Research Review MS Vol.4 No.1, 1966 THE MULTILINGUAL ENVIRONMENT IN THE ADA DISTRICT M.E. Krcpp Dakubu 1. DffBDDXTICN are rather small villages (see Table 1). Almost every village along the beach ves surveyed, but in inland areas such ccrrplete coverage was not possible. The sanple was constructed according to a quota, the para- nBters being age, sex and degree of formal education, fto one under the age of 15 was interviewed. Table 2 shows the structure of the sanple in terms of the first two of these para- meters. It is rather more evenly distributed than the general population, because of the method of selection. The relatively large proportion of respondents from the oldest age group is probably relatable to the amount of leisure tine people have to spend answering questionnaires. The ratio of men to women in the sanple is also not a reflec- tion of the structure of the population at large. 27.0 of the sanple - 37.4 of the males and 15.7 of the ftmales - had been to school. This is somewhat lower than in the population at large, and probably reflects the disproportionate amber of older people. The Ada District is almost wholly Big AJa is the focus of the rural. traditional political organization, and Ate Fba is the current administrative centre and the home of such institutions as the Ate Secondary School and the govemnent-run Health Post. Otherwise the main occupations are fishing and fish processing and trading, farming and salt extraction and trading. social and The Ala District is an administrative district of Eastern Rsgion of Ghana, which more or less coincides geographi- cally with the traditional Ada state. It is bounded by the Volta River on the east, the sea on the south. To the west it is banded by the Ga-Oangbe District which traditionally speaks the same language as the Ala District, nanely Dangme (also known as Atengne, Dangbe, Adangbe). To the north its closest neighbours are Be-speaking. The Ala area is traditionally regarded as spea- king a distinctive dialect of Dangne, and although there is considerable internal variation, (Apronti and Dakubu, 1972) this view can be accepted. This paper is concerned with a portion of the data collected as part of a sociolinguistic survey of the Dangns- speaking area, which was carried out in connection with the multidisciplinary Dangne Area Survey of the Institute of African Studies, Legon. The AJa District was the first to De surveyed. The data exMined herein were collected by Dangne- speaking studentSLin the long vacations of 1972 and 1973.2 I atterrpt an initial examination of the district as a multi- lingual area, showing the relative status of the various languages spoken within the district. The paper is limited to a definition of the broad linguistic configuration of the area. The data are drawn from a total of thirty-three localities, most of which 35 TABLE I THE SAMPLE - GEOGRAPHICAL Villages Surveyed No. Respondents Population 15 and over (1970 Census)* Coast: East: Afrive Azizanya Otrokpe Totimekope Oeanseykope Anyakpor Patunkope Adedetsekope Songutsoekpa Elevanyo Puteh (Upper + Totope West: Lolonya Goi Akp^abanya Anyamam Wektmagbe Ada: Big Ada Ada Foah (Upper Inland: Togbloku Kunyenya Bedeku Tansatoku Dogdbcm Adokope Bonikope Koluedor Toflokpo Agbedrafor Sege Junction Sege Nakomkope Koni Luta Total Lower) Total + Lower) Total 30 31 48 46 27 26 17 22 9 35 81 29 56 74 65 77 35 708 67 119 186 49 17 41 34 18 20 54 33 84 20 53 51 17 11 105 422 230 591 352 113 124 270 154 262 546 331 561 882 771 964 391 7069 2282 2394 4676 477 ? 322 646 296 492 156 521 215 620 186 ? ? Total 502 1396 over 3741 District: 27,374 Entire * ? indicates that the locality was not distinguished in the Census report. 36 TABLE 2 S1MFLE - BY AGE ANr SEX Coast Adas Inland Total 19 24 29 34 39 44 49 54 59 Age 15 - 20 - 25 -• 30 - 35 - 40 - 45 - 50 - 55 - 60+ Total M . 50 65 55 35 33 32 24 23 21 38 376 F 36 52 53 34 27 23 23 23 33 28 332 M 9 14 13 7 5 ia 5 7 8 13 92 F 14 18 10 8 '9 7 5 6 7 10 94 M 44 41 35 26 18 18 17 17 19 26 261 F 35 39 39 23 21 17 15 12 16 24 241 M 105 120 103 68 56 61 46 47 48 7 729 F 85 109 102 65 57 47 43 41 56 62 667 708 186 502 1396 For purposes of ccrrparison the district is divided into three. The Adas are treated as a geographical unit in virtue of being relatively urbanized as well as being a socio-political centre for the entire district. The coast and the inland villages are distinguished, in virtue of the different occupations of their inhabitants, namely fishing plus sons fanning versus fanning, "me coast is further subdivided into an eastern section, between the Volta and the Songaw Lagoon, and a western section, frcm the western side of the Songaw to Wakumagbe. It is hypothesized that the Lagoon, which is a large one, is a natural barrier to comunication, and that somewhat different configurations may occur on either side of it. Dialect study and an investigation of multi- lingual ism among the unschooled on the coast frcm Ada to Prampram, undertaken as part of the survey, have provided tentative confirmation of this hypothesis Afrive is actually on an island in the Volta opposite Ada, but as a fishing village is included here with the coastal villages. 2. A UNOJISTIC PROFILE 2.1 DC STATUS OF H N 9C Dangns, and in particular the Ada dialect of Dangie, is unquestionably the major linguistic vehicle of the district at least as far as this sample can show. For well over 90% of all respondents, the Ada dialect was "home-town" language (ma gbi), the first language learned in life, and the 37 language spoken best at the time of tne survey. The figure is slightly lower in, the /Was than in the other sections of the- sanple: Coast: Adas: Inland: Total: 688 172 483 1343 97.1* 92.4% 96.2% 96.2% directions. Vhen the language of ethnic identity, the ma gbi, is not the same as the first langua- ge learned, it means (usually) that a shift has occurred in the primary linguistic patterns not of an individual but of gene- rations in a family group, or at any rate between the individual and one of his parents: an individual learned one language first, although one or both parents spoke a different one. The sanple showed one or two shifts of this kind betwen Ga and Ada, Ewe and Ada, Ada and another Dangne dialect, in both directions, one shift from Ada to Akan, one from Ningo to Aten, and one from Fula to Ada. in no case were more than two languages or dialects involved in patterns of shift. That is, ethnic, first and best language were never all different. If the language of ethnic identity was the same as the first language learned, then the language currently spoken best might be one of these or a different one, but if they were diffe- rent, then the current best language was always the same as one of them. It the language of ethnic identity and the first language were not the same, then either the first language •continued in later life'as best language or there was a "return" to the language of ethnic identity. In almost all cases, the "home-town" language or language of ethnic identity can be interpreted as the language of' the previous generation, and specifically as the language of the respondent's father, fine Dangns are essentially patrilineal in social organization.) The exceptions to this are insignificant statistically for this sanple, but interesting as indicators of socio-linguistic patterns that. might repay further study. Two individuals said that their "home town" language was Dangne, which was also their mothers' language but not their fathers', even though they said they had learned their fathers' language (Ewe) first. It may be significant that both were female. Two others said that their father's language, mother's language, Exceptions to this pattern seem to be clustered mainly around Big Ada and the larger settlements east of the lagoon. Anong the fifty-three individuals involved, one or more of an individual's home-town language, first language learned in life, or the language currently spoKen the best was another dialect of Dangne: Ningo, Shai, Krobo or Osudoku; or another southern •Ghanaian language, namely Ga, Akan or 6*>; or another West African language, namely Hausa. One individual reported her "hone- town" language to be Fula, although her first language and the language she thought she spoke best were the Ada dialect of Dangne. For twenty of the fifty-four, the Ada dialect of Dangne ves the language they now spoke best, but twenty-one were completely non-Ada, in that neither ethnic, first nor best language was Ada. This included nine Dangmes, for whom all three categories of language were the Osudoku (1), Shai £3), Ningo (4) or Krabo (1) dialect. Four had B e, three had Ga and one had Akan in each category. Four, living in Ada Foa, had Hausa for all three. In several cases the first language learned and the currently best language were not the same, indicating a shift in the linguistic behaviour of the individual, either frcrn one dialect of Dangne to another or from one language to another. Five people had shifted from the Ningo dialect, which was their ethnic ("Hcme town") and first-learned dialect, to Ada, and one from Ada to Shai. Seven had shifted from the Btte language to Ada. For five of them, Ada was also the language of ethnic identity. One had shifted in the opposite direction. There were also a few cases of shifting between Ada and Ga, Ada and A«n, in both 38 and their own first and best language were all Dangne, but gave something else as their language of ethnic identity - one each for Ga and B e. The actual history of each case is not known, but it seems likely that a shift to Dangne as best language (at least) had already occurred in the previous generation, but that the language remains as a symbol of group identification. Another interesting case was that of a man who gave Dangie as his "home town" language and first language and mother's language, but stated that his father had two languages, Dangne, from the father's mother, and Ga from the father's father. It seems likely that the choice of response to "hcme-tcwn" language in such a case is related to where the respondent decides to settle. In all cases where the "home-town" langua- ge was Dangie but the first language learned was something else, the first language had been acquired outside the Dangne-speaking area, except in two cases where the first language learned (in both cases Ewe) was also the language of the respondent's mother. The interpretation of "home-town" language as language of ethnic identity, identified particularly with the father, is not valid for the four individuals in Ada Foa who gave their "hems-town", first, best, father's and mother's language all as Hausa, but gave Gonja as their "tribe" (and did not make any claim at all to speak the Garja language). The position of Hausa among northerners and especially Muslims (one of the four gave his profession as Imam- butcher) is special, and invites further investigation fcee Dakubu 1977). 2.2 SHXN) LANGUAGES Dangne is not a language that is widely known outside its own area. It is also not particularly large, in the Ghanaian context, in terms of the mmber of its speakers (slightly over half a million). One might therefore expect a fairly widespread know- 39 ledge of other languages. This proves to be so, although it must be stressed that more than a quarter of the whole sample and a third of the women in it are by their own report monoglot speakers of Dangne. Dangne was itself a second langua- ge for 29 respondents, and Ma was a second dialect for 8. Ga, Mean and Ewe, which are the languages that principally appear as occurring with Dangie in the ethnic-first-best language system, are also the major second languages, with the addition of English. Table 3 lists the main languages that respondents reported themselves as speaking, whether they thought they spoke them well or only a little. In the district as a whole, 6a is clearly the dominant second language, but there is also significant geographical variation in this respect. Table 4 compares the two sections of the coast. At the eastern end of the coast Ewe is equally popular, but in the western part this is not the case. There is evidence that the popularity of Ewe continues to fall and that of Ga to rise as one continues west beyond the Ada District. In the Adas themselves, however, near the eastern end of the coast, Ga is markedly more widely known than anywhere else in the district. Polyglottism is in fact generally higher in the Adas, the relatively urbanized centre, than elsewhere, as the figures for average number of languages per speaker and also the figures for monoglottism (i.e., those who know only one language,) show. It is also the only section of the district where the second most widespread second language is not Ewe but English, although in the Inland areas English seems to have a very slight edge over Ewe among man. In the district as a whole English and A«an seem to be just about tied for third place, but again there is considerable variation, this tine related to sex. In all three geographical divisions English is more widespread than Akan among men, but the reverse is true amonq women. The differences TABLE 3 NTM3ER OF SPEAKERS PER LANGUAGE, IN PERCENTAGES OF TOTAL SAMPLE Coast M 100.0 57.9 51.3 6.3 2.3 11.4 35.6 1.5 0.5 0.2 0.2 2.9 21.8 F 100.0 44.2 35.2 8.4 4.5 12.6 11.4 1.5 0.3 0.6 0.3 0.0 36.7 Ga Bnre Akan: Akwaplm Fanti Other English Hausa Nzema DagbanI Yor\iba French Dangme only Adas M 100.0 72.8 47.8 F 100.0 64.8 39.3 5 .7 2 .1 10.8 60.8 11.9 2 .1 3.2 3.2 1.0 6.52 19.1 4.2 17.2 42.5 6.3 0 .0 0 .0 0 .0 2 .1 19.1 Inland Total M 100.0 47.1 34.0 16.8 3.0 13.7 36.7 6.8 1.1 0 .7 1 .1 0.7 29.8 F 100.0 49.7 34.4 17.8 4.9 11.2 13.6 4.5 2.9 0 .0 1.2 0.4 38.5 M 100.0 55.9 44.7 11.5 2 .6 12.2 39.3 4.8 0 .9 0 .8 0 .9 1.9 22.7 F 100.0 49.1 35.5 Total 100.0 52.7 40.3 13.3 4.6 12.7 16.6 3.2 1.1 0 .2 0.5 0.4 34.9 12.3 3.5 12.4 28.5 4 .0 1.0 0 .5 0 .7 1.2 28.5 Average Number of Languages p er Speaker: 2.71 2.19 3.33 2.94 2.63 2.40 2.75 2.37 2.57 TABLE 4 NUMBERS OF SPEAKERS PEE LANGUAGE: THE COAST East West M 100.0 60.7 40.5 7.1 3.2 11.1 33.3 0.6 0.0 0.0 0.0 3.2 26.1 100.0 40.1 23.0 7.8 7.2 11.1 7.8 1.9 0.0 0.0 0.6 0.0 46.0 Dangme Ga Ewe Akan: Akwapim Fante Other English Hausa Nzema Dagbani Yoruba French M 100.0 56.0 58.7 5.8 1 .7 11.6 37.6 2.2 0.8 0.4 0.4 2.6 18.8 100.0 47.7 45.5 8.8 2.2 13.8 14.4 1 .1 0.5 1.1 0.0 0.0 28.8 Dangme only Average Number of Languages per Speaker: 2.77 2.35 2.60 2.00 are of 10 % or more, except for wcmen in the Adas, of whcm approximately 2% more spoke English, and Inland men of whom only about 5% more spoke English. Knowledge of /ten seems to increase from east to west along the coast, although never rivalling Ga, and there is evidence that, as for Ga, this trend continues beyond the Ada District. Fante seems to be distinctly less widely known than other dialects of Akan. Hausa is a minor language, particularly on the coast. Yoruba, Dagbani and French are definitely rare, but are claimed by at least one speaker in each. It is noteworthy that the majority of French speakers are at the eastern end of the coast, the part nearest Togo. The woman who reported Fulani as her ethnic language also claimed to speak it. One person in Ada Foah claimed to have learned Ahanta in Takoradi. In 3ig Ada a man employed as a driver in Tema Harbour claimed to know a little Chinese and Japanese, which he spoke with visitors there. There are further differences of linguistic patterning between the sexes. In every division, including both subdivisions of the coast, women have a lower ratio of languages to speakers, and a much higher proportion of monoglots. Even in the Adas, where women are more polyglot and include fewer monoglots than men elsewhere the difference between men aid women is just as great. Ga, Ewe and English are all consistently spoken by a higher proportion of men than of women, except that Inland approximately the same proportions (2$ more women) speak Ga. Akan on the other hand seems for seme reason to be more popular among women than men. It is spoken almost everywhere^ a very small but consistently higher percentape of woman, and in the Adas by a considerably higher percentage. From east to west along the coast the proportion of women speakers for every language except Akan drops considerably - a trend that continues to the west until one approaches another administrative centre, Prampram. For men, knowledge of English and Ewe decreases, although not quite so drasti- cally as among women, but knowledge of Ga increases. In the Adas more men than women know every language except Akan. inland, however, where polyglottism is generally lowest, except with respect to Akan, and monoglottism highest, the sexes are about equal for every language except English. 2.3 Tl£ LINGUISTIC B M J W ed The linguistic environment of the Ada District can be visualized in terns of a series of receding concentric circles. In the middle circle, at the centre of the collective linguistic consciousness would be Dangne, the first and best language of the vast majority, and their lamiage of ethnic and geographical identity, especially in the form of its Ada dialect. At aoonsi- derable distance out from this centre are found Ewe and Ga. Their relative positions and their closeness to the centre vary considerably between different parts of the district. Akan is beyond these, but like Dangne, Ewe and Ga is one of the lamuaaes towards which shifting occasionally takes place in individuals who regard themselves as Adas. English is not one of these and therefore is to be considered as farther still frcm the centre even though sometimes more people claim to speak it. Ewe Ga and Aten all share geographical boundaries with Dangme, although only Ewe has a boundary in caimon with the Ada dialect. Out beyond these languages we meet Hausa and then Nzema, Dagbani, Ycruba, French. Often hardly anyone in a locality is aware of their presence. In sore localities they are rarely if ever heard. Yoruba and Hausa were reported to be sometimes heard from passing salesmen. On the dim horizon is Fulani, of which there are speakers resident in the district, but which very few thought to mention, and only one respondent spoke.3 A small number of respondents in the Adas were aware of Kbtokoli, "Kroo" and Dagaare in their neighbourhoods, and several recognized "Dahcmeyan Ewe" to be present as a distinct language, that, .two or three also spoke. 3. THE POSITION OF ENGLISH English is the only non-Ghanaian language that is claimed as spoken by more than a quarter of the sample: 28:5$. However, both proportion of speakers and status as L2 relative to other languages are much higher among men than women. It is claimed to be spoken by 39$ of the men, and ranks third after Ga and Ewe, with 14$ more speakers than Aten, but among the women it is spoken by only 16$ and ranks fourth, well below the three Ghanaian languages. It may or may not be significant that while about half the women (52.2$) who claimed to speak English claimed to speak it well, with the rest claiming to speak it only a little, almost two-thirds (63.7$) of the men claimed to speak it well. Just what this claim really means in terms of performance ability for either group is impossible to say. At any rate, not only do significantly more men than woman claim to speak English, but they claim to speak it better. As expected, the vast majority of respon- dents reported having learned English at school. Only about 5$ of each sex reported having learned it at hems $ individuals) or elsewhere. It was suspected that the pattern of multilingual ism obtaining among those who had learned English might conceivably differ frcm the pattern in the sartple at large. The nuitier of English-Ga, English-Ewe and English-Aten pairs were therefore counted, and also the number of speakers who spoke only Dangne and English. The nurber of English vand Dangne) speakers who speak Ga, Ewe or AJan was then compared with the mrster of speakers of these languages in the sarple at large. Hie figures are displayed on Table 5. sanple speakers and the between the English is bigger for wcmen, general although not so drastic. It sens then that among English speakers in this sanple there is a very much higher rate of poly- glottism than in the sanple at large, that this has to do with a more widespread knowledge of Akan and Ga among English speakers and occurs in spite of a reduced incidence of knowledge of Ewe, and that the that obtain among English trends TABLE 5 ENGLISH AND OTHER LANGUAGES % English speakers who speak: M F % of sample at l a r g e: M F Ga 7 2 .1 86.4 55.9 49.1 Akan 31.7 52.2 25.0 29.0 Ewe Dangme only 40.7 29.7 44.7 35.5 7.6 2.7 22.7* 3*.9* * Figures f or monoglot speakers of Dangme. The differences between the English speakers and the general sample are drama- tic. The proportion of English speakers who speak Ga is far higher than in the sample at large, and so is the proportion of Akan speakers, especially among women. The proportion who speak no additional language is also much smaller. Even more dramatic is the reversal of the patterns for men and women. Among the relatively small proportion of wcmen who claimed to know English, that is, knowledge of Ga and Akan were extremely high, much higher than among men, and hardly any did not know at least one of Ga, A|