INSTITUTE OF AFRICAN STUDIES RESEARCH REVIEW Vol. 12 Nos. 2 & 3 1981 UNIVERSITY OF GHANA University of Ghana Institute of African Studies Research Review Vol. 12 Nos. 2 and 3 EWE AS A SECOND LANGUAGE:. A SOOIOLINGinSTIC SURVEY OF GHANA'S CENTRAL VOLTA REGION by James Andrew Ring 1981 Preface This study of the language communities of the central Volta Region— their use of Ewe (their Major Regional Language) and factors relating to its u s e— was carried out under a co-operative agreement between the Institute of Linguistics and the Institute of African Studies of the University of Ghana at Legon, affil- iate organizations. Acknowledgments and thanks are most heartily given to the following: Dr. Anthony J. Naden, Linguistics Consultant of the Institute of Linguistics, who initially introduced me to the rigors of survey work; and who, by example, showed me the value of a good sense of humor. Prof. G, Ansre, former chairman of the Language Centre, University of Ghana, whose personal insight and observations stimulated in me a great interest in the area studied. Dr. M. E. Kropp-Dakubu, Institute of African Studies, University of Ghana, for organizing a workshop on multilingualism in Ghana immediately preceding the genesis of this work, the discussions of which spawned many of the techniques and materials utilized in this survey. Mr.'s Norman Price and Edward Hall, Director and Associate Director of the Institute of Linguistics, whose guidance and supervision during the initial stages of this work enabled its completion. My co-workers from each language group whose hard work and diplomacy great- ly facilitated the survey task. The wonderful people of the central Volta Region who so tirelessly welcomed and aided our research, be they government officers or members of the public. My wife, Kate, who spent hours discussing problems, typing stencils, sub- stituting as survey leader during my sickness, and comforting our children in my absence, All errors or inaccuracies in the text presented remain my own responsibil- ity. An introduction to the survey aims, language communities and methods (Sec- tion l) begins the text. The main body is then broken into a section on compre,.- hension tests (Section 2) and one on the sociolinguistie questionnaires (Seetion ?). Each of these two sections begins with a presentation of the methods and materials used, is followed by a summary of the results obtained, and ends with conclusions which can be drawn from the results of each section* The text ends with a discussion of the implications of the results from the entire study. Footnotes and Bibliography appear following the text. This study is intended as an aid to educators, government officials, cfaurch and communitjr leaders as well as others concerned with coiHHHuiication in the Cen- tral Yolta Region. We hope to provide information that proper authorities can refer to as they face issues involving language planning decisions. J. Andrew Ring May, 1981 Accra 11 Table of Contents \ Preface L i st of Figufes L i st of Tables 1. INTRODUCTION 1.1 T he S u r v ey A im 1...2 Setting and. Background 1.5 Sampling M e t h o ds . 2. T HE COMPREHENSION TESTS A ND RESULTS 2.1 The Multilanguage Test 2.2 Multilanguage Test Results 2.3 The E we Comprehension Test 2.4 Ewe Test R e s u l ts . . . . . .. 2.5 Summary of Comprehension Test Results 2..6 Conclusions from Test R e s u l ts 3. SOCIOLINGUISTIC QUESTTQNAIEES 3.1 T he Individual Interview F o rm . . . . . . . . . . . .. -L2 3.2 Results of t he Individual Interview F o rm - . . . . , .. . . . . .. . . . . . . . . . . . . . .. . . . . . . . . . . . .. i ., 13 12 5 7 7 . 11 ii iv iv 1 1 2 2 2 5 12 1. Personal D a ta 2. T r a v el Experience 3. Language Skills 4. Language Learning Environment . . . . . . . . . . . . . . . . .. 1J5 18 1!3 » . . 2$ 3.3 T he Community Interview Form . . . .. . . . . . . . . .. 35 3.4 Results of the Community Interview Form . . . . . . . .. .'*"38S 1. Language Use During Public Functions . . . . . . . . .. 3®; 2. L a n g u a ge Contact Environment . . . . • . . . • . .. . . ' 4 0' 42 3«5 Summary of Questionaire Results . . . . . . . . . . . .. 44 3.6 Conclusions from Questionaire E e s u l ts . 4. IMPLICATIONS: W H AT D O ES T H IS S U R V EY TELL U S? . 44 4.1 Correlating Factors . . . . .. . . . . . . . . . . . . 44 4.2 Conclusions . 46 Footnotes Bibliography . . . . . . . . . . .. . . . . . .* • • • • • 47 . . . . . • • . " " 49 L i st of F i g u r es L a n g u a ge communities of the Central Volta R e g i on . . . . .. 3 Figure 1. 8 Figure 2. A v e r a ge Levels of E we Comprehension . .. Figure 2.1 E we C o m p r e h e n s i on Averages (According to A g e) . . . . . .. 10 Figure 3. Figure 3.1 E we A b i l i ty Claims v s. Comprehension Scores Travel E x p e r i e n ce , * 17 23 L i st of Tables Personal Data Results . . . . . .. . . P o p u l a t i on Totals Table 1, Table 2. M u l t i l a n g u a ge T e st R e s u l ts (°/o) Table 2.1 Ewe Comprehension Test Results Table 3. Table 3.1 Travel Experience Table 3.2 Average E we Ability Claims {"/") Table 3.3 Average Twi Ability Claims (fi) . . . . . . .. . . . . 20 Table 3.4 Average English Ability Claims {%). . 21 Table 3.5 Averages for Language Ability Claims . . . . . . . . .. 22 Table 3.6 Distribution of Claimants at Ability Levels 25 Table 3*7 Language Claims for Neighboring Languages • , . . . . . . 26 Table 3.8 Settle Arguements Table 3.9 Write Letters Table 3.10 Count Money/Goods . . . . .. . . . . . . . . .. 27 Table 3.11 Dream Table 3.12 Language Learning Environment (Ewe) . Table 3.13 Language Learning Environment (North/South Grouping) . . .. 30 Table 3«14 Language U se Activities , Table 3«15 Frequency of Ewe U se Among Individuals . . . . . . . .. 33 Table 3.16 Opinions on Beplacement of the Local Language by Ewe (%) . . . 3^ 36 Table 3.17 "Is E we replacing the Local Language - Why? " . . . . 27 . . . ... , . . . . . '. 27 27 29 Table 3*18 Opinions on Local Support for Local Language Literature Program . 37 Table 3,19 E we Useage During Public Funtions . . . . , . , . . . . '. 39 . kl Table 3-20 Language Contact Environment . . . . . .. . . . . l fa 4 fa 9 14 19 32 1. INTRODUCTION In May, 1979> my wife Kate and I established our residence in Hohoe, Volta Region. With the help of many local leaders we began a sociolinguistic survey of the ten non—Ewe language communities located between Have and Kedjebi, approximat- ely. By the first of July we were underway; by the end of February, 1980, the field work was complete. 1.1 The Survey Aim r Linguists around the world are developing sociolinguistic survey materials and techniques which can give an inside picture of the language abilities, activ- ities and attitudes of an area's communities. Ghana's Prof. Gilbert Ansre (1975:68) with one of his colleagues said, The major use of sociolinguistic surveys is to provide as accurate and as relevant a picture of the "situation" as possible to the many individuals or groups that have the task of making the decisions on matters related to language and to interpret the implications of their findings for the particular community under study in order that decisions may be based on sound information. It is the aim of this survey to study the practical use of the standard variety of a ma.jor regional language in the environment of a minor regional lan- guage community and also the social ramifications of its use. For the mother- tongue speakers of the minor regional languages, familiarity with neighboring minor languages is explored as well as the comprehension of the most current examples of literature in the standard form of" the major regional language, Stand- ard Ewe. We have approached this field with an "applied" rather than a "theoretical" linguistic viewpoint. The idea behind the project was to field test some meth- ods in order to get some useful information. It is our desire to present data here which will be helpful to people in the area under study. The survey is designed to help answer such questions as: a. How many (or what %) of native speakers of Language A understand Language B? To what level and for what purposes? b. Is bilingualism in B for speakers of A increasing, stabilized, or decreasing? c. What are the attitudes of speakers of A toward the use of B? Toward the use of A? In the course of this study we found it possible to gather enough data in a reasonable length of time to answer these questions. Some of the reasons for this include the fact that the central Volta Region has many language groups in relatively close proximity. The settlement sites are both accessible by road as well as concentrated population centers. In places, for example, where settle- ment patterns are solitary in nature and. scattered (as in a pastoral community) the time involved for visiting homesteads would certainly lengthen. Another factor contributing to the survey's completion was the number of formally educa- ted people among the local town populations. The availability of such educated people allowed a high amount of local participation in the survey work, in areas such as translating the survey materials into the local language, making initial village contacts, gathering data, and interpreting during village visits. In environments where such help is not available, survey methods would need to be adjusted. 1.2 Setting and Background A language map of the central Volta Region could neatly be overlaid onto a topographical map. Language boundaries coincide strikingly with changes in the geographical features of the area. Two diagonal mountain ranges run northeast to southwest in this 80 kilometer long by 40 kilometer wide strip. A 15 to 20 kilometer wide plain divides the two mountain ranges, as shown by the map of Figure 1. This plain is the territory of one of the Volta Eegion's two major lang- uage communities, the Ewes. Surrounding and south of the southern mountain range live more Ewes. North of the northern mountain range live indigenous Twi speakers, members of the other major regional language. In the two mountain ranges of the central Volta Region live ten non-Ewe/non- Twi language communities. Historically, scholars have had difficulty classify- ing at least nine of these. Westermann (1927:96-120) put these nine and four others from central Togo into a class b> themselves, calling them the Togo—Rem- nant Group. Recent studies have maintained this distinction (Mukarovsky 1977:26, Heine 1968:294), with the suggestion by Heine that they show their greatest sim- ilarities to the Guang language family. The tenth language of our study, called Nkonya,. is in the Guang family. The other nine languages (listed from North to South) are traditionally known as Buem, Bowiri, Akpafu/Lolobi, Likpe, Santrokofi, Logba, Avatime, Tafi and Nyangbo. Though they each have other names for their own languages we shall refer to them by these more familiar names. Population figures have been steadily rising for these groups, but as the figures show in Table 1 there is little increase in local resident populations. The explanation for this lies in the phenomenon of rural-urban migration, a fact- or which has been affecting all of Ghana's rural areas (Caldwell, 1969). The largest of these ten groups is Buem with an estimated 25,000 total speak- ers. The smallest groups are Nyangbo and Tafi each with around 3,000 speakers. The figures for Traditional Area Residents include residents whose mother tongue is other than the local languages, thus, the number of local residents sometimes outnumbers the Group Totals (see Table 1 Likpe, i960). 1.3 Sampling Methods ' : Throughout the survey we used quota sampling methods. We chose members of the local community in order to obtain results differentiated according to male/ female distinctions as well as for five different age ranges. Two ranges were chosen above 30 years and three below. By using the quota sampling method we were able to (l) obtain balanced, comparable data from group to group, and (2) observe language learning patterns among the age groups. Survey samples were taken at each separate settlement site in the ten language communities. Only Buem varied in this regard. In Buem we sampled the towns were, according to government policy-1-, Ewe has been used as a medium of instruction in the schools, i.e., Eastern Buem. The western section, where Twi is used in the schools, has been surveyed for Twi comprehension by Prempong and Leonhard (1978) using a simple narrative Twi text. (it was partially because of the low Twi averages reported by them that this Ewe study was undertaken.) One town was chosen from the western area and tested for Ewe comprehension for com- parative purposes. The scores of this town, however, proved much lower than the corresponding "Ewe" side and would misleadingly lower the averages of the towns where Ewe has been used. These results are therefore not incorporated into the Buem survey results. 2. THE COMPREHENSION TESTS AND RESULTS 0n« of the main aims of the survey was to study the language comprehension .skills of individuals from these ten language communities. We investigated two areas in this regard: \ \ . •, -jc. lajor Roads Figure 1. Language Groups of the Central Volta Region Group 1980 (Est.) 18,380 24,265 5,480 .. 8,745 11,628 Nkonya Buem Bowiri Akp/Lol Likpe Santrokofi 5,397 Logba Avatime Tafi Nyangbo 3,404 11,562 2,256 3,241 Totals I960 Census 11,000 14,900 3,280 5,370 7,140 3,230 2,090 6,920 1,350 1,940 Table 1. Population Totals Traditional 1980 1970 (Est.) Census Area Residents I960 1948 Census Census 14,278 11,046 9,789 6,920 32,873 25,431 22,171 13,411 8,889 10,837 11,803 6,877 8,384 9,131 3,452 7,444 9,484 4,401 3,405 2,776 4,038 3,124 2,326 8,211 6,352 5,340 2,549 3,220 1,972 2,491 1,603 1,926 2,856 5,572 4,759 1,648 1,719 4,558 1,148 898 Traditional Area Resident totals in some cases do not reflect actual population increases, perhaps due to rural-urban migration. The 1980 population estimates are based on a pro- jected 2.6$) per year population growth rate. A. The familiarity of a community with the languages of the other nine communities and B. more significantly, the comprehension of Standard Ewe by the individuals from each of these non-Ewe language communities. 2.J The Multilanguage lest The first test developed was called the Multi-language Test. It consisted of ten brief, culturally-relevant action events^ (one from each 'of the,ten central Volta Region languages) spoken onto a tape recording by a member of each of the communities. After playing one of the texts, the interviewer (a mother-tongue speaker of the language community being tested) stopped the tape to ask what .jf,he listener understood. Four levels were used for evaluation: l) Understood All, where the listener could explain the episode; 2) Heard Some Detail, where parts of the event could be distinguished; 3) Name Only, where the listener coi;ld only ident- ify the language; and 4) No Comprehension, where the language was totally un- familiar to the listener. 2.2 Multilanguage Test Results The Multi-language Test yielded the results found in Table 2. The scores shown in that table represent the percentage of those tested who.exhibited one of three degrees of comprehension— either 1.) Understood All, 2.) Heard Some De- tail, or 3.) Name Only (could only identify the language.) The language communities have been placed in the table according to their northern and southern geographical groupings. As the double lines indicate, their scores also group together according to geographical location. These groupings show that the people from the northern mountain range have little language learn- ing contact with the southern language communities and vice-versa. Avatime is the only exception. A look at the horizontal Avatime axis shows that for each of the northern languages a portion of the Avatimes exhibited comprehension. A similar pattern can be seen vertically under the Avatime language comprehension column. Almost all of the northern language communities had individuals who knew Avatime. Admittedly the percentages are low ana only from the lower levels of comprehension, but they could, be indicative of a general involvement of Avatime people in per- sonal contact situations such as trade, commerce, or education iu the region. Another interesting area of the table is the bottom right-hand corner. Of the ten languages, Tafi and Nyangbo have been described as the closest in langu- age structure (Heine 1968:27^). Our results show that 83°/o of the Tafi test group understood the Nyangbo text fully while another 8^o heard some detail. This along with the Nyangbo's 67c/o full comprehension score on the Tafi text represented the highest scores in the table. Such scores indicate either an inherent similarity in what has become two separate dialects of the same language (as Heine suggests) or a high degree of bilingualism between the two groups resulting from extensive cultural contact and intermarriage. Whichever is the case, the higher scores among the Tafi suggest that they would be more at ease using language materials designed for Nyangbo speakers (should such materials prove desirable) than the Nyangbo would be with similar materials in the Tafi language form. Tafi and Nyangbo people also demonstrated a high percentage of Avatime com- prehension at kofo and 3^°/° respectively. Logba had the next highest score for full comprehension of Avatime at l&/o* Avatime scored kfo full comprehension for each of the above three languages. LANGUAGE NK BU BO TAPEB LI A/L LANGUAGES SA LO AV Nkonya Buem Bowiri Akpafu/Lol. Likpe Santrokofi Logba Avatime Tafi Nyangbo A S X A S N A S X A S X A S X A S K A S A S X A S X A S N 1 8 ? 3 1 1 ? 1 1 4 4 5 13 3 16 17 3 6 7 10 7 3 1 4 4 10 16 1 3 3 2 1 3 3 4 4 1 3 4 1 4 4 1 1 1 3 1 1 1 5 5 19 23 1 2 1 3 1 2 7 4 15 1 4 13 7 5 5 7 30 3 1 3 1 1 5 9 1 18 1 5 1 16 16 13 3 5 3 1 3 1 1 4 26 1 5 8 5 1 9 1 1 3 2 1 1 3 3 2 3 16 4 43 28 10 34 11 17 TA' 'NY 1 4 23 83 8 3 1 4 31 67 .8 4 Table 2. Multilanguage Test Besults Three levels of comprehension are indicated: A: Under- stood all; S; Understood Some Betail; N: Could Name Lan- guage Heard. Scores stand for the percentages of the sampled population who qualified at each level. A question mark (?) signifies an untested area. 2.3 The Eve Comprehension Test The second part of the comprehension section, the Ewe Comprehension Test, was used to assess an individual's comprehension of the main regional language. Three texts were taken from draft forms of the modern language versions of the Ewe Bible, These are the two most current examples of Standard Ewe literature being produced by Ewe writers for public distribution at this time. We felt we needed a more complex test than that used for the Multilanguage Test, so we included increasingly complex paragraph types— a narrative, an ex- pository, and a hortatory paragraph. The gradual increase in complexity allowed us to rate the individuals on their depth of comprehension. The major portion of the tape, the narrative paragraph, was a five point description of a visual, concrete event. The expository paragraph, worth three points, presented an abstract idea with a concrete example. The hortatory para- graph contained an exhortation along abstract lines with the use af figurative language (worth two points). This tape was divided into nine natural segments. Each segment contained no more than two full sentences. As the interviewer (a local language speaker) stopped the tape after each segment, the listener was asked to narrate what he or she had understood. This narration was written on the answer sheet in English by the interviewer. Individual scores varied from 100^ (10 out of 10) to zero. If an individual scored ^(p/o or less we found that all information understood was from the narratixre paragraph. Those who encountered difficulty did so as the information became more abstract.^ Scores of 70^ or more suggest the presence of abstract comprehension skills and a good grasp of the Ewe language. 2.4. Ewe Test Be suits The Ewe Comprehension Test produced the scores reported in Table 2.1. The first three columns report three types of overall averages for the ten language groups. The average of the total population sample is given in the first column— "Baw Scores". The second and third columns, entitled "Weighted" are the averages of town scores that have been adjusted in accordance with the proportion of the group's population which the town represents. Population figures for the "Weighted Estimate" were those suggested by local town members for the year 1980. The "Weight- ed 1970 Census" column uses the 1970 census in adjusting the town scores. Because of size variations we would expect the weighted averages to be the most accurate representation of actual Ewe comprehension. However, both the weight- ed columns and the raft scores column prove to be quite similar in Table 2.2. \*o -single average for one language community varies more than 5$> from the other two corresponding column.scores. So, to be consistent with upcoming statements on the raw .scores of the ques'tionaire results, we will quote the "Baw Scores" mean as the Ewe comprehension scores for the ten communities. Figure 2, illustrates the Ewe Comprehension Averages of the central Volta Begion language communities. Logba has the highest mean at 8.7 out of 10 (i.e., 87%) while Buem has the lowest, averaging 4.0 (40^). The other coiununities spread out rather evenly over the interval.2 When the Ewe comprehension scores are viewed in terms of age group averages (Table 2.2) there is a clear rise approaching the age range of 31 to 50. This occurs in each of the language communities with the exception of Santrokofi and Avatime (see Figure 2.l). The scores from Santrokofi drop markedly in the age range of 15 to 20, the present middle school age group. They then reach their height in the 21 to 30 year age group before tapering off in the upper ranges. Level of Ewe Comprehension 100 8? 90 . 80 . 70 60 50 40 30 20 10 75 73 72 66 61 61 48 47 40 Lo Li Ta Ny A/L At Bo Nk S a BU Language Communities Figure 2 Average Levels of Ewe Comprehension (%) Raw Scores OVERAGE AVERAGE Weighted Estimate Weighted 1970 Census Nkonya Buem Bowiri Afcp/Lo 1. Likpe Santrokofi •Logba Avatime Tafi Nyangbo 4.8 4.0 6.1 6.6 7-5 4.7 8.7 6.1 7.3 7.2 5.3 3-5 6.4 6.9 7.1 5.1 8.7 6.5 7-4 7.0 5-2 3.8 6.5 6.9 7.2 5.0 8.5 6.4 7.4 6.8 AGE GROUPS (RAW 21-30 5.6 4.3 6.4 7.0 7.8 6.4 9.3 7.3 7.6 8.0 15-20 4.5 4.0 6.0 6.8 6.8 3.1 8.1 6.4 7-3 7.2 9-13 3.2 2.5 3.6 3.8 5.9 4.8 7.9 3.3 6.3 5.2 SCORES 31-50 6.5 5-7 7-3 8.4 8.8 4.8 9.6 6.5 7.6 8.5 ) 50+ 4.1 3-5 7-3 6.9 8.0 4.5 8.6 7.1 7.8 6.9. 31 Male 7.3 7.1 8.1 9.4 8.3 5.0 9.9 7.4 8.0 8.7 - 50 Female 5.7 4.3 6.5 7.4 9.2 4.7 9.4 5.6 7.1 8.3 \O Table 2.1 Ewe Comprehension Test Results The scores represent an average achieved out of ten (10) possible points. "Raw Scores" represent the average of the total population sample. "Weighted" scores were determined by increasing or decreasing the value of one town's results in proportion with the percent- age of the total population that the town represents. Likpe Santrokofi Akp/Lol Bowiri Nkonya Buern Northern Ages: 9-l3--2Q--30--50-:iW-50-30.-20--13-9 *0 Figure 2.1 gwe Comprehension Averages (According to Age,' Logba Southern 10 The Avatime scores peak among 21-30 year-olds and then drop before rising again in the last category. Tafi also shows a rise in the oldest category. With reference to the sharp drop in the scores of 15 to 20 year-olds in Santrokofi, it is interesting to note that an "immersion" English program was undertaken in the three Santrokofi towns during the late 1960's and early 1970's. In the program first, second, and third year primary school children were exposed to English as the medium of instruction. I was told that the stude-nts who par- ticipated in this early English program are now middle school age and find Ewe difficult. Currently, however, Ewe is being used in primary classes one, two, and three. The last two columns in Table 2.2 differentiate the averages of male and fe- male scores in the 31 to 50 age range. In many of the language communities, es- pecially those with low overall averages, this breakdown shows that the men per- formed better on the Ewe Comprehension Test than the women. In one group, Likpe, the women topped the men with 9-2 versus 8.3- This may indicate a dominant role for women in activities like marketing or trading where contact with Ewes is high. While women of this age group tend to be within a point (10%) of the Group Raw score mean, it is interesting that among the communities that peak in the 30-50 year age range, men average a two point (20%) increase over the group mean in their Ewe comprehension. This suggests a greater involvement for men in affairs requiring Ewe usuage. 2,5 Summary of Comprehension Test Results In summary, the following results are apparent from the comprehension studies. 1. Members of a language community from the northern mountain range (i.e., Nkonya, Buem, Bowiri, Akpafu/Lolobi, Likpe or Santrokofi) registered very little recognition of languages in the southern mountain range (i.e., Logba, Avatime, Tafi and Nyangbo). The same was true of these surveyed from southern communities concerning the languages of the north. Only Avatime people exhibited any consistent familiarity with the languages from the north. An average of kP/o of those tested understood at least one of the northern languages. Interestingly enough, five out of the six northern language communities exhibited similar averages of people recognizing the Avatime language. 2. Tafi and Nyangbo language communities exhibited the highest degrees of mutual language comprehension. It was shown that 83% of the Tafi people tested registered full comprehension of the Nyangbo text, while 67% of the Nyangbo people showed full comprehension of the Tafi text. 3-. On a percentage basis, the Ewe Comprehension Test averages were: 81 to 90% 71 to 80% 6l to 70% 41 to 50% 31 to 40% Logba Likpe, Tafi, Nyangbo Akpafu/Lolobi, Bowiri, Avatime Nkonya, Santrokofi Buem 4. In all but three language communities the Comprehension Means peaked in the 31 to 50 year age range. Two of the groups — Avatime and Santrokofi — peaked in the 21 to 30 range, while Tafi rose to a peak in the 50 and above age range. 11 5. In the 31 to 50 age range men performed better than women on the Ewe test in all language communities except Likpe, There the women averaged 92P/o, a score among women in this age range second only to Logba. 6, Of the language communities whose Ewe scores peaked in the 31-50 range, the 31-50 year old males had scores of an average of 20$ higher than the Overall Raw Scores (e.g., the 31_5O year old Bowiri men averaged 8.1 while the overall mean for the group was 6.l). 2.6 Conclusions from Test Besults From the results of the comprehension tests we have answers to the first part of question one, "How well do native speakers of language A (central Volta Region language communities) understand language B (other central Volta Region languages or Ewe)?". The second part, "To what level?", has also been assess- ed, though it has not taken into account the ability which the people themselves feel they have. The third part of question one, "And for what puz-poses?" is a question we still have to address. 3. SOCIOLINGUISTIC QUESTIONNAIRES Our remaining questions about the language picture in these Volta Region communities requires information other than comprehension test results. To de- termine whether bilingualism is increasing, stabilized, or decreasing, we must look at the activities for which people are using their second language. For instance, we have said there is an overall increase in Ewe comprehension approach- ing the 31-50 age range, yet the low peak levels for some of the language commun- ities may mean that Ewe plays only an auxilliary role in those communities. Commun- ication in Ewe.may be limited to surface levels such as greetings or business trans- actions. Attitudes toward the languages of the area are also an indicator of fu- ture language use in a community. For example, where the younger age groups show high levels of Ewe comprehension we expect a good deal of Ewe use in the future. Negative feelings toward Ewe or a local language preference could, however, change this. Two questionnaires were developed for this part of the survey. The first was administered to the sample of ten individuals whose language comprehension was tested in each town (called the Individual Interview Form) and was used to gather personal data, opinions, and attitudes. The second questionnaire was administered to an assembly of community leaders in a discussion-response format (the "Commun- ity Interview Form") to gather information on public language use and language community interaction. We drew our ideas for the questionnaires from a combination of sources. We incorporated questions used among the Hill Guang by Mary Esther Kropp-Dakubu (197?) and questions used in the Sabon Zongo of Accra by Rex Moser (1979). By doing-this we tried to facilitate any future comparisons of data or analyses. For the same reasons we tried to incorporate appropriate aspects of John Bendor-Saniue 1' s sur- vey outline entitled "A Sociolinguistic Profile" (1979) being used throughout Africa in general. 3-1 The Individual Interview Form The Individual Interview Form covered four areas of information: Section A, Personal Data, included common demographic information such as . sex age church membership, education and occupation. (Church membership was included because of churches' high degree of involvement in this area ) 12 Section B, Travel Experience, contained questions about how often the indi- vidual made trips 1) to a larger market town, 2) to a regional commercial cent- er, 3) to another region, and 4) outside the country. This section proved heip~ ful in explaining unusual claims of language competence. We assumed that the farther a person travels from his or her home town the more likely he or she is to need a second language. Section C, Language Skills, was developed to give us an idea of the individ- ual's opinions of or attitudes toward his or her language abilities. Topics covered included l) languages spoken with some fluency (self-evaluated as 'very good1, 'fairly good' or 'small') 2) languages of recognition and 3) languages used to a. settle arguments, b. write letters, c. count money/goods, and d. dream. One might say this section indicated areas of linguistic confidence rath- er than linguistic competence. Section D, Language Learning Environment, was designed to help us know more about the learner's view of social interactions during the language learning ex- perience. The questions covered, the following points: 1. Of the languages spoken, &_. where they were learned, b_. with whom they are spoken. 2. The activities that can be carried out using &_. only the local language, _b. only the major regional language. 3- The frequency of daily or weekly use of the major regional language. 4. Whether or not the major regional language (Ewe) is replacing the local language, and why. 5. Would local language speakers financially support a literature program using their own language. Questions No. 3> 4, and 5 explore the attitudes of the individuals toward the re- lationships between the local language and the major regional language. Such specificity was only possible because we had a basically bilingual situation. These questions could be modified for use elsewhere. When we found a trilingual situation involving the local language, Ewe and Twi, we focused on the individual's use of Ewe. The use of Twi was noted on the Community Questionnaire. 3.2 Results of the Individual Interview Form Results from the ten language groups, for comparative purposes, are present- ed here simutaneously. A selective picture of some of the results for each lan- guage group is. summarized briefly at the end of the text in section 4.1, Correlat- ing Factors. 3.2.1 Personal Data Section A of the Individual Interview Form, dealt with personal information about the individuals of the survey sample. "Sex" and "age" were pre-established by our quota sampling technique. Also a prerequisite for being chosen for a sam- ple was that the individual have had parents who were both native speakers of the local language. Percentages of Ewe families and mixed-marriages were reco-rded or estimated on the community questionnaire. This first form focused on the local, non-Ewe population." The percentages in Table 3 for "Church Membership" indicate the dominant church groups of the areas. Most of the people interviewed claimed some church affiliation. This does not mean, however, that local cultural rituals and be- liefs are being lost. As will be seen later, nearly everyone including the self- designated non-church members indicated their participation in local customary rites as a language community member. 13 CHURCH MEMBERSHIP EDUCATIONAL. STANDARD OCCUPATION RC •EP OTII NO NO S P M f PS TT~ UN P St GW SK CM NO Nkonya Buem Bowiri Akp/Lol. Likpe Santrokofi Logba Avatime Tafi Nyangbo 53 68 7 43 67 3 42 10 60 23 33 29 80 57 20 80 50 78 38 51 3 1 3 - 11 3 5 6 - 19 11 2 10 - 3 13 3 6 3 7 30 27 31 13 13 27 15 11 40 31 12 22 19 31 25 20 23 40 18 23 49 46 46 44 41 47 57 39 35 40 7 5 2 6 12 3 5 9 8 6 2 1 2 4 — 1 - 3 - - _ 1 - 41 48 58 43 37 53 48 36 40 39 37 39 29 41 47 30 33 33 40 44 3 2 2 3 4 3 5 8 3 - 13 3 6 7 4 - 3 13 5 9 4 2 2 4. 4 3 8 6 - 6 2 6 3 1 4 7 2 4 3 3 EWE COMP. BANK 8 10 .6.5 5 2 9 1 6.5 3 4 KEY: EC: Boman Catholic NO: No schooling EP: Evangelical Presbyterian OTHJ Other NOJ NO church membership P: Primary M: Middle S_ T: Se c ondary/Te chni- cal PS Post-secondary TT: Teacher, Training UN: University F: Farmer or Parm/Pish St: Student GW: Govt. worker (CMB, Teacher, Agric Off.,etc.) SK: Skilled (Tailor, mason, factory worker, driver) CM: Commerce (Business, Trading) NO: Unemployed Table J. Personal Data Results The percentages show that three of the five language communities with the highest levels of church membership are also the three high-scorers on the Ewe Comprehension Test—Logba, Likpe and Tafi. The other two communities, Akpafu/ Lolobi and Buem, were fifth and last respectively in the Ewe Comprehension Rankings.7 In the "Educational Standard" columns, Logba, with the highest percentage of Middle School attendance (at 51°/°) also had the highest group average in Ewe comprehension. Likpe, with the highest percentage of secondary school attendance had the second highest Ewe comprehension. In fact, all but two of the ten groups had 40% or more at the level of middle school attendance. This, along with the next set of columns (occupation) shows the strong emphasis laid on education in these areas. The "Occupation" columns divide primarily into two categories: farmers and students. Students seem to be an inordinately large part of the survey sample in this rural area; however, throughout our time of field work we found practical— ly no children out of school in the morning time. In order to, fill the two age categories 9-13 a nd 15-20 we often had to select children from the classrooms. ' This does not reflect an overbalance of students in the sample group, but rather the actual proportion of young students in the group population which the age group categories were designed to display. Seven of the ten groups also had at least 40% who classified themselves as farmers. (included in this group are three young women who gave their occupation as housewife/farmer.) Occupation categories show little correlation with the Ewe comprehension scores. For instance, the highest and lowest scorers in Ewe, Logba, and Buem, have the same farming percentage of k&/o. Bowiri and Avatime had the highest and lowest farming percentages in the sample and yet scored exactly the same in Ewe comprehension. Thus, being a farmer does not affect one's ability to speak Ewe. Similarly, the percentages of students is equally unpredictable. Buem show- ed the fourth highest percentage of students while Logba tied for second to last'. Only the student percentages of Likpe and Santrokofi seem to reflect any similar- ity to the Ewe Comprehension Rank found on the left .side of the chart. One occupational column shows a statistically significant Rank Correlation with Ewe scores and that is the column designated "Commerce". Here, with the ex- ception of Tafi, the Ewe Comprehension Scores and the percentage of the popula- tion sample involved in commerce indicate that there' is a relationship between trade and the igaHiing of Ewe in the central Volta Region. The absence of any men- tion of involvement in commerce among the Tafi group is perhaps indicative "of either the role that immigrant Ewes play in Tafi communities (as traders) or pos- sibly a minor focus on business ventures other than farm projects in the Tafi area. Their high Ewe scores, however, do suggest that they are involved with Ewe in other aspects of their life. ~5.2.2 Travel Experience The results of Section B, Travel Experience, are compiled in Table 3.*1. Our intention in developing this section was to find out where and how often an individual goes when he/she leaves his/her town. We assumed, as we said before, that the farther and more often an individual travels, the more he/she would be exposed to Ewe as a medium of communication. Thus, by assigning graduated values to the percentages of people who travel, according to distance and frequency, we can plot the pattern for travel and compare it to Ewe comprehension scores. This is done in Figure 3>1» The percentages for travel experience parallel quite closely the Ewe compre- hension scores except in three language communities, Buem, Likpe, and Logba. 15 t' Marke Town Commere 'Cento ial r Anothei Region Outs i de Ghana 2/M 2/Y 2/41 2/M 2/Y 2/4Y 2/M 2/Y 2 /4Y 2/M 2/Y 2/4Y 26 64 64 51 53 63 77 54 ' 36 47 .. 6 - 16 17 19 17 23 1 26 25 26 3 2 1 2 - - : 1 3 4 28 39 26 •41 1 , 7 14 27 30 26 21 38 25 31 7 20 29 40 28' 34 3 13 12 7 3 - 10 4 15 14 13 1 7 3 3 - 3 15 13' 10 27 19 15 31 11 10 14 13 17 11 18 14 19 :6 10 20 11 20 20 2 10 1 1 - - 1 • 4 3 1 4 20 1 11 -• 3 1 5 3 - 2 18 3 9 - 5. 27 15 23" 16 ••Adjusted Percentage Of People Who Travel 29 49 36 43 17 23' , V 39 39 38 . • : •' Nkonya Buem Bowiri Akp/Lol. Likpe Siintroliofi Logba Avatinie Tafi Nyangbo Language Learning Value O\ Table 3.1. Travel Experience Key: 2/M: Two times ptvr moutli. 2/Y: T WO times por year. 2/4Y: Two times por 4 years. Lo •i-i A/L Av Bo Nk Sa 3u Figure 3. Travel Experience When the language communities' actual travel levels are plotted against a line correlating with the Ewe comprehension scores there is a strong relationship except in three communities: Logba, Likpe, and Buem. Key: of people who travel Hypothetical Ewe Comprehension Correlation Line 17 Buem is at the top in nearly all of the four distance scored lowest in Ewe comprehension. Likpe and Logba, on very little travel beyond their traditional boundaries (on the whole; ana y scored quite high in Ewe comprehention. ^ ^ Results for the Buems suggest that either the Buems do not use Ew, i* travel on the average have been able to learn home towns. A possible explanation for these occu^ances ,ay lie in the fact main towns of Likpe and Logba (i.e. Mate and ^™Vetll} **„* nf P w es re«idf there ers. And as shall be seen later, comparatively large group, of Ewes re...icH on a permanent basis. English is also an alternative for those traveling to different region,, es- pecially rfo for those who have reached middle school. In this'"area traveling outside the country usually means going to Togo where Ewe is dominant. For the Buems, however, two non-Ewe language communities lie rlgh over the local Togo border - The Akposor and the Ahlffr. A number oi Buem, mentioned traveling to visit relatives and farms in these areas as well as ctiug these languages as ones they could understand. 3.2,5 Language Skills" Results from the Language Skills Section, Section C, help us evaluate peo- ple'^ opinions about their own language abilities. Four additional areas oi functional competency were also investigated.- These included the languages used to settle arguments, write letters, count money or goods, and dream. Those interviewed claimed their highest degrees of speaking ability in the two main regional languages - Ewe and Twi - as well as English. People also in- dicated varying ..degrees ~of familiarity with 32 other languages, 12 of which are indigenous to the Volta Region and 10 of which have their origins outside of Ghana. Regarding the claims of speaking ability in Ewe, Twi, and English, Tables 3.2, 3-3, and 3.4 respectively, contain the group averages for the men and women, subdivided according to their age groups. The percentages represent the average level of ability claimed by the people in each age group. For example, Likpe males above the'age of.fifty-one claimed 100$ proficiency in speaking Ewe win If the same men claimed an average of 64$ speaking ability in Twi. Young men trom Likpe in the 21 to 30 age range claimed higher ability in English than Ewe. Group averages in the three languages are lined up in Table 3."?. from the? group averages we can see that an Akpafu/Lolobi person or an individual ironi Logba claims more speaking ability, on the- average, in English than Twi» tohere<» a person from Bowiri or Nkonya claims more average ability in Twi than English- (A higher Twi ability would most likely have been claimed in the Western side o Buem also.) For all the language communities except j^ony^ claiHis of £jwts,_j*£L£T ity exceeded claims for either Twi or English, The question that arises here is, "Do the claims of E M» competence eo with the performances on the Ewe Comprehension Test?" The nitsxter is !fj-«s't figures from a Statistical Bank Correlation of the Ewe claims and the Coiapreh sion Test Results (weighted 1970 Census column) fall within the range of sta- tistical significance for cojrrelating data. Jghe similarities may also bo see in the pattern of Figure 3.1 where Ewe Ability Claims are plotted against Ewe Comprehension scores. 18 X 'O XI to o CM o r- CM f"- r- ON r- ON x CM c^ t**" ig~* vo x' r- o r» o r- © •H © X © LO O ON LO © © © © s-O ^j £j*^ LO, I to to CM SI CM ["*- -tf IO 66 r- in oo i^- o r^ r-i © o © ON ON co x r~ © r^ x r- LO 'O x o X © © CM © © C". © X \O © © -H © © C". r-' x Is- X LO X x r^- to r^- ~^o x X 0s- O r- x © © x O4 © ON 03 o -p' •H rH •H << 0 3? OJ to c OJ CM ^f in © CM X OJ m o © r- CM 'O ON o DJ r- in s O O\ 00 o S 0) ft iA •H 00 OJ s -g r- & «s •H •H •H I •rt O CO 20 u > X3 M © -4* o OJ © CM co ©in I T -l >o o to Csl in to •H. to i I i cc i c- r- r- ON CO •H in to CM ON CM to CM CM ji I O to o ON m 1 .* 00 IT- \o *# . ON co ^p oo in © © p~ in in in in OJ r- r o CO oj r»- 00 to CM to m O -H O CO CO in too © CM oj r* in OJ © m in ' r* in r- ON © co to in co g -P •H V) SsD I to OJ to H 01 CO CM co CM a CM © ON r- \o OJ m OJ to to to to m OJ -tf IO CM O -P rO 6D ® O i> H4 5 6H ^ •H © o -P CO 21 Nkonyu Bucm Bowiri Alcp/Lo 1 Likpe Santrokofi Logba Avatime Tafi Nyangbo Table t-3 to Ewe 48 56 77 73 73 77 92 70 97 82 Twi """63 12 65 27 20 42 • 28 15 29 .1.6 for Language 3.5 Averages for Ability 'Claims (out 39 43 3« 44 40 42 56 40 52 - 38 of 100 XQQfo Lo Li Ta Ny A/L AV BO Nk Sa Bu Figure 3»1 Ewe Ability Claims vs. Comprehension Scores Ewe ability claims (though higher than test scores) closely re- semble the Ewe comprehension scores in all communities except- three: Tafi, Bowiri, and Santrokofi. Kev: Ewe Ability Claims Ewe Comprehension Scores The three peaks on the graph of Figure 3. 1 show that the people of Tafi, Bowiri, and Santrokofi claim more speaking ability in Ewe than evidenced by their results on the Ewe Comprehension Test. The difference, however, is probab: not a matter of quantity of Ewe use but rather quality. Speakers of these ?three languages may well use Ewe every day (and we will see if they claimed so in; an upcoming section), but their knowledge of its use in the Ewe Test proved more limited. ' , Though much could be suggested from the claims for Ewe, Twi, and English concerning language preference we will leave these scores and go on. The last comment we will make regarding Tables 3.2-5 is that from the correlation of. Ewe scores and Ewe ability claims we can assume a general accuracy in the people's ability to estimate their own language skills. Our tests show, however, that, when depth of understanding is a consideration, people's opinions - taken alone - can be misleading. The actual distribution of the individuals at the levels of speaking ability that they claimed can be seen in Table 3.6. Besides Ewe, Twi, and English, the two languages most frequently claimed were Ga" - the major language of Accra - and Hausa - the trade language of northern Ghana. Of the 780 "individuals surveyed, lesser competencies were claimed by 21 in French; 15 in Dangme; 13 in Dagbani; 10 in Fanti (in addition to Twi); 9 in Kabiye; 7 in Kotokoli; k in Nzema; 3 each in Krachi, Yoruba and Ahl'o; 2 each in Late and Akp.ossor; and 1 each in Konkomba, Basare, Frafra, Wala, Chokosi, Dagati, Moore, and German. People's claims for ability in their neighboring Central Volta Eegion lan- guages (as seen in Table 3-7) followed a pattern similar to that disclosed by the Multilanguage Test. Table 3-7 shows the claims grouped into northern and southern blocks. Ava- time was the southern language in which the most northern people claimed ability. Avatime people also stood out among southern communities as claiming the most competency in northern languages. (These claims, it should be stated, were made prior to the actual testing or exposure to the Multilanguage Texts.) The last four areas of interest in the Language Skills Section included lan- guages used to 1. Settle arguments, 2. w,rite letters, 3. count money or.goods, and 4. dream. The resulting percentages in these areas for the three major languages (Ewe, Twi, and English) plus the local language are set out in Tables 3.8, 3-9> 3-10 and 3^11 respectively. In Table 3*8 there does not seem to be any correlation between the languages people use to settle arguments and any previous data we fve collected. The figures may be an indicator, however, of the number of "strangers" or other non-local language speakers resident in the different areas - something we will look at later in our population studies. Akpafu/Lolobi, Logba, and Likpe had the three highest claims for using Ewe to settle arguments. Mionya had the highest claims for Twi ussage. Table 3-9 shows the dominant role of English in letter writing among five of the six northern language communities. Ewe, however, is used more than English by the southern communities as well as in Bowiri from the North. The Eastern Buems use their own language nearly as much as English in writing letters. They practically stand alone in this respect among the other nine languages. Table 3-10 shows that these ten language communities still retain their in- dividual counting systems. It also shows that four of the six northern communities rate English as the most popular second system, Bowiri people use Twi more than Nkonya Buem Bowiri Akp/Lol Likpe Santro Logba Avatime Tafi Nyangbo ui Ewe 39 X 9 34 24 45 49 (2) 37 (11) 17 6 (6) 11 43 37 1? 40 42 91 31 90 54 33 63 3 2 3 50 16 10 40 Twi English Ga Hausa X 14 71 (1) 12 (27) 6 4 7 36 34 7 9 20 14 18 (26) 19 (6) . 23 23 10 13 12 5 25 13 9 2 4 (1) 23 (49) 12 7 40 X 4 21 6 47 25 31 19 31 18 20 20 30 43 8 9 34 (1) 14 (5) 15 11 ;ii) 19 (2) 24 7 ' (1) 22 ( :ii) 25 32 16 22 21 12 21 X 4 1 2 3 3 1 3 2 2 5 (6) 2 (1) (1) (10) 1 (2) (3) 3 (46) 7 (13) 3 X 1 3 2 2 2 1 1 3 (2) 1 (3) (2) 2 (34) 4 (4) Table 3.6 Distribution of Claimants at Ability Levels (Number represent the percentage of the sample taken). Key: x Very Good / Fairly Good . Small Bit () Heard only TESTED COMMUNITY CLAIMED LANGUAGES NK BU BO A/L •• LI SA LO AV TA ••- NY / 3 4 5(1) (1 ) 1 1 (1) (2) (1)A2 ri1(6): 4(2)Y3 5(3) • 9 do) 2 2 2(2) id) 3 do) 3 1 10 (4) 1 1 (21 )K (1) 1 (1) "(4) / - 1- • 1 (1) (1). 1 - ( 2) (4) dv ( l) 13 11(1) / 1 (40) / 3 f-1) ' 1 1 1-^21) 1(16) 2 : (?) 4 / Nkonya Buem Bowiri Akp/Lc Likpe Santro Logba Avati Tafi .Vyang X / X / X / • X / X / X / • X / X / • X / • X / • (3) ( 1) 4(7). (3) / Table 3.7 Language Claims for Neighboring Languages POUT levels of claimed language ability are represented.- X - Very Good; / - F a i r ly G o o d. .-# . g^ 3.^ () - Identify Only. The figures stand for the percentages of the sample population who claimed competence at each level* ; 26 E T En LL NK 24 37 23 100 BU • 31 4 14 100 BO 3 3 1 100 A/L 70 24 30 100 LI 2*1 -If 19 100 SA 3 - - 100 LO 49 4 9 100 AV 6 - 2 100 TA NY 6 100 100 E T En LL NK 7 9 63 2 BU 36 - 64 59 BO 74 7 52 4 A/L 54 3 63 - LI 40 - 51 SA 53 7 6? - LO 80 - 51 AV 57 1 50 5 TA 52 - 42 2 NY 64 - 34 Table, 3.8 Settle Arguments Table 3.9 Write Letters E T En LL E T En LL NK BU 17 22 59 61 19 2 31 100 BO 12 49 10 91 A/L 67 20 67 81 LI 31 - 46 71 NK 1 1 2 91 BU BO 2 . - - 100 - 97 A/L 7 2 - 93 LI 7 1 4 94 • SA 10 - 43 63 • . SA - 97 LO 53 2 47 • 89 AV 17 1 31 78 TA 27 - 17 87 NY 41 1 23 71 LO 2 - _ • 98 ' AV TA NY 1 100 - 100 ' - 100 Table 3.10 Count money/goods Table 3.11 Dream 27 either Ewe or English, while most Akpafu/Lolobi people prefer both Ewe and Eng- ; lish equally over Twi. \ Except for the Avatimes, the southern communities rated Ewe as their dom- inant second counting system. The Avatimes cJLaimed English as second to their own. j j | These figures suggest that Ewe is the dominant language at markets among the t southern language communities with English and Twi used mainly among the northern \ j communities. . Some of our data so far gives cause for some interesting speculation about ; the social activity of the Avatime people. Avatimes, with their counting pref- erences, would seem more suited for the northern markets - especially Nkonya and Buem. Their Commerce and Travel figures are high according to Tables 3 and 3.1. (We find Avatime tied for second and third respectively among the popula- = tion groups.) And though they prefer Ewe to Twi, their Ewe comprehension scores and claims were comparatively low - a negative factor for trading in southern markets. If Avatimes do in fact trade in northern markets, it would explain why Avatimes displayed language ability in understanding the northern languages and why a knowledge of the Avatime language was complementarily found in most north- ern language communities. \ Table. 3.11 shows tha"f; very few individuals (of those who could remember) claimed to dream in a language other than their own. This question was included to give us an idea of the) depth to which the second languages have entered the subconscious, areas of the people's minds. 3.2.4 Language Learning Environment Section D of the Individual Interview Form, the Language Learning Environ- ment, dealt with the language learners perspective of the language interaction situation around him/her. The results from the first point of this section (i.e., of the languages spoken, a. where they were learned and, b. with whom they are spoken) basically gave us three language learning environments with variations for (a) the place and (b) the participants. The first environment was (a) the home or hometown and (b) the family or townspeople. This was the environment where the local language was learned. The second environment was (a) school and (b) schoolmates, teacher, or na- tive speakers. This environment was where, both the major regional language (Ewe) and the national language (English) were learned. The third environment was (a) market (either a town name to which the indi- vidual had traveled for business or the local town market) and (b) friends or na- tive speakers of the language. This environment included the many other languages in which people claimed competence including, sometimes, the major regional lan- guage. Each language was thus designated as having been learned in one of the three environments which we will refer to as 1. home/family, 2. school/mates, and 3- market/native speakers. The context most relevant to our study here is the environment in which Ewe was learned. The percentages of the ten group samples which learned Ewe in each of the last two environments is presented in Table 3.12. The Ewe comprehension level is also presented at the right of the chart for comparative purposes. No correlations are apparent according to the ranks designated in Table 3.12. A pattern does, however, emerge when the language communities are divided along a northern/southern parameter. As was done with the Haitilanguage Test results in Table 3.7, A new set of .ranks for the two groups and their comparative Ewe Test ranking {withxn their new grouping) are displayed in Table 3.11. 28 Age Groups 9-13 15-20 21-30 31-50 51 SM M 29 42 29 54 33 62 12 87 1? 79 SM SM SM S Nkonya Buem 56 19 81 - 44 37 37 31 19 62 Bowiri , 65 35 75 %5 65 35 50 50 "15 85 Akp/Lol 93 - 100 - 93 - 71 29 14 71 Likpe - 96 4 100 - 87 12 67 33 54 42 Santro 83 17 67 33 83 17 33 50 17 83 Logba 50 50 33 67 42 58 33 67 8 92: Avatime 69 25 94 6 87 12 .81 19 50 44 Tafi 100 - 100 - 62 37 25 75 50 50 Nyangbo 79 7 86 14 79 21 29 71 7 93 fo Average S M Rank S M 24 65 10 2 47 54 30 46 74 20 81 18 57 40 33 76 67 21 67 32 56 41 8 7 3 l 5 9 2 4 6 •7 3 9 10 5 1 8 6 4 Ewe Test Rank 8 10 7 5 2 9 1 6 3 4 Table 5»12 Language Learning Environment (Ewe) Percentages of the sample are given under the environment in which Ewe was learned (S=school/ mates, M=market/native speakers). N Nkonya 0 H Buem T H Bowiri E • • •' E Afcp/Lol N Llkpe Santro S Logba 0 U Avatime r H Tafi E R Nyangbo N Language Community Ewe Test Average Bank Rank s 24 47 54 74 81 57 33 76 67 56 M 65 30 46 20 18 40 67 21 32 41 S 6 5 4 2 1 3 Miiiii 4 3 2 4 6 3 2 1 5 1 4 2 3 Table 5.13 Language Learning Environment (Nortli/South Grouping) When divided along a northern/southern parameter, the ten language communities show a marked similarity between where they learned Ewe (in the north at school and in the south at the market) and how well they learned it. S = School/mates M = Market/Native speakers ' . . In the northern division where Ewe proves to be learned best in the schools the correlation would be even closer if Nkonya were left out. Since many of the schools in Nkonya towns use Twi as the vernacular language of instruction they do not fit well on a table which isolates the school environment column. The southern division shows that Ewe is learned best at the market place. Logba leads the southern communities, and those with mainly the school environment (Avatime in particular) trail. • V, ' The second point of section D, "activities'that Can be carried out using (a) only the local language and (b) only major regional language",- helps us see the different activities through which the individual is exposed to Ewe and the lo- cal language in the community. Table 3^1^ is a compilation of all of the activ- ities mentioned by individuals interviewed. Often people from the same language community mentioned the same activity but cited different languages which they used in order to participate. This points to,an activity where either both local speech and Ewe speech are separately used (such as "Singing"), or where different individuals use one or the other languages exclusively (as in "Praying"). The activities in,Table 3-1^ proceed down the chart, from the uniform 'use of the lo- -cal language to the total use of Ewe. ' ' . , The activities can be grouped into three categories: Traditional Rites (1-10), Public Cultural Expression (11-20), and Externally Introduced Activities (21-29). A few of the activities, marketing (2l) and prayer (l5)-public or pri- vate was not specified- for example, might be better included elsewhere. But most of the items are so clearly similar and follow the gradual:language, change pattern of the table that they've been left where they stand... . '.. The exclusive use of the local language is largely confined to activities in the first category, Traditional Bites. The third category, Externally Introduced Activities, is dominated primarily by the use of Ewe. Most of the groups use both languages for Public Cultural Expression, category two. The third point of the Language Environment section, ".the frequency of daily or weekly use of the major regional language", produced the results displayed in Table 3.15. The table shows the distribution of individual claims over the range of six possible answers. The numbers in each column represent the percentage of the population that was sampled. Value points were given in this table for the frequency of Ewe use which the individuals claimed. The highest score was &_ where the person claimed to use Ewe several times a day. The average answer for all the individuals tested appears in the column third from the right. Converted to a percentage figure, out of a possible 6 points, this score compares significantly to the claims of Ewe Ability reported earlier for the same people in Table 3-5 (see the two right- hand columns of Table 3«15» In the same manner, these scores show Bowiri, San- trokofi, and Tafi as rating themselves nearly 20% higher than their Eve Test scores. This, therefore, gives some veracity to the results gathered by the In- dividual Interview Form, since two unconnected but related questions have given correlating information. Opinions on the fourth point, "whether or not the major regional language (Ewe) is displacing the local language", are presented in Table 3-16. Two interesting results stand out on this table: one, that the highest percentage in the "yes" column belongs to one of the lowest scorers on the Ewe Test (i.e., Santrokofi); and two, the second highest percentage in the "yes" column is from the highest scorer on the Ewe Test (i.e., Logba). The second result mentioned is understandable: Logba people have proven quite bilingual, using the local language exclusively in only four speech events in Table 3-14. 31 1 2 3 4 5 b 7 8 for Privacy (advice j • Traditional Meetings — Funerals/Festivals Culture Traditions iaistooung Marriage Uusternary itites (l i hati on) Fetish Worship Teach Children Plays Games Arbitration Dances/Drumming Prayer Storytelling Pr ove rbs/Eiddle s Singing Talking Public Gatherings Marketing Write Greeting Strangers Preach Church Bead School 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 2b 27 28 Catholic Youth Groups (CYO.) — 29 Circumcision Bites Li Bu A/L Sa Bo Nk Lo L T, L L L L/E L/E L L/E L L L/E L/E L/E L/E L/E L/E L/E E E E E E L L L L/E L L L/E L/E Ii L L/E L/E E L L/E L L L L L L/E L/E L L/E E E E E L L L L L L/E L L L/E L/E L/E E L/E L/E E E E E L/E E E E E E , -p, L L L L T,/P E E L/E L L/E E E E E E L L/E L L/E L/E L/E L/E L L/E E B Ta Av L L L L L L L L _i/E l/TC L T./TC T,/K L T, I,/R E T/E -L/E T'/F E L/E E E. E L L L/E L/E L/E L/E L/E L/E L/E L/E L/E .E L/E E Table 3.14 Language Use Activities (individual claims for the Local language or Ewe.) 1-10 Traditional Bites 11-20 Public Cultural Expressions 21-29 Externally Introduced Activities 6 SD 19 24 75 49 58 5 ID 24 12 - 9 1 Nkonya Buem Bowiri Akp/Lol Likpe Santrokofi •53 10 Logba Avatime Tafi Nyangbo 73 32 81 32 5 22 16 51 4 SW 16 9 14 31 25 23 13 7 - 12 3 IW 18 22 - 7 1 3 - 22 3 4 2 SM 4 2 5 - 5 - 8 6 - - 1 IM 3 17 - - 1 - - 1 - - 0 Never Point Aver. Average Answer °/o Out of six Ewe claims 14 14 6 4 3 11 0 10 0 0 3.6 3.3 5.2 4.8 4.8 4.7 5.3 4.6 5-8 5-0 SW IW ID ID ID ID ID ID SD ID 60 55 86 80 81 78 88 77 97 83 48 56 77 73 73 77 92 70 97 82 Table 3. 15 Frequency of Ewe Use among Individuals (%) Key: SD _ several times a day SW _ several times a week SM _ several times a month JD - once a day IW _ Once a week IM — once a month Is Ewe replacing L? Yeg No Undecided Nkonya Buem Bowiri Akp/Lol Likpe Santrokofi Logba Avatime Tafi 11 24 32 19 35 63 50 6 2 Nyangbo • H 87 76 65 81 61 30 48 92 95 89 2 - 3 _ 4 7 2 1 2 10 fable 3.16 Opinions on of the! Local Beplacement Language by Ewe One would expect the major regional language to be taking a forefront position here. But in the case of the first result, Santrokofi, where 63% of the sample felt Ewe was replacing their local language, it is surprising that their Ewe comprehension scores were low with a kTfo average. Their claims from Table 3*5 were higher at a 77% average and so were their claims for frequency of Ewe use in Table 3.15 (at 78$). The explanation for this discrepancy may lie in the fact that the sample size was too small to accurately assess this smallest of the groups (smallest, that is, if Tafi and Nyangbo were to be put together and considered dialects of each other). There are only three Santrokofi towns from which we took a 30 person sample. The next smallest group with six towns was Logba with a 60 person sampling. This attitude was, however, supported with valid reasons by the individuals during their interviews. Table 3-17 shows the reasons given for the "yes" or "no" answers in Santrokofi as well as in the other language areas. Education (in Ewe) and the adoption of Ewe vocabulary were the two factors most commonly cited on the affirmative side. That Santrokofi has its own mother tongue was given as the main reason against. The general pattern shows that these ten language communities recognize their own languages (mother tongues) as being different from Ewe. The majority of the people except in Santrokofi and Logba, feel their language will not be re- placed in the years to come. The fifth question of this section - the last question of the Individual Interview Eorm - "would local language speakers financially support a literature program using their own language?" produced the results recorded in Table 3«18. Tafi (at 67/*) recorded the lowest percentage in the affirmative or "yes" column. It is followed by the two communities whose majority felt Ewe was re- placing their local language, Logba and Santrokofi. Tafi, however, led the list in Table 3>l6 with the feeling that Ewe would not replace its mother tongue. This indicates a general satisfaction among Tafis with the use of Ewe in the realm of written materials, while at the same time showing a predominance and preference for their own mother tongue in local affairs. In fact, the results-of Table 3«17 have a strong negative correlation with the Ewe Claim Scores (see Table 3>5) where Tafi also leads. In other words, those communities who feel most confident in their use of spoken Ewe feel the lowest need for a literature program in their own language. 3.3 The Community Interview Form The Community Interview contained two sections— A. Language Use During Public Functions and B. Language Contact Environment. Section A, Language Use During Public Functions was used to determine which languages are used in the community at different public gatherings or gathering places. The list of speech activities covered is as follows: l) government offices, 2) political rallies, 3) shops or stores, 4) market, 5) hospital or clin- ic, 6) church a. liturgy b. readings c. sermon d. announcements e. public prayer f. singing, 7) school a. primary 1-3 b. primary 4-6 through middle 8) customary ceremonies, 9) local council of leaders 10) town meeting. The results should give us a picture of the extent to which, and the lines along which, a second language has entered the local community. Section B, Language 'Contact Environment, provided estimates for l) local- language residents, 2) major-regional language residents, and 3) other-language residents. Estimates for categories 2 and 3 (the non-local language speakers) were qualified according to three parameters: residency duration (either per- Nil Bu Bo A/L Li Sa Lo AT Ta Ny 2 17 10 6 29 4 43 21 7 7 2 1 2 2 7 2 Yes: Education is in Ewe Ewe is commonly used Ewe vocab. creeping in Ewe is widely spoken Near to Ewe town People say so It is trying Intermarriage Everyone travels Ewe number system is more complete Church uses Ewe Our town is mostly Ewe people I don't know No: 13 1 2 7 1 i 2 6 3 1 7 1 3 1 1 1 1 8 4 3 24 1 53 2 2 3 1 It is our Mother Tongue They have separate roles Ewe is not spoken well here We are growing Local language used more Local language used with children I dream in my local language Arbitration is done in the local L. We use it for privacy 1 don't know We are proud of ours—-we want it Ewes are learning our language Social activities are in our language Pew Ewes live here The Ewes will move from our land We are many Maybe Twi will replace it 6 6 2 1 6 7 2 7 9 1 11 5 2h 4 10 19 58 1 1 46. 1 1 9 3 1 17 3 1 1 45 3 18 7 4 4 67 5 17 5 57 6 7 1 11 2 1 1 4 56 1 1 4 1 1 1 21 Table 3.17 (Percent of Population Sarnie) "Is Ewe Replacing the Local Lqn/rqage?' -—Why?" •• Skoiiya Baem Boviri JJqj/Lal Likpe Saatrokofi Loglja Awfeiiae fafi Nyangbo Yes 98 No 2 100 - - 92 8% 90 80 /O 90 67 91 2 9 - 10 15 2 2 - - - 6 7 10 10 7 7 30 9 Table 3*18. Opinions on Local Sujiport for J.-ocal I-nnguage L i t e r a t u re Program, ($>) manent or temporary), integration (either mixed or isolated, and location (either town or farm). Each of the factors chosen for this section gave us insight into the social interaction of the community. Out of this section we hoped to develop an indicator of the amount of interaction a local language speaker can be expect- ed to have with speakers of uon-local languages (ajid Ewe in particular). 3.4 Results of the Community Interview Form The settlement pattern in this central Vblta Region area, where families and clans are gathered in clusters around a village nucleus, enal)le each town site in a language community to exhibit its own pattern of Public Language Use as well as its own particular Language Contact Environment. For our purposes, the i~esults from each town will be combined to represent the "average" town site within each Language community. It would require a second treatise to present the language groups as individual case studies. 3.4.1 Language Use During Public Functions Results from Section A of the Community Interview Form are presented in Table 5.19. In this section the chief of a town and his elders were asked to list the languages used, locally during a particular public function. After discussing the matter among themselves, a spokesman gave the answer to the interviewer. The answers tore recorded for each town and then compiled onto a matrix of the entire language community* Some of the speech events were broken down, into their readily recognizable sab-units - church activities in particular, where different aspects of thc> ser- vice have different.functions. The amount of -time spent using the different Ian- i guages mentioned was not assessed, though it mar be relevant in some ca^os (e.g. where summary—translations are given). Because of governmental policy there is a natural break, too, in the use of ; vernacular languages in primary school. Though all schools throughout Ghana are encouraged (where possible) to use the child's mother tongue in the first three years of education, these ten groups have no literacy materials to be used. The : teachers have the.•sometimes difficult task of teaching the young child iu another vernacular language (mainly Ewe) before introducing the national language, English, as the medium of instruction from Primary 4 onward. The scores on line; 7a.. 01 Table 3,19 show the frequency (or infreqneney) with which a teacher has to resort to using the local Language - not Ewe - to assist the children in these oarJy stages. (Similarities can also be noted between this line and the •j.ennral pat- tern of Ewe Comprehension Scores.) The breakdown of evotits under line 6, "Church", of Tubli? 1.19 .show ut< thut when the content of the message is in focus, rather than the form of worship, many of the language communities resort to the use of their own mother tmi^is.'*. The percentages for Hne= 6c,d, and e (i.e., %hv ,-^rmmi, Hmionnawnlv, and pray- er) are considerably lower inmost cases than linr^ n,h, and f (i.e., i h<> Minf- gy, readings, and singing). Customary Ceremonies (liiu- a) c-otil.J proteinJy have been broken down into similar events except for thf ovflrwhe.lm.iuu *>vid..»oo Unit aJl -such activities arc. carried out sojfijy in the locaJ lnu<±nn&'. 1. Government Offices 2. Political Rallies 3. Shop:-' or Si ores h. Market 5. Hospital or Clinic 6. Church - H. 1iturgy b. readings Co sermon (1. announcements e. pub!ic prayer f. singing 7. School - a. primary 1-3 b. prima ry 4-m id.* 8. Customary Ceremonies 9. Local Council of Leaders 10. Town Meeting \!k 50. _!l 85 92 ' 100 62 11 21 79 100 8 1~ 29 liu 21 31 l'i 50 37 31 56 37 3 I 3 I. 37 62 T«i Ny 100 100 1 00 50 87 100 100 100 100 62 87 87 75 100 0 0 93 86 93 100 100 100 71 57 79 86 79 100 0 0 7 1.00 81 100 100 0 8 0 6 50 19 12 A/I.. I.i Sa Ho 2l 211 m 100 100 71 86 8 86 100 100 ;i3 78 hi i >\i\ > 100 0 11 39 50 '->() 0 100 66 100 0 0 0 *The medium of instruction gradually changes from F.we to English. Table 5.19 Mwe Usage During Public Functions The numbers represent the percentage of touns in ouch community which claimed to use 1'hve as a. moans of communication during Public inunctions. A score of 50 could denote either the simultaneous use of Ewe and the local, language through translation during a function, or that either language is used with equal, frequency. Numbers that are underlined denote acticities in which at least one-third of the towns claimed to use Twi as well as IJwe. Two irregularities under line 5, Hospitu] or Clinic, occurred because both of the low scorers, Buem and Logba, have clinics staffed by local language speakers The use of translation among these communities is most clearly seen in the statistics recorded for line 2, Political Rallies. In this event the people them- selves determine whether or not they understand the people addressing them and thev obtain translators accordingly. There is a strong positive correlation between these scores and the Ewe Ability Claims of Table 3.5. Tafi claimed a complete use of Ewe (1O($>) while Buem (at 31%) claimed a major dependance on its own language. 3.4.2 Language Contact Environment Our study of the different language populations resident among those ten central Volta'Region communities was aimed at determining the amount of potential interaction a local language speaker would have with an Ewe speaking person or pop- ulace resident in the local area. It was with this in mind that three parameters were set up to grade the quality of contact one could expect from non-local-lan- guage-speaking residents. The ratios of the language populations in these areas are presented in Table 3.20. Five ratios are presented in Table 3>20. The first shows the number of Ewe speakers as a whole compared to the number of local "language speakers. The second adjusts the ratio according to the typo of residence duration pattern kept by the Ewo speakers, i.e., whether they are permanent residents year-around or whether they are transitory and will move or come and go irregularly. The third ratio rep- resents Ewe speakers who live in either mixed (among loca.1-language speakers) or isolated (as in a. zpngo) integration patterns. The fourth ratio weighs the Ewe speakers according to whether they live within the town nucleus or outside its boundaries in farm cottages. The fifth line combines all three parameters to de- termine their total effect on the contact which local-language speakers can be ex- pected to have with Ewe speakers. The ratios show that on a straight percentage basis Logba, Bowiri, and Tafi have the most Ewo speakers living on their land per hundred local-language speakers. These communities lead the others until the ratios are adjusted according to "Resi- dence Integration." Here Tali shows' that the Ewes there live predominantly apart from the local townsfolk in their own groupings. The next line "Residence Location" shows that the Ewes on Tafi land also live in farm villages outside Tail towns. On the whole, however, Tafi rises again to its place just above Xyangbo in the final adjusted ratio. Bowiri drops, as Tafi did, at the line adjusted for residency integration. Only Nkonya (among the higher ratios) maintains a mixed living pattern with its locally resident Ewes. "Residence Location," the next line, shows that the Eves live mainly in the farm villages at Nkonya where only a few of Nkonya's predom- inately town dwelling populace receive the benefits "of extensive contact. The integration ratios of Buem and Santrokofi point to situations which possibly explain the low Ewe comprehension scores of these language communities. There are few Ewes living in a mixed environment with the local speaker-s. Even the final line of ratios show these communities as having little potential "home- town" contact with Ewe speakers. A language community like Bowiri, however, or Nkonya seems to be having more contact with Ewes. They mix freelv with Ewes and presumably could increase their Ewe comprehension scores in the future. The typical pattern from line four, "Residence Location," shows that the town centers are basically local-language centers. The ratio of Ewe speakers to 100 local-language speakers only exceeds 20 in one language community - Bowiri 40 No. of ftwe speakers* present per 100 local language speakers No. adjusted according to a. Residence Duration (Permanent or Temporary) b. Residence integration (Mixed or Isolated) c. Residence Location (Town or Farm) d. Duration, Integration, and Location Ni Bu Bo A/L Li Sa Lo Av Ta Ny 26 22 80 26 27 5 124 9 58 33 26 24 13 1 1 78 37 10 6 7: 28 8 12 2 1 5 124 64 19 6 6 4 56 17 17 21 10 47 11 U 3 69 5 30 28 ^Figures are based on survey enquiries into non-local language population totals. :L980 estimates were used lor local language totals. Table 3.20 Language Contact Environment (the skewing effect of Kwamikrom, a predominantly Ewe town, is in part responsibl for this). The major part of the immigrant Ewe populations are farmers renting the Jai on a share-cropping basis. Some of the land "has passed into native Ewe hands . through purchase, as was mentioned often by the Buem people, but most of the Lwej are expected to move on after the land they are now using wears out. j Those Ewes who live in the towns (the 20%) are mainly teachers.- Others uji elude traders from the North, carpenters, medical' workers, postal agents and coco produce buyers. Pew of these are permanent residents. 3-5 Summary of Questionnaire Results . ' ' The following information was evident.from the questionnaire resuJts. 1. High church membership correlated with high Ewe scores in Logba, Liipe, and Tafi. The two highest communities- under church membership, Akpafu/Lolobi and Buem, were fifth, and last in Ewe comprehension scores. 2. Educational levels were high for the first two scorers on the Ewe Comprehension Test, Logba and Likpe, The statistics also showed an overall pattern of extensive educational coverage•- among these ten language communities. 3. Occupation results showed farming to be the chief activity in each community with nearly 40% claiming it as their chief occu- pation. Farming proved to be neither a help nor a hinderance in acquiring Ewe comprehension ability. . Both Logba, the high scorer and Buem, the lowest scorer, had the same farming per- centage at 48%. Bowiri and Avatime, both of which averaged ., 61% on the Ewe Test? had the highest and lowest farming per- centages . k. Percentages in the_ Commerce column of the statistics on "Occu- pation" showed a relationship between a community's involve- ment in trade and their' Ewe comprehension ability. 5. The peoples' travel experience paralleled their Ewe compre- hension scores in all but three communities. The three - " Buem, Logba, and Likpe - were so divergent from the compre- hension score patterns that it would prove, fruitless to use "travel experience" as an indicator of a community's Ewe ability. ' „ - . .- 6. Individual's claims for 2nd language speaking ability were highest in Ewe, Twi, and English. Of these- thr.ee Ian- .. ' guages all communities except Nkonya claimed their highest abilities in Ewe. ' Nkonya had its highest claims in1 Twi. 7. Besides Ewe, Twi, and English, the next most spikes. Ian- " guages were Ga (the major language of'Accra) and Hansa (the trade language of northern Ghana)* '• 8. The highest £laims_ for speaking ability in the neighboring Central Volta Begion languges fell into northern and south- ern groupings similar to the results of the Multilanguage Test. In the same way, only the Avatimes from the southern communities claimed any noticeable familiarity with the northern languages, and only Avatime was claimed as a lan- guage which northern communities knew to any extent. 9. Akpafu/Lolobi, Logba, and Likpe had the three highest claims for using Ewe to settle arguments. Nkonya had the highest claims for Twi useage in this matter. 10. Letter writing among the northern language communities is pre- dominantly done in English (Bowiri was the only one to claim a higher Ewe use). Buem stands out because it claimed to use the local language nearly as much as English. The south- ern communities all rated Ewe as the language they use most in writing. .11. The languages used in counting money or goods also followed a northern/southern grouping pattern. After the general mention_ of their own_ counting, systems^ Ewe was. named as the pfelerre'd counting system in the" south "and English in the north. Only Avatime in the south mentioned English more than Ewe, while Bowiri from the north claimed more Twi useage. 12. Few individuals said they found any languages other than their mother-tongue entering into their dreams. 13. The highest scorers on the Ewe Test from the northern six lan- guage communities claimed to learn Ewe mainly at school. The highest scorers from the southern communities claimed to learn Ewe mainly in the market. 14. Claimants nearly all agreed that traditional rites (or customs) vere carried out exclusively in their local language. Activ- ities of public cultural expression (such as singing, plays, or storytelling) could be done in either Ewe or the local lan- guage. Externally introduced activities were mainly done in Eve. 15. Claims for the frequency of Ewe use paralleled the communities claims for Ewe proficiency. Tafi was highest (as in the Ewe claims), claiming the Ewe useage of several times a day, and Buem was lowest, with an average claim of once a week. 16. Santrokofi stood out as the language community with the high- est claim t?aat Ewe is replacing their language. Logba follow- ed with a 50% affirmative answer. 17• Buem expressed the most positive support for a literature pro- gram using their own language. Tafi, the community with the highest Ewe ability claims, had the lowest positive support claim. 18. The town site claims for "Language Use During Public Func- tions" pointed to the fact that teachers in some language communities others when working with children in the primary classes 1 to 3. The statistics on this subject followed the averages for Ewe comprehension quite closely. resort to using the local language more than 19. During a public function where the message of an activity was in focus and not just its function (e.g., in a church sermon or church announcement as opposed to the church liturgy), many communities- showed a marked dependance on the local, lan- guage rather than Ewe. 20. The use of the local language during political railies-pattern- ed in a. way similar to the Ewe Aba Jit;, Claims'of Table 3.5. Tafi claimed a complete use of Ewe (100°c) while Buem (at ~~>l°/c) claimed a major dependance on its own language. 21. The highest residence ratio of Ewe.- per 100 local-language speakers was in Logba, where the LWPI- outnumbered the local people- 1.24 to J. "In-town" residence figures show, however, that not even 20 Ewes per 100 local people reside in the aver- age Logba town. Nearly all resident Lwes live in farm cottages (or villages) outside the towns of these ten language communi- ties. • , ! ! , I 22. integration residence figures show that Beum and Santrokofi have : as? few as one Ewe-speaking person per 100 local residents living ', m a "mixed" living pattern in their town sites. •).() Conclusions from Questionnaire Besults •; It was t-Jear from the results in this third section, that the local lan- guages are actively used and that each language community feels like .a viable entity. Eve does have a place, however, in the language activities of most, of the communities and is readily acknowledged by most individuals as a lan- guage which they can speak with proficiency. " . The residence patterns of our study show that few of the communities have any extensive contact with Ewe speakers on a day-to-day basis. Of the northern language communities (where Ewe scores were low), only Bowiri and Nkonya mentioJ ed a mixed living pattern with Ewes among them.- The southern communities seem to have developed a rather stable .situation with Ewe used in certain activities and tie local language used in others; More likely than not, this situation toil persist unless integration patterns change and a factor such as intermarriage (not really focused on in this survey) becomes more prevalent. r Another question which might have proved helpful in evaluating the lan- guage use attitudes in this area would be, "What language(s) would you like to know better?". Though I doubt it would have given us any further insight into the ralationship between the local language and Ewe, it may have given us an ids about people's preferences for languages of wider communication, say among Ewe, Twi, English, and French. '•f. IMPLICATIONS: WHAT DOES THIS SURVEY TELL US? Hopefully the information from this study will be clear enough to answei questions that officials and executives have about the sociolinguistic pic- ture m this portion of the Volta Begion. Some of the relevant facts that stand out about each language community, and which give a cohesive view of the people in their present position with re- gard to their use_of Ewe, are presented below in alphabetical order. 4.1 Correlating Factors Akpafu/Lolobi: Group averages show relatively low Ewe comprehension (b<#)- — The men, however, in the 31 Lo '">0 year age group demons 46 Footnotes 1. The government policy for instruction in the schools in eight of these ten central Volta Region language communities is to use Ewe in primary school with a gradual change over into English as the medium of instruction from Class 4 going on. Buem towns use either Twi or Ewe, while Nkonya towns use both (i.e. Catholic school use Ewe, Presbyterian schools use Twi, and the Local Authority schools use a mixture.) • • 2. A discussion of text types can be found in E. Casad's work, Dialect .Intelliffi- gibility Testing, 1974. 3. I am grateful to Prof. Christian Baeta, chairman of the Ewe Project for the Ghana Bible Society, and Mr. George Kom, head of the Ewe Project under Living Bible International, for their permission to test with these early drafts of their work. The express aim of both translators was to convey the meaning of the Greek text into the idiom of the Ewe population, coloquial Ewe. When we played the tape re- corded versions of the texts for mother-tongue Ewe speakers, the response was quite favorable. Not only were the texts easy to understand (test scores are pre- sented elsewhere), but they were acclaimed as tremendous improvements over the former versions. Of interest is the fact that native Ewe speakers chose one text over the other as being the most eloquent. Though both were considered "standard", they said the first had been read by a southerner. The other text they claimed was read by a Ewe from the north of the Volta Region. In fact, both had been read by the same man, a native of Bzodze but resident at Hohoe, Mr. S.K. Dewotor. Each new pair of survey assistants from the non-Ewe language groups chose the text written in the northern style (not the most eloquent) as the one their people would most easily understand. Simplicity and clarity, they claimed, were its strengths. 4. When the Ewe Test results were tabulated without the third paragraph, the ab- stract hortatory one, all ten language communities showed a slight rise in their average scores. (The greatest increase was 4^.) This illustrates the increase in complexity which the test was designed to include. 5. Two "control" tests were run; one to check the validity of our sample size (10 people per town), and a second to determine the averages of native Ewe speakers on the Ewe Comprehension Test. The first test showed us that the average score from a forty member sample did not vary significantly (statistically) from the average scores of two separate groups of ten chosen from the same community. Still, to guard against the possible skewing effects of gathering a non-representative sam- ple, we took test samples from every population center (an average of 7 towns per language group). The second control test gave us an average score of 9.25 (92.5^) out of 1@ on the Ewe Comprehension test for a sample group of 40 Ewes (chosen equally from the five age ranges). To adjust the non-Ewe scores using the Ewe score as 100^ would mean adding 7.5$> to each group average. The range of scores would then be from 9.45 to 47.5 for Logba and Buem respectively. This would make it awkward, however, when comparing scores from the questionnaires (e.g. claims of Ewe ability, etc.), since no comparable^data was taken on the Ewe group. To maintain unifrom- ity I will continue to quote the raw scores throughout the text. 47 6. The totals reported in Table 3 and the upcoming tables are representative of the people whose Ewe comprehension scores were reported here earlier. The survey results do not claim to represent all speakers of these languages, but only those resident in their traditional areas from whom the sample was taken. 7. Why is there such a discrepancy in the correlation of high church member- ship and high Ewe Comprehension among the Akpafu/Lolobi and Buem? The discrepancy is not quite so radical for Akpafu/Lolobi which actually lies second to Likpe among the northern communities for Ewe comprehension. But for Buem we must look elsewhere for an explanation. Certainly the high interest among Buems in church activity can be attributed, to a great extent, to the policy of local language use in the dominant E. C. church and the practiced pattern of Ewe/Locaj Language translation found in the Evangelical Presbyterian church throughout Eastern Buem (see Table 3.19 for figures on the dominant use of the local lang- uage during church activities in Buem as compared to the other language commun- ities ). It was the work of Hev. Pa. A. 0. Dougli, Africa's first Catholic Priest, (and a Buem man), that brought written Biblical material to the Buem area. The Catholic Missal (1956) is still in use throughout Eastern Buem. 48 BIBLIOGRAPHY Ansre, G. and Ohannessian, S. 1975. Some reflections on the uses of sociolinguis- tic surveys, Language surveys in developing nations, ed« by Ohannessian, Fer- guson, and Polome. Center for Applied Linguistics: Arlington, Virginia Bendor-Samuel, J. 1979* A sociolinguistic profile. Unpublielied worksheet. Sum- mer Institute of Linguistics: Horsleys Green, High Wycombe Caldwell, J. I969. African rural-urban migration: the movement to Ghana's towns. McGill-Queens University Press: New York (Columbia) Casad, E. 197^> Dialect intelligibility testing. Summer Institute of Linguistics: University of Oklahoma, Norman Dougli, A. 1956. Koyu uburjgun. Catholic Missal. Out of print. Frempong, J. and Leonhard, B. 1978. A survey of Twi use in the northern Volta Re- gion. Unpublished report to the Ghana Institute of Linguistics: Tamale Heine, B. 1968. Die verbreitung und gliederung der fogorestsprachen. Dietrich Reimer Verlag: Berlin Rropp-Dakubu, M. 1977« Report on the Hill Guang linguistic survey. Unpublished manuscript: Legon Moser, R. 1979. Sociolinguistic data concerning the Sabon Zongo of Accra. Paper submitted at the Workshop on Multilingualism in Ghana: Legon Mukarovsky, H. 1976 & 1977-. A study of Western Nigritic, Vol. 1 & 2. Institut fur Agyptologie und Afrikanistik: A-1090 Wien, Frankgasse 1 Westerman, D. 1927. Die westlichen Sudansprachen und ihre Beziehungen zum Bantu. Mitteilungen des Seminars fiir Orient. Sprachen, Beiheft zu Jahrgang 30 .49