D thesis ( Brindle 2011 ).

[2] See also Rattray (1932b:516-7).

[3Brindle (2011:46) explains why Chakali is not spoken in Bulenga, Chagu, Bisikan, Kandia, Dupari, and Gilan. These villages have been excluded from the study because Wilks (1989:17) and Salih (2008:64) suggest different origins for their ancestors, and elders from Bulenga, Ducie, and Tuosa are quite clear in saying that Chakali is not and has never been an ‘identity language’ in these villages, perhaps only spoken by a minority, mainly women, who were raised in Tiisa, Sogla, Tuosa, Motigu, Ducie, Katua, or Gurumbele. As far as one can tell the common denomination originates from a defense cooperation between the people of Bulenga and surrounding villages, including Chakali speaking villages, then reinforced by a political association imposed by the colonial administration (see § 3.1.1 ).

[4Rattray (1932b:525) writes that the Awuna, a Kasem dialect also known as Aculo (Naden 1989:147), has earned its appellation based on a habit of “prefacing an observation with the words” a wun a ‘I say’. It is indeed the case that a Chakali can open a sentence with  ̀ ŋmá kaà, ... ‘I say that’, with the last word being a complementiser introducing a new clause. To hear the Ghanaian English opening expression ‘I say eh, …’ à sé ɛ͂̀ɛ͂̀ in Wa is not unusual.

[5Goody (1954:14-15) reports a ‘Black Waala’ and ‘White Waala’ division, the former being the dominated group, that is commoners and pagan, while the latter being the dominant group, that is members of the chiefly lineage and Muslim. Tony Naden (p.c.) confirmed to me the existence of ‘Black Dagomba’, with no correlative ‘White’, and suspected it to refer to the descendants of the original inhabitants in contrast to the aristocracy, therefore roughly Black = ‘commoner’ vs. White = ‘aristocracy’. In the case considered here, the interviews with ‘Black Chakali’ individuals tell us about the resources people have available for telling their world and creating an identity. Assuming that the connotation of the division black/white is ruled/ruler, dominated/dominant, or commoner/chief, then it appears that despite being labeled as ‘black’, one can exploited this sense of the concept in order to associate one’s group with more positive cultural implications. This social categorisation is in need of further study.

[6] A Sogla-Motigu lect could be situated somewhere in the middle of such continuum. The Katua lect is known to be highly influenced by Pasaalɪ. One version of the creation of Katua, Tiisa, and Tuosa is presented in Salih (2008:74-82).

[7] Abbreviations used are ART – article, HAB – habitual, FOC –focus, sg –singular, IDENT – identificational verb, and ATR – advance tongue root.

[8] One reviewer found intriguing the presence of the etymologically Bambara/Dyula word súŋgúrú ‘mature girl’ in the Gurumbele lect since “no western Mande language is spoken in Ghana”. The details need to be worked out but there must have been contact with Mande languages in the area. To the best of my knowledge the word súngúrúng ‘adolescent girl, virgin’ is found in Gonja, as well as Vagla and Tampulma. In addition, historians have documented Mande/Dyula influx in today’s northern Ghana, e.g. the migration waves from the Northwest started around the 17th century (Wilks 1961:26-27) and several Mande groups settled in and around Wa during the same period (Salih 2008:162-171, Levtzion 1968:145), so it is more a possibility than a curiosity. Also, there is a western Mande language known as Ligbi (iso 639-3:lig) spoken in the area east of Sampa and south of Banda in Ghana (Dakubu 1976:64). The language apparently arrived there in connection with the gold trade to the north, before the Europeans diverted it to the south (M.E.K. Dakubu p.c.; see also Levtzion 1968:8 and Goody 1954:12).

[9] The document listed as Reimer, Jean & Regina Blass 1975. Chakali survey report. GILLBT. unpublished in Tompkins et al. (2002) could not be found at the GILLBT headquarters in Tamale when I visited in 2008, nor obtained from one of its authors, the late Regina Blass.

[10] To save space in the body text, I do not reprint the nine factors and their descriptions and gradings. The reader can access the document by using the link that appears in the bibliography.

[11] In discussing the nine factors below, a word in italics reproduces an evaluative term used in UNESCO (2003) and bears the sense found therein.

[12] “In the 1984 General Census (Boateng 1984) the Chakali were numbered at 4,000. Using this 1984 figure and factoring in a 2.3% annual growth rate would make the current population an estimated 5,000.” (Tompkins et al. 2002:6). I could not findBoateng (1984).

[13] For the survey in Katua (2008-03-11), I was assisted by Idriss Haruna and Dakurah Francis. For the three other villages, Jamila Osman and Mr. Abraham assisted me (2008-01-15).

[14] A household is defined as a group of people who normally eat together and has its own cooking area in the compound.

[15] At the same time, it is debatable whether or not this subtraction is motivated. These are indeed generalizations, but my observation is that the Ducie residents waste no effort in speaking another language than their own to the Fulani. In consequence, after some time Fulani adults do speak a little Chakali, and their children, who do not attend school, learn Chakali with a relatively good proficiency. Contrastingly, in Gurumbele, the Dagaaba settlers, both adults and children, do not attempt to learn Chakali and the residents speak Waali to them. The children do attend school and communicate with Chakali adults and children in Waali. Again, both Fulani and Dagaaba settlers do not stay within the community but in their own section, at a certain distance from the village. By ‘Waala settlers’, I do not include married women who were raised in a Waali speaking village, but only the few individuals who were not raised in Chakali speaking village and live in the Waala chief compound, if there is a chief on site. The decision is based on the type of exposure to Chakali and the will to learn the language.

[16] I do not discard the population of Ducie’s section Gbanwale. Even though they speak a variety of Tampulma very close to Chakali, the majority of Tampulma children and young adults converse using Chakali outside their section.

[17] The population of the Upper West Region (i.e. 573,860) is taken from GSS (2002) with an annual growth rate of 1.7%. The Ghanaian population (i.e. 23,350,927) is taken from the World Bank, World Development Indicators (http://data.worldbank.org/country/ghana. Accessed on July 26, 2010). In GSS (2012:21, 103) the population of the Upper West Region was estimated at 702,110 and the one of Ghana at 24,658,823.

[18http://waeast.ghanadistricts.gov.gh/?arrow=atd&_=114&sa=4649 (Accessed on July 26, 2010)

[19] Materials are available in ten of the 81 Ghanaian languages (Lewis et al. 2014), i.e. Akan, Ewe, Ga, Dangme, Nzema, Dagaare, Gonja, Gurenne, Kasem, and Dagbani. The total number of Ghanaian languages varies across sources.

[20] Ghana has ratified the African Charter on Human and Peoples’ rights (http://www.achpr.org/ instruments/achpr/#a2) and article 2 should inform policies and monitoring against linguistic discrimination, at least in governmental institutions like primary schools.

[21] One explanation is that the villages under discussion were not covered by mobile phone networks at the time of this observation. Local assembly men are thus taken by surprise when representatives of the District Assembly pay a visit to the villages.

[22] In April 2014 there was electricity in Bulenga and utility poles reached Tuosa.

[23] I suspect that they have in mind the GILLBT literacy programs for Vagala, Tampulma, Pasaale, Sisaala, Hanga, and Gonja, and the Baptist Mid-mission literacy program for Waali.

[24] See references on Grusi in § 1.1 , and Rattray (1932b: 98), Manoukian (1951:2) and Wilks (1989: 13-17) for views on the heterogeneity of ‘Waala’ as an ethnic group.

[25] A value within square brackets applies to E1, E2, and E3 villages as a whole.

[26Wilks (1989:133) writes: “Zabarima occupation of Ducie occurred probably early in May 1897.” On the same event, see Holden (1965: 6-77).

[27Rattray (1932b:515-524) may be an exception, but the place names mentioned in this article do not appear in his text. The map in Köhler (1958:56-58) shows four small Tampulma areas, and only two small areas appear in the map of Manoukian(1951), which differs significantly from the one in Goody (1954).

[28] Nyanga was the first Gonja capital (Wilks 1989:121). The Yagbonwura’s skin was moved from Nyanga to Damongo in 1944 (Goody 2005:17). I am still to visit the abandoned villages Kau and Saala to obtain their coordinates. In Figure 1, the name and location of Salawaku come from the Ghana Wildlife Division (Wildlife Division 2005), but I am told that it could be the villages of Saala and Kau mistakenly put together. Unlike the other villages Gbantala was moved to its current location in the 1990s.

[29] Reported by Mr. Dɔga, an elder of Ducie’s Gbanwale, eye-witness of the exile to Ducie.

[30Sigmaa is very complex and cannot have one point of origin. It is well-known that those specialising in the carving of masks are Tampulmas, but a synchronic or diachronic study of Sigmaa masks, and of their exact relation among the relics found within today’s Upper West Region as well as those found in similar traditions as far as the Dogon (Mali) and Bobo (Burkina Faso) traditional areas, is yet to be made.

[31] Like many other groups of the area, they generally practice patrilocal residence whereby the woman comes to settle at the husband’s village.

[32] “Every person is entitled to enjoy, practice, profess, maintain and promote any culture, language, tradition or religion subject to the provisions of this Constitution” (Article 26-1) and “[t]he State shall foster the development of Ghanaian languages and pride in Ghanaian culture” (Article 39-3).





There are no comments yet.

Do you have any additions or comments?

Submit comment.