Political Discourses, Women’s

Voices: Mismatches in
Representation
Copyright © 2010
Critical Approaches to Discourse Analysis across Disciplines
http://cadaad.net/ejournal
Vol 4 (1): 73 – 92
ISSN: 1752-3079

JACINTA NDAMBUKI
University of the Witwatersrand
[email protected]

HILARY JANKS
University of the Witwatersrand
[email protected]

Abstract
This paper is part of a larger project whose overall aim is to investigate the representation
of women’s issues in Makueni District, a rural district in Kenya, using Critical Discourse
Analysis (CDA). The study explores the mismatches between the way politicians select and
represent these issues and the way women construct these issues in women’s groups. This
paper focuses on representations of women’s agency. How women construct their agency is
contrasted with that of politicians and community leaders. This social science research is
multidisciplinary and crosses the fields of language, gender studies and politics. Data was
collected by use of focus group discussions, political speeches and interviews. The data for
the entire study consisted of eleven focus group discussions with women’s groups, four
political speeches and ten interviews with politicians and other community leaders. This
article is based on four focus group discussions, and four interviews. The analysis focuses
on the use of pronouns and modality. Each of these linguistic features provides a different
lens on the data which enables us to understand the construction of agency. While women,
politicians and other community leaders construct women’s agency within deficit
discourses, these discourses do not match women’s enacted practices or what political and
community leaders say they expect of women. The contradiction inherent in the study is
that everyone constructs women as lacking in agency, yet these women act as agentive
subjects.
Key words: Agency, representation, political discourse, women’s issues, women’s group.

1. Introduction
This paper deals with the construction of women’s agency by both women in
women’s groups as well as by politicians and other community leaders in
Makueni District-Kenya. Here, the focus is on women’s agency based on their
construction of their issues with particular emphasis on the use of modality
and pronouns. Critical Discourse Analysis (CDA) serves as a means to
uncover the subtle ways in which language reveals issues of power and
Ndambuki & Janks P a g e | 74

ideology. Although women’s issues appear to be at the core of sustainable

development discourse, the crucial role of language as a possible source of
understanding their agency has been ignored. Yet text and talk are the most
important ways in which people construct reality and have come to be core
units of analysis in the area of CDA. Texts are the products of linguistic
actions (Wodak 2001: 66) and talk produces spoken texts. Much work on
discourse as social interaction focuses on conversation and dialogue as found
in face-to-face encounters (van Djik 1997; Brown and Yule 1983). The term
‘agency’ is used in this study with reference to Fairclough (2003: 22 ) who
defines it as the capacity of people to act freely, while recognising the
limitations of agency thus, ‘social agents are not ‘free’ agents, they are socially
constrained, but nor are their actions totally socially determined’. The term
representation is used in this paper in two senses: the first is ‘political
representation’, how women are represented in the political system in terms
of numbers. The second sense is ‘discursive representation’ which is
concerned with how women are portrayed or constructed and how they
construct themselves.
In this paper we argue that both women and community leaders construct
women’s agency within deficit discourses that do not match either women’s
enacted practices or the expectations that political and community leaders
have of them. The contradiction inherent in the findings of the study is that
everyone constructs women as lacking in agency, yet these women act as
agentive subjects. In view of this, we argue that the way women construct
themselves and are constructed by others enables us to understand their levels
of involvement in both the political process and in social action.

2. Agency as a Women’s Issue in Rural Kenya

Global perspectives addressing women’s issues can be traced from the World
Conference on women: the first in Mexico 1975, the second in Copenhagen
1980, the third in Nairobi 1985 and the fourth in Beijing in 1995. In Kenya,
this period saw the development of welfare associations at the local and
national levels. For example self-help women’s groups popularly known as
merry-go-rounds were working at the local level; Maendeleo ya Wanawake
Organization (Women in Development) focused on specific women’s
objectives at the national level. At the same time, a number of professional
organizations began to interrogate the unchallenged value and belief systems
that reinforce the subordination of women in society. These included the
Association of African Women for Research and Development, the National
Council of Women in Kenya (NCWK), the Forum of African Women
Educationists (FAWE), the International Federation of Women Lawyers
(IFWL) and Kenya League of Women Voters (KLWV). The growth of women’s
groups in Kenya is positively connected to support notably from the
government and Non Governmental Organizations (NGOs). Though several
scholars provide insightful analyses of women’s groups in Kenya (Chitere
1988; Lewa 2002) viewing the women’s groups in the context of the women’s
movement as most of them do is erroneous. This is a test case of the
treatment of women in a monolithic way and it often led to the obscuring the
actual needs of women’s groups at the expense of the national organization.
75 | P a g e CADAAD

The feminist movement as conceived by the elite women’s organizations while

aimed at transformation differs significantly from women’s groups whose key
pre-occupation is sustainability.
Women constitute 52% of the adult population and 60% of the voting
population in Kenya, making them the majority especially in rural areas
(Khasiani 2000). Due to their numeric advantage one would expect them to
play a significant role in elections since their voter turn-out would be higher
than that of men. However, the Electoral Commission of Kenya registration
figures for the 2002 elections in Makueni Constituency show fewer women
than men (ECK 2002): 38,446 women were registered compared to 40,100
men. Consequently, women’s numerical strength does not seem to translate
to a higher vote and therefore to increased attention to women’s concerns and
interests. In other words, despite having good numbers in terms of political
representation, there is not a matching increase in voting patterns. Khasiani’s
study also showed that a third of the women in Makueni District cannot read
or write. Research by Zubair (2003) in Pakistan suggests that high literacy
increases women’s agency while low literacy diminishes women’s agency.
Since most women in the district are illiterate this has serious implications for
their participation in the political process and their increased agency.
A historical perspective is important in trying to understand women’s agency.
Citing evidence from history, Adhiambo-Oduol (2001) argues that African
communities are either matrilineal or patrilineal. Amadiume (1987) in her
book Male Daughters, Female Husbands: Gender and Sex in an African
Society’, based on research on the Igbo community in Nigeria, argues that
economic changes in colonial times undermined women’s status and reduced
their political role such that patrilineal tendencies persist today to the
detriment of women. Kanogo (2005) expresses similar sentiments based on
research in the Kikuyu community in Kenya observing that on the eve of the
colonial era, women were embedded in gendered constructions of power,
authority and ownership of and access to property in a manner that publicly
diminished their individual agency. While there is more change in the urban
areas, for the present study, women are constructed in a more rural,
traditional set up with fixed roles for men and women. Due to rural-urban
migration, most of the men have left to the towns in search of formal
employment.
A review of the literature has revealed agency as a major theme in the
discussion of key issues that affect women such as education, leadership,
poverty, water and rape (Khasiani 2000; Kanogo 2005; Ndambuki 2006). A
needs assessment conducted by Khasiani (2001) among women in Makueni
showed that socioeconomic and cultural governance (defined as access to and
control of resources, improved social status, and participation in economic
decision-making to ensure equity) remains elusive. The trend in Kenyan
political discourse has over the last two decades been characterized by efforts
towards gender equality and working towards fairness for both genders
especially in making opportunities available for leadership roles, yet women
continue to be excluded in the political process. Thus women’s participation
in political decision-making is mostly peripheral. Attempts to explain this
trend in Kenya, for instance by Khasiani (2001), have focused mainly on the
Ndambuki & Janks P a g e | 76

social-economic and political factors that contribute to the marginalization of

women.
In known contemporary societies, power relations are often asymmetrical
such that women’s interests are systematically subordinated to men’s. In
Wodak’s view, many empirical studies have neglected the context of language
behavior and have often analyzed gender by merely looking at the speaker’s
biological sex (1997: 1). As a result, she proposes that a context-sensitive
approach which looks at gender as a social construct would lead to more
fruitful results i.e. a look at gender in connection with the socio-cultural and
ethnic background of the interlocutors, and in connection with their age, their
level of education, their socio-economic status, and the power-dynamics of the
discourse investigated; some of the factors which this study has taken into
consideration.
At the beginning of 2008, the situation of women in Kenya was exacerbated by
the post-election violence that followed the December 2007 elections.
Women’s ability to sustain their communities has been further eroded by their
displacement. The ‘National Accord and Reconciliation Bill’ subsequently
passed in parliament was designed to restore human rights and to secure a
safe environment for vulnerable groups mainly women and children. This
study comes amid calls for ‘issue-based politics’ in Kenya that will address the
needs of the electorate. Agency is linked to the question of power which is
realized in two spheres: in macro-level civic politics and in everyday micro-
level interaction through social practices. In Kenya, both of these are shaped
by patriarchal discourses. The politics of gender intersect with civic politics
such that representation in the political sense and representation in the
semiotic sense intertwine. Representation in language and discourse is
fundamental to the articulation of policies and actions for the public good.
Gendered social relations contribute to the prevailing conditions for the
production and reception of texts. Our argument is that the possibilities that
exist for women’s semiotic representation of themselves affect their political
representation.

2.1 Popkewitz’ Critique of Agency

In contemporary discourse studies, it is assumed that anything working

towards increasing self-actualising agency is a good thing. It is believed that
human agency is necessary for people to act in order to transform their
conditions of life. This is part of the motivation for this study. However,
Popkewitz (2007) in his book Cosmopolitanism and the Age of School Reform
provides a useful critique of agency. In Popkewitz’ view, cosmopolitanism
brings to the surface the importance and limits of the qualities and
characteristics of modern life that order what ‘we’ are, what ‘we’ should be and
who is enabled to be that ‘we’. He argues that these very practices meant to
include simultaneously exclude. Today, that ‘other’ is placed in a space inside
but recognized to be included and different (Morrrison 1992, cited in
Popkewitz 2007: 35). In his words, ‘today, the abject are given the categories
of disadvantaged, urban, at risk, and the left behind, recognized for inclusion
and paradoxically radically cast out as different’ (Popkewitz 2007: 172). This
view of cosmopolitanism succinctly brings out the paradox of agency in that
despite being seen as the way to empower those seen as being in need, it
77 | P a g e CADAAD

actually excludes them further. Popkewitz’ critique is typical of how most

development discourses have represented women’s marginalization; this
study opens up different possibilities by investigating women’s issues from the
point of view of the women themselves.

3. Methodology
3.1 Research Design

The study was carried out in Mbitini Division, a rural division in Makueni
District in the eastern part of Kenya. The data for the entire research consists
of eleven tape–recorded focus group discussions (qualitative group
interviews) of between 40-60 minutes each, ten interviews with politicians
and other community leaders and four political speeches. This paper is based
on only four focus group discussions and four interviews. After explaining the
research to a meeting of each of the women’s groups they then agreed among
themselves who would remain behind for the focus group discussion. The
meetings were held at places where the participants felt comfortable such as
the women’s project sites, at their homes and at the market place. The
Kikamba texts were first transcribed into English and after systematic
thematic content analysis, purposive selection was used to identify key texts
for CDA analysis.

3.2 Data Analysis

The data is analyzed within a Critical Discourse Analytic framework, an

approach that advocates increased awareness in the use of language to
promote the welfare of marginalized groups (Fairclough 1989). Key scholars
who have contributed immensely to debates in CDA include (Fairclough, 1995;
van Djik 1997, 2001; Wodak 1997, 2001; Blommaert 2005) etc. In
Fairclough’s approach to CDA, individuals are seen as agents capable of
constructing their own agency in their daily interaction. The framework
presents power as embedded in social relations. CDA is able to show that the
semiotic representation of social actors and agency is based on linguistic
choices. Fairclough’s model of CDA involves a description of both the social
processes and structures that give rise to the production of a text. He
conceptualizes these relations using a three dimensional view of discourse that
includes analysis of text (spoken or written), discourse practices (process of
text production and interpretation) and an analysis of the socio-cultural
conditions that affect the production and interpretation of texts. In this paper
the focus is on pronouns, voice and modality as analytic lenses.
In addition to Fairclough’s framework, we use the concept of habitus
developed by Bourdieu (1990, 1991). One’s habitus is a set of predispositions
which incline agents to act and react in certain predisposed ways. These
dispositions generate practices, perceptions, and attitudes which are regular
and which help us to understand how and why the women in this study see
themselves as capable of transformative social action or not.
Ndambuki & Janks P a g e | 78

4. Women’s Construction of their Agency

This section analyses transcripts of the women’s focus group discussions in
order to provide examples of how women construct their agency.1

4.1 Pronouns as an Analytical Lens

A critical examination of the data indicates that participants characterize

themselves in a ‘discourse of suffering’. For example, one of the women in a
focus group discussion describes it thus;

Text (1)
Nituthinaa ovaa tutena mundu wautwonia mbee. Na nengi andu makaleaa
kwonua mbee komesa kumbuka.

We suffer here with no one to show us ahead. And if people are not shown
ahead, can they really emerge?

By using of the plural object -infix ‘-tu-’ (we), the women construct themselves
as a ‘suffering community’ that needs a leader to show them the way. Despite
a culture of community, in which women’s group are deliberately constructed
to give members the support of the collective, women construct themselves in
a discourse which focuses on the centrality of an individual leader. In other
words, they do not see the power that exists when they work together
collectively. The women do not understand that community action underpins
their sustainability not individual power. Their agency is based on the mutual
support that women give one another, not on the power of an individual
leader.
Morrison and Love (1996: 59) underscore the role of the pronoun ‘we’ used to
define ‘who we are’. They contend this is especially pronounced in periods of
upheaval (as in the problematic times of the women’s national organization in
the last decade in Kenya) or national resurgence during which there is usually
an attempt to redefine or reassert a particular identity construction. This
involves identity differentiation between ‘us’ as the powerless women versus
‘them’ (the powerful national women’s organization’ in the context of polarity
that exists between the rural women and the urban presumably elite women in
Kenya. More specifically for this paper, women also construct themselves as
the ‘other’ as in the development discourses that Popkewitz critiques.
The women represent themselves in an impersonal way. They refer to
themselves as ‘people’ and use the third person plural pronoun suffix ‘-ma-’
‘they’ which gives women a generic reference. According to Fairclough (2003:
150) generic reference is often associated with the universal and hence by use
of the generic pronoun ‘they’, the women construct a particular ‘we-
community’ which exhibits the suffering of rural women in general in the local
and global community. What is fore grounded is their suffering, not their
actions that sustain their families. For example in the area of study, the
majority of men have left for the urban centers to look for employment; the
result of this is that men are absent leaving the women to fend for themselves.
The women pool their resources together and look out for each other. They do
what needs to be done with limited resources in order to survive. This action
sustains them but just, for them it is survival.
79 | P a g e CADAAD

4.2 Active and Passive Voice as a Lens

In the following text, the women use a combination of passive and active
voice.

Text (2)
Indi nengi neeko kakyama, nambaokutunga muvea, nathi naa…nakuu Wote,
nambo kutungomuvea. Ninanengwe okindu, tinathi tinaaka ngolova ndaaona
ngolova kuu kiima ninamyona. Vu nasya nuseo.

Then I performed a miracle. I want to give thanks, I went …I went to Wote and
I was given something (support). Then we went and built a storied building. I
had never seen a storied building here at the hills, now I saw it. For that I was
grateful’

The speaker emphasizes her individual agency by the constant use of active
verbs ‘neeko’ ‘I did’, ninathi ninanengwa’ (I went and I was given). The
speaker’s emphasis on her individual contribution to change in the group
might be interpreted to indicate a struggle to express her leadership role as
one of the committee members in the women’s group. Whitely and Muli
(1962) note that in the Kamba language, it is necessary to insert an object –
infix whenever the object of a verb is not mentioned; e.g. –n-eeko (I did it),
ni-n-athi ni-n-anengwa (I went and I was given something). The shift in the
use of the singular pronoun ‘-n-’ ‘I’ to the plural pronoun ‘ti-’ ‘we’ within the
same utterance represents a shift in her construction of agency as within the
collective. Further, the shift from the use of the active to the passive presents
a contradiction in the way the women represent themselves. The use of the
passive ‘was given’ constructs the women as dependent on the support or
handouts of others. They beg for handouts constructing themselves as being
unaware of their rights to services such as water and healthcare. What is
interesting is that this quotation from the data shows the women as taking the
initiative, as capable of action. There is other evidence to show that women
make things happen. For example:

Text (3)
Nundu wakwithia miaka yi ovau inavitie tinai thinani tuteavuanisye.
Tinavuaniw’a tioona ngwatanio noyo itonya utuokoa’. Na tinoona tikwatane
twake nyumba tenu … (points to a house) nakiwe no kikundi...

because some years back we were in problems because we were not enlightened,
and then we saw, we got enlightened and saw that unity is the one that can save
us…and we saw let us unite or let us build a house like this one and that other
one I was built for by the group).

Statements like the one above show that women are aware that they have been
able to ‘do’ something for themselves and the community and that this
impacts positively on their lives. The choice of the pronoun ‘ti-’ ‘we’ also
reflects the women’s sense of collective action. This collectivity is seen in the
use of action words such as ‘unite’ and ‘build’ in the expression ‘tikwatane
twake’ (let us unite and build a house). Coppock et al. (2006) in their study
on the creation and governance of women’s groups in arid Northern Kenya
underscore the importance of collective action as a basis for transformation of
women’s lives and the lives of other community members. The action women
Ndambuki & Janks P a g e | 80

take to sustain themselves and their communities demonstrates their agency

and constitutes them as the ‘rock’ that hold things together, not just
individually but collectively. Another participant put it this way:

Text (4)
Yu ukethia ni kana kaaiwa ni viisi tukombanisya inyia tukamutwaia viisi.
Nitutwaite twana tuna ivinda yiu vala tunaamukile natuyuka tuyikala nthi
tuyasya kana kaa keeititwa sukulu va na tuka tuyikata? nituukainua tunenge
inyia mbesa nethaake mbesa atwae kana sukulu’.

Now if there is a child without fees we collect it for the mother and take it to her.
We have taken four children since the time we awoke and we sit down and say
this child has been taken to this school, and let’s do what? Let us uplift him/her
and give her mother and her father fees to take her/his child to school.

From the text above, use of the pronoun ‘we’ combined with the active verb ‘-
twaite’ meaning ‘taken’ indicates the women’s active contribution to the
process of social transformation. They construct themselves as agents capable
of change.
On the whole, however, women construct themselves as lacking agency. Their
sense of their dependency on NGOs appears to erase their sense of agency.
Let us look at the text that follows;

Text (5)
Inter: Nuu uminite kuete utethyo?
Resp: Utethyo usu twanengiwe ni serikali kutukwatania na donor yakuma
Denmark, va nivo serikali yatueteie donor isu vaa na niyo yatunenge maujusi
asu.

Inter: Who has supported this group in any way?

Resp: We were given that support by the government through a donor from
Denmark. And that is when the Government brought the donor here and that is
how it (donor) has given us those skills.

In this text, the close collaboration between the government and NGOs is
shown. However unlike text 3 where agency was with the women, agency in
text 5 may be said to be with the NGO and the government and not with the
women themselves. They construct themselves as done-to’s in the use of the
pronoun ‘-tu-’ (us) in the expressions ‘yatuteie’ (it brought us) and
‘yatunenge’ (it gave us).
Prompted on the form of support, the women indicated that they have
acquired skills in various areas but ironically another problem has been
created, the problem of where to market their finished products.

4.2.1 Passive Voice

Fairclough (2003) observes that where the passive voice is used to construct
social actors, their subjugation to processes is accentuated and they are seen
as being affected by the actions of others.
81 | P a g e CADAAD

Text (6)
Syiasani kana ila tusyitaa politics aka aingi nimakothete kuimwa nafasi kati ka
kusakuwa mautongoi ukethia yu twiw’a ta ivinda yimwe twalilikana siasa inai
vau iina nthini wa nthi yaitu, mama ula wetasya kivila kya usumbi akilasya
aungama aneena munduume ula ekwiw’a ena vinya wauneenania nake atasya
ndesa unyuva mama ta musumbi, tiw’o ningi?

In politics, many women are denied an opportunity in leadership positions, like

for instance if we remember the last elections in our country, the woman who
was vying for a presidential post, whenever she stood to speak any man who
thought they had the energy to speak also stood up and said that he would never
choose a woman as a president, isn’t that true?

In text 6, the expression ‘Syiasani… aka aingi nimakothete kuimwa

nafasi’[In…politics, many women are denied an opportunity] uses the passive
voice to construct women as helpless in the face of their political
marginalization by men. All men need is the energy to open their mouths.
Rather than seeing this patriarchal discourse as a challenge, the women allow
it to shape their horizons of possibility and to prevent their participation in
politics. In the context of patriarchal discourses, women are constructed in
deficit discourses that show negative subject positions for women whereby
they are not allowed out of their traditional roles despite the fact that men
have abrogated responsibility.

4.2.2 Modality

Modality is another of the grammatical resources for interpersonal meanings

identified by Halliday (1985). In the Kamba language, modality is realized by
the term ‘ukethia’ with variants such as ‘uketha’ and ‘ketha’ all of which
generally mean ‘if it were possible’. Despite women taking an active role in
their lives, as seen in previous texts, it is interesting the way they repeatedly
represent themselves in deficit terms as seen in text 7.

Text (7)
Resp: Taketha notwona mundu eutusyaisya atakatumanthia order na muthenya
muna niw’o muukita kwitawa na kuetewe mbesa. Kau to kaindu?

Resp: Now if we had someone to get orders for us and on such and such a day
you will be brought the money. Isn’t that something?

Despite the fact that the women manage to take orders for their produce, here
the respondent laments the absence someone else to do this. This reflects a
constant refrain in their talk: ‘Who will do this for us? The choice of the words
‘tutietewe mbesa’ [we be brought money] implies that buyers should bring
money rather than that the women should sell the produce. Syntactically, the
women position themselves in object position rather than as subjects.
Further, the use of the pronoun ‘you’ constructs women as the ‘you-
community’ referencing the universal community of rural women as in deficit,
supporting Popkewitz’s views on agency. In addition, the use of the Kamba
term ‘taketha’ equivalent to the English expression ‘if it were possible’
suggests wishful thinking. In this text, the speaker positions women as
lacking in agency by her use of syntax and modality.
Ndambuki & Janks P a g e | 82

The majority of the women indicate that support from government for
education would help them to begin income generating activities and alleviate
poverty.

Text (8)
Inter: Mavata menyu kisioni kii ta aka nimo meku?
Resp: Mavata maitu kisioi kii tukwatanitye kikundi kii twakee na tuisomethya.
Inter: mm
Resp: Undu ula muneene viu wimaana kwitu ni kisomo yu ukethia nitukwata
utethyo kisioni kii, ukethia andu ala manengae andu kindu nimatuliikakana,
makatwikisya kindu (itheka).

Inter: What are your needs as women in this division?

Resp: Our needs in this division we have come together as a group to build and
educate
Inter: Mm
Resp: The greatest concern for us is education. If we could get support in our
area, if the people who give people something can remember us and give us
something (laughter).

In text 8, women identify education and building [houses] as key needs. The
use of the plural pronoun ‘tu-’ ‘we’ in the word ‘tukwatanitye’ (we have
united) indicates the collective which enables them to take action to achieve
their goals. In this text women recognize the value of education which a lot of
studies have established as a precondition for development. Further use of
the double modal ‘ukethia’ constructs their desire for education as conditional
on their getting support.
Together these extracts suggest that there is a mismatch between what the
women say and what they do. There is a disjunction in the way the women
produce a highly negative construction of themselves as unable to act without
hand-outs and what they manage to achieve. In Text 8 women refer to
themselves as ‘unremembered’, a forgotten constituency, particularly in
relation to funding. Yet they do so with laughter. Laughter sustains them.
What is interesting is that in reality the women do educate their children
without support from an unnamed source. Yet their talk reproduces them as
dependent subjects, rather than as agentive.

5. Representation of Women’s Agency by Politicians and

Other Community Leaders
The previous section has looked at the construction of women’s agency by
women in women’s groups. This section now turns to look at the
representation of women’s agency by politicians and other community leaders
in interviews. The first part focuses on pronouns while the second part
focuses on modality.

5.1 Pronouns as a Lens

On the whole, the politicians appear to perpetuate the use of deficit discourses
just like the women themselves to construct women’s agency. Text 9 provides
an example.
83 | P a g e CADAAD

Text (9)
Resp: yeah, mostly women are not very good in politics. They are not very good.
Uangalie kama mama Ndetei, (like when you look at mother Ndetei), I would
like to give you that example. Alikuja akawa mbunge (she became an MP) and
we had a lot of backing for that lady. Na siasa yake ilikuwa nzuri sana (and
politics was very good). Lakini you can be played. But politics can be played on
you.
Inter: you mean on women?
Resp: Politics are played on women, unaona (you see)… Lakini (but) women,
you see the other time there was this funny story about Ngilu. Unaona ooka
(you see she came) na ni siasa anafanyiwa, siasa (and it is politics being played
on her).

The use of the plural form of the third person pronoun ‘they’ constructs
women in generic terms as poor in politics. The claim that ‘women are not
very good in politics’ represents the feeling of most male politicians and
leaders that women are deficient in politics as a male dominated domain. The
text constructs women as ‘done-tos’ as seen in the expression ‘politics are
played on them’. The use of ‘them’ shows women as ‘non-agents’ in politics.
In other words, they are objects on which action is taken (by men who are
dominant). Similarly, the speaker constructs women as ‘done-tos’ in the use
of the pronoun ‘her’ in object position. This implies that ‘politics is played on
them’. The two examples ‘Mama Ndetei’ and ‘Ngilu’ cited in the text refer to
two women parliamentarians who were involved in scandals. The respondent
chooses to omit the scandals by referring to them as the ‘funny’ story. The two
women had allegedly been involved in sex scandals. Sexualization of scandals
involving women is common in politics and often seeks to discredit female
politicians and generally portrays them an unfit for public office.
The politicians and other community leaders have constructions similar to the
women; they see women as agentive in collective action but represent them as
having no power in the collective. For example:

Text (10)
Women are pro-unity. Yes eh women are more pro-unity than men. A good
example is majority of the self-help groups I have are of women. And you
cannot have these merry-go-rounds; in fact majority of them 99% of them are
women. It is just in few cases, it’s just a rare case that you find a merry-go-
round for men where you find a merry-go-round for men, but when we come to
trust not with men. Not with men at all at all. In fact majority of the women’s
groups, all the women’s groups in general, people feel more comfortable when
the treasurer is a woman.

In the repeated naming of ‘women’; ‘women are…women are pro-unity…’ the

leader recognizes the collective in women as seen in the merry-go-round,
however the speaker does not see women as having power in the collective.
The reason for failure by this respondent to acknowledge women’s power in
the collective might be because politicians know the danger in acknowledging
this power.
In terms of representation, women are constructed as non-agents:
Ndambuki & Janks P a g e | 84

Text (11)
I think so far ladies have not woken up to realize that they can represent
themselves. Two, ladies do suffer most of the ladies suffer most because they do
not know their rights and three, they do suffer because they are not properly
represented and four because of shyness and shyness is brought about because
of brought about by lack of education, lack of exposure, and many other aspects.
Inter: Can you please explain those points especially on representation?
Resp: One representation, let’s come down to the women’s groups.
Inter: m
Resp: when they happen to come and they want to forward their needs, they feel
more comfortable when sending a man than when sending their fellow ladies.

In the three texts above, women are constructed in a discourse of suffering

similar to that used by the women. This creates a similarity between the way
women and politicians represent women using the third person plural
pronoun ‘they’. This constructs women in generic terms, associating them
with the universal conditions of poverty and inability generally in the world.
This kind of representation denies women agency. The use of the term ‘ladies’
instead of women might also be interpreted to indicate the speaker’s attitude
that politics is a male domain which females ought to stay out of. Further use
of the first person pronoun by the speaker ‘I’ reflects an individualistic
representation of politicians as social actors. It gives them a sense of ‘power
and ownership’ over the women. The politicians produce an ‘us’ versus ‘them’
discourse, a binary opposition in which politicians are constructed as
educated, intelligent and all knowing while the women are constructed as
unintelligent (they do not know their rights), uneducated (they suffer due to
lack of education) and poor (in need). All these discourses maintain and
reproduce deficit discourses that continue to present women as ‘non-agentive’
in the political process.

5.2 Active and Passive Voice

The findings of the study indicate that in their use of voice, the politicians on
the one hand represent themselves as all-knowing; they are the ones who
know and have the answers for the women. They are also dismissive of
people’s ability to think things out for themselves. As a result they represent
women as powerless, illiterate and ignorant about their own issues. This
denies women agency and yet recognition and voice are values that women
need to have the confidence to articulate and represent their concerns. One
leader said:

Text (12)
Resp: makitaa kwaiwa undu mekwika yu tayu mikopo tunengawe kuu no Kenya
Women Finance Trust, na syindu mbingi, kuo tucompanies twingi tuunengane
mikovo. Na mikovo inu ve iveti itonya kwosa ikeka nesa muno ikaola mathina
maingi muno ma musyi. Ona thina uu wa ukua ndoo kilometer na kindu kuite
Emali akaola na akekia kawila vu soko ila vu thome kwake, lakini utithiwa
ndaaallowiwa ni muume kwosa mukopo. Ewa ‘osa mukopo, mikopo itosawa ni
ak’a, utithia nituuvinyikikia vau.

They only do not know what to do because the loans we are given here by the
Kenya Women Finance Trust and many other things, like there are other
companies giving loans. And with these loans, there are women who can get
85 | P a g e CADAAD

them and do very well and reduce many problems in the home. Even reduce
this problem of carrying a bucket for about a kilometer and a half to Emali when
it is raining and instead start a business nearby at the market centre. But she
cannot be allowed to get the loan by her husband. If she is told to get a loan,
‘loans are not gotten by women’ we are pressed there.

Politicians represent women as constrained by patriarchal discourses as seen

in the expression ‘she cannot be allowed to get a loan by her husband’. In
other words, through passivization women are subjugated to patriarchy which
limits their action as social actors. These patriarchal discourses portray
women as enslaved by patriarchy as evidenced in the words ‘loans are not
gotten by women’ implying that ‘loans are got by men’. The current research
also confirms similar hegemonic discourses of ‘development’ which
stereotypically present women as poor, powerless, backwards, illiterate,
suffering; a discourse that constructs the Third World generally in similar
ways (Verma 2001). These discourses continue to portray women in deficit
terms which in turn lead to women being seen as deficient and therefore
unable to take transformative social action.

5.3 Modality

In general, both the women, politicians and community leaders construct

women’s condition as one of possibility in their use of modality. Both the
leaders and the women agree to a large extent that women have been
overburdened by household duties, a finding supported strongly in the
literature. The following are some of the responses from the leaders;

Text (13)
Leader 1: Kila kimavinyiie muno, mm generally andu aka mavinyiiawa ni
syindu sya musyi. Mundu muka niukaa ukethia ona nutonya uthukuma
noyithia ndenawia museo wa kuthukuma.

What is pressing them much is generally things to do with the home house hold
chores, it is as if a woman might do work (household chore but may not have a
good job)

Leader 5: men look at it as if it is a very dirty thing, so most of the women are
now taking up that responsibility.

The leaders appear to be very much in agreement with the women that women
have truly been overburdened by household chores. In the use of the modal
‘ukethia’ leader 1 constructs women’s condition in the possible mode and yet
in reality women are the ones who take the larger proportion of household
chores. The leader in the first part of the utterance acknowledges women as
being overburdened by household chores but does not do so in a firm way in
the second part of the same utterance where he goes on to say that ‘it is as if a
woman might do household work’ yet in reality it is the women in the Kamba
community who do most of the household chores. Women are constructed as
dependent and responsible for household chores such as gathering firewood,
fetching water and farming which they rely more heavily on. In other words,
women are constructed within a patriarchal traditional discourse of women’s
responsibility.
Ndambuki & Janks P a g e | 86

The second utterance by leader 5 might be interpreted as resistance by the

men in view of the fact that in terms of division of labour, the practice of child
rearing was mainly a responsibility for the women in the Kamba community
and continues to be so. Hence the use of the expression that ‘most of the
women are now taking up that responsibility’ constructs women as just
starting to recognize their work in household chores, yet they have done do
since time immemorial. In view of changing conditions of production, this
study recommends focus on this area to enlighten both genders about the
importance of involvement in child upbringing for the good of the community.
This is because this practice appears to be unique to the Kamba community as
one of the participants who is from a different rural community expressed
surprise that men in this community are largely uninvolved in household
chores.
Modality also features in the construction of false promises made by
politicians during elections which was identified as a key issue especially by
the women community leaders such as this one:

Text (14)
Lakini nitwithaa yu ta ivinda yii tweteelle campaingn. Nimokaa makatwia
nimeututetheesya, natuilea kwisa kwona undu meutwika. Maikita uneena tu.
Yii mekaa uneena na munyuka.

But like at this time when we are waiting for the campaign. They come and tell
us that they are going to support us but we never see what they do for us. They
just talk. Yes they just talk with their mouths.

The choice of the term ‘mouth’ in the expression ‘they just talk with their
mouths’ implies a lack of commitment on the part of the politicians.
Politicians’ use of language for coercion has been widely researched. Findings
by political discourse analysts have established that politicians use persuasion
as a strategy to make the electorate to vote for them (Chilton and Schäffner
1997; Schäffner 1996)). In the Kenyan context, voters characterize this
persuasion in terms of ‘false promises’. Majority of the politicians always
promise that if they are voted for, they will provide water and roads to the
people, but in reality, the roads in the study area remain some of the worst in
Kenya.

5.4 A Frequency Analysis of Pronoun and Modality Choices

A frequency analysis of pronouns and modals was done to give an indication

of women’s sense of agency. This was done in relation to the personal
pronouns, ‘nyie’ (I), ‘ithyi’ (we) and the possessive pronouns ‘syitu, maitu’
both variants meaning (our). In addition, analysis of the modals ‘ukethia’
with the variants ‘kethia and ketha’ which express modality and may be
glossed to indicate ‘possibility or if it were possible’ was also done. All these
are captured in the table 1 below.
87 | P a g e CADAAD

Table 1: A frequency count of pronoun and modality choices by women in the corpus:

Pronoun Number of
occurrences
Nyie (I) 65
Ithyi (we) 103
Maitu/syitu (our) 99
Modal: ‘Ukethia’ 271
(modal for perhaps)
Total 538

5.4.1 Pronouns

From table 1, the pronoun ‘nyie’ (I) is the least commonly used pronoun which
might be interpreted to mean that there are fewer representations of women
as individual actors. ‘We’ (ithyi)’ is used more frequently than any of the other
pronouns (more than double). This use of ‘we’ indicates that women construct
themselves as a collective. All their achievements are based on mutual
support within a collective yet they think the answer to agency lies in an
individual. In other words, they fail to recognize the power the collective
represents. In the table above, the use of the possessive pronoun variants
‘syitu’ and ‘maitu’ (our) indicates a remarkable sense of agency among the
women particularly in terms of ‘labour ownership’.
From the table above, modality constitutes the highest usage of the terms
selected for analysis in the corpus. This explains why the women appear to
attribute agency away from the self. In other words, the women presented
their issues as if they were helpless and not in control and yet in practice they
did a lot of things like provision of child care, food and health especially as
caretakers in relation to HIV. Most of the respondents describe their action in
the possible mode even when something has actually happened and in this
way appear to limit their sense of action.

6. Construction of Women’s Agency by Politicians and

Other Community Leaders
As already seen in the section on community leaders’ construction of women’s
agency, the leaders’ construction of women’s agency is in certain ways similar
to that to the women in women’s groups. In certain ways however, their
construction differs remarkably from that of the women themselves. While
the women use words that show them as lacking agency despite being very
agentive in the actions they take in the community, the leaders construct
women’s agency in very overt terms as ‘lacking’. The leaders express concern
over the women’s lack of agency. In order to get an in depth understanding of
this construction, this section provides a quantitative analysis of the leaders’
construction of women’s agency. As done at the beginning of this paper where
pronouns and modality are analyzed to give a sense of agency, a similar
Ndambuki & Janks P a g e | 88

approach is used in this section. The table that follows contains the number of
occurrences and their totals in the corpus for the personal pronouns ‘nyie’ (I),
plural ‘ithyi’ (we) and the possessive pronoun realized by the Kamba variants
‘maitu/syitu’ ‘our’. These pronouns were chosen because use of pronouns is
one of the most important ways through which agency is exercised. Lastly the
modal ‘ukethia’ (if) is also analyzed in order to compare the leaders’ and the
women’s construction of agency. As already noted in the previous section on
women’s construction of agency, an analysis of modality is useful as it is an
important part of how people identify themselves; the question of what people
commit themselves to when they make statements, ask questions make
demands or offers is crucial to the construction of identity.
Table 2: A frequency analysis of pronoun and modality choices by politicians and other
community leaders in the corpus

Pronoun Number of
occurrences
Nyie (I) 126
Ithyi (we) 14
Maitu/syitu (our) 29
Modal ‘Ukethia’ 161
(if it were
possible)
Total 330

From table 2, the politicians’ and other leaders’ construction of women’s

agency indicates a very similar trend to that of the women in focus group
discussions especially in relation to the high frequency of occurrence for the
term ‘ukethia’. This might indicate that both categories perpetuate the use of
deficit discourses in the representation of women’s issues. They both describe
women’s issues in the possible mode with no likelihood of transforming
women’s conditions of possibility into reality.

6.1 The Pronouns

Nyie (I)
Analysis of the personal pronoun nyie (I) indicates a much higher frequency
for the leaders than that for the women in the focus group discussions. This is
not surprising and may be interpreted to mean that the leaders have a greater
sense of agency as individual actors in the political process as compared to the
women. The possible explanation for this is also that politicians have greater
access to resources as they are some of the best paid civil servants in Kenya. It
might also be seen as an indication of their authority as leaders which gives
them a greater platform for exercising power over the grassroots women.
89 | P a g e CADAAD

Ithyi: We
The frequency of occurrence for the use of the plural pronoun ‘ithyi’ us
indicates a minimal sense of ‘collectiveness’ with most of instances of the
pronoun being used by the female leaders. This confirms studies on gender
differences between men and women that indicate women have a greater
sense of the collective than men. This pronoun is remarkably high among the
women in the focus group discussions and the reason for this is the fact that
women groups are essentially female.

Ukethia (Ethiwa) ‘if it were possible’:

In terms of modality, the frequency of the modal ‘ethiwa’ ‘if it were possible’ is
low and the possible explanation for this is the certainty with which the
leaders speak compared to the women. Unlike the women who are
constructed in terms of deficit discourses, the politicians and other leaders
construct themselves as contented actors who are in command in the political
process.

6.2 Contradiction in Overall Construction of Agency

Table 3 provides a summary of the activities that each group engages in.
Table 3: Activities in women’s groups

Group No. of Hours Total hrs Activities

membe of tape and
rs minutes
Group 1 13 45 mins 1 hr 10 Merry-go-round; farming; cultural
mins dance
Group 8 45 mins 1 hr 20 Merry-go-round; farming
2 mins
Group 12 45 mins 1hr 5 mins Kerosene selling; poultry keeping
3
Group 12 50 mins 1 hr 30 Merry-go-round; goat keeping;
4 mins horticultural farming

Source: author’s fieldwork

Table 3 shows the various economic activities that the women’s groups engage
in. Essentially all the women groups practiced small-scale farming, mainly
growing maize and beans. Two groups were involved in house construction,
while one group was involved in adult education and another was a dance
group. All groups except one practiced the ‘merry-go-round’. The term has
come to be used to refer to the rotational practice of contributing money or
labor for one another in turns within the group. The term is seen in practices
like the ‘myethya’ (working group meetings) where people come together to
Ndambuki & Janks P a g e | 90

assist one another in times of need as in everyday activities like crop

harvesting, house construction and organization of various social functions
such as marriage ceremonies, burials, fundraising and the establishment of
social networks. If any member has a problem the group takes it up
collectively.
It is clear that the women’s groups are all engaged in local action to sustain
their families collectively. All but one of the groups is involved in mutual
support among its members. In other words there is clear evidence that at the
level of action these women assume agency to survive and improve their local
condition. Despite this, in their talk, the women represent themselves as
having no agency, hence a contradiction in their construction of agency.

7. Conclusion
The study has shown CDA as one way of understanding how prevailing
discourses impact on the participation of women in the political process in
Kenya. The women appear to attribute agency away from the self. The
women presented their issues as if they were helpless and not in control and
yet in practice they did a lot of things like provision of child care, food and
health especially as caretakers in relation to the HIV pandemic. Most of the
respondents describe their action in the possible mode even when something
has actually happened and in this way appear to limit their sense of action.
The paper reveals that politicians and other leaders in the community
continue to perpetuate the use of deficit discourses in their construction of
women’s issues. Women are represented against a backdrop of discourses of
patriarchy, rurality and poverty that construct them as poor, ignorant and
illiterate; constructions which seem to perpetuate unequal power relations
between men and women in society. However, while both women and leaders
construct women’s agency within deficit discourses, these discourses do not
match either women’s enacted practices, what political and community
leaders say they expect of women. The contradiction inherent in the study is
that everyone constructs women as lacking in agency, yet these women act as
agentive subjects. They also point to a need for the expansion of the
scholarship on gender politics in the African Diaspora and Kenya in
particular.

Notes
1 Transcription Conventions:
( ) To give additional information
… To indicate pauses
, To indicate hesitations
. To indicate end of statement
? To indicate questions
_ To indicate a word cut in delivery
91 | P a g e CADAAD

References
Adhiambo-Oduol, J. (2001). The Socio-cultural Aspects of the Gender Question. Nairobi:
Constitution of Kenya Review Commission (CKRC).
Amadiume, I. (1987). Male Daughters, Female Husbands: Gender and Sex in an African
Society. London: Zed Books Ltd.
Blommaert, J. (2005). Discourse: A Critical Introduction. Cambridge: Cambridge University
Press.
Bourdieu, P. (1990). The Logic of Practice. Cambridge: Polity Press.
Bourdieu, P. (1991). Language and Symbolic Power. Ed. by J.B. Thomson. Oxford: Polity
Press.
Brown, G. and Yule, G. (1983). Introduction to Discourse Analysis. Cambridge: Cambridge
University Press.
Chilton, P. and Schäffner, C. (1997). Discourse and politics. In T.A. van Dijk (ed.), Discourse
as Social Interaction. Discourse Studies: A Multi Disciplinary Introduction. London:
Sage. pp. 206- 230.
Chitere, P.O. (1988). The women’s self-help movement in Kenya: A historical perspective,
1940 – 1980. Transformational Journal of History (17): 50-68.
Coppock. D.L., Desta, S., Wako, A., Aden, I., Gebru, G., Tezera, S. and Tadecha, C. (2006).
Women Groups in Arid Northern Kenya: Origins, Governance and Roles in Poverty
Reduction. Conference documentation: Available Online at
http://www.ilri/research/theme1/
Electoral Commission of Kenya (ECK, 2002) Registered Voters per Polling Station (Samuel
Kivuitu, Chairman). Nairobi: ECK.
Fairclough, N. (1989). Language and Power. London: Longman.
Fairclough, N. (1995). Critical Discourse Analysis: The Critical Study of Language. London:
Longman.
Fairclough, N. (2003). Analysing Discourse: Textual Analysis for Social Research. London:
Routledge.
Khasiani, S.A. (2001) Enhancing women’s participation in fovernance: The case of Kakamega
and Makueni districts, Kenya. In E.M. Rathgeber and E.O. Adera (eds.), Gender and the
Information Revolution in Africa. Canada: International Research Center.
Kanogo, T. (2005). African Womanhood in Colonial Kenya 1900-50. Oxford: James Currey.
Lewa, R.M. (2002). Capacity Building Programmes (CBPs) Amongst Women Groups in
Mombasa District Kenya. Unpublished MA thesis. Nairobi: University of Nairobi.
Morrison, A. and Love A. (1996). A Discourse of disillusionment: Letters to the Editor in two
Zimbabwean Magazines 10 years after independence. Discourse & Society 7 (1) : 39-79.
Ndambuki, J.M. (2006). Foregrounding women’s agency in Africa’s democratization process.
CODESRIA Bulletin. Special Issue on The African Woman 1-2: 17-19.
Popkewitz, T.S. (2007) Cosmopolitanism and the Age of School Reform: Science, Education
and Making Society by Making the Child. London: Routledge.
Schäffner, C. (1996). Editorial: Political speeches and discourse analysis. Current Issues in
Language and Society 3 (3): 201-204.
van Dijk, T. A. (1997). Discourse as interaction in society. In T.A. van Dijk (ed.), Discourse as
social interaction. Discourse studies: A multidisciplinary introduction, vol. 2. London:
Sage. pp. 1-37.
Verma, R. (2001). Gender and Livelihoods in East Africa: Through Farmers’ Eyes. Ottawa:
International Development Centre.
Ndambuki & Janks P a g e | 92

Whiteley, W.H. and Muli, G. (1962). A Practical Introduction to Kamba. London: Oxford
University Press.
Wodak, R. (1997). Some important issues in the research of gender and discourse. In R.
Wodak (ed.), Gender and Discourse. London: Sage. pp. 1-20
Wodak, R. (2001). What is CDA about: A summary of its history, important concepts and its
developments. In R. Wodak and M. Meyer (eds.), Methods of Critical Discourse Analysis.
London: Sage. pp. 1-13.
Zubair, S. (2003). Literacies across generations: Women’s religious and secular identities in
Siraiki villages. South East Asia Research 23 (2): 135-151.