'9 N _w!‘!

013:

 

.. a. :n .s .. .u,
“Eli. .u. I... -2 .. .15.», .1}:
5... 3’ a §1IAV§I ;.lr
.7luflvf

3.11.313:
4 v ilxll ‘

gram”.

:2
.u .3 .1! :u

$1.. v 1 .

 

. wanna fl
3...???
2.213229%}? ‘

1.3.11. .. 5?}:
5

"My:

.5 . .Q
m».
, .

naiihw. ,

e .1 \O

4.33. , yfluiixfi. 5. .

IL; 6.5.13, .3.

x. .raxxa.... .M.?u.z..«fl.r~
.

4:: at: I, y
. . g

.5? 1.
adiihrixma... r3.
. . 5?. ..

.... ..

 

'3‘ :5'.
'1 xv. 1. ‘bg:
f5!

 

V C3; . . . .
.0! i... . L.
l; . . 4-1.5:
. ‘ 33.31.

v ‘ .Vy in...

.1:

 

 

 

 

 

 

 

:1621rlil
ul: :5"... A

L 31.. .‘
45er {.4
o... 2......

2,1 ....

“53’s LIBRARY
3 Michigan State
6113 91.59 University

This is to certify that the
dissertation entitled

SHENG AND VARIATION: THE CONSTRUCTION AND
NEGOTIATION OF MULTIPLE IDENTITIES

presented by

PETER GITHINJI

has been accepted towards fulfillment
of the requirements for the

PhD degree in LINGUISTICS

 

 

 

 

Major Professor’s Signature

2005

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

— -..-.-._.---.-.-.-.-.--.----—

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

of; Int12 280.3

 

0519‘? WI

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/CIRCIDateDuo.indd-p.1

SHENG AND VARIATION: THE CONSTRUCTION AND NEGOTIATION OF
MULTIPLE IDENTITIES

By

Peter Githinji

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Linguistics and Germanic, Slavic, Asian and Afi‘ican Languages

2006

ABSTRACT

SHENG AND VARIATION: THE CONSTRUCTION AND NEGOTIATION OF
MULTIPLE IDENTITIES

By
Peter Githinji

This dissertation addresses the issue of how variation in Sheng — a contact linguistic
code spoken in Nairobi and other urban areas in Kenya — represents the different
identities of the speakers of those different varieties. My chief goal is to underscore the
relationship between language and identity by capitalizing on Sheng’s high lexical
variation and the implication of different symbolisms attached to competing lexical items
by speakers from different backgrounds. While structural variation is relevant to this
study, my major focus is the lexicon, where I demonstrate how social categories adopt
specific lexical items as part of a norm that marks them as members of a distinct social
category. I account for Sheng’s variation by illustrating the interaction between different
word formation processes and people’s social realities.

The guiding logic in this study is that individual and groups exploit lexical
variation to negotiate, confirm, contest, activate or modify existing identities. If words are
the symbolic markers that signal a given identity, then the group’s preference for certain
lexical items and opposition to the use of others follows from the deliberate desire to
either project desired identities or to reject contested identities. This study recognizes the
‘baze,’ a local hang-out joint, as the minimal social unit that exhibits a high degree of
cohesion and through which the most dynamic interaction practices can be observed. I
argue that while different bazes have unique linguistic markers that distinguish them from

other groups, they are very fluid and vulnerable to outside influence. The movement of

members in and out of their bazes during inter-group interaction results in the adoption of
linguistic markers identified with the haze in question. In the process, members gain
multiple identities, which are activated when the necessary interaction conditions demand
it. While language is the main focus of this dissertation, attempts are made to show how
linguistic practices act in tandem with other discursive practices in enforcing group

norms and hence enhancing group identity.

DEDICATION

This dissertation is dedicated to the late Director of Starehe Boys Center
Dr. Geoffrey William Griffin 1933-2005
an educationist who raised eagles to aim and fly high
creating wholesome individuals from characters destined to perish.
to whom I owe a great deal of my academic achievements.

iv

ACKNOWLEDGEMENTS

The dissertation is the culmination of so many years of schoolwork, right from nursery
school, through primary school, secondary school, and university. As I celebrate the
achievement of my intellectual dream, I would like to acknowledge all who have played a
part in nurturing my academic abilities. From my primary school teachers at Ngorika
Primary school who gave me the academic foundation, the secondary school teachers at
Starehe Boys center who inculcated discipline and character in me and finally my
university professors who sharpened my skills in the conduct of intellectual inquiry. If I
am the finished product, then all these people and others are the producers in the
knowledge industry where I was produced.

This dissertation could not have been possible without the assistance from various
quarters. Thanks to all my professors at Michigan State University for their instructions
and advice, my fellow students for their enthusiasm and support, and all administrative
staff in the linguistics department, beginning with the chair, David Prestel, Julie Delgado,
the graduate secretary, and other administrative secretaries Cathy Fields, Jennifer Nelson
and Becky Rhodes. I cannot forget all my former Swahili students whose interest in
Swahili guaranteed me the teaching assistantship that covered my tuition and upkeep.

Special thanks go to my guidance committee for all the support and comments
during my comprehensive exams which shaped the way I looked at issued discussed in
this study. I gained a lot from Dennis Preston the co-chair of the committee, not only in
terms of intellectual guidance, but also professional development. Deo Ngonyani, my

other co—chair was magnanimous. Not only do I owe him my teaching skills, but he also

initiated me to scholarly publication though co-authored work. I also acknowledge
Professors Grover Hudson and David Dwyer for their constructive comments and
suggestions that have informed the structural description and ethnographic account of
vital arguments presented in this dissertation.

Thanks also to all my research assistants, the principals, headmasters and teachers
who allowed me to conduct research in their schools and the support they gave me to
ensure that I succeeded in my mission. I would not forget all the school students who
voluntarily agreed to participate in this study, and the mebeshte, mababi, wanati na pia
mafala. Without them, there could have been no data for this research. Finally, I
acknowledge Professor Kimani Njogu and the staff at Twaweza Communication for their
assistance during the collection of data. I cannot forget my host Kimani, my sister
Wambui, and nieces Nyambura and Gathoni, for making my stay in Nairobi enjoyable.

To you all, thanks for believing in me.

vi

Table of Contents

LIST OF TABLES .................................................................... xii
LIST OF FIGURES .................................................................. xiii
CHAPTER 1

SETTING THE STAGE: UNDERSTANDING THE CONTEXT OF THE
STUDY ................................................................................... l
1.0. Introduction ......................................................................... 1
1.1. Making choices ..................................................................... 3
1.2. The consequences of choice: Sheng and the Kenyan challenge .............. 5
1.3. Guiding questions .................................................................. 8
1.4. Hypotheses ........................................................................... 9
1.5. Major assumptions ................................................................... 9
1.6. Definition of terms ................................................................ 10
1.7. Plan of work ........................................................................ 13
CHAPTER 2

SHENG’S SOCIO-DEMOGRAPHIC BACKGROUND AND

RESEARCH METHODOLOGY .................................................... 15
Part I. Linguistic and demographic backgrounds .................................. 15
2.0. Introduction ........................................................................ 15
2.1. Languages of Kenya .............................................................. 16
2.2. Multilingualism in Kenya ........................................................ 18
2.3. Socio-demographic description of Nairobi .................................... 20
2.4. Demographic facts: A history of residential disparity ........................ 22
2.5. Informal settlements ............................................................... 25
2.6. The Language Policy in Kenya ................................................... 28

vii

2.6.1. After independence ...................................................... 29

2.6.2. Sheng as linguistic capital in an alternative linguistic market... . .36

Part 2. Literature review and Methodology ........................................... 40
2.7. Brief outline ......................................................................... 40
2.7.1. What is Sheng? ............................................................................. 40
2.7.2. Categorization of Sheng within variation ............................ 42
2.7.3. The politics of Identity: the definition ................................ 48
2.7.4. The relationship between language and identity .................... 52
2.7.5. Identity in previous studies on Sheng ................................. 56
2.8. Methodology ........................................................................ 59
2.8.1. Data collection, instruments and interview procedure ............. 60
2.8.2. Data interpretation ...................................................... 65
CHAPTER 3
A DIAGRAM OF VARIATION: EXPLORING FORM AND NATURE
OF VARIATION IN SHENG ......................................................... 67
3. Linguistic variation and social reality ............................................. 67
3.1. Sheng’s structure: two varieties or one? ..................................................... 68
3.2. Drawing from Swahili grammatical patterns ................................... 71
3.3. Word formation processes in Sheng ............................................. 74
3.4. Synonymy: some of the lexical differences .................................... 85

3.5. Morphophonological variation and its connection to speakers’ identity. . .87
3.6. Some background on the users of these synonyms ............................ 90

3.7. Dominance of certain words ...................................................... 91

viii

3.7.1. The gender factor in variation ........................................ 92

3.7.2. The schools and variation ............................................. 93
3.7.3. The age and neighborhood factor in variation ...................... 95
3.7.4. Who are the innovators? .............................................................. 97
3.8. From the estates to the bazes ................................................... 100
3.8.1. The Bazes .............................................................. 102
3.8.2 .The relevance of bazes in lexical variation ........................ 108
3.9. Other factors responsible for variation in Sheng ............................ 112
3.9.1. Socioeconomic factors: The south C baze ........................ 113
3.9.2. Residential factors .................................................... 115
3.9.3. Relevance of occupation in Shengs variation ..................... 116
3.9.4. Ethnicity as a factor .................................................. 118
CHAPTER 4
LEXICAL CHOICES, DISCOURSE AND OTHER DISCURSIVE
PRACTICES IN THE COPNSTRUCTION OF IDENTITY ................... 122
4. Introduction ......................................................................... 122
4.1. Lexical items as linguistic signs and symbols ................................ 123
4.1.1. Ambiguity, bazes and identity ....................................... 124
4.1.2. Age determined interaction .......................................... 126
4.1.3. Signifiers of identity: frozen in oldskool ........................... 129
4.1.4. Gender signifiers: revisiting variation in gender terms ............ 134

4.1.5. Definition by the other: derogatory gender metaphors. . . . . . . . ....137

4.1.6. Rejecting negative identity by contesting derogatory
gender labels ............................................................ 140

4.1.7. Positive self-defining signifiers ...................................... 144

ix

4.2. Inter-group categorization: comparative labels .............................. 145

4.2.1. Socioeconomic symbolism: The wanati vs. the mababi

dualism ................................................................. 146
4.2.1.1. Babi: a person from Ubabini ...................................... 149
4.2.1.2. Babi: People in white-collar employment .............. 149
4.2.1.3. Babi: Mom’s boy .......................................... 150

4.2.1.4. Babi: Just anyone you aspire to be economically. . ....150

4.2.1.5. Babi: One who speaks inferior Sheng ................... 152

4.2.2. The cosmopolitan symbolism: Beshte vs. fala distinction ........ 157

4.3. Performance and negotiation of multiple identities ........................... 162
4.3.1. Non-conscious activation of layered identities ..................... 162

4.3.2. Negotiating multiple identities by accommodating to

other’s linguistic behavior ............................................. 169

4.4.1. Vernacular discourse: A different perspective on local reality ........... 181
4.4.2. Michongoano: Performance of identity in verbal contest ................. 191
4.4.2.1. Mchongoano within the general discourse of verbal art ....... 194
4.4.2.2 Structure of michongoano .......................................... 198
4.4.2.3. The art of kuchongoana: Playing by the rules ................... 201
4.4.2.4. Michongoano as face threatening acts ............................ 205

4.4.2.5. Affirming identity: Michongoano as an in-group activity. .....206

4.4.2.6. The language of delivery ........................................... 209
4.5. Other ways of constructing identities .......................................... 210
4.5.1. Diet and dining habits ................................................. 21 1
4.5.2. Dressing ................................................................. 213

4.5.3. Style of music .......................................................... 215

4.6. Construction of identity in other discourse sites .............................. 216
4.6.1. Popular culture: Sheng hip-hop ....................................... 216
4.6.2. The media ............................................................... 220
4.6.3. The intemet ............................................................. 224
4.7. Attitudes and value judgments ................................................... 226
4.7.1. Respondent’s subjective judgments: Relative comparisons
and identity ............................................................ 226
4.7.2. The question of proximity ............................................. 235
4.8. Summary ............................................................................ 236
CHAPTER 5
CONCLUDING REMARKS .......................................................... 238
Conclusions .............................................................................. 238

Apppendix A. Questionnaire 1: For school children under 18 years old. . . . . ....247

Appendix B. Questionnaire 2: For Subjects over 18 years ........................ 248
Appendix C. Rifi Valley Academy (RVA) Lingo .................................. 250
Appendix D. Synonym frequencies ................................................... 251
Bibliography .............................................................................. 247

xi

 

 

LIST OF TABLES

Table 1. Population growth in Nairobi, 1906-1999 ....................................... 22
Table 2. The choice between buda and mbuyu among the students ..................... 94
Table 3. Ages when respondents claim to have started using Sheng .................... 95
Table 4. Demographic composition of the bazes .......................................... 103
Table 5. Frequency of variants for the word ‘boy’ ....................................... 135
Table 6. Frequency of variants for the word ‘girl’ ........................................ 136
Table 7. Names for ‘girl’, ‘breast’ and ‘butt’ .............................................. 142

Table 8. Different bazes ’ comparison of their Sheng with that of other estates ...... 228
Table 9. Respondents comparisons of their Sheng with that of other estates .......... 230

Table 10. Respondents judgments of the variety of Sheng in different estates. . . . . ....232

xii

LIST OF FIGURES

Figure 1. Languages of Kenya ............................................................... 17
Figure 2. Urban expansion ................................................................... 21
Figure 3. Diferent estates in Nairobi ........................................................ 24
Figure 4. Kibera slum ........................................................................ 153
Figure 5. A house at Riverside drive in Westlands ....................................... 154
Figure 6. The reply of a sound is a sound .................................................. 201
Figure 7.0utsiders neither know the safe topics nor the rules in ritual insults ........ 207
Figure 8. Verbal duels are an in-group thing ............................................... 208

xiii

Chapter 1
Setting the Stage: Undertanding the context of the study

1. Introduction

This dissertation addresses the relationship between linguistic variation and identity. I
will discuss how variation in Sheng, a linguistic code spoken in Nairobi and other urban
areas in Kenya represent different shades of the identity of its speakers. I investigate
variation in terms of lexical choices and discourse types, as different speakers engage in
social interaction. Identity will be regarded both as something that one has, as well as
something that can be acquired. I will argue that individuals have and display multiple or
multifaceted identities, which are mediated through language and other social practices,
and are in a constant process of change. Since identity is tied to self-presentation as well
as others’ perception, I will focus on how identity is constructed through performance
and how it is negotiated in social interactions. By negotiation here, I mean both the overt
and covert interactional strategies that individuals use to cope with conflicting realities
that result from their multiple identities. I have in mind strategies such as affirmation, self
and other categorization, self-presentation, as well as contestations, all of which involve
the modification or partial erasure of individuals’ prior identities.

The major issue addressed in this dissertation is how people in small groups in
Nairobi make use of the lexical resources of Sheng in projecting their various identities.
Previous work on Sheng’s lexicon has been confined to a compilation of vocabulary. Yet
looking at Githiora’s (2002) Sheng glossary, or Mbaabu and Nzuga‘s (2004) dictionary,

one encounters many synonyms for certain words. There has not been any explanation of

this synonymy in a systematic way. This study provides an explanation for the synonymy
by tying lexical variation to different realities that shape speakers’ identities.

The majority of variationist works dealing with identity (e.g, Fischer’s (1958)
qualitative study of [13]/[n] variation in a New England Village and Labov’s (1972a)
famous study of glide centralization in Martha’s Vineyard have focused on phonological
variation. Fisher, for instance, showed that the velar nasal [13] was the preferred variant
identified with the ‘model’ boys while the alveolar nasal [n] was identified with the
‘typical’ boys. Labov, on the other hand showed that the respondents who used the raised
onsets of [an] and [ai] were island-oriented regardless of other static sociolinguistic
identities.

Vocabulary has had very little such treatment, and when it has, it is often an
insider relying on knowledge about their own speech community. A good example is
Smitherman’s (1977) study of various aspects of Black semantics. In reference to the
different levels of meanings of some words used by African Americans, Smitherman
remarks;

For blacks, English words can have potentially two levels of meanings, one black,

one white. Since blacks share in the consensus dialect of the American

mainstream, on one level a word’s referent is the same for blacks and whites. But
since blacks share a linguistic subculture outside that mainstream, on another
level (the Black Semantic level) the same word has multiple meanings and

associations. P 59

As an example, the word bad could be interpreted with the the mainstream
negative meaning e.g I got a bad cold, but among the African Americas the word can

have a positive meaning as in He is a bad dude. Another notable study is Githiora’s

(1999) study of the social racial terms and identity in an Afromexican community.

 

 

Githiora discusses how terms like indio, blanco, negro and moreno could be used as
footing devices as speakers align and realign themselves in relation to their interlocutors.
While these two works have made important contributions to the study of lexical
variation, much still remains to be done. Since different societies experience realities
differently, different languages always come up with interesting peculiarities. This study,
therefore, apart from addressing a gap in the literature on vocabulary variation, also
introduces interesting data from a very dynamic urban code. Since one of my major
arguments is that linguistic choices are made in order to achieve specific outcomes, 1 will
open this introductory Chapter with a brief discussion of the factors and motivation

behind the making of choices with relation to identity.

1.1. Making Choices

The choice of linguistic codes by bilinguals or bidialectals is not always, perhaps not
even usually, intended to accomplish communicative ends. When a speaker has an option
of using either of the languages or codes available, then there must be a rationale for
choosing one and not the other, although “rationale” is not meant to suggest that such
choices or their causes always operate at the conscious level (some are made at the
subconscious level). Code selection is usually guided by the need to optimize the
interaction outcomes depending on prevailing circumstances. During codeswitching for
instance, speakers evaluate potential code choices along a markedness scale depending on
the goals of interaction. The marked choices are those that are not predicted by a
community’s interactive norms in a given situation, while the unmarked choices are those

that are predictable in given situational contexts. According to Myers-Scotton’s (1993b)

markedness model (MM), the choices that speakers make are indices of the desired
Rights and Obligations (RO) sets between participants which might either be marked or
unmarked. These (RO) sets are well known by the interlocutors and they vary depending
on different interaction types. Markedness is relative; what is marked in one interaction
type might be unmarked in a different interaction type and vice versa. The argument that
speakers’ code choices are goal-oriented and that the rational actors engage in decision-
making to come up with the most optimal choice has recently been the central argument
of the Rational Choice (RC) model (see Bolonyai 2005, Myers-Scotton and Bolonyai
2001). Expounding on the RC model, Myers-Scotton and Bolonyai write;

The RC theory is based on the assumption about human cognition that actors are

oriented to seek optimality of an interpersonal nature in their actions, including

linguistic choices. The overall assumption is that the way speakers choose to
speak reflects their cognitive calculation to present a specific persona that will
give them the best “return” in their interaction with others, in whatever ways are

important to them and are rationally grounded. (2001 :23)

Optimatization of interactional outcomes in all the seven functions of language1 is
the rationale behind linguistic choices. Without downplaying the outcomes of facilitation
of interpersonal communication associated with the interactional function of language, I
restate that linguistic choices are not entirely intended to serve communicative function
alone. Language also serves other symbolic functions that might even sometimes
overshadow the communicative fimction. In this study, 1 will be dealing with how
speakers of an esoteric linguistic code manipulate lexical variation for the symbolic

purpose of indexing different aspects of their identities. It will be demonstrated that

although lexical choices are used to highlight different identities, they do not always

 

' Halliday (1973) identifies the seven functions of language as instrumental, regulatory,
heuristic, interactional, personal, imaginative and informative,

follow along the RC model. In fact, some lexical items that are subconsciously selected
seem to challenge the very logic of the RC model. This is due to constant performances
that have caused these choices to become engrained in the speakers’ habitus, such that
speakers become disposed to use them in their group interactions, in the process
activating some aspects of their multifaceted identity. This study goes beyond the notion
of “activating a specific persona” and will provide in more detail an account of linguistic
choices, which are consciously and deliberately adopted, as Myers-Scotton and Bolonyai
put it, “in whatever ways, important to them”, as well as the counterargument for
conscious activation of identities through lexical choices.

In this introductory Chapter, I will give the reader an idea of what this study is all
about. This begins with a background of Sheng, the linguistic code and the topic of this
study, from its evolution to the present situation. This will be followed by the statement
of the questions that this study seeks to answer, and then the general hypotheses that
guide my investigation. I will then define some of the terms that the reader might
encounter in the course of the dissertation which 1 think should be known early enough,
before concluding the Chapter with an account of the general organization of this

dissertation. Let us now turn to Sheng.

1.2. The consequences of choice: Sheng and the Kenyan challenge

Sheng, usually regarded as an acronym for “Swahili-English” slang in previous literature
(e.g, Mazrui 1995), is believed to have evolved in Nairobi in the 19605 and 19705. Its
evolution has been attributed to an inadequate knowledge of the standard languages
(Swahili and English), a need for a language of identity (Osinde 1986, Samper 2002,

2004), its ability to obscure meaning as a secret language (Mbugua 2003, Githiora 2002),

and its cognitive efficiency (Kangethe 2004) among other reasons. Though a product of
multilingualism and language contact phenomena, Sheng defies the traditional
pidgin/creole classification of similar contact languages due to the fact that there existed
a widely spoken lingua franca (Swahili) at the time of its evolution. However, if its
phonotactics and morphosyntax are used as the criteria for its classification, then it can
confidently be regarded as a variety of Swahili. I will revisit this issue in a later Chapter.
Originally, Sheng was restricted to the poor residential areas of the Nairobi
Eastlands, and then it gradually spread to other poor residential areas and has today
spread to all areas of Nairobi and beyondz. The initial hostile attitude towards Sheng was
due to its association with wayward youth, drug users, and generally people who engaged
in antisocial behavior. Nowadays, Sheng is not restricted to the youth as was the case
during its formative years. Many young people claim that they speak Sheng with their
parents, though they also add that the kind of Sheng they speak with elders is quite
different from the variety that they speak with their age-mates. As recent as July 2005,
President Kibaki (in his 705) was reported in the Kenya media as calling people to have
fun by using his now-popular idiom mji—enjoy (enjoy yourselves) while in 2003, Justice
Samuel Bosire, a senior judge in Kenya was reported in the media saying poa (OK, fine)
to a witness during a commission of inquiry, a serious public proceeding. While poa is
derived from a Swahili verb ku-poa ‘to cool down’, mji-enjoy is a combination of two
Swahili morphemes; m, a nominal classifier, and ji, a reflexive morpheme attached to the

English verb enjoy. This underscores the pervasive nature of codeswitching in Sheng.

 

2 See appendix (III) for some Sheng words that appear in the Rift Valley Academy
(RVA) Lingo. RVA is an American Christian school located about 50 km from Nairobi
whose student population predominantly comes from overseas.

While isolated cases such as these that surface once in a while provide evidence of adult
use of Sheng, we can not tell with certainty to what extent adults use Sheng since no
conclusive study has yet been done to address the issue.

Sheng has been blamed for students’ poor performance in national exams. Due to
the predominance of English and Swahili (both the official and national languages) in
Sheng’s lexicon, students normally blur the boundaries that demarcate Sheng and these
languages. This is not helped by the fact that the bulk of the Sheng speakers are from
poor residential areas where language learning resources as well as qualified instructors
are lacking. Writing in the Sunday Nation of January 9’“, 2005, Kimani Njogu, a leading
language pedagogue and Chairman of CHAKITA3-Kenya, lamented that primary school
teachers are expected to teach all subjects, irrespective of whether they had passed the
language exams or not. The result is a situation whereby students are taught language by
teachers who are ill-prepared for the task of language pedagogy. This affects students’
general performance because other school subjects are taught in English.

Since Sheng is just not going to fade away, it is time that the language planners in
Kenya came up with practical ways of addressing the issue rather that just saying that
parents and teachers should discourage students from speaking Sheng. Considering that
Sheng has various social advantages outside the four walls of the classrooms, of which
the most important is serving the identity function, its study will provide insights for
language and education planners when formulating language policies. Too much time is

normally wasted lamenting the corrupting nature of Sheng, but hardly any effort is made

 

3 Acronym of CHAma cha Klswahili cha TAifa, (National Kiswahili Association) An
organization of Swahili scholars and writers in Kenya involved in the development and
promotion of Kiswahili.

to understand why its use persists in spite of its obvious disadvantages. It is high time

those language planners appreciated Sheng’s social advantages and work out how it can

exist alongside other mainstream languages. Apart from contributing to the general

advancement of knowledge and addressing such language planning issues, I seek to give

a clearer understanding of the dynamics of an urban code by looking at the different

patterns of identity that Sheng and its speakers represent. As a starting point to my study,

I begin by stating the general question that will be addressed in my discussion.

1.3. Guiding questions

Although other issues may arise, this study will be guided by the following principal

research questions:

How is the use of Sheng socially demarcated and what role does it

play in the daily lives of each category of its speakers?

How distinct are the Sheng varieties in Nairobi, and what are some of the

feature(s) that distinguish one variety from the others?

Under what circumstances do speakers shift from one variety to

another; what are the major motivations for such shifts?

What is the relationship between these distinct varieties and the identity of the

individuals or groups that use them?

What is the place of different local varieties of Sheng in the identity
of the speakers of those varieties vis-a-vis the broad identity of all

Sheng speaker in Nairobi and Kenya, in general?

How do the groups enforce norms in their varieties and how do the
individuals who subscribe to the group’s norms refine and modify the

identities of others in the group?

I will not try to answer each question in isolation since the answers sometimes overlap.
Even so, I will make as much effort as possible to stay focused on these questions. To

answer these questions, I will adopt the following hypotheses;

1.4. Hypotheses

1. Different identities are reflected by the use of different names for the same
referent.

2. Different meanings for the same referent are a reflection of different identities.

3. Sheng speakers can learn and use the words that are identified with groups other
than their own.

4. Speakers shift identities both consciously and uncounsciously.

5. Shifting of identities is greater in densely populated residential areas as opposed
to sparsely populated residential areas.

6. Sheng speakers identify with Sheng varieties that are similar or close to their own.

1.5. Major assumptions
The guiding assumption of this study is that the identity of Sheng speakers in Nairobi is
multifaceted. These different facets of the Sheng speakers’ identity manifest themselves
as they (speakers) weave through different social units with their unique Sheng variety.
Rather than being in opposition, these different aspects complement each other akin to
the different parts of the body, without which the body cannot be whole.

I also assume, following Githiora (1999:4), that the lexicon is the most important
area in the projection of identity as opposed to syntax or phonology, at least partly
because words carry a whole system of history and meaning and other levels do not seem

to be as varied. The morphosyntactic variation, though crucial in lexical coinage does not

exhibit a great variation disparity. In any case, phonotactics and morphosyntax are
intertwined with lexical manipulation in various Sheng varieties. Therefore, if lexical
items are the most important markers of the group’s identity, then the speakers who wish
to gain acceptance into any prospective group have to learn the lexical items associated
with the group in question. This does not necessarily lead to erasure of prior identity but
rather the modification of earlier identity. In fact, earlier identity can be regained through
the speakers’ re-entry into the earlier group.

In addition, even within a given social category, all the people do not occupy the
same position. There are some who are at the core of the group and others are at the
periphery. The position one occupies within the group is usually reflected by that
speaker’s mastery of the group’s linguistic code. This mastery is itself determined by (a)
the duration of time one has identified with the group, (b) the level of participation in
other social activities that are considered important by the group, and, for the case of
outsiders, (c) the group’s willingness to accept him/her among its ranks. This creates a

hierarchy of identities in any group.

1.6. Definition of terms

Having provided the general background of the study, I will proceed to the task of
defining some of the terms that will recur in my general discussion, and which might be
confusing to the reader. 1 will first define those terms and then explain how I use them in

my discussion.

10

Matatu [ma4+tatu]. Tatu means ‘three’ in Swahili. When the matatu began operating as
an alternative means of transport way back in 1953 (Mbugua 2003), they were charging
three cents per trip, hence the name ma-tatu (plural of tatu). Matatus are the privately
owned commuter vehicles that supplement the organized transportation in Kenya.
Although the passenger capacity should be 14 passengers for the small vans and 26
passengers for the minibuses, the rules are not normally adhered to and rampant
overloading of passengers is common, thanks to the traffic policemen who overlook such

transgressions alter receiving bribes from the matatu crew.

Manamba [ma+namba] Namba is borrowed from English ‘number’. These are the crews
who operate the matatus and are called so because their? work involves shouting the
matatu ’s route number, loading people and their luggage, as well as charging fares. In the
common Kenyan mindset, the manamba are known for their transgressive acts such as
pick-pocketing, extorting sexual favors from female passengers whose fare has been
stolen on a matatu by a conductor’s accomplices, and physical confrontations between
crews and/or passengers (Mbugua 2003). The matatu work, according to Mbugua has
been a predominantly male affair, though some women are nowadays taking up jobs as

matatu drivers or manamba.

 

4 A convention in Bantu language study (e.g. Bleek 1862:97-100, Meinhoff 1932, 39-40)
has been to assign numerals to noun class prefixes. Prefixes m-, wa-, and ma- are class 1,
2, and 6 nominal classifiers respectively (Bleek p.189). As in other Bantu languages,
Swahili classifies all animates under class 1 (singular) and class 2 (plural). Most
borrowed nouns are classified in class 5 or 9 (singular) and 6 or 10 (plural). From this
classification, Babi is a borrowed clipping and thus belongs to class 5 (singular, babi) and
class 6 (plural, mababi), but it is also animate, which explains the confusion when some
people classify it as class 1 (singular, Mbabi) and class 2 (plural, Wababi).

ll

Mababi (ma+babi): Generally, these are the people from the affluent neighborhoods in
Nairobi. The name babi is a clipping from Babylon, a pointer to the influence of reggae
culture. The term is attributed to the teachings of Marcus Garvey, who equated the
enslavement of Africans in the Caribbean and Americas with the Biblical enslavement of
Jews in Babylon. The metaphor captures the socioeconomic domination according to
which the wealthy oppress the poor. I will discuss various significances of this metaphor.
Babi will be used alongside Mbabi, and mababi will be used alongside Wababi because
the respondents alternated between the two during the interviews. The use of
babi/mababi and Mbabi/Wababi results from speakers’ confusion as to whether to

classify them in either class land 2 or 5 and 6.5

Maheshte (ma+beshte): From the English word ‘best’. Although beshte like babi has

various significances, its most unmarked sense is ‘best friend’ or ‘very close buddy.’

Baze: This term was given to me by the Sheng speakers themselves. They used it to refer
to the local place where they usually went to pass time. Although I do not have the
etymology of this term, there is a possibility that it is derived from the English word
‘base’. I will use the term to refer to the local network comprising of people who know

each other and interact with each other often via shared linguistic norms.

Although there are various other terms that need to be defined, 1 will restrict my

definition to these five terms on account of their uniqueness and relevance to my study.

 

5 See footnote 4 above.

12

Matatu and manamba are especially problematic to people who have never had the
Nairobi transportation experience; on the other hand, the terms babi and beshte have been

isolated due to their opacity. Other terms will be defined in the course of the discussion.

1.7. Plan of Work

The five Chapters of this dissertation are organized as follow; Chapter 1 has given insight
into the issue at hand. I have generally talked about language choice and then given a
background of Sheng as an example of a linguistic code that involves making linguistic
choices. The definitions of the problems as well as the issues that will be dealt with in
this study have also been addressed. In Chapter 2, 1 will provide a description of the
setting of this research, review the literature, and elaborate on the research and analytical
methodology and theoretical frameworks. That Chapter will also give an overview of the
linguistic background, the demographic patterns and Kenyan language policy, all of
which have implications for the overall problem of Sheng, its status, use, and users. A
literature review and background research on the issue of identity, the relation between
language and identity, and how the identity issue has been treated in previous work on
Sheng, will also be discussed. The section on methodology and theoretical frameworks
concludes the Chapter

Chapter 3 presents the picture of variation, beginning with Sheng as a linguistic
code for its speakers as members of social categories. Beginning the discussion with
structural variation, 1 will then proceed with a discussion of various Word Formation
Processes (WFPs) that promote lexical variation in Sheng. The outcome of these WFPs
(data) will be presented as lexical variables, which are treated as synonyms. This will be

followed by the patterns of Sheng variation according to sex, schools, and

13

neighborhoods. My unit of analysis — baze or local network and its relevance in
variation and identity will also be discussed. The Chapter concludes with the discussion
of other factors such as ethnicity, socioeconomic status, as well as residential areas,
which are responsible for Sheng’s variation.

In Chapter 4, I discuss the construction and performance of identity in both
discourse and non-discourse sites. I begin with the role of ambiguity in lexical signifiers
and the reality they represent to both insiders and outsiders. These will be discussed in
terms of the young/old, male/female, urban/rural, rich/poor dualism, all of which mirror
the insider/outsider or us/them distinction. I will show how insiders categorize outsiders
and how the categorized groups sometimes contest the negative labels that have been
attached to them. I will then illustrate how identities are activated non-consciously and
not necessarily as outcomes of what might be called “rational choice” in spontaneous
discourse. Cases of linguistic crossing will be discussed as the deliberate efforts to adopt
the linguistic norms identified with a different social category. Vernacular discourse such
as vulgar speech and verbal contest will be shown to be important diagnostics for social
positioning within a social category. This will be followed by an examination of other
non-linguistic ways of constructing identity and the other discourse sites where variation
in Sheng reflects differences in identities. I conclude the Chapter by trying to combine
the construction of identity with people’s attitudes.

Chapter 5 will sum up the major findings in Chapters 3 and 4 and highlight the
contribution that this study has made to the advancement of our knowledge about Sheng.
I will also point to the challenges that Sheng poses in the field of language pedagogy in

Kenya’s education system and suggest areas for future research.

14

Chapter 2

Sheng’s socio—demographic background and research methodology

Part I. Linguistic and demographic backgrounds

2. Introduction
This Chapter will deal with the linguistic, historical, and demographic, as well as the

methodological aspects of the dissertation. The linguistic background begins with a brief
overview of the languages of Kenya, starting with major language groups and other
classification issues. I then touch on multilingualism and its impact on the interaction
patterns in Kenya’s urban areas. This will be followed by a discussion of the historical
background of Nairobi, the setting of this research from the times it became an urban
center, through its various expansions to the present development, with special emphasis
on demographic distribution. This is crucial because, in later discussions, I will show the
relevance of demography on the interaction pattern that have an impact on Sheng and the
whole question of identity. The discussion of language policy and its contribution to the
development of Sheng in Nairobi will complete part I of this Chapter. Part II, will be a
review of the literature, beginning with the definition of identity to the relationship
between language and identity and the way identity has been treated in previous work on
Sheng. This will be followed by a section on methodology, in which I outline the
methods and procedures that were used in the collection of data. This will pave way for
the discussion of the methods of data interpretation beginning with the quantitative
methods and the qualitative methods, which are inspired by various theoretical

frameworks. Afier a brief explanation of how these theories will be used, I will then

15

conclude this Chapter with a brief summary. To begin this Chapter, 1 will now turn to a

discussion of the languages in Kenya.

2.1. Languages of Kenya

Indigenous languages in Kenya have been classified into three language groups — Bantu,
Nilotic and Cushitic (Abdulaziz and Osinde 1997, Myers-Scotton 1993a). According to
the Ethnologue classification based on Greenberg (1966), Bantu languages belong to the
Niger Congo family, Nilotic to the Nilo-Saharan family, and the Cushitic to the Afro-
Asiatic family. Bantu speakers are the majority at 65%, followed by the Nilotes at 30%,
and the Cushites at around 3%; the rest are speakers of foreign languages (Whiteley
1971, Rhoads 1977, F awcett 1970), see also Githiora (2002). The major ethnic groups are
Gikuyu 22%, Luhya 14%, Luo 13%, Kalenjin 12%, Kamba 11%, Kisii 6%, Mom 6%,
other Africans 15%, non-Africans (Asians, Europeans, and Arabs) 1% (CIA factbook,
2004). These seven major ethnic groups make up 84% of the total Kenyan population.
The exact number of languages is contentious. Ethnologue, for instance, gives the
number as 61, but other sources (e.g. Mbaabu 1996, Chimerah 1998, and Rhoades 1977)
give the number as 40. This is due to the grouping of mutually intelligible languages
under a common label, as is the case with the Luhya, Kalenjin and Meru. The Bukusu,
Nyore, and Idakho, for instance, are regarded as Luhya dialects, yet in Ethnologue map
(Figure 1) they are coded as separate languages. Similarly, Tharaka and Mwimbi are
normally classified under Meru, while in the map they are represented as separate
languages. The Sabaot, Pokot, and Tugen, normally classified under Kalenjin, are coded
as autonomous. Ethonologue representation of Kenyan indigenous languages is shown in

Figure l.

16

 

 

       

        

 

 

S . .
\tfl Ii ._\\.w..\\:.fl1\¢.>\
Ssss\\\
..x...m.\\\\.....\......mmxx».... . :3

  
  

    

do"

          

......)

2.THEWH"‘EARFA IWTMLLY

  

 

   

E .. . ..x V\. . ...\\
. .\....\.x..
w .. :H.\\....m.\\ M
. _ ... x m
a
m
..m
m
... m
4..” ..... n ...... m
E N“
T ..rUT E W I I A
meatfmmmm 3.;an
muwwwmmmm A. awmammmmrmm
flan mmmmaxmwm mononuéaamn

k GAR R EDI-N U RAN
IO IDAOIO BU (uh-TIMI“

13 EN 00

17GUSI

01% SIL

C USN ITC

LANGUAGE FAM LIES
BANTU

NILOTC

Or [ANGUAGE A“ as.
u N IN HABITED . A
3) 1m 15) 21)

 

Figure 1. Languages of Kenya (source: Ethnologue)

 

 

DJ

in,

“IL

 

Non-indigenous languages spoken in Kenya are European languages such as English,
German, French, Italian and Spanish; Asian languages such as Hindi, Gujerati, Punjabi,
and Urdu, spoken exclusively by the Asians (see Gorman 1974); and Arabic, mostly
spoken at the coast. English is the most important foreign language, having been
inherited from the former colonizers. Other European languages such as German, Italian,
French, and Spanish, together with the upcoming Asian languages such as Chinese and

Japanese, have no major significance outside the tourism industry.

2.2. Multilingualism in Kenya

Multilingualism in Kenya is a norm rather than an exception. Language contact,
especially in urban centers has resulted in widespread codeswitching and the
development of an urban mixed code known as Sheng. It is difficult to determine the
extent of multilingualism in Kenya before colonialism. Since the traditional political
economy, save for nomadic pastoralism, was self-sustaining at the local level and
provided limited opportunities for mobility, there were few possibilities for language
contact. Therefore, the worldview of most Africans was restricted to local ethnic
boundaries with occasional contact with immediate neighbors.

Colonialism transformed the linguistic landscape as Africans were gradually
integrated into the colonial economic system, though at a subservient level. In order to
accommodate the British settlers, Africans were confined to native reserves, a
development that led to the enhancement of interaction beyond the clan or kinship level.
Urbanization, one of the features of colonial economy, brought Africans from different

linguistic backgrounds together, and they had to confront communication barriers as one

18

of their challenges. Although the ensuing language contact accelerated multilingualism,
these urban centers did not become linguistic melting pots because the colonial structures
were not designed to fully integrate Africans into the economic system. Instead, the
colonial system designed strategies that were geared towards the perpetuation of the
unequal economic model where privileged colonizers occupied the top of the hierarchy.

At the linguistic level, a language was only important as long as it propagated the
status quo; otherwise it had to be subverted. In 1920 for instance, Harry Thuku, one of
Kenya’s pioneer nationalists, used Kiswahili, the lingua franca that was becoming
increasingly popular in the urban areas, to organize people from diverse linguistic
affiliations against the kipande (ID Cards) system (Mkangi 1982). Recognizing the power
of Kiswahili as a uniting language, use of ethnic languages was encouraged by the
colonial government in order to cultivate ethnic mistrust and thus curb the threat of
interethnic unity against colonialism. The colonial ideology guided the language policy,
where Kiswahili, vernacular languages and English, the language of the colonizers,
operated under circumscribed domains. This is the genesis of multilingualism, a
characteristic phenomenon in Kenya’s urban areas.

Multilingualism in Kenya posed a challenge to the independent government’s
unity agenda. The failure to redress the colonialists’ mistake through lukewarm
commitment to the promotion of Kiswahili as a national language left the urbanites
groping in the dark in the multilingual jungle. As urbanites waded through this confusion,
multilingualism, which was at first a disadvantage, was turned into a powerful resource
for self-identity, an identity that drew from the rich cultural linguistic diversity, while at

the same time transcending the restrictive boundaries of ethnicity. Sheng, a linguistic

l9

code born through a combination of language contact and poor language policy, was
weaned on codeswitching, linguistic borrowing and lexical innovation and has been
perpetuated by the functional and social benefits attached to it by its speakers. It is these
functional and social benefits that are the subject of this study.

Having discussed the linguistic background, I now turn to socio-demographic
reality in sections 3 and 4 and 5. Section 6 discusses Kenya’s language policy and its
contribution to the emergence of Sheng and winds up Part I of this Chapter. In part 11,
section 7 is a literature review, section 8 contains the methodology, and section 9 gives a

summary of the Chapter.

2.3. Socio-demographic description of Nairobi
Nairobi started as an internal depot for the Kenya-Uganda Railway in 1899. Its name is
derived from "Nyarobe", a Maasai word that means 'the place of cool waters,’ due to the
presence of the Nairobi River. This river was important to the Maasai herdsmen whose
economic lifeline depended on good pastures and watering places. Like the Maasai
herdsmen, the Railway constructors found Nairobi to be a convenient place for an
internal depot due to its suitable climate and landscape. It was more centralized and
cooler than the drier areas of Ukambani, which made it an ideal resting place before
descending the Great Rift Valley.

After completion of the railway, the imperial administration shifted its
headquarters from Mombasa to Nairobi. An urban center measuring 18km2 was created,
with its main features being the railway buildings, and different residential quarters for

Europeans and Asians. The Imperial British East Africa Company (IBEACO), which was

20

a predecessor of the British colonial administration, moved its headquarters from
Mombasa to Nairobi, which culminated in the elevation of Nairobi as capital of the then
British East Africa Protectorate in 1907.

The extension of Nairobi has been at the expense of the neighboring districts of
Kiambu, Ngong, and Machakos, which have been annexed to pave way for its rapid
expansion. The original 18km2 urban boundary of 1900 was extended to 77km2 in 1927,
with other minor expansion during the later half of the 19405. This was firrther increased

to 686 km2 in 1963, and there has not been any other increase since then. These

extensions are illustrated in Figure 2.

1:. k'fi
' ./
\ r;- /'
I‘\,\ ,j "' x
:5 "s“ .4 I ’4'”. \
/' ‘1“! RT“ s, . r'”‘\ _J \
Inn-co’. x; ...q
P“-.——‘ 2
.-’ ”a”, . ---\ xv»: \
N4 t .l “1 I,
V” 7 ’-l- L. “ ...-3
. I
I, t. :1 I) “ "Io/X. r]
I ‘s L ...... - _/ I ‘; b’m-u5. I:
5/ "’\._.. ‘\.a"“\—" ..J \

 

 

 

O.’..
"\\ ./ I. seem:
._ c
K.“ \\ Boundarie-
‘0‘ ..fl .

w ‘AK. \H 1900 _
V‘s-q” ‘g‘ .920 ————————
9 s 2 in nu. KN... \\ I927 -..-.. --
h M-“ Present -..—.....-

Figure 2. Urban expansion (Source: Kenya Bureau of Statistics)

In 1906 the population stood at 11,000. Four years later when Nairobi became a
municipality, it had only increased by 4,000. By the time of the first extension in 1928,

the population had almost doubled to 29,864. On the eve of the country’s independence

21

in 1963, the number had increased to 266,975, thus prompting another extension.
Currently, the population is estimated to be close to three million. Table 1, adapted fiom
Olima (2001), traces the population growth from the time Nairobi became incorporated to

the most recent census data.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year Area in hectares Population % increase
1906 1,813 11,520 -
1928 2,537 29,864 159.4
1931 2,537 47,919 60.5
1936 2,537 49,600 3.5
1944 2,537 108,900 1 19.6
1948 8,315 118,976 9.3
1963 68,945 266,795 124.2
1969 68,945 509,286 90.8
1979 68,945 827,775 62.5
1989 68,945 1,324,570 ‘ 60.0
1999 68,945 2,143,254 61.8

 

Table 1. Population growth in Nairobi, 1906-1999

While the huge leap in 1928 can be attributed to the increase in the size of Nairobi, the
rapid increase from 1963 is a result of both migration and natural population growth. The

relevance of this increase to this study will be addressed in the following section.

2.4. Demographic facts: A history of residential disparity

Nairobi’s original territory had no residential areas designated for the Afiicans (Mitullah
2003). This changed with the entrenchment of the colonial economy that required Afiican
labor to perform tasks that were considered too low for the Europeans and Asians. By
then, the Indians who had come to Kenya as casual laborers during the construction of the
railway had already established themselves as the trader class, though they were still

below the Europeans with regard to social status. A three-level hierarchy based on race

22

ensued which manifested itself at all levels, including the residential set up. The
European quarters, for instance, were located in the spacious areas west of the Central
Business District (CBD), and the Asians lived close to the CBD, in such estates as Ngara
and Parklands, or in their business premises within the CBD itself. The Africans were
confined in crowded quarters in the Eastern side of the CBD. The independent
government inherited this residential pattern and perpetuated it, though the hierarchy was
no longer based on race but on socio-economic disparity.

After independence, there was an unprecedented influx of people into the city,
driven by the following two factors: 1) the urban movement of Africans was no longer
controlled, as was the case during the colonial period: 2) the search for blue-collar jobs
and the allure of the city life attracted many people to Nairobi. This stretched the ability
of the post-independent government in the provision of basic services — a trend that has
persisted to date.

Currently, it is estimated that there are more than 194 urban centers in Kenya with
a total population of about 10.0 millions, representing 34.8% of the country’s population.
Nearly half of this urban population lives in Nairobi (Olima 2001 , quoting GoK sources).
As the population figures in Table (1) show, the Nairobi population almost doubled
within a span of 6 years, and has been increasing steadily ever since. It is expected to hit
the 6,000,000 mark by 2015. This grth has not been matched by the infrastructural
development to cater to the expanding population. The original estates (those close to the
Central Business District (CBD), have better road network compared to those away from

the CBD which were added during the expansion of Nairobi, as shown in Figure 3.

23

 

 

 

 

 

 

 

 

 

 

 

 

E
Duel-cl..- 2; GOCSBCQ‘

Figure 3. Different estates in Nairobi

Figure (3) shows some of the estates in the expanded map of Nairobi that was shown in

Figure (2). The ‘original’ Eastland estates (only those close to the CBD) are Ofafa
24

Jericho, Makongeni, Buruburu and Umoja, while the Westland areas close to the CBD
are Westlands, Parklands, Kileleshwa Lavington and Kilimani. Generally, people who
live in the western suburbs are affluent while the lower and the middle-income elements
still dominate the eastern suburbs (Mitullah 2003). The government’s lack of a sound
housing policy, official corruption, land grabbing, failure of the City Council to enforce
the by-laws governing the construction of structures within the city, as well as the city
residents’ inability to afford quality housing has resulted in the mushrooming of informal
settlements (shanties and slums). It is estimated that 60% of the total population of
Nairobi lives in informal settlements. These informal settlements occupy roughly 5% of
the total residential area, which means that over half of the urban population lives on 5%
of the total residential land area. The characteristic features in the informal settlements
are unregulated housing structures, filth, unemployment, illiteracy, and rampant crime. A
brief look at the informal settlements sheds more light on the popularity of Sheng and the

kind of reality it represents.

2.5. Informal settlements

Rapid urbanization in Kenya has exerted pressure on the government’s provision of basic
facilities. The population increase has not been matched by similar expansion of services.
The original map of Nairobi was well planned in terms of infrastructure; however, it was
only planned for a small population, and the city was unable to cope with the huge influx
of people, especially after attainment of political independence. Even after the city’s
boundaries were expanded, the government was still unable to provide basic

infrastructure such as housing, which led to the mushrooming of informal settlements.

25

 

 

SUC

 

The problem was exacerbated by the land grabbing phenomena, where public land meant
for road reserves or public facilities such as public toilets, parks and social halls were
fraudulently allocated to well-connected individuals who were allied with public servants
such as grassroots administrators, i.e. chiefs and the local councillors. These land
grabbers then erected structures in areas unfit for human habitation and became absentee
landlords. Since the owners of these structures did not live in them, they did not feel the
need to provide basic facilities for the tenants. It is thus common to find a block of mud-
walled or corrugated iron-walled 10ft x10fi rooms with no bathroom or public toilet, as
was the case in Sinai, one of my interview sites.

The situation in Kibera is much more complex, and the contention regarding the
status of the residents has always confounded the policy makers in terms of service
provisions. The owners of these plots do not have the title deeds to their pieces of land.
This means that they cannot invest money in the development of the plots since they do
not know what the government’s intentions are. The owners’ suspicion of the government
has always been manifested when plans are suggested about a Kibera slum-upgrading
project. Every time such plans are proposed, the owners think it is a government trick to
drive them out, and they always respond by demonstrating in opposition to these noble
plans thus perpetuating their squalid living conditions.

People go to the slums because the cost of housing and living there is generally
cheap. Here the residents are cut off from regulated society. The areas are not patrolled
by police, nor are the various informal businesses regulated. As a result, the owners of
these small-scale businesses do not pay taxes, and this makes their services to be rather

cheap. This might explain why my respondents in Sinai said they would not consider

26

moving out of the slums because houses and food are cheap, and they can smoke
marijuana and drink chang’aa‘5 with impunity. As a result of this, crime is rampant in
these slums since criminals can commit crimes and then disappear with the knowledge
that police will not be able to penetrate deep into the slums. Opposition to upgrading of
the slum is thus usually viewed as a way of making life difficult for the poor, who cannot
afford decent houses in regulated neighborhoods. In this study, I consider these realities
in the slums are the raw materials with which the slum dwellers construct their identity.

As far as Sheng is concerned, the slums provide a favorable setting for intense
interaction and conduct of social discourse. Not only do people from various linguistic
backgrounds live together side by side, but we also see that the general neglect that
characterizes their lifestyle is reflected in the educational system, which denies the
children of slum residents the resources to learn standard languages. It is not the use of
Sheng that inhibits the learning of standard languages, but poor instruction in standard
languages is partly responsible for the persistence of Sheng.

This brief description suffices to demonstrate the socio-demographic reality of the
setting of this research. However, the demographic issues are not solely responsible for
Sheng’s persistence. I now shift to the issue of language policy as a contributing factor to

the development and perpetuation of Sheng.

 

6 An illicit alcoholic drink locally manufactured and consumed in people’s homes. It is
popular because it is cheap and very potent. Deaths and loss of eyesight after its
consumption have been reported in local dailies. (See also http://news.bbc.co.uk/ on Nov
20 2000.)

 

27

2.6. The language policy in Kenya

Language policy is the official action of governments to determine the status of
languages and the nature of their makeup. The actions include executive pronouncements
and legislative bills or court judgments on matters affecting learning, use and
maintenance of language. Mbaabu (1996, 2000) argues that language policy is part of
public policy as formulated by a government to affect its nationals. Mbaabu further
argues that both action and inaction by the government constitute a policy. This
definition will suffice in my brief survey of language policy in Kenya.

Most writers have maintained that Kenya has a trifocal language policy (Habwe
2000:168, Mbaabu 1996, 2000:224, Whiteley 1969), where English is the official
language and Kiswahili, the regional lingua franca, is the national language (Harries
1975, Mbaabu 1977, Habwe 2000, Barasa 2000). The demarcation between the official
function and the national function of these two languages is not clear-cut and is
characterized by constant overlap. In addition, there are other languages of ethnic
affiliation, but again, there are some people who cannot identify with any of Kenya’s
ethnic languages and therefore use only the two languages for their social interaction.

During the colonial era, Kenya’s language policy oscillated between Kiswahili,
English and the vernacular languages. In the 19205, for instance, the administrators, the
educational leaders and the missionaries all pursued different language policies. The
Phelps Stokes commission7 of 1924 recommended the use of vernacular languages in
education, such as Kiswahili, Gikuyu, Luo and Nandi (Mbaabu 1996). Twenty vernacular

languages in which teaching materials were to be developed were identified. As a result,

 

7 Set up by the colonial office in London and charged with examining the educational
systems in the colonies and advising how education of the ‘natives’ could be improved.

28

some ethnic groups were taught in non-local languagess, a situation that was met with
opposition in some cases. Towards the end of the colonial period, the policy was shifting
in favor of English, which greatly reduced the role of Swahili in education (Musau 2000).
This issue continued to dog the post-independent governments, who established various

commissions to look at the issue of languages.

2.6.1. After independence
At independence English had already established itself as the language of upward
mobility that everyone was keen to acquire. It was only accessible to the minority class
that had managed to acquire formal education and took up jobs previously held by the
Europeans. This class, who now controlled government policy, ensured that the colonial
trend that had benefitted them was retained. English thus continued to be the medium of
instruction in school from primary to university level — a trend that has continued to
date. This does not mean that the government has been inactive in propagating an official
policy, but it has been poorly articulated (Barasa 2000:194). To the contrary, there have
been sustained calls for an official language policy by language lobbyists and official
pronouncements, as well as the setting up of commissions that have come up with various
recommendations on language issues.

In 1969 for instance, the ruling party KANU (Kenya African National Union),

decided that Kiswahili would be the national language by 1974 (Mbaabu 1977). This was

8 Giriama, a coastal language for instance, was chosen to serve all the nine tribes that
comprised the miji Kenda group (Giriama, Kauma, Chonyi, J ibana, Kambe, Ribe, Rabai,
Dumma and Digo), while Gikuyu, was chosen to serve all the tribes around mount Kenya
region, such as Gikuyu, Igembe, Chuka, Tigania, Tharaka, Embu, Mbeere, etc). The logic
of mutual intelligibility overrode all other considerations—such as cultural hegemony
and linguistic domination of other tribes.

29

opposed by the then Attorney General, who insisted on the need for internationalization
and efficiency in the use of the legal language (Harries 1975). In spite of this, on July 4,
1974, President Kenyatta announced that from that day on, Kiswahili would be the
national language in Kenya. The opposition by some parliamentarians, who lacked good
command of Kiswahili, coupled with the technicalities in drafting and debating of bills,
prevented this presidential decree from being completely implemented. In addition,
English had already assumed a symbolic index of sophistaction on top of the economic
rewards it afforded the country’s burgeoning elite. Since then, language policy has been
addressed in various commissions set up by the government, as well as in various other
arenas such as mass media, religion, music, constitutional reform, economic affairs, and
lobbying.

Starting with the commissions, the Ominde report of 1964, for instance,
recommended the use of English as a medium of instruction in schools. It noted the
importance of Kiswahili in building national solidarity and promoting regional
communication with neighboring countries. The commissioners recommended that
Kiswahili should be taught as a school subject, but students did not have to sit for a
Swahili exam at the end of the primary school cycle. Vernaculars were dismissed as
unimportant in education because they were deemed poorly equipped for pedagogic
tasks. The Wamalwa report of 1971 went beyond pedagogy as the driving force of
language policy by introducing economic and administrative angles. Noting the
importance of German and French in tourism and international conferences that were
constantly being held in Nairobi, the committee called for the enhancement of training

capacity in those languages. It also recommended the teaching of Kiswahili in civil

30

service institutes to equip administrators with a suitable communicative medium with the
public they were being trained to serve.

Four years later, the Gachathi report of 1975 recommended the language of the
immediate catchment area9 to be used as the medium of instruction for the first three
years of school. English as a medium of instruction was to begin in the fourth grade. At
the same time, Kiswahili was to be introduced as a class subject. The report further
recommended that Kiswahili proficiency be tested at the end of the primary school cycle
to stop teachers from viewing it as expendable. Still, it was not until 1985, following the
Mackay report of 1981, that this recommendation regarding Kiswahili was implemented.
The Mackay report was titled “Report on the Presidential Working Party on the
Establishment of the Second University in Kenya.” It recommended the restructuring of
the educational system from a 7-4-2-3 system (7 years primary, 4-secondary 2-high
school, 3-university) to an 8-4-4 system (8 years primary, 4-secondary, 4-university). It
further recommended the establishment of the African language division, incorporating
Kiswahili and making Kiswahili compulsory in the Second University.

From these commissions, both in their frames of reference and in their
recommendations, it is evident that Kenya has not had a clearly articulated language
policy; we also see that these commissions were geared toward inculcating linguistic
attitudes that would evolve a language policy that reflected the official aspirations.

Attempts have however been made by lobbyists to have the language policy entrenched

 

9 “Catchment” refers to the community where the school in question was located

31

in the constitution — a feat that was accomplished in the recent constitutional review as

affirmed in Chapter 2 section 9 of the zero draft").

(1) The national language of Kenya is Kiswahili

(2) The official languages of Kenya are Kiswahili and English, and all official
documents shall be made available in both languages

(3) The state shall respect and protect the diversity of languages of the people of
Kenya and shall promote the development and use of indigenous languages
and sign languages

(4) The state shall promote the development and use of Braille and other
appropriate modes of communication for persons with visual and other
impairments.

It is unclear what ‘development and use’ means in the case of indigenous languages.
Most likely, they will be relegated to the traditional roles they have always occupiedI '.
The intertwining of language policy and public policy brings us to other arenas
where language issues are played out. In the mass media for instance, Kiswahili and
English dominate other languages (King’ei 2000, Musau 2000). In radio and television,
English takes the lion’s share followed closely by Kiswahili. FM stations such as Nation,
Capital, Kiss, and Metro all broadcast in English while East FM broadcasts in English
and Hindi. In contrast radio stations such as Iqra, Biblia Husema and Citizen Radio
devote a considerable amount of airtime to Kiswahili. As for the vernacular stations,
Gikuyu dominates with three FM stations—Kameme FM, Caro FM and Inooro FM. In

television broadcasting, English dominates, with little competition from Kiswahili. No

local language is used on TV. Driven by entrepreneurial considerations, most radio

 

'0 The final report of the 2003 draft constitution. While there are many issues that have
not been resolved to date, there is little likelihood that the section on language will be
altered since it is not one of the contentious issues.

” These roles include traditional dancing troups to welcome foreign dignitaries at the
airportsand entertaining tourists and guests during political rallies (Musau 2000: 130).

32

stations adopt the most optimal language. The only logic behind language choice is how
to maximize listeners’ satisfaction in their enterprises. It is therefore no surprise that
Sheng, a language that plays no part in Kenya’s official language considerations so far, is
widely used in these radio stations, which are dominated by youth from program anchors
to the audience.

English dominance is sustained in the print media. For instance the four major
dailies, Nation, Standard, People and Kenya Times, are all in English, while the only
Kiswahili paper, T aifa Leo, has limited readership. Compared to a vernacular language
like Gikuyu, which boasts around four magazines, we see Kiswahili being outperformed
by the vernacular languages in the dissemination of news. While we cannot talk of any
mainstream media that broadcast or write in Sheng per se, we find Sheng material in
various media outlets, as the following sentences from leading newspapers show:

(a) so I decided to dandia (hang onto) the matatu’s door. (Pulse, East African
Standard, April 8 2005)

(b) We are surprised odieros (whites) are not the only peOple in this resort (Buzz,
Sunday Nation, May 15 2005)

(c) During the Easter holidays I was bila plot. . .(without any plan) (Pulse April
22 2005)

Another interesting area to look at linguistic practices in Kenya is music. Until the early
19903, vernacular and Swahili music competed for airplay with foreign music in the
Swahili service of Kenya Broadcasting Corporation. Political themes during the clamor
for a multiparty political system at the beginning of the 19905 attracted the wrath of
government censors, who invaded production studios and confiscated music. Citing
linguistic limitations, the govemment-controlled Kenya Broadcasting Corporation ceased

playing vernacular music, while giving considerable airtime to foreign music. The irony

33

that foreign music in languages such as English and Lingala (a Congolese language) was
inaccessible to the majority was lost. Some vernacular musicians’ attempts to compose in
Swahili to cheat the language barrier were unsuccessful”. This situation persisted until
the late 1990s with entry of independent broadcasting stations that were sympathetic to
vernacular musicians like Radio Citizen and later Kameme FM.

On a positive note, the implicit ban on vernacular musicians created a void that a
new breed of artists attempted to fill. Musicians such as Poxi Presha, Hardstone, and the
group Kalamashaka adopted African American hip-hop style and blended it with
traditional beats while employing the linguistic medium Sheng that their audience could
identify with. Emboldened by the success of these musicians, other aspiring young
musicians employed Sheng in their music, and it received a warm reception. Since then,
Sheng has become a characteristic of Kenya music, which has witnessed a kind of rebirth.
Through Sheng, and its adaptability to hip hop, these upcoming local artists have
succeeded in overcoming the language barrier that had dogged their vernacular
predecessors. The casualty has been vernacular music and foreign music, especially
Congolese Lingala. By using the language that the youth identify with, these musicians
have not only won fans, but also the support of the media that highlight what is
happening on the entertainment scene. These musicians have also been able to link with
Kenyans in the diaspora, who feel connected to their music by the use of a linguistic
medium and styles that reflect the global trends.

In people’s everyday life, religion plays a very central role. Use of language in the

religious arena thus becomes a matter of interest to anybody interested in linguistic

 

’2 A good example is Peter Kigia (W a Esther), Simon Kihara (Musaimo) and Albert
Gacheru (Wamaitu)

34

variety in Kenya. In the Christian churches, for instance, rural services are conducted in
local vernaculars. In the urban centers, services may be conducted in a vernacular if the
congregation is comprised of mono-ethnic members, which is not uncommon in Nairobi.
Kiswahili is used in multiethnic congregations. This contrasts with youth services and
also the educated middle class, where English is widely used. In mixed congregations
English is used with a Kiswahili interpreter or vice versa. In the mosques, Kiswahili
dominates. According to Habwe (20002173), to accept Islam is to accept the use of
Swahili. This affinity between Kiswahili and Islam derives fiom a long history of contact
between the Swahili people and the Arab world, the cradle of Islam. As a result, most
cultural religious terms in Kiswahili have Arabic origin. Other aspects of traditional
African religion are carried out in the linguistic medium of the community in question.
Finally, I will look at language policy in the economic sphere. As the official
language, English is used for all legislation pertaining to economic policies. In addition,
English is associated with high status, due to its appropriation by the upper and the
middle class. Since these are the people who control all means of production, it follows
that English is the dominant language in formal economic circles. In the informal
economic sphere, Swahili and Sheng are dominant in the urban areas, although
vernacular languages are also used. This again reflects the socio-economic divisions,
where people of low social economic status lack the means to cultivate competence in
English and hence use the linguistic mediums that are readily available to them.
Following Mbaabu’s (2000) argument that language policy involves both action
and inaction about language, we can argue that Kenya’s language policy has been very

dynamic. However, its fuzziness has also given room for unbounded linguistic practices

35

as well as freedom to choose the medium that the speakers prefer. It is in this context that
Sheng was created, nurtured and sustained. Gradually, it has cut its niche in Kenya’s
linguistic market by filling the linguistic resource gap created by dire economic situations
and poor language legislation. As Kenya’s urban situation became more and more
complicated, this linguistic creation, once frowned upon, assumed a kind of value among
its users. If Sheng is viewed in terms of Bourdieu’s (1991) concepts of language and
symbolic capital where languages are regarded as items with symbolic value, it can then
be argued that it gave the users the accumulated prestige and honor in which to build their
identity. With the new dynamics where Sheng’s influence in cultural creation is being
felt, its symbolic value as it competes with other mainstream language cannot be ignored.

I expound on Bourdieu’s concept of symbolic capital in the proceeding section.

2.6.2. Sheng as linguistic capital in an alternative linguistic market

The linguistic marketplace model regards language as a signs of wealth regarded as a
form of cultural and social capital which can be cashed in economically (Downes 1998).
Competent language users are viewed as possessing linguistic capital capable of
procuring a certain material or symbolic profit (Bourdieu 1991166). ‘Utterances’,
according to Bourdieu, ‘receive their value (and their sense) only in their relation to a
market characterized by a particular law of price formation’ (p. 67). The price formation
imposes linguistic hierarchies relative to the social relations of power in multilingual
situations. In formal interactions for instance, the languages used by those in authority are

allocated the highest value while the minority languages are allocated the lowest value.

36

Bourdieu’s concepts of languange and the symbolic capital can be traced to his
idea of different kinds of capital. He names the three types of capital as (1) economic
capital that deals with economic resources such as cash and assets, (2) social capital
revolving around group membership, relationship, network of influence and support and
(3) cultural capital associated with socialization processes where knowledge, skills,
education are imparted in order to enable individuals to interact with others successfully.
Symbolic capital on the other hand is associated with prestige, honor, fame and
reputations which signal the difference in identities. What is different between the
different types of capitals is the defference in terms of values and how one kind of capital
can be converted into another. Economic, cultural and social capital can all lead to
symbolic capital. Since the linguistic market is part of the broader symbolic market
(Eckert 2000), the symbolic power represented by language should in theory be
congruent with economic, social and cultural resources. This logic seems to have
influenced Sankoff and Laberge (1978), who formulated a model of calculating a
linguistic market index by determining the role that a form of language played in
speakers’ daily lives. Their model equated the index of economic activity with
competence in the legitimized language. Linguistic capital in this context can be regarded
as part of cultural capital, which is manifested through fluency and competence in the

language in which individuals carry out their transactions.

A mismatch may result if people endowed with linguistic capital of legitimized
languages lack its accompanying social symbolism, as in the case of educated but jobless
people. These people seek an alternative linguistic market where they can spend their

linguistic capital without creating a mismatch. Codeswitching and language mixing

37

should be viewed as speakers’ attempt to spend their linguistic capital in order to gain
symbolic capital. This allows them to participate in the mainstream linguistic market as
well as the alternative social markets. Such people are able to participate in both markets
with minimal contradictions to their identities. We can also regard these people as

employing different strategies in order to satisfy the market requirements.

This concept of symbolic capital can be tied to the distinction between the
languages of status and solidarity as the two antagonistic alternative markets. The higher
you go up the status scale, the lower you descend on the solidarity scale, (see the model
developed by Ryan and Giles (1982) in Downes (1998)). According to this model, the
more a language or a speech variety is stigmatized in the mainstream linguistic market,
the more its status rises in the alternative market. A good illustration is given by Preston
(1986), who has shown how speakers of the stigmatized Southern United States dialects
evaluate their own speech positively in terms of social solidarity. Preston argued (p 315)
that preference for local norms along affective lined was greater in areas where there is
linguistic insecurity. Aware that their speech was stigmatized, the Southerners
experienced linguistic insecurity and they sought to escape this linguistic insecurity by

tapping the pride associated with southern cultural values.

Sheng speakers can be viewed along similar lines. Their participation in the
linguistic market would be along the affective dimension. The core speakers in the
ghettoes lack the symbolic capital to participate in the mainstream linguistic market. To
overcome their linguistic insecurity, they create an alternative market that is not governed

by the norms of that mainstream market. In this market, they define what is acceptable

38

and what is not. Since language represents, manifests and symbolizes authority (Bourdieu
19823109), it follows that the alternative linguistic market that Sheng operates in has its
own hierarchies and values. In this alternative market, the roles are reversed. The
speakers of authority languages in the mainstream are subjected to a different value
system which puts them at a disadvantage. All the people who make up the ‘Nairobi

speech community’ participate differently in these two markets.

Considering Nairobi’s multilingualism, the linguistic market is very versatile with
the national and official languages competing with the vernacular languages for space in
the market. This kind of linguistic environment, in combination with other non-linguistic
factors, makes bilingualism inevitable. Kiswahili has dominated in both ethnic and cross-
ethnic interactions. In formal domains, English occupies the apex of the linguistic
hierarchy and competence in that it not only symbolizes wealth (both material and
symbolic), but also authority (power). On the other hand, ethnic languages have always
been used as languages of solidarity. In ethnically heterogeneous urban localities where
youth have low mastery of their ethnic languages, Sheng becomes the default local code
of solidarity in the place of ethnic languages. This has raised its status in the alternative
market. Notice that Kiswahili cannot serve this role because it is also associated with
formality, due to its recognition as a national language. This study focuses on the
dynamics of Sheng in the alternative linguistic market.

The preceding discussion enhances our understanding of the context and
background surrounding the development and use of Sheng. This sets the stage for a
critical evaluation of Sheng from the classification perspective, as well as literature

review and discussion on research methodology.

39

Part 11. Literature review and methodology

2.7. Brief outline

This section begins with a clarification of what Sheng is and how it will be treated in this
study by distinguishing it from Standard English or Standard Swahili, proceeding with an
attempt at classification of Sheng within various linguistic categories. The discussion
surrounding identity politics will mostly focus on how the issue has been treated in
linguistic study. While the review of previous work is intended to show what has inspired
this study, the review of previous study on Sheng will point to gray areas that this study
seeks to address. Finally, the methodology section describes how data was collected, and

the way it was analyzed.

2.7.1. What is Sheng?
Sheng means different things to different people. To some, any non-standard features in
spoken discourse constitute Sheng. In section 1.2, I mentioned Kibaki’s codeswitching
utterance mji-enjoy which is normally regarded as Sheng. Kibaki might have given his
whole speech in English, or Swahili, or switched between both, but it is the use of just
one word mji—enjoy, that makes the people conclude that he spoke Sheng. To others,
Sheng refers to speakers’ alternation between English and Swahili with varying degree of
vocabulary contribution from the local languages. In this case, everyone who is unable to
maintain a conversation in one of these languages for a stretch of time and thus resorts to
codeswitching is also regarded as using Sheng.

Careful examination therefore demonstrates that Sheng is purely a lexical issue

with additional interplay between phonology and morphology. A non-Sheng speaker can

40

easily follow the conversation between two Sheng speakers through syntactic cues, but it
only takes the insertion of a strange vocabulary item or the manipulation of a familiar
vocabulary item to keep that outsider off-track. This explains why people who are not
familiar with these words constantly claim that young people speak in a language they
cannot understand. A brief look at some examples validates this claim.

a. ...I had strutted through the building with a sizeable chunk of gum on my

manyake (butt)
b. Having lived in state (United States) for long
From the Pulse magazine 8th October 2004

In (a) it is evident that the sentence is basically English, but the utterance is regarded as
Sheng just because of the word manyake (butt) while in (b) the word stato qualifies the
whole utterance as Sheng. Someone who is not familiar with Sheng could follow the
issue until such words are encountered. This kind of Sheng is common in written genres
where English syntax and phonology dominates. The final vowel in manyake is

pronounced as /e/ instead of /e/ as the Swahili derivation would dictate. Stato is also

pronounced /ste:to/ instead of /sta:t3/ which provides further evidence that the structure

is heavily influenced by English. This contrasts with spoken and musical genres, where,

Swahili is dominant, as in (0) below

c. nakupenda mpenzi usiyependa chapa
nakupenda mpenzi usiyependa doo
tikisa!
shore wa kweli
manzi wa kweli

41

Translation;

I love you my girlfiiend who does not like money

I love you my girlfriend who does not like money

shake it!
The real girl
The real girl
From the track ‘mpenzi’ by Dudubaya

In this track, the first and the second lines are the same except for the last Sheng words,
both of which mean ‘money’. If we replace these Sheng words with their Swahili
equivalent, pesa, the sentence becomes pure standard Swahili. Similarly, the fourth and
flfih lines are also the same apart from the first words, both of which mean ‘girl’.
Replacing them with the Swahili word for girl msichana, would render them Swahili.
Hence, it is the insertion of these words that makes the verses be regarded as Sheng.

There is also a semantic level on which familiar words deviate from standard
meaning, similar to the way African Americans use the word bad mentioned in section
1.0) above. A case in point is the verb Iengae which in Swahili means ‘aim’ or ‘target’,
but ‘evade’ or ‘avoid’ in Sheng. As my data will show, the sentence structure is mostly
Swahili, with few a cases of English syntax. A linguist writing a Sheng grammar will

essentially be writing a Swahili (or English) grammar. It is due to the realization that

Sheng is to a large extent a lexical issue, that words are the major focus of this study.

2.7.2. Categorization of Sheng within variation

A discussion of Sheng’s classification within linguistic cateegories is important because
it promotes a better understanding of what Sheng is which in turn makes it easier to
Understand what we mean by Sheng’s varieties. The reader might have noticed my

deliberate avoidance of the term “language”, or any other mainstream terms such as

42

dialect, pidgin/creole, jargon or slang (apart from the derivation of the name Sheng as
Swahili-English Slang) in my reference to Sheng. A critical look at Sheng shows that
while it does display aspects of all these linguistic categories, it cannot conclusively be
classified into any of then without raising some problems of definition, as will become
clear through a quick review of all these categories.

In the literature (e. g. Meyerhoff and Niedzielski 1994) pidgins are usually defined
as contact languages that arise in trade settings where the speakers do not have sufficient
time to learn each others’ language. Normally, one language, known as the superstrate,
dominates in the vocabulary of pidgins because of the social and economic dominance of
its speakers. The nondominant languages are called the substrate and their speakers are
motivated to learn the language of the dominant group in order to derive both social and
economic benefits. Classic examples cited in literature are the Chinook Jargon spoken by
the Native American, British and French traders in the Pacific Northwest during the
nineteenth century (Thomas 1970), and Tok Pisin, spoken in Papua New Guinea (Holm
2000). A creole, on the other hand, is usually defined as a pidgin that has been used for a
long period and has acquired native speakers — although there are some creoles that
cannot be linked to any specific pidgin

Due to Sheng’s linguistic hybridity, it fits the definition of a pidgin, though
‘grammatical simplification’ criterion as argued by Meyerhoff and Niedzielski (1994), is
not evident since Sheng’ inherits the full grammar from the participating languages as

can be seen in the example below.

43

Sheng: Buda—ko hu-hep-a job ku-swing na gate
9 father-Zposs hab-evade fv job inf-swing with gate

‘Your father skips work in order to swing with the gate’ '3

Swahili: Baba-ko hu-epuk-a 9kazi ku-pembe-a na m-lango

9father—2poss hab-avoid-fv job inf—swing—fv with 3door

‘Your father skips work in order to swing with the gate’
Of interest here is the grammatical similarity in these constructions. In bold are some of
the grammatical elements that appear in both Sheng and Swahili. In both constructions
the subject is in 9th nominal class, suffixed with 2“d person singular possessive —ko. In the
predicate, Swahili habitual tense is marked by hu- and the verb ends with the —a vowel
associated with all Swahili verbs of Bantu origin. The Sheng construction inserts an
English object job, and and English verb swing in the adjunct as well as the English noun
gate in the prepositional phrase. Grammatical simplification therefore cannot be used to
qualify Sheng as a pidgin. Another problem comes with the concepts superstrate and
substrate because in Sheng, Swahili is the dominant language and not English. Regarding
its creole nature, there are young people who claim that they were socialized in Sheng,
which again justifies its creole definition; however, there is no conclusive data to support
this claim. In any case, the fact that not all ceroles pass through the pidgin stage means

we can not rely on ‘native speaker’ criterion. Mufwene (200328) warns that “the

k

'3 The following abbreviations will apply for all the examples in this dissertation. Arabic
numerals = nominal classes, poss = possessive pronoun, hab = habitual tense, fv = final
Vowel, inf = infinitive, neg = Negation, sm = Subject marker, om= object marker, past =
past tense. pl = plural, sg = singular dem = demonstrative, passive= passive morphology,
cop = copula, pres = present tense, cond = conditional, agr = agreement, caus =
causative, perf = perfect tense, fut = future tense, de = deverbal affix, int = interrogative,
inv = inserted vowel

44

dynamics of language contact are so complex that people should resist the temptation of
accepting the explanation that ‘jumps out’ first’

The distinction between the register and dialect, according to Brizuela, Andersen
and Stallings (1999) is that a register is a language variety that reflects the characteristics
of the situation in which language is being used while a dialect is a language variety that
reflects characteristics of the users. Downes (1998224) gives a clearer definition of
register as the varieties defined by social function of language. An advertisement would
be different from scientific prose or a piece of poetry, because both are governed by
different stylistic conventions. Although linguists define a dialect in terms of
geographical as well as social factors, the popular definition of a dialect as an inferior
subdivision of language still persists.

While Sheng’s use is more widespread in informal situations, the fact that it can
be used almost everywhere (e.g, workplace, clubs, music, home, school, writing, etc.)
creates problems with the register definition. It is highly probable that the different
varieties of Sheng can be classified as registers, but the construct “Sheng’ in its broad
sense cannot be. Regarding the dialect, Sheng fits the linguistic definition, but not the
popular use of the term. Unless Sheng is regarded as a variety of Swahili varieties as
Githiora (2003) proposes, then the dialect definition is also lost.

The two other categories that I will mention are jargon and slang. Elder (1954)
equated jargon with technical vocabulary that is deliberately constructed to serve the
needs of members of a given field. It aids in describing concepts that might otherwise
require long explanations. A broader meaning of a jargon would then be a lingo or

specialized vocabulary used by a group that deviates from the mainstream language such

45

that an outsider cannot understand it. Doctors’ professional vocabulary and thieves’
paracodes are classic examples. When doctors use medical terminology, non-medics
cannot get the gist of their talk. Likewise, thieves with special vocabulary can conspire to
rob a victim within earshot of their prey, thank to their effective lingo. According to
Reves (1926) the pioneer dictionaries associated slang with illiterate people, inelegance,
lowness and vulgarity. Nowadays, it is usually associated with informality and
discreetness thus putting it on a par with jargon. While the current definition endeavours
to remove the negativity of the term, the demonization of slang has not completely
withered.

Viewing Sheng as as jargon or slang has its advantages. It identifies its speakers
who use it for specialized functions in both professional and social circles. The
discreetness that Sheng offers promotes social solidarity that pr0pagates the ‘us vs. them’
boundaries. However, problems arise when we consider the codeswitching elements of
Sheng in contexts where codeswitching has become the ‘unmarked choice’ (Myers-
Scotton (1993b). In such a speech community, the issue of specialized usage does not
factor in because everyone engages in the practice. As an example, president Kibaki’s
favorite, mji-enjoy, cannot qualify as jargon, or even slang. The same applies to Sheng
innovations such as mbuyu and buda, both meaning ‘father’, which are so widespread
that they can hardly be identified with any social group.

Examination of the merits and demerits of these traditional classifications exposes
flaws which can present descriptive problems in the study of Sheng. In light of these
Shortcomings, I propose an alternative way of looking at Sheng which takes its various

dynamics and peculiarities into account. This alternative perspective is the ‘order of

46

discourse’ mentioned by Bourdieu (1990), and it is concerned with how different
discourses are interwoven together. In an unpublished work ‘The dialectics of discourse’
Nonnan Fairclough (2001) writes, “an order of discourse is social structuring of semiotic
difference —— a particular social ordering of relationships amongst different ways of
making meaning” (p 2). In view of the different domains of language use in multilingual
Nairobi, Sheng qualifies as an order of discourse laden with symbolism and ideology that
characterize other orders of discourse that it relates to. If different bazes in Nairobi are
taken as quasi-institutions, where different discourse types takes place, then it becomes
possible to talk of discursive practices within these institutions with their own sets of
norms governing text production and interpretation.

While Sheng as an ‘order of discourse’ is well justified, this study does not focus
on Sheng speakers as a meta-category, but on various subcategories of speakers whose
discourse types are dictated by contextual factors. In this respect, Bourdieu’s sense of
field as social context, as elaborated by Thompson in the introduction of Bourdieu
(1991), is useful in the discussion of Sheng’s variation and identity with respect to
difel‘ent bazes. Thompson (1991214) writes:

A field is always the site of struggles in which individuals seek to maintain or

alter the distribution of forms of capital specific to it. The individuals who

participate in these struggles will have differing aims — some will seek to
preserve the staus quo, others will change it — and differing chances of winning
or losing, depending on where they are located in the structured space of
positions. But all individuals, whatever their aims and chances of success, will
share in common certain fundamental presuppositions.

Tal(ing the capital in this case to be the symbolic capital associated with different

dlScOurse types, it can be hypothesized that as Sheng speakers engage in discursive

Practices associated with different fields, their discourse is modified to conform to the

47

 

acceptable norms. It is these norms, which are based on ‘common fundamental
presuppositions’, which in this case refers to the norms governing lexical usage that
characterize and preserve the uniquness of the baze. It should be pointed out that the
transition to different discourse types does not necessarily have to result in a total
overhaul of their prior identities, hence the different identity hierarchies among members

of bazes.

2.7.3. The politics of identity: definition

The term ‘identity’ is not very easy to define. The reason is that it is used in different
works for different purposes. The problem is compounded by the fact that researchers
have come up with very different forms of identity; there is individual identity, group
identity, ethnic identity, national identity and so on. This lack of consensus on what can
be regarded as identity, or what it does, has further complicated its ambiguity because the
definition is usually lefi to the discretion of the researcher and the discipline in which the
researcher operates.

In scholarly work, identity has been treated from different perspectives of self.
There is the standard view, in which identity is regarded as the ‘self-definition’ and
‘Other-definition’ of individuals according to the shared social meanings associated with
SIKECific roles in a social structure (e.g. Stryker and Burke 2000, Burke and Reitzes 1991,
see also Berger and Luckman 1966). In this case, individuals are perceived as static
entities that map into socially constructed categories such as ethnicity, age, sex, social
Status, etc. (see Turner 1982). This view has dominated the definition and identification

of Social variables in sociolinguistic work such as Labov (1972a) and Trudgill (1974).

48

In contrast to the standard view, the dynamic view of identity (Le Page and
Tabouret Keller 1985, Turner 1982) rejects looking at identity as frozen in stable
categories. Instead, identities are regarded as fluid and in a continuous process of
negotiation and construction (Blackledge and Pavlenko 2001, Bucholtz 1999). This view
is increasingly gaining ground, especially in the communities of practice theory (Eckert
and McConnell-Ginet 1992, 1999, Eckert 2000,) and other studies on language and
gender. Identities are seen as being constructed through practice, where individuals
recognize themselves as part of a larger complex entity, endowed with choices and
capabilities to cast and re—cast themselves in different identities.

While these two ways of looking at identity appear to be in conflict, their
boundary is indeed very thin. According to Tajfel (1982), it is difficult to divide them
into two separate packages. Le Page and Tabouret Keller propose that they are in a
SYIDbiotic relation, both mutually benefitting each other. This becomes clear when we
recognize that the negotiation of identities does not take place in a vacuum, but that they
are constructed through discursive practices in social interaction. Categories are therefore
needed to act as reference points for the individuals who are continually modifying their
identities.

‘Brubacker and Cooper (2000) criticize both approaches by claiming that the
essentialist view (static) is too strong, while the constructionist view (dynamic) is too
Weak The thrust of their argument is that a strong conception of identity preserves the
colllrnon meaning of the term, by assuming homogeneity, which is important for
analytical generalization, but its disadvantage is that it is full of reified connotations. On

t1'le other hand, the weak sense of identity erodes the word’s everyday meaning, which

49

makes the term so elastic that it becomes almost useless in theoretical generalization.
Brubaker and Cooper’s solution is fragmentation of the term identity so as to “unbundle
the thick tangle of meanings that have accumulated around the term ‘identity’ and to
parcel out the work to a number of less congested terms” (p. 14). These less congested
terms are (a) identification and categorization (14-17), (b) self-understanding and social

location ( 17-19), and (c) commonality, connectedness and groupness (19-21).

The advantage of identification and categorization is that theare derived from two
are processual active verbs (i.e., identify and categorize) that lacks the reifying
connotation of ‘identity’ that invites us to specify the agent that does the identifying and
categorizing without necessarily implying internal sameness, distinctiveness or bounded
gr Oupness, Although they point out to the dialectic interplay between the internal self-
identification and categorization, they also mention a form of external identification that
is a ‘ formalized, codified, objectified systems of categorizations developed by
authOl‘itative institutions’ which has no internal counterpart.

Regarding self-understanding and social location, they argue that self-
understanding is a dispositional term designating ‘situated subjectivity’ which they
explain as ‘one’s sense of who one is, one’s social location and how they are prepared to
act’(Sic). Brubaker and Cooper claim that self-understanding lacks the reifying
connotation of identity without being restricted to situations of flux and instability. This

is because self-understanding can vary across time and person, yet still exhibit a kind of
S‘Elbility, because individuals have a conception of who they are.
Commonality, connectedness and groupness are the final clusters that Brubaker

and Cooper propose. While commonality refers to the sharing of a common attribute,

50

connecetedness refers to the relational ties that link people. Groupness denotes the sense
of belonging to a distinctive, bonded solidary group.

Brubaker and Cooper endeavour to come up with terms with descriptive adequacy

is a radical step in the right direction. Their proposed terms provide a great deal of
insight, but sorting out the identity politics is beyond the scope of this study. In spite of
their insights, I have still relied on the constructionist model while also borrowing the
concept of self-understanding. If we are to talk about construction and negotiation of
identity, then the subjects must first understand who they are in social space, and they
must also recognize the existence of other distinct categories. Only then can they
determine the terms under which they will relate with them.

Following Chomsky’s, (1977254) idea that language operates at a high level of
abstraction, which involves idealization (e.g., a homogenous speech community) in order
for its study to be scientifically grounded, it becomes obvious that the reified identity
canJil()t be discarded, in spite of what Brubaker and Cooper recommend. While I do not
wh011y agree with Chomsky when he downplays the importance of sociolinguitic
val‘iations, I agree with him that we need abstraction for meaningful intellectual work.
Idealization of social categories is important because it is inconceivable how an academic
discipline can advance if attention was paid to specific peculiarities of every individual

Witl101R some basic assumption about the sameness of social categories under study. If
individuals belong everywhere, then they belong nowhere; that is, if they have all these
identities, then they have no distinct identity in the true sense of the term. This however

does not imply that individuals are automatons who map into social categories without

co“tl‘ibuting any input to these categories. Yet, the essentialist’s strong view of identity

51

 

implies predictability of a subject’s behavior given the prevailing context, akin to the
behaviorists’ interpretation of human behavior in terms of stimulus/response dualism.
The behaviorist approach has been faulted for ignoring the fact that humans are

’14. In terms

rational creatures whose actions cannot be fully studied in the ‘Skinner’s box
of language use, for instance, people make rational linguistic choices to optimize their
interactional outcomes. This is the major reason behind the shifting of linguistic codes
(read ‘linguistic crossing’) as well as adopting the discourse norms of outsiders. The
addressors’/addressees’ dualism in human interaction means that these outcomes must be
conceived both psychologically and sociologically, a dilemma that will always dog the
politics of identity. In this study, I will mostly rely on Sebba and Tate’s (2002) definition
of identities as ‘texts of social practice’ based on identifications made in interactions
between individuals whose key strength lies on it emphasis on interaction. I will analyze
linguistic material to demonstrate how identity is constructed through social discourse,
although I will also employ more traditional, quantitative methodologies.

Having defined identity as well as reviewed the politics of identity, I now turn to a
discussion of the relationship between language and identity, before giving a brief

background of previous work of identity in Sheng. This will complete the background

necessary for my later discussions.

2'74. The relationship between language and identity
Latlguage and identity is a multidisciplinary concern that has occupied the attention of
Psychologists, anthropologists, sociologists, and educators, as well as linguists—

\

M. Used here to refer to the observatory apparatus used by psychologist B. F. Skinner in
his study of operant behavior.

52

especially sociolinguists and second language acquisitionists. Since identities are formed
through social interaction, I place language at the center of the identity project. I follow
Le Page and Tabouret-Keller’s (1985) declaration that all linguistic utterances are ‘acts of
identity’, in which speakers highlight particular aspects of their multifaceted identities by
selectively using linguistic resources at their disposal. Speakers’ choice of languages,
words in a language, or certain ways of pronunciation, as well as their other linguistic
practices, can either act as a key to their identities, or as the very representation of those
identities. Underscoring the importance of language in identity, Blackledge and Pavlenko
(200 1) stated that identities are constructed, negotiated and renegotiated through
linguistic practices.

Languages, and especially native or ethnic languages, are usually regarded as
Speakers’ ‘identity cards’ (Macaulay 1997) and have been used to index speakers’
Ethnocultural identity (e.g. Fishman 1994). Although the relationship between language
and ethnocultural identity cannot be denied, there is another view. Myhill (2003), in a
rejOiIlder, rejected Fishman’s X/Xish15 claim by pointing out that there are other ways of
constructing identity. Citing the Jewish diaspora in different parts of the word, Myhill
arglled that religion, common ancestry, etc., could work just as well in ethnocultural
identity. While Myhill offers very valid arguments, the role of language in identity cannot
be ignored. Bourdieu (1991) for instance remarks that “it is very rare in everyday life for
a 1anguage to function as a pure instrument of communication”. Apart from

cotIllnunication, languages serve other symbolic functions, such as identifying the

X

is . . . . .

This captures the argument that speakers are identified by their languages, e.g. English
§Peakers are Englishmen, and German speakers are Germans etc. The counter-argument
Is that a community’s culture can be expressed in a non-native language.

53

speakers who speak them. Identification is accomplished through engagement in
discursive linguistic practices that reflect the multiple positions that speakers hold and
hope to hold in terms of their identities, both within and across multiple discourses
(Giampapa 2001). In the extreme, a language can lose its primary communicative
function and remain only with its symbolic function (see Edwards 1985).

There is a vast amount of literature on the relationship between language and
identity in monolingual and multilingual situations and also in intra-linguistic and cross-
linguistic studies. In multilingual contexts, for instance, and especially in language
contact situations, the widespread code-switching phenomenon is usually attributed to the
speakers’ desire to index co-membership in different speech communities with their
accompanying symbolisms.

A study done on children of Dominican immigrants in Rhode Island (Bailey
2000) for instance, showed that those children switched between English and Spanish to
highlight their Hispanic, Dominican or American identities. Other studies in which
Iangllage switching was shown to index multiple identities include Italian-English code-
Swit(:hing by Italian youth in Toronto (Giampapa 2001), English-Chinese-Japanese-
Korean codeswitching by second generation Asian youth in Los Angeles (Lo 1999),
Hungarian—English (Myers Scotton and Bolonyai 2001, Bolonyai 1998, 2005), English-
Tonga code-switching by transgendered males in Tonga (Niko 2003), Swahili-English in
Nail‘Obi (Myers-Scotton 1974, 1992, 1983, 1993b, 1998), and the use of Maori, a local

‘anguage, by Anglos (called locally “Pakehas”) in New Zealand (Bell 1999).
In a majority of studies where only one language is involved, dialectal variation in

t(”511118 of lexical and phonological differences have been associated with the identity of

54

the speakers of those variants. Variation in glide centralization in Martha’s Vineyard,
Labov (1972a), was attributed to the degree of loyalty to island identity. Meyerhoff
( l 998) demonstrated that in Bislama, interlocutors could manipulate pronouns yumi
(inclusive) and mifala (inclusive) and yu (generic) to confirm or contest a shared identity.
Along similar lines, Ito and Preston’s (1998) study of the Northern Cities Vowel Shifi in
northern Michigan demonstrated that an unconscious choice of vowels reflected the
identity of the individual in terms of their desire or lack of desire to remain in the local
area. In another study, Schilling-Estes (1998, 2000) showed that use of the local dialect,
called the ‘brogue’ on Ocracoke lsland and the ‘Lumbee dialect’ by Lumbee Indians in
Robeson County in North Carolina, was motivated by the need to enforce the identity of
their respective speakers. Schilling-Estes’ study can be compared to Hazen’s (2002) work
in Warren County, North Carolina, where copula absence (e.g., She gonna go to the
Store), levelled was (e.g., You was the man) and levelled forms of past negative be, wont
(e.g., You wont gonna go) were regarded as markers of the local cultural identity. Hazen
argued that the rate of use or non-use of the stigmatized non-standard forms marked one
as either oriented to local identity represented by those forms, or expanded cultural
identities that deviated from those forms. In Eckert’s work (1998, 2000), stylistic use of
language was regarded as one of the identity features that distinguished the ‘Jocks’ from
the ‘Bumouts’ in Belten High — a suburban school near Detroit. While the Jocks spoke
the more standard variety, the burnouts’ language was characterized by high use of
vernacular features. In a study on fraternity men, Kiesling (1998) showed how standard
features could be manipulated by different speakers in the same setting to project the

identities that speakers wished to in display. In addition, Bucholtz (1999) studied

55

linguistic practices paired with social practices in a community of ‘nerd’ girls in the
construction of nerd identity.

Outside the United States, (Hoare 2001) has discussed how the Breton accent is
used in Brittany as a mark of Breton identity. Gardner-Cloros (1997) has also claimed
that codeswitching between French and a local Germanic dialect in department stores in
Strasbourg, Alsace, distinguished between ‘Alsatians’ and the ‘French from the outside.’
Likewise, a study by Dirim and Hieronymus (2003), showed how adolescents in the
German city of Hamburg created new identities by code-switching between Turkish and
German, even when none of the speakers was Turkish. In the UK, Sebba and Tate (2002)

Studied how the British born Caribbeans used the patois or Creole to project a
transnational non-territorial identity of the African/Caribbean Diaspora and the localized
Black English to project urban identity within the British Black culture. Earlier,
Trudgill’s (1974) Norwich study had shown that the use of the coronal and velar nasals,
[n] and [-y] respectively identified the socioeconomic class and sex of the speakers.

All these studies emphasing the centrality of language in the construction and
negotiation of identity provide a rich background for studying how different varieties of
Sheng are used to construct different identities in Nairobi. Before plunging into the task

at hand, I briefly review how the issue of identity has been treated in previous work on

Sheng.
2.7.5. Identity in previous studies on Sheng

Various works on Sheng mention its function as an identity marker of its speakers.

Beginning with Osinde (1986) and Abdulaziz and Osinde (1997), identity was advanced

56

as the major factor behind its formation. They actually claim that Sheng was invented by
the urban youth for the specific purpose of maintaining solidarity among members of
various peer groups. Although Mazrui (1995) does not claim that Sheng was invented
specifically to serve an identity function, he mentions identity as one of its functions,
among others, such as serving as a criminal para-code. In Githiora (2002), respondents’
description of Sheng included such phrases as ‘it’s a gangster language,’ ‘it’s a dirty
language,’ and ‘it’s a secret language.’ Considering that linguistic categorization reflexts
the categorization of the speakers of the language, we can rightly say the Sheng speakers
are associated with attitudes expressed in the descriptive phrases attributed to their
langauge. This is an indication that the external identification of Sheng speakers follows
along a hegemonic discourse identified with the dominant institutions of authority. A
Comment like ‘Sheng inaongewa na wakorawakora' — ‘Sheng is spoken by dodgy
people’, made by one of Githiora’s respondents is not likely to have come from the Sheng
Speaker, unless the speaker has surrendered to the ideology propagated by the institutions
of authority.

Like Githiora, I (Githinji, 2003) showed how the folk used descriptive labels to
Categorize Sheng speakers. In addition, the majority of my local Nairobi respondents’
hand-drawn identification maps showed that Nairobi people could easily identify areas
where Sheng is spoken. Although some respondents indicated Sheng being spoken in
almost all areas of Nairobi, it was apparent from a majority of the maps that Sheng is
identified with specific localities. Respondents also claimed that a person could be
identified as a resident of certain areas by speaking a variety of Sheng stereotyped with

that area. Even Spyropoulos (1987) associated Sheng with a certain group of speakers

57

(people engaged in the informal sector) such as shoe shiners, curio-sellers, hawkers and
parking boys”. It is interesting to note that all the works cited above take the static view
of identity, where identity is viewed in terms of social hierarchies based on membership
in idiosyncratic groups.

The recent dissertation by Samper (2002), on the role of Sheng in the construction
of identity and youth culture in Nairobi, is the pioneer project that focuses solely on
identity. In Samper's work, identity is constructed through Sheng, but can also be
mediated by other aspects of popular culture such as hip-hop music, the matatul7 culture

and the mass media. Regarding Sheng and identity, Samper (20032 4) writes;

Sheng has become a badge of a new, dynamic, and cosmopolitan youth culture
that looks to the west for inspiration. Sheng is a linguistic strategy through which
they can articulate and enact different forms of identity outside the ethnic
hegemony.
Samper appreciates the complexity of the identity of Sheng and its speakers. His
awareness of the existence of different varieties is captured by his statement that “Sheng
Varies by the speakers’ estate, age, peers, economic level, education, ethnicity, gender,
occupation and time” (p8). Samper works within the framework of hybridity theory as
propagated in the works of Bhabba (1994) and Nedervene-Pieterse (1995). He defines
hybridity as the merging of two different systems to come up with a new system, which
combines elements of its different components. Viewing Sheng in similar terms, we can
think of it as an entity that combines local and global fragments, both in terms of

linguistics and culture, to project an identity that reflects both. Sheng inherits all the

symbolism associated with its components. It indexes status because one of its linguistic

'6 Young men who direct the motorists to empty parking spots for a ‘fee’ and also guard
the vehicle against vandalism.
'7 Refer to definition in the previous Chapter.

58

components, English, symbolizes status, and it indexes solidarity because its principle
language, Swahili is a language of national solidarity.
Although Samper’s dissertation is a milestone in Sheng research, there are still
some areas that were beyond the scope of his study. On (p.9) for instance, he writes ‘in a
day, an individual may move between several identities and use several Sheng lexicons
for each’. He also notes that young people in Nairobi express their identity as members of
different groups by using the Sheng associated with those groups. Knowledge of a
group’s variety of Sheng enables individuals to be included, thus maintaining group
Solidarity and self-identification. Throughout his dissertation, Samper gives very many
good examples of Sheng vocabulary, but apart from some few cases, he fails to link those
Words with the groups that use them. This might be due to the orientation of his work,
Which was to look at Sheng from a holistic perspective. In this study, I build on Samper’s
work by focusing on Sheng’s lexical variation and associating the linguistic markers of

variation with the groups that use the variants. It is these issues that my hypotheses seek

to address.

2.8. Methodology

My methodology involved the filling in of open questionnaires, recorded interviews, and
personal observations. In addition, articles from newspapers were especially crucial in
providing background information. I also did online research where I found some of the
michongoano (verbal duels) from various Kenyan online forums as well as some cartoons
that are used for illustration. Although there is some quantitative data, a huge part of this

work is based on ethnography. Below is a description of how I went about collecting my

data, and how these data are interpreted in this study.

59

2.8.1. Data collection, instruments and interview procedure
My interviewing techniques involved written questionnaires (see the appendix) both in
English and Swahili. I also recorded around 20 hours of oral interviews where all groups
were asked to respond to specific questions, though sometimes I took the freedom to
deviate from the written questionnaire to follow up the respondents’ answers. Personal
communications with other researchers who have worked in related fields such as Kimani
Njogu and Mbugua wa Mungai provided helpfiil insights. Reconnaissance tours to the
research areas allowed me to make my own personal observations. Articles relating to
Sheng were obtained from the library of the East African Standard newspapers, while the
auothor of the dissertation on matatu provided a digital copy.

Two types of questionnaires were used. One had only l2 questions and targeted
primary and secondary school subjects aged 18 years and below (appendix 1). These
questionnaires elicited socio-demographic information, such as age, sex, and linguistic
backgrounds, such as the languages they speak, when they started speaking Sheng and
with whom they speak Sheng. The educational background was not intended as an
analytical variable but to show the school locations, which were considered important
because they would demonstrate whether the lexical terms used by the students draw
from the area in which the school was located. Since the residential areas in Nairobi are
demarcated along socio-economic lines, they were regarded as cues to socio-economic
status as disccussed on the section on Nairobi’s demographic pattern.

To establish whether the respondents were aware of variation in Sheng, they were
asked a flat question about whether all people spoke Sheng the same way. The

questionnaire was then modified to include a list of selected lexical items against which

60

the respondents were asked to write down the Sheng equivalents that they used. This was
intended to elicit the synonyms which have been used to show that different words are
used for the same lexical item in different neighborhoods in Nairobi.
These questionnaires were filled out after the students had signed the University
Committee for the Research Involving Human Subjects (UCRIHS)—approved consent
forms. For the school students, the class teacher also read a statement to them explaining
that the interview was not a school activity and that they were free not to participate.
Apart from some few cases, the questionnaires were filled out in class during the lunch
break or in the evening afier the normal classroom sessions. Authorization to conduct
research in Kenya was obtained from the Ministry of Education, Science and
Technology. Additional authorization to conduct research in secondary and primary
Schools had to be obtained from the Provincial Director of Education and the City
Directorate of Education respectively.

The subjects who were over 18 years old were all non-students, and their task
involved filling out more detailed questionnaires (see appendix II) as well as being tape-
recorded in an oral interview. They also signed (UCRIHS) consent forms. The first part
of the interview was similar to that of the subject under 18 years. In addition, they were
asked to compare themselves with Sheng speakers from other estates and mark whether
they spoke ‘exactly like me’, ‘little like me’, ‘different from me’, or ‘I can’t understand
them’. They were also asked to rate how Sheng speakers from other estates spoke Sheng
by using qualitative judgments such as ‘very well’, ‘not so bad’, ‘not well’ or ‘bad’. This
was prompted by my assumption that speakers would draw the boundary between the

Sheng variety that they spoke and the variety that others spoke. As a result, they would

61

not identify or seek to acquire a variety that they despised. This was indeed confirmed by
Shauri Moyo respondents who claimed that they could not speak the Sheng of
Makongeni because they considered it ‘childish’ and ‘old fashioned’, in spite of the fact
that the two residential areas neighbor each other. In any case, attitudes towards a variety
were likely to be influenced by other factors which had nothing to do with linguistics.
The prediction was that a variety could be labeled ‘good’ if it approximated the raters’
identity, or the identity they aspired to acquire. Similarly, a variety could be labeled ‘bad’
if it deviated from the raters’ variety, or was associated with values that that raters would
not want to be associated with.

Unlike the respondents below 18, who were all students, the respondents aged
over 18 were all non-students and they were not provided with a list of lexical items.
Instead, they were asked to explain the ways in which their Sheng was different from that
spoken by others. In tape-recorded interviews, they were urged to give the words that
they used in their local networks, which were not used in other areas or by other groups.
As the interview moderator, I constantly intervened to seek clarification and ensure that
the interview did not get out of hand. The term baze was adopted for this study because it
was considered possible to perceive the interaction networks of small clusters of Sheng
speakers without the dangers of erroneous generalization which could have resulted if
estates were takens as the units of analysis. Estates were seen as being characterized with
different clusters of speakers with each cluster displaying their unique lexical variations.
As a minimal category, the respondents in a baze could be regarded as exemplar speakers

whose speech behaviour represented the speech habits of the entire baze. This could not

have been the possible to do with the category ‘estate’.

62

I had a number of assistants who introduced me to my respondents similar to the
‘ friend of a fiiend’ method adopted by Milroy (1980) in gaining contact to the subjects. A
friend of mine introduced me to his friend who works in a barbershop in Nairobi’s
Eastlands. This barber asked his friends to come to his barbershop where I conducted the
interview. Since these respondents came from different residential areas, they offered to
introduce me to other groups in their respective residential areas. This turned out to be a
big advantage because it reduced suspicions and made the oral interview sessions
informal, even with the presence of a tape-recorder. This casualness was due to the fact
that the task of filling out the questionnaire had already convinced them that these were
just harmless questions afler all. I also made efforts to communicate in Sheng whenever
possible to encourage them and also to try to cultivate a sense of closeness with them. In
some cases, the research assistants also participated in the interview, though at a very
marginal level. In a majority of bazes, the interviews were markerd by overlapping and
shifting identities, through the the use of English, Swahili and the inserted Sheng words.
Interaction networks were considered crucial in this study because individuals and
groups identify with other individuals and/or groups that they interact with. To
investigate the nature of interaction networks, some of the questions were tailored along
Milroy’s (1980) conditions for calculating network scores, though this study was not
interested in network scores.
Although I chose the social variables that are used in some of the quantitative
analysis, such as age, residential areas, bazes, and school, the linguistic variables came
from the respondents themselves. My analysis focused on the interaction between these

social factors and the lexical variables among different groups of Sheng speakers, and

63

their implications for identity. From the onset, the interview sessions themselves had
relevance in terms of identity. As a person who understands Sheng, though an ‘outdated’
version, I asked the respondents to feel free to use Kiswahili, English or Sheng. I had
expected that since Sheng follows Swahili structure, the respondents would be very
comfortable completing the Swahili questionnaire; however, I realized that the majority
preferred to fill out the English questionnaires, even in cases where their mastery of
English was poor. Some preferred to seek clarification in Kiswahili but write in English.
When it came to the oral interviews, all communicated very well in either Swahili or
Sheng. Hence even without considering the substance or their responses, the linguistic
choice during the interviews displayed this element of shifting identities. Recall my
earlier discussion on language use in Kenya where I pointed out that English is the high
status language. Since these respondents were filling questionnaires in front of a Kenyan
who studies in America for the highest academic degree, their preference of English can
be viewed as the negotiation of status or an attempt to accommodate to the perceived
linguistic preferences of the interviewer. If the choice of a language is a statement of the
Speaker’s affiliation in the sense of La Page and Tabouret Keller (1985), it can rightly be
argued that the respondent’s preference of English was a way of telling the interviewer
that they have another persona symbolized by the language of status.

We might also attribute it to the fact that these interviews were carried out after
the respondents had just signed the consent forms — a very formal exercise usually
associated with English. Having already assumed a responsive identity demanded by the

formality of the exercise, the switch back to a different identity was just not quick

enough. Hence their preference for filling out the English forms as opposed to the

Swahili ones was possibly a carry-over from this formal exercise.

2.8.2. Data interpretation

Both the traditional quantitative approaches as well as qualitative analysis of
ethnographic data have been used in the interpretation of data. The quantifiable data was
entered into Excel spreadsheets, which were then converted into Excel 3.0 workbook
files. These Excel 3.0 files were then imported into Systat, a statistical software program
for frequency and percentage tabulations. Most of the statistical data and the summaries
presented in form of tables in this study come from the Systat tabulations.

A great part of my interpretation, however, is qualitative where the conversations
with different Sheng speakers are analyzed to demonstrate what it reveals about their
identity. Critical Discourse Analysis (CDA)18 that looks at language as a social practice,
using Fairclough’s (1995297) multidimensional approach that combines linguistic
description of the language text, interpretation of the relationship between the
(productive and the interpretative) discursive processes of the text, and the explanation of
the relationship between the discursive processes and the text has been used. By text here,
I do not just refer to written texts alone, but also the spoken text. The relationship
between the text and the social practice was captured in Fairclough (1995: 133):

The connection between the text and social practice is seen as being mediated by

the discourse practice: on the one hand, processes of text production and
interpretation are shaped by (and help shape) the nature of the social practice, and

 

'8 Although CDA, has mostly been used to study ideology and power (e. g Fairclough
1995, Ngonyani 2000), the question of power is beyond the scope of this study, although
we can clearly see the representation of power dynamics in Sheng texts.

65

 

on the other hand the production process shapes (and leave ‘traces’ in) the text,

and the interpretative process operates upon ‘cues’ in the text.

It is important to consider that language mediates between the group and its texts because
social discourse is to a large extent conducted through language. Since the ideology of
the speakers is encoded in their texts and expressed through language, it becomes crucial
to tie the texts to the identity of their producers. It is out of recognition of this fact that the
‘cues’ in the respondents’ texts will be analyzed for the ‘traces’ of the social practices,
which pervade their discourse. These social practices will be shown to reveal various
aspects of the speakers’ identity, while also influencing their personalities and in the
process modifying those very identities.

The findings are discussed within the framework of different sociolinguistic
theories such as network theory, facework and politeness theory, social identity theory,
accommodation theory, linguistic marketplace theory, communities of practice theory, as
well as the audience and referee design theory. A brief explanation of these theories will
be given within the context of the issues they are intended to address. All the tapes
collected in the field were not transcribed, but were coded and their main ideas
summarized. Although their data does not feature in the discussions, they have influenced

some of the claims and generalizations that I have made in this study.

66

 

Chapter 3
A diagram of variation: Exploring form and the nature of variation in Sheng
Language is a guide to “social reality.” Though language is not ordinarily thought
of as of essential interest to students of social science, it powerfully conditions all
our thinking about social problems and processes. Human beings do not live in
objective worlds alone, nor alone in the world of social activity as ordinarily
understood, but are very much at the mercy of the particular language which has
become the medium of expression of their society. It is quite an illusion to
imagine that one adjusts to reality essentially without the use of language and that
language is merely an incidental means of solving specific problems of
communication reflection. The fact of the matter is that the ‘real world’ is to a
large extent unconsciously built upon the language habits of a group. No two
languages are sufficiently similar to be considered as representing the same social
reality. The worlds in which different societies live are distinct worlds, not merely
the same world with different labels attached (Sapir, Language 1929)
3. Linguistic variation and social reality
The above quote will serve as a preamble to my discussion of variation in Sheng within
the general discourse of identity construction. Sapir’s powerful words show how
languages play a central role in peoples’ adjustment to their social reality. The social
reality is not constructed by the language that people speak; instead, language plays the
role of ‘a guide’ to understanding that social reality. 1 use Sapir’s quote above to
construct my central argument that different varieties of Sheng represent different social
realities. Githiora (199924) writes that lexical choices serve a symbolic function in
indexing the social identities and relationships being constructed during face-to—face
interactions. I follow this argument in demonstrating how lexical variants in Sheng are
used in both self and other categorization and identification. Speakers create their desired
world through lexical choices and in the process confirm or contest prior identities. In

this study the widespread variation in lexical items is taken as evidence that different

social groups use Sheng for different purposes. This will be the bulk of my discussion in

67

 

the next Chapter. To lay the foundation for discussing the construction of identity in the
next Chapter, I will present the diagram of Sheng’s variation in the rest of this Chapter.
Variation in Sheng is both structural and lexical. Although there is variation at
other linguistic levels, such variations are not the object of this study. I will begin by
discussing the structural variation and its implications for the issue of identity; however,
my focus will be the lexicon. My discussion of variation begins with a brief account of
variation in Sheng’s structural templates. I will then proceed with the examination of
lexical variation, whose dependence on elements such as word formation processes and
different backgrounds, such as education, occupation, ethnicity or neigborhood is the

foundation of identity-making in Sheng.

3.1. Structure of Sheng: two varieties or one?

lt is widely accepted that there are two structural varieties of Sheng. One structure is
English-based and the other is based on Swahili grammar. Abdulaziz and Osinde (1997)
call the Swahili-based variety Sheng and the English based variety Engsh (Sheng with
syllable rhyme and head inverted). On his part Samper (2002) chooses the folk
terminologies (maghetto for Sheng and mababi [Babylonians] for Engsh). The two

varieties are shown in (l) and (2) below.

1. Sheng/maghetto
a. Si-na ganji si-na ganji, si-na njaro za ma-ganji
lneg-with money, lneg-with money, lneg-with means of 6-money
‘ I got no money, I got no money, I got no means of getting money’
fi'om the track Superstar by Delicious.

b. Hi-yo munde i-li-kuwa ma-two thao hivi
9-Dem-that money 9sm-past-be 6-two thousand like
‘That money was like two thousand’
from a story narration by a respondent in Githurai

68

 

2. Engsh/Mababi
a. There is this chile I used to vibe
There is this girl I used to seduce.

b. I hope your matinas are better...
I hope your 6-tina- pl are better...
‘I hope your butts are better...’
(from w. mashada. com)

In (1) we see the typical Sheng/maghetto characterized by Swahili structure with
insertions of coined words while in (2) we have a typical Engsh/mababi variety where
either Sheng words or English slang is inserted. Another distinction according to Samper
is between basic and deep Sheng. While basic Sheng requires only bilingual knowledge
of the languages involved, deep Sheng is a codified resource of an in-group that shares a
certain fluid and intersubjective reality and thus requires initiation into the group and
participation in its activities in order to comprehend its discourse.

Adopting a completely different view, Githiora (2002) rejects the two varieties and
claims that there is only one variety — Sheng, the Swahili-based structure. He dismisses
the English-based variety as slang— English laced with Swahili slang. There is a good
deal of support for his argument. The following sentences from mashada.com intemet

forum are examples of full English structures with one or two slang words.

3. My zaks beat a guy once
‘my father beat a guy once’

4. You need to graduate chuo kwanza
you need to graduate college first
‘you need to graduate from college first’
In (3), zaks, a slang coined from Swahili mzee ‘old man’, is inserted into a complete
English structure. The word final —s violates the rules of Standard English because it

pluralizes a singular noun. In (4) two Swahili words end the English construction. Again,

the exclusion of the preposition fiom shows the deviation from the rules of Standard

69

English. Further support for Githiora’s position comes from the respondents themselves.
During my field study in 2001, for instance, my respondents from the University of
Nairobi, when asked to name the languages they spoke claimed they spoke Sheng and
slang among others (Githinji 2003217).
Nevertheless, it should be pointed out that the English structure goes beyond lexical
insertion as shown in sentences (5) and (6) below.
5. The thought of being vunj-w-a-d. ..
the thought of being break-passive—fiz-past.
‘The thought of being beaten up’
6. Incidentally, the mboches was karibu hapo
‘incidentally, the housemaid was near there’
We note in sentence (5) above that a Swahili verb is not only inserted into the English
syntax, but is also passivised and concatenated with English past tense morphology while
the inserted verb from Swahili vunja ‘to break’ undergoes a semantic shift to mean ‘to be
beaten up badly’. In contrast, sentence (6) inserts the coined word mboches (housemaid)
and Swahili locatives, but the structure remains basically English. More importantly, the
way the Swahili or the Sheng insertions are pronounced have their own peculiarity that
deviates from Swahili, English, or any of the other donor languages. While the two
examples might be regarded as just cases of codeswitching, they go against the system
morpheme principle (Myers-Scotton 2002259):
The System Morpheme Principle: In the Matrix Language, + Embedded
Language constituents, all system morphemes which have grammatical relations

external to their head constituents (i.e. which participate in the sentence’s
thematic role grid), will come from the Matrix Language.

70

In (5) for example, the syntactic relation in the Noun Phrase (NP) indicates that English is
the matrix language and Swahili the embedded language. Pasivization of —vunja in
Swahili, followed by English past tense morphology complicates the classical idea of
codeswitching because the system morphemes come from both languages, unless we
argue that vunjwa comes into the grammar already passivized. This indeed seems to be
that case because its semantic features deviate from the ordinary use of the word. We can
therefore argue that in (5) it is the semantic shift of the verb that signals that the
construction is Sheng rather that just a case of ordinary codeswitching. In spite of this
limitation, Myers-Scotton’s claim is very important in deterrning the baze languages of
Sheng as discussed in this study. The functional or grammatical morphemes (system
morphemes) in Myers-Scottons words, such as inflections, prepositions, articles as well
as auxiliary verbs are the most obvious signal of base languages in codemixing. A brief

look at Swahili verbal and nominal forms shows that Sheng is a Swahili-based code.

3.2. Drawing from Swahili grammatical pattern

A simple Swahili sentence follows a Subject Verb Object (SVO) order. The subject and
object are both nominal forms, and are morphologically marked unless they appear as
pronominal affixes (i.e., personal pronouns). The nominal forms include nouns and

adjectives Mohammed (2001). The noun contains the noun class prefix, marked in

numerical numbers, the root and sometimes, deverbative elements as shown below.

7

a. m-toto wa-toto
l-child 2-child
‘child’ ‘children’
b. m-fug-a-ji wa-fug-a-ji
l -breed-fv-de 2-breed-fv-de
‘livestock breeder’ ‘livestock breeders’

71

c. u-shind-i

l4-win-de

‘victory’
In (7a) we have a noun root toto that can be prefixed with either a class 1 singular marker
(m) or class 2 plural marker (wa). In (b) the deverbative mfugaji is derived from the
suffixation of a nominalizer —ji and prefixation of the verb fuga with class 1 and 2
nominal prefixes. Finally, (c) is also a deverbative fromed from the verb shinda ‘win’
which is prefixed with class 14 noun prefix to yield an abstract noun victory.

The verbal constructions are also characterized by morphological complexity due
to the agglutinative nature of the Swahili language. The verbal structures display more
complexity than do nouns. Inflected verbs can have the following elements in linear order
— a subject marker (SM), the tense marker (T), the relative pronoun (REL), an optional
object (OM), verb root (VR), and the verbal extentions depending on various derivations
as shown in the sentence below.

8 U-huru tu- li- o- u-pig- an- i- a

l4-free lpl-past-rel- om-hit-recipl-appl-fv

‘The freedom that we fought for’
In the predicate above, the verb root is preceeded by four morphemes and followed by
two more. All verbs do not have to have all these elements, but the extension part can
include various derivations, such as applicatives, passives, statives, reduplications,
causatives, reciprocatives, etc. that either add or reduce the verb’s arguments. During
Swahili-English codeswitching, verbal extensions might be present or absent, but
normally, an English verb root is inserted while the pronominal subject marker, tense and

the verbal endings which are all part of grammatical relations come from Swahili. The

72

same happens with all other foreign material. In the examples below, taken from a
conversation between South C respondents, the Swahili grammatical frame is evident. .
9. a. Wa-show
3p] sm-show
‘Tell them’
b. Tu-na-meet
Ipl sm-pres-meet
‘We always meet’
c. Ma-boy-s
6-boy-pl
‘guys,
Looking at the examples above, with the exception of (c) we see that all the morphemes
that contain grammatical relation come from Swahili. In Myers-Scotton’s (2002) 4M-
model, the reason for double morphology in (c) is because the two plural morphemes are
activated at different levels during the production process. I will not go into the details of
the Myers-Scottons production model, but I will note that these examples suffice as
illustrations of the Swahili base in Sheng.

I have shown this structure because of its implication in Sheng’s innovation. In
the section on word formation processes below, it will become evident why Sheng should
be viewed as an offshoot of Swahili. Not only is the frequency of Swahili elements
higher, but there is the dominance of Swahili morphosyntactic pattern in both noun and
verbal derivations as outlined above. Even so, I restate my earlier claim that Sheng is
basically a lexical issue and one should not be confused by the high rate of Swahili

material. It can also be recalled that I mentioned that Swahili words can have a totally

different meaning as is common in slang cross-linguistically.

73

While making a categorical distinction between the two varieties is beyond the
scope of this study, I should point out that since Engsh was created by the youth from the
affluent areas as a conscious reaction to Sheng (Abdulaziz and Osinde 1997), it should
not be treated autonomously outside the general Sheng discourse. On the same note, any
discourse on Sheng would not be complete without touching on the issue of Engsh. In
this study, I will treat the two structures within the broad umbrella of Sheng. This said,
juxtaposing these two structures is of utmost importance because they form the templates
that lexical items are mapped into. I have already pointed out that lexical variation forms
the most important manifestation of variation. This is not to downplay variation that may
manifest itself at other levels of linguistics such as phonology, morphology and
intonation. However, this study focuses on the lexicon because of its breadth, which at
times incorporates areas of morphology and phrasal categories. This claim is made with
the assumption that even the lexical examples that exhibit syntactic and phrasal properties
are mostly uttered as a chunk, as will become clear in the discussion of word formation

processes below. I begin with word formation processes.

3.3. Word formation processes in Sheng

Like in all human languages, lexical coinage in Sheng operates under cross linguistic
word formation processes. In this section, I examine some of the ways that are used in the
formation of Sheng’s vocabulary. I discuss cases of affixation, compounding, borrowing,

coinage, calquing, syllabic inversion, and reduction.

74

Affixation: affixation might change the lexical category of the word in the derivation of
another term. In other cases, idiosyncratic morphemes are used to give the new word a

completely novel appearance; as the words in (10) shows.

10. a. m-less English ‘less’

l-less
‘child’

b. ma-finishings English ‘finishing’
8-finishing-pl
‘butt’

0. ni-a-sa-je Swahili slang ni-aje ‘what’s up?’
cop-iv-is-int
‘how are you/how is it?’

(I. anti-nyita English and Gikuyu
anti-catch
‘fool’

e. un-bwog-able English-Dholuo-English
un-shake-able
‘unshakable’

In (a), mless, an English adjective less is prefixed with the Swahili class 1 nominal
classifier m to derive a noun. In (b) an English gerund ‘finishing’ is prefixed with Swahili
class 6 nominal classifier ma to yield a noun, while in (c), niasaje, an idiosyncratic
syllable sa is infixed to the Swahili slang item niaje ‘what’s up’. This syllable is derived
through infixation of s followed by a vowel that harmonizes with the preceding vowel
(see Gathigi 2000211)”. In anti-nyita, an English negation prefix anti ‘against’ is affixed
to a Gikuyu verb nyita ‘catch’ to denote a fool — one who does not catch or understand
anything. Finally, an English adjectival morpheme is suffixed to a Dholuo verb bwogo ‘to

scare, or to frighten’ (Hofrneyr, Nyairo and Ogunde 20032 374) to yield the adjective

 

'9 Other examples from Gathigi are husuko from huko ‘there’, husuyu from huyu ‘this’
(animate), hasapa from hapa ‘here’

75

bwogable which is in turn prefixed with an English negative adjectival morpheme to give

the final derivation unbwogable.

Syllabic inversion: Syllabic inversion is similar to metathesis in phonology. In
metathesis, the expected ordering of sounds is reversed such that if the expected order
was .....xy., we get ...yx... (see Hume 2004: 203). In Sheng, the focus is not on the
phonological sequence but rather the syllable sequence in words. Githiora (2002) has
called it Pig Latin while Mazrui (1995) and Gathigi (2000) call it backslang. I prefer the
term syllabic inversion because it avoids the baggage involved in Pig Latin or the more

confusing backslang. Examples in (11) are typical cases of syllabic inversion in Sheng.

11. new word old word M
a. m-di-ki m-ki-di ‘kid’
b. mba-nyu nyu-mba ‘house’
c. e-nda nda-e ‘car’
(1. di-ngo go-ndi ‘thief’
e. le-mbe mbe-le ‘front, first, ahead’

In the first example, English kid is made to conform to Swahili CVCV syllable structure,
and then affixed with Swahili class 1 animate marker m. During syllabic inversion, the
noun class markers remain unaffected. The words mbanyu and lembe are borrowed
directly from Swahili and then swapped, while enda and dingo were previous Sheng
words with similar meaning that have undergone syllabic swapping. This is a very
dynamic process in the creation of new words. Any Swahili word or previous Sheng
word has a chance of being relexicalized through syllabic swapping, although the nouns

swap more than the verbs.

76

Compounding: This process involves the combination of two or more words to create a
new word. In Sheng, the two words do not necessarily have to come from the same
language as illustrated in the examples in 9 below.

12. compound gloss
a. zeiks-man zeiks (mzee ‘old’ ) + man
93m old-man
‘father’

b. ku-uma vako kuuma (to bite) + vako (fake, lie)
inf-bite fake
‘resting’

c. ku-la njaro kula (to eat) + njaro (quietness)
inf-eat quiteness
‘resting’

d. ku-tia ndani to put inside
inf-put inside
‘to eat’

e. ch-ombo ch-a kazi tool for work

7sm-tool 7-agr-of work

‘butt’
In (123), zeiksman, the resultant Sheng word is a second level Sheng derivation. The
former term zeiks (see zaks in 3 above for its variant) combines with man to derive a
word for ‘father’. When zeiks combines with woman, the resultant derivation means
‘mother.’ In examples (b) and (c) Swahili words kuuma ‘bite’ and kula ‘eat’ combine
with Sheng words vako and njaro respectively to yield Sheng phrases. Interestingly
kuuma and kula are some of the Swahili verbs that are commonly used to derive Swahili

idioms”. These two terms are still used in Sheng; though some speakers regard them as

oldskool or ‘oldschool’2'. Examples ((1) and (e) show ordinary Swahili constructions

 

20 Examples are kula njama ‘conspire’, and kula rushwa ‘take a bribe’,
2' This term is applied to refer to Sheng lexical items that have gone out of use and are no
longer fashionable. It should be pointed out that what is oldskool among some Sheng

77

being assigned meanings that differ from their literal meaning. This is crucial since one
might think that the conversation is going on in Swahili while in real sense the semantic

shifi has turned the words into Sheng lexical items.

Borrowing: Word borrowing is the most common linguistic phenomena in language
contact situations (Stojiéic’ 2004: 30). In linguistic borrowing, the donor language does
not lose its words, nor is there a possibility of giving back the borrowed word. Borrowed
words are usually treated in order to conform to the morpho-phonological constraints of
the recipient language. Without prior knowledge, one might not know that the Sheng
word pare, is a borrowed word from the English word ‘party’ because of its complete
adaptation into the Swahili phonology. This is not always the case as the words sniper

and kahiu in 13 below demonstrate.

 

l3. borrowed word gl_o;s_ origin
a. manga ‘eat’ French
b. sniper ‘thief’ English ‘sniper’
c. dishi ‘eat’ English ‘dish’
d. lenga ‘evade’ Kiswahili ( aim, target)
e. kahiu ‘money’ Gikuyu (knife)

In (13) manga, is borrowed from the French verb manger ‘eat’. It is ambiguous and may
either refer to eating or having sex. This is not unique to Sheng. In Gikuyu, the verb ri'a
/rea/ means ‘eat’, but also can have a metaphorical meaning of ‘sex’ or ‘kill’. Sniper is a
very recent derivation following the sniper scare in Washington DC in 2003. Since this

was an international topical issue, the name sniper generally came to symbolize a very

 

speakers might still be in common usage among a different group of Sheng speakers. See
Chapter 4 for more on ‘oldskool’.

78

antisocial and dangerous person. Considering that the Washington sniper had demanded
payment as blackmail, the term’s association with ‘thief’ is hardly surprising. In (c) dish
has undergone semantic shift from a noun into a verb while lenga yields the exact
opposite meaning from its Swahili counterpart. Finally, kahiu” remains unchanged but like
the rest, has shifted.

Normally, nouns are borrowed more often than the other lexical categories. This,
according to Trask (1996:23) is due to the following reasons; (a) nouns are more
numerous than other classes of words, (b) new things are more likely to be denoted by
nouns than other words, (c) new nouns are often easier to accommodate within the
grammatical system of the borrowing language. In Sheng, borrowing is the most

important word formation process.

Coinage: Coinage still remains a very important way of forming words. As one
respondent said, word coinage can be spontaneous while at other times it is a deliberate
activity that has to be justified and thoroughly explained. The examples in (14) show

recently coined words, though the extent to which they are used cannot be established.

 

l4. coined word giog
a. murungaru . ‘police chopper’
b. weng ‘wheel of a car’
c. wanoto ‘mean guys’
(1. makabelo ‘teeth’

The word murungaru” was coined from the name of the former minister in the office of

the president in charge of the internal security docket, Chris Murungaru, under which the

 

22 The said minister has since been transferred to the Ministry of transport and
communication.

79

police department falls. The invented word weng is an example of spontaneous coinage
by the car washers at Shauri Moyo baze whose daily activities involve dealing with
vehicles, hence the numerous descriptive terms for car parts. In the following

conversation, Stevo, a Shauri Moyo respondent explains to me (PG) about the

spontaneity of coinage in Sheng.

1. Stevo: ....haina ati mwanzo ........ haina ati setwork

2. PG: mm-hm

3. Stevo: ...Sheng inakujanga vile siku zionaenda

4. PG: mm-hm

5. Stevo: ...every new day you discover something new...
6. PG: yeah

7. Stevo: You form a word for it unaona?

8. PG: yeah

9. Stevo: ...Sheng inaweza kuwa hata huyu. ..

10. PG: mm-hm

ll. Stevo: ...kwa mfano anakimbia akaenda kuanguka

12. PG: yeah

13. Stevo: . . .akamake sound. . .unaona?

14. PG: mm-hm

I5. Stevo: ...next time itaita kuanguka hiyo sauti alimake. Do you understand? That

how Sheng huja

The following is my translation

1. Stevo: it has no beginning. . .it has no setwork

2. PG: mm-hm

3. Stevo: ...Sheng comes the way days are passing

4. PG: mm-hm

5. Stevo: ...every new day you discover something new...
6. PG: yeah

7. Stevo: You form a word for it, you see?

8. PG: yeah

9. Stevo: ...Sheng may be even this guy...

10. PG: mm-hm

Il. Stevo: for example he is running and as he goes to fall
12. PG: yeah

l3. Stevo: he makes a sound. . . you see?

14. PG: mm-hm

15. Stevo: ...next time it (Sheng) will call falling that sound he made. Do you

understand? That is how Sheng works.

80

This exchange coming from a member of the group that has already coined new Sheng
words is a good illustration of how metaphor and onomatopoeia operate in spontaneous
coinage of words. Other examples are makabelo ‘teeth’ ((1) and wanoto ‘mean guys’ (0),
coined by two upcoming musicians who I interviewed. While the coiner of makobelo
claimed that he was inspired by a newscaster who had ‘funny teeth’, the coiner of wanoto
did not offer an explanation for his coinage. I was not able to establish whether these
words are used outside those musicians’ local networks. However, they are testimony to
how spontaneous coinage contributes to the array of lexical choices that Sheng speakers
can make. The use of these coined words alongside the already existing words expands
the Sheng vocabulary in addition to providing Sheng speakers with linguistic resources to

make their group distinct.

Clipping and phonological reduction: Clipping involves removing some parts from a
word. The term ‘shortening’ (e.g. Kouega 20032519), might also apply in this word
formation process, but his application of this term to Camfraglias — a creole language in
Cameroon also includes acronyms and abbreviations, which I treat as separate processes.
In Sheng, this word formation process manifests itself in the clipping of syllables where
some elements of the previous word remain. These words may either be new loanwords
or previous Sheng words. In some compounds, the second part of the compound may be
discarded while in others; morpho-phonological reduction takes place. After these
processes have taken place, phonological modification gives the Sheng words new

appearence as can be seen in (15) below.

81

15. reduced word gloss

a. paroz ‘parents’

b. hao ‘house’

c. daro ‘class’ (cf. darasa in Swahili)
d. dashi ‘female breast’ (c.f. dashboard)

e. maflai ‘police’ (c.f. ma—flying squad)

In the Swahili-based Sheng variety, words ending with a consonant, such as paroz, are
not so common, but they are very common in the English-based variety. However, what
is interesting is how the word ‘parents’ has undergone phonological reduction in the
derivation of paroz” . Hao is clipped from the English word house, and dam is clipped
from the Swahili word darasa with a similar meaning. In the word for female breast, a
Sheng term dashboard, coined by matatu people, has lost the second part of the
compound to derive dashi, the new term for the Shauri Moyo group. Shedding of the
second part of the compound also applies to the term maflai ‘flying squad’ (British
English), an elite police division that deals with violent crimes. Other cases of clipping
involve chopping off of phonological material from proper names eg, 0mosh from
Omollo, T ash from Gitau, Nai from Nairobi etc. although the subsequent phonological

treatment might lead us to regard only Nai as “pure” clipping in these words.

Calquing or loan translation: Sometimes, words might not be borrowed from the donor
languages but the meaning might. Calquing involves translation of the semantic
components of words and phrases from the donor language into the recipient language.
According to Trask (1996) a foreign word is taken and translated morpheme-by-

morpheme. The English words superman and world-view are borrowed from German

 

2’ Pam is the singular form of this word. Pam is sometimes pluralized into mapero thus
sticking to the Swahili structure instead of the English structure that derives paroz

82

Ubermensch and Weltanschauung respectively. In Sheng, this morpheme-by morphemes
translation might not be transparent because of the agglutinative nature of its Swahili
base language. I regard literally translated words that retain their similar meaning as
calques. My examples in (16) below might not be regarded as classic examples of calques

in Trask’s definition.

 

calque compare with
16. a. ji-to-a English expression ‘remove yourself

refl-remove-fv
‘to take leave/go’

b. genge English popular music genre ‘gangster’
6 gang
music
c. throw wood earlier Sheng version ‘ku-tup-a mbao’
‘be mad’ inf-throw-fv 9timber/wood
‘to be mad’

Jitoa has been used by the Sheng speakers for a long time. It is a calque from the English
verb ‘remove’. Literally translated, it means to ‘remove oneself’. On its part, genge is a
very recent case that can be attributed to the popularity of hip-hop music. The concept is
most likely a borrowing from the African American concept of ‘gangster’ in hip-hop,
which can be translated into Swahili as genge. The use of genge in popular culture should
not be confused with the mainstream Swahili semantics of genge as ‘gang of thieves’z".
The final example from Mbaabu and Nzuga’s (2003) dictionary is the most transparent
example of a real calque. Kutupa mbao, a widely used term, has been translated into
English morpheme for morpheme with the same semantics. Examples of calques are
however not very common and this study did not come across many of them. As I

pointed out, they might be contested as ordinary agglutination in Standard Swahili.

 

2" Genge also means ‘crowd’. It is possible that the use of genge in pop culture comes
from its ability to attract huge crowds.

83

Initialismzs: Although this term usually applies to the word formation process involving
more that two words e.g. DA (district attorney), MIA (missing in action), I will extend
its meaning to cover cases of single words, including situations where more than two
letters are taken from a single word so long as they are pronounced separately. A few

examples will do.

1 7. CD condom/chapatti
TA teacher
D Dandora estate

The formation of CD fiirther shows how complex lexical coinage in Sheng can be. The
condom reading derives from taking two letters from the name ‘condom’, but the chapatti
reading derives from the round shape of the chapatti, whose ”shape resembles the CD. TA
is derived from the phonetic realization of the clipping of the word ‘teacher’ yielding
/tie/, while D for Dandora is the best example of the most common form of initialism
because it involves only one initial.

These word formation processes over-generate Sheng’s vocabulary, thus resulting
in many variables for a single lexical item. This semantic relation is referred to as
synonymy and is responsible for the instantiations of lexical variability. Synonymy here
will be regarded as the semantic equivalence between different lexical items such that
one can replace the other without altering the meaning. Although the fact that full
synonymy is extremely rare or even non-existent, as claimed by Pustet (2003:249); is

taken into account, this study focuses on the ability of some words to substitute or replace

 

25 This should not be confilsed with acronym. In acronym, the letters picked form a word,
e. g., laser from light amplification by stimulated emission of radiation or scuba from
self-contained underwater breathing apparatus

84

others without altering the key semantics and not whether the lexical items are full
synonyms or not. The discussion in the next section provides authentic data for better

elaboration.

3.4. Synonymy: some of the lexical differences

To elicit synonyms, I asked students from different primary and secondary schools in
Nairobi to write down the Sheng words they used for the following concepts: father,
mother, girls, boy, policeman, to rest, to eat, to sleep, to go, music, thief, house, money,

fool, car, child, teacher (see appendix 1). Below are synonyms of selected words.

Father: buda, budeng', budeng’deng, fatha, fathe, fusebox, gadabu, kize, mbulala,
mbuyu, mdabu, mdagala, maze, oldman, papa, tifar, warinzu and zeiksman

Mother: gadama, mama, masa, matha, mathako, mathe, mathor, mdhema, mnyaka,
mnyasa, mokoro, monthere, monthero, msa, mtassi, mthama, mthas, munthre,
muthama, muthes, mwothes, themaa, therma, tima and zeikswoman.

Rest: bangaiza, bentle, gichi, kubanjuka, kubarizi, kubazz, kubreeze, kubuaga, kuchill,
kudozi, kugandiza, kujibanza, kukeyo, kula njaro, kumanya, kumorea, kumoss,
kupoa, kupoza, kupozi, kupumz, kupumzika, kurelax, kurest, kutulia, kutulix,
kutuliza, kutuna, kuuma vako, kuwaka, liatu, poa, swach, take rest, tinare, tired,
tulia, tuliatulia, tuliz and tuliza

Police: afuade, antimegech, babby, banga, gava, goshogi, gwang, karao, mabeast,
manyani, mapolisi, mban'ga, msoro, oraka, pai, ponyi, sanse, sarka, sinya, tipoli,
vedi and wachi

Girl: azguoba, bogola, buss, chibela, chik, chile, cholosa, dame, dem, gala, gwaru, hoe,
jaba, kemfaa, kenta, kff, kichuna, kidalipo, mama, man, mandu, manyake, manzi,
mbas, mbus, miogar, mlosho, mresh, mroro, mshe, mshi, msusu, mtasi, mtoi,
mtoto, nyambizi, nyang'we, nzima, roro, shore, shwala, siste, supuu, toto and
wanyina

Boy: beshte, boi, boy, boyboy, boyii, bro, chali, charty, dume, hommie, huyu boi,
jamaa, kaboy, kijana, licha, mdhi, mru, msee, mthi, muthee, muthii, mzee,

85

Eat:

Sleep:

Thief:

bonya, dema, difu, dishi, ficha, finya, fweba, fyamba, fyenfa, hog, jibonda, jirock,
jistack, jitibu, kumachez, kumatema, kupiga, kusosi, kustack, kutia ndani, manga,
minya, munch, ngama, sosi, stack and venga

K.O., bleki, kudoro, kudoroi, kudorome, kudosro, lcudoz, kudoze, kudozi,
kuduna, kukaita, kukata, kukoma, kulaka, kulala, kunatu, kushonga, kuswatch,
kutulia, kutuna, kuuma vako, natu and swachi

bandi, bandishi, bandit, chizi, conmen, dindo, dingo, dongi, goda, gondi, goya,
jambazi, jambo, jangili, kauzi, mgosa, mnati, mwizi, pinji, shifta, sinya, sniper
and thegi,

House: boma, cube, digz, ghetto, hao, jake, kej, krib, kubalo, maskan, mbanyu,

moroto, mtaani, riggs and won

Money: alpundez, capital, chapa, chantu, cheda, chumbs, daf, dongera, doo, ganji,

Music:

Fook

Go:

Child:

Car:

gashai, greens, hela, kenge, macheda, mafuta, majani, mapesa, monyo, munde,
munke, murangi, ngoma, niado, niadu, paper, pesa, sape and winchi

beats,-chizi, debe, genge, gogo, hewa, lyrics, maflavouz, mahewa, malyrics,
mavitu, mogithi, music, ngoma, reggae, rwembo, song, timus, vokol, wimbo,
wingnimbo and ziki

anti-nyita, bani, barhe, bore, bure, chizi, chodha, chununu, danda, dula, dwan,
dwanzi, fala, fika, fyatika, fyatuko, gamlez, gwethe, gweze, jamaa, kamande,
kemothe, kerothi, kichoyo, kwaya, lafa, mbatopas, mbonye, mdongo, mjinga,
mjeshi, mshakee, mwenda, mwere, nduwele, ng'ea, ngamres, ngamwe, ngoie,
nongwe, pankass, punju, quir, shenzi, shoga, thati, tiwetini, tost, tothi, upende,
weri and zushu

bail, bounce, chak, cheza l, chora, dark, gich, gushia, gwan, inua, ishia, jiches,
jicurve, jikata, jikate, jikov, jipanga, jitoa, jivuta, kacha, kugo, kuhepa, kula vako,
kung’oka, kuwiz, kwenda, lembe, mkidi, ndae, nyaa, nyorosha, ramus, shoot,
tinago, tuende, tukore and uiz

chali, doli, kijanaa, mdiki, mdoido, mjunior, mkid, mless, mniaju, mtoi, myoung,
ndai, tich and wothes

bima, break, dinga, gari, mathree, mboko, mbokonya, mode, moti, mots, murenga,
ndai, nganya, rack, ride, ticar and wheels

Teacher: D.H, mchals, mdie, mdijo, mode, mtiche, mwalimu, odich, odicha, odicho,

odido, odijo, odisa, odish, odiso, odojo, ondiso, t.a, tiche, tie and udicha.

86

From these words, it can be observed that the words with the highest number of
synonyms are ‘rest’ (42), ‘girl’ (45) and ‘fool’ (52), while ‘child’ (14), ‘house’ (15), and
‘car’ (17) have the lowest number of synonyms. It can be argues that words with the
highest number of synonymy are the words that dominate the ingroup discourse of
speakers as opposed to those with the highest numbers of synonyms. Speakers would, for
instance, have various words to refer to their activities in the bazes, hence the high
variability for the word ‘rest’. Since sex issues feature prominently in the discourses of
various groups, synonyms for ‘girl’ are many. Finally, outsiders are regarded as fools;
hence the name ‘fool’ displays high variability because it assumes in-group relevance. In

contrast, ‘car’ ‘child’, and ‘house’ are not in—group terms, hence the low variability.

3.5. Morphophonological variation and its connection to speakers’ identity

In these examples, some synonyms display very minimal orthographic variation. This can
partly be attributed to the fact that Sheng lacks a ‘standardized description of its own
rules’ (Mbaabu and Nzuga, 2003: introduction). As such, the respondents were at liberty
to write the words the way they deemed best, which sometimes yielded unmotivated
variation. Taking the variants for ‘mother’ for instance, mathor and matha should be
considered as one word, just like m-thama and mu-thama. Similar cases include dem,
dame and m—she, m-shi for ‘girl’. Other cases of variation can be attributed to the
presence or absence of nominal classifiers, e. g,

18. m: mroro, roro for ‘girl’, mtiche, tiche for ‘teacher’
ma: mahewa, hewa and malyrics, lyrics for ‘music’

’

mapesa, pesa and macheda, chenda for ‘money .

87

The presence or absence of an infinitive marker ku in the case of verbs can result in
variation, e. g;

19. ku—sosi, sosi and ku-stack, stack for ‘to eat’
ku-swach, swach and ku-natu, natu for ‘to sleep’.

Such differences should not be considered lexical variation since they only involve
orthographic or grammatical realization or lack of it. However, word-final and word-
initial phonological differences are regarded as variants because Sheng thrives on
relexicalization and phonological manipulation. To illustrate this, we note that there is a
big difference between mbele and lembe, ‘in front, before’, m-kidi and m-diki ‘child’, keja
and jake, ‘house’, pesa and sape ‘money’, chali and licha ‘boy’, manga and ngama for
‘eat,’ though the latter is derived from the former and they bear the same meaning. As we
saw earlier, syllabic inversion is a word formation process that forms words from existing
ones by transposing their syllables while retaining their meaning. Within optimality
theory, swapping of syllables can be regarded as the violation or reranking of the linearity
constraint, which has implication, for the word’s external appearance. Speakers have a
choice between using words that remain faithful to the linear order of syllables or using
words that violate it. Swapping of syllables was used with a higher frequency by the Sinai
girls that any other group, which sets it as linguistic marker of their identity, though it is
not resticted to their group.

Similarly, 0dich and 0dicho should be considered as two variants because their
realization depends on the speakers’ lexical choices. In my discussion of word formation
processes above, I mentioned that borrowed or coined words are normally made to
conform to the Swahili well-formedness rule of CVCV syllable structure. Words ending

in clusters or consonants are marked. It thus follows that while odicho remains faithfiil to

88

this constraint, 0dich violates it. If we revisit the Sheng—Engsh distinction, where we
observed that Sheng uses Swahili structure while Engsh uses English structure, then we
can rightly conclude that the choice of 0dich/odicho is deliberate. 0dich ’s deviation from
Swahili phonotactics signals a parallel word formation process. This argument for 0dich
and odicho can be extended to other lexical variants such as mots and moti for car and
bandit and bandi for ‘thief’. The relevance of phonotastics with regard to identity is that
those who choose English-sounding words are usually the affluent speakers from the rich
neighborhoods, while those who use the Swahili phonotactics are mostly from the
ghettos.

Apart from the difference in the word-final vowel, the other difference between
matha and mathe, or fatha and fathe is the syllable weight While matha [maoa] and
fatha [faoa] ends in light syllables, mathe {mazoe’zl and fathe [fazoéz] ends in heavy
syllables with rising intonation, while the middle syllables are stressed. Although the
word initial 0 in nouns is usually attributed to Dholuo influence (Abdulaziz and Osinde
1997, Mazrui 1995, Kang’ethe 2004)”, the choice between mdijo and odijo depends on
preference of the o-initial innovation or conforming to the Swahili requirement of
classifying all animates under class 1 nominal classifier prefix m. I will also regard the
choice of fricatives in odiso, odijo or odicho as being driven by phonological

manipulation; hence they can be regarded as variables.

 

2" This word initial 0 in nouns is usually attributed to Dholuo influence (Abdulaziz and
Osinde 1997, Mazrui 1995, Kang’ethe 2004). According to Hilary K’owino (personal
communication) the occurrence or word initial 0- is not characteristic of Dholuo.

89

3.6. Some background on the users of these synonyms

Turning now to the distribution of these synonyms among different categories of
speakers, we note that the most relevant categorical variation is manifested in the
residential areas where speakers come from, the school location or the school itself.
When drawing generalizations, it is important to take into account that there is a
mismatch between the names of these three categories and the actual localness of the
respondents. Taking the case of students for instance, it can be observed that due to their
mobility, it is not always the case that they come from the school location’s immediate
neighborhoods. This is especially so with the high schools. Arya Girls High School in
Ngara is a good example where the respondents were drawn from Huruma, Satellite,
Ngong, Utalii Village, Uthiru, Kibera, Juja Road, Zimmerman, Parklands, South C,
Amboseli, Highrise, Lower Kabete and Buruburu neighborhoods. Only 2 out of 14
respondents come from the immediate neighborhoods of Parklands and Juja Road. The
rest are all spread out across different neighborhoods in Nairobi. At the opposite scale is
the Busara primary in Kayole where all the pupils come from the immediate
neighborhood of Kayole. In other schools, it is possible to have students who come from
the same neighborhood. Buruburu High has for instance 4 students from Buruburu and
several from other immediate neighborhoods, though there are some who come from as
far away as Eastleigh, Savannah, Kayole and Komarock. Like Arya Girls, the case of
Muthangari primary might be misleading because although the school is located in the
up—market Lavington residential area, hardly any pupils from the immediate area attend
that school. The majority of its pupils come from the outlying slums, such as Congo,

Kawangware, Kangemi and Uthiru.

90

The above discussion shows that the relationship between the residential area,
school location and the school name should be treated with caution when dealing with
variation. In spite of these shortcomings, the three categorizations I mentioned can be
used to draw broad generalizations about the distribution of selected lexical items and

how they can be tied to the identity of their speakers.

3.7. Dominance of certain words

For each of the lexical items, two or three synonyms are widespread amongst the
majority of the student respondents. In some cases, however, there is a huge gap between
the most dominant synonym and its closest competitor. In the example below I show the
leading variables for the words for ‘father,’ ‘house,’ ‘to rest,’ ‘money,’ ‘child’ and ‘to
eat.’ Most of the other variables for these words had only one or two respondents. (refer
to appendix IV for a full illustration of the frequency of different variables for each
lexical item). The numbers represent the frequency of use of each term by the 212
respondents. The reader might notice that some additions exceed 212. This is because

some respondents used more that one variable for a single lexical item.

20. father mbuyu 120, buda 92, fathe 17
house keja 132, hao 66, mbanyu 11
to rest kupozi 34, kutuliza 29, kutulia 26, kupoa 12, kurelax 12,
bangaiza 10, kurest 5
money doo 84, ganji 76, chapa: 26, cheda 12, munde 6, niado: 6
child mtoi 141, mkidi 49, mjunior 12, mdiki 11
to eat manga 129, dema 34, dishi 20, bonya 6, hog 5

Among these five examples, mbuyu and buda dominate for ‘father’, keja and hao for

‘house’, kupozi and kutuliza for ‘to rest’, doo and ganji for ‘money’, mtoi and mkidi for

91

‘child’ and manga and dema for ‘eat’. Buda and mbuyu will be used in my discussion of

these types of variations and dominance according to sex and school.

3.7.1. The gender factor in variation

The use of buda amongst the girls is slightly higher (26%) as compared to the boys’
(21%). While low participation of girls in this study might have affected the general
outcome, we also reason that Sheng has for a long time been a boys’ phenomenon. We
note that buda is an older variable as compared to mbuyu. If the boys have been the
leading users and innovators of Sheng, then we can attribute this difference to the fact
that as girls catch up with Sheng, boys move a step away in order to create distance.
Girls’ low participation in Sheng has usually been attributed to their indoor activities that
keep them away from the street where they could acquire the vernacular. A different
explanation is that Sheng had been stigmatized for so long and girls who used Sheng
were viewed with suspicion. Samper (20022154) reported about a young woman called
Teeka, who told him that Sheng was associated with poor people and with thugs and
prostitutes and she chose not to use it in order to avoid such associations. The
sociolinguistic explanation for this, following Trudgill (1974) is that vernacular features
are normally associated with the covert prestige of masculinity and toughness, which are
desirable characteristics for males but not for females who on the contrary are more status
conscious due to the pressures from society. Women thus tend to stick to the standard
features rather than adopt the vernacular. The use of the older synonym for ‘father’
among the school girls can be attributed to their low integration into the Sheng networks,

either because boys keep them away or because they do not want to participate.

92

Trudgill’s arguments have been challenged by Cameron and Coates (1988) who have
remarked that sociolinguistic explanation tends to lump women from all social classes
together, ignoring inter and intraclass variations. They also criticized what they called the
‘intrinsic maleness of the norm’, where men provided the standard from which women
could only deviate. Their arguments are well-supported in this study. The Sinai girls, for
instance, used more Sheng forms that the Kibera or South C males. There was also a high
degree of difference between the frequencies of Sheng forms used by the Sinai female
respondents as compared to their female counterparts from Kariobangi. I have already
mentioned the pervasive inversion of syllables in their speech, which can be regarded as
an innovative strategy in their verbal transactions. This shows that women are also
innovators, but we have to go into their local networks and approach their discourse
devoid of preconceived generalizations. This calls for a need to be more specific that the
pattern of conservative norms among females is not uniform among all females. While
schoolgirls might be more conservative, girls outside school might not. Futhermore, the
school might have a mixture of girls from different social background, while the local

networks consist of females from the local networks who share a similar social reality.

3.7.2. The schools and variation

The distribution of mbuyu and buda in schools also reveals that mbuyu is the more
dominant variable; however, the secondary school students seem to use it slightly more at
65% as compared to primary school students at 57%. Table (2) summarizes the choices

according to schools.

93

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

School Sec/Pry buda mbuyu out of others
Arya secondary 3 l 0 l 4 l
Buruburu secondary l l 7 22 4
Eastleigh secondary 9 l 3 25 3
Highway secondary 6 l4 2 l 1
Busara Primary 7 2 20 l l
Kamukunji secondary l l4 l6 1
Muthangari Primary 4 14 20 2
Muthurwa Primary 8 14 23 l
Ofafa Jericho secondary 3 l l 23 9
St. Brigids Primary 7 22 29 0

 

 

 

Table 2. The choice between buda and mbuyu among the students

In this table, the column for ‘others’ represents tokens where variables other than mbuyu
and buda were used (see above for the other terms for ‘father’). We note that in both
Busara and Ofafa Jericho, the two dominant variables for ‘father’ are not unanimous;
there are many variables that fall outside the buda/mbuyu synonyms — Busara 11 and
Ofafa Jericho 9. Furthermore, it is only in Busara where the use of buda surpasses that of
mbuyu. We can explain the case of Busara to the fact that it is located in Kayole estate, a
more recent residential area, which is far removed from other estates in Nairobi. As a
result, older term for ‘father’ has not changed. In addition the high number of ‘others’
might indicate that the speakers have come with different words that were used in their
former residential areas. The high number of ‘others’ in Ofafa Jericho and Busara might
also be a pointer to the existence of other variables in these schools. Reverting to my
earlier argument that words are coined to serve certain purposes, then we can make the
assumption that different synonyms reflect the use of a variable by different categories of
speakers, or in different interactional contexts. If this assumption is correct, then we can
rightly conclude that the term for ‘father’ has more in-group use or different contexts of

use among the students of these two schools. This conclusion follows from the fact that

94

terms that are widespread can not serve as distinctive features of groups. Lexical
uniformity signifies shared linguistic norms which means that the speakers belong to the
same speech community in the classical speech community definition (e.g Labov 1966).
On the other hand, high lexical variability signals different norms or different speech

communities.

3.7.3. The age and neighborhood factor in variation

Age is an important factor in the study of linguistic variation because certain variants
seem to be used at a higher rate by certain age groups as compared to others. Sheng for
instance is usually claimed to be spoken more intensively by the adolescents and pre-
adolescents. The relevance of age in identity is that speakers of Sheng can be categorized
into broad groups that are identifiable. Furthermore, the assumption was that acquisition
of synonyms follows an age progression where speakers who began using Sheng at a
lower age have a high likelihood of using more variables than those who began to do so
at a later age, or those still in the process of acquiring Sheng. To establish the veracity of
this assumption, the respondents were asked to say at what age they began speaking
Sheng. The results in table (3) excludes data from respondents who claimed they begun

speaking Sheng below age 3 and after 17 years because they were judged unrealistic.

 

l0

9 ' :8T 10
12 15
2

     

Table 3. Ages when respondents claim to have started using Sheng

95

Table 3 shows that a majority of the speakers acquire Sheng between 3 and 12 years.
From the sample of 212 students for instance, 176 (83%) claimed to have begun speaking
Sheng between the ages of 3 and 12 years while 34 (62%) out of 55 respondents over 18
claimed to have begun speaking Sheng by age 12. By age 12, the majority of students
have already picked up the Sheng necessary for general communication. In table 2, we
saw that the primary schools, apart from Busara, do not display a lot of lexical variability
as compared to the secondary schools. If we take into account that most people join
primary school between the ages of 6 and 7, it follows that the majority acquire Sheng
during the 8 years in primary school. Therefore, by the time they join secondary school
between 14 and 16 years they have already acquired Sheng. It is also worth noting that
most primary school students tend to attend schools near their local neighborhoods, hence
the kind of Sheng that they speak is more localized, with few variants. Unlike primary
school students, secondary school students do not necessarily attend school in their local
neighborhoods. When students from different neighborhoods converge in the secondary
schools, different variables are brought into contact, which results in synonymy. We thus
expect more lexical variation in secondary schools as opposed to primary schools. That is
why in table 2 primary schools (apart from Busara) seem to have fewer tokens in the
column ‘others’ as compared to the secondary schools.

The secondary/primary distinction also makes sense in terms of age in that
language learners normally have fewer synonyms as compared to competent speakers.
The youngest respondent in primary school was 11 years while the oldest was 17; the
youngest respondent among secondary school student was 13 while the oldest was 19.

The average age of the primary school students was lower as compared to that of

96

secondary school students. The relevance of this age disparity is that younger Sheng
speakers (primary) exhibit little variation as they are still in the process of learning, as
compared to older Sheng speakers (secondary) who have already learnt. In addition, we
expect that transgressive behavior and sub-cultural elements that require obfuscation such
as discussing sex issues, smoking, and going out with friends are practiced more by the
older secondary school students as opposed to primary school students. This then could
partly explain the higher lexical variability among secondary school students as
compared to primary school. More studies focusing on the specific words that display
high variability would confirm this claim. It is also probable that secondary school
students had the academic edge in understanding interview instructions. As a final
comment, in this discussion on age and neighborhood factors, I note that although
schools, especially secondary schools, strive to teach the students standard languages,
they (schools) also provide the environment not only to learn Sheng, but also for people
who use different variants to compare notes. As far as Sheng is concerned, therefore,
secondary schools act like two-way traffic, where students bring with them the lexical
items from their neighborhoods. If other students find them cool, they adopt them and

take them to their neighborhoods.

3.7.4. Who are the innovators?

I have already discussed the different word formation processes that are responsible for
the high variation in Sheng’s lexicon. What remains is a description of where these words
are created and who creates them. While schools have their own kind of Sheng (Mbaabu

and Nzuga 2003), the most intense creation of words takes place outside the schools. The

97

manamba in the matatu industry (Mbugua 2003) are a good example. According to
Mbugua, Sheng constitutes an instance of hidden transcripts where the matatu crew
exploits its opacity to cover their antisocial behavior. Viewing the matatu phenomenon
within the context of a subculture, we realize that the members have every need to create
the words that are useful to mask their activities. Such activities include raising fares,
impressing female commuters and cutting off commuters in their conversations. Raising
fares without passengers’ knowledge, for instance, explains why we have such a high
number of enumerative variables for money, which facilitates their coded communication
(Mbugua 2003). Another group responsible for the high rate of innovation is the parking
boys” who, according to Spyropoulos (1987), invented various words for policemen
because of their constant confrontation with authorities. Looking at my list of lexical
items above, ‘police’ and ‘money’ have a very high rate of synonymy. It should be
pointed out that the fact that some groups seem to lead in innovation of certain terms does
not exclude other groups from innovating other terms for similar lexical items. Indeed, it
is the very cornerstone of my argument in this study.

Money terms further complicate the picture of variation in the variables that refer
to different denominations of money. I will provide a few examples that this study
collected, and which are corroborated by Gathigi’s examples (2001:16-17) and Githiora’s
(2003) glossary, as well as Sheng dictionaries such as Moga and Fee (1993, 2004) and

Mbaabu and Nzunga (2004).

 

27 See footnote l3. Nowadays, the concept of parking boy has been replaced with ‘street
boys’ or chokoraa (derived from the Indian name chokkra for ‘boy’ (Spyropoulos 1987,
Githiora 2002).

98

21 5 shillings ngovo, kobole

10 shillings ashu, ashara, imiku, kinde
50 shillings fije, guoko, jet, ndege, kaufifty, hamsa
1000 shillings mote, tenga, ngiri, kavu, ndovu, thao

Variation in enumerative terms for money also provides evidence of the use of Sheng in
various commercial activities. This lexical productivity can be attributed to the informal
nature28 of these commercial activities. Operating outside of formal boundaries, they are
not bound by the mainstream language of commerce. The variation of the activities as
well as the informal nature of these transactions corresponds to the variation and
informality of the linguistic mediums in which these transactions occur.

It is therefore obvious that the creation of words, to a large extent occurs outside
the school compounds. To explain this, we recognize that while school-life is
characterized by routines, life outside the school is more diverse and dynamic. Schools
can be viewed as an extension of the students’ social reality; or a temporary stopgap in
their daily transactions that involve interactions with significant others who speak Sheng.
This requires them to learn how to balance their different personas, which might either be
complementary or in conflict. While the social reality in their neighborhood demands that
they use Sheng to cultivate solidarity with their significant others”, the school imposes
another reality characterized by inhibitions through prohibition of vernacular usage and
other non-mainstream features of languages. Even in cases where the school might not

impose overt proscriptions, the language policy that prescribes a certain medium of

 

2" jua kali means ‘hot sun’ in Standard Swahili (see Samper 2003, Spyropoulos 1987) is
the common name given to the informal sector in Kenya. It derives from the fact that they
are conducted in the open under direct sunlight.

29 Kariobangi female respondents claimed that using standard Kiswahili or English
among their friends, some of whom were not well-educated, would be considered
showing off.

99

instruction as well as the social expectations has an impact on their level of mastery of
the linguistic codes spoken in their neighborhoods. The school reality remains a key
element of their complex persona and does not necessarily erase their identities as
members of neighborhood groups. In any case, they will eventually return to these
neighborhoods, unless they are lucky to secure white-collar employment or join the
university. It is with this in mind that I now turn to look at these neighborhoods that are
responsible for the existence of different variables in the lexical forms. I move away from
the schools for a brief discussion of the bazes where I conducted the recorded interviews.
I am guided by the logic that although the bazes are the main discourse site where
interaction in Sheng takes place, they are also innovation centers in their own right. I

begin by defining the concept baze.

3.8. From the estates to the bazes

Various studies and commentaries have mentioned the variation in Sheng according to
different residential estates or neighborhoods (e.g., Ngesa 2002, Abdulaziz and Osinde
1997, Githiora 2002, Samper 2002). According to Githiora (2002), the ‘estate’ in
standard Swahili is the same as mtaa which translates as ‘town quarter’ or ‘surburb’, but
in Nairobi it bears the narrower and more specific meaning of the low income city
council housing estates concentrated in the Eastlands, a section of the city that emerged
during the colonial period as a reserved area for African workers in the city. This study
deviates fi'om Githiora’s definition to include a broader definition in which the estate is
regarded as the ‘housing development’. Many suburban houses, especially those whose

genesis can be traced to the colonial period began as housing developments; so were the

100

former workers quarters. Today the need for a more gereral term has extended the use of
the term ‘estate’ to refer to any residential area with a certain peculiarity, whether in the
suburban or the informal settlements. The popular view is that each estate has its own
variety of Sheng, which is incomprehensible to residents of other estates. Some of these
claims go as far as implying that you can tell a person’s estate from their version of
Sheng. While not disputing the place of residential areas in the whole diagram of Sheng’s
variation within the general discourse of identity practices, I conceive of estates as
macro-entities comprising territorial or administrative categories that lack the social
cohesion necessary for the members to perceive themselves as cohesive social categories.
Conceptualizing a speech community in terms of residential estates is problematic
because members are not bound by the same linguistic norms, nor is it possible to build a
strong network necessary for enforcement of linguistic norms within estates. The nature
of unrestrained mobility, occasioned by the squalid conditions of the residential areas in
Nairobi, prevents residents from having deep emotional attachments30 to their estates.
This mobility means that linguistic habits that served in one residential area might not be
useful in a different residential area. In any case, it is not possible to place a linguistic
boundary between one residential estate and the next. In dialect geography, for instance,
the areas of convergence of two dialects are usually characterized by bundles of
isoglosses, to show the overlapping features of both dialects. Such overlapping of
features results in the vanishing of isoglosses (Chambers l995:150). This means that

insisting on inter-estate variation implies assigning clear-cut boundaries that goes against

 

30 Moreover, the majority of the city residents do not consider the city their home. To
them the city is only a temporary place to live while making money before retiring in
their rural homes.

101

the dialect continuum. As such, estates as the units for delimiting a speech community for
the purposes of analyzing speech behavior in Nairobi are not very helpful. Perceptual
inter-estate variation is probably a case of the Sheng continuum being overplayed, an
abstraction that require a closer examination to analyze its dynamics.

There is, therefore, a need to look at more realistic categories for the investigation
of speech behavior as far as the use of Sheng is concerned. These categories must be
local enough to allow for cohesion among members. They must also be fluid enough to
accommodate outsiders who aspire to join the group. Furthermore, they must be very
dynamic to ensure that the bonds that bind the members are not affected beyond repair by
the departure of some members, or admission of outsiders. These categories must,
therefore, be self-sustaining with a common engagement that acts as a reference point. In

this study, I identify these categories as the bazes.

3.8.1. The Bazes
Baze is a folk term that came to my attention during the course of my research. Members
of small groups that hang out in the local neighborhood comers referred to themselves or
their counterparts in the same or other neighborhoods as bazes. I define the baze as a
local ‘joint’ where friends meet to relax, or work, similar to the concept of the ‘set’ in
African Americans’street culture. Writing about the playing of dozens amongst African
American, Horton (1972) wrote;

The more or less organized center of the street life is the ‘set’ meaning both the

peer group and the places where it hangs out. It is the stage and central
marketplace of activity, where to find out what’s happening” (p. 21)

102

Horton’s definition of a set fits perfectly with the baze. Placing the bazes at the center of
the discourse practices of Sheng speakers enable us to observe both the group and
individual interactive practices. This category fits well into the analytical model that I
advocate in this study because it is small enough to allow a researcher to focus on various
discourse practices. The advantages of focusing on a small group were pointed out in
Bucholtz’ (1999) study of language and identity practices in a community of nerd girls in
a US high school. A small group allowed her to pay closer attention to some interactive
practices that would have passed unnoticed if a large group was involved.

It is these advantages that inspired my preference for a few respondents from each
baze. The bazes that I visited had between 3 and 8 members. Table (4) shows the bazes

visited in this study, as well as the age, sex and ethnic composition of their members.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Baze No: Age sex of resps. ethnicity

Buruburu 6 ‘ 23-27 males Luhya

Eastleigh 4 21-23 1 male, 3 females Luhya

Kabete 4 20-30 males Luo, Luhya, Kamba

Kariobangi (F) 5 18-21 females Luhya, Luo, Gikuyu
. Kariobangi (M) 4 22-24 males Luhya, Gikuyu

Kibera 5 18-32 males N ubian

Shauri moyo 4 20-23 males Luo, Kamba, Kisii, Gikyu

Makadara 4 23-28 males Luhya

Montecarlos 4 18-23 2 males, 2 females Luo, Kamba, Luhya, Gikuyu
figara 3 20 males Luo

Sinai 3 18-21 , females Gikuyu, Luo

South C 8 19-24 males Nandi, Luhya, Luo, Somali, Gikuyu

 

Table 4. Demographic composition of bazes

Table 4 shows that a typical baze in has about three or four respondents. The concise
information summarized needs to be expanded in order to show its relevance. I will only
describe three bazes for illustrative purposes. Like the discussion of synonyms above,
which mostly focused on primary and secondary school students, this description will be

another way of veryfying Hypothesis 1, which states that people of different social

103

groups use different lexical items as identity markers. The description of the bazes is
intended to show that members construct narratives of identity that reflect the realities of

their residential areas and their daily activities.

Sinai: The Sinai baze is located in the outer Western flanks of an expansive slum and is
also refered to as ‘paradise’. This term is a paradox when one reckons with its
inaccessibility, the general lack of hygiene, and the ramshackle structures in which they
live. The respondents in this baze were three young females Joy, Kui and Liz, all aged
between 18-21 years. They were all unemployed and made their living through selling
illicit alcohol and also engaged in petty prostitution. They were almost always drunk and
admitted to using marijuana regulary. They described their Sheng as ‘perfect’, ‘very
good’ and ‘good’. All claimed to have gone to secondary school, though it was not clear
whether they completed it or not. Kui admitted to having a child who was being raised by
her mother, and said if she became pregnant again, she would procure an abortion. These
respondents seem to have resigned themselves to their fate and have accepted struggling

as part of their lives.

South C: This baze is located in the middle class residential estate. It is comprised of
young jobless males who normally hang out near a barbershop constructed on a road
reserve. Some of them could even do hair. Although this was their convergence point,
they all had different engagements and schedules, with some being members of some
private clubs. Three had attended school outside Nairobi, and one claimed to have

traveled out of the country for further education. Unlike the Sinai respondents, they were

104

living with their parents, which explain the mama’s boys’ label given to them by the
Sinai respondents. All had good command of English and mostly rated their Sheng
mastery as ‘fair’, though some claimed to speak Sheng a lot. South C brings in a mixture
of people from different ethnic groups as can be see in Table 3. This is because ethnic
dominance is not a feature in such estates, and in any case, ethnic languages are not used
a lot. This might also explain why there is little innovation considering that most of the
lexical innovation is inspired by ethnic languages. The affluent backgrounds of South C
members manifest itself in the kind of Sheng they speak, and its deviation from other

varieties, especially those spoken in the ghettoes.

Shauri Moyo: This baze is located in the Shauri Moyo estate. The respondents were all
‘self employed’ as car washers, which is a feature of Kenya’s informal sector known as
jua kali. All attended primary schools in Nairobi, but one attended an upcountry
secondary school, which is not uncommon in Nairobi. Not only are upcountry schools
cheap, but some parents in Nairobi think it is important to send children to school in the
rural areas in order for them to experience their cultural roots. In addition, lack of school
upgrading has meant that the city schools do not have adequate space to accommodate
the ever rising number of students. The Shauri Moyo members had a high degree of
solidarity, which can be attributed to the fact that they were always together while
working. When asked whether they would consider moving out of Shauri Moyo, they
claimed that it was not likely, but even if they did, they would always come back to see
their friends. This solidarity extended to their linguistic practices, where they considered
themselves the best. Their discourse was mostly influenced by their occupation activities,

both in terms of setting as well as lexical innovation. Hanging out together, whether

105

working or not, resulted in the establishment of a close-knit network, thus making it easy
to retain their Sheng norms. That is why one speaker (Stevo) dominated the interview

without being inten'upted or contradicted by his collegues.

Describing all the peculiarities for each baze would be too big a task, but I hope that the
few I have described will be sufficient to show that the bazes were located in various
places throughout the Nairobi neighborhoods. Coming back to Table 3, it needs to be said
that the baze normally has more that three or four members. I chose bazes with three or
four people because I reasoned that it would not be possible to carry out a coherent
interview with many subjects. The majority of the members of these bazes belong to the
Luhya, Gikuyu, Kamba and Lao ethnic groups. This is not surprising considering that
these are some of the largest ethnic groups in Kenya, as the section on Kenya languages
discussed in Chapter 2 illustrated. When Sheng borrows from the ethnic languages, these
four languages are the most important donor languages (Mbaabu and Nzuga, 2003
introduction, Githiora 2002). The summary of the composition of the bazes in Table 4 is
crucial because lexical innovation and the subsequent construction of identity is
influenced by ethnicity, sex, age, housing, occupation and socioeconomic background, as
I will demonstrate later.

Baze members refer to each other as ma-beshte (singular beshte) — a truncation
of ‘best fiiend’, which has been phonologically manipulated. It is when Mabeshte are
socializing in the bazes that the intensity of interaction and social bonds are cultivated
and maintained. Some of the social activities that take place in the bazes are drinking

either legal or illicit alcohol, smoking cigarettes or marijuana, and chewing khat or

106

muguka — green twigs used as stimulants. In various bazes, khat also goes by the name
veve or gomba; and chewing khat is called ku-chana, ku-veveka or ku-chonga. The term
kuchonga seems to have gained in popularity since its competitor like ku-chana has been
used for a long time and ku-veveka is too transparent, which has weakened its opacity
advantage. Sheng’s lexical items are used as the ‘official’ language of interaction during
these social activities. Sometimes these activities constitute the subject of their discussion
or a joke among the baze members. Other topics run the gamut of general existential
issues and include personal experience narratives dealing with police escapades, the
merits and demerits of certain things, memorable parties or adventures, rides in matatus,
sexual exploits, deaths of friends and encounters with criminals. Sometimes, traditional
stories are retold in Sheng, while in other times, well known stories are retold with
modification to suit the discourse norms of the group.

Bazes are not just restricted to hanging out places. The Buruburu baze for instance
doubled as a hang out joint and a small business. As a barbershop, it provides a very
convenient place for the Mabeshte, who are jobless, to meet. Since most do not have
fulltime employment, they come to the baze to relax when they are not working. When
they are out of a job, the baze becomes the place where they learn of places where they
can get part time jobs. At the same time, when the barber has more clients than he can
handle, they can assist him to cut hair for commissions. Similarly, the Sinai baze doubles
as a workplace as well as hang out joint. The girls who hang out in the alcohol dens for
leisure also engage in petty prostitution as an economic activity. The Shauri Moyo and
Ngara bazes were purely occupational, while the Eastleigh, Kariobangi, Kabete, South C,

Kibera and Makadara bazes were ‘estate comer’ hang-out. Although Montecarlos is a

107

reggae dancehall, I regard the respondents I interviewed there as part of a baze because
they admitted to being regulars who constantly interacted. However, the fact that these
revelers come from different neighborhoods in Nairobi should not be lost.

It should be pointed outthat bazes are very fluid and members do not necessarily
belong to one baze. There are cases where members belong to more that one baze. For
instance, one member of Buruburu also doubled as a member of Kabete baze, his former
place of residence. Similarly, another member of Buruburu baze was also a member of
Kariobangi baze, his residential neighborhood. To these two members and probably
others, Kabete and Buruburu baze are convergence points just like the school where
different lexical variables compete for space in the members’ discourse. As much as they
introduce lexical material from their respective bazes, they also carry with them the

lexical material that they have acquired to their respective bazes.

3.8.2. The relevance of bazes in lexical variation

The thrust of this study as seen in Hypothesis 1 is that different bazes employ unique
lexical items that are relevant to their different everyday realities. These lexical items
constitute the core linguistic norms within the discourse practices of the Mabeshte who
make up the baze. The word formation processes discussed earlier afford Sheng speakers
a resource for virtually infinite lexical coinage or the modification of pre-existing terms
when situations demand. Mastery of these norms is the major criterion for distinguishing
those at the core or the periphery in a given baze. We can view the lexical items as well

as meaning peculiar to a given baze as network markers" akin to the phonological

 

3 l The term is used here to refer to the linguistic features and social norms that make a
certain group distinct and are hence regarded as symbols of the group’s identity.

108

variants [a] and [s] in Belfast (Milroy 1980, Milroy and Milroy 1985, 1992). Only the
core members with strong orientation towards the baze have full access to these network
markers and their various realizations because of constant participation in the discourse
practices of the baze.

We are thus able to see how individuals negotiate their positions within the group
through lexical choices or the meaning attached to lexical items. As a classic example,
Eckert (2000) reports about how meaning was spontaneously negotiated during an
interview with a group of seven high school students in California. When asked about the
skull insignia, which was a symbol of their identity, they said it symbolized ‘death’.
However, one boy said he thought the skull’s symbol meant ‘pirates’. The ensuing
discussion attempted to make the connection between ‘death’ and ‘pirates’ in order to
accommodate his views. This showed that the boy was a peripheral member of the group,
a pointer that although people might share a similar identity, we do not necessarily have
uniformity since members participate differently in groups’ practices. Such cases are
unavoidable due to co-membership in different bazes as well as the different position that
different members occupy within the group hierarchy.

Bearing in mind that some of the activities in the bazes are illegal explains the
secretive nature of the core members of that baze and the adoption of opaque lexical
items. In a baze where members smoke marijuana for instance, a newcomer has to know
the entire stock of the disguise terms that members use when communicating about their
indulgence in the presence of outsiders. An outsider who needs to join the baze regulars
in the smoking of marijuana must quickly conform to the baze members’ collective

linguistic values as a show of loyalty to the baze. These outsiders should not be regarded

109

as mere conformers but also as the agents who, in their own way, modify the identity of
the group since they come with features of their previous groups.

Gaining admission into a baze is, however, not automatic. An outsider needs to be
introduced by a well-known regular. Even so, complete acceptance takes time and
personal efforts must be made by the outsider seeking admission to master all the
communicative norms of the baze in order to fit into the baze. Baze regulars assess the
sincerity of a person before he/she earns their full trust, as illustrated below in the
comments of my female respondents in Kariobangi:

1. PG: ...kama mimi nimekaa katika mtaa tofauti hapa, na mmeniona leo ni mara
ya kwanza nikija pengine hapa kwenye kikundi chenu, ee, mnaweza
mkanikubali kuwa mi ni mmoja wenu hata tukizungumza sawasawa?

Shiro: Ahh, itatake time,
PG: itatake time?

Pame: ee lakini with time tutakukubali tu. Ata wewe utaanza kufeel comfortable.
Shiro: tena inadepend na njia umecome nayo

.U‘PP’!"

The English translation:

1. PG: ...like me I have stayed in different residential areas here, and you have
seen me for the first time today coming here probably to your group,
would you accept that I am one of you if we speak the same way?

Shiro: Ahh. . .it will take time

3. PG: it will take time?

4. Fame: yes, but with time we shall just accept you. Even you, you will begin to

feel comfortable.

5. Shiro: again it will depend with the way you have come

1"

From this conversation, we understand that an outsider will have to interact with other
members of the baze as s/he acquires the network markers identified with the baze. In the
process s/he relinquishes part of his/her former identity (albeit temporarily) and exposes
himself/herself to the hegemonic influence of the baze members’ linguistic norms. We

explain these respondents claims within the communities of practice (e.g Eckert and

110

McConnell-Ginet (1992, 1999), Bucholtz (1999) and Eckert (2000) where one has to be a
members and participate in all other activities, even the non-linguistic ones in order move
to the core of the group. The communities of practice, according to Ettiene Wenger, one
of its pioneer, is defined as ‘a group of people who share a concern or a passion for
something they do and learn how to do it better as they interact regularly’. He lists the

three components that define a community of practice as:

1. What it is about—— its joint enterprise as understood and continually renegotiated
by its members.

2. How it functions — the relationships of mutual engagements that bind members
together into a social entity.

3. What capability it has produced— the shared repertoire of communal resources
(routines, sensibilities, artifacts, vocabulary, style, etc.) that members have
developed over time.

With regards to the bazes, we have seen that the baze-specific variation in Sheng results
from the different shades of reality that pervade the narratives of the baze members.
These narratives are laced with linguistic artifacts manifested by their lexical innovation
that results from their shared enterprise as members negotiate meaning. Living together,
going to school together, working together, suffering together, looking for jobs, smoking
and drinking are just some of the mutual engagements that bind the members of the baze
together. Their linguistic artifacts cannot be retricted to the issue of synonymy, but also
include the words that are used in different discourse types. This is important because the

different bazes conduct their discourse in different times and spaces. Consequently,

shifting of identities by baze members through the use of lexical material associated with

111

other bazes, as a result of contextual variation, signals participation in other communities
of practice at one time in their lives.

Still, it is important to emphasize that although mastery of linguistic code is
important, sharing communication code is only one part of the identity project. There are
other non-linguistic identity practices that also need to be taken into account. These will
be discussed later in the next Chapter, where I will demonstrate how multiple identities of
the members are activated and negotiated in the discourse practices of the haze members.
But first, I will look at some other elements that are crucial in the diagram of Sheng’s

variation.

3.9. Other factors responsible for variation in Sheng

Earlier, I pointed out that the various word formation processes in Sheng are a resource
exploited by Sheng speakers in enriching its lexical stock. The majority of the coined
words do not just happen; rather, they mostly result from the relexicalization of previous
lexical items that constitute the raw materials from which new vocabulary is constructed.
This follows from the fact that the previous linguistic habits of the speakers (especially
with respect to their respective ethnic languages) have been internalized. These
internalized linguistic behaviors influence word coinage and manifest themselves in the
kind of lexical items preferred by certain groups of speakers. The factors that drive the
peculiarity of lexical items in various bazes have their background in social, economic,

residential, occupational and ethnic bases.

112

3.9.1. Socioeconomic factors: The south C baze

In Chapter 2, I pointed out the socioeconomic disparity between the Eastlands and
Westlands areas. According to Abdulaziz and Osinde (1997), the linguistic code in the
Westlands is Engsh while Sheng pervades the Eastlands. However, I point out that these
two are just a matter of categorical generalization. Strict demarcation between the
Eastlands and the Westlands only applies in the original map of Nairobi before the
expansions (see figure 2 in Chapter 2). In the current map, there are slums in the West
side of Nairobi such as Kangemi, Kawangware and Uthiru. Likewise, we have plush
residences in areas east of the CBD. One such area is the South C estate, one of the bazes
visited in this study. I expected their variety to be Engsh, but this was not attested in the
structural sense, I did find some terms in this baze that were deeply influenced by

English, hoever, consider the following examples:

 

 

22. South C baze English commonly used word
yumbus father buda, mbuyu
spleng girl manzi, shore, dem
shoot to go ishia
frush bad thing
sneed bad thing
chums money doo, ganji, chapa, munde

If we follow the idiosyncratic categorization of Sheng varieties along socio-economic
lines, we would expect South C to follow along Engsh lines. This study only found out
this influence in their word coinage. For instance, in the Swahili-based structure, words
are forced to abide by the Swahili word formation rules that follow the CVCV
phonological structure. Regarding the morphosyntax, the nouns display a class marker —
nominal root ——- aflix order while verbs follow the subject marker — tense — verb root

—— verbal extensions order. Consonant clusters are also not very common in the Swahili-

113

based structure; however in the examples I have given above, we find word-final sibilants
such as [8, s], nasal [13] and also coronals [t, (1]. That is as far as Engsh — the English-
based structure goes. In narratives, we find the Swahili morphosyntactic template in
operation as in the following exchange at the beginning of a story between FF and MM:

FF: ebu washow vile ilienda.

MM: sa inakuwa niaje Friday tunameet hapa nyuma baze. Tulikuwa maboys

watatu.

F F 2 tulikuwa gomba si ndiyo?
Translated as:

FF: just tell them how it went

MM: Now it was just on Friday we were meeting here at the baze. We were

three boys

BB2 we were chewing khat isn’t it?
In this brief excerpt, we witness the syntactic structure as well as some word formation
processes that follow the Swahili template. These include dropping of [h] in ebu (just),
codewitching from Swahili to English wa-show ‘tell them’, tu-na-meet ‘we normally
meet’, ma-boys ‘guys.’ We also witness the phonological reduction in sa ‘now’ from the
Swahili word sasa. This is a clear demonstration that the Swahili structure is very much
alive. To explain the ambivalence of these respondents, we notice that in South C, the
residents are from the middle class families with better access to the standard language.
The prevalence of English vocabulary in their discourse practices shows that although
they adopt the Swahili structure, they still set themselves apart from other children in the
slums by using these English-based lexical items. I do not completely dismiss the concept

of Engsh in South C, but I claim that in the above example it is only manifested in word

structure.

114

Whether these youths represent the general trend is not obvious. We see that they
are all unemployed youth. They refer to themselves as ‘anti-Mababi ’, the term Mababi to
them meaning their rich employed counterparts. It is possible that the employed youth in
South C use English structure in their sentences, though I have no data. Use of Swahili
structure as well as English-based words represents the dilemma facing the South C
subjects. On the one hand, they identify with other poor unemployed youth in the
ghettoes, but, on the other hand, they are born into a middle class privileged lifestyle.
This represents a conflict of identities, which is well represented in their linguistic
practices. The uniqueness of terms used in South C and other affluent residential areas is
evidence that socioeconomic factors are crucial in lexical coinage. I have attributed this
to the advantage these residents have in being able to affOrd quality instruction in English

as opposed to the residents of lower class neighborhoods.

3.9.2. Residential factors

Most respondents in the lower class residential areas claim that Sheng from all the
ghettoes is similar, and that they can easily adopt features of Sheng used in other bazes.
Still, they agree that there are some words that they use which are unique to their
respective groups. The Kariobangi baze for instance, is a run-down residential area whose
inhabitants belong to lower social economic class. The subjects (girls) have different
words, which are peculiar to their baze. If we look at the words they use for ‘girl’, they
are very different from those used in everyday Sheng. The words below were not attested

in other bazes, even among the other females interviewed in this study.

115

23. Kariobangi (I) baze English common Sheng

msupa/chakachaka girl manzi, shore dem,
mafinishing/madusko butt rasa

kizee boyfriend chali

jiweka eat dema, dish, manga
matinye/machungwa boobs dashboard, matuzo
mbaru/maflai police gava

In addition, the words githeri mboga and waitherero are also used by the group to refer to
a ‘girl’. Chakachaka is a South African songbird who is very popular in Kenya, and a
role model to lots of girls. This label is consciously selected to point to a very positive
image of themselves that the girls would like to project. Mafinishings for ‘butt’ might be
an idiom of beauty as in ‘finishing’ — something that finishes a piece of work as in
artwork or handcraft, and to which great attention is paid. It could also be that ‘finishing’
just refers to the rear part that one can see as a person passes by. Compare this to the
Shauri Moyo baze (all males), who use the term chombo cha kazi ‘tool of work’, a rather
derogatory term for the same reference. The term machungwa ‘oranges’ for breast was
not attested in any other estates. 1 will juxtapose these uses of referential terms when
dealing with the construction of gendered selves in the next Chapter. Still, I need to add
that the fact that there are some terms that are peculiar in certain neighborhoods does not
preclude the use of other terms used in other neighborhoods. However, it is the
peculiarity of these terms that shows how neighborhoods can be important factors in

lexical coinage as well as discourse practices.

3.9.3. Relevance of occupation in Shengs variation
The lexical items employed by the Shauri Moyo car washers are the clearest

manifestation of occupational influence in lexical variation. One example is the word

116

weng for ‘wheel’ mentioned earlier, which can be attributed to their daily occupation that
deals with vehicles. No other baze would require the use of such specialized vocabulary.
Other words that display the relationship between lexical items and occupation among
Shauri Moyo respondents are mawamba ‘water’ and mapapa ‘undercarriage of a car’.
From this, we can conclude that their discourse revolves around their daily activities such
as whether they will get a car to wash, the kind of water to use on a car, places to wash
and so on. The lexical items below are a selection of occupation-related terms to

demonstrate how economic activity influences the linguistic practices of Shauri Moyo car

washers.

24. ma-papa undercarriage of a car
ma-weng car wheels
ku-duwacha work
ku-funga kazi close work
ku-pamba do a shoddy job

The coined words among the Shauri Moyo guys help communicate meaning in the course
of their work. Take ku-funga kazi for instance, the literal meaning of the term is ‘to close
work’. However, in their interaction, it does not mean closing work in the transparent
sense, it usually mean getting a lot of money from a certain job as payment
compensation, which is enough to take care of the whole day; worth a normal day’s
earnings. Consider also the word ku-pamba; if we get rid of the infinitive marker ku-, it
means to ‘cover’ or ‘adorn’ in Swahili, but in Shauri Moyo baze, its meaning is accessed
at the metaphorical level. It exploits the metaphor of covering to mean doing work for
show only while actually doing a shoddy job hence ‘metaphorically covering’ the eyes of

the person who gave them the job.

117

3.9.4. Ethnicity as a factor

The dominant ethnic group influences local varieties. According to Eckert (2000222), if a
specific ethnic group dominates the local marketplace, local forms will derive from the
ethnic repertoire. This has also been pointed out by Githiora (2002), who noted the heavy
influence of Gikuyu on Sheng varieties in Githurai and Uthiru areas. If an ethnic group is
dominant in a certain neighborhood, the influence of its language on the Sheng variety
spoken there is enormous. Since ethnicity is not subject to negotiation, it constrains the
kind of identity practices that go on. In Kibera baze for instance, there is heavy use of
Swahili in their Sheng coinage, which can be attributed to its ethnic composition. My
subjects in Kibera were all from the Nubian ethnic group, the majority of whom are
Muslims. I have already pointed out the affinity between Kiswahili and Islam, such that
to accept Islam is to accept Kiswahili, as argued by Habwe (2000). Kibera, being a
Muslim enclave”, is likely to exhibit this trend. Interestingly, the construction of ethnic
identity among the subjects in the Kibera baze is not conducted using an ethnic language
like Nubian as expected, but through Kiswahili, the national language, due to its religious

importance. The following words will serve as examples:

 

25. Kibera baze English common Sheng
jaza gauge/zuia eat manga, dema
sheikh father buda, mbuyu
zama/ruka to go ishia
manga steal/eat manga
msumbiji, maangola thieves dingo/gondi

 

32 Other examples of Muslim enclaves in Nairobi include Kawangware and Majengo. In
these Muslim enclaves a coastal version of standard Swahili is spoken. Indeed some of
the Shauri Moyo respondents, while saying that their Sheng is the best in ‘the whole of
the world’, claimed that they admired the Majengo version of Sheng.

118

In Standard Swahili, jaza means fill, hence they use the word jaza gauge for ‘eating’, or
filling the stomach. On the other hand we notice zama ‘drown’ and ruka ‘jump’ are used
for ‘go’. Although the ‘eat’ reading of manga is widespread in all bazes, its ‘stealing’
reading was only specific to Kibera. Notice also that Ma-Angola ‘the Angolans’ and
Msumbiji ‘Mozambique,’ are names commonly used to denote people who are always
fighting. Msumbiji recalls the F relimo movement in Mozambique’s war of independence
and has been appropriated in Kenya to denote a dissident group that was fighting to
overthrow the government especially in the late 1980s”. The metaphorical use of these
terms to denote anti-social activities is quite striking, and in my study only attested in
Kibera. This semantic shift from Standard Swahili shows that although subjects in Kibera
baze also use lexical items widely used in Nairobi, it is possible to stylize others who do
not use the Sheng terms that they use or with a meaning different from theirs. Ambiguity
in Sheng and its place in identity will be revisited in the next Chapter.

To sum up this section on factors that drive lexical variation in this study, we bear
in rrrind that identity practices in different bazes are manifested in the nature of discourse
types, which is in turn dependent on the social reality of the members of these bazes.
Social and economic influences surface in the lexical coinage by subjects in the South C

baze as a result of their privileged status, which endows them with the linguistic capital

 

33 This is what the president said in a public gathering, but according to Danson Esese
(personal communication) the Msumbiji issue was a very complex one. Basically, it was
a gang of well-organized robbers with very sophisticated weapons. The group mostly
operated from Luanda in Western Kenya and was composed of unemployed people and
former police and military officers who had been dismissed from the service. The group
terrorized the area residents in the middle of the night and the police seemed to be
helpless when it came to dealing with them. It is possible that when the Kibera
respondents used this term, they used it in the sense of thives and not the political
connotation associated with it.

119

of the authoritative language —— English. Residential areas also come into play because
different residential areas have their own different realities. If we take language as a
social construct, then it is not surprising that lexical coinage is subject to the dictates of
the social environment. In addition, occupation as a part of the people’s daily reality
provides a background that results in different lexical terms. Finally, we find that Sheng
borrows from different languages, and is highly accommodative to new lexical items. In a
multilingual environment, it cannot avoid the infiltration of words borrowed from the
ethnic languages of the speakers. All these factors do not act in isolation and are
characterized by various instances of overlap. For instance, the coinage of novel
vocabulary by subjects of a certain residential area might be as much the result of ethnic
composition as by the occupation or socio-economic orientation of the members.

In this Chapter, I have attempted to present the diagram of Sheng’s variation. I
began by discussing the templates in which Sheng as a linguistic code is based. 1 also
discussed variation in Sheng by showing how lexical items are created. From selected
lexical items, variation that bred synonymy was attributed to the linguistic word
formation processes available to the speakers of Sheng. Synonymy was also attributed to
the fact that there are convergence points, such as schools and some bazes, where the
different lexical variables are used. Although there were dominant variables for every
lexical item, the existence of other variables, though not very widely used, was regarded
as a pointer to the use of these words among different group of speakers. Finally, other
factors that promote lexical variation in different bazes were examined. These factors

were tied to the word formation processes discussed in earlier section of this Chapter.

120

In the next Chapter, I will return to the theme of identity construction by
discussing both discourse and non-discourse activities that are involved in the identity
projects. I will argue that the multifaceted identity of Sheng speakers in Nairobi, and in
different bazes in particular, is a combination of their social practices, some of which are
non-linguistic, and that linguistic features are overt manifestations of how these practices

are negotiated and preformed.

121

Chapter 4

Lexical choices, discourse and other discursive practices in the
Construction of identity

4. Introduction

In the previous Chapter, I outlined variation in Sheng. I traced the variation in terms of
structure, demonstrated specific lexical variation, and explained how such variation
comes about. I then focused on the speakers and argued that the reason variation arises is
due to the fact that the speakers come from different areas of Nairobi and belong to
different social categories. In this Chapter, I demonstrate how lexical items function as
‘identity cards’ (Macaulay 1997) of members of a given baze. Taking bazes as the
reference points, I will discuss how individuals, both insiders and outsiders, negotiate
their place, either at the core or the periphery of the group, through conscious or non-
conscious choice of certain lexical items. I will discuss the concept of negotiation and
positioning within Bell’s audience design framework (1984, 1991, and 1999) that looks at
how speakers modify their speech to approximate that of the audience with whom they
wish to identify.

I begin by looking at words as linguistic signs and discuss the relevance of
semantic and lexical variation in identity in Section 1. This Section will look at variation
in lexical usage according to age, gender and other different categories and how different
lexical items are given the signification that draws the ‘us’ vs. ‘them’ distinction. In
Section 2, I will discuss how certain labels are used in different bazes with different
senses, and how an out-group term can be used positively within an in-group. Section 3
will discuss the performance of multiple identities in spontaneous discourse and how and

why speakers of certain baze-specific varieties of Sheng accommodate to other varieties.

122

In Section 4, I will discuss how different identities are performed through the use of
vernacular discourse. I will argue that vernacular discourse is an in-group activity
because it involves shared background that may not make sense to an outsider. Section
five will examine other non-linguistic ways of constructing identity, which act in
combination with linguistic varieties to complete the identities of Sheng speakers. In
Section 6, I will discuss other discourse sites such as the intemet, popular culture, and
mass media, and how Sheng is used in these alternative discourse sites to construct
certain identities. Section seven will conclude the Chapter by discussing the relationship
between people’s attitudes towards other varieties of Sheng and their identity. I

summarize the Chapter in Section 8.

4.1. Lexical items as linguistic signs and symbols

Saussure (1959: 16) conceived of language as a system of signs in a semiological system.
He conceived of the linguistic sign as the union of a concept and a sound image (65-67).
Rather than looking at the sound image in phonetic terms, he saw it as a psychological
imprint of the sound or the impression it makes on our senses. Put‘ differently, the sound
image is the signifier while the concept is the signified. Saussure also pointed out that the
relationship between the signifier and the signified is arbitrary. Just as one signified can
be represented by more that one signifier (synonymy), a single signifier can be associated
with more than one signified (ambiguity) as will be discussed below. In fact, this

discussion of ambiguity will serve as my verification of Hypothesis 2.

123

4.1.1. Ambiguity, bazes and identity
Like synonymy — semantic relations where different wirds have the same meaning
which I discuseed in Chapter 3, semantic relations in which a single word bears different
meanings (ambiguity) are importants marker of identity also. If a baze can construct its
identity by using different structural elements, it can also construct its identity by using
different semantic elements. In this section, I will examine how a similar form is
interpreted differently in different bazes. I begin with Samper’s (2002) examples of
magong ’o, shore and sanganga. According to Samper, the meaning of these words varied
according to estates;
26. magong’o a stupid person, male or female, in Kayole
a lazy person in Umoja,
a stupid girl in Mbotela
shore a cute girl in Kayole,
an illiterate girl in Jericho,
derogatory for stupid girl in Majengo
sanganga buttocks in Majengo,

gold in Umoja

[Umoja, Mbotela, Kayole, Jericho and Majengo are residential estates in Nairobi’s
Eastlands (some of these estates are not shown in figure 3)

Out of Samper’s three words, the only word that I obtained was shore, and it failes to
display the semantic variation that he noted. Instead, the Ngara respondents claimed that

shore was a term used in Ubabini. However, I got other ambiguous words such as;

27. mubatari policeman as used by manamba,

manamba as used by Sinai girls

manyake women (South C) women and/or butt as used in Ngara
butt and/or female genitalia as used in Sinai

manga eat and /or sexual intercourse, especially as used in Sinai
steal as used in in Kibera

kizee boyfriend as used in Kariobangi and Sinai
father as used in Shauri Moyo and Kibera

ghetto slums as used in the majority of bazes

house as used in Kabete

124

According to Mbugua (2004), policemen are referred to as a-batari (the needy people) by
the matatu crew because of their persistent habit of extorting bribes from them.
Ironically, the same matatu men who are regular clients of the prostitutes in Sinai are
referred to as the needy. Nyake (meat in Sheng) is derived from the Swahili word nyama
with similar meaning. With the addition of the pluralizing prefix ma-, its semantics has
changed over time, initially referring to big butts before the meaning extended to women
with big butts. The popularity of the song ‘Juala' by Circuite and Jo-el extended its
meaning further to female genitals. Manga, a Romance borrowing, is a widely used term
for ‘eating’, but the Sinai women used it to refer to having sex. The sense ‘stealing’ by
the Kibera respondents was new to me and to the best of my knowledge only unique to
Kibera baze. Ki-zee“ is a modification of m-zee, a Swahili word for ‘elder’ or ‘old man’.
In Kenyan upcountry Swahili, m-zee is used by women to refer to their husbands, and by
the children to refer to their fathers. It seems that in the course of the innovation of this
term, its extended semantics were transferred such that the use of kizee for boyfriend
mirrors the use of m-zee for husband. While ghetto generally refers to the ‘slums’, its
‘dwelling place’ sense in the Kabete baze is hardly surprising, considering that Kabete is
an informal settlement where most of the houses are in all respect ghettos. We also note
that m—dosi and dongera, the two terms that normally refer to ‘boss’ or a ‘rich person’,
are also used to mean ‘father’ in South C. This can be understond in the context that most

of the fathers in South C are indeed rich, literally.

 

3" Ki is a class 7 nominal prefix (see footnote 4 on nominal prefixes). While this nominal
class is generally referred to as the class of ‘things’ it also includes all the Swahili
diminutives formely in nominal class 13. In this case, mzee, a class 1 animate, goes to
class 7 as a diminuitive. Sometimes, diminutives are used to indicate the primitive state
of something; but they are also be used in an affective sense.

125

The cases I have cited might not all be considered as typical cases of ambiguity. This is
because a lexical item might not be ambiguous in one specific baze; however, this still
presents a problem to an outsider who wishes to gain admission in a baze and might use
the term with a completely different signification. In fact, ambiguity is a very useful
devise, because where different senses for a word exist in the same baze, outsiders will
easily be cut off from a discourse when ambiguous terms are used to signify something
other that what outsiders expected. For the baze members themselves, however, familiar
discourse and shared meaning disambiguate polysemous words. Since meaning making is
a common engagement tied to the group’s identity, ambiguity will reflect different

hierarchies depending on how group members decode ambiguous words.

4.1.2 Age determined interaction

Since Sheng is extensively associated with young persons, I begin the investigation by
discussing the relationship between variation and age. The two important issues are
whether the use of Sheng exhibits an age dependent pattern, and whether there are lexical
varieties associated with different age groups. Previous studies (e.g., Samper 2002,
Kangethe 2004264) have pointed out that Sheng is used as a language of youth identity,
but they also add that its age domain is always expanding (Githiora 20022160, Kangethe
2004264). In this study Shiro, a respondent from Kariobangi, claimed that her parents
understood and spoke Sheng. Likewise, Kui, a respondent from Sinai, claimed that her
grandmother was very competent in Sheng. This calls for an investigation of the

categorical differences in the use of Sheng according to age.

126

In order to know who the students normally spoke Sheng with, I asked them to underline
all that applied from the following list: friends, brothers and sisters, parents, teachers,

policemen, anybody (see Q 8 appendix). Below is the summary.

Friends 189 out of212 or 89.1%
Brothers & sisters 128 out of 212 or 60.3%
Anybody 25 out of 212 or 11.8%

We see that 89% of the students speak Sheng with friends and 60% with their kin, but not
with parents, teachers or policemen. Only a small percentage (12%) claimed to speak
Sheng with just anybody. These results are exactly what we anticipated. We assume here
that the fiiends and brothers and sisters belong to the same age group. Recall from
Chapter 3, that my student respondents’ ages range between 11 and 19 years, but we
expect parents, teachers and policemen to be much older. When I asked students to
underline the people with whom they would not interact in Sheng from the list of parents,
teachers, friends, policemen and nobody. I got the following results:

Parents 130 out of212 or 61.3%

Teachers 133 out of 212 or 62.7%

Policemen 105 out of 212 or 49.5%
Although we can extend the argument above that the majority of the students would not
speak Sheng with parents, teachers and policemen due to age difference, I prefer to view
these data by what they do not show directly. It is very significant that nearly 40% of the
students (N=212) would speak Sheng with their parents or teachers, while approximately
50% would do so with policemen. This might seem to counter my argument above that
use of Sheng is restricted to people of similar age groups and that people of different age

groups do not interact in Sheng. However things become clearer when we recognize that

127

speakers exploit Sheng’s variation to interact with different categories of people
according to the prevailing contexts.

Use of Sheng across age categories can be explained by exploiting Samper’s
(2002) distinction between deep and basic Sheng. If basic Sheng only requires bilingual
knowledge of English and Swahili, then it is not surprising that parents and teachers
speak Sheng, at least the basic version. When the students say that they speak Sheng with
these people, they probably refer to the Swahili-English code-mixing that characterizes
the everyday speech practices of the Nairobi people rather than the specialized in-group
variety that they normally use among close friends. According to Samper, the primary
purpose of deep Sheng which contains very little English and is highly coded is to
obfuscate and confuse non-group members (p. 8). On ' the other hand, basic Sheng
‘employs lexical and phrasal codeswitching and slang without the complex semantic
shifts, relexification and syntactical changes’ (p. 8). The examples below from Samper
(2002) distinguish deep (a) from basic (b) Sheng:

28. a. Na-taka ku- m- kat- i- a hu- yu vi-pepe

lsg-want inf-om-cut-appl-fv dem-lsg 8-butterfly
I want to flirt with this girl (p. 6)

b. I’ll do everything hata ma-sikio ni- ta- ku-ramb-a
I’ll do everything even 6sm-ear lsg-fut-om-lick- fv
I’ll do everything; including licking your ears. (p. 7)
It is also probable that the parents generally use a variety that has gone out of fashion,

since they have moved out of Sheng networks, which in effect alienates them from the

institution of Sheng’s innovation. As a result, they are unable to keep up with new

128

innovations and remain in generational or social ‘relic areas’”. The concept of oldskool,

which I discuss below, further clarifies this age-determined variation in Sheng.

4.1.3. Signifiers of identity: frozen in oldskool
It is possible in many societies to tell the generation of a person by looking at the kind of
slang words that a person uses. This also applies to Sheng, where words fall into disuse as
soon as they become widely known (Mbugua 2003). New generations of Sheng speakers
coin their own vocabulary to reflect their users’ reality. Words of yesteryears are no
longer regarded as ‘cool’ since they have been neutralized through constant use. Just the
way it is possible to identify a Sheng speaker’s residential area by the lexical items that
s/he uses, it is also possible to tell one’s generation by paying attention to the lexical
items that a person uses. KJ’s quote below in Samper (2004) illustrates this fact;
“[y]ou never realize just how much ‘Sheng’ changes until you meet that shady
guy talking the ‘Sheng’ of the eighties!” (KJ. 2003) then the shady guy is quoted
as saying to a young fellow “...hiyo moti ya ponyii ilikacha mbio kama dush
ikanipita kama na-buy puruu. Ilikuwa inafukuza mgodii ule alikuwa ameiba bano
[‘that police car dashed like bird and passed me buying sweet. It was chasing the
thief that stole bananas] (p 147, translations mine)
In the quote above, moti (clipped from motor-car) seems to have gone out of fashion and
is replaced by dinga, ndae or murenga. Ponyi (policeman) has also gone out of fashion

and is replaced by the more popular terms mabeast, karao, or magava. In addition mgodii

(thief) has other variants like mdingo, and mgosa (see my list of lexical variations for

 

35 Areas excluded from new innovation. Old linguistic features persist in spite of the fact
that new forms are used in the surrounding areas (see Downes 1998). The term is used
here to refer to the social isolation of older folks who are unable to master the innovative
Sheng features being used all around them by the youth.

129

‘car’ ‘thief’ and ‘police’ in Chapter 3). Such people, who continue to use words that are
out of fashion, are considered oldskool.

Some lexical items that are considered oldskool in one baze can still be in use in a
different baze. Consider buda and mbuyu, the two leading terms for ‘father’. In Chapter
3, we saw that mbuyu was used by 56% of the students compared to its competitor buda
at 29%. Buda then can be considered to be drifting towards oldskool, at least as far as its
use among school students is concerned. When we come to the bazes themselves, we find
that some bazes have a preference for certain terms. In Makadara baze for instance buda
is the preferred word for ‘father’, but respondents in Ngara baze preferred mbuyu instead
of buda, saying; ah ‘buda’ hiyo ilikwisha ‘ah buda, that one died out’. When I pointed
out that Makadara people were still using buda, they said it depended on the place and
when they acquired it. The existence of some different lexical items for a single referent
can be attributes to the fact that some people remain frozen in oldskool.

Coming back to KJ’s claim above, we note that the Sheng speakers who know
about current Sheng and who have abandoned oldskool terms in favor of the trendy ones
will easily draw the distinction between themselves and those who retain oldskool terms.
Using an oldskool term in a baze points to the position one occupies in a baze in a
particular interactional context. One risks losing face for not being trendy through
continued used of the oldskool terms, as Mosh, a respondent fron the Ngara baze does
when he gives the oldskool word moti for ‘car’:

Josh: kuna ndai,

PG: kuna lingine? (pause)

Mosh: moti

Josh: aa duh! (laughs) hapana moti ni ya kitambo
Mosh: lazima... (unclear)

130

Otish: moti ni go
Josh: moti ni go, hapana, tunaita tu ndai

Translated as

Josh: there is ndai

PG: is there another one? (pause)

Mosh: moti

Josh: aa duh! (laughs) no, moti is an old one

Mosh: it has to... (unclear)

Otish: moti is gone

Josh: moti is gone, no we only call it ndai
I had asked them to give me the names they used for car and although it was Mosh who
originally suggested ndai, he was later derided by his Mabeshte for suggesting moti, a
word they no longer use. In this instance then, he can be regarded as having moved to the
periphery of the baze and taken the identity of an outsider, most likely an older person.
His attempt to explain is blocked by his mates who say that they only use ndai. He is able
to reclaim his face when I asked for their other terms, and he quickly volunteers a term
and is supported by his Mabeshte. This episode demonstrates that one can shift from
‘new school’ to ‘oldskool’, but such shifts that deviate from a group’s interactional norms
must be checked because they threaten the prevailing identity of the group. A group that
uses oldskool terms risks losing their esteem among other Sheng users, as was the case
with the Shauri Moyo respondents who regarded the variety of their Makongeni
counterparts as ‘childish’ and ‘weak’ because ther Sheng variety contained words that
they (Shauri Moyo) had already abandoned. It is in this context that we view Josh and
Otish’s rejection of Mosh’s oldskool term moti.

Oldskool also expresses a very outdated claim as regards Sheng. In the cartoon

strip Head on corrishon (Oct 3 2004), K] draws a caricature of an old man standing on

the railroad track in front of the on-coming “Sheng express” train shouting frantically

131

‘stop this thing’. An arrow indicates that this mbuyu (father, old man) is oldskool. The
reason mbuyu is referred to as oldskool is because he has fallen behind the times.
Opposing Sheng is an outdated and useless engagement that is associated with old
people. Sheng as a reality in Kenya cannot be stopped or wished away. KJ later makes
the following statements in Vile 2naifeel ‘the way we see it’
Hakuna lugha mavijanaa wana understand na wanaweza ku-communicate nayo
bila pressure ya Sheng! Hiyo ni reality [there is no language the youth can
understand and can communicate without pressure apart from Sheng! That is the
reality]
Kwamba hii ni lugha ambayo mizizi yake ishajikita kwa hii society na hakuna
mtu anaweza kui-wish away [that this is a language whose roots are deeply fixed
in this society and nobody can wish it away]
Lyrical revolution inanuka na oflicial language yake ni Sheng! Itakuwa ni poa
kama hautajiweka mbele ya hiyo wave isije ikaku-sweep away! [lyrical
revolution can be scented and its official language is Sheng. It is better if you will
not put yourself in front of that wave lest it sweeps you away!]
T ume-learn hii lugha mtaani. Mtu yeyote angetaka tujue lugha ingine hakuwepo
wakati mtaa ukitufisndisha hii lugha.[we have learnt this language in the
neighborhood Anyone who had wanted us to learn a different language
wasn ’t there when the neighborhood was teaching us this language]
In these cartoon strips, KJ normally identifies himself with ma-vi-janaa36 (the youth). He
portrays the reality about Sheng and then shows that Sheng is unstoppable, just like the

old man cannot stop a moving train. Pleas to ban Sheng were constantly expressed in the

l980s and 1990s when Sheng was so stigmatized, but such calls have waned. Therefore a

 

36 Vi-janaa is a Swahili noun which is already pluralized with the class 8 prefix. Futher
pluralization with the class 6 prefix renders it a case of double morphology. From Myers-
Scottons 4-M model (2002) double morphology in codeswitching occurs because the
system morphemes of the donor and receipient language are activated at different levels,
some early and others late. This one is a clear illustration that even when the words
externally appear as Swahili, the identity of the morphosyntax cannot always be regarded
as an indicator that a single system is in operation.

132

person who still thinks they can stop Sheng is using the reasoning of the 80s and 908,
hence the label oldskool.

I have used the frequency of variants for mbuyu and buda to determine what is
oldskool or not; however, I must point out that this is not always a good diagnostic test.
Current Sheng words that are used in mainstream discourse, though not oldskool, are a
threat to group identity. Considering that Sheng is valued by its speakers for the
transmission of coded messages, we cannot always say that what is trendy in general is
also trendy in different bazes. I quoted Mbugua’s claim that Sheng words fall into disuse
as soon as they become widely known. What is trendy at the mainstream level does not
imply it is trendy in the in-group circle. Indeed what is trendy at the mainstream level
might be oldskool for the group. Consider the word for ‘to go,’ where 133 students out of
212 or 62.7% gave ishia as its Sheng equivalent, but most South C respondents claimed
that ishia was oldskool. The fact that the South C baze preferred to use kukaa teke,
kudondoka, kushoot and kuinua not only support the claim that identity is constructed in
smaller social units, but also that popular Sheng terminology, though not oldskool, cannot
function as identity markers of small groups.

The case of Mosh’s use of moti cited above should serve as a reminder that
however innovative a social group is, oldskool will always be lurking in the background.
When defending himself from the accusation of using oldskool terms by his buddies, a
South C respondent said “oldskool haiwezi kosa...yaani huwanga” ‘you can’t avoid
oldskool. . .it is always there’. The reason for oldskool surfacing in speech interactions can
be attributed to the fact that people need it to interact with their interlocutors who are still

frozen in oldskool. These include people such as brothers and sisters or parents who do

133

not belong to their networks. The surfacing of oldskool terms in dynamic groups indicates
the fluid nature of the bazes, and also the multiple identities of baze members, who have
to interact with people outside their bazes in a variety of contexts. We must also bear in
mind that some baze members are themselves former users of those oldskool terms when

they were in vogue.

4.1.4. Gender signifiers: revisiting variation in gender terms
Like age, sex is one of the more important categories in language choice studies. I noted
above that males relish constructing the category of women with many descriptive terms,
which are in most cases derogatory; we also find, however, that the women do not allow
themselves to be defined by the men all the time. They give themselves more positive
labels, which keep them still within the Sheng discourse. The signification that the males
attach to their labels is modified when women apply them. We thus note that while there
are terms that cut across gender lines, there are others that are more apt to be employed
by a certain gender.

One popular view is that women have been at the periphery of Sheng discourse.
This would explain why we have had so many terms that refer to women while we have
very few that refer to their male counterparts. In this study, I asked primary and
secondary school students to give me the names they use for mothers. There were 27
names for mother as compared to 18 for father. Some of these variants are not important
since they simply deviate from other terms in a minimal way. This could be due to the

fact that the young people who speak Sheng construct parents almost the same way. Were

134

the parents participants in Sheng discourse, then there could have been variation arising
from how men construct women through the labels they use and vise versa.

This changes when we look at the labels for boys as compared to those for girls.
There were 45 variants of the term for girls as compared to 23 for boys. This is a clear
pointer to who does the labeling in these cases. Interestingly, most variants for ‘boy’ are
derived from generic vocabulary, while those for ‘girl’ are mostly descriptive, some with

very negative connotations. Below are variants’ frequencies for ‘boy’:

 

Word Variant’s frequencies

Boy beshte l, boi 6, boy 1, boyboy l, boyii l, bro 1, chali
178, charty 1, dume l, hommie 1, huyu boy I, jamaa 8,
kaboy 1, kijana 1, licha 1, mdhi 1, mm 1, msee 3, mthi l,
muthee l, muthii l, mzee 1, ndaume l

 

 

 

 

 

Table 5. Frequency of variants for the word ‘boy’

Since Sheng lacks a standardized grammar, the orthography is usually left to the speakers
whims. As as result, we end up having variants which are not variants at all. If we assume
that very similar items result from the mismatch between phonology and orthography,
then a real picture of the variant will emerge by careful reclassification of these labels by
collapsing the phonologically or morphologically related words.

(1) boy (11) boi, boy, boyii, boyboy, huyu boy, kaboy
(ii) muthii (3) muthii, mthi, mdhi
(iii) msee (3) msee, muthee, mzee
(iv) chali (179) chali, charty

(v) dume (2) dume, ndaume.

(vi) beshte (1)

(vii) bro (1)

(viii) jamaa (1)

(ix) kijana (l)

(x) licha (1)

(xi) mm (1)

(xii) hommie (1)

135

This leaves us with only 12 lexical variants. The term chali by 179 respondents is almost
unanimous, and there is only a slight attempt at syllabic inversion in licha, which could
be grouped with it. Apart from muthii and msee, which in Sheng refer to ‘people’, and
beshte ‘fiiend,’ all other terms are ordinary borrowings without change in meaning. This
can be seen in the Swahili words dume ‘male’ jamaa ‘person’ and kijana ‘boy’as well as
bro and hommie which are most likely borrowed from the AAVE/hip-hop terms bro and
homey. Hommie might as well be from English slang homie, disguised by the addition of
an extra m. There is no information about the origin of the term mm (and it might be a
misspelling of Swahili mtu ‘person’). However, we can say that, generally, the terms for
‘boy’ are positive or neutral and probably come from the boys referring to each other.
Otherwise, how would you explain the use of mzee ‘old man,’ a Swahili title of respect?
We note that no terms bear any negative connotation. In stark contrast, names for girls
are legion, and very diverse, even if we attempt to collapse them. This can be seen in the

variant frequencies in Table (6) below;

 

Word Frequency

 

Girl azguoba 1, bogola l, buss l, chibela 3, chik 1, chile 3, cholosa 2,
dame 3, dem 36, gala l, gwaru l, hoe l, jaba 1, kemfaa 1, kenta l,
kff 1, kichuna 1, kidalipo 1, mama 1, man 1, mandu 1, ma-nyake 2,
manzi 132, mbas l, mbus 5, miogar 1, m-losho 1, m-resh l, m-roro
13, m-she 3, m-shi 5, m-susu 2, m-tasi 1, m-toi 1, m-toto 2,
nyambizi 1, nyang'we l, nzima l, roro 1, shore 17, shwala l, siste 2,
supuu 1, toto 1, wanyina 1

 

 

 

 

Table 6. Frequency of variants for the word ‘girl’

In the terms that refer to girls, we still retain a large number of variables, even after
applying the rule of collapsing phonologically close ones. Only 5 can be collapsed; these

are:

136

mbus (7) mbus, buss, mbas

mtoto (4) mtoto, mtoi, toto
mroro(14) mroro, roro
mshe (8) mshe, mshi

dem (39) dem, dame
After collapsing these terms, we are still left with 33 variants, with manzi being the
lexical choice of 132 respondents. Dem with 39 tokens comes in a distant second while
shore with 17 tokens comes in third. In a majority of cases, a variant appears only once,
indicating the intensity of coinage or the variation within speakers’ codes. If this is the
case, then we can say that while there is a variety spoken by the majority, variation that

cuts across speakers of different gender categories exists.

4.1.5. Definition by the other: derogatory gender metaphors

Coming back to the lexical construction of gender, we find that even though we might
lack the resources to decode the semantics of most of these variables, the fact that they
exhibit such diverse variation from the ordinary terms for ‘girl’ in mainstream languages
is telling — that coining of terms for girls is an ongoing process. Unlike the terms for
boys, terms for girls are descriptive and are not the general terms. The in vogue label
manyake for instance, is derived from the former Sheng word nyake, itself a derivation
from Swahili nyama ‘meat’. Thanks to the song manyake by the local rap duo of Circuite
and J o-el, the term has spread with a variety of meanings. At times it may refer to plump
women who have a lot of flesh. However, it is also used to refer to a woman’s butt. At the
extreme, it refers to the female genitals. Whether hoe is ‘whore’ in disguise or colloquial
hoe is derogatory because it bands all girls together and comments about their sexual

promiscuity is not an issue because it is derogatory no matter how it is construed. It is not

137

clear whether reference to girls as mtoto, ‘baby’ is affectionate in the diminutive sense or
derogatory.

Interesting too is the way girls are metaphorically constructed through the
generalization of one aspect of their physique. The butt and the breasts are the two parts
of a woman’s body that are labeled with a large number of variables. Manyake above
gives a good illustration. In the Kabete baze for instance, cianda is a Dholuo borrowing
that refers to butt, but its plural counterpart (macianda) could refer to women. In the
excerpt below, the Kabete respondents tell the researcher that labels for women are

descriptive.

l. Karis: ...unajua kwa hawa wasichana utakuta. . .unajua ile shape yake ndiyo
itam-describe
. PG: enhe
. Karis: ...hata pengine ako na titi kubwa utasikia watu wakimwita Tuzo.
. PG: Tuzo?
(laughter)

5. PG: Tuzo ni yule ana matiti?
6. Karis: mmh

(laughter)
7. Dau: yaani inatokana na 81 unajua hii factory ya maziwa hii ya Tuzo?
. PG: 000 000 000 000!
(laughter)
9. Dau: . . .ako amebeba the whole factory. (laughs)

43MB)

00

Translation:

1. Karis: ...you know for these girls, you will find that their shape is the one that
will describe them

2. PG: enhe?
3. Karis: ...like maybe she has large breasts, you will hear people calling her
Tuzo
4. PG: Tuzo?
(laughter)
5. PG: Tuzo refers to the one with breasts?
6. Karis: mmh
(laughter)
7. Dau: that is, it derives from, you know this milk factory called Tuzo?

138

8. PG: 000 000 000 000!
(laughter)

9. Dau: she has carried the whole factory (laughs)
Having been out of the country for some time, I was unaware of the existence of T uzo
milk factory. My inability to make the obvious connection aroused giggles from the
members of this group. The similarity between the term matuzo and KCC (Kenya Co-
operative Creameries), another giant dairy company, as used by the Kariobangi and other
respondents in Nairobi, is quite striking. The fact that physical characteristics play a big
role in the way women are perceived shows how gender is constructed by the men in the
Sheng speaking areas. The Ngara subjects who used the term matyre from the English
word tyre to refer to an undesirable girl validate this argument. The metaphor of the car
tyre is used to comment on bulging flesh.

With this wealth of terminology, we can say with certainty that the category ‘girl’ is
in most cases constructed by the males in their discourses rather than by the females
themselves. Male dominance in the use of Sheng has also defined how the women who
speak Sheng are perceived in society. Commenting on the gender disparity in the use of
Sheng, Samper (2003: 154-155) wrote:

Men derive cultural capital from the use of Sheng (see below), but women lose
those benefits. Sheng for men is a source of interpersonal power, and women’s
avoidance of it reflects women’s lack of power in Kenyan society. Friedman and
Todd 1994 argue that women in Kenya lack personal power because girls are
socialized to be silent, subservient to men, and not to voice their needs. Moreover,
the socioeconomic condition in Kenya and the enduring strength of traditional
gender roles make it impossible for women to change their current low status of
their interpersonal power. Sheng could change this perception, but the patriarchal
symbolic order closes this avenue of access to symbolic power by equating

women’s skillful use of Sheng with prostitution and low moral values. In these
terms, Sheng does not carry the same advantage for women as it does men.

139

This asymmetry explains the reasons behind women’s low participation in Sheng, which
results in a virtually one way coinage of gender-descriptive terms. But women need not
cede their right to define themselves passively. They also try to alter the existing power
relations by creating an identity different from the way they are categorized by the men
through the coinage and use of labels that portray their identity more favorably. In the
next section, I will discuss how women contest these categorized identities by paying

attention to some of the lexical terms for different parts of women’s bodies.

4.1.6. Rejecting negative identity by contesting derogatory gender labels

Using positive labels for self-reference within an in-group is one way of contesting the
negative labels by the out-group. I have shown how women are given many terms by
their male counterparts, many of which are derogatory. I will now show how women
contest the negative identity given them by men by using labels that are more acceptable.
When talking about the peculiarity of lexical terms in certain bazes, I pointed at selected
labels that focus on certain parts of their anatomy. I will here look at the terms for ‘girl,’
‘butt’ and ‘breasts.’ In table 5, below, I show the terms used by the women compared to
the ones used by the men.

Most of the names for ‘butt’ are borrowings from the Swahili or local languages,
though we have some that are borrowed directly from English. The majority of these
names imply massiveness, either the admirable size or a despised one. Also important is
that the majority of these terms refer to the female butt while the male butt goes by the
widespread term rasa, although rasa can also refer to female butt. Women generally do

not have many names for ‘butt’; we note that when they use ma-tina, borrowed from

140

Gikuyu with similar meaning, it lacks the negative connotation inherent in the labels used
by men. However, when the Gikuyu slang ma-thutha or Swahili slang ma—nyuma is used
to refer to butt, they become gender circumscribed as well as laden with negative
connotations. In Chapter 3, I said that ma-finishing could be viewed in an aesthetic sense
— delicate and beautiful, as in finishing in furniture or artwork. It might as well be
literal, i.e., the rear of the woman’s body. However, we note that whichever way we look
at it, it does not bear as negative a connotation as the other terms used by males. The
pervasiveness of the prefix ma-37 in these terms is striking

Notice too that the women preferred to use maduda and madusko for ‘butt,’ both
derived from the Gikuyu slang mathutha. Phonological modification has yielded one of
the variant maduda. A root madu- is then extracted from maduda, which is then suffixed
with -sko to derive the other variable madusko. This seems to strip ma-thutha of the
negativity it bears when used by the males. Table 7 below presents some of these terms

with their source and meaning, as used by different genders.

 

’7 The ma-prefix — according to Ashton (1969265), some f the semantics of the ma-
prefix are (a) the names of things that occur in quantities but which may be thought of
singly, (b) words expressing an amplicative idea and (c) words that express a collective
idea. We may assume that the derogatory sense of these terms follows from the
augmentative semantics of the ma-prefix. But augmentations may not always be negative.
Think of the terms Mabeshte for ‘best friends’ or ‘close buddies’ where augmentation is
actually positive. The semantics of the words must therefore be taken on a case-by—case
basis, especially considering that there are other semantics of the class 6 nouns.

141

 

Name Women (self) labels Male terms (other) labels

 

Butt ma-finishings, ma-tina [G ma-tyre, ma-nyake [S. meat], ma-cianda
butt], ma—duda, madu-sko, [Dh. Butt], carrier, mudguard, nyungu [S.
[GS ma-thutha, butt] pot], ma-sitting allowance, chombo cha

kazi [S work tool], ma-thutha [GS. butt],
ma-nyruna [S. behind], manyu [see
nyuma], ma-ketho ma-loso [Dh-butt?]

 

 

Girl waitherero [proper name] ma-nyake[meat Sh], m-toto [baby], first
chakachaka[proper name], lady, kitu yangu[S my thing], ma-ketho
msupa [unknown] ma-loso [Dh but?] , wifey, ma-cianda [Dh

butt], matyre

Breast ma-chungwa[S oranges], KCC, ma-tuzo[milt factory], number

 

 

 

 

ma-tinyo [S matiti=breasts] plate, ma-earphones
[Keyz Sh = Sheng, S = Swahili, Dh = Dholuo, G = Gikuyu, GS = Gikuyu slang]

Table 7. Names for ‘girl’, ‘breast’ and ‘butt’

 

In Table 7, the names for ‘girl’ by the male respondents show that the names for ‘butt’
are usually extended to refer to ‘girl’ or ‘women’. Invoking African aesthetics to account
for this, we recognize that many Africans considered women with big butts as beautiful.
This conception of sexiness has prevailed, although the western concept of sexiness is
affecting the Kenyan urban youth. If a woman’s sexual appearance is concentrated in one
area of their anatomy connected with sexual appeal, then extending this meaning to cover
the entire woman reflects the fact that the men, or at least the Sheng speaking males who
do the labeling, mostly perceive women in sexual terms. We could expect the female
genitals to be used in similar terms, but the taboo on mentioning human sexual organs
inhibits the extendibility of these lexical items. The butt, being the most visible part

receives this extra attention in terms of negative labels.

While the terms in Table 7 are selected from a wide range of variables, it is
obvious that women are usually perceived as people without individuality. Take kitu

yangu for instance; in Swahili it means ‘my thing,’ which implies possession. Whenever

142

it is used by the males it denies them their humanity. Wifey from English ‘wife’ also
implies possession by males, i.e, as girlfiiends or wives, and not as women. First lady
may seem to be a positive label, but on a closer look, we realize that the first laay is
supposed to operate under the shadow of her mighty husband, doing everything for his
glory. Looked at in this way, the label is no longer positive. One might argue that kitu
yangu is a diminutive and is used in an affectionate sense. I have no evidence to dispute

this, but if so, then this would be one of the few exceptional cases.

In contrast, women give themselves labels derived from proper names;
Waitherero from Gikuyu, and Chakachaka, a popular South African musician who has
inspired many. Use of proper names reclaims personality, by focusing on the entire self
rather than one part of their anatomy. I have no explanation for msupa at the moment,

except the fact that it is very different from what the males use —— supuu (now oldskool).

Finally, we come to the names for breasts. Women seem to have accepted the
name dashboard just like the men because it does not seem to bear any negative sense.
Number plate is also likely to go this way once it gains widespread use. However, we
find that women do not use matuzo (private dairy company) or K C C — a government co-
oporation) to refer to their breasts. This is probably because they are derogative terms
used to refer to large breasts. I mentioned that the men have for a long time emphasized
the butt for its sexual appeal, but very large breasts are usually not regarded as sexy, as
captured by musicians Nonini and Juacali in the track “Vile Tafanya” that extols the

6

preferred shape of beauty for a girl as... awe na kifua ndogo, na nyuma mahaga’ she
should have a small chest, and big butt at the rear’. In a culture where huge breasts are

not preferred, augmentation that equates them with giant milk factories is therefore not so

143

polite. It is thus understandable why women contest these terms by not using them.
Instead, we find the women choosing a name like machungwa (oranges) —- an
aesthetically appealing metaphor through reference to color, shape and size. The other
variable used by the women to refer to breast is matinyo or matinye. These variables are
derived from the Swahili word for breast matiti (singular titi). Phonological reduction
results in mati-, which is then suffixed with -—nye or —nyo for the purpose of obscuring the
original meaning. I must also add that since matinye or matinyo are derived from Swahili
standard vocabulary with similar meanings, they also lack the negative connotation that

we find in the derived terms coined by males.

The choice of positive terms by the women is not similar to euphemism, but it is a
way of contesting the identity that has been thrust upon them by outsiders of the opposite
gender. As they contest these terms, they reclaim their individuality and define their

identity in more favorable terms.

4.1.7. Positive self-defining signifiers

Amelioration of labels applies to socially organized groups as well. In the introduction of
their Sheng-English dictionary, Mbaabu and Nzuga (2004: vi) write that some schools
chose good names for themselves, whereas others are given less glorious names by
students of other schools. They also noted the generation of proper names, especially in
the names of various residential areas. I will only give the cited names of the estates

relevant to this study;

29. Isich Eastleigh
Jeri Jericho
Kasa Kasarani
Kibich Kibera

144

In addition, the following names were used in the bazes where I collected data; Bongo
and Bangla for Kariobangi, Buru for Buruburu, Monte for Montecarlos. I did not get the
names for Kabete and Sinai from the respondents. Very few groups gave themselves a
name that was different from the name of their respective estates. The Shauri Moyo car
washers called their baze M0. W. At the first mention, the Kibera respondents, rather that
using Kibich, referred to Kibera as Motherland, while Makadara respondents called their
baze Harlem in addition to the widely used Mahadu. In the Shauri Moyo case, M.O.W,
which are the initials for Ministry of Works, is quite fitting and identifies the respondents
with work, in this case washing cars. In Kibera, derived from Kibra (bush or forest) in
Nubian, the respondents were all Nubians whose parents or grandparents were settled on
the then isolated bushy area at the periphery of Nairobi by the British colonial
government. The successive governments have never issued them with the title deeds to
guarantee ownership of those plots, which means that their occupancy can be contested at
any time. The choice of the tag motherland is a way of creating an identity that contests

the marginalization of their ethnic community by the mainstream institutions.

4.2. Inter-group categorization: comparative labels

In his distinction between the sign and the symbol, Saussure (1959268) argues that unlike
the sign, the symbol is never wholly arbitrary, since a natural bond exists between the
signifier and the signified. While words are usually regarded as arbitrary linguistic signs,
negotiated conventional signification may remove this arbitrariness, which makes them
function as symbols. I am interested here in the symbolic use of descriptive labels to draw

the distinction between the insiders and the outsiders. Since meaning is negotiated at the

145

local level, the symbolism attached to different labels as far as in-group and out—group
dualism is concerned varies from baze to baze. The out-group labels are negative and
derogatory as compared to in-group labels, which are normally favorable and positive. In-
group labels might symbolize the exact opposite when applied to an out—group and vice
versa. In the following discussion, I will look at symbolic labels such as Mbabi, beshte,
Wanati and fala and discuss how they are juxtaposed to signify different identities.
Evidence from their discourse will be used to reveal more about the identities that

different respondents attach to these labels.

4.2.1. Socioeconomic symbolism: The Wanati vs. the mababi dualism

The ghetto people normally refer to themselves as Wanati (plural for nati) derived from
English “natty”. Its counterpart Mababi/Wababi (plural for babi or Mbabi, see Chapter 1
definitions) refers to the rich people, soft people or weaklings. These two labels are
borrowed from the Rastafarian movement and were transmitted through reggae music
that dominated the airwaves before the advent of the hip-hop music. The biblical analogy
of the suffering of the children of Israel under the yoke of the Babylonians resonates well
with the urban poor. Taken within the context of Kenya’s political economy that displays
unequal distribution of wealth, the urban poor is the disadvantaged group. In a recent
survey titled ‘Pulling Apart: Facts and Figures on Inequality in Kenya’ by the Society for

International Development (2004)”, Kenya was reported to have one of the biggest gaps

 

38 The full report can be accessed online at
http://www.sidint.org/Publications/Docs/pulling-apartpdf

146

between the haves and the have-nots. It ranks among the top ten most unequal countries
in world, and 5th in Africa (p. 7)

The vicious cycle of urban poverty is sustained by this unequal distribution of
resources. The children from the ghetto have no access to quality education, which limits
their chances in life. The yoke of Babylon can be seen as a metaphor of oppressive living
conditions similar to the biblical captivity of the Israelites. In such a harsh environment
the Wanati survive by their wits, going through great pains to put bread on the table. This
explains why the word ‘suffering’ comes up constantly in their discourse. On the other
hand, the affluent Wababi, having been used to the life of ease and comfort, know no
suffering. This was well-articulated by the Sinai girls when I asked them to describe who
was a Mbabi according to their definition. Giving the examples of residential areas of

Westlands and Kileleshwa, they said;

1. Joy: unajua hao ni mababi. . .ma...mahomie sasa. Juu unajua watu wa ghetto

ni Wanati.
2. PG: mh?
3. Kui: wacha nikupe sasa for example, unajua mtu wa ghetto ni mtu anajitegemea
4. Joy: ee
5. Kui: ...lakini mtu Mbabi. Ni mtu anategemea mamake, kila kitu ni mamake,

lakini unaona huyu wa ghetto, wewe hata hushughulikangi na mamako,
hukulangi kwa mamako, mamako hajuangi unalalanga wapi. . .kitu ka
hiyo

6. Joy: Wanati ni wale watu wa ghetto wamezoeanga ku-suffer

Translated as;

1. Joy: you know those are the mababi. . .ma. . .mahommie now. The ghetto people

are Wanati

2. PG: mh?

3. Kui: let me give you an example, you know a person from the ghetto depends
on herself

4. Joy: yes

7. Kui: ...but the Mbabi person,... is a person who depends on his/her mother.

But

147

you see this one from the ghetto, you do not bother with your mother, you
do not eat at your mother’s, your mother doesn’t not know where you
sleep. . .something like that

8. Joy: Wanati are the ghetto people who are used to suffering.

Despite the fact they these girls engage in petty prostitution as well as selling illegal
alcohol, they take pride in the fact that they are independent of their parents. From their
deprived form of existence, any kind of moneymaking activity is valid. With their poor
education, very few options are open to them, and selling illicit alcohol as well as
engaging in commercial sex offers them an avenue of making a living. They perpetuate
the vicious circle of poverty that they have inherited from their mothers, who abdicated
their family responsibilities because of lack of means to raise their children in an
atmosphere of hope. Unfortunately, this makes them vulnerable to sex predators and
criminals, who expose them to hard drugs to ensure that they are eternally dependent on
them. After turning them into zombies, these criminals can then inflict all forms of
physical abuse on them with impunity. Yet the girls are resigned to their fate and stick
with these abusive ‘boyfriends’ because they are their only means of daily sustenance.
Independence from the parents is therefore not independence at all.

Comparing the rough life of the Wanati to the comfortable life of the Mababi explains
why Wanati view the Wababi as soft or weak in terms of both physical and psychological
strength. Hardness or toughness to withstand suffering has covert prestige among the
Wanati because you can only survive by having the strength to persevere. The symbolism
of Wanati is consistent across the board; but the label babi has various symbolic

meanings attached. I will now look at some of the different senses that go with this label.

148

4.2.1.1 . A person from Ubabini

Anyone from Ubabini (an affluent neighborhood) qualifies for the label Babi or Mbabi.
For this reason my respondents from South C fit the bill, although their babi identity is
overshadowed when compared with others from the more prestigious neighborhoods like
Muthangari, Lavington, Runda and Spring Valley. This residential making should be
used with caution because of the presence of domestic workers who live with their
families in servant quarters. The role of domestic servants in the spreading of Sheng and
Swahili in affluent neigborhoods has been mentioned in Abdulaziz and Osinde (1997).
Recall also that in Chapter 3, I mentioned that Muthangari Primary School is located in
the affluent Muthangari area, but it serves the children of domestic servants and students
from the surrounding slums. Identifying a Mbabi on the basis of his or her residential area

can sometimes be misleading.

4.2.1.2. People in white-collar employment

The South C respondents rejected the label Mbabi and instead referred to themselves as
anti-mababi. This was rather interesting considering that South C is an affluent
residential area. Kids from those areas attend good schools; one of them even claimed to
have been in Italy for education. Being unemployed, they considered themselves as the
lowest among their peers in the same residential area. Their conception of a babi was in

reference to their rich employed peers.

149

4.2.1.3. Mom ’s boy

In contrast, the girls at Sinai considered the South C respondents babi since they lived in

an affluent estate. However, they added another dimension in their definition. Their

conception of mababi as people who relied on their mothers makes it possible to have a

Mbabi who does not necessarily hail from Ubabini.

4. 2. 1.4. Just anyone you aspire to be economically

I informed the Shauri Moyo respondent that the South C respondents do not refer to

themselves as mababi, and instead used the term to refer to the employed people or

people from Westlands. Stevo opined that the term babi referred to those people better

off than you economically. In this conversation, I avoid my back channeling in order to

better capture Stevo’s argument.

PG:

Stevo:

PG:

Stevo:

Pish:

Deno:

Tamslation:
PG:

Stevo:

Unajua nimekuwa South C, na hata wao nikaona jamaa wanasema
wanazungumza juu ya mababi wakisema wale watu wa kutoka Westlands
ni mababi!

...ok its like this, society yetu imekuwa hivi. . .society yetu always
Una-aspire to. . .to go another level ...so mtu yuko that level una-aspire
kwenda kwake ni babi. . .that’s why pia hao utasikia wanaita watu wa
westlands babi. . .ukienda westlands wanaita watu wa Runda

Babi. . .eeee. . .unaona?

yeah

Inaenda hivyo

lakini sasa kwetu sasa vile sii hujua. .. because us 3181 pia. ..

hao wote ni mababi,

You know I have been to South C, and even them, I saw that they
say...they talk about the mababi, saying the people from Westlands are
mababi

...ok it’s like this, our society is like this. . .our society you always aspire
to. . .to go to another level...so the person at the level you aspire to go is
babi. . .that is why you also hear them calling people from Westlands babi.
If you go to Westlands, they call people from Runda babi. . .eee, you see?

150

PG: Yeah

Stevo: it is like that

Pish: but now to us now the way we see it. . .because we also...

Deno: all of them are mababi
Stevo’s argument exploits social stratification and the human endeavor to better oneself.
It is similar to claiming that every babi has his/her babi. However, to the majority of the
Wanati, the criterion is anyone who is better off than you are, hence Pish and Deno’s
assertion that all of them are Mababi. This stems from the fact that the Wanati occupy the
lowest rung in the social ladder. That said, we must also recognize that different ghettoes
can also be ranked. However, just like the South C respondents contested the label babi,
the respondents from the better off ghettoes also contested the label Wababi even when
the contrast is so obvious. Korogocho for instance is an informal settlement of lower
standards compared to Kariobangi, still, the female respondents from Kariobangi baze
objected to being called Mababi by Korogocho people in the exchange below.

PG: Sasa wakati watu wa Koch huko wanasema nyi ni mababi mnasikiaje?

Shiro: Ni uduh juu juu wewe ni mtu wa ghetto kaa hao

Edee: Ni wa ghetto unakula tu ile shida wanakula

Shiro: Eee

PG: Sa kwa nyinyi, Wababi ni kina nani?

Edee: Tuseme watu wa. ..

Shiro: Watu wa flato

PG: Now when Korogocho people say you are mababi, how do you feel?

Shiro: It’s nonsense, because you are a ghetto person just like them

Edee: we are ghetto, you experience the problems they are experiencing

Shiro: Eee

PG: Now to you who are the Wababi?

Edee: we have said people from...

Shiro: People living in flats
Shiro and Edee reject the label Wababi because they do not live in flats. This affirms

Stevo’s claim that babi is the person you aspire to be socioeconomically. When we

151

compare the difference in terms of the socioeconomic standards and general lifestyle
between Kariobangi and Korogocho, we find minimal differences; however, Kariobangi
enjoys better infrastructure because it is a city council estate, which has been run down
after years of neglect. Hence the Korogocho people would consider moving to

Kariobangi as a step up.

4.2.1.5. One who speaks inferior Sheng
Ngara respondents consider South C respondents as Mababi because they are rich kids
and speak an inferior version of Sheng, which they dubbed Ntiology due to the
prevalence of the /nt/ sound sequence in their speech”. This can be attributed to the
stereotype of the prevalence of English in the upmarket estates. The perceived English
competence of the Wababi can be attributed to their economic status that enables them to
attend good schools where they acquire quality English instruction. Their parents also
provide any relevant material that can promote their proficiency in English — the
language that denotes status. This fact is responsible for the English influence in their
variety of Sheng —— Engsh, which is recognized by other Sheng speakers. Pragmatic
reasons prevent the Wababi from acquiring ghetto Sheng. Not only are their houses
spread out, but they also do not make efforts to learn Sheng due to its stigmatization.
Since their families take care of all their needs, their survival does not depend on
solidarity; hence there is no pressure to maintain networks with friends.

These different senses of babi illustrate how a term can either affirm or contest an

identity. The South C respondents, while rejecting the babi label, would certainly not

 

39 This was not attested in my interview with the South C respondents.

152

refer to themselves as Wanati“, and they understand that there is a clear difference
between them and the ghetto people. However, while they do admire the cool linguistic
habits of the ghetto people as well as their ‘freedom’ (e.g., no overprotective parents),
they would still not like to be associated with the other side of Sheng speakers,
epitomized by poverty, disease, pronriscuity and urban crime.

The label babi is also juxtaposed with beshte to express the difference in lifestyle.
While beshte should be a close friend whether from Ubabini or the ghetto, ghetto people
have used this label to express the lack of interaction in Ubabini. The residential patterns
of these two places that inspired my hypothesis 5 illustrate this fact. In the pictures
below, I show an example of a typical slum, as well as a typical suburban residence in

Nairobi.

 

Figure 4. Kibera slum (source: www.tripgalle§y.com)

 

4° Although in his track Bamba the late E-Sir, an artiste from South C, refers to himself as
Mnati in the album Maisha Vol. I . However, it is not easy to establish the extent this term
is applied by middle class people outside of popular culture.

153

 

source: ht_tp://www.homes2000kenya.com/Properties/HC66.htm

Figure 5. A house at Riverside drive in Westlands

Comparing Figure 4 and Figure 5, we expect the rate of interaction in the ghettoes to be
much higher, due to residential dynamics. While living close together in the slums
promotes intense interaction and at the same time solidifies the bonds of solidarity, the
same is not the case in the isolated neighborhoods of Ubabini. It is therefore easier to
build and maintain dense interaction networks in Kibera than in Riverside. Consequently,
Wababi are not able to become competent in Sheng — the deep Sheng because they lack
resources to learn it and/or to build proficiency due to lack of inter-neighbor interaction.
If Mabeshte are the people who usually hang out together and perform their activities
together, then the residential dynamics explain why they are found in the ghettoes but not
in Ubabini. As such, the Mabeshte/mababi distinction has a socioeconomic angle.

These residential dynamics can be explained within the network theory (e.g.,
Milroy 1980). Of interest here are the concepts of dense multiplex networks and weak
uniplex networks. The closeness of the houses in the Kibera slum, for instance, promote
the dense multiplex networks where Sheng speakers establish multiple ties in areas such

as next door neighbor, fellow local drinker, fellow smoker, jua kali colleague, and so on.

154

In effect, they are able to cultivate and enforce linguistic norms. On the other hand the
residents of Riverside are unable to enforce linguistic norms because they share very few
ties with other local residents. From the outside, this might appear as though there is a
conflict with the core argument of the network theory, where the expectation is the
opposite of what I have just described. Middle class speakers are normally regarded as
having weak network ties and hence more receptive to innovation, while the strong
networks of the slum dwellers are supposed to serve as norm enforcement mechanisms
thus ensuring the retention of conservative norms. What needs to be taken into account is
the fact that there is very minimal interaction between these two socio—economic classes,
which block the diffusion of new innovations and their consequent adoption by the
middle class. In any case, these invations are deliberately created to draw group
boundaries. As such, the middle class only accesses the lexical items that have become
widespread. Seen in this light, the ghetto varieties of Sheng exhibit intense variation as
compared to those of the middle class neigbourhoods.

The relationship between the Wanati and the Wababi is symbiotic rather than
acrimonious. When we look at the current hip-hop music, the so-called Wababi musicians
(there are hardly any from the ghetto) have become the ghetto people’s advocates. The
themes their songs tackle sometimes highlight the issues in ghetto life. While this might
be seen as the extension of the exploitation of the ghetto people by the well off people

through commercialization of their culture and language for financial gains,4 we must

 

4' Such a claim has been made against Eminem, a white American rapper who was
accused by Ray Benzino, an African American rap rival and co-founder of The Source
magazine, of stealing black people’s culture for commercial gains (The New York Sun,
Nov 25/2003); see also Jerser. com, Nov 21/3003, Detroit Free Press, January 12/2004
and Garvin ’3 Blog. com April 4/2004.

155

not stereotype musicians as unscrupulous people who romanticize the ghetto for
economic mileage; instead, they call for revitalization of the ghettoes so that the ghetto
people can lead a more humane existence (see Samper 2004). A good example is Nazizi
and Wyre of the group ‘Necessary Noize’ and their tracks ‘Jah children’ and ‘Bless ma
room, ’ where the plight of the ghetto youth is highlighted. The late E—Sir, a hip-hop artist
from South C constantly made reference to Wanati and Wababi in his songs where he
identifies with the Wanati.

The symbolic label beshte bears different references to the ghetto people. It can
refer to a casual acquaintance, a close buddy, or anti-babi. The label babi also bears
different senses depending on the socioeconomic scale of the pe0p1e using the term. All
these categorizations are mediated by the form of Sheng that is used by the people who
use the term. Although babi, in most cases, is tied to the residential area a person comes
from, mom’s boys for instance, whether from Ubabini or not are poor speakers of Sheng.
They are thought to be under the protective care of the mother, and they might not want
to offend their moms, considering that Sheng bears negative connotations, especially
among women. The white-collar workers are too tied to the mainstream languages of
power and status to bother about mastering Sheng.

It is important to bear in mind that being a babi is only one aspect of someone’s
identity, and depends on the context of the interaction. If the context changes, a babi
might be transformed to a beshte when interacting with a fellow Mbabi. It is also possible
that a Mbabi who has interacted with ghetto people for so long that s/he has become
accepted might be referred to as a beshte. Moving away from babi, I will now focus on

another distinction, that between beshte and fala.

156

4.2.2. The cosmopolitan symbolism: Beshte vs. fala distinction

I have already defined beshte as a very good friend. The respondents gave various
qualities of a beshte, all of which emphasize shared engagement in various activities in
their daily lives, and all of which fit the definition of the ‘communities of practice’ as
explained in Chapter 2. These activities included playing together, joking together, going
to school together, watching videos together, chewing khat together, drinking together,
working together, and living together. Another name that stands in opposition to beshte is
fala, which in standard Swahili, means ‘a fool.’ In the Sheng discourse, fala also means a
stupid, rural person, as in the label fala kutoka ushago ‘stupid person from the country’
(Samper 20022161). Generally, these are the people who are considered to be outside the
Sheng speakers’ communities of practice. Informal settlements that border the
countryside are not spared in this fala reference. The Sinai respondents, for instance,
called the residents of Uthiru and Kawangware mafala (plural) for sharing a border with
the countryside, although these areas are normally regarded as part of the wider Nairobi
(see Figure 3).

I had originally assumed that the poor youth from the ghetto would identify more
with the simple rural people instead of the rich kids from Ubabini. Instead, my findings
confirmed Samper’s (2002) claim of Nairobi cosmopolitan identity, which is inconsistent
with the country life. To an urbanite, the rural evokes images of dullness, traditionalism
and ignorance, which are contrary to the enlightened trends of modernity. The stereotype
of a rural monoglot as opposed to an urban polyglot becomes the basis for discrimination
against rural people. Rural folks are believed to speak their mother tongue only, while the

urbanites are either regarded as multilingual or poor speakers of their native tongues.

157

(Notice here that mastery of a native tongue is not held in high esteem by such urbanites).
This perceived urban bilingualism, which as I said earlier forms the basis for Sheng,
together with other discursive practices, constitutes the shared urban norms. These shared
norms distinguist the two groups as a community of practice as defined in section
(2.8.2.7) that blurrs the boundaries between the rich and the poor urbanites resulting in a
shared cosmopolitan identity.

With this information, we can conclude that rural people lack the linguistic
resources necessary to negotiate their positions in urban social groups. This confines
them to the periphery of the Mabeshte ’s discourse practices. Samper (20022161) gives an
example of an upcountry guy who visits relatives in Nairobi and has to keep his mouth
shut lest he reveals his incompetence in urban lingo and earns the label of fala. This
perpetuates his/her continued exclusion from the urban communities of practice.

In reality, the stereotype of a rural monoglot is an exaggeration, because a
majority of the Kenyans speak the national language, Kiswahili. On a different note, the
ghetto speakers claimed that they understood Mababi and modified their speech (mixed
English and Swahili) in order to accommodate Mababi ’s linguistic norms. It is not clear
why they did not extend a similar gesture to the rural folk, who might even be more
competent in English and Swahili than they are. This is an indication that the major
reason behind looking down upon the rural people has nothing to do with languages;
rather, the disadvantaged rural guy who is trying to gain his/her footing upon encounter
with modernity is detested because he personifies a contested identity. The ghetto people

would rather identify with the affluent because theirs is the kind of life they aspire to

158

lead, but they cannot. In the conversation below for instance, although Shiro would not

like her children to grow up in Kariobangi, her aspirations are in Ubabini and not ushago:

PG: Mnatosheka na hapa Bangla?

Fio: Mimi natosheka

Shiro: Mimi sijatosheka. Mimi as in my life,

PG: mm-hm

Shiro: ...like ka watoi wangu siwezi (unclear) nataka wa-change

PG: mm-hm

Shiro: walelewe place ingine different na hii ya hapa. Siwezi taka
wa-experience. . .

PG: Kule ushago ama?

Shiro: Eee si ushago,

Edee: Ubabini

Shiro: ...yaani place poa tu ubabini...awe na life different..ee...asi-lack,

PG: enh..

Shiro: bado days zake zisiwe ka venye mi nilipitia hardships. Yeye asipitie
hardships. . .everything iko

Translation:

PG: Are you satisfied with Kariobangi?

F102 1 am satisfied

Shiro: I am not satisfied. Me as in my life,

PG: mm-hm

Shiro: ...like my children I would not like (unclear) I want them to change

PG: mm-hm

Shiro: They be raised in another place different from this here. I would not like
them to experience...

PG: in the countryside isn’t it?

Shiro: Eee. . .not in the countryside

Edee: Ubabini

Shiro: I mean a nice place in Ubabini, to have a different life. . .ee. . .not to lack
anything

PG: enhe

Shiro: then his/her life should not be like the way I passed through hardships.

S/he should not pass through hardships. . .to have everything.

Shiro, like most other ghetto dwellers, aspires for a different way of life. Recall that in
Chapter 2, I discussed the way many people come into the city with big dreams of

making it in life. The urban dream that lured people into the city is still very much alive.

159

This dream cannot be realized in the ghetto, but in Ubabini. A Mbabi therefore is not
someone to be despised, but someone to aspire to be. Notice that when I try to suggest the
countryside as an alternative, there is a quick objection by both Shiro and Edee. The
ghetto person understands that the rural areas are not his/her solution; indeed, the
countryside suffers from a similar depravity, if not worse. Going back to the countryside
will be like drifting back to the kind of life the urbanites have already turned their back
on. This explains why the ghetto person is unsympathetic to his/her fellow poor guy from
the countryside, who evokes images of a past they would rather keep buried.

Even the label fala does not always bear a negative connotation. Sometimes the
in-group may use the label in an affective way. When used by the in—group such as the
Mabeshte in a baze, fala bears the opposite meaning — beshte. According to Kabete
respondents, a friend who treats his buddies to drinks until they are completely wasted is
a real beshte. When a person who enjoyed such a beshte’s generosity is narrating about
the incident to his fellow buddies, he might say something like ‘huyufala aliturogajana’
‘this fool poisoned us yesterday’. Use of fala in this sense can be compared to the English
insults ‘son of a bitch ’, or ‘dumb ass ’ or even ‘nigger’ among African Americans that
can be used affectively among close friends, but with serious consequences when used by
an out-group. The ability to use insults without tension is a diagnostic test of how much a
person has moved up the the in-group hierarchy. An outsider who does not understand
the polysemy of fala might misinterpret the intention and hence threaten the stability of
the group. The symbolism associated with fala can be seen in the rightward or leftward

movement along the linear scale as in the diagram below.

160

Out-group <——fala-—>in-group
(negative) (positive)

This fala example shows that use of insults is an in-group/out-group thing. Whether the
insult is used affectionately or disparagingly depends on who uses it in terms of ingroup
or outgroup. Smitherman (1977271) illustrates this very well with a 1974 cartoon from
G.B.Trudeau. In the cartoon, a white boy (Bobby) who has an overprotective mother has
been warned to keep away from black kids. But on the bus to school, he usually sits
beside a black boy (Rufus). He naively asks Rufus what ‘honky’ means and how he was
expected to respond if someone called him honky. Rufus, aware that it was a disparaging
term for whites used by the blacks is evasive and tells Bobby that he will pick it up (know
it somehow). Unfotunately, Rufus uses the term on a black boy, gets beaten up badly and
has to be admitted to the hospital.

Insults therefore bear different meanings depending on whether they are being
used by the out-group or the ingroup. Their semantics change according to the contexts
and speakers who are using them. At one extreme, fala is negative, while at the other
extreme, it is positive. An outsider negotiates his/her position within a group depending
on the symbolism s/he attaches to fala along the linear scale. A country person who
adjusts to urban life gradually moves up the scale from fala to a beshte, but we must bear
in mind that since identity is negotiated at a local level, a fala may gain acceptance by
some but not all, thus giving him/her multiple identities within the same baze. This
juxtaposition of fala and beshte, show us how symbolism associated with an identity can
change depending on context and users. This meaning-making is extended to other

discourse practices that might constitute face-threatening acts.

161

4.3. Performance and negotiation of multiple identities

After looking at how various Sheng lexical items are manipulated to symbolize different
identities, I now move to identity performance and negotiation. I will analyze speakers’
discourses and discuss how Sheng speakers from different bazes perform multiple
identities in ordinary conversation. I will show how identities are affirmed and contested
at both the conscious and subconscious levels. I first discuss the non-conscious
performance of multiple identities in spontaneous discourse and then discuss how Sheng
speakers fi'om different bazes shift between various Sheng varieties and in the process

shift between the identities associated with those varieties.

4.3.1. Non-conscious activation of layered identities

Individual identity is a complex of layered sub-identities. Some of these can be activated
consciously while others, which have been internalized through constant enactment and
re-enactment, may surface non-consciously. The term salience has been used in social
identity theory to refer to the activation of an identity relevant to a given situation (Stets
and Burke 2000: 229). I am not interested in the whole definition of salience and
activation as explained by Stets and Burke, but my interest here is in the non-conscious
activation of identities that surface in spontaneous discourse. I focus on how these deeply
lodged identities are performed in naturally occurring conversations, and in the process
confirm or contradict various aspects of the identities that speakers claim they possess. I
pay attention to what speakers say about themselves, and then analyze their discourse for

confirmation of what they claim they are. When we pay close attention to discourse, we

162

realize that what respondents say about their identity might not necessarily be congruent
with how they perform those identities in discourse.

In personal experience narratives, we can see how narrators perform multiple
identities at different parts of their narration through interjecting different lexical items in
their discourse. In the following short except, the narrator alternates between English,
Gikuyu (note Gikuyu orthographes of the front mid high vowel [e] as /'1/), Sheng and
Swahili. The story is about an old man who was trying to sell stolen shoes to the narrator
and his fiiends. Unknown to the seller, someone from the group who had recognized the
stolen shoes had informed the owner. The story opens with the narrator (Tash) and his
friend bargaining, just as the owner of the shoes arrives and a fight ensues.

....sasa sisi bado....niaje (Sheng) we deal (English) we (Sheng) chukua

rwa...rwabe (Sheng). Analia aa eee. Jamaa hii_nimwahinyi§iria... he (Gikuyu).

(laughs) wakati anabongabonga (slang).....hii' (Gikuyu), mbo...bo...(ideophone)

enyewe mimi nimeona ni kama nimeona movie (English), ndiyo yule mwenye hii

jumu (Sheng) akakamu, hakubonga, alikamu tu na washa (Sheng). Alichapa yule
mzee washa za kichwa,(laughs). Mzee ile kustagger stagger,
Translation:
...now we were still ...come on man, let’s deal, take two hundred. He cries aa
eee, men oooh, you are exploiting me...eh! (laughs). While he was talking, man
mbo...bo it was like I had seen a movie. Then the owner of the shoes came, and
he did not speak, he just came with a club. He struck this old man on the head
with the club (laughs). The old man started to stagger...
My account for the above follows Bucholtz (1999: 209) that individuals engage in
multiple identity practices simultaneously and that they are able to move from one
identity to another. In this story, which is told in Sheng, the narrator uses Swahili

structure, which affirms his ghetto identity. Moreover he identifies with his group

through the use of its Sheng vocabulary. Support for this comes from the words rwabe

163

‘two hundred shillings’ and jumu ‘shoes’. In some bazes, soo mbili for ‘two hundred’ or
chuja for ‘shoes’ might have been used. English insertion in non-English discourses is a
characteristic of a majority of Nairobi discourses. It is the unmarked code that indexes
education and status. Even when one is not well educated, or has lower socioeconomic
status, people still use it to claim these identities (see Myers-Scotton 1993a). Writing on
codeswitching in Kenya, Meyers-Scotton argued that people use Swahili to project a pan-
Afiican identity, while still using English to project the status that English is associated
with. The use of Gikuyu in the above discourse might indicate two things: either the
narrator was affirming his Gikuyu identity”, or he was alluding to the common
stereotype of the Gikuyu tribe being the tribe of thieves. It could as well be that he was
only imitating what the man said verbatim as a stylistic device. Even so we can take this
as a case of commentary about the ethnicity of the thief, which in itself constitutes an
identity exercise. From this short excerpt, we see that the narrator projects different
aspects of his layered identities as well as the identity of others. We can however not tell
whether this was deliberate or was an unconscious act.

A similar case was observed during the interview with Sinai girls. In one instance,

one respondent was commenting on the closeness of the slum dwellers as a result of the

 

42 It is possible for somebody from another tribe to use Gikuyu vocabulary since Sheng
borrows from different ethnic languages. However, the construction nimwahinyiri'ria is
not just the insertion of Gikuyu lexical items, but also a display of mastery of Gikuyu
morphosyntax, as we can see below;

Ni-mwa-hiny-ir-ir-i-a
Cop- l sm-press-AP-AP-CAUS-FV
You have exploited me.

Although we can argue that the two applicative morphemes have been gramaticalized
such that hiny- as a root might not make sense by itself, the use of the copula and covert
tense inflection shows that the speaker is well versed with the Gikuyu grammar.

164

nature of their shacks. With one house separated by corrugated iron sheet walls, there is
no secrecy, and people can talk across walls, as Kui claims;
Kui: kila mtu ni sistako juu ndiye anakusaidia na mnasaidiana. Mna-cooperate
vizuri, unajua ni hii mabati na hii ingine kusema “haiya wa kamau rehe

maguta ” (laughs) sasa wakamau akiwa hapo si maguta si unajua lete hiyo
matagu, hata hakurushii mbali tu pale tu kwa mbao na unashika. ..

Translation:

Kui: Everyone is your sister since she is the one who assists you and you assist
each other. You cooperate well, you see between these corrugated iron
sheet walls you can say “ Hey! Wakamau (either Kamau’s mother or
Kamau’s wife) pass the fat/oil” (laughs) now if Wakamau is there, she will
not have to pass it over from afar, she passes it over by the wallboard. . .

Kui affirms her group identity by highlighting their solidarity. She is also affirming her
identity as a ghetto girl by narrating about the poor living conditions in the ghettoes as
portrayed by the poor houses (houses separated by corrugated iron). Like Tash above,
Kui introduces a Gikuyu expression haiya wa kamau rehe maguta ‘hey! Wa Kamau, pass
the 011’, while she was well aware that one of the respondents was not a Gikuyu. I regard
this as a ‘negative identity practice’— those practices that are inconsistent with the group
identity (Bucholtz 1999) — because she may not be deliberately trying to show her
Gikuyu identity, but by deviating from the common linguistic code she projects her
ethnic identity. Evidence of Kui’s Sheng performance of group identity comes from the
use matagu, derived from maguta (‘oil’ or ‘fat’ in Gikuyu) through syllabic inversion
mentioned in Chapter 3. I noticed that syllabic inversion was an important characteristic
in the Sinai girls’ version of Sheng.

Compared to the other two girls in the Sinai baze, Kui was the most talkative. She

was also very fluent in Swahili. This is very relevant because she claimed that she was

born in Kibera and that she had Nubian parentage. Recall that in Chapter 3, I mentioned

165

that the Nubians in Kibera have appropriated the Swahili culture to project their identity.
Kui performs this part of her identity by using her mastery of Swahili. She also said that
she went to Kibera often to see her boyfriend. In a different interview session, her friend
Joy called her Mfinfibi iiyzi ‘you Nubian’ (Gikuyu orthography), which shows that her
Nubian heritage was a matter of common knowledge.

In my next illustration, Joe, a respondent from Makadara, tells about his
encounter with criminals on his way home from town. He and his girlfriend had alighted
from a matatu and were headed to his house when a gang of criminals who had a gun
accosted them. He had initially thought they were policemen, but when they beat him up,
stole his mobile phone, money, cap, and bags of cheap liquor, he realized they were
thugs. His claim that he initially thought that they were police made Bill question his

courage. Joe defends himself by saying that he was overpowered because he was alone.

Joe: ...unacheki unaona ni usiku. Halafu kitu ya pili wasee wamekuja...kirende
ya mtu tano. Hata wewe kirende ya mtu tano wewe utatoboaje wewe u. ..
man solo?

Bill: Si ndiyo nakwambia wewe uliingia upepo. . .mapema tu.

Jim: ni noma

Joe: haki ni vile unacheki unaona maze hasapa, hauna otherwise. Nikacheki hii
unajua nilionanga tumeanzilia na mama, nimekuta yaani ile chuja
nilikuwa nayo, matope ilikuwa imefika huku juu. Trouser yaani ungecheki
hivi ungedhani ni ya juzi...

Translation:

Joe: . . .you look and realize it is at night, then the second thing the guys have
come. . .a group of five people. Even if it was you, a group of five people,
how could you prevail when you are alone?

Bill: That is why I am telling you that you got scared. . .too early.

Jim: it’s hard

Joe: really, it is the way you can look and realize, man, here you have no
alternative. I saw now you know I saw the way we had started with the
girl. I found the shoes that I had were full of mud. If you looked at the
trouser, you could think it was an old one...

166

In a place like Makadara where the rate of crime is high, one has to display a lot of
courage to repulse the petty criminals. When fellow Mabeshte challenge you on account
of your lack of courage, it is a real face threatening act, hence the spirited attempt by Joe
to justify why he was unable to fight back. It is important that the person who makes the
challenge is Bill, a military trainee, who spends his weekends with fellow Mabeshte in
the baze. From his military orientation, he would have expected Joe to make a quick
assessment of the situation and fight back. This is how we get Bill’s identity as a military
man. Joe succeeds in convincing Jim, who now accepts that it was hard for Joe to fight
back. This allows Joe to regain his face and hence salvage his self-esteem among his
buddies.

Equally interesting is Joe’s use of chuja for shoes. During the interview, he had
said that they used jumu in their baze. At the beginning of his narration, he had referred
to his girlfiiend as mbevo, a name he had attributed to the Dandora estate. He also used
kavu moja for one thousand shillings, while in the interview he had said that they used
tenga moja. This mismatch between what was said and what actually happened in
discourse is a pointer to the fluidity of the bazes. Use of lexical materials that are not
generally thought as belonging to the speaker or the group can be viewed as a case of
language crossing (Rampton 1995, 199721) and can be attributed to interaction with other
fiiends from outside the baze. “Crossing” here is taken in its narrow sense to refer to the
baze-specific varieties of Sheng instead of Rampton’s ethnic languages. During linguistic
crossing, identities may be consciously activated to suit situational characteristics, but I
am interested here with the non-conscious surfacing of lexical features that speakers

produce in their discourse, which affirms their multiple identities. When Joe inserts

167

items, which he claims are used in other places, it shows how deeply ingrained these
identities have become, so much so that he does not even realize it when he is performing
them. If what he says, that these lexical items are used in different bazes, is true, then we
can conclude that by using them he is displaying a mosaic of identities that reflect the
different bazes he has interacted with.

These pieces of conversation show how the participants move across different
identities, even when they are engaged in a single utterance because they have become
imprinted into their personalities. I explain this in terms of Bourdieu’s (1990253) concept
of habitus, a disposition to doing something which he defines as;

...systems of durable, transposable dispositions, structured structures predisposed

to function as structuring structures, that is as principles which generate and

organize practices and representations that can be objectively adapted to their
outcomes without presupposing a conscious aiming at ends or an express mastery
of the operations necessary in order to attain them. Objectively “regulated” and

“regular” without being in any way the product of obedience of rules, they can be

collectively orchestrated without being the product of the organizing action of a

conductor.

While this statement does not downplay the importance of agency among the Sheng
speakers, we see that they do not have to consciously attempt to perform the identities; it
has become ingrained through constant performance. When these out-group identity
practices are being enacted, the speaker is not trying to adopt the identity of outsiders, but
such usage demonstrates how the outside influences form an important part of their own
identity.

In summary, we have seen that identities are not just asserted but also performed
by the speakers in their discourse. I discussed how people confirm and contradict their

identities through non-conscious identity practices in spontaneous discourse. I have

concluded by claiming that this is so because their interaction, characterized by shifting

168

speech behavior, is ingrained in their habitus. In the next section I will be discussing how

speakers perform identities through conscious choice of linguistic features.

4.3.2. Negotiating multiple identities through accomodation

Shifting of identities does not always follow from a set of dispositions that drive agents to
subconciously generate practices as the the concept habitus (see above definition).
Speakers sometimes act deliberately when they want to achieve specific communicative
goals, as proposed in speech accomodation theory (e.g., Giles, Taylor and Bourhis 1973,
Giles and Powersland 1975, 1997, Giles eta], 1979). Speech accommodation theory deals
with the adaptive strategies that speakers adopt with respect to their interlocutors. Key
concepts in the literature on accommodation theory have been speech divergence and
speech convergence. In speech divergence, speakers increase perceived dissimilarities
when they intend to distance themselves from undesirable others while in speech
convergence, speakers increase perceived similarities with their interlocutors with whom
the wish to identify.

The central logic advanced in accomodation theory also features in Bell’s (1984:
158-164, 19912 105) ‘audience design’ framework, in which the argument that speakers
linguistic productions are a reflection of their audience’ speech norms, is advanced. In
other words, speakers are seen as styling their speech habits to approximate to those of
their listeners. This model works well in radio broadcasts, where the maistream language
is propagated as the preferred broadcasting language. The opposite scenario, where
speakers allocate themselves the role of setting linguistic standards is argued in the

‘referee design’ model (Bell 1984). The major claim of the referee design model is that

169

 

 

 

when speakers regard themselves as the reference point, they modify their speech in order
to be unlike the out-group. Broadcasters who believe they are using the model speech
will maintain their speech habits for their audience adopt. If the reference group shifts to
the out-group, speakers may either modify their speech to be similar or different from the
out-group depending on what they wish to achive. Take Meyerhoff's (1998) analysis of
the use of pronouns yumi (inclusive), mifala, (exclusive) and yu (generic) in Bislama as
an example. One of her female subjects used yumi to emphasize their shared gender
identity (affiliative43), but Meyerhoff deliberately used the generic yu, to dissociate
herself from her interlocutor’s perceived shared identity. Her subject finally adjusted her
use of pronouns to reflect their non-local identities (dissociativefl). Describing a similar
phenomenon, Bucholtz (1999), talks of positive and negative identity practices. Negative
identity practices define who the speakers are not, while positive identity practices define
who the speakers are. Negative identity practices are dissociative while positive identity
practices are affiliative.

Analysis of Sheng discourse reveals instances of affiliation as well as dissociation
as baze members negotiate identities. When Sheng speakers from a baze wish to affiliate
themselves with speakers from a different baze, they consciously model their speech by
adopting the salient lexical features of that baze. My respondents claimed in the oral
interviews that they could easily adopt the kind of Sheng spoken in other bazes. In
Makadara for instance, when I asked the respondents why they spoke Sheng associated

with other estates, they responded with a Swahili saying: Ukienda kwenye chongo lazima

 

43 I borrow these terms from Vermeij (200424) who talks about affiliative and dissociative
crossing.
44 See 5 above.

170

ufimgejicho lako ‘if you go to people with monocular vision, you have to shut one eye’.

They were unanimous that they do so in order to adapt to the situation demands when

they visit other bazes. As outsiders, the onus is usually upon them to speak the variety

spoken in the bazes that they visit. This was confirmed by the Kariobangi female

respondents in response to the question whether they ever spoke the Sheng spoken by the

Wababi in the following conversation.

1. PG:
2. unaimous:
3. Shiro:
4. PG:

5. Shiro:
6. PG:

7. Shiro:
8. PG:

9. Sera:
10. Pame:
11. PG:
12. Pame:
l3. Sera:
14. Pame:
15. Shiro:

Translation:
1. PG:

Na kuna wakati nyi mnazungumza Sheng ya Wababi?

...labda tukiwa nao

Labda ukienda na yeye akuongeleshe hiyo kizungu yao mnaenda
juu kidogo umuonyeshe ume-educat—iwa. Unajua Sheng yao
inakuwanga English-Kiswahili

enhe?

ka ume-educatiwa unaweza mix hizo vitu mbili.

kwa mfano tuseme una chali huko. . .huko. . .ubabini
ubabini eh?

(laughs) wakati mnazungumza mnatumia nini?
tunatumia zote

hiyo yao

yao, siyo yako

Sheng yao

juu yako ukiongea. ..

hataelewa

hatashika, na wewe utashika yake juu wewe umesoma ata ka wewe
ni wa ghetto.

...and is there a time when you speak the Sheng of the Wababi?

2. unanimous: maybe when we are with them

3. Shiro:

PG:
Shiro:

5"?

PG:
Sera:
PG:
Sera:

©0040

maybe if you go with her and she talks to you in their English you
go up a little bit in order to show her that you are educated. You
know their Sheng is usually English-Swahili?

enhe?

if you are educated you can mix those two things

for example let’s say you have a boyfriend there in Ubabini
Ubabini?

(laughs) when you are speaking, what do you use?

we use all

171

10. Pame: theirs

11. PG: theirs, not yours?

12. Pame: their Sheng

l3. Sera: because if you use yours

14. Pame: he will not understand

15. Shiro: He will not understand and you will understand his because you

are educated even though you are from the ghetto

We note that the identities of these female respondents are not just affirmed through
speech, but also performed. In line 2 all the respondents claim that they do speak the
Sheng of the Wababi when they visit Ubabini. Lines 10, 12—15 restate this besides
providing the reason why they cannot speak the Sheng of the ghetto — the Wababi
cannot simply understand the ghetto variety. In 3, Shiro displays her attitude towards the
Ubabini Sheng by saying hiyo kizungu yao” ‘their English’. The codemixing in the word
ume-educat—iwa ‘you are educated’ and ‘mix’ in lines 3 &5 performs a dual task: 1) It
transmits the intended message, and 2) since English is associated with the speaker’s
education, use of English words may also be seen as performing the identity of an
educated person. Therefore, the ghetto girls accommodate to the speech of the Wababi
because they want to show that they are educated, not because they admire the Wababis’
kind of Sheng. This is a way of identifying with the status associated with the English
language exemplified by the Wababi.

Earlier we saw Shiro’s desire to move to Ubabini because she would like a better
life for her children. She associates Ubabini with the good life, and speaking Sheng laced
with English is a way of symbolically claiming that life, which she cannot get in the

ghetto. The claim that the girls from the ghetto are the ones who accommodate to the

 

45 Non-proximate demonstrative hiyo is used here to signify distance and lack of
attachment to the subject by the speaker.

172

style of the Wababi was supported by the South C respondents. Not only don’t the
Wababi understand ghetto Sheng, but to them there is nothing much to be gained from
competence in that kind of Sheng. Therefore there is no motivation to perform the
identity of the ghetto people, although below I will discuss just such an attractiveness of
ghetto Sheng. The girls’ speech habits when they are interacting with the Wababi are
responsive in the sense of the Bell’s (1984) audience design model mentioned above.
They engage in ‘positive identity practices’ (Bucholtz 1999), in order to identify with the
individuals they wish to affiliate themselves with. Ubabini, in the sence of Bell’s referee
design model, may be regarded as Shiro’s out-group reference because she claims that
she modifies her linguistic habits to affiliate herself with her interests in Ubabini.
Contradicting the Kariobangi females are the Shauri Moyo respondents, who
claimed that the Wababi are the ones who normally accommodate to their form of
speech. Although they echo Kariobangi females’ sentiments that English is the language
spoken in Ubabini, they explain that the closed lifestyle in Ubabini makes it hard for the
Wababi to develop a ‘cool’ language. In the conversation below, Stevo, Pish and Joe

claim that Wababi copy their language because it appeals:

1. Stevo: ...lugha inaongelewa kwao ni kizungu. Sasa unapata mtu anatoka
huko, labda unapata kapesa kidogo ukahamia huko,

2. PG: mm-hm

3. stevo: ...wale wakiona vile unaishi,

4. PG: mm-hm

5. Stevo: .. .vile unaongea, ina-appeal kwao, sasa pia wanataka kujihusisha
ndani...

6. Joe: Alafu unajua watu ka wa Westlands wale...unajua si tuko hapa
wengi wetu kuna ma—baze.

7. PG: eee

8. Joe: Westlands. . .hapa kuna baze pahali chali tunapatana

tunakaa...huko ni ngumu. Huko hakuna baze. Huko kila mtu
anajishughulikisha mambo yake.

173

9. PG: sawa

10. Joe: Sa ku-form hiyo language ni ngumu sana.

ll. pish: so unaonanga hao wakikujanga huku wanadhani hii lifestyle
tunaishi...

12. stevo: . . .ina-appeal,

Translation:

1. Stevo: ...the language spoken at their place is English. Now you find a
person fi'om there, probably you get some little money you move
in there...

2. PG: mm-hm

3. Stevo: When they see how you live,

4. PG: mm-hm

5. Stevo: ...the way you speak appeals to them, now they too want in...

6. Joe: Then you know people like from Westlands that. . .you know here
for most of us there are bazes

7. PG: eee

8. Joe: Westlands. . .here there are bazes where us guys meet and
relax. . .there it’s hard. There, there are no bazes. There everyone
minds their own stuff.

9. PG ok

10: Joe: Now to form that language is very hard.

11. Pish: so you see when they come here they think this lifestyle we lead...

l2. Stevo: appeals. . ..

Like Shiro in the preceding dialogue, Stevo in line 1 affirms that in Ubabini people speak
English. Regarding who accommodates to the other, the claims of the female respondent
are now reversed. However, note that I had given the females a scenario of having an
Ubabini boyfriend. They said they would use his variety because he couldn’t understand
theirs, though they could understand his. A similar answer would probably have come up
if I had given the Shauri Moyo males a scenario of having an Ubabini girlfriend. Joe in
line 6 attributes Ubabini Sheng’s lack of appeal to isolated homesteads“, individualistic
tendencies of the residents, and lack of hang-out joints (line 8), which makes it difficult
to form Sheng (lineIO). Following network theory, we can argue that the isolation in

Ubabini prevents formation of dense networks where innovation can spread. Since close-

 

46 Refer to the two earlier pictures of Kibera slums and a house at Riverside.

174

knit networks act as norm enforcement mechanisms (Milroy 1980), the lack of the bazes
in Ubabini checks the formation and spread of Sheng. It is very interesting that this na‘r’ve
respondent pinpoints one of the major features of network theory so accurately and
concisely.

The social status of the Wababi and the implicit requirement to conform to social
convention prescribes their lifestyles. On the other hand, the ghetto people, to an extent
live outside the boundaries of social conventions. They lack the material means necessary
to fulfill all the requirements prescribed by the conventional standards of social status.
They have little or nothing to lose by transgressing the social boundaries, but what they
lack in status, they compensate with their high sense of solidarity. It is this freedom from
the yoke of conventions and their deep sense of solidarity that inspires inventions, both in
terms of languages and popular cultural forms. This freedom to transgress the social
norms with impunity, and the appeal of popular cultural forms, appeals to Ubabini youth,
who in spite of the lavish lifestyle and the external aura of contentment, nurse a void in
their social lives. We can draw a parallel with white youth in America, who appropriates
the images of African America popular culture. A good illustration comes from Cutler’s
(199922) description of Mike, a white teen in New York. Mike employed linguistic
features associated with African Americans and engaged in practices like wearing baggy
jeans, reversing his baseball cap, and developing a taste for rap music, all associated with
African American culture.

These examples from Kariobangi and South C bazes show that accommodation
goes both ways, though for totally different motivations. The Kariobangi females did not

claim to accommodate to the variety of the Wababi because of its appeal. The do so

175

because they yearned for status associated with Ubabini lifestyle in general. However, as
Stevo (5) puts it, the Wababi are attracted by cool lifestyle in the ghetto, but whether they
are accepted or not is another issue altogether. In the above-mentioned case, Cutler
(1999: 8) wrote about Mike’s subsequent resentment towards African American peers,
who ‘always hang together’ and ‘separate themselves’. This followed their failure to
admit him into their subculture, thus shattering the dream of an exotic identity he was
trying to construct. Although Mike’s language crossing was successful, his crossing in
other cultural aspects was a failure. Similarly, the Wababi cannot construct for
themselves an accurate image of the ghetto people due to lack of assimilation into the
ghetto culture. By itself, the speaker’s linguistic accommodation is not sufficient to
complete the construction of identity. Equally important is the degree to which one is
willing to get engaged in the general lifestyle, a criterion that would not be so easy for the
Wababi who also enjoy the perks that go with their social status. As such they only
engage in short term accommodation, which restricts them to the periphery of the baze.
Lack of total acceptance finally leads to frustration and reappraisal of the failed
conscious identity practices. The lifestyle ceases to appeal and the original boundaries are
redrawn. This may be explained in the referee design model (Bell 1984) as the shift in
reference point. When the reference shifts from out-group back to the in-group, it is
accompanied by modification of linguistic practices that prompt the shifter to magnify the
differences with the out-group. Alternatively, a composite identity with aspects of both
may be created — such as South C respondents, who called themselves anti-mababi, but

still claimed that they could not understand the ghetto people.

176

One question that comes to mind is what happens when the language crossing
involves two ghetto varieties spoken by people from different bazes. In view of my
discussion of the Mabeshte/mababi contrast above, I asked the Shauri Moyo respondents
whether there were times that they desired to be members of another baze on account of

their Sheng. The following conversation captures some highlights of the discussion.

PG: Kuna wakati unajisikia kuwa member wa Mabeshte wengine kwa sababu
ya Sheng yao?

Unanimous: Sa hapo twasema eee.

Stevo: Kuna saa zingine wapata unaongea Sheng na watu, kama wabonga Sheng
tamu. Watu kama hawa Kijey waskia wabonga Sheng tamu.

PG: yeah

Stevo: Sheng yao yaani, unasikia kuna words wanaongezeaongezea hapo
unasikia yaani ladha yake ni tamu

PG: Enhe

Stevo: Eee, ingawa pia si tunajua Sheng yetu ni tamu lakini tunasikia yao kwanza
unakaa karibu nao unasikia enyewe ...ni kitu new... ni kitu tofauti.
unasikia utamu yake ata ukiingia, hata wewe mwenyewe ukiwa unajua
Sheng, lakini unanyamaza unaanza kusikia

Translation:

PG: Is there a time when you feel you would like to be a member of another
group of Mabeshte because of their Sheng?

Unanimous: Now there we say yes

Stevo: There are times when you are speaking Sheng with some other people,
their Sheng is sweet. Like these people from kijey (Majengo) you hear
them speaking very sweet Sheng

PG: Yeah

Stevo: Their Sheng, you hear, there are words they add... you find their flavor
is sweet.

PG Enhe

Stevo: Yes, although we know our Sheng is sweet, but (when) we hear theirs we
first stay close, you feel like. . .it is something new, it’s different. You
feel its sweetness getting to you, even if you join them and you know
Sheng, you pause and begin listening.

The issue here is why Shauri Moyo respondents who later claim that their Sheng is ‘the

best in the whole world’ would find any other form of Sheng attractive and sweet. One

177

possible explanation is that, as mentioned earlier, Majengo is one of the Muslim enclaves
where Swahili culture is so much in evidence. Majengo estate borders Pumwani ——
another Muslim dominated estate (Burja 1974). According to Burja, Pumwani’s Swahili
had a very distinctive coastal flavor, and the same applies in Majengo. The Shauri Moyo
respondents in this case are therefore not making a comment about Sheng per se, but are
projecting another aspect of their identity; the Kenyan national culture transmitted
through Swahili language and culture, and whose standards are set by the coastal version.
What is interesting is that Stevo does not talk about speaking Kijey Sheng; instead, he just
likes listening when people from Kijey speak. This is evidence of his in-group loyalty,
which prevents him from engaging in negative identity practices such as denying his
group the prestige they claim for themselves as speakers of the best Sheng.

Unlike the Wababi, the Shauri Moyo respondents are not trying to appropriate the
perceived ‘coolness’ of the Kijey ghetto since they are also ghetto. However, they are
able to portray an aspect of their Kenyan national identity by affiliating themselves with a
variety of Sheng that approximates to the language that propagate national identity. This
case might be regarded as a counter argument to the accorrnnodation theory discussed
above. These respondents do not claim to modify their speech to either converge or
diverge from the speech habits of desirable others. Hence, the possible explanation would
be that they have such deep loyalty to their in-group that they would not modify their
linguistic norms to suit those of outsiders.

Majengo’s case is also interesting when approached from a very different angle. I
mentioned in Chapter 3 that a large part of Sheng involves codeswitching with Swahili as

the matrix language in the sense of Myers-Scotton’s (1993b) Matrix Language Frame

178

(MLF) model. This is especially so with the kind of Sheng spoken in Majengo, which
like that of Kibera, is heavily influenced by Swahili. As a result, the Majengo people
exhibit a triple identity, at least where the ethnic Swahilis are involved. These include
national identity indexed by using Swahili that approximates to the Standard version,
ethnic identity, also indexed by Swahili as their ethnic language, and being Nairobians,
they get the cosmopolitan identity expressed through Sheng by default. We must also not
forget that Majengo is a residential estate, and as such, we expect there to be different
bazes in the estate with their own distinct variations.

The above illustrations have dwelt on the motivation behind the conscious
shifting of identities. I will tie the construction of these identities to the efforts and
commitments that the people seeking to gain admission in a group invest in the
enterprise. Reitzes and Burke (1991: 243) argued that greater commitments result in
shared meanings that affirm, and are consistent, with an identity. Commitment to out-
group ideals leads to activation of the identities that are consistent with the norms of the
desired group. Commitment to the in-group retains the status quo. Equating commitment
to a cybernetic control system, Reitzes and Burke (19912243) wrote;

People can learn which behaviors are effective in maintaining congruence

between their identity setting and the reflected appraisals. In addition, they can

learn how to stabilize lines of action to maintain congruity, provided that the

meanings and objects and behaviors do not shift or change while they are engaged
in that learning — that is, provided that they are in a stable semantic relationship.

Commitment leads people to behave in ways that are consistent with the groups they seek
to be affiliated with. While not taking the group as a collection of static entities, I
recognize that there must be some sort of stability that defines the hierarchy of all the

members of the group. This hierarchy maps down the boundaries and norms within which

179

the members operate. Whether expressed overtly or covertly, these norms are well
understood by each member of the group and are responsible for the maintenance of
order and identity of that group. Existence of these boundaries determines the criteria for
the admission of outsiders and the requirements they must fulfill before they become
acceptable. Commitment therefore must be matched by the group’s willingness to accept
those who seek admission and who are perceived as ‘safe’— those who will not
endanger, or lead to the disintegration of the group. A group of marijuana smokers will
for instance not knowingly admit a policeman, no matter how committed that policeman
is. Indeed we find that the reverse is true as evidenced by the many terms for the
policemen listed in Chapter 3. The rejection of Mike by the Afiican American sub-groups
in Cutler (1999) can be interpreted in the same way.

The discussions above verify Hypothesis 4, that identities are shifted both
consciously and unconsciously. Non-conscious shifting of identities shows that not all
choices are consciously made as the RC model in its strict sense would predict. However,
its applicability cannot be ignored as can be seen in speakers’ uses of associative and
dissociative strategies to negotiate a desired identity.

Although Shifting of identities and the identity hierarchy has been linked to the
commitment to be congruent with the desired out-group, it should be understood that
commitment is not in itself sufficient to guarantee automatic admission to desired groups.
It must be consistent with other factors bordering on shared realities such as
socioeconomic, ethnic, gender, and age and so on, all which enhance shared meanings in

which identity is negotiated and constructed. In the following two sections, I examine

180

some discourse practices that show that participants construct their identities by using

their shared meanings that revolve around the realities of their daily lives.

4.4.1. Vernacular discourse: A different perspective on local reality

Many sociolinguists use the Labovian definition of vernacular as the style in which
minimum attention is given to the monitoring of speech (Labov 1972a2208). Labov’s
definition has been applied extensively in the study of linguistic structure, especially
phonological features. However, unmonitored speech is not restricted to the study of
linguistic structures alone, but also covers other discourse styles, including those that
transgress the boundaries of social norms of civil language. Unmonitored speech is
usually frowned upon when taboo themes such as sex are discussed. In this Section, I
analyze the discourse of two young women from Sinai baze, who dabble in petty
prostitution. It will be demonstrate that what is taboo in the mainstream society is not
necessarily considered taboo in the subculture. I use De Meis (20022 15) metaphor of the
‘whore’ and ‘the sex worker’, where the ‘whore’ is defined as one who accepts the
definition that the society has accorded her as just a bad woman, who sells her body for
money. The ‘sex worker’, on the other hand, goes beyond seeing herself as a person who
enmeshed in the flesh peddling vice, but also as a worker whose profession guarantees
liberty, independence and self reliance. The use of vernacular in the narratives by these
two petty prostitutes in performing their heterosexual, ‘whore’ and ‘sex worker’ identities
will be discussed. It will also be argued that using vernacular allows these girls to claim

modesty as prescribed by mainstream norms.

181

Sinai is the only place where I went back after the initial visit. The two young
women, Kui and Joy, had impressed me with their forthrightness during my first
interview, and when they invited me back, I immediately took the chance. Since we had
already broken the ice in our first encounter, they were more comfortable with me during
our second meeting, and they could say anything with little or no embarrassment. I
recorded their conversation as they talked about various issues in their lives. Most of their
themes revolved around prostitution, consumption and sale of keroro (illicit alcohol),
illegal abortions, poverty, and violent deaths.

In a number of episodes, Kui and Joy expressed disgust at each other’s gross
behavior, but the accused simply shrugged it off with ni pesa ndiyo hufanya ‘it is because
of money’. Throughout the exchange, the issue of peddling their flesh for money comes
up constantly, and they say it casually with no embarrassment, which may surprise one
used to politically correct terms in the discussion on sexual topics. My analysis of their
discourse rests on the assumption that only insiders, whose daily interactions are
conducted in such a vulgar language would feel free to engage in vulgar discourse in the
presence of a stranger. In my first illustration, Joy had accused Kui of setting her up to
the police, and Kui had justified her action by claiming that she had to do it because Joy

was attracting more men than her. The conversation opens with Joy’s promise to revenge:

1. Joy: ...hiyo lembe unataka twende, tutaenda. Na mimi magava si tisho kwangu
unaona?

2. Kui: tutaenda. Hakuna noma. .. sawa hata wanakamu ninawapatia nyake hiyo...
na hiyo inaisha.

3. Joy: hiyo ni umbwau eee.

4. Kui: si ndiyo eee.

5. Joy: hiyo ni ushoga

6. Kui: eee, ee nai. . .(laughs)

182

Translation

1. Joy: ...if you want us to go to the authorities, we shall go. And police do not
scare me, you understand?
2. Kui: we shall go. no problem. . .its ok even if they come, I will give them the
vagina and that will be the end of it.

3. Joy: That is being bitchy

4. Kui: It sure is, yes,

5. Joy: that is homosexuality

6. Kui: yeees, yees and it... (laughs)
Joy is telling Kui that the policemen do not won'y her, and that she can also set her up.
Kui casually says that if the policemen come for her, she was going to buy her freedom
by sleeping with them. Every prostitute is aware that they can buy their freedom with sex,
when they do not have money to bribe policemen. Even when they have the money,
sleeping with policemen is a better option than giving away their ‘hard earned money,’ a
fact Joy should know too in spite of her accusations. One important thing to note here is
that Sheng has removed the taboo associated with the female genitals through use of
words like nyake. It should also be recognized that in mainstream discourse, this kind of
exchange would not take place. We would also expect Kui to defend herself against such
insults from Joy. Instead, she opts to keep mum, and actually laughs. This can be
attributed to the prevalence of such vulgarities in their everyday speech, to an extent that
they do not carry the insult as they do in mainstream discourse. This is borne by the fact
that Joy accuses Kui of homosexuality, while there is no evidence of that considering that
the majority of police are males. Hence, it is not possible that Joy is refening to them as
homosexuals, unless there exists a social meaning that views policemen as feminine. The

more likely explanation is that ‘bitchy’ and ‘homosexuality’ are socially descriptive

labels for ‘promiscuity’ and ‘whoredom’ among Sinai females.

183

A noticeable characteristic of these girls’ discourse is their heavy use of syllabic
inversion. In Chapter 3, I mentioned that Swahili words, previous or current Sheng words
have a chance of having their syllables inverted. Examples of such words in the above
exchange are:

30. Tu-ta-enda lembe from tu-ta-enda mbele

lsg-fut-go ahead
‘we shall go to the authorities’

u-mbwau from Swahili u-mbwa (obsolescent ‘dogness’)

15-dog

‘bitchy’
Considering the nature of their activities, where they survive by their wit, syllabic
inversion allows quick communication of messages among themselves in the presence of
their clients. It allows them to use words known to the clients but in a twisted form under
their very noses. Notice too that Kui uses nyake — a contested name among females, for
‘vagina’, instead of the widely used word senye. This means that taboo words and other
vulgarities do not bother these women. Indeed, we can say that society expects them to be
“whorish”, and they are just affirming what the society expects of them. If such terms
constitute part of their vernacular, then this should be a good illustration of their whore
identity.

In the next conversation, I we see how the differences between the whore and the

sex worker are played out. Kui had told Joy that she flies everywhere like a Nairobi fly,"7

in reference to her whoring activities. Joy justifies that she does so in search of money.

 

47 paederus crebinpunctatis or paederus sabaeus —an orange black and white beetle that
does not sting, but releases pederin toxins that sting the victims and leave an ugly rash.
The bug was attributed to the El Nifio weather phenomenon that devastated Kenya in
1997.

184

Having an upper hand, Kui proceeds to describe those kinds of Nairobi flies, and Joy
decides to turn the heat on her by shifting the topic to Kui’s mother.
1. Joy: ...eh mathako ni poko... mathako ni poko
2. Kui: ...hata wewe mathako ni poko. Mathako ni poko. Si ndiyo?
3. Joy: matha yangu si poko. Mathako ni poko, mathangu huenda upoko
akinitafutia unga. Hiyo ni job. Hiyo ni. . .hiyo ni janta...hiyo ni janta.. ..
4. Kui: eee. Aaa si ndiyo? Hata mi matha yangu anafanyanga job akinitafutia unga

si ndiyo?

Translation:

1. Joy: ...your mom is a whore...your mom is a whore.

2. Kui: ...Even you your mom is a whore. Your mom is a whore. Isn’t she?

3. Joy: My mom is not a whore. My mom is not a whore, my mom goes whoring

in order to feed me. That is a job. . .that is work. . ..that is work.

4. Kui: yees, ah, isn’t it? Even me my mother, work to fend for me, isn’t it?
The metaphor of the Nairobi fly is a good example of a dozen (more in the following
section). This, according to Labov’s (1972b) distinction between ritual sounding and
personal insult is a personal insult because it comments on reality. However, this reality
is accepted by both prostitutes, since the only way to earn money, is to be on the move
looking for ‘victims’. Therefore, Joy does not have to defend herself, instead she starts to
diss Kui’s mother for her whoring activities. In (1) she repeats ‘your mom is a whore’.
The reason for repetition is to cut off Kui who was still elaborating on the Nairobi fly.
Since Kui had put Joy at a disadvantage, she was not ready to cede the floor to Joy. When
claiming the floor, Joy had to repeat her statement to signal that she was now changing
the topic. This repetition works, and Kui, who was unprepared, had no good answer, so
all she could say was that Joy’s mom was a whore too. Her tag question ‘isn’t she?’ (2) is

evidence that that fact was known to both of them. This can be seen as a strategy through

which Joy draws the distinction between being a whore and a sex worker in (3).

185

Joy first contests that her mom is a whore, but she also knows that Kui is well
aware of her mom’s activities, so she decides to mitigate that by saying that her mom
does so in order to feed her. To her, her mom’s activities are justified because she is
engaging in a form of economic activity that sustains the family (line 3). In an
environment where poverty and deprivation stare you in the face, it is not possible to be
bound by the mainstream discourse of moral logic when it comes to matters of life and
death. The moral question becomes whether one can be able to put food on the table, and
any means towards this end is acceptable. In this regard, Joy’s mother is not just a whore,
but also a worker like any other. In line 4, Kui is still not impressed, and she says that her
mother also works to take care of her, but not in prostitution. Kui does not question the
legitimacy of prostitution as an economic activity, but she questions their mothers’
involvement. Elsewhere, she challenges her friend, si uko na senye kama mimi? ‘don’t
you have a vagina just like me?’ to imply that they can prostitute themselves instead of
letting their mothers do it for them.

Throughout the interview, Joy constantly rebukes Kui for not taking her ‘job’
seriously. She cannot understand why Kui would sleep with men who pay her nothing, or
give her very little money. Although to Joy everything goes so long as she is getting
money for it (ni pesa ndiyo hufanya ‘it is because of money’), there is a limit to what she
can do. She for instance displays her heterosexual preference and challenges Kui for
sleeping with a ‘boyfriend’ who engages in homosexuality. Note that she does not accept
Kui’s defense of money, because to Joy, what he gives Kui is too little to justify her

sleeping with a man with ‘abhorrent’ sexual behavior.

186

1. Joy: aaaaa kwenda huko! Ndio nakuulizaje? Doo gani hiyo hukupea?
2. Kui: ...hanipeangi doo lakini. ..

3. Joy: hiyo fmje... finje, finje ndiyo unai..unaita ganji?

4. Kui: hiyo ni ya chai na si na tunakunywanga na wewe? Senye wewe!
5. Joy: ...finje ndiyo una..unaita ganji?

6. Kui: si ndiyo? Hiyo ni pesa.

Translation

1. Joy: aaaaa forget it! I am asking you. . .how much money does he give you?
2. Kui: ...he does not give me money, but...

3. Joy: 50 bucks... 50 bucks, 50 bucks you are...you call that money?

4. Kui: that is for the tea, and don’t we drink it with you? You vagina!

5. Joy: ...50 shillings! You are...you call that money?

6. Kui: that’s right. . .that is money.

Compared to an American dollar, fifty shillings is about sixty cents. In line 1, Joy is just
expressing her disgust that Kui would sleep with someone who has gay tendencies. Kui
in line 2 is at first reluctant to admit that she is paid, because that would mean that she
does not get value for her money. But when Joy in line 3 tauts her that she gets only 50
shillings, Kui tries to save her face by telling Joy that it is just a token and reminds her
that they normally spend it together (line 4). Joy maintains that fifty shillings is
negligible, and Kui holds on to her position that it is still money (better little than
nothing). Such accusations continues to the very end of the interview, and ends with Joy
telling Kui that her so-called boyfriend cannot buy her a 20-shilling panties (about
quarter of a dollar), to which Kui replies that she does not really care.

This exchange shows Kui moving from the realm of commercial sex to intimate
friendship. After all, prostitutes or not, they have personal feelings like other people, and
money does not count in relationships; however, these relationships are not stable
because they are not good for business for a sex worker. Hence, they should be taken
within the context of Kui’s whore aspect. To Joy, the whole activity is looked at in

economic terms. We can even interpret her disgust for homosexual males from an

187

ecomonic perspective. A gay male is useless to a female prostitute, because there is no
money to be made from him.

Returning to Kui’s defense of her mom, we can argue that she is still bound by
some laws of mainstream morality when it comes to the social perception of prostitution.
She has not yet been emancipated from the stigmatization of her profession by the
mainstream society. She knows that prostitution dehumanizes the prostitutes, and that it is
better to use legitimate ways of making a living. The woman still remains a mother figure
who should be respected, something impossible if one is engaging in prostitution. On her
part, Joy, sees nothing wrong. Later, she chides Kui that the only reason that her mom
quit prostitution was because she became ill and ugly and could no longer attract men. On
the other hand, Joy has been liberated from the mainstream logic, which has denied their
lot a chance to get into legitimate money making activities. She does not seek to excuse
her mom’s prostitution, but justifies it as a profession.

Joy and Kui’s exchanges display their mastery of Sheng, which debunks the
common assumption that women are not good users of vernaculars. Because their clients
are Sheng speaking males, these women have to be up to date with their Sheng so that
they do not become disadvantaged in their exchange encounters. In addition, Sheng
affords them with a communication medium to mask the taboos associated with their
occupation. The sentences below are just a small sample amongst many that appeared in
their one-hour exchange. They show how Sheng is used in negotiation of ‘fees’ (1&5),
explaining the nature of their work (2&4), and mitigating their lifestyle (3).

l. kama unaweza mpatia thao mbili na soo punch, sawa. . .lakini ati ummange sare

if you can give her two thousand and five hundred, ok...but not that you screw her
for nothing... (Kui to a man who wanted to talk to Joy)

188

2. wenye mi huenda nao, mi huwanga biarasha nao
the ones I sleep with, it is usually on business terms (what Joy tells her potential
clients)

3. mimi ni malaya nimezoea kujitafutia na senye ya mine...
I am a prostitute who is used to earning my living with my vagina (What Kui told
a man who wanted to take her by force)

4. unaenda unadozi na yeye anakupatia ganji. ..

you go and sleep with him and he gives you money (Kui telling Joy how she likes

it with her clients)

5. asubuhi ananiachia rwabe an soo punch

in the morning he leaves me 200 shillings or 500 shillings (Joy talking about how

much she charges)

In 2, syllabic inversion comes up once more. This persistent strategy serves a double
advantage; it not only makes their variety unique, but it also allows them to conspire
against their clients by manipulating a familiar medium. The word mubatari [mofatari] —
‘the needy’ (Gikuyu orthography), for example, comes to mind. I said earlier that the
matatu men use it to refer to the police, but the prostitutes use it to refer to the matatu
men. Moreover its ambiguity is very appropriate because everyone is needy in one way
or another (Mbugua 2003). Even so, these Sinai females reconstitute it to mtafari, which
completely masks a familiar term that empowers them to speak about their stigmatized
profession and still save face. They are not the one in need, they only satisfy a need.

A look at the different senses of the name poko justifies the use of De Meis’
metaphors of the ‘whore’ and the ‘sex worker’. When Kui is informed that her
‘boyfriend’ sleeps around with other women and gives them money while he gives her
nothing she responds yeye nifala na ni poko ‘He is a fool and a whore’. In her anger she
says she was going to call it quits with her boyfriend and was going to tell him wewe

upige upoko yako na mimi nipige yangu ‘you whore yourself and I whore myself’.

Referring to Frankie, a friend who passed away, Joy asks ‘umeshawahi sikia tu poko

189

alikuwa amezaliwangwa na upoko ‘ever heard of a whore who was born a whore?’ And
finally, when Kui says she would prefer to remain with girl children if she leaves her
“husband”, Joy ask her mapoko? ‘whores?’ refening to the girl children. From these uses,
we see that a whore does not necessarily engage in the vice for money. Even males with
multiple sexual partners are regarded as whores, and girl children who grow up in the
ghetto are whores because that is the most likely opening for them. This contrast with the
upoko ‘prostitution’ as a commercial enterprise when these prostitutes make statements
like ...animange lakini anipe ganji ‘he can screw me but he must give me money’

From these prostitutes’ vulgar discourse, we see that we can isolate different
forms of identities that surface in their exchange. These identities range from poor ghetto
dwellers, whores, sex workers, as well as females, with emotional needs like any other
female who lives under mainstream norms. These identities come up in their vernacular
discourse that derives from their social realities, which is distinct from the mainstream
reality. Downes (19982195) remarks regarding vernacular use amongst a group of
speakers is very relevant in this case;

...the positive social significance of vernacular speech resides in the community

and culture of its speakers. This will often diverge from superposed norms.

Although overtly ‘stigmatized’, its actual ‘social meaning’ may be positive from

the point of view of those to whom it is the vernacular, and deeply tied up with
their identity.

Following this logic, we can conclude that the vernacular of the Sinai girls characterized
by the overt use of vulgar vocabulary accords them a unique identity. Although we might
assume that they are performing the expected identity by talking ‘dirty’, we notice that

they negotiate their place in the mainstream by softening taboo words in Standard

190

Swahili by masking the vulgar words in Sheng. I will briefly touch on other aspects of

their vernacular in the course of my arguments in the next Section.

4.4.2. Michongoano: performance of identity in verbal contest

Interpersonal interaction between close fiiends is not always characterized by concerns
about political correctness. Discourse practices that in ordinary discourse would arouse
tension and confrontation if conducted with strangers, can pass without fuss. The rules of
interactions are usually codified into discourse norms, and are well understood by the

participants. Consider the exchange between Joy and Kui below:

1. Kui: Kwenda! wacha hii sura yako inakaa ya Mjaka, Mkamba, Mnubi, kwenda

2. Joy:
3. Kui:
4. Joy:
5. Kui:
6. Joy:
7. Kui:
8. Joy:

Translation:

1. Kui:

2. Joy:
3. Kui:
4. Joy:
5. Kui:
6. Joy:
7. Kui:
8. Joy:

‘mfinfibi fiyfi! ’ ([monofi ojo] —— Gikuyu orthography)

...na vile unakuwa na mapua kubwa kubwa kama za Mlude.
kwenda! na wewe hizi zako zinakaa ka filimbi ya Jakubu.

si uangalie tu poa; zimekaa kama filimbi ya Jakubu?

wacha hizi rasa yako inakaa kama pikipiki.

unaudhika juu ya chali yangu anaipendanga tu mbaya sana.
Eee, wacha waipende.

hata chali yangu akinipiga ananipiganga juu ya rasa zangu....

leave me alone, your face looks like that of a Luo, Kamba, and Nubian,
go away, you Nubian!

...and the way you have a big nose like that of a Luhya.

stop it! And you, yours looks like Jakubu’s whistle.

just look at them carefully; do they look like Jakubu’s whistle?

and your butt looks like a motorcycle.

it pains you because my boyfriend loves it very much.

ok, let them like it.

when my boyfriend beats me up, it is because of my butt (i.e. he can’t
bear to see anyone else touching it)...

(Luo, Kamba, and Luhya are large ethnic groups in Kenya while the Nubians are a
minority ethnic group)

191

Kui’s remark in (1) implies that Joy’s mum is a prostitute. That is why her (Joy) facial
features resemble people from different tribes that her mother has slept with. Joy in (2)
pursues this line of prostitution and claims that Kui’s big nose resembles that of a Luhya.
In addition to tribal references, Kui uses a local reference in (3) ‘Jakubu’s whistle’ whose
meaning was lost to me. The only association between Jacob and whistle that comes to
mind is the Biblical passage on Zechariah 1022-12,"8 though I am not sure whether this is
what the girls had in mind. However, judging from Joy’s reaction in (4) there is no doubt
that it is something quite gross. Finally, Kui changes the t0pic of the diss to Joy’s large
butt, and Joy turns tables and makes the diss work to her advantage in (6&8). Both use
Sheng in their interaction and they are very comfortable with each other’s linguistic use.
The meaning of this exchange can be contextualized only in consideration of their

shared meaning. Joy does not react to the reference to her Nubian parentage because she

 

4" “The diviners envision lies, and tell false dreams; they comfort in vain. Therefore the
people wend their way like sheep; they are in trouble because there is no shepherd. My
anger is kindled against the shepherds, and I will punish the goatherds. For the Lord of
hosts will visit His flock, the house of Judah, and will make them as His royal horse in
the battle. From him comes the cornerstone, fiom him the tent peg, from him the battle
bow, from him every ruler together. They shall be like mighty men, who tread down their
enemies in the mire of the streets in the battle. They shall fight because the Lord is with
them, and the riders on horses shall be put to shame. I will strengthen the house of Judah,
and I will save the house of ‘Joseph’. I will bring them back, because I have mercy on
them. They shall be as though I had not cast them aside; for I am the Lord their God, and
I will hear them. Those of Ephraim shall be like a mighty man, and their heart shall
rejoice as if with wine. Yes, their children shall see it and be glad; their heart shall rejoice
in the Lord. I will whistle for them and gather them, for I will redeem them; and they
shall increase as they once increased. I will sow them among the peoples, and they shall
remember me in far countries; they shall live, together with their children, and they shall
return. I will also bring them back from the land of Egypt, and gather them from Assyria.
I will bring them into the land of Gilead and Lebanon, until no more room is found for
them. He shall pass through the sea with affliction, and strike the waves of the sea: All
the depths of the River shall dry up. Then the pride of Assyria shall be brought down, and
the scepter of Egypt shall depart. So I will strengthen them in the Lord, and they shall
walk up and down in His name, says the Lor .”

192

has already come to terms with it. (She actually claims Nubian parentage.) To retaliate,
she also hit at Kui’s Luhya parentage. Later on in the exchange, Kui admits that her
father was a Luhya, a pointer to the fact that they share mutual knowledge of each other’s
parentage. This is a reminder that mastery of interactive norms and the shared reality
enable the decoding of the intended message especially when making the distinction
between the literal and the idiomatic senses. Jakubu’s whistle, Kui and Joy’s prostitution
as well as their parentage are background knowledge that they draw from in their
interactions. This logic can be extended to all forms of interaction within the Cooperative
Principle (Grice 1975) whose central claim is that any discourse is a joint effort where the
speaker and addressee obey certain rules in order to communicate effectively. Shared
background knowledge (BGK) is the key to understanding the interpretation of discourse
as aptly articulated by F airclough (1995: 44):
The concept of BGK reduces diverse aspects of the ‘backgrounded material’,
which is drawn upon in interaction — beliefs, values, ideologies as well as
knowledge properly so called —— to ‘knowledge’. ‘Knowledge’ implies facts to be
known, facts coded in propositions, which are straightforwardly and transparently
related to them.
The centrality of BGK will become evident in the following Section in my discussion of
a discourse genre where some tenets of the cooperative principle are flouted without
resulting in a breakdown in communication. This discourse genre comes in the form of
verbal contest known as mchongoano, and it will be discussed in the context of

Goffrnan’s facework theory (1967), and Brown and Levinson’s (1978) politeness

strategy. Its relevance to identity will then be discussed.

193

4.4.2.1. Mchongoano within the general discourse of verbal art

The study of verbal art focuses on both cultural and linguistic skills of the speakers. As a
result, it has attracted the attention of both folklorists and linguists, who have studied it
under different titles such as ‘signifying,’ ‘playing the dozens,’ ‘sounding’ and ‘toasting’
among others. Labov (1972b), discusses the rules of ritual insults and their grammatical
properties such as clause embedding amongst the African American adolescents in the
New York ghettoes. On the folklore side, we have examples such as Dundes et al., (1972:
130-160) who looked at rhyming among the Turkish boys’ verbal dueling. Other studies
include Kochman (1972: 241-264), Mitchell—Kernan (1972: 161-179”), and Abrahams
(1964) among others. Most recently Marcyliena Morgan (2002: 84-110) has written about
conversational signifying, and a ritualized routine in face-saving discourse among
Afiican American girls. Morgan’s and Mitchel-Keman’s works are especially crucial
because they pursue the discourse beyond the focus on male. Their work shows that
women are also competent and do participate in these kinds of interactions, sometimes
coming up with their own kind of word play with a distinct structure. These earlier
studies inform my discussion in the rest of this Section.

A precursor to the discourse of michongoano in Sheng is to be found in the rich
African oral traditions. A good illustration is the Gikuyu circumcision activities and
Swahili poetry. Mbugua (2003) writes that amongst the Gikuyu, there was a three-week
period prior to the initiation ceremony when the Gikuyu world was turned upside down.
During this period, young men were expected to behave contrary to the expected social

norms, both before and afer the initiation. Mbugua discusses how during that period, the

 

49 See also in Kochman (1972: 315-335)

194

boys were allowed to address women with vulgar language and to engage in both
physical and verbal contests as the girls cheered from a safe distance. Girls chose as their
dancing partners, boys who won both the verbal and physical contests for demonstrating
their manhood and they shunned the boys who lost. It was thus important for the would-
be initiates to win such contests in order to earn respect among their peers.

In Swahili traditional poetry, two poets could give each other verbal challenges.
Such poetic verbal challenges were very similar to the rhyming among the Turkish boys,
but they differed in topics. Although Swahili poetry avoided the use of vulgar rhymes, it
is striking that both talked about emasculation of their opponents. The verbal duels
among the Turkish boys revolved around topics with sexual themes, but the Swahili
poetry revolved around the self-pride as illustrated by the poem simba mwenda pweke ‘a
lone lion’ by Ahmad Nassir (Ustadh Bhalo) in Malenga wa Mvita (1971):

Simba mwenda pweke

Ndimi simba ndume ndimi, ndimi mwelevu wa nyina

Niwazingae wasemi, kwa huja wakadangana

Ndimi mkingama ndimi, zisipate la kunena

Mungautambika mwina, na kunena jifunzeni

Its me the mighty bull, I am the expert of traps

When I surround the talkers, they gather in awe

I am the tongue binder, so that they can’t talk
If you were to set a trap, teach yourself to speak.

Ustadh Bhalo artistically throws down the gauntlet to his fellow poets by claiming to be
the first among equals. By comparing himself to the lion, regarded as the undisputed king
of the jungle in the oral traditions, he is symbolically claiming a high status while
implying that other poets should pay homage to him. There is the implication that since

the other poets are not his match, they were worthless. In order to salvage their honor, his

195

contemporaries had no recourse but to rise to the challenge, composing their rejoinders
with an equal show of verbal display. Failure to rise to the occasion meant that you had
averred to what your opponent has claimed.

We can look at michongoano within this context, though artistic composition is
not very elaborate. Instead, quick-witted one-liners are the goals of these verbal insults,
which should be viewed within the context of Sheng’s promotion of cognitive efficiency

(Kang’ethe 2004). A good illustration is in order:

Buda-k0 hu- hep-a job ku-swing na gate

9father-poss hab evade-fv job inf-swing with gate

Your father skips work to swing on a gate
In just one short sentence, a host of implications have been communicated. The mental
condition of an opponent’s father is put into question by reference to his immaturity.
Adults are not supposed to swing on gates, and to think that he had run away from work
to do it makes it even more belittling. Here, Sheng serves the purpose of shortening
communication while transmitting deeper sociocultural meaning. Notice the use of the
coined word Buda, from Hindi Buddha (Kang’ethe 2004, Githiora 2002) and the code
mixing in ku-swing. In just one instance we observe words coinage, Swahili structure,
and inserted English material. Following Myers-Scotton’s (1993a) social motivations for
codeswitching, we can View the instances of codeswitching as strategies that allow
contestants to claim local identity through the use of Swahili, while negotiating status
through the use of English. A different example is when Kui disses Joy with mathako
anaendanga kila bayesa ‘you mother visits every bar’, where in very few words, she

signifies that Joy’s mother is a prostitute who moves from one bar to another in search of

clients, a shared meaning retrievable through BGK.

196

Lots of parallels can be drawn between michongoano and the African American
playing of dozens. In my intemet search, I discovered that some of the michongoano are
in fact ‘Shenginized’ versions of ‘yo mama ’ jokes made popular by African Americans in
the United States. However, it would be wrong to imagine that they are directly copied
without any input by the Sheng speakers. In most cases, yo mama is truncated and

9

replaced with you, or given the beginning ati we... ‘that you are... as in examples

below. The Sheng joke is glossed literally, and then compared with its ‘yo mama’ joke
counterpart (in italics).
(a) wewe ni m-jinga mpaka u— li- fail blood test
You cop lsm-fool until ZSg-past-fail blood test

You are foolish such that you failed blood test
Yo mama ’s so dumb, she couldn't pass a blood test.

(b) wewe m-fat mpaka u- ki- va- a raincoat ya yellow people shout taxi!
You lsm-fat until ZSg-cond-wear-fv raincoat of yellow people shout taxi!
You are so fat that if you wear a yellow raincoat people shout “taxi”

Yo mama ’s so fat, when she wears a yellow raincoat, people call her “Taxi ’

(c) ati wee m- so u- ki- jump kwa air u-na get stuck!

That you lsm-big ZSg-cond-jump in the air ZSg-pres get stuck

You are so big that if you jump in the air you get stuck!

Yo mama ’s so fat, she jumped up in the air and got stuck
We can attribute this cultural borrowing to the Kenyan diaspora in the United States, and
the discourse practices in the internet that make ‘yo mama’ jokes accessible to Kenya’s
virtual community. Once they are used in the intemet forums, they easily infiltrate the
normal daily discourse of Sheng speaker in Nairobi and elsewhere. As we can see above,
Sheng speakers modify these jokes to fit the structure of their discourse through
pervasive codeswitching. We also note that some realities are not local but international.

In (b) for instance, reference to yellow taxis indicate that the joke’s origin is foreign,

most probably the United States since in Kenya, taxis are not yellow. It is also

197

noteworthy that the borrowed dozens directly confront the target instead of being directed
to someone’s mother, though in these intemet forums, they are usually not directed at
anybody in particular. This means that no one is likely to lose face, and, even if that
happened, the virtual distance prevents the danger that could ensue in face-to-face
interactions. Before I discuss the relevance of michongoano in the identity project, I will
briefly compare the structure of michongoano with the template adopted by Labov

(1972b) in his analysis of the African American ritual sounding.

4.4.2.2 Structure of michongoano

As a discourse practice, Labov (1972b) proposed that ritual sounding followed a general
syntactic pattern characterized by clause embedding or deletion. He proposed the
following structure;
T (B) is X that P
Where T is the target of the sound, X the attribute of T, which is focused on, and
P is a proposition that is coupled with a quantifier “so...that” to express the degree
to which T has X. B refers to the person the sound is directed at.
A similar attribute/comment structure is manifested in michongoano, where the first part
defines the attribute(s) of the opponent, members of his/her family or anything else
associated with him/her, while the second parts makes a comment on that attribute, as
illustrated in the following examples.
(1) buda-ko ni m- chafu mpaka umbu hu-tumi-a straw
9father-your cop lsm-dirty until mosquito hab-use-fv straw
ku-m- nyony-a damu.
inf-3sgom-suck 9damu.

your father is so dirty such that mosquitoes use a straw to suck blood from
him

198

 

 

(2) ati wewe ni m-chafu mosquito i- ki- ku- um- a kwanza
that you cop Ism-dirty mosquito 9sm-cond—2sg bite-fv first
i-na-end-a ku-brush meno.
9-pres-go-fv inf-brush 6teeth
you are so dirty that if the mosquito bites you, it goes to brush its teeth

(3) hao yenu soo u- ki- tok- a back door u- ko ocha
9house your huge 2sg-cond—comefi'om-fi2 back door 2sg-loc 17 countryside.
your house is big that if you exit the back door you are in the rural area
In (1), the target attribute is the opponent’s father’s untidiness, with the comment being
the joke that a disease-causing insect has superior health standards. The same theme can
be seen in (2), with a slight variation; however, the target switches to the opponent
himself. In (3), the opponent’s house becomes the attribute, while the comment just adds
value-laden information. Ocha is a Sheng word borrowed from Gikuyu gicagi for
‘village’. In Sheng, ocha or its other variants oshago and shaggs refer to the remote rural
areas. We can make the obvious deduction that the participants live in urban areas,
presumably Nairobi. The imagery of a house with the front door in Nairobi and the back
door in the rural area is lefi to the reader’s imagination.
Labov also claims that in cases where the overt structure fails to display the ‘T
(B) is X that P’ sequence, a rule of deletion similar to the syntactic rules of ellipsis has
taken place. In his example “your mother the abominable snowman”, what is lacking,
apart from the copula, is the attribute and the quantifier, but the comparative metaphor is
still there. Participants familiar with such discourse will know right away that the
attribute has to do with the opponent mother’s massive figure. In light of this, we can say
that discourse norms make the meaning accessible to the participants due to the

intersubjective reality of the participants. The shared background knowledge makes it

possible to recover the truncated materials. A similar argument can be extended to

199

michongoano, especially in the jokes that begin with ‘asubuhi asubuhi’ ‘morning
morning’ as in these examples:
(i) asubuhi asubuhi gazeti na harufu ya socks

9moming 9morning Snewspaper and 9smell of socks
You guys eat newspapers and stench of socks for breakfast

(ii) asubuhi asubuhi kokoto na kumikumi
9morning 9morning 9gravel and 9kumikumi
You guys eat gravel and kumikumi for breakfast

(iii) asubuhi asubuhi goodygoody50 na makojo ya ngamia
9morning 9morning goodygooay and 6urine of 9camel
You guys eat goody goody and camel urine for braekfast

(iv) asubuhi asubuhi kokoto moto na glasi ya usaa

9morning 9morning 9gravel hot and 9glass of I 1 pus

You guys eat hot concrete and a glass of pus for breakfast
Although these kinds of michongoano are not laden with deep metaphors, they are rather
problematic because their meaning cannot be accessed through overt syntax. However for
those familiar with these kinds of jokes, they are able to follow the pattern in the recovery
of the missing material. Background knowledge allow speakers to relate the surface
structures with the similar untruncated structure they might have encountered elsewhere,
such as kwenyu nyinyi huamkia maembe na maziwa mala, au glue na kokoto “ in your
home you start the day with mangoes and fermented milk or glue and gravel”. Using such
a cue, we can gloss (i) above as “in the morning, you start the day by eating newspaper
and the stench from socks” This pattern can be applied to the rest of this kind of truncated
jokes. Still we realize that we are not done yet. We have only reconstructed the comment
while the attribute is lacking. Appropriate attributes can either be poverty or foolishness.

The reconstructed structure should go like ‘You are so poor (foolish, gross) that in the

morning, you begin the day by eating newspapers and the stench from socks’, The same

 

50 A kind of bubble gum popular with boys and girls

200

can be said of metaphorical jokes such as sura kama dogi ya Bush ‘face like Bush’s dog’,
or sura kama bakuli ya jela ‘face like a prison bowl’ as comments about someone’s ugly
face. This then is a confirmation of Labov’s claim that these jokes follow from the same

underlying structure.

4.4.2.3. The mastery of verbal art: playing by the shared rules

Having discussed the structure of michongoano along the lines of African American
playing of dozens, we now proceed to draw other parallels in the way this discourse
activity is conducted. Again, I adopt some of Labov’s rules governing ritual insults as I
develop my discussion of their relevance in the identity project. Three of these rules are
1) the reply of a sound is a sound, 2) the proposition in sound must not be true, and 3)
sound are never denied. To understand rule 1, ‘the reply to a sound is another sound’, let

us consider the “Boondocks” cartoon by Aaron Gruder below:

 

(UM-MAY,WEL-

 

 

ml.

   

Source: www.livejournal.com/community/boondocks comic 03/25/2005
Figure 6. The reply of a sound is a sound

In the above cartoon, Caesar, the dreadlocked character has the option of opting out, or

changing the discourse activity; however, he gives a rejoinder that retains the theme of

bad breath. In these cartoon strips, Riley, his opponent is usually portrayed, as a social
deviant who glorifies what is abhorred by the society. Terrorists are his heroes, which is
the reason he accuses Caesar of doing political humor. It is obvious that he does not want
to continue with the discourse that demonizes his heroes, so he says that he does not do
political humor. In a way, he is looking for a way out, or changing the topic, but he wants
to do it without losing face or flouting the rule.

Likewise, a michongoano contest is a dual activity where a person taunted must
reply in another mchongoano. One does not opt out of the conversation, for doing so can
diminish one’s status among his/her peers, especially if there was an audience at the time
of exchange. Failure to respond could either mean that you lacked the necessary verbal
skill to match your challenger or that what was said was true. A participant has no choice
between responding and going away because either case would lead to loss of face. To
resolve this dilemma and save face, the responder must deal with rule 2; ‘the proposition
in the sound must not be true’. Accepting the hyperbolical nature of michongoano
enables an equally exaggerated retort.

Operation of rule 3, ‘sounds are never denied’ in Sheng can be seen in the
exchanges between Ejaab and Hater, and Chokoraa and Baba_ndogo both taken from
kenyaniyetu. com website:

Hater: Ejaab, asubuhi asubuhi close up na ngozi ya ngamia

9moming 9moming closeup and 9skin of 9camel
You guys eat close up and skin of a camel for breakfast

Ejaab: Hater, asubuhi asubuhi, magoti ya mende, sawdust na gazeti ya nation
9moming 9moming, 6-knees of 9roach, sawdust and paper of nation

You guys eat roach joints, sawdust and Nation newspaper5 lfor
breakfast.

 

5' The Nation is one of the leading Newspapers in the East and Central African region.

202

Hater: Ejaab, asubuhi asubuhi kokoto moto na glasi ya usaa
9moming 9moming 9gravel hot and 9glass of l l pus
You guys eat hot gravel and a glass of pus for breakfast

Ejaab: Hater, asubuhi asubuhi jasho ya murungaru na soks za mwiraria
9moming 9moming 5sweat of murungaru and 9socks of mwiraria

You guys eat Murungaru’s52 sweat, and Mwiraria’sS3 socks for
breakfast.

Hater tries to portray the kind of disgusting food that Ejaab and his family eat for
breakfast. Ejaab’s retort is equally effective when he counters that what they eat at their
home is nothing compared to what Hater’s family eats. Fearing the loss of face, Hater
gives a more devastating retort by showing the disgust at Ejaab’s morning dish. Ejaab
effectively counters Hater’s move, while keeping within the breakfast topic, but
introduces political characters that majority of Kenyans detest.

Note that I have used the concept of ‘losing face’ because of the intemet’s
interactive features that enable the audience to participate. When one has been bested,
other intemet users comments on it by quoting the diss and then adding accompanying
smileys (e.g., @, ©,) or comments such as hapo umewezwa ‘there he got you’ or hiyo ni
kali ‘that is a tough one’. These intemet strategies work similar to laughs, sighs and
clapping among the real audience in face-to-face interactions. The intemet, however, has
its disadvantage because we cannot say whether participants have opted out, or
differences in time zone has led to the termination of a michongoano exchange. Still, we
see that a sound is responded to with another sound, mostly on the same topic. On the

same intemet thread, the exchange between Chokoraa and Baba_ndogo represents

 

52 Murungaru: an unpopular cabinet minister known for his sweating and always
associated with high level official corruption.
53 Mwiraria: Minister of Finance, also unpopular with Kenyans

203

another type of michongoano genre. The dyad below is in the ‘challenge and retort’

adjacency pair format.

Baba_ndogo: Chokora, weeee ni m- zeee mpaka vile Noah a- li- jeng-a

Chokoraa:

Baba_ndogo:

Chokoraa:

Chokora you cop Ism old until when Noah Isg-past-make—fv
ark weee ndio u- li- lip- ish- a bu fare

ark you cop (emp) 2sg-past pay-caus-fv bus fare

‘Chokora, you are so old that when Noah built the ark, you were
the one who charged the fare’

Baba_ndogo, wewe m- zee hadi passport yako
Baba_ndogo, you Ism-old until passport your

i- me- andik-w- a kwa scroll

9sm-perf- write- pass-fv with scroll

‘You are old such that your passport is written on a scroll’

Chokora, weee ni m-kondee ati m-vua i- ki- nyesh-a yii
Chokora, you cop Ism-thin that 2-rain -9sm-cond-fall—fi2 it-
u- kuu- hat-a

2sg-you om sweep-fv

You are so thin that when rain falls, it sweeps you off

Baba_ndogo, wewe m- fat mpaka u- ki- va- a raincoat ya
Baba_Ndogo, you Ism-fat until 2sg-cond-wear-fv raincoar of
yellow people shout taxi!

yellow people shout taxi!

You are so fat that when you wear a yellow raincoat people shout
taxi!

The first adjacency pair connects the distant past and modernity in commenting about

someone’s age. Baba_ndogo draws the biblical analogy of Noah and the ark to comment

on Chokoraa’s age. Chokoraa cleverly sticks to the biblical analogy of scrolls. A person

who is conversant with Biblical stories understands that stories of the Noah’s ark were

first written on scrolls. Therefore, if Baba_ndogo’s passport was written on a scroll, then

he has no right to comment on Chokoraa’s age since they both belong to the era of

scrolls. The second adjacency pair juxtaposes the idiom of skinniness with that of fatness.

204

Who is better, a very skinny person or a very fat person? This is lefi to the audience to

judge.

4.4.2.4. Michongoano as face threatening acts

The kinds of topics that pervade michongoano jokes are highly volatile. They constitute
what (Goffman 1967, Brown and Levinson 1987) call face-threatening acts due to their
overt commentaries of absurd attributes of the interlocutors or members of their close
family. Goffman’s facework theory and Brown and Levinson politeness theory provides
good analytical tools for this kind of discourse. The key issue here is that people want to
be positively perceived by others, which requires the participants to be considerate so that
they do not make each other to be perceived negatively. This leads to self censorship as a
politeness strategy, where participants bow to the dictates of certain norms. Since
engaging in discourse that can lead to loss of face may threaten the group’s cohesion,
especially if a member ends up getting hurt, certain discourse types are normally avoided.
Atematively, only the people who share background knowledge to undertand the rules
involved engage in them.

In michongoano discourse, a communicative dilemma has to be resolved. While
on the one hand they are geared towards humiliating an opponent by putting him/her
down, they are on the other hand intended to entertain and not to hurt, but there have
been cases where they degenerate into physical confrontations. In an intemet forum
mashada.com, Chizzy recounts an experience in primary school, when a fellow pupil told
him; ati nyanyako ana laini ya red kwa kunyi ‘that your grandmother has a red line on her

butt’. Unable to bear the insult, Chizzy pounced on the hapless pupil. In this case, his

205

 

anger cannot be attributed to the fact that the instigator commented about the age of his
grandmother since grandmothers are supposed to be old. It is because he implies she is
abnormal (a red line on her butt), as understood within the traditional context of taboo
concerning female nudity —— an abnormality on someone’s butt can be noticed only when
that person is naked. ‘Ati’ here indicates that the instigator is reporting a circulating
rumor. Chizzy’s report shows the lurking danger when playing michongoano in face-to-
face interaction, especially if they touch a raw nerve. To avoid such ugly scenes,
michongoano are normally played among close fiiends who understand the established

norms governing this discourse activity.

4.4.2.5. Affirming identity: Michongoano as an in-group activity

People who are outside the subculture might not be aware that verbal dueling is in session
because it is often spontaneous. In the Boondocks cartoon below, Riley sounds on his
friend Caesar in the presence of his mixed-race neighbor girl, Jazmine. With a father who
shuns black culture and white mother who doesn’t know it, Jazmine lacks the ethnic
heritage and hence BGK necessary for comprehending the rules governing the dozens
discourse. She is bewildered by the ‘insult’ Riley hurls at Caesar. Note that Caesar is not
bothered and does not defend himself because he understands the rules and interprets it as
a sound. He knows that sounds are never denied because they are not supposed to be true.
When Riley pretends to apologize, he throws another barb at Caesar leaving Jazmine

more dumfounded as seen in Figure 7:

206

 

 

 

RILEY! SW M YOU!

 

 

 

(source: www.1ivejoumal.cm“’ "y," J ' _comic, 03/25/2005)
Figure 7: Outsiders neither know the safe topics nor the rules in ritual insults

This cartoon shows that even though these three youth constitute a group of fiiends, they
have different identities. Jazmine’s identity is salient as an outsider both in terms of her
gender as well as her mixed race background.

Relating this to group identity, we find that among strangers, it is hard to know a
safe or unsafe sound (Labov 1972b). Consider my example above of Chizzy fighting with
a boy who said that his grandma had a red line on her butt. If Labov’s three rules had
been have been satisfied, then we can attribute the physical confrontation to the fact that
Chizzy and the boy were not close buddies, or that the context was not appropriate.

Furthermore, Labov also distinguishes between a personal insult and a ritual
insult. Personal insults, are truthful commentaries while ritual insults are hyperbolical
jokes. Unlike ritual insults, therefore, personal insults are characterized by denial, excuse
or mitigation. An outsider from a different baze does not have all the facts to know when
he is operating within the felicity or infelicity frameworks. Failure to make this crucial
distinction may lead to personal insult when none was intended. We can interpret the
altercation between Chizzy and the boy in this context, since in normal circumstances

Chizzy could easily have reciprocated with his own retort. This leads us to conclude that

207

michongoano discourse is an in-group activity. Only the insiders have the necessary
background knowledge to know the topics safe for michongoano, and the taboo topics
that should be avoided at all costs. In a close-knit group of Mabeshte, such as the two
Sinai female respondents, each understands the temperament of her fellow colleague.
They know when to stop when they touch a raw nerve. In the same breath, a beshte is
usually tolerant of the verbal transgression of a fellow beshte so that s/he does not
threaten the solidarity of the group. A beshte who interprets a mchongoano as an insult
where none was intended cannot be a true beshte, and risks losing face. This would be a
clear indication that his/her hierarchy is at the group’s periphery.

Earlier, I defined Mabeshte as close buddies who hang out together, mostly in a
baze. Parents and other older people are usually excluded from these bazes. As outsiders
they cannot participate in michongoano, though they often constitute the subjects of
michongoano. In any case, they would only play michongoano with their fellow age

mates, but not their kids or grandchildren. Again let us look at a Boondocks cartoon:

 

 

 

 

 

cmxwm‘rm-w MW‘MW
, (adv ‘
‘M‘a' .

 

Source: www.1ivejournal.co.m'- "J," ‘ !u_comic 03/26/2005

 

Figure 8: Verbal duels are an in-group thing

Riley gets a whacking for engaging in an age-graded activity with his granddad. He

thinks his granddad does not understand the joke, but the reality is that these two people,

208

 

though relatives, belong to different in-groups and therefore interpret reality differently.

The family identity is not salient as the in—group identity in these kinds of discourse.

4.4.2.6. The language of delivery

The mode of delivery of michongoano is a key ingredient to this discourse activity. Since
Swahili is a mainstream language, delivering michongoano in Standard Swahili or
Standard English, for instance, signals a shift from the joking mode to a serious mode.
Delivery in Sheng, on the other hand, signals that a non-serious friendly discourse is in
session. Following Hymes, (1974:58-60), I propose that Sheng is the instrument, or form
of speech that makes the michongoano interesting and unique. Because a lot of
michongoano are borrowed from English, translation into English will not greatly affect
the joke; however, translation into Standard Swahili often affects the quality of the joke.
As an illustration, let us compare a mchongoano in Sheng with its equivalent in Standard

Swahili. This comparison might not make sense to non-Swahili speakers.

Kui: We kw-enda u- na-flaiflai-ngi everywhere kama Nairobi fly
You inf-go 2sg-pres-fly- hab everywhere like Nairobi fly
Wewe nenda, unapaapaa kila mahali kama inzi/mdudu wa Nairobi
go away, you fly everywhere like Nairobi fly.
To a Sheng or Swahili speaker, the bold italic (Standard Swahili translation) has
implicitly been shed of the humor that characterizes this type of verbal exchange. We
also find that lexical equivalents might be so hard to translate, as in Nairobi fly“, but this

is easily handled in Sheng through codeswitching, which salvages the joke. Similarly, we

note that ‘yo mama’ jokes are delivered in vernacular; a lot of humor would be lost if

 

5" lnzi/mdudu wa Nairobi is ambiguous. It might mean a certain fly, or any kind of fly in
Nairobi.

209

 

they were delivered in Standard English. From this, we say that delivery of jokes in non-
standard idioms seems to be a cross linguistic phenomenon.

In this Section, I have argued that if michongoano is an in-group verbal discourse
delivered in Sheng, then the non-Sheng speakers will always constitute the out-group
every time this verbal discourse is in progress, as a result of their limited verbal skills. In
addition, mastery of lexical features in itself is not sufficient when engaging in these
kinds of discourse. One has to be a regular participant in other baze activities in order to
know the topics that are safe to discuss and those that are not. The michongoano that do
not reflect the local reality are inaccessible to the insiders, but those that reflect the local
reality are inaccessible to the outsiders. However, the michongoano that border on the
national reality are accessible to all people who are conversant with this type of
discourse. Furthermore, we note that the international reality has been incorporated into
the local reality through adoption of African American ritual insults (sounds or dozens).
All these factors have a very big impact in the shifiing of identities among the Sheng

speakers in Kenya.

4.5. Other ways of constructing identities

Although my major focus has been on linguistic variation and the discourse practices in
the construction of identity, I have claimed that linguistic practices alone are not enough
to complete the identity project. They combine with other discursive practices in creating
meanings of any category or group. Dere, a Kabete respondent, talked of groups of
marijuana smokers whose Sheng was incomprehensible because it was highly coded.

Admission into such a group means mastering their Sheng; however, knowing Sheng

210

may not enough; you must also be a marijuana smoker in order to gain membership.
Mbugua’s (2003) account of the meaning created by the matatu crew through their trendy
dressing styles as a way of creating group visibility can be evaluated along the opposite
end of the spectrum with respect to these marijuana smokers. When certain norms and
practices become the distinctive features of a peoples’ patterned life, they come to be
regarded as marker of the people’s identity. In this respect, football, baseball and
basketball are connected with the American identity more than any other country’s
identity due to the role they play in the American national culture. In the same context,
eating vegemite can identify one as an Australian (Rozin and Seagal 2003). In the next
Section, I will discuss some of the non-linguistic practices that are important in the
construction of identity. I will examine identity practices such as dietary and dining

habits, way of dressing, and style and choice of music.

4.5.1 Diet and dining habits

Both the Sinai and Kariobangi female respondents claimed that they could differentiate
between a Mbabi and a ghetto person by looking at the kind of food that they eat as well
as the dining manners. While the ghetto people would eat anything, the Wababi will first
assess the hygienic condition of the place where they are eating, order an expensive dish
of a particular taste, then ‘proudly nimble’ on the food, and throw away the rest. The
ghetto people would consider this wasteful. To them the most important thing is to have
something in their stomachs, as Edee and Shiro, two female respondents from

Kariobangi, state in the following conversation:

211

Nu—o

3.
4.

. Shiro: watu wa ghetto most of them [most of the ghetto people]
. Edee: tumezoea kila kitu mchanganyiko [we are used to a mixture of

everything]
Shiro: eee mchanganyiko isipokuwa mawe. [yes, mixture except stones]
Edee: mawe tu ndiyo hatuwezi kula. [it is only stones we cannot eat]

For these females, choosing and being picky are luxuries they cannot afford. Unlike the

Wababi, they are not concerned about the hygienic environment, in which the food is

cooked or sold:

hWNU—

\OOOQONUI

. Edee: unaona venye sa chakula. . .zile machips za kobole kobole vitu ka hiyo
. PG: aha

. Shiro: chips ni chips.

. Sera: kuna sewage inapita hapa, na zinapikiwa hapa. Sasa mtu wa babi

hataweza kula

. PG: hawezi.

. Pame: lakini mi hukula

: Edee: hujali, umekula germs nyingi. ..

. PG: aha

. Shiro: yeye hawezi kula na sewage inapita.

Translation:

bWNr—

~000an

. Edee: you see like those chips sold for 5 shillings. . .things like that

. PG: yes

. Shiro: chips are chips.

. Sera: there is a sewage flowing here and they are being cooked here, now

for a babi fellow, s/he will not eat.
PG: s/he can’t.

. Pame: ...but I do eat

: Edee: you don’t care, you have eaten a lot of germs

. PG: aha

. Shiro: s/he can’t eat and the sewage is flowing close by.

Ghetto people are sustained by Jua kali (informal businesses) that exploit every open

space available to conduct business — whether by the side of a sewage trench (4) or

garbage dumps. The ghetto people are used to this lifestyle and consider it normal, ‘chips

are chips’ (3, 6, 7), no matter where they are made. For the hygiene-conscious residents

of Ubabini, where sanitary malfunction is reported and repaired immediately, such

212

unhygienic conditions are health hazards to be avoided at all costs (4). Notice too the
price of chips in the ghetto, five shillings (0.07 USD). Higher hygiene standards will call
for higher prices, which the ghetto people cannot afford. This distinction between the
Wababi and the ghetto in terms of diet and dining practices is an extension of the socio-

economic discourse mentioned earlier.

4.5.2. Dressing

Dress style is another way of constructing an identity. At the beginning of this Section, I
mentioned Mbugua’s claim that the matatu crews use trendy dressing style as one way of
constructing their identity. Both Mbugua (2004) and Samper (2002) assert that the matatu
crews are the trendsetters when it comes to dressing styles. Apart from matatu crew, it is
common in Kenya to see the reggae fans from the ghetto sporting the colors associated
with Rastafarianism. The Rasta colors and dreadlocks are visible marks of the Rastafarian
identity — although not every dreadlocked person subscribes to the Rastafarian ideals.
We can also look at the prostitutes dressing in revealing attire as an affirmation of their
identity as flesh peddlers, though this is not always the case.

When drawing the contrast between the mababi and maghetto, I mentioned that the
Wababi have been heavily influenced by western tastes, unlike the ghetto people. This
contrast is very noticeable and is normally overplayed as a feature that can distinguish a
Mbabi from a ghetto person as stated by the Sinai respondents below:

PG: utajuaje Mbabi pale?

. Liz: ee babi namjua

3. Kui: unajua babi huvaanga vitu ya fleya fleya hivi hivi tu zinafagia chini chini
lakini mnati. ..

4. Joy: nikwambie kitu moja, mtu wa ghetto, hata ukiwa wewe ni wa ubabini,
unamwelewa tu. Juu kung’ara yake,

Nr—A

213

. Liz: staili yake ...iko, unamwelewa tu

5
6. Kui: ..., na yeye huyu wa ghetto huvaa nguo za wapi?... za Gikomba.
7. PG: aha

8. Kui: Nguo anavaa inakaa unaona wife yake anavaa imemshika.

9. Liz. (unclear) sisi tunavaa za Gikomba

10. PG: Si lakini wajua sasa kuna mtumba, hata wewe ukienda kule Gikomba..
11. Joy: eh?

12. PG: Unaweza kupata nguo ambayo inakaa kama ya Mbabi siyo?

13. Joy: Ndiyo lakini ni..ngumu, ni ngumu

14. Kui: siwezi buy kwa sabau najua hii ni ya babi.

 

Translation:
1. PG: How will you know a Mbabi?
2. Liz: ehh, I know a babi.
3. Kui: A babi wears things that are flair and they sweep the ground, but a mnati. . ..
4. Joy: Let me tell you one thing, a ghetto person, even if you are from ubabini, you

will understand her. Because of the way she dresses.

5. Liz: Her style. . .will, you will understand her.

6. Kui: ..., and the person from the ghetto wears the clothes from where? ...from
Gikomba”.

7. PG: aha

8. Kui: the clothes he wears, you will see his wife wears a tight one
9. Liz: (unclear) For us we wear from Gikomba
10. PG: but you know that now we have mtumba, even you if you go to Gikomba..

11. Joy: eh?

12. PG: You can get a dress that looks like that of a Mbabi, isn’t it?

13. Joy: Yes, but it is...hard, it is hard

14. Kui: I cannot buy it because I know it is for a Mbabi.
A Mbabi is constructed by the others, not just as a rich person, but also as a person deeply
immersed in modernity as represented by dressing style — flair pants for men and tight
clothes for the ladies (3). It would be wrong to assume that the ghetto people reject these

clothes because they cannot afford them. Unlike food, they can buy mitumba (second

hand clothes) sold cheaply in Gikomba, as I pointed out in (10). They simply reject doing

 

55 Gikomba is an informal open-air market that specializes in the sale of second hand
clothes commonly known as mitumba.

214

that because they associate these clothes with Ubabini (14), and they would make them
look like people that they are not. Just like the lifestyle of the Wababi will prevent them
from completely acquiring the ghetto identity (Section 4.3.2), economic reasons prevent
the ghetto people from engaging in other practices that complete the Mbabi identity.

In view of what I discussed in Section (4.2.2), where we saw the ghetto pe0ple
aspiring to be like Wababi, this seems to present a paradox. But it is resolved when we
take into account that we were then dealing with a different baze (Kariobangi) that
interpret reality differently from the Sinai baze. In addition, identities are influenced by
different factors, one of which is context. A group might construct a certain identity in

one context, but construct a completely different one in another context.

4.5.3. Style of music

Ubabini and ghetto people have different tastes of music. The ghetto people have been
known to prefer reggae music while the Ubabini people have always moved with the
current American musical trends. A Makadara respondent remarked that they listened to
reggae while the Wababi listen to the slow jam only, which they called ‘ma—yo! yo! pekee
yake’ [yo yo only]. Sera, a female respondent from Kariobangi, also expressed similar
sentiments: sisi ni wasee wa reggae ‘we are reggae guys’. This bias may also be
attributed to their socioeconomic differences, which influence their lifestyles. American
popular culture has mostly been transmitted to the Kenyan youth through movies and
music videos. Yet owning a TV, let alone a VCR, is beyond the reach of the majority of
ghetto youths. Those who can afford a TV have no money to buy their own music videos

or pay the subscription fees for entertainment channels. This cuts off from the popular

215

music programs through which they could access these popular American music.
Furthermore, the kind of lifestyle portrayed in American music is alien to the ghetto
people due to its unafordability. In contrast, reggae music strikes a familiar chord within
the deprived lives of the majority of ghetto residents, and this makes them identify with
the sufferings and the aspiration of the African diaspora as espoused in the lyrics of

various reggae artists.

4.6. Construction of identity in other discourse sites

In the greater part of this Chapter, I have located the construction of identity as taking
place in the bazes. I finished by discussing other non-linguistic ways of constructing
identity in various bazes. In the next three Sections, I will briefly discuss the other sites
where identities are constructed such as hip-hop, mass media, and the intemet. Although
these can also be regarded as other ways of constructing identity, they will be treated as
sites where Sheng discourse takes place. I begin with hip-hop within the context of the

popular culture.

4.6.1. Popular culture: Sheng hip-hop

Popular culture can be considered as a discourse site, which incorporates both the input
and output of identity practices as two parts of dialectic. I mentioned in Chapter 2, that
the radio stations have strategically placed themselves to respond to the market forces.
This is similar to Bell’s (1984, 1991) account of how the newspapers in the UK, or radio
stations in New Zealand, adopted the language of their target audience. When radio

stations play the music that is popular with the young people, or relax their language

216

policy so that they accept and employ Sheng in their broadcast, they not only identify
with what is hip, but also participate in creating and enforcing certain identities of the
Sheng users. This makes the media to become important channels for the codification of
certain mainstream identities, which are then re-modified by the listeners to suit their
different needs. The media also participate in word coinage, which then infiltrates into
the daily discourses of Sheng speakers. However, unless such media lexical items are
taken as the norm of an in-group within a given media house, they cannot be used in local
construction of identity since they are too mainstream to serve in-group purposes.

Hip-hop music itself draws from the Sheng that is in vogue, but the musicians are
very conscious of the kind of language that they can use. They want to appeal to a larger
audience and would not like to alienate their potential audience by adopting the variety of
Sheng that is too vulgar or too hard for the audience. Nevertheless, hip-hop musicians
assume different identities as far as Sheng is concerned. In their music, they use the most
basic Sheng, while in live performances, they may use deep and vulgar Sheng (Samper
(2004) — although such usage may not have been premeditated. In different contexts,
these musicians are also members of estates and bazes where they also hang out with
their friends and catch up with what is trendy. Due to their celebrity status, they attract a
huge fan base that crosses categorical borderlines. This requires them to align themselves
differently in order to interact with different fans in different situations.

To appreciate the role that Sheng hip-hop plays in the lives of the Kenyan youth,
one needs to see how much it surfaces in their ordinary discourse. During my interview
with the Kariobangi female respondents, Pame and Koki, poked fun at me by quoting

lines of a popular song and then pretended that that it was their authentic talk:

217

Pame: Mbuyu alinileta juu

(pause)

Koki: Ishia. . .usiletewe noma za budako (laughs)
Pame: Saa hii anajua tu nimeenda laibo na niko baze.

[laughs]

Translation:

Pame: My father was very mad at me

(pause)

Koki: you should leave so that you do not run foul with your father (laughs)
Pame: Right now he thinks I have gone to the library and I am out at the haze.

Most of these words are borrowed from a popular song by Nonini called ‘Wee Kamu’

which has the following lines;

usiseme umeenda wapi do not say where you are going
sitaki noma na mathako, I don ’t want trouble with your mom
sema umeenda laibo say you are going to the library
sitaki noma na budako I don ’t want trouble with your dad

In this song the artist encourages his girlfiiend to lie to her parents that she has gone to
the library so that she can come to his place. These young women in Kariobangi knew
how to appropriate the song’s message in their discourse. Apparently, the song’s message
urging youth to lie to their parents when they want to go out for fun seem popular with
the youth. Quoting of popular lines from a local musician came up once more during my
interview with the Ngara respondents. In the conversation below, Josh narrates about his
encounter with a beautiful girl at a bus stop. The bus had been delayed and got chatting
while waiting. Finally they decided to walk to their destination, and after a while, Josh
had nothing else to tell her,

Josh: stori zinakatika. Nakaa mimi ninamwambia, ‘unacheki hiyo ndae? Ni poa’

(laughs) ananishow, ‘ee ni poa’. Tuna-kamkam namuuliza, ‘unapenda reggae?’
PG: mh
Josh: akanishow, ‘ndiyo’. Si nikajua sa nitakuwa nazo?

218

PG: mh?

Josh: Nikamwambia, ‘ee nina collection nyumbani sa niaje si uzikamie?’
PG: mh

Josh: akanishtua, duuuuh, yeye ana mingi kuniliko.

(Jim laughs)

Josh: anaweza nifunza reggae kwanza.

Translation:

Josh: the stories ended. I paused, and I asked her, ‘do you see that car? Its cool’
(laughs) she told me, ‘yes, it is cool’. We proceeded I asked her, ‘do you

like Reggae?’

PG: mm-hm

Josh: she told me, ‘yes’, then I knew that I have them (reggae music) ok?

PG: mm-hm?

Josh: I told her yeah, ‘I have a collection at home, would you like to come for
them?’

PG: mm-hm

Josh: she surprised me duuuh, she has more than I do

(Jim laughs)

Josh: she could even teach me reggae.

Throughout the narration, Josh was smirking while his buddies constantly exploded into
fits of laughter. Unknown to me then, Josh was structuring his story around the rap group
‘Necessary Noize’s’ popular hit Kenyan Boy Kenyan Girl, but he was making most of his
story up while maintaining the plot and some words used in the song. It was not until I
had heard the song that I was able to draw the parallel. In Josh’ narration, waiting for a
bus, mutual attraction by both parties, the expensive looking girl, the love of reggae,
asking some silly questions and taking a stroll are all borrowed from Necessary Noize’s
song (though in the song the girl is the seducer). The words unapenda reggae, and nina
collection nyumbani above are actual lines in the song. The joke aside, we can see how
the music has come to have a huge impact on the lives of young people in Nairobi. What
the popular local artists say quickly filters into the mainstream discourse. As culture

brokers who feel that they have a role to play in the young people’s lives (Samper 2002,

219

 

2004), musicians present the world for the youth’s consumption. When the youth employ
some words fi'om popular music tracks in their discourse, it means that they have
accepted the values that the musicians stand for. Together they participate in the identity

of hipness, and the modern lifestyles and consumption the musician propagates.

4.6.2. The media

“My radio doesn’t shika anything else” screams the flashing message on 98.4 Capital FM
website. Capital FM is one of the many commercial FM stations in Kenya. The language
mixing in this flashing message reflects the general pattern in the radio’s language use.
Indeed many of the FM stations have taken to using Sheng words as a way or responding
to their audience’ needs. This situation also obtains in print media. Readers of Kenyan
newspapers who are not conversant with Sheng are at a loss when they encounter Sheng
words that have been thrown in by writers in their columns such as the ones below;

1. As I was struggling to fish it out, half of the shirt chomokad. .. Pulse, a Friday
magazine from the Eastafrican Standard

2. Be yourself and avoid ma-fiction mob... interview with a Flex, a local hip-hop
artist appearing on Buzz, a Sunday entertainment magazine from Nation Media.

3. Definitely! T unawakwachu no lyrics zetu. . .Flex again

4. For some reasons they still think they are unbwogable. . .Mutahi Ngunyi, a
columnist with Sunday Nation, a leading weekly from Nation Media.
These are just a few examples of Sheng lexical items one encounters daily in Kenya’s
leading newspapers. While media houses have no official policy that promotes the use of
Sheng, we can say the writers’ indulgences are condoned because they are responding to
the language of their target market. The East African Standard’s Friday magazine Pulse is

full of such Sheng words. In the readers’ column ‘The so called life’, readers narrate their

220

most embarrassing moments. Contributors normally lace their pieces with Sheng word to
the extent that a piece that lacks at least a Sheng word is an exception. Smitta Smitten’s
weekly column in the same magazine is usually full of Sheng words, which would make
it inaccessible to a person with no knowledge of Sheng. As an illustration, I have adapted
one of the articles by Smitta Smitten’s with some sections removed. The highlighted

Shng words (in bold) are glossed at the end of the article.

 

 

 

Thursday March 24, 2005
Pulse

Underground chit chat
with Smitta Smitten

Last Friday, I went underground to Java Coffee House's basement. Down in the
underground, I ran into Bamboo ll — the new, redeemed version. Dude still
operates a black cap and shades that look like he borrowed them from the estate of
the late Ray Charles. The shades I can understand - Bamboo II has seen the light.
and it's blinding. But the cap? That's for shady characters like Smitten on the prowl.
I recommend a white starch shirt and black cotton tie for Bamboo Il, if he's saved.
Plus he was with this hot,jungu chickie.

Later, I rolled into Carni for the Chaguo La Teeniez Nominees' bash with Kitten...
If you grew up in the 19905, New Jack Swing makes sense. If you're below 25,
please forget it. Just read Pulse and tune in to rock ziski (music) with Italia Masiero.

Speaking of which, Nazizi and her bro Feroze seem to have dropped off the face of
the earth. Hey, niaje brathe, niaje siste? When the shit hits the fan, the answer
isn't to pull a ‘Muite' (remember Pattni's ShZOM accusation?) and go underground.
What you do is pull a ‘Debra Sanapei', i.e. news on Friday is that you owe everyone
in town chinky (money) - on Saturday, you show up, in style, at the latest ‘Trendy
Bar' launch in town.

...Prezzo rolled in, in all his glory - half of which was Sheila Mwanyigz looking
resplendent in red. That nigga's got style. Only wish he’d swing me real bling (my
‘gold' chain is already turning a suspicious coppery colour). Bamboo II and hisjungu
spent their time sheltered in the shadows...

 

221

 

 

 

where the Pulse crew was waiting to do our EATV-spot for ‘Word Up' on Satoz at
7.30pm (catch the show kesho, for once). In the background, I almost swear I could
hear Maina Kageni's humongous voice.

...Encouraged, Mdomo Baggy decided to take me on in ‘Mchongwano'. Very
19805. I told Baggy diesel like he grew up in such a rough-ass neighbourhood that
someone even mugged the rainbow.

Keroro shouldn't be sold to tois...

 

 

Some of the Sheng words used in the article are;

redeemed version
jungu chickie
cami

chaguo 1a teeniez
niaje brathe

niaje siste

prezzo
mwanyigz

nigga

bling

jungu

satoz

kesho

mdomo Baggy
mchongwano

keroro
tois

Fl
born again I
white girl '1
carnivore L

teenager’s choice

what’s up, brother

what’s up, sister

stage name of a local hip-hop artist, also Sheng name for president
Mwanyiga, name of Prezzo’s girlfriend

borrowing from African American English; Nigga is used as an
affectionate in-group term in this context

another African American borrowing, showy jewelry

white woman, from Swahili Mzungu

Saturdays

tomorrow, a Swahili borrowing

baggy mouth, probably a Sheng nickname

verbal duels similar to Afiican American playing of dozens (see
Section on Mchongoano in this Chapter

alcohol

children, underage

The reporters’ styling of their language to approximate to that of their audience is the

motivation behind Bell’s audience design model (1984, 1992:110). In his analysis of

language style of the UK’s newspapers, Bell found out that the rate of determiner

deletion in a media outlet varied according to the social stratus of that paper’s readers.

The mainstream papers such as The Times, The Independent, and Financial Times, were

popular with the upper class and had fewer cases of determiner deletion, while the tabloid

papers such as the Sun, and the Daily Mirror, which were popular with the lower class,

222

had a very high rate of determiner deletion. Presence of determiners was associated with
the standard features of Received Pronunciation (RP), while lack of determiners was
associated with non-standard features attributed to the popular culture that looked up to
the United States.

Coming back to Smitta Smitten’s style, his use of Sheng’s lexical items is a
deliberate way of styling his speech to reflect the language styles of his readers. Majority
of his readers are the young people whose life has been greatly influenced by the norms
of popular culture. Smitten’s column, and indeed the whole pulse magazine, borrows
heavily from the English tabloid press in terms of themes, rumors and language. Smitta
Smitten is a typical paparazzo who stalks artists, updating his readers on the most recent
rumor. Somebody who does not keep abreast with the Kenya’s entertainment scene may
be unable to understand his column due to use of Sheng lexical items. In the above story
for instance, one needs to know that Bamboo claims to be a born again Christian and has
a white girlfriend, that Mwanyiga is Prezzo’s girlfriend, that Prezzo is a very rich artist
and that Nazizi had an embarrassing moment on account of her brother. His extravagant
use of Sheng words means that he has conceptualized an idea of an in-group following
that constitutes his fan base with whom he communicates. This overly used of English
laced with Sheng items here and there seems to be a good example of the Engsh version
discussed in Chapter 3. Although we may assume that there is a limit to which his editors
might have allowed him to use his Sheng, it should not be lost that his object, and his
audience largely comprise of middle class youth, who form the majority of the artists, as

well as the audience that attend their shows.

223

Smitta Smitten’s column in the Pulse magazine compares well with KJ’s cartoon
head on corrishon in Buzz — an entertainment magazine from Sunday Nation, both in
terms of themes and the use of language. Like Pulse, Buzz reports what is happening in
the local entertainment scene. The audience is the youth, who as we have said elsewhere,
are the most avid users of Sheng. KJ actually creates a distinction between the Mavijanaa
‘youth’ and the Wabuyu ‘the adults’ and he puts himself into the group of Mavijanaa.
When he is giving an opinion from the youth perspective, he writes sisi Mavijanaa ‘we
young people’. These two illustrations show that both Pulse and Buzz target a certain
audience, and they adjust their language to conform to the language styles of their
readers. They draw their vocabulary from the Sheng that is in vogue, but they sometimes
coin their own words. As a result, they become vehicles through which some of these

Sheng features filter into the wider populace.

4.6.3. The internet

The intemet has become a very dynamic discourse site where identity has been played
out. Stephanie Nilsson (http://www.eng.umu.se/stephanie/web/LanguageBlggspdt) has
carried out a recent study on ‘the function of language to facilitate and maintain social
networks in research weblogs.’ From this work we can see that the form of interaction by
the virtual community can be studied using the same models that we use in daily
interactions. By tracing the intemet links, a clear diagram of the network clusters to
which bloggers belong can be traced. It is with this in mind that I look at the relevance of

the intemet in Sheng’s variation and identity construction.

224

Many kinds of discourse take place in the intemet, such as general discussions
and even michongoano. The intemet’s ability to connect Kenyans at home and those in
the diaspora is the best arena where Sheng, in Samper (2002) words, ‘localizes the global
and globalizes the local’. The intemet can be looked at in terms of discourse output and
discourse input as we saw in our discussion of michongoano earlier on. The intemet users
can be viewed as virtual members of a speech community with their own kind of identity.
Those who constantly log into the intemet and actively participate in Sheng threads —
those threads that either talk about Sheng, or those that use Sheng as their medium, can
be regarded as members of a certain interaction group. Furthermore, the intemet users
can, and do, form their small groups in chat rooms where discourse is conducted in
Sheng. These little discussion groups become so tightly bound as real communities of
people who know each other well. For a discussion group that has been interacting for a
long while, they cultivate their own interaction norms that only people who interact
constantly are aware of. A good example is in the orthographic aspect of words such as it
‘you’ and r for ‘are,’ as used in this joke: ati u r so fat mpaka u r taller while lying down
‘that you are so fat that you are taller lying down’. Such cases show that norms governing
what can be initialized and what cannot, can be one way that participants in virtual
communities construct and maintain their identity, structurally.

Just as one can move from one group to another in real life, intemet users can
move from one group to another in various chartroom. Moreover, the intemet provides
resources for members to reject undesirable members by blocking them from their buddy
lists, or generally ignoring their posts, which eventually isolate intruders from the regular

loggers. This becomes another manifestation of the use of Sheng in the construction of

225

identity in cyberspace among different groups of virtual communities. We can envision
the existence of ‘virtual bazes’ when a group of intemet chatters communicate in private
chat rooms regularly. This can be an interesting and challenging line of research but is
beyond the scope of this study. I will now turn to the psychological aspects that tie
people’s attitudes to their identities. This relationship will be discussed in the next

Section as the final discussion of my dissertation.

4.7. Attitudes and value judgments

A group’s identity cannot be divorced from the attitudes that they hold towards others
and their practices. To complete this Chapter, I will discuss subjective judgments of the
Sheng speakers from different bazes towards other Sheng varieties and relate these
judgments to identity. Although I have already rejected the estate as a viable unit for the
analysis of linguistic practices, it does not mean that the speakers themselves do not make
use of this abstraction. My use of the term ‘estate’ should therefore not be regarded as

conflicting with baze.

4.7.1. Respondent’s subjective judgments: relative comparisons and identity

Earlier on, I rejected the perspective of looking at identity as frozen in static categories in
favor of the perspective that views identities as fluid and in a constant process of
modification. In this section, I adopt the perspective propagated in social identity theory’
(e.g. Turner 1982, Tajfel and Turner 1986) that identities become apparent through
juxtaposition of contrasts. Individuals categorize groups in terms of opposites, and they
use those categories to identify and define themselves. Once they perceive themselves as

members of groups, these individuals see everyone in their group as similar. In their

226

effort to cultivate a ‘positive self-concept’ they compare themselves with other social
groups in order to enhance individual and collective self esteem (Niens and Cairns
(2003). This arouses a sense of intergroup competition where the mere perception of
belonging to two distinct groups becomes a sufficient reason for discrimination against
the out-group.

More insight will be drawn from relative deprivation theory (RDT) that proposes
that individuals can feel personally deprived when comparing themselves with better off
individuals (Smith and Leach 200412). Similarly, people feel group deprived when
comparing themselves as group members with members of a better off outside groups. If
people compare themselves with those they consider inferior, their positive attributes are
confirmed, which raises their self worth (Tajfel and Turner 1986, Smith and Leach 2004:
2). With respect to Sheng, we expect that when speakers in a given baze compare
themselves with other speakers that they consider to be like them, they should identify
with them by rating them positively. On the other hand when they compare themselves
with those they regard as inferior to them, they should magnify the contrast in order to
consolidate their distinctiveness.

To investigate how respondents identified themselves with Sheng speakers from
different estates, I asked the subjects from the 12 bazes56 to rate how Sheng speakers
from Jericho, Kawangware, Kibera, Makongeni, and Westlands estates spoke compared

to themselves. I gave them the following subjective response options (see appendix II

 

5" Shauri Moyo, Buruburu, and Makadara, are part of the original Eastlands, though Sinai
is also regarded as Eastlands. Kariobangi and Eastleigh can be considered as North, while
Ngara, and Kabete are in the Westlands. South C and Kibera are in the South. Monte
Carlos is in the City center. In my generalization of Westlands and Eastlands, Kibera,
South C, Eastleigh and Kariobangi are considered as Eastlands.

227

Q10): ‘exactly like me’ [ex], ‘a little like me’[li], ‘very different from me’[di], and ‘I do
not understand them’[do]. The data below are by no means conclusive and might not
reflect the view of all the members of the ban”. These findings are summarized in

Table 8 below;

Kabete n=4 Kibera n=5
ex li di ' ' ex li di do
1 l 2 l 3
k’ 4 0 0 ~ I 0
0

kibcra

9

3 l 0 l

W

 

Table 8. Different bazes’ comparison of their Sheng with that of other estates N=54

From table (8), I assign index [F] indicating full identity if at least half of the raters rated

the estates as ‘exactly like me’ and [P] indicating partial identity if at least half of the

 

57 While the size of the sample might raise questions about statistic validity, careful
interpretation can overcome this handicap by regarding the speakers as exemplar
members of a baze, but not as representative sample. Bazes have more that four members,
but I only chose a small sample per baze because it was going to be hard to carry out a
coherent interview with more that five people.

228

raters rated as ‘little like me’. This makes it possible to summarize how different bazes

identify themselves with regard to the kind of Sheng spoken in other estates.

 

Baze Nature Qf identity — full [F I oLpartialfP]

Shauri Moyo: Jericho [P]

Monte Carlos: Jericho, Kibera [F], Kawangware and Makongeni [P]
Buruburu: Jericho, Makongeni [F], Kawangware [P]

Eastleigh: Jericho, Makongeni [F], Kawangware, Kibera [P]
Kabete: Kawangware, Westlands [F], Kibera, Makongeni [P]
Sinai: Jericho, Makongeni [F]

Kibera: Kawangware, Langata, Makongeni [F]

Kariobangi males: Kibera [F], Jericho, Makongeni [P]

South C: Kibera, Makongeni [F]

Makadara: Jericho, Kawangware, Kibera, Makongeni, Westlands [F]
Ngara: Jericho [F], Kawangware, Kibera, Makongeni [P]

Kariobangi females: Jericho [F] Kawangware, Makongeni [P]

From these data, we can see that the majority identifies with the variety of Sheng spoken
in Jericho, followed by that of Makongeni, but very few identify with the Westlands
variety. Indeed only the members of the Kabete baze thought that they spoke a Sheng
similar to that of Westlands speakers. 1 explain this by invoking the political connection
that binds Kabete and Westlands. Since Kabete is considered part of the Westlands
constituency, the socioeconomic boundary that sets them apart is deferred and the
Westlands political identity is made more salient. The Makadara baze identifies their
Sheng with all the other varieties spoken in Nairobi. Recall that although Joe in his
narration claimed that some words were spoken in different estates, he nevertheless used
them in his discourse (see section on performing multiple identities). Interestingly, Shauri
Moyo raters did not think there were people who spoke Sheng that was exactly like
theirs. During the interview, the Shauri Moyo raters claimed their Sheng was the best in

the ‘whole of the world’. Recall too, that they claimed that Kijey Sheng is sweet, but only

229

 

to listen to. This is an indication that they are bound by deep in-group loyalty, that even
their innovation is geared toward making them more distinct as a group.

We note that only the Kibera baze rated Langata (a neighboring middle income
estate). Since Kibera was one of the estates being rated, I reasoned that they could not
possibly compare themselves with themselves. Like the case of Kabete and Westlands,
Kibera is in Langata constituency; therefore they share a form of political identity with
their Langata middle class neighbors.

A better generalization comes from Table 9 below. The middle column combines
both full and partial identity as a pointer to people’s orientation towards a variety of
Sheng. From here, we can see that the most favored Sheng variety is that from
Makongeni, followed by that of Jericho, then Kawangware and finally Kibera.
Consequently, we can say that these are the varieties they identify with. On the other
hand, less that half of the respondents identified with the Westlands variety, with the

majority rating it as ‘different from me’ or ‘don’t understand them.’

 

 

 

  

 

 

 

 

 

 

 

exactly a little , different don’t
Jericho 25 18 ' = '7' 10 1
Kawangware 19 20 13 2
Kibera 19 13 14 4
Makorgeni 21 23 2 2
Westlands 10 10 24 10

 

 

 

 

 

 

Table 9. Respondents comparisons of their Sheng with that of other estates (N=54)

These data in Table 9 shows that people still identify with the Sheng of the Eastlands. It
is also possible that since majority of the raters were from the Eastlands, they were rating
their local varieties more favorably. Furthermore, we can conclude that the general

stereotype of Sheng being the language of the ghettoes persists since most of these bazes

230

are located in the poor residential areas; therefore, there exists a possibility that they
compare themselves favorably with their counterparts in other deprived estates with
whom they share an identity.

Some comparisons yield unexpected results. For instance the South C raters, who
could be expected to identify with the Westlands variety on account of their shared socio-
economic status, do not do so. Likewise, it is not always true to conclude that the ghetto
people identify themselves with other ghetto people just because they rate other ghetto
people favorably. They also have their own comparisons, and sometimes, they rate other
ghetto people harshly. The Kariobangi females, for instance, regarded Korogocho, their
neighbors as speaking the worst Sheng. In the conversation below, the Kariobangi

females were discussing about a girl who was murdered by another girl for ‘stealing’ her

boyfriend.

Sera: saa. . .na huyo manzi alimarikanga alikuwa amemaliza chuo? alikuwa
anasomea wapi hiyo shule?

Koki: our lady,

Sera: our lady,

Koki: ali-drop ka ako form. . .form three

Shiro: eee, ni dropper, ee obvious mtu amesomea karibu na Koch una-expect
awe. . .?

Koki: yeah, hiyo ni kawa..

Translation:

Sera: Now... and that girl who died, had she finished school? She was going to
which school?

Koki: Our lady,

Sera: Our lady,

Koki: She dropped out when she was in form three

Shiro: yes, she is a dropper, yeah, it’s obvious, what do you expect a person who
studies close to Korogocho to be?

Koki: yeah, that is norm...

231

Kariobangi females hold very negative opinions of their Korogocho neighbors. In the
sense of relative deprivation theory discussed earlier, it seems that when Kariobangi
female raters compare themselves with those they consider lower than them, they erect
boundaries that portray them as distinct. I will revisit this issue shortly. Meanwhile, I first
compare these data with the value judgments on how the same bazes regarded the way
speakers of different estates spoke Sheng (see Appendix 11 Q 11). Respondents from the
12 bazes were asked to give qualitative judgments of how people from Mathare,
Kileleshwa, Kangemi, Muthurwa and Dandora spoke Sheng. They were asked to rate
speakers according to the following qualitative terms: ‘very well’[w], ‘not so bad’[nb],

‘not well’[n], and ‘bad’[b]. A summary of the general results is in Table 10 below.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Shaurimoyo n=4 Montecarlos n=4 Buruburu n=6 Eastleigh n=4
w nb 11 b w nb n b w nb n b w nb n b
mathare 1 2 0 1 3 1 0 0 5 0 l 0 l 3 0 0
K’shwa l 0 2 l 1 1 1 l 1 l 3 l 0 l 3 0
kangemi 0 4 0 0 2 1 l 0 1 4 1 O l 1 1 0
muthuru l 3 0 0 1 1 2 0 4 1 l 0 0 l 2 0
dandora 3 0 l 0 4 0 0 0 4 l 0 l 2 0 0 l
Kabete n=4 Sinai n=3 Kibera n=5 K’bangi (m) n=4
w nb n b w nb n b w nb 11 b w nb n b
mathare 1 l l l 3 O 0 0 4 1 0 0 0 2 l 1
K’shwa 0 3 0 1 0 0 . 0 3 4 0 l 0 0 2 2 0
kangemi 3 0 1 0 1 2 0 0 3 0 1 l 1 0 l 2
muthuru 3 0 1 0 3 0 0 0 4 l 0 0 l 0 0 3
dandora 1 2 1 0 3 0 0 0 3 0 2 0 l 2 l 0
South c n=8 Makadara n=4 Ngara n=3 K’bangi (1) n=5
w nb n b w nb 11 b w nb 11 b w nb 11 b
mathare 3 5 0 0 1 l 1 l 2 l 0 0 5 0 0 0
K’shwa 0 2 5 1 0 1 2 1 0 0 3 0 0 3 1 1
kangemi l 4 l 2 l 0 l 2 2 l 0 0 l 4 0 0
muthuru 2 3 l 2 2 0 1 l 1 1 0 l l l 3 0
dandora 8 0 0 0 2 0 2 0 3 0 0 0 5 0 0 0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 10. Respondents judgment of the variety of Sheng in different estates, N=54

232

In Table 10, the relevant data are regarded as those qualitative terms chosen by more than
half of the respondents in a baze. A further summary of the above table according to

different bazes is given below;

 

Baze qualitative judgment

Shaurimoyo Dandora [w], Mathare, Muthurwa Kangemi, [nb]
MonteCarlos: Dandora, Mathare, Kangemi, [w]

Buruburu Dandora, Mathare, Muthurwa [w] Kangemi [11b]
Eastleigh Dandora [w] Mathare [nb]

Kabete Muthurwa, Kangemi, [w] Dandora, Kileleshwa, [nb]
Sinai Dandora, Mathare, Muthurwa [w] Kangemi [nb]
Kibera Dandora, Mathare, Muthurwa, Kangenri, Kileleshwa[w]
Kariobangi (m) Dandora, Mathare, Kileleshwa, [nb]

South C Dandora [w], Mathare, Kangemi [nb]

Makadara Dandora, Muthurwa, [w]

Ngara Dandora, Mathare, Muthurwa [w]

Kariobangi Dandora, Mathare [w], Kangemi, Kileleshwa [nb]

This analysis shows that the most favored variety of Sheng is that spoken in Dandora.
Out of a possible 54 times, Dandora is rated 37 times as ‘very well’ and 5 times as ‘not
bad’. It is followed by Mathare 29 ‘very well,’ 17 ‘not bad,’ then Muthurwa 23 ‘very
well,’ 12 ‘not bad,’ and finally Kangemi 17 ‘very well,’ 12 ‘not bad.’ In contrast,
Kileleshwa speakers are rated poorly, only 7 ‘very well’ and 14 ‘not bad.’ This shows
that just like in table (9) the Sheng that the majority of the respondents favor is that from
the poorer estates as compared to that from the affluent estates. This can be explained in
terms of symbolic capital where we can argue that the need for cultural capital is greater
in poorer areas due to the fact that the economic capital is out of their reach.

Surprisingly, five out of eight South C respondents rated Kileleshwa speakers as
‘not well’. Bearing in mind that Kileleshwa and South C share a lot in common in terms

of socioeconomic status, this unexpected rating can partly be attributed to the dynamics

233

of social affiliations in the two estates. While houses in South C are close together, those
at Kileleshwa are isolated (similar to the Riverside house shown in Figure 2). We don’t
have bazes for Mabeshte to hang out in Kileleshwa, but we do have them in South C. If
the bazes are small clusters where a great deal of interaction takes place in Sheng, then
the perceived differences can be explained within network theory (Milroy 1980, Milroy
and Milroy 1992). Lack of such clusters in Kileleshwa, means that the networks of the
residents are very weak. This inhibits the enforcement of vernacular norms, which makes
the residents resort to basic codeswitching, or to insert isolated Sheng words here and
there. In contrast, due to their participation in bazes, the South C residents have very
strong networks, almost similar to their counterparts in the informal settlements like
Kibera. Therefore, the poor rating of Kileleshwa by the south C respondents converges
with that of the raters from the poor residential areas in this respect.

To explain Kibera’s positive rating of the Kileleshwa variety, we must assume
that Sheng does not evoke very deep in-group emotions among the Mabeshte in the
Kibera baze. We must also bear in mind that their Sheng has a lot of Swahili influence.
Most of their coinage, for instance, involves Swahili lexical items that have undergone
semantic shift. Assigning Samper’s (2003) label of basic Sheng to the Kibera variety will
equate it with the Kileleshwa variety except for the fact that while the Kibera variety uses
Swahili, Kileleshwa uses English.

Relative comparisons of the bazes with relation to estates reveal the bazes that can
be considered to be occupying the liminal space between the two marginal extremes.
These are the most interesting settings in which to examine identity construction since

they keep on moving to and fro, as contexts dictate. Although the Kariobangi raters

234

identify with Korogocho when it suits them, they are not blind to the fact that Korogocho
people are ‘fellow sufferers’. However, comparing themselves with their Korogocho
neighbors enhances their self esteem. At the same time, when they compare themselves
with the upper class people who live in ‘flats’, they join their Korogocho comrades in
claiming to suffer a common identity. We can look at the South C raters in that same
light. When they compare themselves with the posh Westlands, Lavington, Kileleshwa
and Runda estates, they perceive a socioeconomic difference, which arouses inter-group
feelings. On the other hand they recognize that they are not like the people from the low
class Eastlands estates or the informal settlements such as Mathare. Depending on who
they are comparing themselves with, they can either project their identity as the rich kids
— not from the ghettoes, or as a deprived group — jobless residents of not so good

estates.

4.7.2. The question of proximity

Varieties spoken near a certain baze are likely to be perceived more favorably because
they would be regarded as similar or close to the variety used in that baze. There is
evidence of this in the Kabete subjects’ positive rating of Kawangware and Kangerrri,
their closest neighbors. Similarly, Makadara raters unanimously regarded their close
neighbors, the Jericho and Makongeni estates, as ‘exactly like them’. The same can be
said of Buruburu and Sinai’s bazes’ positive rating of the Muthurwa variety.
Nevertheless, we must be aware that proximity itself does not always result in positive
ratings. A good illustration is the different ratings for Mathare by the two genders from

the Kariobangi baze. We have also seen how Kariobangi females rated their neighbors in

235

-‘L‘j

Korogocho poorly. Similarly, Shauri Moyo subjects considered the Makongeni variety to
be ‘childish,’ using oldskool words that they had already abandoned. This might be a case
of over-exaggeration, considering that Makongeni is a neighboring city council estate.
Surely, the difference should not be so huge.

While proximity can be invoked to explain Kariobangi females’ positive rating of
Mathare speakers, the same cannot be said for the case of Kariobangi males. Similarly,
proximity cannot explain the positive rating of Dandora speakers or the mixed rating of
Muthurwa speakers by the South C respondents. In the light of this inconsistency, I will
downplay the inconsistency of the Kibera raters’ favorable rating of the Kileleshwa
variety in favor of a more universal generalization. For these raters the most important

distinction is the one between the ghetto and Ubabini speakers.

4.8. Summary

I began the Chapter by invoking the Saussurean conception of words as linguistics signs
with a dual semiotics of the signifier and the signified. I discussed how ambiguity serves
to build identity among speakers of different varieties. I also explained how certain Sheng
words and other practices vary according to age. I used the concept oldskool and argued
that generational variation comes about not only because of actual age, but also as a result
of some people failure to update themselves with new innovations. Oldskool items
sometimes surface in daily interactions, even in highly innovative bazes as a result of the
fluid nature of the bazes. It is this fluidity that is responsible for different shades of
identity of the baze members. I also discussed the lexical construction of the female

gender and how women contest derogatory labels by using labels that portray a positive

236

image. Different categories of people were shown to choose positive terms to refer to
themselves. Discussing how symbolic labels that are used to attenuate the ‘us’ vs. ‘them’
boundaries, I demonstrated how terms such as Wanati, Wababi, Mabeshte and fala vary
according to different bazes and the position one occupies within the group in question.
When discussing the performance and negotiation of multiple identities through
non-conscious and conscious discourse practices, I argued that some identities are
activated non-consciously in spontaneous discourse because they have become ingrained
into the speakers’ habitus. I also demonstrated how Sheng speakers accommodate to the
Sheng varieties of other speakers as per the changing contexts. On vernacular discourse, I
analyzed the case of two prostitutes and illustrated how their daily reality as whores, sex
workers, heterosexuals and mainstream females are performed through discourse. Still
focusing on the vernacular, I examined michongoano — a face-threatening verbal contest
discourse, as an in-group enterprise. I discussed how it can be used as a diagnostic for
inter-group perception and how local, national and international identities are performed
in michongoano. Other non-linguistic ways of constructing identity that complete the
identity project were also discussed. Apart from the bazes, other discourse sites where
identity is constructed and performed such as mass media, intemet and popular culture
were also examined. Concluding this Chapter was discussion of language attitudes and

their relevance to identity through comparison and judgment of other Sheng varieties.

237

”J

Chapter 5
Concluding remarks

Conclusions

Cognitive social psychology conceptualizes identities as cognitive schemas, or internally
stored information and meaning that serve as the frameworks for interpreting experience
(Stryker and Burke 2000:286). This study equated these “cognitive schemas” with the
background knowledge that cues the shared reality that help speakers in interpreting
experience. This was taken within the context of variation in Sheng’s lexicon acting in
tandem with other discursive practices to represent the speakers’ multifaceted identities.
In this pursuit I did not limit myself to the Saussurean sense of words as concrete entities
with arbitrary relationship with their signified; instead I also regarded them as cognitive
symbols with associated meanings. Basing my argument on the logic that meaning
making is a shared enterprise that varies according to different speakers and contexts, I
set to examine the picture of variation and its relevance to the identity discourse.

The overriding issue addressed in this dissertation is how people in small groups
in Nairobi make use of vocabulary resources of Sheng in projecting their different
identities. The various discussions point to the conclusion that due to the ephemeral
nature of some of Sheng’s vocabulary, as well as the high rate of innovation, the identity
of Sheng speakers cannot be said to be complete. It is an ever-changing process that
involves negotiation, renegotiation, modification, construction and deconstruction. Even
when we look at other ways that complete the identity picture, we notice that they are

also in a state of change. Today’s music will be tomorrow’s old music, and while some

238

 

people will remain in both physical and psychological ‘relic areas’58 or what the Sheng
speakers call oldskool, others will move with the tide and adopt the new innovations.
Lexical variation thus becomes the symbol of the Sheng speakers’ different identities as
long as such speakers shift their vocabulary depending on the prevailing situations.

In this study, a case of Sheng as an order of discourse, in relation to other
languages used in Nairobi was made. Various bazes were equated to institution which
justified relating their variants with discourse types. This way of perceiving Sheng was
shown to have advantages over various classifications such as pidgins, Creoles, jargons
and slang as defined in the literatures because of lack of consistency and the confusion
that arises when Sheng data is introduced to such definitions.

Variation in Sheng was regarded as an overt manifestation of different cognitive
schemas that different categories of Sheng speakers exploit to interpret their different
experiences. I demonstrated how the words that speakers use are created to express a
certain reality and are laden with meanings that require the background knowledge for
interpretation. This underscored the importance of paying attention to the way they are
used in discourse. Speakers who share background knowledge are able to disambiguate
ambiguous words in discourse. In addition, the ease with which Sheng synonyms can be
substituted in Sheng discourse promotes intergroup interaction among Sheng speakers
thus continually altering the speaker’s previous identities.

To emphasize the importance of discourse as a cue to shared meaning, vernacular

speech of two female respondents in Sinai was analyzed where it was demonstrated that

 

58 In sociolinguistics these are the areas or places that are left out or change last in
linguistic innovation (Downes 1998: 143): i.e., where old forms persist whereas
surrounding areas have adopted new innovations.

239

their identity as whores or sex workers was reflected in their use of vulgar language that
transcends social inhibitions. The same was seen to apply on the discussion on verbal
contest genre known as michongoano which was seen to rely so much in the shared
background knowledge among the speakers, who know the issues they can talk about and
those that are out of bounds.

Turning to the highlights of this study, I will briefly show how I addressed the
hypotheses stated in Chapter 1. Hypothesis 1 that sates that “different identities are
reflected by the use of different names for the same referent’ was addressed by presenting
data on the productive lexical innovation in Sheng that breeds high rate of synonymy,
defines as a semantic relation where various lexical items bear the same meaning. I
argued that synonymy occurred because speakers prefer to use different lexical items in
order to draw group distinctions. It was shown that some terms have a high frequency of
use among the respondent as opposed to others. I argued that the lexical terms that are
used widely across the board are not very useful in the construction of identity. Such
terms are shunned because they are known by almost everybody and would therefore burr
the group boundaries leading to loss of identity. It is for this reason that words that are
not widely used were regarded as the most important in the study of group’s identity.

Standing in stark contrast with synonymy was ambiguity, a semantic relation
where a lexical item might have different meaning. Ambiguity addressed the Hypothesis
2 that states that ‘different meanings for the same referent are a reflection of different
identities’. Speakers from different bazes were shown to construct their identity around
their shared meaning which deviated with the meanings of other groups. The relevance of

ambiguity in identity was equated with Geneva Simitherman's (1977) example of the

240

 

different level of meaning of some words used by the African American. A word could
have the mainstream meaning for both Whites and Blacks, but its semantics could change
to the exact opposite when used among the African American as a racial category.
Similarly, it was shown how groups of Sheng speakers could build their identity by
changing the semantics of some widely known words as was the case with mubatari
whose mainstream use was needy, but meant ‘police’ among ‘manamba’, and ‘manamba’
among the Sinai females.

Although my discussion on the politics of identity demonstrated the problem with
adopting either the static or dynamic perspectives as exemplified by the critique by
Brubaker and Cooper (2000), this study adopted the dynamic View where the argument
that identities are fluid and constantly changing was advanced. The logic of this was
essential in discussing my Hypothesis 3, which stated that ‘Sheng speakers can learn and
use the words that are identified with groups other that their own’. Capitalizing on the
fact that identities are coded in lexical items in Hypotheses 1 and 2, I argued that it is
through the ability to learn and use the lexical items, as well as their various
significations that Sheng speakers from different bazes shift their identities. Ability to
learn the Sheng from different bazes was attributed to the respondents’ claim in a
statement ‘Sheng from all ghettoes is the same’. This was explained in terms of the
almost similar reality experienced by all ghetto residents. In addition, a claim was made
that variation in Sheng, and indeed Sheng itself was overwhelmingly a lexical issue.
Hence it was easier to learn vocabulary when the speakers are not burdened by learning a

new grammar.

241

 

The motivation behind shifiing identities was attributed to the change in social
situations. Since the majority of the respondents were from the ghetto, identification with
the varieties of other ghettoes was driven by their shared reality, while identification with
the varieties spoken in affluent neighborhoods was driven by the status symbolism that
the ghetto people aspired for.

The discussion on conscious and unconscious shifting of identities addressed
Hypothesis 4 that ‘Speakers shift identities both consciously and unconsciously’.
Unconscious shifting of identities was attributed to speakers’ habitus which have become
ingrained into their linguistic norms. At the same time conscious shifting of identities was
attributed to cognitive processes where the speakers took initiatives to shift their
linguistic norms to suit those of desired others. It is possible to conclude that the
speakers’ habitus is built through these conscious discursive linguistic practices.

Hypothesis 5 that says that ‘shifting of identities is greater in densely populated
residential areas as opposed to sparsely populated residential areas’ was addressed by
relating residential dynamics to the network patterns between the poor and affluent
neighborhoods. The strong networks in the ghettoes promoted heavy interaction in Sheng
which bled many small units. Since members of these small units interacts with members
of other small units within the ghetto then Sheng lexical material are able to spread. The
same was not possible in affluent neighborhoods because of lack of small units of
interaction. Consequently, there was no enforcement of norms that could make a group
distinctive as per the claims of network theory (Milroy 1980). However, due to
intervention of popular culture and mass media in Sheng dissemination, some terms from

the ghetto have trickled to the affluent neighborhoods. Although their adoption, to an

242

extent modifies the identity of the ubabini residents, they do not lead to major shifting of
identity because widespread terms are useless in identity construction.

My final hypothesis was that “Sheng speakers identify with Sheng varieties that
are similar or close to their own”. This issue was addressed by examining attitudes of the
respondents towards other Sheng varieties. It was shown that the ghetto raters, who
comprised the majority of my interviewees, rated the other ghetto varieties better. This
was attributed to the fact that these varieties were very similar to their own varieties.
Divergent cases such as that between Kariobangi and Korogocho, or Westlands and
South C were attributed to relative comparison, where speakers used factors other than
linguistics to draw intergroup boundaries.

The role of individuals as members of social categories must be underscored in
the identity project. This is because the negotiation of identity takes place at a personal
level. Innovations begin with individuals and then spread within the group before
spreading outside the group. It needs to be taken into account that even among the groups
in which innovations originate; they might not have a uniform effect. There are for
instance, members who simultaneously mix the old forms with the new (e.g., Mosh in
Ngara), or outside norms with the group’s norms (e.g., Joe in Makadara), while others
only stick to those forms that are identified with their baze. It should not also be assumed
that belonging to a social category like a baze dwarfs other social categories that those
baze members belong to.

While the major focus of this study was linguistic variation, other non linguistic
practices that teams up with linguistic choices in the construction of identity were also

considered. This is what complicates the picture of Sheng’s variation, which increases the

243

distance in social categories that speaks Sheng. With this lack of uniformity in the social
categories together with the differing pattern of identity among various groups, the issue
of whether whether we can talk of large-scale group identity as in Samper’s work (2002)
arises. How, for instance, can we talk of Nairobi as a meta-category with its own identity
if all groups are different? Even in cases of micro categories like bazes, how can we talk
of identity on one hand while emphasizing individual distinctiveness? These are some of
the issues that should be addressed in future studies. This shows that even though this
study advances various issues not addressed in Samper's study, it has just scrapped the

surface of this broad issue of Sheng and identity.

Implications and future research

As concluding remarks, I will tie the implications of this study to the direction for future
research in Sheng. I mentioned in Chapter 1 that Sheng has been blamed for poor
performance in the standard languages among the school students. First of all, the
languages teachers need to understand the reality that Sheng serves a very important
social function among its speakers. Exhorting student to stop using it is not going to
work It is sad that the language teachers still regard Sheng as a ‘concoction’ or
‘corrupted language’ with no rules. As we have see, Sheng operates like any other
mainstream language at least in terms of syntactic structure and word formation rules;
only its lack of standardization especially in terms of orthography inhibits uniformity.
Appreciating this fact will encourage the teacher training colleges to develop a
curriculum that integrate understanding of Sheng among language teachers in order to

equip them with the pedagogical knowledge necessary for making students understand

244

 

the boundary between spoken and written language. This study has shown that when
there is a clear motivation, Sheng speakers accommodate to other Sheng varieties. It is
my belief that they can also accommodate to the standard language in the necessary
contexts so long as teachers adopt the right attitudes. Since demonizing Sheng has proved
to be counterproductive, strategies that appeal to students local and national identities
should be adopted so that they can see the value of both and make the optimal choices as
situations demand.

Another implication of this study is tied to the growing use of Sheng in the
intemet. Kenyans in the diaspora are interacting with their counterparts at home and
sharing their experiences. In addition, intemet users are blurring the identities associated
with the bazes or neighborhoods since cyberspace connects people of different
backgrounds. This does not mean that networks are eliminated altogether. I believe the
study of the intemet use of Sheng and the network patterns of Sheng speakers will come
up with interesting findings concerning identity interaction in cyberspace.

I also pointed out that once innovation spread to the mainstream, its function as a
symbol of a group’s identity reduces. The study of the use of Sheng in popular culture
and the media will therefore reveal how speakers of Sheng varieties respond to this
invasion of their private spaces by continuous innovation.

One of the major limitations of this study was the amount of time spent on
subjects. I believe that in order to account for different patterns of identity construction in
a particular baze, a single visit is not enough. While this cross-sectional study came up
with interesting findings, a longitudinal study would have revealed more dynamics

among the individual members of the baze. Eckert (2000) was able to come up with such

245

detailed findings in her study of adolescents in a Detroit suburban High School because
of the amount of time she spent interacting with the students. I believe this approach
would yield better results in the study of Sheng. It is just not enough to attend one-
interview sessions and then make generalizations. Now that I have concluded that bazes
are an important locus for Sheng variation, future research should focus on three or four
bazes, perhaps with only two females and two males in each, and spend a lot of time in
such environments. This will allow the researcher to verify the claims that members make
in their interviews. In addition, there will be a chance to observe different discourse
genres, which might be governed by their own rules, even within a single baze.

Without such an investigation, however, this study has shown that the emergence
of a contact variety such as Sheng cannot be properly assessed, not even understood for
what it is in its locale without an understanding of both its influence and dependence on

the emergence of multifaceted identities in such a setting.

246

 

APPENDIX A

Questionnaire 1: For school children under 18 years old
1. Age Sex
2. School (if applicable)

3. Which residential area have you lived in Nairobi (begin with the most recent in that order
Residential areas Duration

 

 

 

 

 

 

 

 

 

 

 

 

 

9999‘s»

 

4. In this scale, how much would you say that you speak any of the given Languages per
day? mark (x)

Langgages All times A lot Little Veg little Never

Sheng
Swahili
English
Others 1.
Others 2

LII-huh):—

5. How good is your Sheng
a. perfect b. very good c. fair (1 not so good e. not good at all
6. How old were you when began speaking Sheng? years

 

7. Where did you learn Sheng
school, home, friends, church, TV, music, radio
8. Who do you speak Sheng with?
Friends, bothers and sisters, parents, teachers, policemen, anybody
9. Who would you not speak Sheng with, even if s/he knew?
Parents, teachers, strangers, friends, policemen, nobody
10. Do all people speak Sheng the same way?
Yes No
11. Write down some of the differences between the Sheng you speak here with your
friends and the Sheng that others speak?

 

 

Father to eat foolish person
Mother to sleep to go

to rest thief child
policeman house car
girl money teacher
boy music

12. Would you like to move out of your estate?

247

APPENDIX B

Questionnaire 2: For Subjects over 18 years
1. Age Sex
2. School (if applicable)

 

 

 

3. Which residential area have you lived in Nairobi (begin with the most recent in that
order

Residential areas Duration

 

   

4. In this scale, how much would you say that you speak any of the given Languages per
day? Mark (X)

Lan a es All times A lot Little Very little Never

Sheng
Swahili
English
Others 1.
Others 2

mthg—r

5. How old were you when began speaking Sheng?
years
6. Where did you learn Sheng

 

7. Who do you speak Sheng with?
8. Who would you not speak Sheng with, even if s/he knew?

8. Do all people speak Sheng the same way?
Yes No

 

 

9. Write down some of the differences between the Sheng you speak with your friends
and the Sheng that others speak?

248

 

10. Identity: How do these people speak Sheng? Mark (X)

 

Exactly like
locations me

A little like me

Very different
from me

I don’t understand
them

 

 

 

 

 

mcow>

 

 

 

 

 

 

ll. Judgment: How do these people speak their Sheng’ Mark (X)
Not so bad

locations V well

Not well

 

12. The Sheng’ Index: How would you rate your Sheng?
a. Perfect b. Very good c. Fair d. Not so good e. Not good at all

13. Where did you go to School? Primary ............ secondary ............

14. Do you have a small group of friends in your neighborhood whom you speak Sheng

with?

15. Who would you consider as people who belong to your group?

16. Are there times you want to belong to other groups?

17. Do you have other family members in the neighborhood who also speak Sheng?

18. Do you have at least two people from the neighborhood with home you work at the

same place?

19. Where you work, do you work with at least two others of the same sex?

20. Are the social activities you engage in restricted to the same neighborhood?

21. How many of your closest friends belong to your ethnic group?
a. 85-100% b. 70-84% c. 55-69% (I. 40-54% 3. Below 40%

22. Do you consider moving into another neighborhood?

249

 

APPENDIX C

Rift Valley Academy RVA Lingo

Record: [26 to 50] of 219

 

30. choo CHOH, n., Swahili. toilet: toilet.

32. chop CHAHP, n., Swahili: kuchapa kazi, meanIng to work
energetically: one who works energetically. ' -- ‘ '
Alternate spellings: chap '

   
 
    

 

36. chuck CHUHK, v., English. to leave

38. chumes CHOOMZ, n., Swahili: uchumi~economyz money.
Alternate spellings. churns, chooms

40. conch KAWNCH, adj., English: conch: hard, difficult.

42. convos kawn-VOHZ, n., English: conversation; see alsO' com/0'.

. cruise

 

50. dirocks DIGH-rocks, n., English: diarrhea.
Alternate spellings: dirox

Adapted from Davids Leber’s website at http ://www.freepgs.com/lebes/ with his
permission.

250

APPENDIX D

Synonym fiequencies N=212

Frequencies for ‘father’

buda [62] budeng' [1] budengdeng [l] fatha [17] fathe [1] fusebox [1] gadabu [l]
kize [1]mbulala[1] mbuyu [141] mdabu [3] mdagala[4] mzae [2] oldman[1]
papa [1] tifar [l] warinzu [1] zeiksman [1]

Frequencies for ‘mother’

buda [1] gadama [1] mama [l] masa [103] matha [22] mathako [1] mathe[55]
mathor [5] mdhema [l] mnyaka [3] mnyasa [3] mokoro [1] monthere [l] monthero
[1] msa[1] mtassi [1] mthama[1] mthas [l] munthre [l] muthama [25] muthes [1]
mwothes [l] themaa [1] therma [2] tima [l] zeikswom [1]

Frequencies for ‘rest’

Bangaiza [10] bentle [1] gichi [l] kubanjuk [1] kubarizi [1] kubazz [l] kubreeze
[l] kubuaga[1] kuchill [3] kudozi [3] kugandiz [3] kujibanz[1] kukeyo [2] kula
njaro [2] kumanya [2] kumorea [1] kumoss [l] kupoa [ll] kupoza [l] kupozi [34]
kupumz [l] kupumzika [3] kurelax [12] kurest [5] kutulia [27] kutulix [l]kutuliza
[26] kutuna [3] kuuma vako [2] kuwaka [1] liatu [2] naWengine [1] poa [l] swach
[1] take rest[1] tinare [2] tired[l] tulia[2] tuliatulia [l] tuliz[l] tuliza[3]
usiogope [l] wasee [2]

Frequencies for ‘police’

afuade [1] antimege [1] babby [l] banga [4] gava [43] goshogi [2] gwang [l]
karao[115] mabeast[l4] manyani [l] mapolisi [1]mban'ga[l] msoro[1] oraka [1]
pai [5] ponyi [l6] sanse [10]sarka [l] sinya [4] tipoli [1] vedi [2] wachi [1]

Frequencies for ‘girl’

Azguoba[l] bogola [1] buss [l] chibela [l] chik[3]chile[3] cholosa[2] dame

[3] dem [36] gala [1] gwaru[1]hoe [1] jaba [l] kemfaa[1] kenta [l] kff[l]
kichuna[1]kidalipo [1] mama[l] man [1] mandu [3] manyake[2] manzi[l32]

mbasjl] mbus[5] miogar[1] mlosho[1] mresh [l] mroro[l3] mshe[3] mshi

[5] msusu [2] mtasi [l] mtoi [1] motto [l] nyambizi [l]nyang'we [1] nzima [l] roro
[1] shore [17] shwala[l] siste[2] supuu [1] toto [l] wanyina [l]

Frequencies for ‘boy’

beshte [1] boi [6] boy [1] boyboy [1] boyii [l]bro[l] chali [178] charty[1]
dume[l]hommie [l]huyu boy[l] jamaa [8] kaboy [l] kijana [5] licha [1]mdhi [1]
mru[l]msee[3]mthi[l] muthee [1] muthii[l] mzee[1] ndaume [1]

Frequencies for ‘eat’

bonya [6]dema [34] difu [l] dishi [20] ficha[1]finya[1] fweba [l] fyamba [2]
fyenfa [1] hog [5] jibonda [2] jirock[l] jistack[1] jitibu[l] kumachez [l] kumatema

251

[2] kupiga [1] kusosi [2] kustack [3] kutia nbani [3] manga [129] minya [2] munch
[l]ngama[l] sosi [1] stack [2] venga[2]

Frequencies for ‘sleep’

KO. [1] bleki [l] kudoro [52] kudoroi [1] kudorome [1] kudosro [l] kudoz [3] kudoze
[4] kudozi [68] kuduna[1] kukaita[l]kukata[l] kukoma[1] kulaka[l] kulala [3]
kunatu [2] kushonga [2] kuswatch [5] kutulia [1] kutuna [59] kuuma vako [l] natu [l]
swachi [1]

Frequencies for ‘thief’

Bandi [5] bandishi [1] bandit [l] chizi [1] conmen[l]dindo [1] dingo [23] dongi [1]
goda [1] gondi [107] goya[l] jambazi[l] jambo[4] jangili [l] kauzi [13] mgosa [37]
mnati [1] mwizi [1] pinji [4] shifia[l] sinya[1] sniper [4] thegi [10]

Frequencies for ‘house’
boma[1]cube [l]digz[3] ghetto[l]hao [66] jake [1] keja [132] krib [5] kubalo [2]
maskan[1]mbanyu[11] moroto[2]mtaani [l] riggs[l] won [1]

Frequencies for ‘money’

alpundez [1] capital [1] chaga [1] chantu [1] chapa [26] cheda [12] chumbs [1] daf [1]
dongera [1] doo [84] ganji [76] gashai [2] greens [1] hela [l] kenge [2] macheda [1]
mafuta [1] majani [1] mapesa [l] monyo [4] munde [6] munke [l] murangi [1] ngoma
[l] niado [6] niadu [1] papers [1] pesa[1] sape [1] winchi [2]

Frequencies for ‘music’

Beats [1] chizi [l] debe [3] genge [5] gogo [I] hewa [1] lyrics [2] maflavou [l] mahewa
[50] malyrics [l] mavitu [1] mogithi [1] music [1] ngoma [134] reggae [2] rwembo [1]
song [I] timus [1] vokol [2] wimbo [2] wingnimb [l] ziki [l]

Frequencies for ‘fool’

anti-nyita [1] bani [3] barhe [3] bore [7] bure [2] chizi [34] chodha [l] chununu [1]
danda [l3] dula [1] dawn [1] dwanzi [l] fala [117] fika [1] fyatika [2] fyatuko [1]
gamlez [l] gwethe [l] gweze [3] jamaa [l] kamande [1] kemothe[1] kerothi [1]
kichoyo [l] kwaya[l] lafa [1] mbatopas [1] mbonye [l] mdongo [l] mjinga[3]
mjeshi [1] mshakee [2] mwenda [2] mwere [l] nduwele [1] ng'ea[1]ngamres [l]
ngamwe [l] ngoie [l] nongwe [6] pankass [1] punju [l] quir [1] shenzi [2] shoga [l]
thati [2] tiwetini [1] tost [2] tothi [8] upende [1] weri [3] zushu [1]

Frequencies for ‘go’

Bail [2] bounce [4] chak [3] cheza [l] chora [1] dark [1] gich [l] gushia [l] gwan [1]
inua [1] ishia [133]jiches [1] jicurve [1]jikata [43] jikate [l] jikov [1]jipanga [1] jitoa
[5] jivuta [l] kacha [l] kugo [1] kuhepa[1] kula vako [1] kung’oka [l] kuwiz [1] kwenda
[2] lembe [1] mkidi [1] ndae [l] nyaa [6] nyorosha [2] ramus [I] shoot [2] tinago [l]
tuende [l] tukore [1] uiz [1]

252

Frequencies for ‘child’
chali [1] doli [1] kijanaa [l] mdiki [l l] mdoido [1] mjunior [12] mkidi [49] mless[l]
mniaju [2] mtoi [141] myoung [3] ndai [1] tich[1] wothes [1]

Frequencies for ‘car’

bima [1] breka [2] dinga [111] gari [2] mathree [4] mboko [3] mbokonya [1] mode [1]
moti [ll] mots [1] murenga [10] ndai [84] nganya [1] rack [6] ride [4] ticar [1] wheels
[3]

Frequencies for ‘teacher’

DH [1] mchals [1] mdie [1] mdijo [1] mode [66] mtiche [1] mwalimu [1] no [1] odich
[15] odicha [1] odicho [l] odido [1] odijo [123] odisa [6] odish [l] odiso [1] odojo [2]
ondiso [4] La [1] tiche [9] tie [4] udicha [1]

253

 

 

Bibliography

Abdulaziz, Mohammed & Ken Osinde 1997. Sheng and Engsh: Development of a
mixed code among the urban youth in Kenya. International Journal of the
Sociology of Language, 125, 43-63.

Abrahams, Roger 1964. Deep down in the Jungle ...; Negro Narrative Folklore from
the Streets of Philadelphia. Folklore Associates, Hatboro, Pa.

Ahmed Nassir bin Juma Bhalo.l97l. Malenga wa Mvita: Diwani ya Ustadh Bhalo,
Tungo za Ahmad Nassir. Oxford University Press, Nairobi.

Ashton, Ethel 1969. Swahili Grammar (Including Intonation). Longman, London.

Austin, John 1962. How to do things with words. In Adam Jaworski and Nikolas
Coupland (eds.)( 1 999), The Discourse Reader. Routledge, London.

Bailey, Benjamin 2000. Language and negotiation of ethnic/racial identity among
Dominican Americans. Language in Society, 29, 555-582.

Barasa, Kukubo. 2000. “Language policy incubation in Kenya” in Kimani Njogu et.al.,
(eds.) Kiswahili Katika Kame ya Ishirini na Moja (Kiswahili in the 21“" Century).
Center for Advanced Studies of Afiican Society, Cape Town.

Bell, Allan 1984. Language style as audience design. Language in Society 13, 145-204.

Bell, Allan 1992. Language of the News media. Blackwell. Oxford.

Bell, Allan 1999. Styling the other to define the self: A study in New Zealand identity
making. Journal of Sociolinguistics, 3, 4, 523-541.

Blackledge, Adrian & Aneta Pavlenko. 2001. Negotiating identities in multilingual
contexts. The international Journal of Bilingualism, 5, 3, 243-257.

Bleek, Wilhelm Heinrich 1862. Comparative Grammar of South Afi'ican Langauges.
Triibner and Co. London.

Bolonyai, Agnes 2005. Who was the best?: Power knowledge and rationality in bilingual
girls’ code choices”. Journal of Sociolinguistics, 9, 1, 3-27.

Bolonyai, Agnes 2000. Elective Affinities: Language Contact in the Abstract Lexicon
and its structural Consequences. International Journal of Bilingualism 4, 1, 81-
106

Bourdieu, Pierre 1990. The logic of Practice. Stanford University Press, Stanford.

254

Bourdieu, Pierre 1991. Language and Symbolic Power. Harvard University Press
Cambridge Massachusetts.

Brizuela, Maquela, Elaine Andersen & Lynne Stallings 1999. Discourse markers as
indicators of register. Hispania 82, 1, 128-141

Brown, Penelope & Steven C. Levinson 1987. Politeness: Some Universals in
Language Usage. Cambridge University Press, Cambridge.

Brubaker, Rogers & Frederick Cooper 2000. Beyond identity. Theory and society 29,
1-4 7.

Bucholtz, Mary 1999. Why be normal?: Language and identity practices in a community
of nerd girls. Language in Society, 28, 203-223.

Burger, Peter & Luckman Thomas 1966. The social construction of reality: A treatise
into the sociology of knowledge. Doubleday, Garden City, NY.

Burja, Janet 1974. Pumwani: Language usage in an urban Muslim community. In
Wilfred.H.Whiteley (ed.) Language in Kenya. Oxford University Press. Nairobi.

Burke, Peter & Donald C. Reitzes 1991. An identity theory approach to commitment.
Social Psychology Quarterly, 54, 3, 239-251.

Cameron, Deborah & Coates Jennifer 1988. Some problems in the sociolinguistics
explanations of differences. In Jennifer Coates & Deborah Cameron (eds.) Women

in their Speech Communities 13-26. Longman, London.

Chambers Jack 1995. The Canada-US border as a vanishing isogloss: The evidence of
Chesterfield. Jornal of English Linguistics, 23, 155-166.

Chimera, Rocha 1998. Kiswahili: Past, Present and Future Horizons. Nairobi University
Press, Nairobi.

Chomsky, Noam. 1979. Language and Responsibility. Pantheon Books, New York.

Colins Michael 2005 What we mean when we say ‘creole’: An Interview with Salikoko
S. Mufwene. Calaloo, 28, 2, 425-462.

Cutler, Cecilia 1999. Yorkville crossing: White teens, hip hop and African American
English. Journal of Sociolinguistics, 3, 4, 428-442.

lDe Meis, Carla 2002. House and the street: Narratives of identity in a liminal space
among prostitutes in Brazil. Ethos, 30, 1-2, 3-24.

Dirim, Inci & Andreas Hieronymus 2003. Cultural orientation and language use among

255

multilingual youth groups: For me it is like we all speak one language. Journal of
Multilingual and Multicultural Development, 24, 1&2, 42-55.

Downes, William 1998. Language and Society (Second Edition). Cambridge University
Press, Cambridge.

Dundes, Allan 1972. The strategy of Turkish boys’ verbal dueling rhymes. In John
Gumperz & Dell Hymes (eds.) Directions in Sociolinguistics: the ethnography
of Communication, 130-160. Holt Reinhert & Watson, New York.

Eckert, Penelope & Sally McConnell-Ginet 1992. Think practically and look locally:

Language and gender as community-based practice. Annual Review of
Anthropology, 21, 461-490.

Eckert, Penelope & Sally McConnell-Ginet 1999. New generalizations and explanations
in language and gender research. Language in Society, 28, 185-201

Eckert, Penelope 2000. Linguistic Variation as Social Practice: The Linguistic
Construction of Identity in Belten High. Blackwell. MA, Malden

Edwards, John 1985. Language, society and identity. Basil Blackwell, London.
Elder, Joseph 1954. Jargon— Good and Bad. Science, 119, 536-538

Fairclough, Norman1995. Critical Discourse Analysis: The Critical Study of Language.
Longman, London

F asold. Ralph 1987. The Sociolinguistics of Society. Basil Blackwell inc. Cambridge,
Massachusetts

Fawcett, RP. 1970. The medium of education in the lower primary school in Afiica
with special reference to Kenya. In T.P Gorman (ed.). Language in Education in
Eastern Africa: Papers from the first Eastern Afiican conference on
language and linguistics 51-69. Oxford University Press, Nairobi,

Fisher, John 1958. Social Influences on the choice of a linguistic variant. Word, 14, 47-56

F ishman, Joshua 1994. The truth about language and culture (and a note about its
relevance to the Jewish case). International Journal of the Sociology of Language,
109, 83-96.

Gardner-Chloros, Penelope 1997. Code-switching: Language selection in three
Strasbourg department stores. In Nikolas Copeland & Adam Jaworski (eds.)
Sociolinguistics 361-3 75 St. Martin’s Press, New York.

Giampapa, Frances 2001. Hyphenated identities: Italian-Canadian youth and the

256

negotiation of ethnic identities in Toronto. The International Journal of
Bilingualism, 5, 3, 279-315.

Giles, Howard, Donald Taylor and Richard Bourhis 1973. Towards a theory of
interpersonal accommodation through speech: Some Canadian data. Language in
Society, 2, 177-192.

Giles, Howard and Peter Powersland 1975. Speech style and social evaluation.
Academic Press, London.

Giles, Howard et al., 1979. Accommodation theory: Optimal levels of convergence.
In Howard Giles & Robert St Clair Language and social psychology 45 -65 .
University Park Press, Baltimore.

Githinji, Peter 2003. Language Attitudes: Nairobi People and Sheng. MA Thesis.
Michigan State University.

Githiora, Chege 2002. Sheng: Peer language, Swahili dialect or emerging Creole?
Journal of Afiican Cultural Studies, 15, 2, 159-181.

Goffinan, Erving 1967. On face—work: An analysis of ritual elements in social
interaction. In Adam Jaworski and Nikolas Coupland (1999), The Discourse
Reader 306-320. Routledge, London.

Gorman, T.P. 1974. The development of language policy in Kenya with particular
reference to the educational system in Wilfred Whiteley (ed.) Language in Kenya
397-455. Oxford University Press. Nairobi.

Grice HP. 1975. Logic and Conversation. In A. Jaworski and N. Coupland (1999),
The Discourse Reader, 76-88. Routledge, London.

Habwe, John 2000. Kiswahili language as Kenyas mobilizing tool: A critical approach.
In Kimani Njogu et.al, (eds. ) Kiswahili Katika Kame ya Ishirini na Moja
(Kiswahili in the 21 3' Century). Center for Advanced Studies of Afiican Society,
Cape Town.

Halliday, Michael Alexander 1973. Exploration in the Functions of Language. Edward
Anold, London.

Halliday, Michael Alexander 1978. Language as social semiotic: The social
interpretation of Language and Meaning. University Park Press, Baltimore.

Harries, Lyndon 1975. “The nationalization of Swahili in Kenya”. Papers presented at the
African Studies Conference at New York University on Feb 17-18, 1975.

Hazen, Kirk 2002. Identity and language variation in a rural community. Language, 78,

257

2, 240-257.

Hoare, Rachel 2001. An integrative approach to language attitudes and identity in
Brittany. Journal of Sociolinguistics, 5, 1, 73-84.

Hofineyr, Isabel, Joyce Nyiro and James Ogude 2003. ‘Who can bwogo me?’ Popular
culture in Kenya. Social Identities, 9, 3, 373-382

Hogg, Michael, DeborahTerry and Katherine White 1995. A Tale of two Theories: A
critical comparison of identity theory with social identity theory. Social
Psychology Quarterly, 58, 4, 255-269.

Holm, John 2000. An introduction to Pidgin and Creoles. Cambridge University Press,
Cambridge.

Horton, John 1972. Time and cool people. In T. Kochman (ed.) Rappin ' and sytlin'
out: Communication Urban Black America 19-31. University of Illinois Press,
Chicago

Hume, Elizabeth 2004. The indeterminacy/attestation model of syllabic inversion.
Language, 80, 2, 203-237.

Hymes, Dell 1974. Foundations of Sociolinguistics: An Ethnographic Approach.
University of Pennsylvania Press. Philadelphia

Hymes, Dell 1964. Towards ethnographies of communication: the analysis of
communicative events. In Paolo Giglioli (ed.) Language and Social Context
(1972) 21-44. Penguin, Harmondsworth.

Ito, Rika and Dennis Preston 1998. Identity, discourse, and language variation.
Language and Social Psychology, 17, 465-484.

Kang’ethe, Iraki 2004. Cognitive Efficiency: The sheng phenomenon in Kenya.
Pragmatics, 14, 1, 55-68.

Kiesling, Scott F abius 1998. Men’s identities and sociolinguistic variation: The case of
Fraternity men. Journal of Sociolinguistics, 2, 1, 69-99.

King’ei Kitula 2001. Challenges and opportunities in the development of African
Langauges in Kenya: A Contemporary perspective. In J .S.Mdee and H.J.M
Mwasonko (eds.) Makala ya Kongamano la Kimataifa Kiswahili Proceedings,
122-133. Taasisi ya Uchunguzi wa Kiswahili, Chuo Kikuu cha Dar es Salaam,
Dar es Salaam.

Kochman, Thomas 1972. Towards an ethnography of black American speech behavior.
In Thomas. Kochman (ed.) Rappin ’ and sytlin ' out: Communication Urban Black

258

America, 241-264. University of Illinois Press, Chicago.

Kouega Jean-Paul 2003. Word Formative Processes in Camfranglais. World Englishes,
22, 4, 511-538.

Labov, William 1972a. Sociolinguistic Patterns. University of Pennsylvania Press,
Philadelphia

Labov, William 1972b. Language in the Inner City: Studies in Black English Vernacular.
University of Pennsylvania Press, Philadelphia.

Labov, William 1989. The exact description of the speech community: Short 0 in
Philadelphia. In Ralph F asold and Deborah Schiffiin (eds.). Language Change
and Variation, 1-57 Georgetown University Press, Washington DC.

Le Page Robert, and Andrée Tabouret-Keller 1985. Acts of Identity: Creole-Based
Approaches to Language and Ethnicity. Cambridge University Press, Cambridge.

Lo, Adrienne 1999. Codeswitching, speech community membership, and the
construction of ethnic identity. Journal of Sociolinguistics, 3, 4, 461-479.

Macaulay, Ronald 1997. Standards and Variation in Urban Speech. John Benjamins
Publishing Company Amsterdam/Philadelphia.

Marcyliena, Morgan 2002. Chapter 4. When women speak: how and why we enter. In
Language, Discourse and Power in African American Culture, 84-1 10.
Cambridge University Press, Cambridge, New York.

Mazrui, Alamin 1995. Slang and Codeswitching: The case of Sheng in Kenya.
Afrikanistische Arbeitspapiere, 42, 168-179.

Mbaabu Ireri 2000. A case for teh use of Kiswahili as a langauge of instruction in lower
primary school in Kenya. In Kimani Njogu et.al., (eds.). Kiswahili Katika Karne
ya Ishirini na Moja (Kiswahili in the 21‘" Century). Center for Advanced Studies
of Afiican Society, Cape Town.

Mbaabu, Ireri 1996. Language Policy in East Afiica: A Dependency Theory Perspective.
Education Research and Publications, Nairobi.

Mbaabu, Ireri 1977. New Horizons 1n Kiswahili: A Synthesis in Developments, Research
and Literature. Kenya Literature Bureau, Nairobi.

Mbaabu, Ireri & Kazungu Nzuga 2003. Sheng-English Dictionary: Deciphering East

Africa ’s Underworld Languge. Taasisi ya Uchunguzi wa Kiswahili. Chuo Kikuu
cha Dar es Salaam. Dar es Salaam.

259

Mbugua wa Mungai 2003. Identity Politics in Nairobi Matatu Folklore. PhD
Dissertation. Hebrew University, Jerusalem.

Meyerhoff, Miriam 1998. Accommodating your data: the use and misuse of
accommodation theory in sociolinguistics. Language and Community, 18, 205-
225.

Meyerhoff, Miriam & Nancy Niedzielski 1994. Resistance to creolization: An
interpersonal and intergroup account. Language and Communication, 4, 313-330

Meyerhoff, Miriam. 1999. Sorry in the Pacific: Defining communities, defining practice.
Language in Society, 28, 225-238.

Milroy, James & Leslie Milroy. 1985. Linguistic change, social network and speakers
innovation. Linguistics, 21: 339-384.

Milroy, Leslie 1980. Language and Social Networks. Blackwell, Oxford: New York.

Mitullah, Winnie 2003. Urban slums, The case of Nairobi, Kenya. In UN-Habitat (2003)
Part IV: 'Summary of City Case Studies'. Global Report on Human Settlements
2003, The Challenge of Slums, 195-228. Earthscan, London.

Mitchell-Keman, Claudia 1972. Signifying, loud-talking and marking. In Thomas
Kochman (ed.) Rappin ’ and sytlin ' out: Communication Urban Black
America, 15834 University of Illinois Press, Chicago.

Mkangi, Katama 1985. The political economy of Swahili: A Kenya-Tanzania
Comparison” in Joan Maw and David Parkin (eds.) Swahili Language and
Society, Beitra'ge Zur Afi'ikanistik, 333-348, Band 23 Wien 1985 .

Moga, J acko, and Dan Fee 2000. Sheng Dictionary, Third edition. Ginseng Publishers
Nairobi.

Mohammed, Abdulla 2001. Modern Swahili Grammar. East African Educational
Publishers, Nairobi.

Musau, Paul 2001. The spread of Kiswahili in Kenya in the new millennium: prospects
and challenges. In J .S.Mdee & H.J.M Mwasonko (eds.) Makala ya
Kongamano la Kimataifa Kiswahili 2000 Proceedings, 134-143. Taasisi ya
Uchunguzi wa Kiswahili, Chuo Kikuu ch Dar es Salaam, Dar es Salaam.

Myers Scotton, Carol 1974. Codeswitching as a “safe choice” in choosing a Lingua
Franca. In William McCormack and Stephen Wurm (eds.) Language and
Society: Anthropological Issues, 71-87. Mouton, The Hague.

Myers-Scotton, Carol 1983. The negotiation of identities in conversation: A theory of

260

markerdness and code choice. International journal of the sociology of language,
44, l 1 5-1 36.

Myers-Scotton, Carol 1992. Codeswitching in Africa: A model of the social functions of
code relations” in Robert K. Herbert (ed.) Language and Society in Afiica.
Witwatersland University Press. Witwatersland.

Myers-Scotton, Carol 1993a. Social Motivations for Codeswitching: Evidence from
Afiica. In Suzzane Romaine and P Mulhausler (eds.). Oxford studies in Language
Contact. Claredon Press, Oxford.

Myers-Scotton, Carol 1993. Dueling languages: grammatical structure in
codeswitching. Clarendon, Oxford.

Myers-Scotton, Carol 1998. A theoretical introduction to the markedness model. In Carol
Myers-Scotton (ed.) Codes and Consequences, Choosing Linguistic
Varieties 18-38. Oxford University Press, New York.

Myers-Scotton, Carol and Agnes Bolonyai 2001. Calculating speakers: codeswitching in
a rational choice model. Language in Society, 31, 1, 1-28.

Myhill, John 2003. The native speaker, identity and the authenticity hierarchy. Language
Sciences, 2, 77-97.

Ngesa, Mildred 2002. ‘And Sheng’ lives on... Life magazine, February 11-17, 2002. East
African Standard online.

Niko, Bisnier 2003. Crossing genders, mixing languages: the linguistic construction of
transgenderism in Tonga. Janet Holmes and Miriam Meyerhoff (eds.). The
Handbook of Language and Gender. Blackwell

Olima, Washington 2001. The dynamics and implications of sustaining urban spatial
segregation in Kenya: Experiences from the Nairobi metropolis. Conference
paper presented at the Lincoln Institute course ‘International seminar on
segregation in the city” on July 26-28.

Osinde , Ken 1986. Sheng: An investigation into the Social and Structural Aspects of an
Evolving Language. MA Thesis, University of Nairobi.

Preston, Dennis 1996. Where the worst English is spoken. In E. Schneider (ed.) Focus on
the USA, 297-360. Benjamins. Amsterdam.

Pustet, Regina 2003. Prototype effects in discourse and the synonymy issue: Two Lakota
postpositions. Cognitive Linguistics, 14, 4, 349-378

Rampton, Ben 1997, Langauge crossing and the redefinition of reality: Implications for

261

research on code-switching community. In Working Papers in Urban
Langauges and Literacies, 1-21. Kings college, London.

Rampton, Ben 2005. Crossing: Langauge and Ethnicity among Adolescents. Longman,
London.

Reitzes, Donald and Burke, Peter 1991. An identity theory approach to commitment.
Social psychology Quarterly, 54, 3, 239-251.

Reves Haviland 1926. What is slang?: A survey of opinion. American Speech, 1, 4, 216-
220

Rhoads, John 1977. Linguistic Diversity and Language Belief in Kenya: The Special
Position of Kiswahili. Syracuse University, New York.

Rozin Paul, and Michael Seagal 2003. Vegemite as a marker of national identity.
Gastronomica — The Journal of Food and Culture, 3, 4, 63-67.

Samper, David 2004: Afiica is still our mama: Kenyan rappers, youth identity and the
revitalization of traditional values. African Identities, 2,1, 31-51.

Samper, David 2002. Talking Sheng: The role of a Hybrid Language in the Construction
of Identity and Youth Culture in Nairobi Kenya. PhD Dissertation, University of
Pennsylvania.

Sankoff, David and Suzanne Laberge 1978. The linguistic market and the statistical
explanation of variability. In Davod Sankoff (ed.) Linguistic Variation, Models
and Methods, 239-250. Academic Press, New York .

Sapir, Edward 1929 The status of linguistics as a science. In E. Sapir (1958) Culture,
Language and Personality. University of California Press, Berkeley.

Saussure, F erdinard 1959. Course in General Linguistics. Philosophical Library, New
York.

Schilling-Estes, Natalie 2002. American English and social dialect variation and gender.
Journal of English Linguistics, 30, 2, 122-137.

Schilling-Estes, Natalie 1998. Investigating “self-conscious” speech: The performance
register in Ocracoke English. Language in Society, 27, 53-83.

Schilling-Estes, Natalie 2000. Investigating intra-ethnic differentiation: lay/ in Lumbee
native American English. Language Variation and Change, 12, 141-174.

Sebba, Mark & Shirley Tate 2002. "'Global" and "local" identities in the discourses of British
born Caribbeans.’ International Journal of Bilingualism, 6, 1, 75-89.

262

Smith, Heather and Colin Leach 2004. Group membership and everyday social
comparison experiences. European Journal of social Psychology, 34, 297-308

Smitherman, Geneva 1977. Talking and T estijying: The Language of Black America.
Houghton Mifflin Company, Boston.

Spyropoulos, Mary 1987. Sheng: some preliminary investigations into a recently emerged
Nairobi street language. Journal of the Anthropological Society of Oxford, 18, 1,
125-136.

Stets, Jan and Peter Burke 2000. Identity theory and social identity theory. Social
Psychology Quarterly, 63, 3, 224-237.

Stojicié Violeta 2004. Sociolinguistic stimuli development of the English lexicon-
langauge contact and special need. Linguistics and Literature, 3, 1, 29-36

Stryker, Sheldon & Peter Burke 2000. The past, present and future of identity theory.
Social Psychology Quarterly, 63, 284-297.

Tajfel, Henri 1982. Instrumentality, identity and social comparisons. Social Identity and
Intergroup Relations. University Press Cambridge. Cambridge.

Tajfel, Henri and John Turner 1986. The social identity theory of intergroup behavior. In
Stephen Worchel & William Austin (eds.) The Social Psychology of Intergroup
Relations. Nelson-Hall, Chicago.

Thomas, Edward 1970. Chinook—A History and Dictionary. Binfords and Morts
Portland.

Thompson, John 1991. Introduction. In Pierre Bourdieu (1991) Language and Symbolic
Power. Harvard University Press, Cambridge Massachusetts.

Trask R.L. 1996. Historical Linguistics. Anold, London.

Trudgill, Peter 1974. The social differentiation of English in Norwich. Cambridge
University Press, Cambridge.

Turner, John 1982. Towards a cognitive redefinition of the social groups. Social Identity
and Intergroup Relations. Cambridge University Press, Cambridge.

Vermeij, Lotte 2004. Ya know what I’m sayin? The double meaning of language crossing
among teenagers in Netherlands. International Journal of the Sociology of
Language, 170:141-168.

Whiteley Wilfred 1969. Swahili: The rise of a national Language. Methuen and Co.
London. '

263

Webgages

www.ethnologe.com

news.bbc.co.uk_/

www.castandard.net

www.ngtionmedi2_r.com

www.macshadacom

www.l_<envanivetu.com

www.t1ipgallerv.com
www.homesZOOOkenvg.com/Properties/l-IC66.htm

www.livejoumal.com/communig/boondocks comic

www.cwenger.com/theory/index.htm

www.sidint.org[P_1_1blications/Docs/pulling-apartpdf

www.engumu.se/stephanie/web/QnguageBlogs.pdf

www.freepgs.com/lebes/

264