Content uploaded by Benjamin Piper
Author content
All content in this area was uploaded by Benjamin Piper on Feb 21, 2016
Content may be subject to copyright.
This article was downloaded by: [Benjamin Piper]
On: 13 November 2013, At: 12:33
Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Current Issues in Language Planning
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/rclp20
Whatever the law says: language
policy implementation and early-grade
literacy achievement in Kenya
Barbara Trudella & Benjamin Piperb
a Director of Research and Advocacy, SIL Africa Area, Nairobi,
Kenya
b Chief of Party, Primary Math and Reading (PRIMR) Kenya, RTI
International, Nairobi, Kenya
Published online: 11 Nov 2013.
To cite this article: Barbara Trudell & Benjamin Piper , Current Issues in Language Planning (2013):
Whatever the law says: language policy implementation and early-grade literacy achievement in
Kenya, Current Issues in Language Planning, DOI: 10.1080/14664208.2013.856985
To link to this article: http://dx.doi.org/10.1080/14664208.2013.856985
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
Whatever the law says: language policy implementation and
early-grade literacy achievement in Kenya
Barbara Trudell
a
*and Benjamin Piper
b
a
Director of Research and Advocacy, SIL Africa Area, Nairobi, Kenya;
b
Chief of Party, Primary
Math and Reading (PRIMR) Kenya, RTI International, Nairobi, Kenya
(Received 1 December 2012; final version received 16 October 2013)
Language policy is generally seen as a national-level decision regarding which
languages the state will support, and in which public domains. However, the reality is
that language policy plays out at regional and local levels as well. In fact, it could be
argued that the most important instantiations of language policy are those which
directly determine local-language behaviors in institutions such as schools,
government and civil society. Using data drawn from Kenya, this article examines the
formulation and implementation of language policy as it plays out in the primary
classroom environment. The relationships between language policy implementation at
the classroom level and students’early literacy outcomes are explored, giving insight
into how the degree of adherence to language policy in the classroom intersects with
student achievement. The article presents findings using language use as a predictor,
school and student-level economic status as control variables and student
achievement as the outcomes. The country-level differences in language policy
implementation between Kenya and Uganda, and the impact of those differences on
student achievement in the two countries, are also examined. The article has
implications for the establishment of a learning environment in the multi-language
primary classroom, and demonstrates the extent to which choices about language
policy implementation can present a serious challenge to effective education.
Keywords: Kenya; language policy; primary school; literacy; mother-tongue education;
African languages; medium of instruction; reading assessment; multilingual education
Introduction
For proponents of mother-tongue-based education, the disparities between language
policy and practice are grounds for endless conjecture and –often –profound frustration.
Certainly the formulation of national language policy itself presents many challenges,
particularly in highly multilingual nations: the difficult politics of choosing one or
more national languages for special status at the expense of others; the implications of
choice for participation in international arenas of education, governance and commerce;
and the uncertainty of being able to carry through on policy aims relating to equal treat-
ment for all citizens.
However, beyond these challenges of national-level policy formulation, issues of local
response to such policy are also complex. Local appropriation of national language policy
tends to reflect the language attitudes and goals of the population of speakers, which may be
© 2013 Taylor & Francis
*Corresponding author. Email: barbara_trudell@sil.org
Current Issues in Language Planning, 2013
http://dx.doi.org/10.1080/14664208.2013.856985
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
either compliant or antagonistic to stated national policy goals; the attitudes and priorities of
headmasters, teachers and parents about language and education generally hold the decid-
ing influence on school language practices. Thus, where national policy upholds a child’s
right to education in a language the child speaks, the reality in classrooms may be otherwise
if local decision-makers prefer the use of other languages. On the other hand, where
national policy precludes education in the mother tongue, resourceful teachers and
parents may collude to make such education possible despite the national policy
environment.
This interplay of national-level statements on the direction of the country and local
appropriation of those statements to fit local priorities is at the crux of the implementation
challenge where language policy is concerned. This implementation challenge has in some
cases all but nullified national language use goals, and understanding the broader context of
policy enactment is key to the successful development and implementation of language
policy.
This article examines this multi-faceted interaction of policy and implementation where
language of instruction is concerned, focusing on data gathered in Kenya. The nature of
national language policy formation is first examined as well as the ways in which local
language use decisions either comply or subvert that policy, particularly as played out in
the formal education system. A case study of primary-school classrooms in two regions
of Kenya then provides a quantitative examination of language use behaviors in these class-
rooms. A particularly useful feature of this study lies in the quantitative evidence it brings to
the discussion of these issues. In an environment where the stakeholders may or may not
openly discuss their motivations, values and practices where language is concerned, the
quantitative description of classroom language behaviors and competencies provides evi-
dence that is difficult to ignore.
Language policy and the nation-state
Language policy may be described as the set of principles formulated and legally estab-
lished by the state, intended to guide language use particularly in public domains. Other
uses of the term include decisions about language use as made by local actors, not only
those made at the national level (Zhao & Baldauf, 2012). The dynamic, informal, negotiated
side of language policy is described by Spolsky (2004) as growing out of people’sown
practices, ideologies and beliefs where language is concerned.
Language policy development is actually just one of several intentional activities that
have a bearing on the use and development of languages. Language planning aims at
modifying the status of a particular language in society as well as the range of
domains in which it is used; this is usually a state-driven activity, although local-language
activist organizations may also carry it out (Trudell, 2010). Language development aims
at increased use of a given language, and involves increasing the domains of use of the
language (Haugen, 1966) through activities such as linguistic analysis, orthography and
vocabulary development, production of written texts, establishment of opportunities to
become literate in the language and so on. Language ideology is described by Spolsky
(2004) as the consensus of a speech community regarding language, language use and
language practice. Woolard and Schieffelin (1994) note that language ideologies are
not just about language; they link to group and individual identity, and to social insti-
tutions as well. In the school context, language ideology plays out in the choices made
by teachers, education authorities and parents regarding the language(s) they want to
see used in the classroom.
2B. Trudell and B. Piper
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
The language policy of a nation is much more than simply an official statement of pre-
ference for the use of one language over another. Language choice is a fundamental means
by which the state defines itself in contrast to other nations and peoples. The development
of this crucial role of language policy coincided with the emergence of the nation-state in
the late nineteenth century. Shohamy (2006, p. 26) observes that
the establishment of the nation-state at the end of the 19
th
century started the association of
nations with language. …Thus the use of language in the service of national ideology is rela-
tively new, as the modern nation-state is a relatively recent phenomenon. It was especially then
that language began to be symbolized with power, ideology, nationalism, loyalty, patriotism
and the drive for assimilation. (May, 2001)
In this context, the national language becomes a symbol of belonging, an identifier of both
inclusion and exclusion where the nation is concerned.
Those languages chosen as official national languages take on a political significance
that is largely unrelated to the number of their native speakers or the demographic range
of their use in the nation. National languages are associated with modernity and progress
(May, 2008), and with specific national goals and identity. One strong example of this
association of a national language with national political goals and identity is the establish-
ment of kiSwahili as the official national language of post-independence Tanzania. The
policy of ‘Swahilisation’(Wright, 2004, p. 79), in which the language became a national
symbol of the state’s ambitious social and political goals, successfully facilitated the
nation-building agenda of the early government for about 20 years.
However, Tanzania’s choice of an African language as the primary communicative rep-
resentation of national goals and aspirations was not the norm across Africa. As African
nations gained independence from the colonial powers, they generally officialized the colo-
nial language for use in governance, commerce and education. The reasons behind this
choice were described at the time as pragmatic, nationalistic and economy-driven
(Muthwii, 2007); nevertheless, given the significant role of the official language in
shaping and expressing national identity, a policy which legitimized expression of the pro-
foundly important cultural and linguistic character of the nation would surely have served
these nations better.
The case of Rwanda provides additional insight into the links between national identity
and language policy. Samuelson and Freedman (2010) argue that the evolution of Rwanda’s
language policy since 1996 is playing a key role in social reconstruction in that country.
Describing Rwanda’s new policy of 2008, in which English replaces French as the official
language of the nation and as the medium of instruction throughout school, the authors
note:
The policy-makers frame the change as a major factor in the success of social and education
reforms aimed at promoting reconciliation and peace and increasing Rwanda’s participation
in global economic development. (p. 191)
Describing the political context in which this policy decision was made, the authors note
that this move was
completing a process that had been started 14 years earlier when the victorious Rwandan
Patriotic Front (RPF) added English as an official language in 1996, soon after taking power
in the wake of the 1994 wars and genocide. (p. 192)
Current Issues in Language Planning 3
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
In this case, one ex-colonial language has been replaced by another, clearly signaling the
state’s intention to move into the English-dominated global community –and specifically
the East African Community, which Rwanda joined in 2007. The fact that Kinyarwanda is
spoken by the entire population, and as a mother tongue by most, raises the question
whether a national identity could be conceived of that is shaped by indigenous language
and culture rather than European and American language; however, the history of inter-
ethnic genocide in Rwanda perhaps nullifies that possibility.
In any case, when a nation’s language policy gives precedence to one particular
language, the languages which are not chosen for such a prestigious role in shaping national
identity come to be seen as non-modern, non-progressive and even obstacles to the political
project of nation-building (May, 2008). They are perceived as a threat to the unity of the
state. In the African context, the result has been to marginalize the hundreds of local and
regional languages which serve as the primary means of communication and identity for
millions of citizens. For example, Heugh (2008) documents the deleterious effects which
the advancement of Setswana in Botswana has had on perceptions of smaller Botswanan
languages.
Thus, the language policy of a given nation is far from randomly chosen; the choice of
national language is freighted with the identity and aims of the state which has made that
choice. Where significant and effective changes have been made to existing national
language policy in Africa, they have been linked to major political change as well: the Tan-
zanian example above, for example, or the case of the Transitional Government of Ethiopia,
which in 1991 took power and set in motion a range of initiatives –including a strong pro-
local language policy which was embedded in the Transitional Education Policy (Federal
Democratic Republic of Ethiopia, 1994).
Given the significance of language choice at the crucial stages of nation-building, it is
important to consider the nature of later changes in language policy, changes which are not
linked to a major shift in national identity. Since 1990, most African states have committed
themselves to greater use of African language in educational contexts to one degree or
another (Heugh, 2008). Kenya’s decision (taken a bit earlier, in 1976) to permit the use
of local languages in early primary school is one example (Kenya Institute of Education,
1992). Another is the Cameroonian government’s move in the 1990s toward legitimizing
the use of local languages as media of instruction; the Constitution of 1996 made reference
to the importance of national languages (Tadadjeu, 1997, p. 23), and in 1998 the new Law of
Education in Cameroon listed ‘promote [Cameroonian] national languages’as one of its
nine educational objectives (Cameroonian Ministry of National Education, 1998).
One of the most striking features of these policies that make significant space for local
African languages is how easily ignored, easily reversed and poorly implemented they are
(Gove & Cvelich, 2010; Muthwii, 2007). Recent changes in Ghanaian language and edu-
cation policy provide a good example: a poorly implemented policy permitting the use of
nine Ghanaian languages in early primary grades was replaced in 2002 with a policy requir-
ing the use of English as the medium of instruction throughout school (Government of
Ghana, 2002; Opoku-Amankwa, 2009). Interestingly, this decision was again reversed in
2009 after protest from academics, educators, traditional rulers and others. However,
implementation of the earlier policy remains weakly supported.
This characteristically weak or non-implementation of pro-local language policy by the
national authorities is a constant hindrance to the development of African languages for use
in learning and communication. And indeed, Shohamy (2010) raises the question of
whether these language policies
4B. Trudell and B. Piper
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
are in fact meant to be implemented, or if they only serve bureaucrats as ideological statements
and evidence of action and intentions without serious concern or their actual feasibility or
meaningful implementation. (p. 182)
It appears that if such policy is not formulated as an integral component of national identity
and aims, it is not likely to be strong enough to withstand the pressures of school failure, the
chimera of economic prosperity through international languages and the forces of
globalization.
National and local-language policies in conflict
National-level policy formulation is not the only, and is perhaps not even the most powerful,
iteration of language attitudes and values. In fact, understanding patterns of local language
use requires recognition of who has the power to influence local-language practices. Par-
ticularly where the state is not exceptionally strong, national language policy is played
out according to the values and beliefs of those local actors with influence, power and/or
language expertise (Trudell, 2007; Zhao & Baldauf, 2012, p. 6). Given Spolsky’sdefinition
of language policy as including local practices and decisions (referenced above), it is
clearly important to consider the people and modalities involved in local acceptance of –
or resistance to –national language and education policy decisions. Language policy
implementation is ultimately a series of decisions by a range of influential stakeholders
regarding which language they will speak in which context. Those stakeholders are
found in multiple sites in society; they include teachers, headmasters, local government
authorities, parents, community leaders and the students themselves (Mutaka & Attia,
2008; Yigezu, 2010).
Language use decisions reflect a range of responses to the tension between the global
attraction and instrumental value of international languages on one hand and the strength
of local cultural identity on the other. The equilibrium between these two forces is not
always predictable; but where school is concerned, language use decisions tend to be
highly pragmatic. Laitin (1993) found that Ghanaian parents who themselves used their
local language almost exclusively ‘could not imagine a world where their children would
not learn through the English medium’(p. 229).
Given these multiple sites and motivations for language policy formation and appro-
priation, conflict between the values and goals of national and local education stakeholders
readily leads to inconsistent implementation of national policy. At the most extreme, local
stakeholders may employ an actively subversive approach to policy, bending or even break-
ing it to serve their own educational goals.
Where language and education policy tends to be restrictive, as in the USA or France,
such subversive appropriation of language policy can be seen in local classroom initiatives
to introduce students’first languages into instructional practice, in order to improve student
learning. These initiatives may be formally sanctioned at the local level, or they may take
place in more informal ways. As an example of the latter, Hélot (2010) describes a Turkish-
speaking student teacher in an early primary classroom in France, a country in which
French has been institutionalized as the language of education (Hélot, 2010). This
student teacher, although he ‘had not planned at first to use languages other than French’
(p. 60), found that this was not an adequate medium of instruction for some of his
pupils. His decision to use the Turkish language to help two of his Turkish students was
unpremeditated; its application was informal and did not extend to any class-wide activities.
Hélot comments:
Current Issues in Language Planning 5
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
When I asked him why he did not use Turkish with [the Turkish student] from the outset, my
question took him by surprise. He had not been aware he had internalized a covert policy for-
bidding the use of the home language. (p. 61)
A case of more intentional appropriation of restrictive language policy is described by
Johnson and Freeman (2010), where school district staff in Pennsylvania, USA, chose to
interpret federal education laws in ways that would permit them to carry out a dual-
language education initiative in their district. The stakeholders in this case played on the
ambiguity and debate surrounding the law to make a space for Spanish–English dual
language programs in the early elementary grades.
However, the norm in Africa is that language policy which prioritizes the use of inter-
national languages of instruction is typically honored at the local level (Ouane & Glanz,
2010)–despite the fact that huge percentages of the children in the classrooms do not
control those languages well enough to use them in learning curricular content. In these
cases, local beliefs about the appropriateness of African and international languages for
education align with national policy regarding language of instruction (Trudell, 2007)to
make the hegemonic position of the former colonial languages virtually unassailable.
The reality of extensive code switching behaviors in Kenyan primary schools (see for
example, Nyaga & Anthonissen, 2012) does not detract from the ideological power of
English in the national education system.
Thus, it is not surprising that the most common subversion of national language policy
in African contexts occurs where the policy mandates the use of locally spoken languages as
media of instruction in the primary grades. As noted above, a number of African nations
currently have such policies in place, generally indicating that local languages (or the
language of the catchment area) may be used up through grade 3 of primary school. In
other cases the national policy is less definite, but leaves space for the use of local languages
at some level in early primary education.
However, these pro-African language education policy choices are nearly always con-
tested –on the grounds that the international language is more advantageous to the students,
or on the grounds that the African languages chosen for use do not reflect the stakeholders’
own local language preferences.
In Kenya, where the policy mandates use of local-language instruction in early grades
and where only an estimated 15% of the citizens speak English fluently, Kembo and
Ogechi’s(2009) study of language use in the classroom found that ‘English still enjoys
solid support among Kenyans, and this support is not restricted to the middle-class
sector of the population but cuts across all socio-economic classes’(p. 169). Where the
language of the students is used in the African classroom, this is generally done on a
very informal basis, and takes the shape of code switching and mixing. In some contexts,
teachers are taught that to resort to such informal use of the pupils’language in order to
convey information is a sign of poor teaching (Trudell, 2004).
The shape this contestation takes depends on the level at which the opposition is found.
Political leaders may not provide local-language policies with the needed supporting legis-
lation. Education authorities may neglect to facilitate the teacher training and materials pro-
vision that would support local-language medium; or they may assign teachers to parts of
the country where the teacher does not speak the local language. Teachers and headmasters
may choose not to use the pedagogical methods associated with the local-language medium
–though it is also the case that pro-local language policies may be nearly impossible to
implement at the classroom level due to the lack of training, materials and teacher
support. Parents may withdraw their children from the offending schools, or send the
6B. Trudell and B. Piper
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
children to schools in a different part of the country where the policy implementation is
more congenial to their own language values (Ambatchew, 2010).
However, by far the most common form of non-compliance with local-language
medium policy is for the teacher simply to ignore the local language in the classroom,
focusing on maximum use of the international language. They do this, generally with
full parent support, based on two mutually reinforcing beliefs: (1) that African languages
are not adequate means for success in formal education and (2) that the most effective
way for the students to learn the international language is by means of ‘submersion’
(Baker, 2006, p. 216) –complete immersion in with no support provided in, or to, the
language that the students speak (Sprenger-Tasch, 2003, p. 362). This two-pronged
language ideology is so deeply ingrained in African societies that only the most strictly
enforced language policies can dislodge it.
Applying this language ideology in the classroom leads to pedagogical disaster for
millions of African children. Describing the impact of French-medium instruction on stu-
dents in Burkina Faso, Mali and Niger, Alidou (2003, p. 17) observes:
The majority of students in both urban and rural schools come from poor, non-literate
families where French is not the means of daily communication. The education experience
of this group of students is characterized by class repetition and a high drop-out rate
between fourth and sixth grades …In francophone Africa, the majority of primary school stu-
dents experience exclusion in the classroom. Owing to a lack of proficiency in French, they are
silenced and spend most of their time listening to the teacher and the very few students who can
speak French.
Diarra (2003), writing of classrooms in Angola, describes a similar situation:
No in-depth study is required to see that the indifferent results generally obtained in teaching in
Portuguese [in Angolan schools] are due in very large measure either to an inadequate
command of the language on the part of both teachers and pupils, or to confusion between
it and the national languages, from the phonetic level to that of syntax and semantics. As a
result, the rate of learning is slowed down and the quality of teaching is noticeably affected.
(pp. 340–341)
In Kenya, Kembo and Ogechi’s(2009) study of language use in primary classrooms docu-
ments the failure of the prevailing English-focused ideology to facilitate learning in science
and mathematics. Even in schools where the local language was supposedly the medium of
instruction through grade 3, English was heavily used throughout the primary grades. The
authors note that
the transit of children from mother tongue to English medium is premature, thereby denying
them the opportunity to develop cognitive/intellectual skills in their first language, which
they would later transfer to English. (p. 170)
The authors also observed serious deficiencies in the teachers’mastery of science and math
concepts, as well as English proficiency that inhibited their ability to teach well in the
language.
Thus, national policy favoring local language use as medium of instruction is often
appropriated in ways that nullify the pedagogical and cultural aims embedded in the
policy. In these cases, the economic and logistical challenges to implementing such
policies are relatively minor compared to the huge challenge posed by a widespread
language ideology that rules out African language use in the classroom in favor of an
Current Issues in Language Planning 7
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
international language which the students rarely speak. The impact of this language ideol-
ogy on literacy achievement in Kenyan classrooms is the focus of the remainder of this
study.
A Kenyan case study of language policy implementation in the classroom
This case study of language use practices in 100 Kenyan classrooms focuses on three
related questions, the answers to which provide a picture of the language-of-instruction
scenario in Kenyan schools.
.Do language-of-instruction practices in the classroom adhere to the national language
policy of Kenya?
.How is language used in the primary classroom?
.How does classroom practice relate to early grade assessment of students’reading
ability?
Kenyan language policy calls for the use of the mother tongue or ‘the language of the
catchment area’in the first three grades of primary school (Kenya Institute of Education,
1992, p. 143). The languages not used as medium of instruction are meant to be taught
as subjects. In the two regions where this study was carried out, implementation of this
policy would have featured the use of either Gikuyu or Dholuo as the medium of instruction
in lower primary in the rural schools studied, and the use of Kiswahili in this role in the
urban schools.
The Kenyan study took place in 2009–2010, when Research Triangle Institute (RTI)
International undertook a large early grade reading assessment (EGRA) study focusing
on language of instruction in Nyanza and Central Provinces. This study was organized
with a primary emphasis on understanding the implementation of language-of-instruction
policies in urban and rural Kenyan schools. The design was such that both student outcomes
and language usage could be examined across settings, and that the relationship between
language usage and student outcomes could be established (Piper & Miksic, 2011). The
design also allowed for a careful description of how and when particular languages were
being used, and individual pupils’reading outcomes were assessed across languages to
better understand what skills pupils had in the two or three languages used in the individual
schools that they learned in. In Kenya, the study used the EGRA tool, which was first uti-
lized in Malindi in a 2007–2008 study. EGRA has been implemented in more than 50
countries and over 80 languages (Trudell, Dowd, Piper, & Bloch, 2012). The Kenyan
EGRA was unique for two reasons, however. First, children were assessed in more than
one language (see Table 1). This allowed an examination of the impact of a multilingual
environment on schooling, and also of the way in which implementation of the Kenyan
language policy affects student outcomes.
Table 1. Number of students assessed by language of instruction and urban/rural status.
Province Urban or rural Languages Schools Students
Central Urban English and Kiswahili 25 501
Rural English, Kiswahili and Gikuyu 25 502
Nyanza Urban English and Kiswahili 25 499
Rural English, Kiswahili and Dholuo 25 498
8B. Trudell and B. Piper
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
Second, a set of classroom observations were done in parallel with the assessment of
student reading ability. These observation data are the basis of the EGRA findings most rel-
evant to this study as discussed below.
The research tools
EGRA consists of several tasks which are adapted to particular language and country con-
texts. The following tasks were assessed in Kenya.
.Letter identification: name: ability to read the names of letters.
.Phonemic awareness: ability to identify the first sound (or syllable) in a word.
.Letter−sound fluency: ability to identify the sounds of letters.
.Syllable naming fluency: ability to identify syllables.
.Familiar word fluency: ability to read high-frequency words.
.Non-word oral reading fluency: ability to read non-words.
.Connected text oral reading fluency: ability to read a passage, about 60 words long.
.Reading comprehension: ability to answer comprehension questions based on the
passage read.
As noted previously, EGRA was implemented in English, Kiswahili, Gikuyu and
Dholuo languages. Given the linguistic differences between these four languages, the
sets of tasks assessed in each language were slightly different (Table 2).
Also as noted previously, the Kenyan EGRA dataset also included a set of classroom
observations done in parallel with the student assessments. Across the 100 schools, 979
observations of classroom instruction were made, with a particular focus on language of
instruction. The aim was to determine the actual language practices in the classroom,
and to assess whether the characteristics of language use differ by locality. National
language policy distinguishes between urban and rural schools, with the expectation
being that urban areas would employ Kiswahili as the language of instruction in Standards
1–3, while rural areas would employ the mother tongue (Gikuyu in Central Province and
Dholuo in Dholuo-speaking portions of Nyanza province) in those grades.
The instrument developed to assess these language practices was based on the Stallings
snapshot tool;
1
for each 30-minute class, 15 data points measured the language used, the
person using the language (teacher or student), the person(s) to whom the communication
was directed and the type of language use (i.e. speaking, reading or writing) (Piper &
Miksic, 2011). This classroom observation information constitutes a substantial set of
language-of-instruction data, and is the foundation of the current examination of how
language policy is implemented in Kenyan primary schools.
Table 2. Kenya EGRA tasks, by language.
Language
Letter
identification:
name
Phonemic
awareness
Letter
identification:
sound
Syllable
naming
Familiar
word
reading
Non-word
reading
Oral
reading
fluency
Reading
comprehension
English ••• ••••
Kiswahili •••••••
Gikuyu •••••••
Dholuo •• ••••
Current Issues in Language Planning 9
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
Language-of-instruction practices in the classroom and national language policy
The first research question addressed is whether Kenya’s classrooms adhere to the national
language-of-instruction policy. Figure 1 shows the proportion of classroom instruction time
spent using various languages in early primary classrooms (Standards 1–3). The data in this
figure are from Central Province; the proportions found are quite similar to those found in
Dholuo-speaking Nyanza Province. In the classrooms studied, Gikuyu, the mother tongue
of the students, was used 18.2% of classroom time in rural areas and 1.6% of classroom
time in urban schools. Kiswahili, the purported language of instruction in urban areas
and a subject in rural areas, is used only 31.8% of the time in urban areas and 27.7% in
rural areas. In all cases English instruction dominates, used 66.6% of the time in urban
classrooms and 53.9% of the time in rural classrooms. These data from Central Province
and comparable data from Nyanza Province (Piper & Miksic, 2011) also show that the
great majority of instruction in these primary classrooms is carried out in English.
This language practice can be compared with the expected language use for each taught
subject. English, Kiswahili and mother-tongue subjects are expected to use the respective
languages of instruction; subject areas such as math, religious education and science are
supposed to be taught in the mother tongue. The data on what languages were actually
used in particular subjects (Figure 2; Piper & Miksic, 2011) shows that English was the
most frequent language used in Standard 1–3 in English class, as expected; but it was
also the most frequent language used in the subjects of social studies (82.7% of the
time), mathematics (77.8%), life skills (75.5%), religious education (72.0%) and science
(70.4%). Kiswahili use dominates only in Kiswahili class, and mother-tongue use domi-
nates only in mother-tongue class. Thus, in contravention to national language policy,
the mother tongue and Kiswahili are confined to the courses called ‘mother tongue’and
‘Kiswahili’, while English dominates instruction in all other classes.
In addition to understanding how language is used across subjects, the data also demon-
strate how language use differs through the primary years. Figure 3 shows that in both
the Central and the Luo-speaking Nyanza schools studied, use of the mother tongue dra-
matically and quickly declines from Standard 1 to Standard 5, starting at 24.5% of the
time in Standard 1 and declining to 2.7% in Standard 5. English use increases just as
dramatically, from 46.8% of the time in Standard 1 to 81% in Standard 5. So even in
Standard 1, the proportion of English use observed was significantly higher than either
Kiswahili or mother tongue; but by Standard 2, the proportion of English used was
higher than that of Kiswahili and mother tongue combined.
Figure 1. Language use in early primary classrooms of Central Province, by urban and rural schools.
10 B. Trudell and B. Piper
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
The data also allowed investigation of whether individual classroom observed fol-
lowed the national language policy. The criterion used here was that individual class-
room practices were considered to ‘follow the policy’when at least 75% of the
classroom’s data points were in accord with the policy for that subject. Figure 4 presents
the percentage of classrooms following the language policy by this definition, the
language stipulated by policy for teaching content subjects in the province and the
overall number of classrooms assessed. This figure shows that classrooms followed
the language policy less than 50% of the time; policy adherence was highest in rural
Nyanza Province, at 43.9%. Urban policy adherence was near 40%, in both urban
Central and Nyanza provinces. Policy adherence was lowest in rural Central Province,
but in both urban and rural Central and Nyanza, adherence to the language policy
was quite low.
Figure 2. Language use by subject in early primary classrooms.
Figure 3. Language use by language and class level in Kenya.
Current Issues in Language Planning 11
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
Language use and pedagogy
Classroom observation data also addressed the ways in which the type of language use
varied by activity and by classroom structure. In examining the classroom structure data,
the question was whether language use happened at the whole class level, in small
groups or at the individual level. Figure 5 shows that more than 70% of teacher–student
interactions observed took place between the teacher and the whole class, rather than
with either individuals or small groups.
A related question is how language is used in the classroom by both teachers and stu-
dents: speaking, reading or writing. Figure 6 presents the proportion of observations, of
both students and teachers, of speaking, reading and writing activities. The figure shows
that the most prevalent activity is speaking, with teacher or student speech accounting
for 43.9% of the language-use time, more than writing (26.5%) or reading (14.5%). This
fact helps to explain some of the low achievement scores on reading outcomes indicated
in the Kenyan EGRA data (Piper, 2010), since it is difficult for children to learn to read
with fluency and comprehend if little time is spent on reading.
Comparing classroom practice to reading assessment data
These observations of classroom language behaviors take on special significance
when compared to the assessment of reading fluency and comprehension among the
students in those same classrooms. The assessment data are presented in more depth in
Figure 4. Adherence to national language policy in individual classrooms.
Figure 5. Type of teacher–student interaction in the classroom.
12 B. Trudell and B. Piper
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
Piper (2010) and some of the conclusions drawn there are highly relevant to understanding the
significance of the classroom observation data presented in this study.
First, given the preponderance of English used as the language of instruction in Central
and Luo-speaking Nyanza schools, it is unsurprising that the EGRA findings indicate that
children ‘read’English (as measured by phonemes or words read orally per minute) more
fluently than they do the other languages in this study (though their scores in English are
still only between 25 and 40 words per minute). More time spent teaching in English
yields more student familiarity with English sounds and words. Where little time is invested
in teaching reading in Kiswahili or the mother tongue, that lack of attention results in poor
ability to decipher the letters and words unique to those languages.
However, when it comes to understanding the EGRA results show that reading compre-
hension scores were significantly higher in the mother tongue than in English. Children in
the rural Central Province scored 465.6% higher on comprehension tests in Gikuyu than
they did in English; those in rural Nyanza Province scored 748.5% higher in Dholuo
than in English.
Where Kiswahili was concerned, the students’comprehension was also significantly
higher than in English. In the urban Central Province, children scored 61.9% higher on
the comprehension test in Kiswahili than they did in English; in the urban Nyanza Province,
they scored 141.5% higher. And in the rural classrooms, children understood what they read
in Kiswahili 292.7% better than in English in the rural Central Province, and 193.3% better
in the rural Luo-speaking Nyanza Province.
This shows that while Kenyan children can ‘read’(that is, decode) English words more
fluently, their understanding of English is very limited. On the other hand, children under-
stand much more of what they read in their mother tongue, even though their skills in
decoding words in the mother tongue are weaker than they are in English.
Discussion of findings
The classroom observation data show that children in Standards 1–3 of the schools studied
are receiving instruction predominantly in English, right from the start of primary school.
The high proportional use of English during classroom time is limiting the use of the mother
Figure 6. Language activities by teachers and students.
Current Issues in Language Planning 13
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
tongue, and even the use of Kiswahili. Even in Standard 1, the proportion of English
use observed is significantly higher than either Kiswahili or the mother tongue; but by
Standard 2, the proportion of English used is higher than that of Kiswahili and mother
tongue combined. This practice is in apparent disregard for the children’s lack of fluency
in English.
These findings confirm the expressed attitudes of many Kenyan parents and teachers as
studied by other researchers (Bunyi, 2005; Kioko, 2013)–namely, that English is the preferred
language of instruction for primary students. The evidence also makes it clear that the Kenyan
national language policy mandating use of the language of the catchment area in grades 1–3is
being widely ignored in local decision-making regarding language of instruction.
The findings also indicate that this practice of using English as the predominant
language of instruction is providing the children with basic skills in recognizing
common English words and very modest skills in decoding new words. However, this prac-
tice is not providing the English language skills needed for the children to read with under-
standing. This aligns with reading research on the crucial role of language fluency in
reading comprehension (see for example, Snowling & Hulme, 2006). In these Kenyan
classrooms, following the national language policy by teaching the students to read in
their mother tongues could result in far stronger achievement in both fluency and compre-
hension, yet that is not what is taking place.
The findings also show that the most prevalent classroom language activity is speaking
–specifically, the teacher speaking to the entire class, and predominantly using the English
language. Given the lack of English fluency among the children, this practice is pedagogi-
cally indefensible. In addition, reading by either teacher or students takes up 15% or less of
classroom time, so that only minimal modeling and practicing reading are taking place. This
fact helps to explain some of the low achievement scores on reading outcomes (Piper,
2010), since it is difficult for children to learn to read with fluency and comprehension if
little time is spent on reading –in any language, let alone a language which the students
do not master.
Conclusions: policy, resistance and agency
What does this evidence mean for language policy formulation and appropriation in Kenya?
The dismal reading scores indicate that the current mother-tongue policy is not resulting in
successful learning; children are unable to decode well in either English or their own
languages, and their English comprehension is sorely lacking as well. National policy is
being clearly flouted in favor of locally held language ideology about the preferential
role of English in formal education, and this resistance is being supported by education lea-
dership. This locally held language ideology, as firmly as it is held, is not leading to success
in the classroom; yet this fact is seldom recognized in such situations of school failure.
Rather than being better supported and enforced, the national policy is blamed –and poss-
ibly replaced.
Is this mediocre uptake of national policy at local levels a source of concern for the
national government? Certainly there is no ready evidence of state-initiated support for
teacher training, local-language materials development or advocacy with educators and
community leaders on behalf of the policy. Not only so, but the normalcy of resistance
to the national language policy seems to indicate that local actors do not think the state
is seriously behind it either. Reflecting on Shohamy’s observation that language policies
may perhaps not even be intended for implementation, one cannot help but wonder
whether the Kenyan language policy falls into this category.
14 B. Trudell and B. Piper
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
A related question has to do with the extent to which local decisions about language use
in the classroom are actually an expression of local agency. Sen (1999)defines agency as
the human capacity to bring about change, the results of which may be judged in terms of
the agent’s own values objectives. The expression of agency with regard to language choice
has been well documented; indeed, as Korang and Slemon (1997, p. 249) note, ‘every
theory of culture puts language at the centre of debates about power, ideology, subjectivity
and agency’. However, agency where local-language communities are concerned has most
generally been described in terms of the determination of speakers to maintain and revita-
lize their languages as part of their ethnic and community rights (May, 2001; Trudell, 2006;
Trudell & Klaas, 2010)–in the face of global and/or national pressures for linguistic
conformity.
What is demonstrated in this study, on the other hand, is a determination on the part of
local-language-speaking stakeholders to reject the use of their own languages in the formal
school system, in favor of a prestige language which their children in fact do not speak and
in which those children are clearly not succeeding. Certainly the stakeholders are expres-
sing agency as defined by Sen, but this decision bears a suspicious mark of influence by
the same dominant culture, language and education system that has marginalized them in
the first place.
Not only so, but the fact that the pro-local language policy originated with the national
government should not be construed to mean that it reflects the position of the national
culture regarding the value of local language and culture, since it certainly does not. In
countries where national leaders embrace the reality of their nation’s multi-ethnic, multilin-
gual identity, the political will to both formulate and implement pro-local language edu-
cation policies does exist; the case of multilingual education in Ethiopia demonstrates
this amply (Benson, Heugh, Bogale, & Gebre Yohannes, 2012). However, where a
language-of-instruction policy collides with national self-identity the political traction
necessary for serious implementation is lacking. This is why advocacy efforts on behalf
of mother-tongue-based education must reach beyond the technical and pedagogical, to
address the realities of national identity and aspirations.
Possible additional research directions could further illuminate the significance of these
data. One fruitful line of investigation might be an examination of the language choices
made by parents, teachers and pupils in these communities, and how those choices are per-
ceived to be meeting local aspirations and expectations where education is concerned.
Examination of the processes of teacher training, and ways in which they support or
neglect the pro-local language education policy in Kenya, would also be useful. Studies
of schools in other language communities of Kenya would also be helpful; the Dholuo-
and Gikuyu-speaking communities are among the largest, but by no means the only
major language communities of the nation; and language attitudes are far from homogenous
across Kenyan communities. A study comparing the impact of a local-language reading
program against that of a Kiswahili and English reading program could also be useful.
Recent history in Kenya clearly indicates that issues of cultural and linguistic identity
exert a powerful influence on Kenyan society. In Kenya’s drive for national unity, its pro-
foundly multilingual and multi-ethnic nature must be recognized and supported as a means
of forging a strong nation. Ensuring that the conditions are in place to make the current
language policy successful would be a good place to begin.
Note
1. www.eddataglobal.org/embedded/stallings_snapshot.doc.
Current Issues in Language Planning 15
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
Notes on contributors
Barbara Trudell, PhD, has worked in the field of language and education in both South
America and Africa. Her research interests include local processes of language development,
language policy implementation, local-language literacy and development, language and
reading, and language choices in formal education contexts. Barbara lives in Nairobi, Kenya.
Benjamin Piper is the Chief of Party of the DFID and USAID funded Primary Math and
Reading (PRIMR) Initiative programs in Kenya, serving 1300 schools and undertaking
applied research on the most effective methods to improve learning in Kenya. Dr Piper
is also the Regional Education Advisor with RTI’s International Development Group. In
this role, he provides policy advice for governments across East Africa. His previous
work investigated the impact of a variety of language-of-instruction educational policies
on student outcomes, with particular focus on how pedagogical differences relate to
early-grade reading acquisition and the transfer of reading skills between the first language
and the second language. He has experience working with classroom pedagogy, director/
principal supervision, policy analysis and student assessments and has worked with a
variety of organizations, including the World Bank, UNICEF, DFID and Save the Children.
References
Alidou, H. (2003). Language policies and language education in francophone Africa: A critique and a
call to action. In S. Makoni, G. Smitherman, A. F. Ball, & A. K. Spears (Eds.), Black linguistics:
Language, society and politics in Africa and the Americas (pp. 103–116). London: Routledge.
Ambatchew, M. (2010). Traversing the linguistic quicksand in Ethiopia. In K. Menken & O. Garcia
(Eds.), Negotiating language policies in schools: Educators as policymakers (pp. 198–210).
New York, NY: Routledge.
Baker, C. (2006). Foundations of bilingual education and bilingualism (4th ed.). Clevedon:
Multilingual Matters.
Benson, C., Heugh, K., Bogale, B., & Gebre Yohannes, M. A. (2012). Multilingual education in
Ethiopian primary schools. In T. Skutnabb-Kangas & K. Heugh (Eds.), Multilingual education
and sustainable development work: From periphery to center (pp. 32–61). New York, NY:
Routledge.
Bunyi, G. (2005). Language classroom practices in Kenya. In A. Lin & P. W. Martin (Eds.),
Decolonisation, globalisation: Language-in-education policy and practice (pp. 131–152).
Clevedon: Multilingual Matters.
Cameroonian Ministry of National Education. (1998). Law of education in Cameroon: Law No. 98/
004 of 14th April 1998. Yaoundé: Ministry of National Education.
Diarra, E. (2003). Choice and description of national languages with regard to their utility in literacy
and education in Angola. In A. Ouane (Ed.), Toward a multilingual culture of education (pp. 333–
348). Hamburg: UNESCO Institute for Education.
Federal Democratic Republic of Ethiopia. (1994). Education and training policy. Addis Ababa:
Author.
Gove, A., & Cvelich, P. (2010). Early reading: Igniting education for all. A report by the early grade
learning community of practice. Research Triangle Park, NC: Research Triangle Institute.
Government of Ghana. (2002). Meeting the challenge of education in the twenty-first century (Report
of the president’s committee on review of education reforms in Ghana). Accra: Adwinsa
Publications.
Haugen, E. (1966). Dialect, language, nation. American Anthropologist, 68, 922–935.
Hélot, C. (2010). ‘Tu sais bien parler Maîtresse!’Negotiating languages other than French in the
primary classroom in France. In K. Menken & O. Garcia (Eds.), Negotiating language policies
in schools: Educators as policymakers (pp. 52–71). New York, NY: Routledge.
Heugh, K. (2008). Language policy and education in southern Africa. In S. May & N. Hornberger
(Eds.), Encyclopedia of language and education (2nd ed., Vol. 1, pp. 355–367). Language
policy and political issues in education. New York, NY: Springer Science.
16 B. Trudell and B. Piper
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
Johnson, D., & Freeman, R. (2010). Appropriating language policy at the local level: Working the
spaces for bilingual education. In K. Menken & O. Garcia (Eds.), Negotiating language policies
in schools: Educators as policymakers (pp. 13–31). New York, NY: Routledge.
Kembo, S., & Ogechi, N. (2009). Linguistic human rights and language policy in the Kenyan edu-
cation system. Addis Ababa: OSSREA.
Kenya Institute of Education. (1992). Primary education syllabus (Vol. 1). Nairobi: Republic of
Kenya Ministry of Education.
Kioko, A. (2013). Language policy and practice in Kenya: Challenges and prospects. In H.
MacIlwraith (Ed.), Multilingual education in Africa: Lessons from the Juba language-in-edu-
cation conference (pp. 117–126). London: British Council.
Korang, K. L., & Slemon, S. (1997). Post-colonialism and language. In M. Msiska & P.Hyland (Eds.),
Writing and Africa (pp. 246–263). London: Longman.
Laitin, D. D. (1993). The game theory of language regimes. International Political Science Review,14
(3), 227–239.
May, S. (2001). Language and minority rights: Ethnicity, nationalism and the politics of language.
London: Longman.
May, S. (2008). Language, education, pluralism and citizenship. In S. May & N. Hornberger (Eds.),
Encyclopedia of language and education (2nd ed., Vol. 1, pp. 15–29). Language policy and pol-
itical issues in education. New York, NY: Springer Science.
Mutaka, N. M., & Attia, L. M. (2008). Introduction: Issues in the teaching of languages as a prop to a
development-oriented literacy. In N. M. Mutaka (Ed.), Building capacity: Using TEFL and
African languages as development-oriented literacy tools (pp. 1–12). Bamenda: Langaa
Research and Publishing CIG.
Muthwii, M. J. (2007). Language planning and literacy in Kenya: Living with unresolved paradoxes.
In A. Liddicoat (Ed.), Issues in language planning and literacy (pp. 46–62). Clevedon:
Multilingual Matters.
Nyaga, S., & Anthonissen, C. (2012). Teaching in linguistically diverse classrooms: Difficulties in the
implementation of the language-in-education policy in multilingual Kenyan primary school class-
rooms. Compare: A Journal of Comparative and International Education, 42(6), 863–879.
Opoku-Amankwa, K. (2009). English-only language-in-education policy in multilingual classrooms
in Ghana. Language, Culture and Curriculum, 22(2), 121–135.
Ouane, A., & Glanz, C. (2010). Why and how Africa should invest in African languages and multi-
lingual education. Hamburg: UNESCO Institute for Lifelong Learning.
Piper, B. (2010). Kenya early grade reading assessment findings report (William & Flora Hewlett
Foundation research report). Retrieved from https://www.eddataglobal.org/documents/index.
cfm?fuseaction=pubDetail&ID=275.
Piper, B., & Miksic, E. (2011). Mother tongue and reading: Using early grade reading assessments to
investigate language-of-instruction policy in East Africa. In A. Gove & A. Wetterberg (Eds.), The
early grade reading assessment: Applications and interventions to improve basic literacy (pp.
139–182). Research Triangle Park, NC: RTI Press. Retrieved from http://www.rti.org/pubs/bk-
0007-1109-wetterberg.pdf.
Samuelson, B. L., & Freedman, S. W. (2010). Language policy, multilingual education and power in
Rwanda. Language Policy, 9, 191–215.
Sen, A. (1999). Development as freedom. Oxford: Oxford University Press.
Shohamy, E. (2006). Language policy: Hidden agendas and new approaches. New York, NY:
Routledge.
Shohamy, E. (2010). Cases of language policy resistance in Israel’s centralized education system. In
K. Menken & O. Garcia (Eds.), Negotiating language policies in schools: Educators as policy-
makers (pp. 182–197). New York, NY: Routledge.
Snowling, M. J., & Hulme, C. (2006). Language skills, learning to read and reading intervention.
London Review of Education, 4(1), 63–76.
Spolsky, B. (2004). Language policy. Cambridge: Cambridge University Press.
Sprenger-Tasch, M. (2003). Attitudes towards Luganda, Kiswahili, English, and mother tongue as
media of instruction in Uganda. In H. Cuyckens, T. Berg, R. Dirven, & K.-U. Panther (Eds.),
Motivation in language (pp. 347–366). Amsterdam: John Benjamins.
Tadadjeu, M. (1997, September 11–15). The challenge of language and education in African devel-
opment: A vision for the 21st century. Paper presented at the Oxford International Conference on
Education and Development, Oxford, UK.
Current Issues in Language Planning 17
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
Trudell, B. (2004). The power of the local: Education choices and language maintenance among the
Bafut, Kom and Nso’communities of Northwest Cameroon (PhD dissertation). University
of Edinburgh, UK. Retrieved from www.silcam.org/download.php?sstid=030401...
ThePoweroftheLocal.pdf
Trudell, B. (2006). Local agency in the development of minority languages: Three language commit-
tees in northwest Cameroon. Journal of Multilingual and Multicultural Development, 27(3), 196–
210.
Trudell, B. (2007). Local community perspectives and language of education in sub-Saharan African
communities. International Journal of Educational Development, 27, 552–563.
Trudell, B. (2010). Language, culture, development and politics: Dimensions of local agency in
language development in Africa. Journal of Multilingual and Multicultural Development,31
(4), 403–419.
Trudell, B., Dowd, A. J., Piper, B., & Bloch, C. (2012). Early grade literacy in African classrooms:
Lessons learned and future directions. Retrieved from http://www.adeanet.org/triennale/
Triennalestudies/subtheme1/1_5_04_TRUDELL_en.pdf
Trudell, B., & Klaas, R. (2010). Distinction, integration and identity: Motivations for local language
literacy in Senegalese communities. International Journal of Educational Development, 30(2),
121–129.
Woolard, K., & Schieffelin, R. (1994). Language ideology. Annual Review of Anthropology, 23, 55–
82.
Wright, S. (2004). Language policy and language planning: From nationalism to globalisation.
New York, NY: Palgrave.
Yigezu, M. (2010). Language ideologies and challenges of multilingual education in Ethiopia: The
case of Harari region. Addis Ababa: Organization for Social Science Research in Eastern and
Southern Africa (OSSREA).
Zhao, S., & Baldauf, R. B., Jr. (2012). Individual agency in language planning Chinese script reform
as a case study. Language Problems & Language Planning, 36(1), 1–24.
18 B. Trudell and B. Piper
Downloaded by [Benjamin Piper] at 12:33 13 November 2013
