viii" : l|}‘_ _ 574% ‘ l .‘ 7. 5)- 71; '77“; R q ‘- 1' h Eiffak'n "'H .31 £1177," . '7 ” 37.3"?" 7;. .1, .1 3:7" ”Wt“ ,7. ..'..,;§‘1,1,;~;~.;.7. 7 5,131}: ‘MT ‘3'}: ”if: H "'3“ 4cm: 6-“ - .I ‘5‘ 5‘: "5'5“ ”J '17 - ; 13;: ‘ .7;-72.. :73 .7 5.27 .7 '. t)- . nu‘ , . I. . 77.71.7977!“ ,gngvuqr ‘33) 7‘77“}; ER“ VIPER “mg, . ,v , fi§%%@%g&il“‘ C W {tzjw‘ .‘Q in. uni" 1’34“: ‘i L ‘ W 55754;”.34" ‘HLLm'H' ' ' ‘ 7.7 L ”'- - ;'~l-“~fi:¢ If 31“ v ' u' “ ‘ g V I V) .9 _‘. .. .L~.b~~.‘§‘ afih'fitzf'fidffi 7&7: f 77:97:71." '5‘ '~ 5‘85““ QS‘E.‘ '1 “c7 y+$§; k . ‘ 7.75.7777: . . ‘ 1,; 7,, 7, a; .. 7L . $526153,“ :7: H”: y. Lu-g-{Rv‘i 3:” 77.3" 58.3% ,. ~ 7.“!77'7“ v nghaé?‘ ‘63 . .1 7 1377.. 5 . ' r ‘ jk {- IL :- - a - A... .2 7 tag... i 43:39:71. Qty-vim .14. l :7; "i 73;» 6-. WV 0- . Wu 1‘5». 75777 ”I . “ 212*, .,, "1.37.437, " 4" 7.}:7M&%§%%L17 ‘52. w ‘7" 55753351577 ‘ 1‘35 ’ 1‘ 'MEfiit'nfi'r‘mfi'f-E74 ‘ i - . .. . 7.. 7‘37 , 7:127. 7 ' 3’13— ’ ‘ L's ' , . ' , 1‘5“" .7 " “:1 if) ‘ ’9‘?!“ I 33.95:; ' . If ‘ MJ :' ‘7». " ' 5&3. $3“ ‘QEJE Wééxnpu‘ufid gs: “3;“ .- 93191331? ‘7‘: "i . .. 7711‘ '1' w a. "’51,... LwhN-‘z 1‘1 7&31. ‘ .. ~ , E5§¢ an. 7-. JP (5712;133 km. was}? ”i“; ‘L {YT “2571;. --.-. .2 1 €111: , Eqég‘ 3:11;”, “Lg-Wu, gibffglm .ngi""5““"“ ‘55".57‘i;'». ‘35 5 5% 5“» £7,773; \5‘ ~ ~- ‘ 77-- {zyl 57,. 2' g» “7‘7 - - u-L‘u ’7‘ _ y ‘ mum: ’~ ‘0’; 77...; ~ ' "5. ,. A " , mg? , $7 "sm%.gx 3' ‘ 2. '3'; "7“?" ‘:. ;17;3;- ‘ " 7. 'u h 1. ‘L 3“". . w" “’7: " 57,237,173: 5f...:;‘7‘:" 727. "“517” 574%? ' .' '1» $1: ‘ 7'51. ;_ .'_. ”it“: 1>-. ""3. ' ~ 4";- 7 5757-337” "5": 7.. ;;‘T§7§v .‘ {“3 4‘ 4’" . . 7777597753" - 7: 5. “7‘ 5 4. 5 ’5 7.7.9: 7.1:“ . . L" J ' 3157:7377. a" §L£k§$y¢2h¢§mr in?“ ‘ .7 . .77 I. -. .7, 7,71. “37!: i, ‘5” . {~77 \.w:‘l:.':£ A7. |~~"~ml~'% ' ' :mEfi ‘3" 0, m“ 3.1g.2;.‘4’1741‘.t;‘3k::k t 5 J11 .- ‘ . . h: -m ~41}- ‘1571557‘5 “573 ”5 b.4777 _ 577.... a”? {$557.1 Ema." :7" ”17115 .Lg.‘ V. £54 , .1. ‘7 7"; L. "3; ‘ Tu. '..‘ 017734111111“ 7 7 v m. q. 1-5.» ’41.; IS H? I; I)ate IIIIIIIII IIIIIII IIIIIIII II I II 0094 .4282 vmaux...‘ Michigan State University This is to certify that the thesis entitled DETERMINANTS OF LANGUAGE CHOICE BY MIDDLE LEVEL MANAGERS IN THE KENYAN PUBLIC SECTOR presented by Mwangi wa Kariuki has been accepted towards fulfillment of the requirements for Master's degree in Communication , ( 7m- (51*... Major professor 2/19/86 0-7639 MS U is an Aflirmative Action/Equal Opportunity Institution bViESI.) RETURNING MATERIALS: Place in book drop to L13RARJES remove this checkout from , your record. FINES wi11 be charged if book is returned after the date stamped be1ow. 1 DETERMINANTS OF LANGUAGE CHOICE BY MIDDLE LEVEL MANAGERS IN THE KENYAN PUBLIC SECTOR BY Mwangi wa Kariuki A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1986 ABSTRACT DETERMINANTS OF LANGUAGE CHOICE BY MIDDLE LEVEL MANAGERS IN THE KENYAN PUBLIC SECTOR BY Mwangi wa Kariuki This study attempted to examine what determines whether the middle level manager in the multilingual Kenyan public sector speaks English, Kiswahili, or first language (mother- tongue) when interacting with other organizational members. A self-administered questionnaire was administered to 119 public officers. Chi-square (X2) statistical tests were computed to find out whether there is a relationship between language competence, age, educational status, occupational status, attitudes, and the language(s) the middle level managers use when interacting with their superiors, peers and subordinates. However, no significant relationship was found between all the variables (above) and language choice. But a pattern of language choice and language use can be identified. When the middle level managers interact with their superiors and peers, they mostly use English alone or in combination with Kiswahili. While when interacting with their subordinates, they mostly use Kiswahili alone or in combination with English. It was concluded that both English and Kiswahili are vital medium of inter—ethnic communication and none of them can be used singly. Dedicated to Waitherero and Wangui ii ACKNOWLEDGEMENTS This thesis would not have been realized without the assistance and encouragement of many people. Special appreciation is expressed to Dr. R.V. Farace, who as my academic advisor and Chairman of the thesis committee a guided and encouraged me throughout the writing of this thesis and throughout my graduate studies. Gratitude is also expressed to Dr. Jack Bain and Dr. C. M. Scotton of the Linguistic Department for their invaluable guidance, encouragement and suggestions throughout all phases of this study. Special thanks to my friend J. Bagakas whose expertise in computer and statistics proved invaluable. I am also indebted to the former principal of KIA, Mr. J. K. Gichangi, and Mr. S. Karigithe the Head of Communication department at KIA for their assistance in the administration of the questionnaire and the many documents they sent me to enrich this study. Special appreciation is expressed to my colleagues at KIA who administered the questionnaire. Gratitude is also expressed to all the public officers who willingly answered the questionnaire. My friends at MSU — A. Mwangi, A. Mwakikoti, Y. Rubanza, P. Ngatagize, C. Langat and P. Biwott are especially worthy of recognition for their emotional support and encouragement. Thanks too to my friends in Kenya who have supported me in many ways for the last two years. I am deeply grateful to Dr. P. Theuri of the Continental Bank, Nairobi, who as my instructor, friend, colleague and mentor encouraged me to persue my graduate studies. I should also thank F. Verillo for providing excellent typing of this thesis in face of the deadlines. I am also deeply indebted to my parents, brothers and sisters whose sacrifice has helped me persue higher studies. Finally, I wish to express my deep appreciation to my wife Waitherero and my daughter Wangui whose genuine love and encouragement transcended the barriers of space, distance and time. Their support and patience encouraged me to make the most of my abilities. In difficult moments, Waitherero and Wangui upheld me, and in times of triumph they shared my joy. AHSANTE SANA. iv TABLE OF CONTENTS List of Tables Chapter 1. I Introduction Statement of Purpose Rationale of the Study Definition of Terms. . Research Question. Description of the Variables II Background of the Study: The Kenyan Language Situation. . . . . . . . . . . . . . . Chapter 2. English. . First Languages. Kiswahili. The Kenyan Language Policy I Theoretical Perspectives An Overview of Language Choice Domain Model . . Accommodation Model. Markedness Model II Review of Literature Chapter 3. Educational Level. Language Competence. . Occupational Status. . Directionality of Interaction. Age. . . Ethnicity. Attitude Methodology. . Setting of the Study . Selection of the Respondents Description of the Respondents Instrument . . . . . . . . Data Analysis. Page vii (”@00me HH M00040 15 16 20 23 28 28 3O 33 34 35 36 37 37 41 42 43 44 46 Chapter 4. Chapter 5. APPENDICES Results. Discussion Language Competence. Age. . Educational Level. Occupational Status. Attitudes Towards Kiswahili. Attitudes Towards English. Ethnicity. Directionality of Interaction. Language Use with the Public Theoretical Implications of the Study. Suggestions for Future Research. Conclusions. . . . . . . . BIBLIOGRAPHY vi 48 63 63 68 69 76 8O 87 90 97 101 107 Table LIST OF TABLES Frequency of Claimed Competence in Kiswahili by Language Choice Frequency of Language Choice by Age. Frequency of Language Choice by Educational Level. Frequency of Language Choice by Occupational Status. Percentages of Attitudes Towards Kiswahili by Language Choice Percentages of Attitudes Towards English by Language Choice vii Page 49 51 53 56 58 62 CHAPTER 1 I INTRODUCTION In all human organizations, physical, human and technological resources are managed to ensure that the goals of the organization are achieved. During the planning, control and coordination of the organizational activities, workers in all organizational levels interact. Interaction therefore, is a fundamental aspect of the organizing processes in an organization. Central to the interaction process is the language that is used by the interactants as the medium of communication. Language, as Graff (1983, p.220) observes "lies in the heart of communication." Effective language use is a vital interactional phenomenon in both monolingual and multilingual organizations. However, in a multilingual organization, effective language use can only be realized if the organizational members choose the language they use appropriately. When choosing one language rather than the other(s), a speaker must take into consideration the varying competence of the other interactants in the heterogeneous linguistic environment. Thus, choosing the appropriate language to use in a given interaction in a multilingual , :‘ix':~r " ‘ organization is an important communication skill that may have positive or negative effects on the overall communication climate in the organization. For example, for a middle level manager in a multilingual organization to communicate effectively , he/she must make an appropriate choice of language when interacting with superiors, peers and subordinates. Statement 9: Purpose The primary purpose of this study is to examine the determinants of language choice by middle level managers in the Kenyan Public Sector. As the Kenyan Public Sector is a multilingual organization, the study will endeavor to identify what determines whether the middle level manager speaks English, Kiswahili, or first language (mother-tongue) when interacting with superiors, peers and subordinates. Rationale :9: the Etggy The rationale for the topic under study is three-fold. First, the language a government officer chooses to use in a particular interaction is important if effective communication in the multilingual organizational environment is to take place. Choosing the most effective language, for instance, may help an officer interpret government policies and regulations effectively when interacting with other organizational members. Second, as the middle level managers are centrally positioned in the organization, they act as "bridges" between the top officers and the subordinates. Therefore, their organizational position requires them to interact with three distinct groups (superiors, peers and subordinates) which requires an appropriate choice of language. Finally, the findings of this study may be practically useful to the Kenyan Government in the formulation of future language policies, especially in respect to the education and training of public officers. Qfieitgge 9: T_e£m_s_ A Language is a linguistic variety that speakers use as a medium of interaction. In this case, English, Kiswahili and first languages will be considered as three distinct languages. Laagaaga Choice is the unconscious or conscious action of selecting and using one language rather than the other(s) in a given interaction. Migala Layal Managers are officers holding middle ranks in an organization. They are supervised by the top officers, and they in turn supervise the subordinates. According to the Kenya Civil Service organizational structure, which is alphabetically marked from Job Group A to T (there is no Job Groups I and O), officers ranked in Job Groups E to M will be considered the middle level managers (Daily Nagaaa, 1985, October, 2; Republic of Kenya, 1985). The Paglia Sagtgg is defined as all government ministries and departments, other than the Armed Forces, that provide direct services to the public. §£QE§ 92 SEE §£EQY QBQ Limiiaiiens The focus in this study is on the middle level managers mar!“- in the Kenyan Public Sector. The top public officers and the subordinates were not included in the study. Although it would have been viable to have a cross-sectional study by focusing on all workers in all organizational levels, time, finance and distance from Michigan State University to the site of the study were prohibitive. Further, although the external organizational environment may have an effect on language use, the study only examines the usage of language within the public sector. This means that the findings of the study may not be generalized outside the public organization. In addition, the questionnaire used in the collection of data may not provide all the information required in such a study, however, other field research methods such as "participant observer" which could have supplemented the questionnaire were not possible because of the constraints mentioned above. Research Question One major research question will be focused on in this study. The question is: Will the middle level managers' language competence, educational level, age, occupational status, ethnicity, attitudes towards English or Kiswahili and the directionality of interaction determine the language they choose to use when interacting with their superiors, peers and subordinates? Description 93 £22 gagiablaa The focus in this study will be on the following variables: educational level, language competence, occupational status, directionality of interaction, age, attitude, and ethnicity. Educational Layal: the level of education a middle level manager has attained prior to joining the public service or while in the public service. The various educational levels are: Primary School Level (7—8 years); Secondary School Level — O'Level (12 years); Higher Secondary School Level — A' Level (13-14 years); and University or College Level (includes Polytechnics, Diploma Colleges, Professional Schools etc.). Laagaaga QQEEEEEEEE is the claimed knowledge and ability to speak a particular language in a given interaction. gggapatagaal Sgagaa is the organizational position of a middle level manager as per the organizational hierarchy. Thus, a middle manager has higher occupational status than the subordinates, equal status with other managers holding similar managerial positions, but lower status than the top officers. Directionality a: interactaga is the pattern of interaction according to the organizational structure. From the middle level managers' position, the interaction may be upward to the top officers, downward from the top officers to the middle level managers; downward from the middle level managers to the subordinates; or horizontal from the middle level managers to peers holding similar positions. Aga indicates how old a middle level manager is. Attitude is the disposition to act positively or negatively toward the use of a particular language, in this case, English or Kiswahili. Ethnicity is the use of a first language in formal situations as an indication of ethnic consciousness and identity when interacting with members of the same ethnic group in a multi—ethnic organizational environment. II BACKGROUND OF THE STUDY: THE KENYAN LANGUAGE SITUATION In this section, Kenya's degree of multilingualism, the three distinct languages (English, Kiswahili, and first languages), and the Kenyan language policy will be briefly discussed to shed light on the research problem. Multilingualism in £2222 The Kenyan public sector is a multilingual organization. The linguistic heterogeneity is marked by a wide variety of languages spoken by members of different ethnic groups. Scholars differ as to the exact number of languages spoken in Kenya (German, 1968, 1974a; Rhoades, 1977; Hofer, 1978; Heine, 1980). For instance, Knappert (1965) notes that there are 22 languages, while according to Harries (1983), over 60 languages are spoken in Kenya. However, giving a precise number of languages spoken in a multilingual society is difficult (Alexandre, 1972). According to Alexandre, there are theoretical limitations because linguists have not agreed on the specific principles to be applied to distinguish a dialect from a language. Furthermore, sociocultural and political criteria, rather than primarily linguistic criteria are used to determine which linguistic varieties belong to a particular language (Penalosa, 1981; Ladefoged, et al., 1972). In Kenya, a triglossia language situation exists, with a first language (mother-tongue) spoken by members of each ethnic group; Kiswahili spoken as a lingua franca, official and national language; and English spoken as a lingua franca and official language (Gorman, 1974b; Mkilifi, 1978; Whiteley, 1979). About two—thirds of the Kenyan population speak a second language(Heine, 1970). Over two—thirds of the second language speakers, speak Kiswahili as their second language and less than one third speak English as a second language (Heine, 1980). Only a small proportion of the Kenyan population speaks English (Knappert, 1965), and most of those who speak English combine it with Kiswahili (Heine, 1970; Ekwelie, 1971; Ohly, 1974; Whiteley, 1974b; Scotton, 1976a). English English is the official language of Kenya. Since the British colonial era, English has been acquired only by those who have gone through the formal education (Scotton, 1976a, 1977; Heine, 1977; Parkin, 1977; Heine, 1980). As Mazrui (1969, p.221) states: the ability to understand even spoken English, let alone written English, presupposes in East Africa a degree of exposure to formal education. The situation is not one in which one can easily pick up English casually by the ear . . . for the time being acquisition of the English language lies in the universe of the literate culture. Although English is the medium of communication of government and education, it is however an elitist language that is a scarce resource to the majority of the population. English, Wallwork (1978. p.56) says "is the elite language, the language of education and power." The majority of the Kenyan population cannot understand and speak English. Therefore, as Mazrui (1969, p.100) notes, "English . . . for intertribal (inter-ethnic) communication at the grassroots is intrinsically and hopelessly ill-equipped to meet the challenge." The Kenyan elites therefore speak English as a symbol of socio-economic status and power that distinguishes them from the majority of the people who have not acquired and mastered it (Bujra, 1974). Thus, English as Scotton (1978, p1732) states, "restricts access to the high status sectors of society to those who have command of the elite language." However, not all those who have attended school are competent in English. Heine (1980) observes that primary school graduates lack competence in English if they do not have the chance to constantly use it. Therefore, the middle level manager works in a public organization where the peers and the top officers are elites who can effectively speak English, but the subordinates whom the officer administers may lack competence in it. §i£§£ LQEQE§H§§ The Kenyan African languages are linguistically categorized into four groups, namely: the Bantu languages (e.g. Kikuyu, Kamba etc.), the Nilotic languages (e.g. Dholuo), the Para—nilotic languages (e.g. Nandi, Maasai etc.), and the Cushitic languages (e.g. Somali, Galla etc.) (Gorman, 1968, 1974a; Whiteley, 1974a; Parkin, 1974b; Scotton, 1978). About two—thirds of the Kenyan population speak Bantu languages (German, 1974a; Whiteley, 1971). There are also languages of Indian origin (e.g. Gujarati, Hindi etc.) spoken by the Indian immigrants (Gorman, 1968). As is the case for any group of related languages, African languages within the same group differ in their syntax, lexicon and phonology. The degree of mutual intelligibility between the languages, however varies according to the particular languages. Of course, the languages of one group are mutually unintelligible to those of the other. For example, the Luyia language, which is of Bantu origin, is mutually unintelligible to the Cushitic Somali language. However, all the Luyia language speakers (who speak over 15 dialects) do not understand each other very clearly (Itebete, 1974). Like the Luyia language speakers, the Kalenjin people speak six dialects (Kipsikiis, 10 Nandi, Keiyio, Sebeei, Tugen and Pokot) but as Toweett (1979, p. xiv) observes, they "understand one another when they intend to be understood. . . . But when the intention is not to be understood then . . . (they) may not understand' one another." Kiswahili is a Bantu language that is syntactically, lexically and phonologically related to all the other Bantu languages. Although the Bantu language speakers have an advantage when learning and speaking Kiswahili, this does not mean that the non—Bantu speakers cannot learn and speak Kiswahili fluently (Halliday, 1972; Itebete,, 1976). First languages are often used in intragroup communication by members of a given ethnic group (Mkilifi, 1978). However, they may also be used as second languages in inter-ethnic situations (Heine, 1980; Scotton, 1982c). According to Heine, in a heterogeneous linguistic environment, the minority group learns and uses the first language of the majority groups as a second language. Kiswahili Kiswahili is a language of wider communication, not only in Kenya, but also in the neighboring East African states (Steere and Madan, 1890; Harries, 1968; Heine, 1970; Scotton, 1976a, 1977; Kihore, 1976; Idakwa, 1978; Datta, 1982). Today, Kiswahili is the "best known, most widely taught and commonly spoken African language" (Hofer, 1978, p.25). However, although Kiswahili is taught in schools, many peOple acquire it informally (Scotton, 1976a, 1982b; 11 Parkin, 1974a, 1977; Heine, 1977). Although Kiswahili has borrowed words from Arabic, Portuguese, English, Persian and Hindi languages, about 71 percent of the Kiswahili words are of Bantu origin (Harries, 1968; Indakwa, 1978; Hofer, 1978). Thus, it is an African language that is genetically related to other Bantu languages. Also, it is spoken as a first language by a small population living along the Kenyan Coast. However, the majority of the Kenyan population speak it as a second language (Heine, 1970, 1980; Mazrui, 1972; Rhoades, 1977; Harries, 1983). As a link language between speakers of different ethnic groups, Kiswahili has been recognized as an effective medium of inter-ethnic communication for a long time (Itebete, 1976). For instance, Steere and Madan (1890, p.22) emphasizing the importance of missionaries learning Kiswahili said, "they (missionaries) will carry with them a key that can unlock the secrets of an immense variety of strange dialects." Though the various East African languages could have been "strange" to the missionaries, the role of Kiswahili as a lingua franca is underscored. Further, in the political arena, Kiswahili has been a vital tool in national integration and solidarity (Mazrui, 1969; Ohly, 1982). The Kenyan politicians used Kiswahili to unify the various ethnic groups during the struggle for independence. The politicians, as Harries (1968, p.416) notes: 12 . saw in Swahili language a unifying factor, and its use was encouraged by official action as being first and foremost an African language capable of strengthening the cause of nationalism. Today, politicians use Kiswahili in linguistically heterogeneous communities when interacting with members of the public. Like the politicians, government officers use Kiswahili when interacting mostly with the subordinates and members of the public who are not competent in English. Generally, Kiswahili is Kenya's most important medium of inter-ethnic communication in government, business, social and political situations. The Kenyan Language Policy The Kenyan language policy has been revised a number of times since the British colonial era (Halliday, 1972; Gorman, 1968, 1974a; Scotton, 1978). Furthermore, language practice often has been contrary to the formulated policies (Scotton, 1978). English became the official language during the British colonial era, and its status has not changed in independent Kenya. As Whiteley (1979, p.45) states: the defacto adoption of English as an official language . . . is in conformity with an essentially elitist ideology, where the country is administered by a professional Civil Service not very different from that inherited from the Colonial Government. The colonial language policy was geared to the promotion of English, therefore, mastering English was not considered as simply the acquisition of a second language, but the "status of having acquired the master language 13 itself" (Mazrui, 1975a, p.90). Consequently, the role of Kiswahili was trivialized by the colonial language policies. However, the policies in independent Kenya have not given Kiswahili equal recognition to English. Furthermore, controversy on the role of Kiswahili and English in national development has been evident since independence. Further, the concern as to which language should be used in government, business and education has been crucial in the overall national development (Prator and Whiteley, 1967). For example, in the education system, the issue has concerned a choice regarding the medium of instruction — English, Kiswahili, or first language. (Gorman, 1968, 1974a; Rhoades, 1977). The language policy tag of war has been mostly between English and Kiswahili. The inconsistency and ambiguity of the language policy has led to an education system that has considerably revised the policies on the medium of instruction and language teaching especially at the lower educational levels (Gorman, 1974b; Itebete, 1976). For instance, in the case of Kiswahili teaching in schools, Gorman (1974c, p.496) observes that "some pupils Speak Swahili well . . others know very little or none, and most have a rudimentary knowledge of some features of the language." Itebete (1976, p.73) also notes that the teaching of Kiswahili has "swung backwards and forwards," but English has gained more prominence in the school curriculum. Today, English is the medium of instruction in all schools and Kiswahili is taught as a school subject (Republic of Kenya, 14 1972). Kiswahili gained its official status in 1974 when it became the official language of parliament (Scotton, 1976a, 1976b; Harries, 1983), but nevertheless it has not been given equal status to English (Mazrui, 1975a; Heine, 1977). Further, although there are no regulations on language use by government officers, English and Kiswahili are the two main languages that are mostly used at work. It is in this complex linguistic environment that the middle level manager in the Kenyan public sector works. The complexity of the language situation is further complicated by the fact that the low level workers have varying competence in English and Kiswahili, the two main languages that a manager may be equipped with. A middle level manager's linguistic repertoire may be comprised of at least three distinct linguistic varieties, namely: English, Kiswahili and his/her first language. Therefore, it is difficult for a manager to always choose a language which will be understood when interacting with the superiors, peers and subordinates. When making the choice, it is vital to consider which language is appropriate, where, when, with whom and for what purpose (Hymes, 1967; Fishman, 1972; Halliday, 1972; Saville-Troike, 1982). The choice of language the middle level manager makes, reflects on him/her as an individual officer and on the entire public organization. CHAPTER 2 I THEORETICAL PERSPECTIVES In this section, an overview of language choice will be briefly presented, and then the theoretical models that may be applied to understand why interlocutors choose one language rather than the other(s) in a particular interaction will be discussed. Ag Overview 9: Language Choice Language choice in a multilingual society can be viewed in two perspectives. First, at the national level, regarding language policy, language choice can be studied by asking the following questions: which language(s) shall be used as the official language(s) of government, business and education?; which language(s) shall be used as the national language(s)?; which language(s) shall be used as the medium of instruction in all educational levels?; and which language(s) shall be taught as a school subject? (Ladefoged et al., 1972). Answering these questions is a conscious attempt to choose the most appropriate language(s) to be used in various aspects of national development. The second perspective is approaching language choice 15 16 at the individual level. The central question is: which language does speaker x use when interacting with addressee y in a given interaction? It is this latter perspective of language choice that will be focused in this study. Language choice is largely unconscious (Gumperz, 1972; Scotton, 1979; Saville—Troike, 1982) and is a consistent process (Fishman, 1972). However, speakers may know which language to use when interacting with particular interlocutors. The choice of language a speaker makes may partly depend on his/her linguistic repertoire and also the linguistic repertoire of the other discourse participants (Parkin, 1974d). Also, it may depend on the image one wants to project of herself/himself. When making the choice, the speaker may find himself/herself in an "overlapping situation" (Herman, 1968, p. 493). According to Herman, a speaker may choose to speak a particular language to satisfy his/her personal needs, expectations and desires. On the other hand, he/she may be compelled to choose the language designated by the social group norms. Thus, the speaker may be in conflict between choosing the language that he/she considers as the most appropriate and the language the societal norms designate as the most appropriate for a given interaction. Theoretical Models Three models, namely: Domain model, Accommodation model and Markedness model will be discussed. 17 Domain Model For over two decades, some linguists have explained language use behavior in multilingual societies in relationship to the domain (situation) that a specific language is used (Hunt, 1966; Mackey, 1965, 1966; Fishman, 1972; Bell, 1976; Whiteley, 1974b; Halliday, 1972; Wallwork, 1978). The domain scholars relate language use behavior to institutions such as the family, school, church or business organization in one society or comparatively between societies. Domains according to Fishman (1972, p. 441) "attempt to summate the major clusters of interaction that occur in clusters of multilingual settings and involving clusters of interlocutors." The domains may be categorized into two levels: societal-institutional (e.g. family, church, business organization etc.) and socio-psychological (e.g. intimate, informal, formal etc.) (Saville-Troike, 1982). However, the two levels may be closely related. For example, the formal level may be closely related to the Business organization domain and the intimate level to the family domain. For example, the language used in the business organization may be totally different from that used in the family. Mackey (1966, p. 77) Observes that: the language used may depend less on the person than on the place . . . many bilinguals (multilinguals) speak one language with their fellow bilinguals (multilinguals) when they have them in their homes, and will address them in the other language at their place of work. 18 According to the domain scholars, the topic of discourse, the purpose, locale and role relationships are the most important elements that determine language choice in a particular domain. Language choice, Gorman (1974b, p. 369) states: . . is affected by conventions associated with the use of particular languages in particular settings, by the role-relations between the participants, by the topic which is being discussed and by the manifest and latent function of the conversation. First, the topic of discourse regulates language use (Fishman, 1972) because certain topics are discussed better or more appropriately in one language rather than the other(s) in a particular multilingual environment. For example, English may be considered a more appropriate language for discussing technical topics than Kiswahili or first languages. Second, the purpose of an interaction may influence language use behavior (Mackey, 1966, Saville- Troike, 1982) because each language has specific functions in a given multilingual community (Penalosa, 1981). According to Penalosa, one language can be functionally differentiated from the other(s). For example, English may be used to give directives as a mark of authority and power, whilst first languages may be used to express feelings and emotions. Third, the locale (setting) of an interaction may influence language use behavior (Fishman, 1972). For instance, the language a middle level manager uses in his/her boss's office, may be different from the language 19 he/she may use when interacting with the boss in a social club. Finally, the role relationships of the participants in a particular domain may have an impact on language use (Fishman, 1972; Bell 1976; Wallwork, 1978). Dyadic relationships such as father-son (Family domain), teacher—student (school domain), and boss— subordinate (business organization domain) may determine which language any of the participants may choose when interacting with the other. For example, a subordinate may only choose the neutral language (English or Kiswahili) when speaking to the boss, even though they might be speakers of the same first language. However, the linguistic preference of the most powerful person in the dyad may be used as the language of interaction. The domain model tends to assume that language choice in a particular domain is clear-cut and unequivocal. The model postulates that a language used in one domain may be exclusively not appropriate in the other. For example, Halliday (1972, p. 9) States that, "people who speak more than one language . . . use each one in certain types of situations only; they do not use all interchangeably, in all situations." This however is too simplistic because as Wallwork (1978, p. 57) notes, there may be "contact with other people with whom there is a potential choice of two or even three languages, and here the choice may be made depending on either the role of the two speakers vis-a-vis each other." Furthermore, in the Kenyan situation, the three 20 distinct languages (English, Kiswahili and first languages) are often used in the same situations (Rhoades, 1977). For example, if we consider a work place as a single domain, we would expect only one language to be used in the Kenyan public sector. However, this is not the case because the three distinct languages (English, Kiswahili and first languages) are all used. Therefore, the domain model does not seem to explain complex linguistic behavior exhaustively. As Scotton, (1972, p. 106) notes: While we accept the possibility of this type of prediction . . . prediction in terms of general trends . . . from the domain or social situation, we see domain or classes of situations as a very weak predictor of linguistic behavior in specific situations. Accommodation Model The accommodation model (Giles and Smith, 1979; Giles, Taylor and Bourhis, 1973; Taylor and Royer, 1979) is deep rooted in four social psychological theories, namely: similarity-attraction, social exchange, causal attribution and intergroup dynamics theories. The model assumes that a speaker may increase the perceived similarities with the addressee to become more attractive or to gain social approval. Conversely, perceived dissimilarities may be increased to signal social disapproval. By increasing the perceived similarities, the social distance between the speaker and the addressee is reduced, but, social distance is increased if the perceived dissimilarities increase. 21 Thus, as Taylor and Royer (1979, p. 185) state, "individuals subtly and indirectly communicate approval or disapproval of one another by altering their speech so as to be more similar to or different from the other." A speaker's speech behavior converges or diverges from that of the addressee (Taylor and Royer, 1979; Giles and Smith, 1979). Convergence takes place when a speaker shifts his/her speech toward that of the addressee, as an indicator of social approval. For instance, choosing a language that the other interactant can speak fluently is a form of convergence. If speaker A can fluently speak languages x and y, but the addressee B can only speak language x fluently, then for A to converge with B, he/she must speak language x. On the other hand, divergence occurs when a speaker shifts his/her speech away from that of the other interactant to signal social disapproval. For instance, choosing a language that all participants in an interaction do not understand and speak is a divergence strategy. If in a multi-ethnic situation all the .interlocutors do not share a first language, then if any of the participants who share this first language choose to speak it, then that is a divergence strategy employed to exclude the non-first language speakers. Intergroup processes have a significant influence on convergence and divergence strategies (Giles and Smith, 1979; Taylor and Royer, 1979). Group members tend to use a convergence strategy when interacting with each other, but a 22 divergence strategy when interacting with the non—group members. Thus, we would expect the middle level managers in the Kenyan public sector to choose English alone, or a combination of English and Kiswahili when interacting with ’ their peers and superiors (convergence strategy), but choose Kiswahili alone when interacting with their subordinates (divergence strategy). Therefore, as Giles and Smith (1979, p. 52) state, "When group membership is a salient issue, speech divergence may be an important strategy for making oneself psychologically and favorably distinct from outgroup members." Further, choosing a convergence strategy is an attempt by the speaker to maximize the rewards of being a member of the in—group. For example, when the middle level managers speak English to their peers, they signal their high educational level and socio-economic status which is symbolized by the English language in a linguistic environment where many people, especially the subordinates do not speak it. In addition, the degree of accommodation may depend on the speaker's personality and his/her role relationship to the addressee, rather than primarily the addressee's social identities such as socio-economic status and educational level (Platt, 1979). From Platt's observation, we would argue that the middle level managers' role relationship to their superiors, peers and subordinates would largely influence whether the managers will use a convergence or a divergence strategy. While the middle level managers might 23 choose to narrow the social distance by using a convergence strategy when interacting with their superiors and peers, a divergence strategy may be employed when interacting with the subordinates. Markedness Model The markedness model of Scotton (1972, 1979, 1982a, 1983, 1985, forthcoming) is deeply rooted in three social psychological theories, namely: social exchange, social norms and role-taking. First, the social exchange theory (Thibault and Kelley, 1959) assumes that participants in an interaction weigh the rewards and costs of the exchange. Thibault and Kelley (p. 13) state: The consequences or outcomes for an individual participant of any interaction or series of interactions can be stated, then, in terms of the rewards received and the costs incurred, these values depending upon the behavioral items which the two persons produce in the course of the interaction. For instance, for the middle level managers, rewards may be derived from a task that is successfully accomplished through effective interaction, while the cost incurred may depend upon the amount of communicative energy needed to get the task accomplished. Further, the relative reward for an individual participant depends on what he/she perceives as the best reward in relationship to the cost (Secord & Backman, 1964). Second, social norms have an impact on the behavior and 24 attitudes of members of a group (Sherif, 1936, 1948; Sherif and Sherif, 1969). Sherif and Sherif (p. 41) define a social norm as "an evaluative scale designating an acceptable latitude and an objectionable latitude for behavior, activity, events, beliefs or any other object of concern to members of a social unit." Thus, norms are formed as members of a group interact, and clearly specify the expected behavior of each member. Therefore, we would argue that in a multilingual organization, there are norms that specify which language to use when interacting with members of different organizational levels. Third, role relationship is an important phenomenon in a group because every group member is expected to behave or act according to his/her role (Secord & Backman, 1964). Slawski (1981, p. 43) defines a role as a "pattern of behavior associated with a position in a society, group or relationship." There are role expectations which are held by other group members to designate the expected behavior that a person holding a particular role should exhibit (Shaw, 1970). Thus, language can be used as an indicator of the role relationships of the participants in a conversation exchange (Nichols, 1984). In the Kenyan public sector therefore, the middle level manager is expected to be competent in both English and Kiswahili by virtue of his/her role in the organization. The markedness model assumes that participants in a conversation exchange interpret code choices in terms of a 25 natural theory of markedness, and that as part of their communicative competence, they recognize all choices as either unmarked or marked in reference to the social norms of their speech communities (Scotton, 1979, 1982a, 1983, 1985, forthcoming). Scotton (1979, p. 360) defines the unmarked choice as "that choice which the norms of society indicate represents the most expected choice for a particular status holder in a particular role relationship in a particular situation." Conversely, the marked choice can be defined as that choice which is unexpected and its use conveys certain social meaning. For example, if a subordinate is not fully competent in English, then Kiswahili may be the unmarked choice for the middle level manager to use when interacting with him/her, but English is a marked choice in this interaction because its use may be an attempt by the middle level manager to indicate his/her high socio—economic status and educational level. Furthermore, participants in a given interaction know the unmarked and the marked choices that may be appropriate for an interaction (Scotton, 1985; forthcoming). For instance, speaking the first language to the boss may not be expected unless he/she initiates it. Further, according to the model, participants in a conversational exchange negotiate a set of rights and obligations in their relationships (Scotton, 1982a, 1982b, 1985, forthcoming). Different code choices mark different needs and expectations of the participants in an 26 interaction. Thus, linguistic choices, Scotton (forthcoming) states, "are indexical of a set of rights and obligations holding between participants in the conversational exchange." The social norms in a particular community designate the consequences of choosing a particular language rather than the other(s). Scotton, (forthcoming) notes that, "speakers are free to make any choices but how their choices are interpreted is not free." This means that a speaker has to weigh the rewards and costs of making a particular choice. For example a middle level manager who chooses to speak English or Kiswahili (unmarked choices) when speaking to the superior may be positively perceived (reward) as an officer who wants to maintain their status differences. However, if the middle level manager chooses to speak a first language (marked choice) to the superior, if they happen to share it, then the superior may interpret that as an attempt to narrow their social distance, which he/she may resent (cost). The choice of language that a speaker makes has an impact on the other interlocutors. Therefore, as Saville— Troike (1982, p. 91) states: the roles which individual speakers assume and the status they are accorded is generally dependent on their relationship to other participants in the communicative event. The participants' relative status in the interaction is signalled by the choice of language. According to the 27 markedness model, a speaker's choice of the unmarked choice calls for the maintenance of the relationship, but the marked choice calls for a change of the role relationship (Scotton, 1979, 1982a, 1985, forthcoming). For instance, if a middle level manager speaks English to the superior, then the choice may be intended to maintain their boss- subordinate relationship. But if the middle level manager speaks the first language to the superior, if they happen to share it, then it may be interpreted as an attempt to change their role relationship (boss—subordinate) by interacting as members of the same ethnic group. The degree of markedness depends on what is the most salient feature in the interaction. In the first case (above), when the middle level manager speaks English to the superior, occupational status, socio-economic status or educational level may be the most salient features. While in the second case, ethnic solidarity, affiliation and loyalty may be the most salient features when the middle level manager speaks the first language to the superior. Thus, we cannot label a particular language as either unmarked or marked because each choice a speaker makes depends on the conversational exchange and more specifically what the participants in the role relationship perceive as the most salient feature in their interaction. The markedness model sees speakers as having a range of options which they can freely choose within a normative framework that designates the consequences of each choice in 28 an interaction. Further, it views interaction as a process in which both the speaker and the addressee are active participants. Thus, language usage and choice is directly tied to the role relationships of the discourse” participants. The role relationships are marked by social identities such as age, educational level, occupational status, language competence and ethnicity. This study will endeavor to find out the relationship between these social identities and language choice in a multilingual public organization. II REVIEW OF THE LITERATURE During the last two decades, linguists and educators have studied the problems of multilingualism in Kenya. However, relatively few studies have hitherto looked at language use in Kenya in an organizational perspective. The few studies that have focused on language use at work situations have been relatively brief and descriptive. In this section, the literature on the language use in Kenya will be reviewed in the light of the variables under study (educational level, language competence, occupational status, directionality of interaction, age, ethnicity and attitude). Educational Level The Kenyan educational language policy has been a 29 controversial issue. During the British colonial era, the educational language policies were inconsistent (Gorman, 1968, 1974a). Mutahi (1983), for example, points out that Kiswahili was used as the medium of instruction in the African schools in the 1930's, but in the 1950's, English became the medium of instruction. In independent Kenya, the educational language policies have also been inconsistent. Although, English is the medium of instruction in all educational levels, the status of Kiswahili and first languages in the school curriculum has been ambiguous. The inconsistencies of the language policies has led to the graduation of students with varying competence in both English and kiswahili depending on the language policy that was prevalent during their school life. Bujra (1974) for example, found that 37.3 percent of the respondents had been taught partly in English, 3.9 percent wholly in English, 25.5 percent wholly in Kiswahili, 35.5 percent partly in Kiswahili, 23.6 percent in the first language and 27.5 percent in both first language and either English or Kiswahili as the medium of instruction. Bujra's findings indicate the inconsistencies of the language policies over time. Although Kiswahili has been used as a medium of instruction or taught as a school subject at one time or another, Whiteley (1969, p.66) says, "no where was it (Kiswahili) surrounded by an aura of prestige comparable to that of English." Therefore, many educated people may have 3O attended schools that did not teach Kiswahili, and as a result, they have a "shaky command of it" (Itebete, 1976, p. 75). Thus, Kiswahili lacks equal recognition to English in the education system, and this may negatively affect a speaker's competence and attitudes towards it. A rise in education means an increase in the use of English, but a decrease in Kiswahili usage (Scotton, 1982b). Kiswahili is the main inter-ethnic work language used by the less educated workers, while the highly educated use English or combine it with Kiswahili (Scotton, 1976a). A study by Parkin (1974a) found that workers who had no formal education had higher claimed frequency in Kiswahili usage than the primary and secondary school graduates. According to Scotton (1982b), the educated people have the Option of choosing one of the two neutral languages, English or Kiswahili. Further, Parkin (1974a) also found that the more years of instruction an officer had in Kiswahili, the less the officer used it. Parkin argues that the highly educated workers who have formal knowledge in Kiswahili work in organizational levels where English is mostly required and the use of Kiswahili suppressed. Language Competence Kenyans have varying language competence in both English and Kiswahili. A majority of the Kenyan population speak Kiswahili as a second language. However, only a small percentage of the population speak English . Parkin (1974a) found that all the respondents (N=349) claimed to know 31 Kiswahili, 42 percent could speak English, but only 22 percent used English at work. Bujra (1974) also found that 47.7 percent of the respondents could not speak English, 27.9 percent could speak only a few English words, 18.6 percent could speak moderately, and only 5.6 percent could speak it fluently. These findings show that the lower organizational workers who were the respondents in both studies can effectively speak in Kiswahili, but not English. Although many middle level managers may not have difficulties when speaking English as do the subordinates, some may lack competence in Kiswahili. Whiteley (1974, p. 344) quotes a Senior Government Officer who said: No one minds speaking English badly because it is a mark of hard work, initiative etc., to have tried to learn it at all. A man who corrects one's Swahili may be no better than one self, this cannot be true of someone who corrects one's English . . . those of us who speak it (Swahili) badly are ashamed or shy of speaking it at all The above view is consonant with Heine's (1977) argument that competence in a vertical code (English) is used as a socio—economic indicator because the speaker's ability to speak the vertical code is positively evaluated by the other interactants. Conversely, the use of a horizontal code (Kiswahili) is neutrally evaluated. Therefore, a speaker who conforms to the norms of the vertical code is regarded as competent and socially successful. The difficulties some middle level managers face when interacting in Kiswahili may be exemplified with the 32 problems the members of parliament encountered when the Kenyan National Assembly adopted Kiswahili as an official language of parliament in 1974. The speaker told the members: I have explained him (the President of the Republic of Kenya) the difficulties we have in switching over to Swahili, . but we shall in due course start sorting out our difficulties (Republic of Kenya, 1974, p. 21). As Kiswahili is an official and national language of Kenya, every government officer should understand and speak it effectively in formal interactions (Gichangi, 1984). According to Gichangi, failure to speak Kiswahili effectively may retard national development because the officer may not interpret and communicate government policies and regulations to the fellow officers and members of the public. This view is supported by Ohly (1974) when he argues that an officer who has no command in the language spoken by the managers and other fellow workers is restricted in participating in national development. The following comment of Le Page (1964, p. 18) serves as a summary of Gichangi and Ohly's view: Wherever the language of the government differs from that of the mass of people . . . . Linguistic diversity therefore acts as a brake on economic progress. In Kenya therefore, competence in Kiswahili is important if government officers are to be effective communicators of government policies and regulations. Thus, as Mwangi (1981), a senior government officer strongly 33 notes, "time has come that no government officer can execute his/her duties effectively, without adequate knowledge and ability to speak Kiswahili fluently." Occupational Status Kiswahili has been often viewed as a language that is mostly spoken by the low status workers, while the high status officers speak it only to signal authority, power and formality (Heine, 1940; Whiteley, 1974b; Scotton, 1977). The high status officers use English mostly because it symbolizes high position, authority, power and high socio— economic prestige (Hunt, 1966; Scotton, 1976a; Parkin, 1977; Mkilifi, 1978;- Scotton & Ury, 1977; Harries, 1983) and is the language of vertical mobility (Heine, 1977) spoken only by the few people who have had formal education. Acquisition and mastery of English is a passport to white-collar jobs and hence a source of high socio—economic status (Parkin, 1974a; Gorman, 1974a; Whiteley, 1979). Since the British colonial era, English, Itebete (1976, p. 76) notes, "has been of immediate profit," as the vehicle for socio-economic mobility. People therefore acquire English for utilitarian reasons such as job opportunities and vertical mobility in an organization. Quite often some job vacancy advertisements specify acquisition and mastery of English as a requirement. For example, an advertisement by the Directorate of Personnel Management for Accounts Assistant Trainee positions stated: Applicants must be in possession of East African Certificate of Education with 34 credit passes in five subjects including English language . . . (Daily Nation, 1979 September, 20). Conversely, such instrumental motivation lacks in the case of Kiswahili (Ngara, 1982). Thus, because of the. instrumental value attached to English, Itebete (1976, p. 74) observes that: Even the lowest placed worker in Nairobi struggles sometimes fruitlessly, to converse and convey his ideas in English and this tends to be a source of pride and satisfaction to him, because in so doing he is perhaps trying to demonstrate that he is capable of getting a promotion because he "knows" English. A study by Parkin (1974a) found that the white collar workers (clerical, professional, highly technical and supervisory) had a lower claimed frequency in Kiswahili usage as a first language at work than the unskilled workers. Parkin's findings show that the higher the status of an officer, the less the officer speaks Kiswahili as a first language at work. Therefore, to distinguish themselves from the unskilled workers who are relatively uneducated, the white collar workers speak English to maintain their elite status. However, like English, Kiswahili may also be used as a symbol of authority and status when a person in a senior position speaks to the subordinates. Directionality 9: Interaction The language the middle level manager uses may be an indicator of his/her role relationships with the superiors, peers and subordinates. When peers interact, Parkin (1974b) found that 71.6 percent used Kiswahili, 25.4 percent used 35 English, and only 0.8 percent used their first language. These findings show that Kiswahili is mostly used in horizontal interaction than English or first languages. Scotton (1976a) supports these findings when she points out that speaking English alone among the peers is stigmatised because English increases the interactants' social distance. According to Scotton, most speakers tend to combine English with Kiswahili when interacting with their peers. For upward interaction, a recent study by Scotton (1982c) found that 61 percent of the respondents interacted with their superiors in English, and only 13 percent used Kiswahili. These findings are congruent with Whiteley's (1979) observation that English is used mostly with government officers who are senior to oneself, and Kiswahili with the junior officers. Thus, English is mostly used for vertical interaction by middle level managers, and Kiswahili for downward and horizontal interaction. _A_ga Speakers of different ages tend to differ in their language use behavior. Heine (1970, 1980) found that the younger people in the age group between 20 and 39 used English more than the older people in the age group 40 years and over. Whiteley (1979) also found that 77 percent of the respondents under 30 years old claimed competence in their first language, Kiswahili and English, while only 4 percent of those over 30 years old claimed to be fluent in all three. Further, Bujra (1974) also found that 18.6 percent 36 and 5.8 percent of the respondents who were moderate and fluent in English respectively were all under 30 years old and non were over 50 years old. These studies show that Kiswahili is spoken mostly by the older people and English by the younger generation. Bujra (1974, p. 244) notes that the widespread use of English by the young people is the "mark of the elite in training." As the young people tend to be more educated than the old people, they speak English mostly to conform with what is expected of people in high socio-economic positions (Parkin, 1974a). Ethnicity Although there are many symbols of ethnicity such as dress, artifacts, dance and music, Fishman (1977, p. 26) notes that language is one of the most important symbol in the transmission of ethnic messages, and has "a prime ethnic value in and of itself". First languages are largely ethnic- bound (Mazrui, 1975b), hence they are used in intra—group interaction as symbols of ethnocultural identification, ethnic solidarity, affiliation and loyalty (Parkin, 1974c; Mkilifi, 1978; Whiteley, 1979) and differentiate members of one ethnic group from the other (Bujra, 1974). Whereas first languages create ethnic cohesiveness and solidarity, cleavages may be created and nurtured between members of different ethnic groups (Parkin, 1974c). According to Scotton (1983), speaking a first language when all the participants in a conversation exchange do not share the same first language is an attempt to exclude the non- 37 first language speakers by withholding information from them. The participants excluded by the first language speakers negatively interpret and perceive its use in a multi-ethnic situation (Scotton, 1976b; Parkin, 1977), and therefore, its use is highly stigmatised. Speaking first languages in a multi—ethnic work environment is associated with ethnic parochialism which is considered a social malady, especially in the public sector, an institution that is supposed to forge national unity. In Kenya, first languages are not formally sanctioned as work languages, but they are widely used when officers of the same ethnic group interact (Rhoades, 1977; Whiteley, 1979; Heine, 1980). Whiteley notes that speaking first languages in a work situation reduces the level of formality. Further, according to Scotton (1970b), the need for a neutral language diminishes when members of the same ethnic group interact. However, speaking a neutral language like Kiswahili may unify the interactants and minimize ethnic differences (Soba, 1984), and no members of any one ethnic group may be alienated from the other during the discourse (Karigithe, 1982). Attitude Debates on the role of Kiswahili and English in national development have involved academicians, politicians, civil servants and members of the public. According to Whiteley (1979), there are three schools of thoughts based on people's views on Kiswahili and English. 38 First, supporters of English hold that English is a symbol of socio—economic prestige and status, while Kiswahili symbolises lower socio—economic achievement and status. Although politicians have emphasized the role Kiswahili plays in national development and integration, Ohly (1982) notes that this has not minimized the attitudinal problems imposed on Kiswahili by various socio—economic groups. According to Ohly, some people in high socio—economic status groups have a negative attitude towards Kiswahili, and strongly prefer English. However, a study by Janice (1971) shows that highly educated people prefer Kiswahili as the national language. Janice found that 71 percent of Kenyan students in the United States of America preferred Kiswahili as the national language, and only 20 percent preferred English. Further, some advocates of English consider Kiswahili a foreign language like English (Khalid, 1977) and also inferior to English (Knappert, 1965). Consequently, some people in high socio—economic status groups are reluctant to speak Kiswahili because it is "inferior" to English which is a symbol of their status. For example, Bujra (1974) observes that some government officers are consciously reluctant to speak Kiswahili because they want to distinguish themselves from the majority of the uneducated people whom they administer and serve. Second, supporters of Kiswahili, strongly maintain that Kiswahili is a vital tool in national development (Khalid, 39' 1977; Karigithe, 1982) and is a symbol of national identity (Rhoades, 1977). They argue that English is nationally limited because the majority of the population cannot speak it (Mazrui, 1969). Further, recognizing and giving Kiswahili the status it deserves does not trivialize the status and vitality of English because Kiswahili plays a cardinal role that does not conflict with that of English (Karigithe, 1981). The concern that any language policy changes in favor of Kiswahili will trivialize English has not been founded. As Wallwork (1978, p. 56) observes, ”the English educated elite, among others, have not been too anxious for this changes to be brought about with haste." Regarding the English supporters contention that Kiswahili is an inferior language, the kiswahili advocates point out that Kiswahili like all other languages can develop to be a useful tool in all human endeavor, and therefore should not be considered too poor or inferior to be an effective medium of communication (Ekwelie, 1971; Halliday, 1972). Thus, Ekwelie (1971, p.90) strongly states: the argument that Kiswahili is not capable of absorbing the complexities of the age appears time—bound and inorganic. It is time—bound because its proponents judge it on the basis of available facilities at a given point in time, and inorganic because they appear to discount the possibility of growth. The views of the Kiswahili supporters have been strongly articulated for the last two decades. For instance, Bujra (1974, p. 249) quotes a then Acting Secretary General of the ruling party KANU who said: 4O Swahili should be spoken by all people at all times both officially and socially. Any Kenyan professing a knowledge of English . . . and denying knowledge of Kiswahili should be known as a quisling. The strong views in support of Kiswahili may be well ' summarized by the late Shaaban Robert's Kiswahili poem Titi la mama l; tamu, hata likiwa la mbwa (One's mother's breast is sweetest, canine even it be) which shows his love and pride of the Kiswahili language. Titi la mama litamu, hata likiwa la mbwa, Kiswahili naazimu, sifayo iliyofumbwa, Kwa wasiokufahamu, niimbe ilivyo kubwa Toka kama mlizamu, funika palipozibwa, Titile mama litamu, jingine halishi hamu. (One's mother's breast is the sweetest, canine it may be, And thou, Swahili, my mother—tongue, art still the dearest to me. My song springs forth from a welling heart, I offer this my plea. That those who have not known thee, may join in homage to thee. One's mother's breast is the sweetest, no other so satisfies.) (Jahadhmy, 1975, p.3). The third school of thought maintains that both English and Kiswahili have vital roles to play in Kenya's national development and should not be seen as conflicting with each other (Karigithe, 1981). Therefore, both should co-exist in a stable relationship in which one language strengthens the other rather than weakens it (Halliday, 1972). From the above views, we would argue that a speaker's attitude towards Kiswahili or English will largely be determined by the stance one takes, which will consequently affect the speaker's language choice and use. CHAPTER 3 METHODOLOGY Survey research techniques were used in the execution of this study. In the section below, the setting of the study, and the procedures used to collect the data and obtain the results, and the hypotheses will be discussed. The Setting 9: the Study This study was conducted at the Kenya Institute of Administration (hereinafter KIA), in the Republic of Kenya. KIA is the only government institution in the country that trains middle level managers in the public sector. The institute organizes a wide variety of management and administrative seminars, conferences, workshops and courses designed to meet the practical needs of various target groups in the public sector. The institute organizes the training programmes under the auspices of the Directorate of Personnel Management in the Office of the President. The duration of the training programmes range from approximately two weeks to two years. Although the institute runs pre-service courses to prepare candidates for entry into government service, its main focus is in-service training designed to improve the managerial and 41 42 administrative skills of officers already working in the public sector. However, the pre-service trainees work for a short period of time (about one year) prior to the training. Further, due to the variety of the training programmes at one given time, the course participants represent a wide variety of occupations in the public sector. Therefore, in this study, the public officers on training at KIA were considered to have characteristics of language use behavior that reflect those of the entire public sector. Selection pf the Respondents The survey population was composed of 508 officers attending 18 training programmes (Appendix I) at the time of the Study (June—July, 1985). Of the 18 training programmes, 3 were pre—service programmes and 15 were in-service programmes. The respondents were selected by use of systematic sampling technique with a random start. First, a nominal role for all the course participants in the institute was constructed based on the duration of the courses as the criterion. The course participants were listed from those attending the longest courses (two years) to those attending the shortest course (one month). Second, the first respondent was randomly selected between numbers 1 and 10 on the list. From the randomly selected participant, every fourth (4th) course participant was selected. A total of 127 course participants were selected in the sample. The 43 procedure used in sampling the respondents was adopted because the course participants were randomly listed on the course nominal roles and also to ensure that all the courses were proportionally represented in the sample. Description _f gha Respondents All the respondents were public officers working in various Government Ministries and Departments. Of the total, 30.5 percent (N=36) worked in the Ministry Headquarters, 5.9 percent (N=7) in the Provincial offices, 28.0 percent (N=33) in the District Offices, and 35.3 percent (N=42) in other government departments, such as training institutions, Exchequer and Audit, Airport and Aerodromes. The respondents represented seven occupations as functionary categorized in this study. Of the total, 2.5 percent (N=3) police officers, 38.7 percent (N=46) administrative officers, 22.7 percent (N=27) financial officers, 15.1 percent (N=18) social workers, 5.0 percent (N=6) agricultural officers, 2.5 percent (N=3) health officers, and 10.9 percent (N=13) secretarial officers. The respondents' educational level varied considerably. Specifically, 2.5 percent (N=3) had primary school education, 32.2 percent (N=38) had secondary school education (O'Level), 19.5 percent (N=23) had higher secondary school education (A'Level), and 45.8 percent (N=54) had college or university education. The respondents working duration in the Public Service ranged from 28 years 44 to 1 year, with a mean of 8.3 years. Further, the sample was composed of 70.6 percent (N=84) males and 29.4 percent (N=35) females. The respondents ages ranged from 28 years to 49 years, with a mean age of 30.8 years. All the respondents speak English and Kiswahili, however they were speakers of 15 first languages (mother- tongues) of African origin, and one first language of Indian origin (Appendix II). Instrument A self-administered questionnaire was used in gathering the data (Appendix III). All the questions on the 30-item questionnaire were closed-ended, and were of a structured nature, designed to be answered in a short period of time. According to the format of the questionnaire, items 1,3,4,5 and 7 were to elicit demographic information such as place of work, job description, first language, the duration the officer has worked in the public sector, and the sex of the respondents. Items 8 through 10 were to provide information on the frequency of speaking English, Kiswahili and first language at work. To elicit the information needed for the specific variables under study, item 2 was to provide information on occupational status; item 6, the age of the respondents; items 11 and 12, the educational level and the highest level of formal instruction in Kiswahili respectively; items 13 through 16, the directionality of interaction; items 17 and 45 18, ethnicity; items 19 and 20, language competence in English and Kiswahili respectively; items 21 through 24, attitude towards Kiswahili, and items 25 through 28, attitude towards English. On the attitude related questions, the respondents were supposed to show their responses to each statement on a five—point Likert scale ranging from "Strongly Agree" through "Don't Know" to "Strongly Disagree." The last two items on the questionnaire, items 29 and 30 were language policy related questions designed to elicit information as to which language is the most effective in communicating government policies, rules and regulations to the subordinates and members of the public respectively. All the respondents were supposed to answer all the questions except item 18 which was a contingency question. The questionnaire was pretested on a similar group of five (5) Kenyan public officers on study leave at Michigan State University, and revised for clarity before sending it to the specified respondents. The questionnaires were administered to the respondents by lecturers at the KIA, under the auspices of the Department of Communication at the institute. This ensured a high response rate because of the 127 sample, 93.7 percent (N=119) responded. The 6.3 percent (N=8) who did not respond were not at the institute at the time of the study. All the respondents provided usable responses, hence none of the questionnaires were discarded. 46 Data Analysis 2 Chi—Square (X ) tests were computed to find out whether there is a relationship between educational level, language competence, occupational status, age, ethnicity, attitudes and the language the middle level managers use when interacting with their superiors, peers and subordinates. This non-parametric test was used because language competence, ethnicity and directionality of interaction were categorical variables measured on the nominal scale; while age, occupational status, educational level and attitudes were measured on the ordinal scale. Therefore, Chi-square tests were used because the variables were either measured on nominal or ordinal scales (Williams, 1979). Further, an alpha level of .05 was chosen for testing the significance levels for all hypotheses. Thus, using Chi—square (X2) statistical tests, the following hypotheses were tested: H1: Middle level managers with high competence in Kiswahili will interact frequently in Kiswahili with their superiors, peers and subordinates. H : There are differences between the younger and the older 2 middle level managers in their language use behavior. H : There is a relationship between educational level and 3 the language the middle level managers use when interacting with their superiors, peers and subordinates. 47 There is a relationship between occupational status and the language the middle level managers use when interacting with their superiors, peers and subordinates. There is a relationship between attitudes towards Kiswahili and the language the middle level managers use when interacting with their superiors, peers and subordinates. There is a relationship between attitudes towards English and the language the middle level managers use when interacting with their superiors, peers and subordinates. CHAPTER 4 RESULTS The findings of the study and the interpretations of each hypothesis will be presented in this chapter. As stated in the preceding chapter, Chi—square (X2) statistical tests were computed to test all the hypotheses. The findings of each hypothesis will be individually analyzed, and all pertinent data briefly discussed. An alpha level of .05 was used to test the significance levels for each hypothesis. Hypothesis 1: Middle level managers with high competence in Kiswahili will interact frequently in Kiswahili with their superiors, peers and subordinates. There is no significant relationship between competence in Kiswahili and the frequency of Kiswahili usage by middle 2 level managers when interacting with their superiors (X = 2 9.14887; df = 9; P > .05), peers (X = 6.6858; df = 6; P > 2 .05), and subordinates (X = 11.07913; df = 9; p > .05). Competence in Kiswahili does not seem to have a significant relationship with the language the middle level managers choose to use. These results are depicted on table 1. As table 1 shows, the respondents who claimed to be "very good," "good," and "adequate" in Kiswahili tend to use English alone or they combine it with Kiswahili when 48 4S3 Table 1. Frequency of claimed competence in Kiswahili by language choice. I I I I I I I I Kiswahili I Directionality I English I Kiswahili I English/ I First I Kiswahili/I Row I Claimed I of I Only I Only I Kiswahili I Language I First I Total I Competence I Interaction I I I I Only I Language I I ...... I _____I _-I ______I-____-_--__I_______-__I_______-___I_-___-__I I I I I I I I I I superiors I II I 0 I 10 I 0 I - I 21 I Very | I I I ----------- I ---------- I ----------- I -------- I I peers I I2 I 0 I 9 I - I - I 21 | Good I I I ——————————— I ----------- I ---------- I ----------- I -------- I I subordinates I I I II I 9 I - I 0 I 21 I I" I- *— I I" I I I" I I superiors I 29 I 0 I I6 I l I - I 46 I I I-- I- I--- I I I -------- I Good I peers I 33 I 0 I I3 I - I - I 46 I I I— I—-— —-I ——I —————————— I ----------- I -------- I I subordinates I 4 I 19 I 22 I - I l I 46 I ----------- l'IIII I I I superiors I 35 I I I ID I 0 I - I 46 I I ---------------- I --------- I ----------- I ----------- I I—- I -------- I Adequate I peers I 33 I 2 I II I - I - I 46 I I ---------------- I ————————— I I I I--- I I I subordinates I 4 I I6 I 23 I - I 3 I 46 I I -—-I ————————— I ——————————— I ----------- I ---------- I ----------- I -------- I I superiors I 3 I 0 I 0 I 0 I - I 3 I Not I ---------------- I-—— I- I ----- I ---------- I ----------- I -------- I I peers I 3 I 0 I 0 | ' I ' I 3 I Good I ---------------- I --------- I ~~~~~~~~~~~ I ——————————— I ---------- I ----------- I ———————— I I subordinates I 0 I 0 I 2 I - I l I 3 I ___________ I____--_______-__I_______-_I-_____---__I-_-_--__-__I_________-I__-_ I I 2 X raw score = 9.14887 with 9 df; significance = .4236 for superiors. 2 X raw score = 6.68582 with 6 df; significance = .3509 for peers. 2 X raw score =11.07913 with 9 df; significance = .2703 for subordinates. 50 speaking to their superiors and peers. However, English alone is used more frequently than a combination of English and Kiswahili. Surprisingly, except for those who claimed to be "adequate," the others claim not to use Kiswahili alone when interacting with their superiors and peers. But with the subordinates, English alone, Kiswahili alone, and a combination of both are used. However, a combination of English and Kiswahili is mostly used than Kiswahili alone. Those who claimed "not good" in Kiswahili tend to speak English alone to their superiors and peers, and a combination of English and Kiswahili to their subordinates. None of the respondents claimed to be "very poor" in Kiswahili. Also, first languages are rarely used, but the few respondents who use them, claim to use them mostly with their subordinates. Therefore, although the respondents claimed relatively high competence in Kiswahili, that does not seem to influence their frequency of using it alone, especially when interacting with their superiors and peers. Hypothesis g: There are differences between the younger and the older middle level managers in their language use behavior. There are no significant differences between the younger and the older middle level managers in their use of 2 Kiswahili (X = 1.78396; df 4; p > .05), first languages 2 2 (X = 1.69861; df = 8; p > .05), and English (X = 4.06988; df = 2; P > .05). These results are presented on table 2. As the table indicates, there is no significant relationship 51 Table 2. Frequency of Language choice by Age. I I I I I I I Frequency of I More than I Once A I Once or I Once or I Never I Row I Age I Language Use I Once A Day I Day I Twice A I Twice A I Speak I TotalI I I I I Day I Month I I I I I I I I I- I ------ I I English I 69 I O I - I ~ I - I 69 I I I I I I I I ------ I 20-30 Years I Kiswahili I 65 I 2 I 2 I - I - I 69 I I I I I I I ------- I ------ I I First I 30 I 7 I 12 I 4 | 15 I 68 I I Language I I I I I | I I I I I- --—I -I I —————— I I English I 38 I O I — I - I - I 38 I I I I I I I ------- I ------ I 31-40 Years I Kiswahili I 37 I 1 I 0 I - I - I 38 I I- I -—-I I I -—-—I ------- I ------ I I First I 11 I 6 I 6 I 2 I 7 I 38 I I Language I I I I I I I - I ---—I ———————————— I -------- I—- I I I ------ I I English I 10 I 1 I - I — I - I 11 I I -------------- I--- I I I— I I I Above 41 I Kiswahili I 11 I O I 0 I — I - I 11 I Years I -------------- I---- I --I-- I I I I I First I 5 I 1 I 1 I I I 3 I 11 I I Language I I I I I I I -I _______ I ____________ I ________ I ......... I ......... I _______ I ______ I 2 . . . _ x raw score = 4.06988 with 2 df; s1gn1f1cance — .1668 for English. 2 . . . _ X raw score = 1.78396 with 4 df; sign1f1cance — .7754 for Kiswahili. 2 . . . _ X raw score = 1.69861 with 8 df; Sign1f1cance - .9889 for first languages. 52 between age and the use of the three distinct languages (English, Kiswahili and first languages). It seems however, that all the respondents regardless of their age group claimed higher frequency of English and Kiswahili usage, but lower frequency of first language usage at work. But English was claimed to be used slightly more frequently than Kiswahili. The older managers (above 41 years old) claimed a slightly higher frequency of Kiswahili usage than the younger managers (20-30 years old). But more younger managers claimed that they never speak their first language at work than the older managers. Hypothesis a: There is a relationship between educational level and the language the middle level managers use when interacting with their superiors, peers and subordinates. There is no significant relationship between educational level and the language the middle level managers 2 use when interacting with their superiors (X = 3.99953; df 2 = 12; p > .05) and subordinates (X = 15.33073; df = 9; p > .05). However, there is a significant relationship between educational level and the language the middle level managers use when interacting with their peers (X2 = 22.71028; df = 6; p < .05). These findings are depicted on table 3. As the table shows, the language the middle level managers use when interacting with their superiors and subordinates does not vary with the educational level. When speaking to the superiors, English alone or a combination of 513 Table 3. Frequency of language choice by educational level. I I EducationalI Directionality I English I Kiswahili I English/ I I I I First I Kiswahili/I Row I I | Level I of I Only I Only I Kiswahili I Language I First I Total I Interaction I I I I Only I Language I ______ I- ___I__-_____-I___-______-I____-_____-I__-__-_-__I_-___-__-__I__--_-__I I I I I I I I I I superiors I 2 I 0 I 1 I 0 I - I 3 I Primary I I I— I— I -I ——————————— I -------- I I peers I I I I I I I ‘ I - I 3 I Schooi I I--- I I ----------- I ---------- I I I I subordinates I 0 I 3 I 0 I - I 0 I 3 I I I I I ----------- I ---------- I ----------- I -------- I Secondary I superiors I 25 | 0 I 11 I 1 I - I 37 I I I —-—I ----------- I ----------- I--- I I—- I School I peers I 22 I 0 I 15 I - I - I 37 I I -I --------- I I ~-I ---------- I ----------- I -------- I (O'Level) I subordinates I 2 I 13 I 21 I - I 1 I 37 I . I I- I ----- I ---------- I I I I Higher I superiors I 14 I 0 I 9 I O I — I 23 I Secondary I --I I- I- I ---------- I I -—I School I peers I 17 I O I 6 I - I - I 23 I I ----- —-I —I —--—I ——————————— I --I ------ I -------- I (A’Level) I subordinates I 5 I 9 I 7 I - I 2 I 23 I I I --------- I ----------- I I— - I I I I superiors I 37 I 1 I 16 I 0 I - I 54 I College/ I ---------------- I --------- I ----------- I ----------- |-- -I I -------- I I peers I 41 I 1 I 12 I - I — I 54 I UniversityI ---------------- I --------- I ----------- I- | --I--- I I I subordinates I 2 I 21 I 28 I - I 3 I 54 I ___________ I___--__-____--_-I____-____I-_-____-_-_I---___-____I________--I__--_____-_I_-___-_-I 2 X raw score = 3.99953 with 9 df; significance = .9114 for superiors. 2 X raw score =22.71028 with 6 df; significance = .0009 for peers. 2 X raw score =15.33073 with 9 df; significance = .0822 for subordinates. 54 English and Kiswahili are both almost equally used by respondents of all educational levels. While when speaking to the subordinates, respondents with primary school education use Kiswahili alone, whilst those with at least secondary school education tend to use Kiswahili alone or they combine it with English, although the combination tends to have more weight. However, the language the middle level managers use when interacting with their peers vary with education. Although English alone or in combination with Kiswahili is mostly used with the peers, it however seems that the higher the educational level of the manager, the more the manager interacts with the peers in English alone than a combination of English and Kiswahili. However, Kiswahili alone and first languages are rarely used with the peers by respondents of all educational levels. Hypothesis 5: There is a relationship between occupational status and the language the middle level managers use when interacting with their superiors, peers and subordinates. The Chi—square (X2) tests show that there is no significant relationship between occupational status and the. language the middle level managers use when interacting with their superiors (X2 = 18.08843; df = 12; P > .05), peers (X2 = 17.42350; df = 8; P > .05), and subordinates (X2 = 20.42350; df = 12; p > .05). These results are depicted on table 4. As the table shows, the language the middle level 55 managers use when interacting with their superiors, peers and subordinates does not vary with their job groups. Thus, regardless of the respondents occupational status, English alone or a combination of English and Kiswahili are mostly . used with the superiors and peers. But English alone is more dominant than a combination of English and Kiswahili. Further, when speaking to the subordinates, Kiswahili alone or a combination of English and Kiswahili is mostly used. A combination of English and Kiswahili is more dominant than Kiswahili alone. In addition, first language alone or in combination with Kiswahili is rarely used by respondents of all the occupational status groups. However, a combination of Kiswahili and first language is used more frequently than first language alone when the middle managers speak to their subordinates, but not to the superiors and peers. Hypothesis 9: There is a relationship between attitudes towards Kiswahili and the language the middle level managers use when interacting with their superiors, peers and subordinates. To test this hypothesis, each item on the questionnaire that sought to know the respondents attitudes towards Kiswahili was tested individually (see Appendix III). On item 21 ("Kiswahili is as foreign to most Kenyans as English"), no significant relationship was found with the language the middle level managers use when interacting with their superiors (X2 = 12.84031; df = 12; P > .05), peers (X2 2 = 8.67353; df = 8; p > .05), and subordinates (X 5€5 Table 4. Frequency of language choice by occupational status. ------—--—--—-~——---—--—-—————--—-———-—- ——--——-—————————-——- I I I I I I I I Occupational I Directionality I English I Kiswahili I English/ I First I Kiswahili/I Row I Level I of I Only I Only I Kiswahili I Language I First I Total I I Interaction I I I I Only I Language I I --- I -I I ........... I ........... I .......... I--- I l I I I I I I; I I Job Group I superiors I 2 I O I 3 I I - I 5 I I I I I I 'I" I I E or I peers I 1 I 1 I 3 I I - I 5 I I I I I ----------- I I I ---I Below I subordinates I O I 3 I 2 I I O I 5 I I I —I ----------- I ----------- I ---------- I—-- I I Job Group I superiors I 34 | O I 18 I O I - I 52 I I ---------------- I --------- I I I -—I I I F, G, or I peers I 37 I O I 15 I - I - I 52 I I ---------------- I --------- I-- I I I ----------- I -------- I H I subordinates I 3 I 22 I 23 I - I 4 I 52 I ------------- IIIIIIII Job Group I superiors I 33 I O I 9 I O I - I 42 I I ---------------- I --------- I ----------- I—-—- I I I I J, K, L or I peers I 31 I O I 11 I - I - I 42 I I I I- -—-I ----------- I-—- I- I I M I subordinates I 2 I 14 I 25 I - | 1 I 42 I I —-I --------- I ----------- I-—-— I- I I I I superiors I 1 I O I O I O I - I 1 I N or I ---------------- I --------- I-- I I --I I I I peers I I I 0 I 0 I ' I ' I I I Above I ---------------- I --------- I-—-- I- I — I I I I subordinates I 1 I O I O I - I O I 1 I I I I I I I --—-—-c—-- -——————-—-— --———-——--— -—-—- 57 13.12712; df = 12; p > .05). On item 22 ("A person who speaks Kiswahili fluently is generally either uneducated or is of low status"), no significant relationship was found with the language the middle level managers use when interacting with their 2 superiors (X = 6.90428; df = 9; p > .05), and subordinates 2 (X = 10.47890; df = 9; p > .05). However, highly significant relationship was found with the language the middle level managers use when interacting with their peers 2 (X = 23.86537; df = 6; p < .05). On item 23 ("kiswahili lacks the vocabulary needed in many work situations"), no significant relationship was found with the language the middle level managers use when interacting with their superiors (X2 > .05), peers (X2 = 5.71266; df subordinates (X2 = 10.10193; df = 12; p > .05). 10.27995; df = 12; p 8; p > .05), and On item 24 ("It is more difficult to learn and speak Kiswahili than English"), no significant relationship was found with the language the middle level managers use when 2 interacting with their superiors (X = 11.12475; df = 12; p 2 > .05), peers (X = 3.91771; df = 8; p > .05), and 2 subordinates (X = 9.09285; df = 12; p > .05). From the above findings, we could say that the middle level managers' attitudes towards kiswahili have no significant relationship with the language they use when interacting with their superiors, peers and subordinates. Table 5 shows the respondents' attitudes towards Kiswahili. 558 Table 5. Percentages of Attitudes towards Kiswahili by language choice. ATTITUDES TOWARDS I I LANGUAGE CHOICE KISWAHILI I I-- I I I I I I I -------- I---_- - _-- I I I I I I I I I I Items I I I IDirectionalityIEnglishIKiswahiliIEnglish/ IFirst IKiswahili/I I Positive INeutralINegativeI of IOnly IOnly IKiswahiliILanguageIFirst I I(Disagree)I I(Agree) IInteraction I I I IOnly ILanguage I _I I I I ------ I _- I- I --------- I -------- I ---------- I I I I I I I I I I I I I | I superiors I 66.7 I .9 I 31.6 I - I 9 I I I I I I I I- "I I I Item 21I 77.8 I .9 I 21.3 I peers I 69.2 I 1.7 I 29.1 I - I - I I I I I I I-- I I -------- I ---------- I I I I I subordinates I 7.7 I 39.3 I 47.9 I 5.1 I - I I I I I I I I I I I I I I I superiors I 66.4 I .9 I 31.9 I - I .9 I I I I I I " I I" I "I I Item 22I 94.8 I O I 5.2 I peers I 69.0 I 1.7 I 29.3 I - I - I I I I I | I I I" I I I I I I subordinates I 7.8 I 38.8 I 48.3 I 5.2 I - I I I I I I“ I I I I I I I I I superiors I 66.4 I .9 I 31.9 I - I .9 | I I I I I I 'I I I I Item 23I 55.2 I 12.0 I 32 8 I peers I 69.0 I 1.7 I 29.3 I - I - I I I I I -------------- I I I I-- I I I I I I subordinates I 7.8 I 38.8 I 48.3 I 5.2 I - I I-- I ------- I--- I I | I I I I I I I I superiors I 67.0 I .9 I 31.3 I - I .9 I I I I I I I I I I I Item 24I 74.7 I 7.0 I 18 2 I peers I 69.6 I 1.7 I 28.7 I - I - I I I I I —————————————— I ——————— I --I I I I I I I I subordinates I 7 8 I 38.3 I 48.7 I 5.2 I - I -------- I---------- I -----_- I ------- I _---_--------- I I I--------- I-------- I -__-----_- I Item 21: Kiswahili is as foreign to most Kenyans as English. Item 22: A person who speaks Kiswahili fluently is either uneducated or is of low status. Item 23: Kiswahili lacks the vocabulary needed in many work situations. Item 24: It is more difficult to learn and speak Kiswahili than English. 59 As the table indicates, the respondents generally show positive attitudes towards Kiswahili. Specifically, 77.8 percent of the respondents do not View Kiswahili as a foreign language like English; 94.8 percent do not see a fluent Kiswahili speaker as either uneducated or of low status; 55.2 percent do not consider Kiswahili as deficient of the vocabulary needed in many work situations; and 74.7 percent do not consider Kiswahili as being more difficult to learn and speak than English. However, although the respondents show positive attitudes towards Kiswahili, they rarely use it alone when speaking to their superiors and peers as table 5 shows. Instead, they mostly speak English alone or in combination with Kiswahili to their superiors and peers. While with their subordinates, they speak mostly Kiswahili alone or in combination with English. Hypothesis 93 There is a relationship between attitudes towards English and the language the middle level managers use when interacting with their superiors, peers and subordinates. To test this hypothesis, like hypothesis 5, each item on the questionnaire that sought to find out the respondents' attitudes towards English was tested individually (Appendix III). 0n item 25 ("English is the only language that should be used in offices"), no significant relationship was found with the language the middle level managers use when 60 2 interacting with their superiors (X = 11.12475; df = 12; p > .05), and subordinates (X2 = 11.92420; df = 12; p > .05). However, highly significant relationship was found with the language the middle level managers use when interacting with their peers (X2 = 32.11748; df = 8; p < .05). On item 26 ("All office workers should be encouraged to use English at work"), no significant relationship was found with the language the middle level managers use when interacting with their superiors (X2 = 15.14165; df = 12; p > .05), peers (X2 = 13.87805; df = 8; p > .05) and subordinates (X2 = 16.02804; df = 12; p > .05). On item 27 ("only effective use of English can contribute to an officer's career development"), no significant relationship was found with the language the middle level managers use when interacting with their superiors (X2 = 8.05729; df = 12; p > .05), 'peers (X2 = 9.20064; df = 8; p > .05), and subordinates (X2 = 3.4236; df = 12; p > .05). On item 28 ("Good command of English ensures respect at work"), no significant relationship was found with the language the middle level managers use when interacting with their superiors (X2 = 20.04557 df = 12; p > .05), peers (X2 = 12.13115 df = 8 p > .05), and subordinates (X2 = 9.43262 df = 12 p > .05). From the above findings, we could conclude that the respondents' attitudes towards English have no significant relationship with the language they choose to use when 61 interacting with their superiors, peers and subordinates. These results are depicted on table 6. As the table indicates, the respondents generally show negative attitudes towards the use of English at work. Of the total, 93.1 percent of the respondents do not consider English as the only language that should be used at work; 77.6 percent do not think that all office workers should be encouraged to speak English at work, 66.5 percent do not consider effective use of English as the only determinant of an officer's career development; and 49.5 percent do not think good command of English ensures respect at work. However, although the respondents show negative attitudes towards English, they use it alone or in combination with Kiswahili more frequently when speaking to their superiors and peers. Also, when speaking to their subordinates, although the respondents use Kiswahili alone, still a combination of English and Kiswahili is common. 62 Table 6. Percentages of Attitudes towards English by language choice. I I I I ATTITUDES TOWARDS I LANGUAGE CHOICE I I ENGLISH I I ------- I-_-_-_------ I --- ---____-----__--__------------I I I I I I I I I I I Items I I I IDirectionalityIEnglishIKiswahiliIEnglish/ [First [Kiswahili/I IPositiveINeutralI Negative I of IOnly IOnly IKiswahiliILanguageIFirst I I(Agree) I I(Disagree)IInteraction I I I IOnly ILanguage I _______ I I I--- I I__--_--I I-_ I- I I I I I I I I I I I I I I I I superiors I 66.4 I 9 I 31.9 I - I .9 I I I l I ------------- I ------- I I -I- I I Item 25I 6.0 I .9 I 93.1 I peers I 69.0 I 1.7 I 29.3 I — I - I I I I I I I I I I ’--'I I I I I subordinates I 7 8 I 38.8 I 48.3 I 5.12 I - I ------ I I I I I I "-I"--"---I"--"-’I'-'-""’*I I I I I superiors I 66.4 I .9 I 31.9 I - I .9 I I I I I ------------- I ------- I I I I ~---I Item 26I 20.6 I 1.7 I 77.6 I peers I 69.0 I 1.7 I 29.3 I - I - I I I I I I I ----I I I I I I I I subordinates I 7.8 I 38.8 I 48.3 I 5.2 I - I *I I ------- I I I I ‘I I I-' I I I I I superiors I 66.4 I .9 I 31.9 I - I 9 I I I I I I I I I — I “I Item 27| 28.4 I 6.0 I 65.5 I peers I 69.0 I 1.7 I 29.3 I - I — I I I I I ----------- I I I I "I ---------- I I I I I subordinates I 7.8 I 38.8 I 48.3 I 5.2 I - I I I I--- I I' I I I I I I I I I superiors I 67.0 I .9 I 31.3 I - I .9 I I I I I """"""""" I """"" I -------- I -------- I I I Item 28I 42.6 I 7.8 I 49.5 I peers I 69.6 I 1.7 I 28.7 I - I - I I I I I ---------- I ------- I ‘I --"I I I I I I I subordinates I 7.8 I 38.3 I 48.7 I 5.2 I - I _______ I---_-_--I-------I-----_--__I---------_----I------_I_-__-_---I----__--_I-_------I_-----_---I Item 25: English is the only language that should be used in offices. Item 26: All office workers should be encouraged to use English at work. Item 27: Only effective use of English can contribute to an officer's career development. Item 28: Good command of English ensures respect at work. CHAPTER 5 DISCUSSION In this chapter, the findings of each hypothesis will be independently addressed, and the relevant data analysis discussed. In addition, language use with the public, theoretical issues pertinent to these findings, and suggestions for further research will be presented. Finally, general conclusions will be drawn and their relevance and significance briefly discussed. Language Competence Middle level managers with a high degree of competence in Kiswahili were expected to use Kiswahili more frequently than those with low competence in it when interacting with their superiors, peers and subordinates. However, as the respondents were highly educated, their competence in English was not expected to vary considerably. This study, like earlier studies (e.g. whiteley, 1974b; Parkin, 1974a; Bujra, 1974) based competence to speak a particular language on the respondents' claimed competence. As table 1 indicates, there is no significant relationship between the degree of competence in Kiswahili and the language the middle level managers use when interacting with their superiors, peers and subordinates. 63 64 This may be due to the fact that the respondents claimed a high degree of competence in Kiswahili. Of the total, 97.4 percent of the respondents claimed to be "very good," "good" and "adequate" in Kiswahili. None of the respondents claimed to be "very poor." We would partly attribute this high degree of Kiswahili competence to the fact that most of the respondents had formal instruction in Kiswahili. Of the total, 94.9 percent had formal instruction in Kiswahili. Specifically, 18.6 percent had formal instruction in Kiswahili upto primary school level, 61.9 percent secondary school level (O'level), 10.2 percent higher secondary school level (A'level), and 4.2 percent college or university level. Another plausible explanation for the high degree of competence in Kiswahili is that most of the respondents had worked for a long period of time in the multilingual public organization. The respondents had a mean working duration of 8.3 years, which means that those who had not mastered Kiswahili in school, could have informally mastered it after being exposed to the multilingual organization where Kiswahili is frequently used. The respondents' claimed competence in English was also very high. Thus, of the total, 99.1 percent claimed to be "very good," "good" and "adequate," and only .9 percent claimed "not good." None of the respondents claimed to be "very poor" in English. However, it seems that there is a slightly higher degree of claimed competence in English than 65 in Kiswahili. Specifically, 86.9 percent of the respondents claimed to be "very good" and "good" in English, while only 58.1 percent claimed a similar degree of competence in Kiswahili. This probably means that a high percentage of the middle level managers would claim to speak English more effectively than Kiswahili. As this study wholly relies on data collected by use of a questionnaire, it is difficult to know whether the high degree of competence especially in Kiswahili reflects the actual situation. This problem is clearly articulated by Whiteley (1974, p. 3) when he observes that: If you want to know what language a person uses in a particular context, and how he uses it, it is not enough simply to ask him. He may tell you what he thinks you want to hear, or what he thinks he does say, or what he would like others to think he says To overcome the problem Whiteley mentions above, participant observation technique can be used to supplement the questionnaire, but as stated in the introduction of this study that was not possible. However, it is worth noting that although the respondents were highly educated, and the majority having had formal instruction in Kiswahili, in reality there is a small percentage of middle level managers that has not mastered Kiswahili. This may be more common with the young officers who have not been exposed to the multilingual organization for a long time, and might have attended schools (both primary and secondary) that never taught Kiswahili at all (Itebete, 1976), or they come from linguistically homogeneous communities where the need for a 66 lingua franca like Kiswahili is minimal. For example, a young administrative officer who had just graduated from the university of Nairobi commented after she had been assigned the responsibility of chairing a meeting for all the drivers in the Ministry Headquarters: I don't know Kiswahili at all. The drivers could speak better than me. They discussed the agenda of the meeting while I embarrassingly gazed at them I hope I will never be assigned such a duty again (Personal Communication, 1982). The above case supports the arguement that there are few middle level managers who cannot handle a formal discourse in Kiswahili effectively. This predicament may be attributed to the Kenyan education system which over the years has trivialized Kiswahili in the school curriculum. The lack of incentives for those who master Kiswahili (Itebete, 1976) has led to a situation where some schools, do not teach it. It is vital to note that there are public officers who may claim to be highly competent in Kiswahili because they can casually use it, but they may find it difficult to transmit complex organizational information to those who do not speak English. However, the officers who are relatively incompetent in Kiswahili realize its organizational importance, and consequently, they informally master it over time. Aga The older middle level managers were expected to speak their first language and Kiswahili more frequently than the 67 younger managers. Conversely, the younger managers were expected to speak English more frequently as a mark of their elite status that differentiates them from the relatively uneducated older generation. However, as table 2 shows, no significant differences were found between the younger and the older middle level managers in their use of English, Kiswahili and first languages. This finding is contrary to the findings of earlier researchers (e.g. Heine, 1980; Bujra, 1974; Parkin, 1974a; Whiteley, 1979) who had found that young people tend to speak English more frequently than the older people. All the respondents in this study regardless of their age claimed high competence in the three languages. It is vital to note that the respondents in the earlier studies mentioned above were relatively uneducated, working in lower organizational positions or self employed in small businesses. Thus, it is possible that the younger people who were found to speak English more frequently than the older people had only a few years of formal education. However, in the current study, all the respondents were higly educated and worked in relatively high organizational positions. Therefore, both the young and the old had high linguistic competence in the two lingua francas. Although it is possible that the younger managers may be more educated than the older managers, their competence especially in English is not higher than that of the older managers. The older managers' educational status may have 68 been low when they joined the public service, but their exposure to a work situation where English is constantly used may have enhanced their competence. This argument is consonant with Bujra (1974) and Parkin's (1974a) observation that people who have had a chance to constantly use English even though their educational level is low are able to competently interact in it. Educational Level It was expected that middle level managers who are highly educated would choose to speak English more frequently than Kiswahili and first languages. Conversely, those with low educational status were expected to speak Kiswahili or their first languages more frequently because they might not be highly competent in English. As table 3 indicates, no significant relationship was found between educational level and the language the middle level managers use when interacting with their superiors and subordinates. But significant relationship was found with the language the middle level managers use when interacting with their peers. Contrary to the findings of earlier researchers (e.g. Parkin, 1974a; Scotton, 1982b) that as education rises, English usage increases, but the use of Kiswahili decreases, in this study, the rise in education does not seem to affect the frequency of English or Kiswahili usage. As the respondents were highly educated, with 97.5 percent having had at least secondary school education 69 (O'level), their competence in both English and Kiswahili did not vary considerably. This means that 97.5 percent of the respondents had formal instruction in English for at least 12 years, and 94.9 percent had formal instruction in Kiswahili. Therefore, unlike in the earlier studies (e.g. Bujra, 1974; Parkin, 1974a; Whiteley, 1974b) where the relatively uneducated respondents had only a few years of formal instruction in English and Kiswahili, most of the respondents in this study had formally acquired high competence in both languages. Every middle level manager spends a substantial amount of time either giving instructions and orders to the subordinates, or informing and educating the public on government policies and regulations. This means that. those who have acquired and mastered Kiswahili in school utilize their skills when interacting with the relatively uneducated officers and members of the public. Therefore, Kiswahili is widely used by the highly educated, although they tend to combine it with English. In sum, the middle level managers in this study are highly competent in both English and Kiswahili, and they tend to use them frequently regardless of their educational level. Occupational Status Middle level managers with high occupational status were expected to speak English more frequently as a symbol of their high socio—economic status, power and authority. Conversely, those with low occupational status were expected 70 to speak Kiswahili and first languages more frequently when interacting with other organizational members. However, as table 4 shows, no significant relationship was found between occupational status and the language the middle level managers use when interacting with their superiors, peers and subordinates. Regardless of the middle level managers' occupational status, English alone or in combination with Kiswahili is mostly used with the superiors and peers, and Kiswahili alone or in combination with English when speaking to the subordinates. To maintain the status differences, the middle level managers mostly use English alone when speaking to their superiors. On the other hand, they mostly use Kiswahili alone when interacting with their subordinates as a mark of authority and power. However, the middle manager may use his/her first language when speaking to the superior or subordinate to reduce the level of formality, and consequently interact as equal members of the same ethnic group. Fluent English speakers are presumably evaluated as organizationally highly productive officers. Thus, it is probable that the middle level managers reinforce this notion when they choose to speak English more frequently to their superiors. However, while this may be possible in written communication regardless of its failure to communicate effectively to the non-English speaking organizational members, it is highly equivocal whether 71 effectiveness in verbal interaction can be wholly based on an officer's competence in English. One plausible explanation for the lack of significant relationship between occupational status and the language use behavior of the middle level managers might be their relatively high competence in both English and Kiswahili. Moreover, for the few whose educational level is low, their middle level managerial positions calls for high competence in the two lingua francas, therefore, by constantly using them, their competence has increased over time. Attitudes Towards Kiswahili Middle level managers with positive attitudes towards Kiswahili were expected to speak it more frequently than those with negative attitudes towards it. As the attitude items were designed to represent peoples' views and beliefs about English and Kiswahili, this assumption was based on the premise that speakers with negative attitudes towards Kiswahili would perceive it as: (1) a foreign language like English, (2) inferior to English, (3) difficult to learn and speak than English, and (4) that those who speak it fluently are either uneducated or of low status (see Appendix III). As table 5 indicates, no significant relationship was found between the middle level managers' attitudes towards Kiswahili and the language the managers use when interacting with their superiors, peers, and subordinates. However, the respondents' responses show positive attitudes towards the use of Kiswahili at work. 72 As already noted, most of the respondents were highly educated, and apparently, a high percentage (94.9 percent) had formal instruction in Kiswahili. Although Kiswahili has no equal status to English in the education system, its recognition as a school subject is an indication of its national importance. Furthermore, as most of the respondents were relatively young, they went through the education system in independent Kenya when the government has been showing increasing concern on the role Kiswahili plays in national development and integration. For instance, the adoption of Kiswahili as an official language of the National Assembly in 1974, is an important landmark that may have changed the respondents‘ attitudes towards it. Thus, as stated in the Kenyan constitution: . . the official languages of the National Assembly Shall be Kiswahili and English and the business of the National Assembly may be conducted in either or both such languages (Republic of Kenya, 1979, p. 30). The use of Kiswahili in the National Assembly indicates its importance as a medium of communication between the government and the public (Soba, 1984). Kiswahili today has greater standing in government and education than during the colonial era. A more recent recognition of Kiswahili in the education system is its being one of the ten examination subjects in the Kenya Certificate of Primary Education (KCPE) under the new 8-4—4 education structure (The Weekly Review, Nov. 22, 1985). For the first time in Kenya's history as an independent nation, 73 Kiswahili was examined in the primary schools in november, 1985. This is an attempt by the government to raise the linguistic competence of the youth in the language that has cardinal role to play in national development and ' integration. Therefore, we would reasonably argue that the policies the government has instituted in favor of Kiswahili may have contributed to the middle level managers' positive attitudes towards Kiswahili. Further, although government official correspondence is usually in English (Rhoades, 1977; Heine, 1977), it is vital to note that many official documents and forms such as birth certificate, passport, national identity card, driver's license and medical forms are written in English alongside Kiswahili. Therefore, given that these documents are structured, issued and managed by the middle level managers, it is possible that they realize the role Kiswahili plays in making the documents understood by the relatively uneducated Kenyan populace. In addition, most of the respondents, especially those in provincial administration, social work, police and probation use kiswahili mostly in their daily organizing activities. Certainly, if Kiswahili is effective in their implementing government policies and regulations, that may contribute to their perceiving it positively. For instance a senior government officer, Diah-Wilson (1984) speaking on behalf of participants of Advanced Public Administration course who had successfully completed a Kiswahili speech 74 writing and presentation project at KIA said: We as government administrators realize that if we are not fully competent in the language that those we administer and serve speak, serious communication problems may arise, and consequently, the goals and objectives of our government in national development may be adversely affected. The cardinal role Kiswahili plays in nation building has been verbally articulated from many quarters of the public sector. Public statements made by senior public officers and politicians in favor of Kiswahili has been evident for the last two decades. Probably, this realization of the importance of Kiswahili in Kenya has encouraged public officers to speak it more frequently, and consequently perceive it positively. For instance, for the provincial administrative cadre, who were the largest occupational group (38.7 percent) in this study, their ability to implement and articulate government policies and regulations largely depends on their competence in Kiswahili, the language most members of the public speak and understand. However, the factors that affect how an individual in a particular speech community uses a given language, as Gorman (1974a. p. 397) observes are also a product of "Complex historical, social and economic forces . . .." For example, the widespread use of Kiswahili as a tool of unifying the various Kenyan ethnic groups between the 1920's and 1950's was a result of the growing nationalism which was eminent before Kenya attained her independence. Thus, despite the colonial language policies that suppressed Kiswahili usage 75 (Mkangi, 1982), its role as a lingua franca was realized at the time. This recognition of Kiswahili as a vital medium of inter-ethnic communication has ensured in independent Kenya. Thus, although government policies in regard to the use of Kiswahili at work have not been formally documented, historical, social and political forces enhance its national importance. Despite the increase in official recognition of Kiswahili during the last two decades, English still remains the language of vertical mobility. It is obvious that the day of Kiswahili attaining equal status to English is yet to dawn. It is not surprising then that there is a small percentage of the middle level managers in this study who show negative attitudes towards Kiswahili. For instance, of the total, 21.0 percent consider Kiswahili a foreign language like English, 5.1 percent consider a fluent Kiswahili speaker as generally either uneducated or of low status, 32.2 percent consider Kiswahili as lacking the vocabulary needed in many work situations, and 17.9 percent consider Kiswahili as more difficult to learn and speak than English. We would expect members of this group to be reluctant to speak Kiswahili unless one is in a desperate situation where he/she and the addressee(s) have no other choice but to speak it. Such managers would probably not speak Kiswahili to their superiors and peers, but only to the subordinates who lack competence in the prestigeous English language. Probably, when they speak Kiswahili to the 76 subordinates, they combine it with English regardless of the subordinates' competence in English. Attitudes Towards English Middle level managers with positive attitudes towards English were expected to speak it more frequently than those with negative attitudes towards it. This assumption was based on the premise that interlocutors with positive attitudes towards English would perceive it as: (1) the only language that should be used in offices, (2) the language that all office workers should be encouraged to use at work, (3) the only language that can contribute to an officer's career development, and (4) that good command of it would ensure respect at work (Appendix III). As table 6 indicates, no significant relationship was found between the middle level managers' attitudes towards English and the language they use when interacting with their superiors, peers and subordinates. However, as table 6 shows, most of the respondents' responses generally indicate that they have negative attitudes towards the use of English alone at work. This may be partly attributed to historical, social and political forces which have been working in favor of Kiswahili as already discussed. Since the British colonial era, English has been associated with high socio-economic status and prestige. For example, when English was introduced as a medium of instruction in primary schools, the then gag: African Standard (now the Standard) on the 13th may 1960 77 stated, "if mothers of tomorrow could bring up their children with English in the home, progress would be swifter and surer" (Quoted from Mkangi, 1982, p. 15). The above view which probably could have been held by many Kenyans at the eve of Kenya's independence and immediately after, seemingly has changed. English is no longer viewed as the only medium of communication that is vital for Kenya's national development. Although English still remains the vehicle of socio-economic mobility, it is however evident in today's Kenya that even those who have not acquired this prestigeous language are equally productive members of the Kenyan society. Thus, the increasing government recognition of Kiswahili and widespread use of it by politicians and senior public officers has to a large extent increased the awareness of many Kenyans that English alone is deficient in many aspects of national development. Therefore, the socio- political situation that the middle level managers find themselves today, perhaps reinforce the fact that using English alone in a multilingual nation where the majority of the population lack competence in it is inappropriate. Furthermore, educational opportunities have been available to more Kenyans than was the case before independence. For example, in 1957, there were 3,134 African pupils in secondary schools (East African High Commission, 1955-58), but in 1977 there were 280,388 pupils (Republic of Kenya, 1975-77). Although this means that more Kenyans have acquired and mastered English, it is possible that the 78 values previously attached to English have changed. During the colonial era, and the first decade of Kenya as an independent nation, the few public officers who were fluent in English were considered a prestigeous group. But today, all the public officers, except some in the lower organizational levels are relatively highly educated, hence socio—economic prestige can only be derived from other sources other than competence in English. Thus, although English does play a role in appointments and professional growth, competence in Kiswahili is equally important if a public officer has to interact effectively with the subordinates and members of the public. Therefore, the realization that English alone cannot fully meet a public officer's communication needs in a multilingual organization, may also have contributed to the respondents' negative attitudes towards English. However, the argument is not that English is totally inadequate or unnecessary, but that on its own, most of the respondents perceive it inadequate. Possibly, this has shifted the respondents' favor towards Kiswahili, which has been trivialized for a long time. By viewing Kiswahili as a vital language for inter—ethnic communication in the public sector, it could be argued that the officers consider English and Kiswahili as languages that should be used together, rather than English being used at the expense of Kiswahili. We may perhaps say that the middle level managers with negative attitudes towards English would choose 79 Kiswahili more frequently when speaking to other organizational members. Also, we would expect them to be reluctant to speak English to any fellow officers, especially the subordinates who might be incompetent in it. However, there is a significant proportion of the middle level managers who generally show positive attitudes towards English. Of the total, 5.8 percent of the respondents consider English the only language that should be used in offices, 21.2 percent think that all office workers. should be encouraged to use English at work, 28.0 percent believe that only effective use of English can contribute to an officer's career development, and 42.7 percent believe that good command of English ensures respect at work. Most of the respondents who have positive attitudes towards English, View it as a language that can contribute to an officer's career development and is a symbol of social prestige at work. Possibly, they view English as a language that has greatly contributed to their career achievement, and hence should be strengthened while relegating Kiswahili to a lowly position. Further, it could be argued that middle level managers with positive attitudes towards English would probably choose to use it even when the addressee is not fully competent in it. To such officers, as Bujra (1974) found, speaking English differentiates them from those who have not mastered it. On the other hand, we may perhaps say that those who have positive attitudes towards English probably 80 perceive Kiswahili negatively. Such managers would choose to speak English more frequently to their superiors and peers, and even to the subordinates who have a rudimentary knowledge of it. Ethnicity First language usage was assumed to be an indicator of ethnic consciousness, identity, loyalty and affiliation (Parkin, 1974c; Mkilifi, 1978; Whiteley, 1979). Although no specific hypothesis was tested, it is important to discuss first language usage in the Kenyan public sector in relationship to language choice and use. As earlier researchers (e.g. Rhoades, 1977; Whiteley, 1979; Heine, 1980) noted, first languages are a very important medium of communication in work situations. In this study, 68.9 percent of the respondents claim to speak their first languages at work when interacting with members of their ethnic group. Of these, 23.8 percent speak to their peers, 26.2 percent to subordinates, .12 percent to superiors, and 47.6 percent to any fellow officer who speaks their first language. It seems that first languages are used mostly when interacting with peers and subordinates, rather than superiors. A first language reduces the interactants' social distance and level of formality (Whiteley, 1979), therefore, unless the superior initiates it, a middle level manager would choose to maintain the status differences by choosing a neutral language (English or Kiswahili) when speaking to the superior. 81 The use of a first language in a multi—ethnic environment is stigmatised (Parkin, 1974c, 1977; Scotton, 1976b, 1983), therefore, it is possible that the 31.1 percent of the respondents who claimed that they "never speak it," consciously refused to do so to avoid the stigma. First languages, as Parkin (1977, p. 193) clearly states, "connote ethnic inclusiveness and solidarity . . . and, conversely, exclusion and opposition when used in ethnically mixed contexts." Thus, first languages are accused of perpetuating ethnic parochialism, sectionalism, nepotism and other organizational ills which are ethnic—centered. The stigma associated with ethnicity in the public service is reflected in a comment made by a senior public officer who said: there are scores upon scores of people who only think in terms of how they can achieve maximum benefit for themselves and their own kith and kin. We have small pockets of Civil Service 'clans' which tend to work for the benefit of only those who belong to their families or ethnic connections (Nyamu, 1980, p. 117). The degree of ethnic parochialism and loyalty shown in the above comment is believed to be enhanced when public officers withhold information from other interlocutors who do not share their first language. Therefore, given that public officers are supposed to project a positive image of the public organization, it is no surprise that although the use of first languages at work is widespread, they are however resented by those who are excluded in the discourse. From a communication standpoint, one major limitation 82 of the use of a first language in a multi-ethnic work situation is the restriction it imposes on organizational information flow. Only those who share the first language can receive and send information using that medium. However, although the information transmitted may be only pertinent to the in-group members, the out—group members may question and resent the use of that medium. Apparently, the use of neutral languages such as English, and Kiswahili which ”penetrates through ethnic boundaries" (Karigithe, 1981) is encouraged in the Kenyan public sector. First languages should not be viewed wholly as a medium of communication that threatens organizational stability. They do play an insurmountable role of transmitting information to those who may not be fully competent in either English or Kiswahili. For example, a middle level manager who realizes that a subordinate is not competent in both English and Kiswahili may use the first language if the two share it. Although such a language choice may be negatively perceived by other interlocutors, from an organizational standpoint, the choice is appropriate if pertinent information is effectively disseminated and received. Thus, while we agree with Soba (1984) that ethnic differences can be minimized if interlocutors use Kiswahili and not first languages in inter—ethnic interactions, we could however emphasize that any language used in transmitting information, one of the major resources of an organization is important. 83 Moreover, as an organization is a social unit, members of the same ethnic group may form informal groups which may work in support of the formal organization. Such informal groups which are enhanced by the sharing of a first language satisfy workers' social needs which may not be satisfied by the formal organizational structure. By sharing a first language, a member of the informal group feels socially accepted by other ethnic members, and a sense of belonging to a particular group is a positive feeling to all organizational members. Directionality pf Interaction Several discernible patterns of language choice and use can be identified in this study depending on whom the middle level manager is speaking to. When the middle level managers interact with their superiors, English alone or in combination with Kiswahili is mostly used. Specifically, 66.7 percent of the respondents use English alone, 31.6 percent use a combination of English and Kiswahili, .9 percent use Kiswahili alone, and .9 percent use first language alone. This finding is consonant with Whiteley (1979) and Scotton's (1982c) observation that English is used mostly with the senior officers, and Kiswahili with those who are junior to oneself. Kiswahili alone, and first language alone, or a combination of the two is rarely used when interacting with the superiors. For horizontal interaction when middle level managers interact with their peers, a similar pattern as when 84 interacting with the superiors is seen. Of the total, 69.2 percent of the respondents use English alone, 29.1 percent use a combination of English and Kiswahili, and 1.7 percent use Kiswahili alone. It seems that when interacting with the peers, both English alone and Kiswahili alone are used slightly more frequently than with the superiors, but first languages are rarely used with the peers. Thus, for upward and horizontal communication, the middle level managers mostly speak English alone or they combine it with Kiswahili. One plausible explanation for this similar pattern of language choice and use when the middle level managers speak to the superiors and peers is that both have relatively high educational level and socio—economic status in comparison to the lower organizational workers. Thus, given that they are relatively highly competent in both English and Kiswahili, then the language use behavior when the middle level manager interacts with the superiors and peers does not seem to vary considerably. For downward interaction, when the middle level managers interact with their subordinates, Kiswahili alone or in combination with English is mostly used. Specifically, 39.3 percent use Kiswahili alone, 47.9 percent use a combination of English and Kiswahili, 7.7 percent use English alone, and 5.1 percent use a combination of Kiswahili and first language. Further, for upward interaction when subordinates speak to the middle level 85 managers, a similar pattern of language choice and use is evident. Of the total, 41.0 percent of the respondents claim that their subordinates use Kiswahili alone, 42.7 percent use a combination of Kiswahili and English, 13.7 percent use ' a combination of Kiswahili and first language, 1.7 percent use first language alone, and .9 percent use English alone. It is worth noting that middle level managers use English alone more frequently when interacting with their subordinates than vice versa. 0n the other hand, more subordinates use Kiswahili alone or first language alone more frequently when interacting with the middle level managers than vice versa. One may perhaps say that there are middle level managers who speak English alone to the subordinates although the subordinates might not be fully competent in it. It could be argued that such language choice to a large extent projects a negative image of the middle level managers because the subordinates may not perceive them as credible speakers. The middle level managers as an elite group would prefer to use English more frequently to assert their high socio-economic status, authority and power. One plausible explanation is that when the middle level managers speak English to the subordinates, the relatively uneducated junior officers may not interact at par. Thus, the middle level manager has an advantage in the discourse. While the educated people can effectively interact in both English and Kiswahili, as Scotton (1982b, p. 80) notes, the relatively 86 uneducated workers ”do not have the same options in their repertoire". Speaking Kiswahili alone to the subordinates is almost equally preferred as combining it with English. However, we can detect some discrepancy between this pattern of language choice, and the language(s) the middle level managers perceive as the most effective that a manager should use when interacting with the subordinates. Of the total, 65.5 percent of the respondents consider a combination of English and Kiswahili as the most effective choice that a middle level manager should make when interacting with the subordinates, 22.2 percent Kiswahili alone, 7.6 percent a combination of Kiswahili and first language, and 6.7 percent English alone. The major discrepancy is that Kiswahili alone is used more (39.3 percent) than is preferred (22.2 percent), while a combination of English and Kiswahili is used less (47.9 percent) than is preferred (65.5 percent) as the most effective choice when interacting with the subordinates. What is eminent in these findings is that both Kiswahili and English are important for disseminating information downwards by the middle level managers, and seemingly, none can be use singly. From the above discussion, it is clear that a middle level manager has multiple identies when interacting with the superiors, peers and subordinates. By using English, Kiswahili or first language depending on whom the middle manager is speaking to, as Scotton (forthcoming) puts it, it 87 is "as if the speaker were saying, not only am I x, but I am also Y." The middle level manager relates with the superiors and peers by speaking English alone or combining it with Kiswahili, and using Kiswahili alone or combining it with English when speaking to the subordinates. Both English and Kiswahili are more linked to occupational roles, while first languages are used as a link with members of the same ethnic group in the multilingual organization. Language Use with the Public Middle level managers like other public officers are supposed to inform and educate the public on government policies and regulations pertinent to national development (Soba, 1984). Such responsibility can be effectively executed if public officers speak the language(s) the public understands and speaks. Thus, using the most appropriate language will ensure that information is effectively transmitted from the government bureaucracy to the public, and feedback from the public is received by the government. Given the high degree of multilingualism in Kenya, effective interaction between public officers and the public can largely be realized if the two lingua francas, English and Kiswahili are used. However, as most of the members of the public have not acquired and mastered English, then Kiswahili remains the sole candidate. This is indicated by the respondents' response on the item (see Appendix III) that sought to know which language they would consider as the most effective when speaking to the public. Kiswahili 88 alone or in combination with English was considered the most effective. Specifically, 37.0 percent of the respondents consider Kiswahili alone to be the most effective, 30.3 percent a combination of Kiswahili and English, 29.4 percent a combination of Kiswahili and first language, 1.7 percent English alone, and 1.7 percent first language alone. It seems, English alone, or first language alone are not considered effective unless combined with Kiswahili. In the light of this finding that most respondents consider Kiswahili alone or in combination with English the most effective choice when speaking to the public, it is reasonable to assume that most public officers are willing to speak Kiswahili to the public. This argument contrasts with Bujra's (1974) assertion that public officers are often reluctant to speak Kiswahili to the public, and consequently the public has to rely on brokers who are competent in English when dealing with public officers. If there are public officers who speak English alone to members of the public who might not be competent in it, then such a language choice and use is inappropriate and mars the image of the public organization. But given that only 1.7 percent of the respondents consider English alone an effective choice when speaking to the public, we would expect a small percentage of public officers to make this inappropriate language choice. It is worth noting that, as the rural areas are now the centers of national development under the Rural District 89 Focus Development Strategy, the role of Kiswahili has become more viable than ever before. The success of this development strategy to a large extent depends on effective dissemination of information by public officers to the rural Kenyans. This can be realized if the public officers and the public can attempt to create a "homogeneous linguistic community" (Le Page, 1964, p. 15) by mostly speaking Kiswahili. For example, the provincial administrative cadre, who are also the Chairpersons of various rural development committees have to use Kiswahili effectively if the public is to participate fully in rural development. Further, according to the government code of Regulations (Republic of Kenya, 1966), public officers are "liable to be posted to any station within Kenya to discharge the usual duties of their office." This means that many public officers work away from their home districts where different first languages other than their own are spoken or in urban areas which are largely multilingual. Thus, although a combination of Kiswahili and first languages was considered effective by 29.4 percent of the respondents, such language use is only possible for middle level managers working in their home areas where their first languages are used, or those in the urban areas when they interact with people from their own ethnic group. Therefore, Kiswahili remains the main inter-ethnic language most of the public officers use when serving the public. 90 Theoretical Implications a: the Study The theoretical implications will be discussed in the light of the three theoretical models (domain, accommodation and markedness models) presented in Chapter 2 of this study. Domain model Our findings in this study show that all the three distinct languages (English, Kiswahili and first languages) are used in varying degree in the work domain of the Kenyan public sector. This is contrary to the central assumption of the domain model that in a multilingual environment, one language may be exclusively used in one domain but not in the other(s) (eg. Hunt, 1966; Mackey, 1965, 1966; Fishman, 1972; Whiteley, 1974b; Halliday, 1972; etc.). For instance, although the first languages might be predominantly used in the home domain, like English and Kiswahili, they are significantly used at work. However, this study did not attempt to find out whether the topic of discourse, purpose of interaction and role relationships have an impact on language choice as the domain model stipulates. But several implications can be discussed from the study. First, in regard to the topic of discourse as a factor in determining language choice, it is possible that English is used more frequently when discussing technical topics than Kiswahili or first languages. But rather than the topic of discourse per se 91 being the sole determinant, the interactants' role relationship with the addressee may be a significant factor. Thus, dyadic relationships such as middle level manager/superior; middle level manager/peer; and middle' level manager/subordinate seem to have an impact on language choice. Therefore, as already discussed, the middle level managers mostly use English alone or combine it with Kiswahili when interacting with their superiors and peers and Kiswahili alone or in combination with English when speaking to their subordinates. In addition to the role relationships of the interlocutors, we would say that the purpose of interaction also partly influences language choice as the domain model assumes. For example, a middle level manager uses English when speaking to the superiors and peers when his/her objective is to maintain formality in the discourse, but uses Kiswahili if the objective is to interact informally. While when interacting with the subordinates, the middle level manager uses English if the objective is to emphasise his/her high educational level and high socio—economic achievement that differentiates him/her from the uneducated subordinates. But the manager uses Kiswahili to assert authority, power and formality when speaking to the subordinates. On the other hand, first languages are used when the middle level manager intends to interact informally with other members of his/her ethnic group. It is important to note that language choice is a 92 complex process especially when interlocutors are equally competent in at least two languages that are potential choices in the discourse. For instance, when speaking to the superiors and peers, there are two potential choices (English and Kiswahili), but three potential choices (English, Kiswahili and first languages) with those who are members of the same ethnic group. This means that choosing one of the languages is not as clear—cut as the domain model stipulates. Accommodation Model The assumptions of the accommodation model (Giles and Smith, 1979; Giles, Taylor and Bourhis, 1973; Taylor and Royer, 1979) to a large extent are congruent to the findings of this study. According to the model, we would reasonably say that the middle level managers choose a given language as a convergence strategy to reduce or maintain the level of formality, or a divergence strategy to increase the level of formality. Therefore, the middle level managers employ a convergence strategy mostly when interacting with their superiors and peers, who are in-group members with relatively high educational level, high occupational status and high linguistic competence in both English and Kiswahili. Thus, English alone or in combination with Kiswahili is mostly used with the in-group members (superiors and peers) as a convergence strategy. While when interacting with the out-group members (subordinates) who 93 have relatively low educational status, low occupational status, and low linguistic competence especially in English, the middle level managers mostly use Kiswahili alone or in combination with English as a divergence strategy. Further, the middle level managers use their first languages as a convergence strategy when mostly interacting with their peers and subordinates, but rarely with the superiors. Markedness Model In respect to the markedness model (Scotton, 1972, 1979, 1982a, 1983, 1985 etc.), the role relationships of the discourse participants seem to strongly influence the language use behavior in the Kenyan public sector. Although our hypotheses were not fully confirmed, it is however implicit from the preceding discussion that the language(s) the middle level manager chooses to a large extent depends on whom he/she is speaking to. When interacting with the superiors and peers, English and Kiswahili are the unmarked choices. English is an unmarked choice because the superiors and peers are highly competent in it. Therefore, as officers in relatively high organizational positions, they are expected to speak English as a mark of authority, power and high socio-economic prestige. On the other hand, Kiswahili is an unmarked choice because it is an official and national language that Kenyans are expected to speak in inter—ethnic situations to signal national identity and solidarity. Thus, the unmarked choice English is mostly used by the middle level managers when 94 speaking to their superiors and peers. However, although Kiswahili is an unmarked choice, it is rarely used alone except in combination with English when speaking to the superiors and peers. When speaking to the subordinates, Kiswahili, is an unmarked choice because most of the subordinates are not competent in English. But English is a marked choice because its use may be interpreted as an attempt by the middle level manager to signal his/her high educational status that differentiates him/her from the relatively uneducated subordinates. In addition, first languages are marked choices regardless of whom the middle level manager is speaking to. As first languages signal ethnic solidarity, affiliation and identity, using them is often interpreted as a conscious attempt by the speaker to exclude the non-first language speakers as already discussed. The three distinct languages (English, Kiswahili and first languages) tend to be used by the middle level managers in varying degree when speaking to their superiors, peers and subordinates. This may partly indicate that the choice depends on what the interlocutors perceive as the most salient feature in their discourse. When power, authority, educational level and socio-economic prestige are salient, the middle level managers mostly use English with their superiors and peers. They may also use Kiswahili when national identity and solidarity are salient. The use of English with the superiors and peers ensures formality, and 95 Kiswahili ensures informality of the interaction. Further, when interacting with the subordinates, if socio-economic prestige, authority and power are salient, the middle level managers use Kiswahili. But when educational status is salient, the middle level managers use English to signal their high educational status. However, regardless of the addressee's organizational level, when ethnic solidarity, affiliation and loyalty are salient, the middle level managers use their first languages with members of their ethnic group. The use of the three languages in varying degree may also be explained by the consequences the speaker anticipates from using any of them. When interacting with the superiors and peers, speaking English projects a positive image of the speaker, but its use may be negatively perceived when used with the subordinates. On the other hand, speaking Kiswahili to the subordinates is positively perceived, but a middle level manager may be reluctant to speak Kiswahili to his/her superior, if the superior is not fully competent in it. Further, although first languages are vital medium of communication especially with subordinates who might not be fully competent in both English and Kiswahili, the stigma attached to their use may reduce the frequency of their usage. For instance, as already noted, first languages are rarely used with the superiors because probably the middle level managers fear that their bosses may resent the attempt to reduce the level 96 of formality, unless the superior initiates it. In essence therefore, as the markedness model stipulates, speakers weigh the rewards and costs of the language they choose in a given interaction. Suggestions for Future Research Few studies have previously been undertaken directly under the rubric of language use in a multilingual organization. As language use is a vital communication phenomena in a multilingual organization, a few suggestions for future research will be presented. First, it would be important for future researchers to focus on the effects of language use on organizational activities such as decision-making, supervision and production. For example, what are the effects of using one language rather than the other(s) in decision making, supervision and production? Does the language used have an impact on the effectiveness of those organizational activities? Second, an attempt should be made to assess the impact of language choice and language use on organizational information flow. For instance, does formal information flow effectively when one language rather than the other(s) is used? Which language is effectively used on the grapevine? How does management effectively manage the information disseminated in different languages? How does the varying linguistic competence in the official language(s) affect the dissemination and storage of information? Third, does the choice of one language rather than the other(s) have an 97 impact on the speaker's communicative credibility? Finally, as this study only focused on the middle level managers, it would be viable to have a study that could include all the organizational members (subordinates, peers and superiors). Such study for instance would attempt to find out what determines the language members of each organizational level choose in a given interaction. The topics suggested in this section and many others if addressed could contribute to the knowledge needed before bold language policies in multilingual organizations like the Kenyan public sector are made. Conclusions All the hypothesized relationships in this study were not fully confirmed. The study failed to find evidence that each variable studied wholly determines language choice on its own. However, as eminent in the preceding discussion, it seems that each variable has an impact on language use behavior, and probably, the variables interact rather than a single variable per se being the sole determinant of language choice in a given interaction. There is no much variation in the independent variables. But the results do show that there are clear unmarked choices for interaction with different groups. However, three conclusions can be drawn in this study. First, middle level managers with similar social identies, for example, educational level, language competence and occupational status tend to make similar 98 language choices. Thus, it is clear that a pattern of language choice and language use exists in the Kenyan public sector when the middle level managers interact with their superiors, peers and subordinates. The study, for instance, shows that middle level managers mostly speak English alone or in combination with Kiswahili when interacting with their superiors and peers. On the other hand, they mostly use Kiswahili alone or in combination with English when interacting with their subordinates. These patterns form a continuum - English alone, English/Kiswahili and Kiswahili alone. Second, both English and Kiswahili are invaluable media of communication in the Kenyan public sector. It seems, none of them can be used singly without the other. However, although primarily one should have the knowledge and ability to speak each of the two lingua francas fluently, it is equally vital to know which one to use when interacting with the various groups in the multilingual public organization. While English alone might be more appropriate when interacting with the superiors and peers, it however does not seem to be a very appropriate choice when interacting with most of the relatively uneducated subordinates. Kiswahili is not only effective when speaking to the subordinates, but also to most members of the public. Further, although first languages are organizationally inferior to English and Kiswahili, they are important medium of communication especially when interacting with the 99 subordinates and members of the public. Third, for effective communication to take place in the multilingual public organization, all public officers should be highly competent in both English and Kiswahili. However, a high degree of competence in the two lingua francas cannot be assured if the Kenyan education system continues to emphasise the vitality of English as a symbol of high socio— economic achievement, while trivializing the role Kiswahili plays in national development and integration. This calls for an effective national language policy that should elevate the status of Kiswahili, so that both lingua francas can be used as equal tools of communication in all aspects of national development. Further, given that some middle level managers may not be fully competent in Kiswahili as in English, there is a need for the Kenyan government to emphasise the mastery of Kiswahili by all public officers. This can be done by equally rewarding those who are competent in Kiswahili as their English competent counterparts, and also ascertaining that the teaching of Kiswahili is an integral part of the curriculum of all government training institutions like KIA. However, as Kiswahili has become a compulsory subject in primary schools, we would reasonably predict that the overall use of Kiswahili will increase as more people acquire a high degree of competence in it. In closing, it should be noted that the three distinct languages (English, Kiswahili and first languages) used in 100 the Kenyan public sector have complimentary functions. Therefore, as the primary objective of communication in an organization is the dissemination of pertinent organizational information, we would reasonably say that the role of each language is crucial in achieving this end. Thus, from a communication standpoint, choosing the most appropriate language when interacting with a particular addressee is vital if the intended message is effectively disseminated and received. The choice the speaker makes affects the overall communication with other organizational members. APPENDICES APPENDIX I APPENDIX I Courses a; KIA during the Study (June-JulyL 1985) I I COURSE I Duration ofI No. of I the Course | Participants' I I I I 1. Social Work- No. 16/85. I 2 Years I 24 l I 2. Probation Officers— No. 10/85.I 2 Years I 25 I I 3. Probation Officers— No. 9/83. I 2 Years | 18 I I 4. Certified Public Accountants I 1 Year I 29 III - Commercial. I I | I 5. Certified Public Accountants I 1 Year I 26 II — Commercial. I I l I 6. Certified Public Accountants I 1 Year I 25 II - Graduate. I I I I 7. Certified Public Accountants I 1 Year | 27 II - Central Government. | I I I 8. Certified Public Accountants I 1 Year I 26 II - Local Government. | I I I 9. Certified Public Secretaries I 1 Year | 27 II - Central Government. I I I I 10.Certified Public Secretaries I 1 Year I 26 II - Local Government | I I I 11.Diploma in Supplies ManagementI 1 Year I 24 l I 12.Advanced Public Administration] 6 Months I 25 I I 13.Basic French I 3 Months | 22 I I 14.1ntermediate French I 3 Months I 16 | l 15.Advanced French I 3 Months I 16 I I 16.Project Development ManagementI 3 Months I 30 I | 17.Secretarial Management | 2 Months I 22 Training I I I I 18.District Focus for Rural I 1 Month I 100 l I I | Development 101 APPENDIX II 10. ll. 12. 13. 14. 15. 16. Appendix _I First Languages Spoken py the Respondents Dholuo Kikuyu Giriama Ekegusii Luhya Kalenjin (Nandi, Kipsigis, Keiyo etc.) Kamba Kiswahili Meru Dawida (Taita) Digo Teso Somali Kuria Borana Gujarati (An Indian Language) 102 APPENDIX III Appendix III MICHIGAN STATE UNIVERSITY College pf Communication Arts and Sciences Department pf Communication Questionnaire We are doing a study on language use at work, and we would like to ask you a few questions. Be assured that your responses to the questions will be strictly confidential, and will be used only for statistical purposes. You are not required to sign your name. Most of the questions can be answered by circling the most appropriate answer, a few questions require brief written answers. Please, use a pap to answer all questions. Your participation and cooperation are highly appreciated. 1. Where do you work? (Circle one). 1. Ministry Headquarters. 2. Provincial Offices. 3. District Offices. 4 Others (please specify) 2. What is your job group category? (Circle one). Job group E or below. Job group F, G or H. Job group J, K, L or M. Job group N or above. Others (please specify) UIIFQNH 3. What is your job description? (For example, Personnel officer, social worker, district officer, etc.) 4. How long have you been working? years. 5. What is your sex? (Circle one). 1. Male 2. Female 6. What is your age? years. 7. What is your first language (mother tongue)? 8. On the average, how often do you speak English at work? More than once a day. Once a day. Once or twice a week. Once or twice a month. Never speak it. (DIP-(ONH 103 10. ll. 12. 13. 14. 15. On the average, how often do you speak Kiswahili at work? (11133me More than once a day. Once a day. Once or twice a week. Once or twice a month. Never speak it. On the average, how often do you speak your first language (mother tongue) at whork? More than once a day. Once a day. Once or twice a week. Once or twice a month. Never speak it. O'IIhODNI—A What is your educational level? 1. Primary school. 2. Secondary school (0' level). 3 High School (A' level). 4 College/University. 0 what educational level did you learn Kiswahili? Primary school. Secondary school (O'level). High school (A' level). College/University Never learned it in school. (ii-38001014»; Which language(s) do you speak when sending official messages to your immediate supervisor? English only. Kiswahili only. First language (mother tongue) only. Both English and Kiswahili. Both Kiswahili and first language (mother tongue). (J‘fiCDlOi—t Which language(s) do you speak when sending official messages to other officers at your level? English only. Kiswahili only. First language (mother tongue) only. Both English and Kiswahili. Both Kiswahili and first language (mother tongue). 0:4:me Which language(s) do you speak when sending official messages to subordinates (eg. drivers, office messengers, etc.)? English only. Kiswahili only. First language (mother tongue) only. Both English and Kiswahili. Both Kiswahili and first language (mother tongue). (J'luhCONI-A 104 16. 17. 18. 19. 20. Pleas Which language(s) do the subordinates (e.g. drivers, office messengers, etc.) speak when reporting to you? English only. Kiswahili only. First language (mother tongue) only. Both English and Kiswahili. Both Kiswahili and first language (mother tongue). (”#00401-4 Do you sometimes speak your first language (mother tongue) at work? 1. yes. 2. No. If Yes continue to question 18. If Np go to question 19. When do you mostly speak your first language (mother tongue) at work? 1. When speaking to officers of your rank from your ethnic group. 2. When speaking to subordinates from your ethnic group. 3. When speaking to senior officers from your ethnic group. 4. When speaking to any officer (senior, subordinate or peer) from your ethnic group. How would you rate your knowledge and ability to speak English at work? 1. Very good. 2. Good. 3. Adequate. 4. Not good. 5. Very poor. How would you rate your knowledge and ability to speak Kiswahili at work? 1. Very good. 2. Good. 3. Adequate. 4. Not good. 5. Very poor. e read each of the following statements carefully, and by circling one of the numbers, indicate whether you Strongly Agree (SA), Agree (A), Do Not Know (DK), Disagree (D) Strongly Disagree (SD). 105 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. (SA) (A) (DK) (D) (SD) Kiswahili is as foreign to most Kenyans as English. 5 4 3 2 1 A person who speaks Kiswahili fluently is generally either uneducated or is of low status. 5 4 3 2 1 Kiswahili lacks the vocabulary needed in many work situations. 5 4 3 2 1 It is more difficult to learn and speak Kiswahili than English. 5 4 3 2 1 English is the only language that should be used in offices. 5 4 3 2 1 All office workers should be encouraged to use English at work. 5 4 3 2 1 Only effective use of English can contribute to an officer's career development. 5 4 3 2 1 Good command of English ensures respect at work. 5 4 3 2 1 Which language(s) do you think is the most effective in communicating governmental policies, rules and regulations to the subordinate officers? (Circle one). English only. Kiswahili only. First language (mother tongue) only. Both English and Kiswahili. Both Kiswahili and first language (mother tongue). O‘IhQNI-J Which language(s) do you think is the most effective in communicating governmental policies, rules and regulations to the members of the public (Wananchi)? (Circle one). English only. Kiswahili only. First language (mother tongue) only. Both English and Kiswahili. Both Kiswahili and first language (mother tongue). 01I§QNH 106 BIBL I OGRAPHY BIBLIOGRAPHY Alexadre, P. (1972). fig Introduction pp Languages and Language 1p Africa. London: Heinemann. Asre, G. (1977). Four rationalisations for maintaining the European Languages in Education in Africa. Kiswahili: Journal a; the Institute pf Kiswahili Research, 47, 55- 610 Bell, Roger T. (1976). Sociolinguistics: Goalsy Approaches and Problems. London: B.T. Batsford Ltd. Berger, Charles R. (1979). Self Consciousness and the study of interpersonal interaction: Approaches and Issues. In Giles, H.; W. P. Robinson, & P. M. Smith (Eds.), Language: Social psychological perspectives. Oxford: Pergamon Press. Bourhis, R. Y., & F. Genesee (1979). Evaluative Reactions to code Switching Strategies in Montreal. In Giles, H.; W. P. Robinson, & P.M. Smith (Eds.), Language: Social psychological perspectives (335-343). Oxford: Pergamon Press. Bujra, J. (1974). Pumwani: Language use in an Urban Muslim Community: In Whiteley, W. H. (Ed.), Language pp Kenya. Nairobi: Oxford University Press. Daily Nation (1979), 20 September. Daily Nation (1985), 2 October. Datta, A. (1982). Sociolinguistic Behaviour and Social Change: 5a illustrated py the Swahili and Bengali Speech Communities ; A Comparative Study. Calcutta: Indian publications. Diah - Wilson, J. (1984). Hotuba ya Shukrani Kwa niaba ya Maafisa wa Utawala Kwenye sherehe ya Kuwakabidhi Maafisa wo Utawala Hotuba za Hadhara. In Bakhressa, S. (Ed.), Hotuba ga Hadhara (Public Speechesl. Unpublished manuscript, KIA. East African High Commission (1955-58). Statistical Abstracts: the Colony and Protectorate pf Kenya. Nairobi: Government Printer. 107 108 Ekwelie, S. A. (1971). Swahili as a Lingua Franca. Geneva 3 Africa, 10, 86-99. Fishman, J. A. (1972). Domains and the Relationship between Micro- and Macrosociolinguistics. In Gumperz, J. J. & D. Hymes (Eds.), Directions 1p Sociolinguistics: The ethnography pf communication (435-453). New York: Holt, Rinehart and Winston Inc. Fishman, J. A. (1977). Language and Ethnicity. In Giles, H. (Ed.), Languagez Ethnicity and Intergroup Relations (15-57). London: Academic Press Inc. (London) Ltd. Gichangi, J. K. (1984). Hotuba ya Prinsipali wa Taasisi Kwenye Sherehe ya Kuwakabidhi Maafisa wa Utawala Hotuba za Hadhara. In Bakhressa, S. (Ed.), Hotuba aa Hadhara (Public speeches). Unpublished manuscript, KIA. Giles, H.; D. M. Taylor; & R. Bourhis (1973). Towards a theory of interpersonal accommodation through language: some Canadian data. Language 1p Society, 2, 117-192. Giles, H., & P. Smith (1979). Accommodation Theory: Optimal Levels of Convergence, In Giles, H., & R. N. St. Clair (Eds.), Language and Social Psychology. Baltimore: University Park Press. Graff, Robert D. (1983). Communication for National Development: Lessons from Experience. Massachusetts: Oelgechlager, Gunn & Hain Publishers, Inc. Gorman, T. P. (1968). Bilingualism in the Education System in Kenya. Comparative Education, 4, 213-221. Gorman, T. P. (1974a). The Development of Language policy in Kenya with particular reference to the Education System. In Whiteley, W. H. (Ed.), Language la Kenya (397-453). Nairobi: Oxford University Press. Gorman, T. P. (1974b). Patterns of Language use among school children and their parents. In Whiteley, W. H. (Ed.), Language 1p Kenya. Nairobi: Oxford University Press. Gorman, T. P. (1974c). The teaching of Language at secondary school level: some significant problems. In Whiteley, W. H. (Ed.), Language 1p Kenya. Nairobi: Oxford University Press. Gumperz, J. J. (1972). Introduction. In Gumperz, J. J. & D. Hymes (Eds.), Directions la Sociolinguistics (1—31). New York: Holt, Rinehart and Winston, Inc. 109 Halliday, M. A. K. (1972). National Language and language planning in a multilingual society. East African Journal, 9, (8), 4—13. Harries, L. (1968). Swahili in Modern East Africa. In Fishman, J. A., et al. (Eds.), Language Problems 9: Developing Nations. New York: John Wiley & Sons, Inc. Harries, L. (1983). The Nationalisation of Swahili in Kenya. In Kennedy, C. (Ed.), Language Planning and Language Education (118—128). London: George Allen & Unwin (Publishers) Ltd. Heine, B. (1970). Status and Use pf African Lingua Francas. Munchen: Weltforum Varlag. Heine, B. (1977). Vertical and Horizontal Communication in Africa. Afrika Spectrum, 12(3), 231—238. Heine, B. (1980). Language and Society. In Heine, B., & W. J. G. Mohlig (Eds.), Language and Dialect Atlas _§ Kenya, Volume 1 (61—78). Berlin: Dietrich Reimer Verlag. Herman, S. R. (1968). Explorations in the Social Psychology of Language Choice. In Fishman, J. A. (Ed.), Readings 1p ppa Sociology 93 Language (492—511). The Hague: Mounton. Hofer, B. (1978, April). Toward a Continental African Language? Swiss Review a; World Affairs, 25-29. Hunt, C. L. (1966). Language choice in a multilingual society. Sociological Inquiry, 36, 240—253. Hyden, G. (1972). Social Structure, Bureaucracy and Development Administration in Kenya. African Review, 1(1), 118-129. Hymes, A. (1967). Models of the interaction of language and social setting. Journal a; Social Issues, 23, 8-28. Indakwa, J. (1978). A 'Lingua Franca' for Africa. Kiswahili: Journal a; the institute a; Kiwahili Research, 48, 57— 73. Itebete, P. A. N. (1974). Language Standardization in Western Kenya: the Luluyia Experiment. In Whiteley, W. H. (Ed.), Language 1p Kenya (87-114). Nairobi: Oxford University Press. Itebete, P. A. N. (1976). Some Thoughts on Reversion to an Indigenous Language as the Official Language of a Nation: The Swahili case in Kenya. The KIA Journal. 110 Jahadhmy, A. A. (1975). Anthology pf Swahili Poetry. Nairobi: Heinemann. Janice, S. M. (1971). The Influence pf Tribalism) Lingpa Francaz and Mass Communication pp National Development 9: East African States pf KenyaI Eganda and Tanzania. Unpublished Ph.D. Dissertation, The University of ' Texas, Austin. Karigithe, S. M. (1981). Hotuba ya mkuu wa Idara ya Lugha na Mawasiliano kwenye sherehe ya kuwakabidhi Maafisa Utawala Hotuba za Hadhara. In Bakhressa, S. (Ed.) Hotuba aa Hadhara (public speeches). Unpublished manuscript, KIA. Karigithe, S. M. 1982). Hotuba ya mkuu wa Idara ya Lugha na Mawasiliano Kwenye sherehe ya Kuwakabidhi Maafisa wa Utawala Hotuba za Hadhara. In Bakhressa, S. (Ed.), Hotuba aa Hadhara (public speeches): Unpublished Manuscript, KIA. Khalid, A. (1977). The Liberation g: Swahili from European Appropriation. Nairobi: East African Literature Bureau. Kihore, Y. M. (1976). Tanzania's Language policy and Kiswahili's Historical Background. Kiswahili: Journal a; the Institute a: Kiswahili Research; 46, 47-66. Knappert, J. (1965). Language problems of the New Nations of Africa. Africa.Quarterly, 5, (2), 95-104. Ladeforged, P.; R. Glick & C. Criper (Eds.) (1972). Language 1p Uganda. London: Oxford University Press. Le Page, R. B. (1964). The National Language Question: Linguistic Problems pf Newly Independent States. London: Oxford University Press. Mackey, W. F. (1965). Bilingual interference: its analysis and measurement. Journal a; Communication, 15, 239-249. Mackey, W. F. (1966). Measurement of Bilingual behavior. The Canadian Psychologist, 7a, (2), 75-92. Mazrui, A. A. (1969). Social Engineering and Language policy in East Africa. Proceedinga a: the Seventh Symposium, the East African Academy, Makerere University, Kampala, 210—228. Mazrui, A. A. (1972). Cultural Engineering and Nation Building 1p East Africa. Evanston, Illinois: Northwestern University Press. 111 Mazrui, A. A. (1975a). The political sociology p; the English Language: 5p African Perspective. The Hague: Mounton & Co. Mazrui, A. A. (1975b). Racial Boundaries of the English Language. In Savard, J. & R. Vigneault (Eds.), Multilingual Political Systems and Solutions. Quebec: Les press De' L' Universite' Laval. Mkangi, K. G. C. (1982, April) The political ; Economy 9; Kiswahili: a Kenya: Tanzania Comparison. paper presented at the Swahili Language Workshop, School of Oriental and African Studies, University of London. Mkilifi, M. H. A. (1978). Triglossia and Swahili - English Bilingualism in Tanzania. In Fishman, J. A. (Ed.), Advances la the Study a: Societal Multilingualism (129- 149). The Hague: Mounton Publishers. Mutahi, Karega (1983) (Review a: Swahili ; The Diagram a: Crises by Ohly, Rajmund - 1982). Journal 9: African Languages and Linguistics, 5, 212-215. Mwangi, D. M. (1981). Hotuba ya Shukrani Kwa niaba ya Maafisa wa Utawala Kwenye sherehe ya Kuwakabidhi Maafisa wa Utawala Hotuba za Hadhara. In Bakhressa, 8. (Ed.), Hotuba aa Hadhara (Public Speeches). Unpublished manuscript, KIA. Ngara, E. A. (1982). Bilingualism) Language contact and Language Planning. Gwelo: Mambo Press. Nichols, P. C. (1984). Networks and Hierarchies: Language social stratification. In Kramarae, C., M. Schulz & W. M. O'Barr (Eds.) Language and Power. Beverly Hills, California: Sage Publications. Nyamu, H. J. (1980). Aspects pf Kenya's Development: a participant's view. Nairobi: East African publishing House. Ohly, R. (1974). The Influence of Language on Manpower Resources. Kiswahili: Journal a; the Institute a: Kiswahili Research, 44, 54-58. Ohly, R. (1982). Swahili : The Diagram pf Crises. Dar—es- Salaam: Afro-pub. Distribution in Tanzania by University Press. Parkin, D. J. (1974a). Status Factors in Language Adding: Bahati Housing Estate in Nairobi. In Whiteley, W. H. (Ed.), Language 1p Kenya (147—165). Nairobi: Oxford University Press. 112 Parkin, D. J. (1974b). Language Shift and Ethnicity in Nairobi: the speech community of Kaloleni. In Whiteley, W. H. (Ed.), Language la Kenya (167-187). Nairobi: Oxford University Press. Parkin, D. J. (1974c). Nairobi: Problems and Methods. In Whiteley, W. H. (Ed.), Language 1p Kenya (131-145). Nairobi: Oxford University Press. Parkin, D. J. (1974d). Language Switching in Nairobi. In Whiteley, W. H. (Ed.), Language 1p Kenya. Nairobi: Oxford University Press. Parkin, D. J. (1977). Emergent and Stabilized multilingualism: polyethnic peer group in Urban Kenya. In Giles, H. (Ed.), Language, Ethnicity and Intergroup Relations (185—210). London: Academic Press (London) Ltd. Penalosa, F. (1981). Introduction pp the Sociology pf Language. Rowley, Massachussetts: Newbury House publishers, Inc. Platt, J. (1979). The Lingue Franche of Singapore: An investigation into strategies of Inter-ethnic Communication. In Giles, H.; W. P. Robinson, & P. M. Smith (Eds.), Language: Social Psycholagical Perspectives (171-177). Oxford: Pergamon Press. Prator, C. & Whiteley, W. H. (1967). Survey of Language use and Language teaching in Eastern Africa. African Language Review, 6, 159-167. Republic of Kenya (1966). Code 9: Regulations. Nairobi: Government Printer. Republic of Kenya (1972). Educational Structure 1p Kenya. Nairobi: Jomo Kenyatta Foundation. Republic of Kenya (1974). Kenya National Assembly Official Report) Vol. XXXV. Nairobi: Government Printer. Republic of Kenya (1975-77). Kenya Statistics Division: Statistical Abstracts. Nairobi: Government Printer. Republic of Kenya (1979). The Constitution _3 Kenya. Nairobi: Government Printer. Republic of Kenya (1985). Civil Service Review Committee 1985: New Terms and Conditions pf Service. Nairobi: Government Printer. 113 Rhoades, J. (1977). Linguistic Diversity and Language Belief 1p Kenya. Ph.D. dissertation, Maxwell School of Citizenship and Public Affairs, Syracuse University. Saville—Troike, M. (1982). The Ethnography a: Communication. Oxford: Basil Blackwell: Scotton, C. M. (1972). Choosing a Lingua Franca 1p ap African Capital. Edmonton, Champaign: Linguistic Research, Inc. Scotton, C. M. (1976a). Strategies of Neutrality: Language choice in Uncertain Situations. Language, 52, 919—41. Scotton, C. M. (1976b). The Role of Norms and Other Factors in Language Choice in Work Situations in Three African Cities (Lagos, Kampala, Nairobi). In Kjolseth, R., & Verdoodt, A. (Eds.), Language and Sociology (201—231). Louvain: Editions Peters. Scotton, C. M., & W. Ury (1977). Bilingual Strategies: The Social Functions of Code—Switching. Internatioal Journal pf the Sociology a: Language, 13, 5-20. Scotton, C. M. (1978). Language in East Africa: Linguistic Patterns and political ideologies. In Fishman, J. A. (Ed.). Advances la paa Study a; Societal Multilingualism (719-759). The Hague: mounton publishers. Scotton, C. M. (1979). Explaining Linguistic Choices as indentity Negotiations. In Giles, H.; W. P. Robinson, & P. M. Smith (Eds.), Language: Social Psychological Perspectives (359-366). Oxford: Pergamon Press. Scotton, C. M. (1982a). The possibility of Code-Switching: motivation for maintaining multilingualism. Anthropological Linguistics, 24, 432-444. Scotton, C. M. (1982b). Learning Lingua Francas and Socioeconomic integration: Evidence from Africa. In Cooper, R. L. (Ed.), Language Spread (63-95). Bloomington: Indiana University Press. Scotton, C. M. (1982c). Language Use in Kenya: an Urban- rural Comparison of the Luyia. International Journal a: the Sociology a; Language, 34, 121-136. Scotton, C. M. (1983). The negotiation of identities in Conversation: a theory of markedness and code choice. International Journal pf the Sociology a; Language. 44, 115—136. 114 Scotton, C. M. (1985). What the heck, Sir: Style Shifting and Lexical colouring as features of a powerful language. In Street, R. L., & Cappel, J. N. (Eds.), Sequence and Pattern 1p Communicative Behaviour. (107- 117). Scotton, C. M. (Forthcoming). Code Switching as Indexical of Social Negotiations. In Monica, H. (Ed.), Code Switching; Anthropological and Linguistic Aspects. Cambridge University Press. Secord, P. K. & Backman, C. W. (1964). Social psychology, 2nd ed. New York: McGraw-Hill Book Company. Shaw, M. E., & Constanzo, P. R. (1970). Theories pf Social Psychology. New York: McGraw-Hill Book Company. Sherif, M. (1936). Psychology pf Social Norms. New York: Harper & Brothers Publishers. Sherif, M. (1948). fig Outline pf Social Psychology. New York: Harper & Brothers Publishers. Sherif, M.; & Sherif, C. W. (1969). Social Psychology. New York: Harper & Brothers Publishers. Slawski, C. J., (1981). Social Psychological Theories: é Compative Handbook for Students. Glenview, Illinois: Scott, Foresman and Company. Soba, Omar (1984). Hotuba ya Mgeni wa Heshima Kwenye sherehe ya kuwakabidhi Maafisa wa Utawala Hotuba za Hadhara. In Bakhressa, S. (Ed.), Hotuba aa Hadhara (Public Speeches). Unpublished manuscript, KIA. Steere, E., & Madan, A. C. (1890). A handbook a: the Swahili Language aa Spoken a: Zanzibar, London: Society for promoting Christian Knowledge. Taylor, D. M., & Royer, L. (1979). Group processes Affecting Anticipated Language Choice in Intergroup Relations. In Giles, H.; W. P. Robinson, & P. M. Smith (Eds.), Language: Social Psychological Perspectives (185-191). Oxford: Pergamon Press. The Weekly Review. (1985). 22 November, Nairobi: The Weekly Review Ltd. Thibault, J. W.; & Kelley, H. H. (1959). The Social Psychology 9; Groups. New York: John Wiley & Sons. Toweett, T. (1979). Study g; Kalenjin Linguistics. Nairobi: Kenya Literature Bureau. 115 Wallwork, J. F. (1978). Language and People. London: Heinemann Educational Books. Whiteley, W. H. (1969). Swahili: The rise a; a National Language. London: Menthuen & Co. Ltd. Whiteley, W. H. (1971). Some Factors influencing Language policies in Eastern Africa. In Rubin, J. & Jernudd, B. H. (Eds.), gap Languages pa Planned? (141-158). The University Press of Hawaii. Whiteley, W. H. (1974a). The classification and distribution of Kenyan African Languages. In Whiteley, W. H. (Ed.), Languages 1p Kenya (13-61). Nairobi: Oxford University Press. Whiteley, W. H. (1974b). Some patterns of Language Use in the Rural Areas of Kenya. In Whiteley, W. H. (Ed.), Language 1p Kenya (319-347). Nairobi: Oxford University Press. Whiteley, W. H. (1979). Sociolinguistic Surveys at the National Level. In Pride, J. B. (Ed.), Sociolinguistic Aspects a; Language Learning and Teaching (44-56). London: Oxford University Press. Williams, F. (1979). Reasoning with Statistics, 2nd ed. New York: Holt, Rinehart and Winston. IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII