203 BILINGUAL! SM AND LANGUAGE SHIFT IN (paper r e c e i v ed 1979) A«K. Awe dob a* In tro due tion . • • The aim of this paper is to outline the nature of •bilingualism and language shift found in Navrongo. "Bavrongo" has several meanirgs, but it should be understood here to refer to that portion of Kasena-Nankani territory under the jurisdiction of Bavropio, chief of Navrongo.2 Territorial units of comparable status axe Paga, Chiana-Katiu, Kayoro, Kolgu, etc. Navrongo, also known, as "Uavoro" and "Navoringu" by -fee Kasena and Nankani people respectively, is believed to have been founded by a man called Butu, who, according to legend, gave the place i ts name. -> Navrongo comprises the following sub-territorial units: t) Havoro 3) Gean 5) Manyoro. 2) Pungu 4) Pinda Each of these sub-territorial units is a grouping of clans and maximal lineages in a continuous state of segmental opposition.* KL -though Pinda and Manyoro now owe allegiance to the Navropio, this used not to be so before the intervention of the British. Navrongo went to war with Paga, Pinda and Manyoro stored out of the conflict. In former df^s when In almost all of these gub-territorial units tw> groups of people the autochthonous group and the immigrant group. The are identified: autochthonous groups have custody of the earth shrines and claim to have lived in caves before the advent of the immigrants who are believed to have taught them to live in mud huts. The immigrants brought along with them same chiefly powers believed to be capable of bringing prosperity to the locality and were therefore accepted by the authoch- thonous group as politico-religious leaders. *Research Assistant, Institute of African Studies, uhiversily of Ghana. 20& 0*1 -^.ayoro,, comprises two important composite clans: Zigia and Gwala bia> There are also groups from Biu, fo and Chuchuliga in Navoro. The Zegia come from Zekko, a place believed to be near Bcngo, in Nankani country. The founder of the clan, Butu, migrated to present day Navoro in the company of his sister. He was later joined ty the founder of Doba, another Zekko man. Today members of the Zegna Clan constitute a majority in a number of village-settlements in Navoro. These include Nogsenia, Wuru, Nangalkinia, Doba and Gcngonia. The Gwala clan is believed by some to be autochthonous* The following ' • ' •• •• - • 5 - - . - . . - • . • • ". village settlements, are occupied by lineages of the Gwala Clan: Korania, Bwaaia, Nawogiia, Bundunnia, Yigbania, Vonania, Gaani and Taampala. Bwania contains attached lineages believed to have originated from Chuchuliga, seven miles west of Navrongo. The land ovoaing lineage in "Wuru is believed to have descended from a cave-dweller* These formerly- cave- dwelling lineages are believed to be of Biu origin. Gaane, Vbnania and Janania are supposed to have elements from Po or Tiebele in Southern Upper Volta» The origin of Mamolo is not very eleaf as in forman ts ar e n ot in agr eemai t en - it» While some con sider the N amato as belonging to the same clan as Vonania and Jaaaania, others relate them to the Wusim clan in Pungu. The Ban in i a lineage of Namolo is a blacksffliiii1 s lineage and is considered an off-shoot of Punyoro in Pun^i* Nayagiia is intercalary between Pungu and Navoro. The lineages inhabiting ihe land there are related to the Wusim sub-^clan in Pun guy but used to consider Namolo and Nangalkinia as their landlords* 0.2 £,HLSJ, ^as a s ^ ' d i - ef v;no owes allegiance to Kavropio. Before British intervention, the Punga chief may have,been of comparable status with Navropio. BDth Punga and Navoro plus other satellites united to fi^at paga. Villages coming under the jurisdiction of J>ungii pio include Tclania, Bavogiia, B^wiu, Manchoro, Nembasaiia, Bisungu, Nyangoa, punyoro, Yiteria and Saboro. The ohiefship of Puhgu is vested in the Bavogiia lineage while the Telania lineage controls the earth shrine* The arrangement is ^mptomatLc of the diverse origins of the lineages in Pungu« Telania ^ -. are b'elievecl to be authochthonous in Pungu* J ancestors, they claim, ®ieir were a (cave-dwell ing grouo. 2 05 •;••:. .. 0*3 £®21 comprises three settlemaats, N an gwao..-, Batiu an d N abwio « The first owes allegiance to 5favropios bit the remaining two come under the jurisdiction of paga Pio. The whole of the territory of Gean belongs to the Adinyina lineage of Nan gwao and since this lineage owes allegiance to Havropio, ax that strength, the whole of Gean has been included under Havrongo in-this discussion* On other grounds Gean. may be considered as an intercalary -unit between Paga and Navrongo. Custody of the earthshrine in Gean vests in the Adinyina lineage* They claim to be the autochthonous groups their ancestor was a cave- dweller of Biu origin. The Kwo Songo lineage controls the office of eldership in Gean. This lineage is affiliated to the Zegia clan in Navrongo; ancestor of the Zegaa clan. i ts ancestor was a (great) grandson of .Batu, the putative Nabwio and Batiu are considered as stranger lineages founded by immigrants from Baloo, Kajelo, Nyannia and Paga respectively- There is also an accessory lineage - s i s t e r 's offspring- who migrated from Naao, now in ruins* to Gean. It is also considered a very old settlement - perhaps as old as 0.4 Pinda is a small entity vis a vis units of comparable political status* Biu. The legend about the founding of Paga states that Panda was already in existence before.Nave, the founder of Paga came into that locality from Kampala* The history of Manyoro is not knova to the writer but tradition in Navrongo relates it to Punyoro in Pungu and Kayoro. All three settlements are traditionally blacksmithing settlements but a curious legend suggests that Kayoro was the cradle from which some blacksmiths migrated eastward to found Punyoro and Manyoro. It can be seai from the brief historical sketch based an oral tradi- tion that Navrongo was a meeting place of peoples of diverse origins and, probably, speaking a variety of languages. We now group the various villages according to dominant language in use and discuss to what extent language shift can be said to have takoa place in some of the villages. 206 •j.O. Languages in Use Languages with significant numbers of speakers in Navrongo include the local languages - Kasem9 Nankani and. Buli and languages of wider conmunication. The latter group of languages include the official lan- guage of Ghana, English, as well as Hausa and Akan. Kasem and Nankani are the most widely spoken languages in Navrongo but although they are spoken side by side sometimes by the same people, they are not mutually intelligible. Kasem is a Grusi language but Nankani, mutually intelligible with Frafra or Gurreni, is a Mooxe-Dagbani language. 8 9 It is a Moore-Gurma language. Buli is the language of the Bulsas whose territory lies to the south- It is not widely spoken west of Navrongo. in Navrongo despite the fact that there is some amount of intermarriage between the people of Navrongo and the Bulsas. It is important to note that in Biu, Buli is widely spoken - perhaps more widely spoken than Nankani which is the other f i r st language there. Some Akan is understood by the people in this corner of Ghana owing to a seasonal migration to southern Ghana (popularly known as in quest of jobs and the "good things of l i f e ". to speak Akan was the hallmark of the "been-to", in Kasena-Nankani t e r r i- tory. ISitil recently Hausa is understood by a sizeable number of Navrongo people, Formerly many learnt Hausa outside Navrongo but there is now anew breed of Hausa users who acquired facility in this language through daily attendance in the market at Navrongo. English is understood and used by the l i t e r a te in a variety of domains- It is used even in %he discussion of topics for which Kasem would be equally, if not more} suitable. For many a l i t e r a t e, Efaglish serves'as a convenient bridge between Ka^em and Nankani, being more neutral. comBCn phenomenon with the l i t e r a te sector of the population but some i l l i t e r a te or marginally l i t e r a te people frequently affect it whenever they wish to put on ajxs. Interlarding is a 1-1 Grouping of Villages The lineages which combine to constitute the various clans in Mavrongo axe based in villages- These villages will now be grouped according to language use- Group "A" villages are monolingual while Group «B" villages axe bilingual. Group "A" villages have been further sub- Inhabitants of Group "Aj" villages grouped into Group "Ai« and Group "A2"« are monoglot Kasem speakers but those of Group "A2n villages speak Nankani. There are also Group "Cw villages where one section speaks Nankani said the other Kasem« •1*1 Group :A1 Tillages Membership of Group »A-|» includes the following villages- •]) Gean ( a ll sections) 2) All villages of Pungu except Punyoro 3) The whole of Pin da 4) Wuru 5) Hegsenia 6) N amolo 7) Nawogaia 8) Yogbania. An examination of oral tradition § suggests that although Kasem is now being spokto and used in nearly a ll domains in these villages. There was a time whai other languages were &l$o used. In the case of Hogsenia andHamolo it was only recently that there wag a complete shift to Kasem. In the K&bagiia section of Nogsenia and in Hamolo nearly everybody above the age of 30 years understands and cm speak Nankani fluently. Many persona^ namea are s t i ll in Nankani but the majority of young people below the age of 30 have Kasem rather than Sankani names- In both villages sacrifices t$ important fetishes and shrines are accompanied by prayers spoken wholly or partly in Kankgni- This is because the aacestors of these people once spoke Naikani rather thai Kasem. •The case of Hogsaiia Is particularly interesting. The founder was Bogse* the sen of Butu. Nogse .vss.3 supposed to have had several wives and many childraa who in turn were the ancestors of the various line&ges in Nogsaiia. Members of the J a ^ i af Bswria iad Bagwe major Xtaaagot |jj Neggfl&ia replaced Kmkaal with fesea pa?oba,|!y & vtry long time age* fh© tt&fntS Of the founders |ait>ia.j Bawia aid Bagvsre ape tmm »M ftot Nankmi aai v*sj§r ftw aembers of these. liB.ea.ggg fi§w speak Mmk8ni» Yet a il agfnliers of lineages agree that %H§f a.*e descendants Of aMmkattl »6fti lit«.» eslaiaticti $.ym_ is ihs<"te th# a©tti§» if p o t h ^i ef #i# fouaders of iiB-eages wea?e lasaaa v^jp## fh§ i&iigiiage of thg moths? of aa aacestor Is howeverj pot ®.©ii^i to aoeomt |§f Wholesale language shift tualess sracceedJng gmeratiaas esdiibited a propaasity to map^ womga who ^pokfc* fihg game laiguage as ^jig moth# 0f the lineage raaee,sto£» The Kabagiia aeotion of Hoggenia ig al§o bflieved 1y |f idiis Is 11 persons in. Nogsmia to have descended frooj Nogs««( ' laikani poxsbably survived there for a longer period because it§ mm folk Qf this lineage exhibited a p asp easier t© na-jgey women #10 ^okeHaikani rather than.Kas.em. Qa -the oths?r hand the Kab&ipia lineage may well be a difft-rmt group alto-geftier* lids perhaps explains t h s ir custody of the earth shrina? They may in faet b§ %e indigaious Btu ^habitants fouad there fcjr the Zekko iimiigrants» According te EWme versions Of ttie Hitu legends the Zekko inrnigraats lived with the Blm oave-dwellgrs for a while before building a separate conjjcund for lineages a^e descendants of a Biu men. than they may have spokaa Naakani oy as f i r st language, giv^i that, avaa todaiy Nmkani and Bill are the main languages spokm to Biu. 13 the Kabagaia ttiemselves. If !£he tentative accotmt of lQngt»ge shift provided above for could be replicated in the case of =raany other Group >1A.1' villages, fhis .would eertajhiy be the ease if furu and Gean were examined from Hie point of view of language shift. Both places ence had Biu elemeats before the advent of -fti© 2egia lineages there. Ihe Jiame IWiru cannot be of Nankani etjnuology because we find'places with similar names in other parts of Eaaemland but none in Nankanl land. The name Ge^i, howevesr, is of Naakani elymology and means "ebony tree" (diospyros mespillformis)» !Ehie name, .w givsi .by the founds? of the Zegia lineage in Ge9n« Yet today Gean is . • 2 09 .. wholly Kasem speaking; a ll personal names are in Kaseia including even the names of imports.t earth shrines l i ke "Ohichira" and ^rpuwura". "Eunzcngp" is probably the oily shrfne with a n on--Kasem name* Nawogpia is K as en*-speaking* The name derives from a Bali phrase Naariwogi ijhich is an. 'insult meaning "long mouth" or "strong mouth'*, aid was acquired through hostile interaction with -Bu.lsa.s- from Chuchuliga in the days ih&t the ancestors of Sawognis,, i.e» Nawoga, lived on the banks 14 of Nawog Tono r i v e r. Like i ts siblings? the ancestors of Hawogiia must have spokga Nanksni. and Bali also? as a result of interact!en with the Balsas* "Whatever language or languages were originally spokai by the ancestors of lawognia, these had beaa replaced by Easem. This shift coincided with the migration of Hamoga and h is lineage from the eastern banks of Tono, also kno?M. as .Nawog - Tono, to their present territory about 1 mile south of Harrange, onto land ence belonging to Eamolo. 1 Pinda is probably the oily Group 'M village inhere Kaseia has sftokai - perhaps ihe cradle of Kasem in the locality of Navrcnpn 15 >2 Gpoup _ It is hard to say whether there are villages in- Havrongp where inhabitants are maioglot Haokani speakers* Possible candidates are however Gaani aid Doba. Much of 2oba is bilingual but people living in compounds bordering -n Kadiga to tho east toad to sp&ak and understand only Uaakani* Doba wi?" be discussed more fully under Group -B villages* Gaani is located about six- miles south ofMavrongo town, at the southern periphery of what is here known as ^avrongo**.. Althou^a the majority of ueople in Gaani probably speak only Bankani, legmd r e l a t es that their founding ancestor was aa imnigrant hunter from Tie bele. a Kasem-speaking territory in the Upper Volta= The founder of Gaani was 1 settled near Lake Kulnaba by a friendly man from Eius probably in the l a t t e r 's capacity as earth p r i e s t. From the mset the cradle of Gaani probably spoke Kasem but inter- uia,-rriage saad gmeral intercourse with Hankani people in ihe neighbourhood soon led to the replacement of Kasem 210 As in Group 'A-] villages wtifirelankani may be retained for use in. a limited number of domains Gaani people-may use some Kasem in certain rituals, particularly in sacrificial prefers to the founding ancestors* A§.ih§u,gh the shift fjQjn Kag«m fe I eaakani may have t%m place a long there Is new an Ift^aati^e 4fc the a h^f.lte, Mahg|fc| Funeral dirges are probably sung in both Kaseifl aid Nankani end so ar'e sacrificial prayers- D^ba people would however compose f@lk $Ehgs in Nankani rather than in Kasem. and. Yigbania constitute a subs-group of Group ••£'• villages* Together with Korania and Nawogiia etc* The authentic" lineage groups in these villages trace descent patrilineally to Gwala. fhey probably spoke Kankani in the past but Kasem now predominates as f i r st laiguage and Nankani-is a second language* Most inhabitants of Bwania and Yigbania •speak .both languages although knowledge of Nankani is declining fast<• Some Bull is also spoken _in Bwania by an accessory lineage there- Nankani personal names s t i ll exist but Kasem ones predominate' The accessory lineage maitioned above inhabits the section of It is attached to the Bay an i a lineage. This Bran i a knova as Agaia^biji^. expression is a Bull one aad suggests the ancillary status of the group concerned" The name Bwan_ia, however, is a Kasem phrase. I t 's ©iymology is unknown to the writer but the belief that it is Kasem is supported by the existaice of the same name in other parts of Kasemland viz- Ghiana. 21 Folktales found in Bwaaia contain Bali songs and Kasoia soags .and "both narrators and l i s t m e rs have no problems singing these songs- The presaice of Buli la folk tales is however the. result of borrowing. 5ater~ marria§§@ bgtwesi Bwania and Chuehuliga, the nearest Buli-spsaking ijexri- iery, **i& the frequent v i s i ts by village farmiaig and funeral parties to eith«8? ttd© has aada it possible to assimilate each others language. Ohuchuliga like Bwmia is bilingual in Buli and Kasem. 21 the multilingualism prevalent in Bwaiia, the people compose folk icngs in Kagem« The choice of what language to compose is may The place where it pays best in terms of rewards to the ia Havrcngo market. Theres Kasem is the predominant language. Ohuohuliga majket is conparatively small aid Saidema market is too far away to attract nusioiais from two villages can be grouped in this category. Th^r see? Naigalkinia aadNaya^ia. . NangalkJnia occupies a texritoay which borders en Nawogj-i& aad VH Ja iije nor-taa, Yogbaiiia aid Ko rani a in the west, Gongtnia in -tfie south aid Nogsataia in the east. In a words it lies south of Kasem-speaking villages But north of NaUcani-speaking villages. that cue half of Nangalkinia speaks Nankani while the othe? speaks Kasem. iJaokani is gpokm as in-gaoup language in the southern half of Naigalkinia i»e» the section occupied by the Afabia lineage but the northern half is mixed* The Nagriu and Yikuri lineages now use Kasem for aost purposes but the Bangia lineage has retained Nankani, which is used occasionally as an in-'group language. It is not therefore surprising Naigalkinia was once wholly Nankaii-speaking. The name of the village derives from the Nankani phrase - naam galisiya "chiefship abounds". The . founder of the village was called Naamgali and was believed to n?-ve been bne of the sons of Butu. The authentic lineage of 5i angalkinia was aNaikani lineage by origin. Almost every compound in Nangalkinia.still' bears a Nsnkani name and nrast people in the village haveNankani names. Kasem names are rather few here. Nankani is s t i ll used throu^iout the village for sacrifices to shrines. 22 The p -esent dsy language situation is that Kasem is understood all ihe 7.iliaga^althou^i Nankani is s t i ll understood by the older people 'j>y.«- people of 4^ years and above,younger people in northern Nangalkin.ia only understandNankani without being fluent in i t. i It is not hard to see wfcy Kasem has replaced Nankani in northern 23 ?b:e frequent intercourse with northern neighbour ing villages has Na; galkinia* The importance of Kasem as the language of the wider v coimuniiy nakes it necessary for nearly everybody in Navrohgo to understand; it- reEJulted in language shift in morthem Nangalkinia. The sputhem NangaL- kls-\ia prtcole whose territory is separated from that of the northerners by a Btr^iifc of marsiiy land find more convenience in intercourse with . neighbour in g southern villages like Gongoaia* This accounts for their retention of The Bangia lineage of Nagalkinia live in compounds located doae to the Wuru— Nangalkinia border but they have retained Nankani althou^i ixo Nankani is spoken in Wuru. The reason l i es in the fact that the Bangia lineage is the one in which the joint office of headship and Iandprietship is vested* These offices are usually held by the oldest 111 ty both ages and geti-eraticn and only male members of the Bangia lineage are qualified to hold them. However, "the incunibgit should be able to speak Ea*kari^; 1 : S"iiice Naamgali1 s shrine should be + sacrificed to usingilanl have felt compelled: to speak Nankani for this ireason • This is particularly true for (foajporingu1 s compound where even -today parents feel compelled to " • ' -. use NankaQi to children* ' • Parents , l i es t° "^e south-east tf Nswraigo tosaa. ^t shares I'ungu^villages to the north, 3)oba< to the ~ast and Jaanania to the ^ south* I t 's -western border is^sV^ed ty Nogsenia. Li&e is surrounded by both KasenKspeaking ^dNankani-spe^king. HHhereas Uang&lkini:. is usually ctnsidered by i ts people and by du^tsiders as a simple unit, Nayagiia is a complex unit* I t 's name, which - -. ••" -• -. •" • ,. ••• • • . ' •' is iri Kasem? implies the diverse origins of i ts inhabitants* Ooisequent3y does not bmd^its :"obnpcQent..3^ea^s?--;---Tib^a>iatv-:Karaaia^ Badania ' • '. ' " 2 4" aid Kungw?siia. • 2 1 / 5- Everybody in Nayagnia can speak and -understand K&sem and nearly every- body can understand Nankani* Perhaps not all Nayagiia inhabltaits can speak Neikani tolerably well but the majority have that facility*. . . ,. Kungwania aad Badania Members' spe^k Kasem in nearly a ll domains but some of them tin dears tan d find speak Hankani which they would use to a n cn-Kasem-speaking per sen • They would not normally useNankani. The people of Karania are believed to have had their origin iii Saboro* A lineage from K as em-speaking Saboro had migrated to to found the Karania line&ge. As .evidence, the exogamy rule s t i ll this, lineage to Saboro* IThe present day language situatim .in Karania does hot in the least lankani is the suggest that i ts people had originated from Saboro« laiguage used for comnunicatioi among the majority of lineage mates but very l i t t le Nankani is how understood iff: Saboro. Kasem is however stood and spoken fluently* Personal r>ames are in both Kasem but the l a t t er predominate* Iblksongs are composed in Kasem but this is probably the result of the higher premium the wider cdmtiunity places -an Kasem songs. Rial era! dirges are -aitii-ely in Kasem and probably sacrificial prayers would be in Kasem too* The shift from Kasem to Nrnkani as f i r st language can perhaps be attributed to the mothers of most, lineage members• We shall discuss this point further in. a sabsecp.ent section* Yibanabia is perhaps a migrant lineage from Punyoro and may sacrifice to vo^ro, the blacksmith's fetish* Members r£ the lineage speak both Kasem andlankani in the home, ibr e sat eric rituals Naakani is likely to-be used, being the language of their paternal ancestors* As in Punyoro funeral dirges are sung in Kasem. Personal names are in both Bankani and Kasem. Although it is true to say that .in Navrongo almost everybody under- stands some amount of Kasem, the kind of Kasem spoken varies from place to place* Broadly two geographical varieties-may be distinguished; H-K and N-K 25 N~K is the dialect f i r st .developed by the Gatholic Missionaries and l a t er given a boost by the Kasem branch of the Bureau of Ghana languages- 215 is spoken in Group Most l i t e r a t es educated in the mission schools taad to speak a variety of t h is dialect. As a geographical dialect, N-K '&' and It has been more heavily influenced by Nankani which is the •'0 Tillages. other language spokm in the villages in question. A phonological marker of N-K syllables. The only consonants that may be labialised in N-K are the velars; k, g, q. All other consonants may only be labialised in. syllables where the vowel is low i-e« / a / o r / 3 /. labialised. is the absence of lip-rounding or labialisation from certain In N-K all consonants may be N-K. kwx pwa bwe b ut N-K kwi pwa be 'bones' •chiefs' 'to ask' It is obvious that the restricted occurrence of labialisation in N-K 1 is a result of phonological interfermce from Nankani speech habits. Naikani has velars as well as labial velars and the l a t t er are the only instances of double articulations. The labialised consonants of Kasem are comparable to double articulations; qw, kw, and gw in Kasem are comparable to Naikani 13m,kp and gb respectively but pw, bw, nw etc. have no homophones in Nankani hence the dropping of the second consonant from, these clusters in N-K, 1 N-K is the dialect found in most Group 'ky villages Though different from the Paga variety of Kasem, N~K_ is l e ss divergent- Harmony which N-K does not have in i ts full form. It has a full vowel "While Kasem can be said to have beai influenced by Naikani the reverse is also true- The varieiy of laakani is so different from other varieties that i ts speakers frequently prefer to use a neutral language like English when speaking to speakers of other varieties like Irafra in Bolgatanga etc* On the other hand speakers of other varieties do not hide t h e ir scom for Navrongo N aikani» 2.0 She outline sketched above was aimed, at indicating the extent of .language shift. The observable trmd is towards replacing Nankani with Kasem in many settlements. The eiymology of place names? lineage names as well as personal names now aid in the past taken together with language use in selected domains suggest that Nankani was cnce more widely spokai thai today* Below, language use in selected domains will be examined with a view to ascertaining the extsit to which these have contributed to language maintenance or shift. Domains to be selected include the following; •j) Entertainment 2) RituaL 3) Market • 4) Education and Church 5) Kinship and Marriage. Entertainment may take various forms with various aims* It may or may not involve verbalisation. Here, attention is focused en entertain- ment involving Verbal Art. . • , Folktales are s t i ll told and enjoyed by both adults ssid childrai particularly in rural settings. The language of narration is Kasem in Group s easily. • Riddles and singing games have not been observed in Grou,D 'A2' 217 villages but in the other groups they tend to be performed in Kasem although it is also likely that in some Group 'B • villages they -would be enjoyed inEankani too. Folktales* riddles and singing games are more or le'ss an in-group affair except whsi performed in educational institutions* Jtolk.-singing is performed for both personal md. group entertainment. Within the village the f i r st language would be the noaaal language to use except for the educated who sing Ihglish songs for personal entertainment. For out-group situations e«g» at the durbar grounds or on festive days when villages meet in Navrongp market, the language of folk scngs would be Kasem. 2«2 Situal Domain Hituals of various kinds can be distinguished and they vary in the size of the congregation they draw. Personal shrines, donga, may be sacrificed to by the individual personally or by his or her son. The shrine of a dead parent is usually sacrificed to using the first language. In sacrifices to lineage founders a larger congregation is expected but the language used is usually the one the lineage founder spoke during his life time* For very remote ancestors it m3y be necessary to use the language those ancestors were supposed to have spoken at their time* This oftai means using a language different from the one now being used by living descendants* For most of Navrongo, Naakani would be used although in many villages Kankani is no longer spoken. $here the person pouring the libation cannot speak Nsnkani it is enough to murmur the i n i t i al words of the prayar in Nankani, a thing which everybody caa do, before switching oode to the more familiar language. Foreign shrines and fetishes or magical objects must be addressed in the language understood by the shrine in cjuesticn. A Mossi shrine will be addressed in 1/foore. The Kwara, the chiefly fetish, must be sacrificed to in Nankani because it came from Zekko in ftankani country. The yoro, black-smith's fetish, must be addressed in Uankani. Because of -the linguistic demands imposed on office holderss vine ar« in control of important lineage shrines, lineages in which is vested certain traditional offices sometimes try to prevent the complete replace- 2L8 ; . ment of the ritual language with mother, such as Kasem. Curiously most of Navrongo sing funeral dirges in Kasem. The explana- tion for this is not obvious however most of such fimeral dirges are recm.t* They make references to recent events l i ke slave raids, visits to Gambaga, one time headquarters of the Northern Territories, aid wars with Paga. These songs are stereoiyped for almost all villages and are not therefore indices of language shift. 2*3 Market Domains The market is an instance of wider-*settings since different groups meet there now. It serves economic, social and political ends. In the past the market congregation was smaller and more localised and the event used to be held in the l a te afternoon. The most important market in Navrongo is the central market found in It was probably a small market in pre-colonial dsgrs* Now Navrongo torn. it attracts people from Navrohgo and beyond. Situated in the hub of Group -A-]*' villages? it has boosted Kasem at the expense of Naakani since Kasem is the language most commenty used in Navrongo market' Although Nankani, Buli, Hausa and even Akan, not to mm t ion English, are spokai in the market, affective marketing demands knowledge of Kasem - preferably a local dialect of Kasem. Other markets in the neighbourhood are Maiyoro market where the most important languages are Hankani and Kasem, Paga and Kejelo markets where Kasem is spoken, Chuchuliga were Buli and Kasem are used and ELu market were Nankani, Buli and Kasem are used. These markets have only local importance aid their influence on language shift or maintenance is less important. 2*4 Education and Church Apart from a minority of local people and some outsiders, most people in Mavrongo are either members of the Catholic Church or followers of the traditional religion. The Catholic fathers arrived in Navrong around 19^6 *d set up their mission there almost immediately. Soon afterwards they established a school. 219 Today Navrcngo is the seat of Bolgatanga-Navrongp Diocese. Kasem was, and s t i ll i s, the language of evangelisatian in many- places in 'the Navrcngo parish. liturgy, parts of the Bible, hyims e tc have beai translated into Kasem. The Sunday sermons are made in Kasem and English* Catechism is conducted in Kasem and the In tiie early mission schools Kasem was the medium of instruction at the early stages but it was also taught as a subject ad literacy material was prepared in it« The Kasem wing of the Bureau of Ghana Languages and the School of Ghana Languages have since continued with the woik initiated by the Catholic missionaries. The Institute of Linguistics in Tamale is also doing a lot in this direction. Because most of the schools in Navrcngo are located in the Kasem- speaking areas and as the language of instruction was, and s t i ll i s, Kasem, children from Group B' villages and elsewhere who went to school in Navrcngo were compelled to leam Kasem. The i n i t i al emphasis put ai Kasem may have also created a more favourable disposition towards Kasem at the expense of Naikani thereby hastening further the spread of Kasem and the decline of Nankani in Navrongo. 2-5 Kinship and Marriage The rule of exogamy binds lineages believed to be related or believed Ibr example, Nawogiia to have other kinship t i es not easily demonstrated. is bound by exogamy to i ts sibling lineages - Koraiia, Bwaiia and Yigwania. They cannot marry fromNamolo however, because they have been settled en Namolo lands* it>r most villages like Nawogiia, there are several neigh- bouring villages with which marriage is impossible. On the other hand the Kasena scorn marriages with nearby villages with which no exogamy actually forbids marriage. Because of exsgamy and other a-fcfcitudes, marriage is often desired with far eff villages and in many cases this means marriage between Kasena a n d N e n k a n is or Bulsas* 27 Once such a marriage bond h as been e s t a b l i s h ed o t h er marriages a re s et in motion. The d e s i re of every woman is to b r i ng in as many of h er "sisters11 into her husband's clan as possible. She may start by fostering "junior sisters11. Her husband may decide or be invited to render services 220 (agricultural in nature) to his in-law This however is usually done at village level; young men from the husband's clan or village visit his wife's village and stay there for a couple of days working en the in-laws' farms* Usually they return home with more daughters taken from the host village. Before long, the husband's village abounds in women drawi from the wife's village* This pattern of marriage end Kin ship which allows the accumulation of girls from a village speaking one language in another village whose language is differmt is one source of language shift• Mothers inva- riably teach their children their own languages and parents may evai contribute to the shift by showing toler mce to "toe wife's language* One instance of this is the practice of chile naming in the language of the wife • Inter-tribal marriages like those outlined above abound in Navrcngo* Their main contribution to language shift aid bilingualism lies in the fact that they break language prejudices- This is true forNavrongo en the whole? where H ankani speakers and Kasem speakers have no bones to pick. fjhat matters is not what local language one speaks but where one come from* Gomreon language is not a centripetal force in Navrongo. Throughout the discussion no attempt was made at assigaing Manyoro to a group. This is b a cause of the writer's igiorance of the language situ a I? oa prevailing there' According to Oardinall, both Kasem arid H ankani are spokai in Manyoro* ' This assertion is true basically but further details are required to oaaLle Manyoro to be classified in one of the groups set up here* On the whole the traid towards replacing!ankani with Kasem is likely to continue into the future> unless the prestige accorded Kasem is minimised through change in Govemraaat policy* 221 IDOTNOTES 1) Mo systematic field research,-was conducted for this paper aid the material presented here is based en first hand observations and impressions as well as bits and pieces of information gathered over time* 2) Eefer to the map attached, which was culled from the 196O Gaisus map of Northern Ghana- The map is inaccurate in some respects? some of the villages msxtioied in this paper are either omitted altogether or wrongly located- Another limitation of the map is that i ts Enumeration Areas (E«A.) do not always tally with village settlements; large villages have been divided up so that certain sectiens can be classified with neighbouring villages to constitute E«A« -while other villages discussed here are not mentioned at all, having been included in some other villages to constitute an E»A. 3) A version of the legend in question is also found in St. John Parsons (195^5 1-10)« The version contained in St. John-Parsons is more elaborate that many others found in the field but it also appears to imply that the Nogsenia lineages are the cnly descai- dants of Butu, also knowi as Gebutu. In fact Gebutu is usually not mentioned by informants unless specifically questioned about . him. Whai asked about Butus Achulba, elder of Gongpnia, asserted that he was their ancestor but that he never set foot in N&vrcngo. According to i.c:mlba it was Butu's grandson Asoo who migrated to Navrcngo because he ha-d been denied the chief ship. 4) The method employed in this description involves utilising clan histories to account for language shift. The terms "domain", "out-groups -v.i~g:ccup" nwider conmuni cation" e t c, employed in the paper were takgi from Socio-linguisties and where their meaiings have beaa modified this has been indicated- On the other hand some terms found in Social Anthropology such as »%§lann, "maximal", "major", "accessory", "attached" and "authentic lineage" have been used too. the same as in Ibrtes ( 194 5) • Ill e me an ings of these forms are more or less 5) The pioxy headman of Nangalkinia described Gwala as a cavedweller, but Ateyiri of Nawognia, whose village is a Gwala lineage, asserted that the father of Gwala descended from the sky and that Gwala himself was neither a cave-dweller nor aa immigrant. On the other hand Zegolas the man next in succession to the custodianship of Gwala' claims that Gwala was also aa imnigrant from Zekko. 6) The v i l l a ge of the w r i t e r 's mother. 7) Several versions of the legmd which describes the founding of . Paga exist" One of these is found in St. John Parsons op «cit» p* 53 - 56, and anoitier in Callow and Teviu. 222 8) There is the story that a /Eaakaniman and aKasaaaman were once given- a fowl for joint seprices they had raidered somewhere* At 1b© place of partingt en their way hose, the Nmkanimaa suggested in Haakani, jji, JM| (well* shall we roast i t) to whioii the Xasena- man nsfao did not -uaderataadlaikani replied, aawo, di wu wo. go. (l am afraid we mist roast it)« Both had misunderstood each other end "only the timely interval ticn of s, bilingual pre^" sited an uglj sa&xe* This stoay may not have happgaed as reported but ut emphasises the lack of mutual intelligibility "between Kasem aidNankani. 9) Baader-Samuel (i97i) notes'that lexically Bali u& a l i t t le closer to Iiagbani-Mampruli than to Naudm but that it retains a noxsi class ^stem -typical of Pilapila and lauds. writer has never studied the language sd ma ticn in Biu at first hand but the majority of people who have been to Biu usually assert that although some Kasem andHaikaai are spokga therei Buli is the commonest first language. Bin is however under the juria- dicticn of the chief of Kolgu ir Haikani country. It should also be noted -that CardJn-all •-)921 identifies Biu with Biuk in Akuna of Bagwe lineage in Mogsaiia described the Kabagaia group as aa authentic IMeage descended fromNogse but other informants from other lineages in Uavrcngo have refuted Akuna's claim. "Ehese informaits asserted that Kabagiia is ai attached lineage snd they eaplained that custody of the earthshrine was vested in the lineage by qae of the Navropios as a sign of favour* Akuna, see . . * 11, mentioned that a cave-dwelling lineage was present in Nogsenia before the advent of the Zegia group and according to o-ftier informants Butu first lived with cave-dwellers at Waala, which is not far from Kabagiia. If there were cave- dwellers at all it is hard to find out who these people were find whether their descendants have survived. Curiously enou^x Oar din all op.cit*, p. 21 affirms the existence of cave-dwellers at Nogsenia. Befer also to P-' 11 above. Zegpla of Bwaaia considers the name to be of Naakaai e'tymologf. % attention has Just been drawn to the relationship betwem the chiefly lineage in Pin&a aad i ts counterpart in Pungu viz* Bavogiia. If these two lineages are truly related'as has beffi said, then most probably there are Zekko elements in ]?inda too. "The presence of the chiefly shrine5 ^Kw&ra", the exclusive preserve of the/Zegia (9.an further supports this suggested relationship, furthermore only the presence of a Zfegaa ruling lineage in tinda O£K. explain the political relationship between Pinda sttdUavrongo despite the fact that the £0after" is nearer to P&ga, 12;' 13) 14) 15) 223 According to Zegola, Gaani Is a Gwala lineage, but Cardinall op» pit, p. 13 mentiens a legend which claims that the founder of the settlement was a him ten? from Po, a K as em- speaking locality in Upper Voltas who was married to the daughter of a friendly Bin man aid settled there by him at Biu« On the surface these two accounts conflict but it is possible that Gaani like most village settlements contains at least two differ ait groups (if not more): the Po group which ¥£>uld sacrifice to i ts aacestor using Kasem but initially speaking probably Bull, and the Gwala group mjhich was preceded by the l--o group. •The widespread use of Hankani can be explained by the presaice of the Nankani-speaking Gwala lineage aid the proximity to other Nankani-speaking villages in the neighbourhood. The place name "Gaani" or "Gaaiia» sf* cardinall's MGania», op. oit. p. -]3 suggests that the name has a Kasem etymology viz. (bush;, 2i_ia (people) i«e« "people living in the wilderness". Gongonia sing theirprai.se dirges ~ Inse•- in Bill but the 16) •j?) g dirges recounting tribal bravery* are sung in Kasem. It is interesting to observe that the former are sung in praise of deceased female members ' i . e. wives and mothers. It appears that in tygcne days whgi the rale of exogamy was extmsive Gongcnia had recourse to Bull-speaking Ohuchuliga for wives5 hence the use of Buli. Other Group 'B- lineages may also sing Bali luge but it possible to make an exhaustive enquiry. Certain sections, of Bwaiia sing Buli ^use aid the reasm givai by Zegola of Bwmia is that Bali luse are more melodious heace the temptatioa to use these dirges* 18) Zegola maitlcned Vonsnia as one of the Gwala lineages aid yet most people assert that the original home of Von an i a- was-at Vona, a sacred spot on the western border of Hamolo. Kalu Kambonseyia whose patri lineage is at Vonania agrees that Vonania came from Vona but she also motioned that Vonania and Korania sre very closely related villages. Other sources also-related Vonania to Telania the group who, even today, s t i ll sacrifice at their ancestral caves? Teiania are supposed to consider Vona as one of their former homesteads. The traditions vary from village to village and seem to confuse the picture.- deduced that Vonania has more than me maximal lineage - which probably spoke Kasem, Bali and Hankani in the remote past. However as of presmt Hankani is the dominant language followed- by Kasem. Bali may however be retained for certain esoteric rituals but this is left to be found cut» It cai however be 19) The ruins of Hogsaiia are at Kaasi near Doba andNogsenia and Doba s t i ll observe a joking relationship described in a p a t r i- lineal idiom. Other sources assert that some of the lineages 224 in Doba were descended from exiled Kogsaiia people yet others namely Mkana ©plained that the founder of toba was also an immigrant from Zekto, possibly related to Butu. 20) The authaatic lineages in fastngu trace descent to TSusim, .also of the Zekko stock; howeverf todcf the f i r st language, of lugurxgu. is Kasea and for practical purposes Nenksaii is non-existait» 21) "Hae legendary founder of Ghuchuliga to Bulsaland was a, hunter from Kasemr speaking Jiebele in Upper Volta called Achuloa. CM.efsb.ip has passed out of ny hands"-* w as translated, by other informants as "the 23) term «wider comimani1y» should b© understood here to mean a aw as defined above* 24) la cant investigations suggest that the name la^agiia derives from the personal name, Tagm, and has nothing to ob with the mixed oojqposition of .-die settlement complex^ Yagm was one of 14ie n a n i s ms of fusim aad'ha4 migrated to present day Nayagiia fxom ilxsini's oou^oxaid in Pungu» 25/ N-X is the shortaaed fojjn for the term Wavrongo Kasem", i.e* the •'dialect of Kasem ^okaa in Navrango. This vaxiely is a sub- varie%" of Southern Kasenij one of the*three principal dialects distinguished by _Zwe»iemam (1962)* 26) Basrxcngo Parish extends beyond Navoro as defined here* , * 27) Before Hie time of Navropio, Kwara, the exogamy rule was binding en maiy more lineages in Navrongo aid Pungu but Kwara p'roclaimed. •Ktiat the rule should be restricted to vaiy closely related lineages, .rt is likely, this p roolaroaticn was made about half a a g o. • • . - 28) Cardftiall op» pit. B E F E H I H C ES 1971 "Niger-Cong), Gur" frgids in Linguistlea Tol • 71 Ed», f» Sebeok et al, Jfoutan• PP« 139-179 • Callow, J. and Teviu, n.d» of ,funding of Paga g Stadiesj*: Cardinall, A.if. Fortes, M-. Schott» Bidiger St» John-Parsons> Zweanemann, J. 225 1921 1945 1977 1958 The n a t i v es of ...the. Northern T e r r i- t o r i es of the Gold Coast, Ion don. The lynamics of Clanship Among the Talien s i, Anthropological Publicatiois} Oxford. "Sources for a history of the Bulsas in Northern Ghana1*, paideuma 23 Legends of Northern Ghana; London • "Hemarques Prelimanaires sur le verbe du Kasem et du Nxana" Actes du Secoade Colleque In tern at ion ale de 3Dakar«