See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/248167871

Classroom interaction: Potential or problem? The case of Karagwe

Article  in  International Journal of Educational Development · March 2010

DOI: 10.1016/j.ijedudev.2009.03.007

CITATIONS READS

29 3,331

1 author:

Åsa Wedin
Dalarna University
66 PUBLICATIONS   355 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Mother tongue tution and studying guidance in Swedish compulsory school (2015-2017) View project

Flexibilitet och ökad individanpassning i SFI-undervisningen View project

All content following this page was uploaded by Åsa Wedin on 01 November 2018.

The user has requested enhancement of the downloaded file.

This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

International Journal of Educational Development 30 (2010) 145–150

Contents lists available at ScienceDirect

International Journal of Educational Development

journal homepage: www.elsevier.com/locate/ijedudev

Classroom interaction: Potential or problem? The case of Karagwe

Åsa Wedin *
HumES, University of Örebro, Örebro, Sweden

A R T I C L E I N F O A B S T R A C T

Keywords: This paper discusses interactional patterns in classrooms in primary school in rural Tanzania, based on
Educational policy an ethnographic study on literacy practices. The paper argues that the official policy of Swahili-only in
Development primary school, together with the huge gap between high expectations on educational outcome and lack
Classroom interaction of resources, have resulted in the creation of safety strategies among pupils and teachers. These safety
Primary school
strategies include interactional patterns that also constitute a hindrance for students’ learning. However,
Call response
I claim that these interactional patterns could constitute a potential for educational development, if
research findings from bilingual education were taken into account.
ß 2009 Elsevier Ltd. All rights reserved.

1. Introduction not advised, through curricula, schoolbooks and manuals, how to

handle the fact that Swahili is a second language for the majority
It is important that resources devoted to education are used of the children. English is used as the medium of instruction from
effectively and that the education is relevant for students. This is secondary school level and the status of English in Tanzania is
particularly important in poor countries, such as Tanzania, as lack increasing. That, in combination with the later assumption, that
of resources is a big problem. In Karagwe district in Tanzania, an effective way of teaching children English is to make it the
educational outcome is low, as in many parts of Africa. This is well- medium of instruction also of primary school, has resulted in a
known and there are many factors behind that, such as lack of mushrooming of English medium primary schools in Tanzania
teaching materials, well-trained teachers, and relevant school around the turn of the millennium. In this paper, I will challenge
buildings (Brock-Utne, 2000; Buchert, 1994; Wedin, 2004). The the first one of these assumptions, that children in Tanzania
role of language as a medium of instruction in education in Africa is master Swahili when they begin school. I also aim at showing that
discussed in many contexts. Many researchers have discussed the existing policies have resulted in interactional patterns in school
fact that the use of English for instruction in secondary school in that constitute a hindrance for students’ learning. I also aim at
Tanzania causes great problems for educational outcome in showing that if these interactional patterns were developed,
secondary school, as neither teachers nor students master English based on relevant research on bilingual education and in
at a sufficient level (see for example Trappes-Lomax, 1990; accordance with the ways similar patterns are used in society,
Lwaitama and Rugemalira, 1990; Rubagumya, 1994). However, the they could be turned into an asset for educational development in
role of language as a medium of instruction in primary schools in Tanzania.
Tanzania is seldom discussed.
Researchers have shown that there are two common assump- 2. Methods and approaches
tions that seem to be held by educational policy makers in
Tanzania (Brock-Utne, 2000; Rubagumya, 1990, 1994; Blom- This paper is based on a study I carried out during the period of
maert, 1996). The first one is that all children in Tanzania master 1999–2003 on literacy practices in Karagwe. I studied people’s use
Swahili when they begin school and the second is that using a of and attitudes to written language and tried to find ways to make
foreign language such as English as medium of instruction in literacy education in school more relevant to the students’ future
school is a good way of teaching students that language. Thus, lives, by linking what happens in school with what happens in
primary school education is carried out in Swahili and teachers are society. I spent 10 months in the area, divided into five field studies
during 2000–2003, observing literacy practices in different
settings, such as in schools, offices, homes, churches, market
* Correspondence address: Department of Education, Örebro University,
places and in other domains where people spent their days. One
Fakultetsgatan 1, SE-701 82 Örebro, Sweden. Tel.: +46 19 30 12 81. important obstacle for effective and relevant literacy education I
E-mail address: [email protected]. identified, and that is the topic of this paper, was interactional

0738-0593/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijedudev.2009.03.007
 Author's personal copy

146 Å. Wedin / International Journal of Educational Development 30 (2010) 145–150

patterns in classrooms that I claim are a result of language patterns 2.2. Theoretical background
and language choice in primary school (see Wedin, 2004, 2005).
The study was carried out using an ethnographic perspective. The linguistic situation in Karagwe, a multilingual setting, with
This means that it was concerned with the study and cultural one language used mainly in private domains and one mainly used in
interpretation of behaviour in naturally ongoing settings. Ethno- official domains, while other languages are used parallel to the main
graphic methods, which strive to give a holistic/contextual, languages, is quite common in the world, particularly in poor
comparative and cross-cultural picture on the phenomenon in settings. Similar situations, where children master one language
focus, are particularly appropriate where deeper understandings when they enter school and are supposed to master another
are searched for. Ethnography concerns issues of human choice dominant language when they finish primary school, are found in
and meaning, and in that sense it concerns issues of improvement several settings in the world and this is particularly common in
in educational practices and is thus highly relevant for educational former colonies. Research on bilingual education is quite impressive
research (Ericksson, 1986). The interpretative approach enables today and have been carried out by among others van Lier (1996),
the creation of information that is valuable in educational Cummins (2000, 2007), Heller and Martin-Jones (2001) and
planning. Hornberger and Chick (2001). A study by Thomas and Collier
Interviews were carried out with children, teachers and (1997, 2002) showed very impressively that the most important
parents both formally and informally in different settings in the factor for children’s success in bilingual education was that their
society. I carried out formal interviews with children from mother tongue was strongly supported throughout their schooling
different classes, apart from all informal contacts I had with (see also Cummins, 1996, 2000; Gibbons, 2002, 2006; Schleppegrell,
numerous children. By interacting with children inside and 2004). It should not be very controversial to claim that children learn
outside schools I tried to get an understanding of their views. As best when they are taught in a language that they understand. This is
the society is strictly hierarchical, formal interviews with children particularly important during the first years of schooling, when
particularly in school settings are problematic. Observations in children are supposed to learn the three R’s (reading, writing and
different areas where people spend their time, was also an arithmetic) as this is the base for later schooling (Cummins, 2000).
important part of the study. In order to understand how students’ This means that, in the case of Karagwe, children should gain a lot if
learned, including their level of understanding of what was taught they where to be taught Swahili as a second language, following
at school, I observed classes, studied the pupils’ exercise books research findings on bilingual education, while they should
and carried out follow-up interviews a few days after observed simultaneously have support in their mother tongue, Runyambo.
lessons. I also studied teachers planning books and interviewed This, however, is not the case. Following the official policy of using
teachers before and after lectures. Swahili only from pre-school, very little Runyambo is used in
classrooms, as this is against official policies.
2.1. The setting Another important factor for children’s success in second
language acquisition in school is that what is done in school is
Karagwe district is located in Kagera in the north-west of based on their earlier knowledge and experience (Säljö, 2000).
Tanzania, bordering to Uganda and Rwanda. The Banyambo, who Research drawing on the findings by among others Shirley Brice
inhabit Karagwe, number about 500,000. This was traditionally a Heath in Piedmont Carolina, USA (Heath, 1983), has stressed the
hierarchical society, with ruling pastoralist clans and a number of importance of linking what happens in classrooms with what
agriculturalist clans. The language of the Banyambo, Runyambo, happens in the children’s homes and surroundings. The impor-
is one of the Western-Lacustrine Bantu-languages together with tance of considering interactional patterns in classrooms has been
neighbouring languages such as Ruhaya, Kinyankole and stressed by many researchers, such as Mehan (1992), Au (1980),
Kinyarwanda. The Arab traders introduced Swahili and as the Ericksson (1986), Cazden (2001) and Watson-Gegeo (1992). In
area was later colonized, first by the Germans and then by the Hawaii, Au showed how the use of local patterns used in narration
British, both German and English influenced language patterns in also in school, led to increased achievement in the education in the
the area. lower classes. In Kwara’ae (Solomon Islands), the local interac-
In the beginning of the third millennium and 45 years after tional pattern ‘‘shaping the mind’’ was introduced in schools with
independence and the creation of the state Tanzania, I estimate positive results (Watson-Gegeo, 1992).
that most Banyambo over 12 years of age are functionally bilingual
in Swahili and Runyambo. Both languages are frequently used and 2.3. Children’s language skills in primary school in Karagwe
most adults can use them for their daily communicative needs.
English is a high status language that is mastered by only very few This study showed that nearly all children in Karagwe master
individuals in the upper classes, such as some higher officials and a Runyambo when they begin school but not Swahili (Wedin, 2004,
few other highly educated persons. 2005). Nearly all children in the countryside and perhaps more than
Tanzania is among the poorer countries in the world and the 90% of the children in the towns know only a few words of Swahili
educational challenges are tough. The situation for the primary when they begin school. I also found that children in Karagwe
schools in this study, which were five among the best rural schools generally do not master Swahili until the last classes in primary
in Karagwe, was challenging in many ways. Most classrooms had school. According to my results it is not until in standard four or five
more than 50 students, in some cases more than 100. The highest that most students know Swahili well enough to be able to use it in
number of students was found in preschool and standards one and ordinary conversation. This was verified by teachers. I asked
two. As the drop-out rate was high, classrooms in standard seven teachers, for example, if they told children in lower classes stories
often had only 20–30 students. In the beginning of the study the but they answered that they could not do so as the children would
lack of school books was high. Only some subjects had books and it not understand a story told in Swahili until higher classes. This
was not unusual that four students had to share one book. During means that teachers’ perceptions in this case are similar to mine.
2002 and 2003 the number of schoolbooks increased. As school However, I have found that probably most students who finish
buildings is the responsibility of parents, in many cases four standard seven, i.e. the whole primary school, master Swahili on a
students had to share a school desk designed for two and some functional level. I interviewed some of the students in standard
classrooms had earthen floor. There were also cases were teaching seven with the lowest results and they could all discuss with me in
had to take place outdoors. Swahili on a level that I estimated as proper for their age. Some of the
 Author's personal copy

Å. Wedin / International Journal of Educational Development 30 (2010) 145–150 147

most low-performing students I also met in follow-up interviews 1 part of a system that Hornberger and Chick call safe-time (2001). I
year after they had finished school and they were still fluent in also noted that students were very skilled in guessing the answers
Swahili. This shows that most students in my study acquire Swahili expected by the teacher. Teachers are poorly trained and in many
at least on a functional level through primary school as children in occasions I noted that teachers not only made mistakes but also
these rural areas rarely meet Swahili outside school. The domains were inconsistent. Thus, I occasionally had problems in predicting
where Swahili is used in Karagwe, apart from in schools, are mainly what answer the teacher was expecting. However, many times
adult domains, such as offices, churches and market places. during my observations, I noticed that students managed to guess
the ‘‘correct’’ answer when I did not.
2.4. Interactional patterns in classrooms I prefer to call the different strategies used to hide failure in
classrooms safety strategies to include other strategies such as
Through the observations and interviews in schools I under- those mentioned earlier, and to call the prompt-answer interaction
stood that generally, the understanding of what was taught was call–response as in all these cases the teacher, or another speaker,
very poor. Students often did not seem to have understood what prompts or calls for response.1
had been taught at all. This of course has many reasons, but I I found four types of call–response used by teachers in primary
conclude that one reason why most children could not retell schools in Karagwe:
anything of what had been taught even the day after the lesson had
been given, or answer any questions on the content, is that thay are 1. Repetition: Pupils repeat words or parts of teachers’ speech.
taught through a language that they do not master, at least not 2. Confirmation: Teachers ask questions of the type: ‘‘Are we
during the four or five first years of schooling. together?’’, ‘‘Do you understand?’’ and ‘‘Isn’t it?’’ The answer is
What also became clear to me was that pupils and teachers had invariably: ‘‘Yes!’’
developed what I call safety strategies. These are strategies that are 3. Content-question: Teachers ask questions on the content taught.
developed by teachers and pupils to create a picture of successful 4. Completion: Pupils are requested to complete words, clauses or
learning and to hide failure. In a situation as in Karagwe, where there sentences in the teachers’ speech.
are extremely high expectations on education while at the same
time the conditions are extremely poor and thus also the outcome, Teachers mark pupils’ turns prosodically by using high tones and
teachers and pupils create ways to hide this failure to escape stress on the last syllable. (In Swahili stress is normally on the
punishment and to ‘‘save their faces’’ (see also Hornberger and Chick, penultimate syllable) Often the teacher repeats the answer, or gives
2001 for similar examples from South Africa and Peru). Education is the correct one if pupils fail, in low tone. Pupils are also encouraged
perceived as very important in Tanzania and particularly in to shout their responses out loudly, which they do with great
Karagwe. In the earliest policies of the new state of Tanzania, the enthusiasm, and the pattern is explicitly trained in lower classes. I
importance of education was stressed heavily (Nyerere, 1968). will give some examples from my observations. The first one,
Education was expected to be the key to development and a better repetition, is commonly used in the lower classes. (Underlining
financial situation. Through my study I found that this was very denotes talk with high tone and italics denote talk in low tone.)
much the case in Karagwe. People on all levels, both high and low Example 1: Repetition
educated, stressed the importance of education as a key for a better
future. Teachers are expected to produce well-educated citizens and Mw: Hii ni a! T: This is a!
children are sent to school to be able to improve their families’
W: Hii ni a! Ps: This is a!
economy in the future. In such a situation, with a cinsiderable gap
between expectations and actual outcome, and in a hierarchical
system where both teachers and students are frequently punished In this example we see that the teacher uses stress and high
for failure, they are likely to create ways to hide their failures, i.e. tone on the last syllable to call for pupils’ chorusing. The second
safety strategies. This is what has happened in schools in Karagwe. type, confirmation, is frequently used throughout primary school.
Teachers in the study tended, for example, mainly to address Example 2: Confirmation
successful learners and not to follow up when students did not show
any understanding. Students were supposed to hand in their Mw: Tunakunywa T: We drink clean water, don’t we?
exercise books after each lesson. This was usually only done by the maji safi, sivyo?
more successful students, a fact that teachers seldom reacted on. W: Ndiyo! Ps: Yes!
Less successful students tended to hide in the back of the classroom Mw: Eee? T: What?
and even not to show up at school at all. There is a saying among
W: Ndiyo! Ps: Yes!
children in Karagwe that many of their mates wanajificha kichakani,
they hide in the bushes. That means that they dress in the school
uniform in the morning and leave their homes. However, they stop Pupils’ answer is invariably affirmative. They should never
somewhere on the way to school and ‘‘hide in the bushes’’. This way answer no. The third type resembles typical questions of test-type,
they hope to escape the punishment they expect in school due to low also frequently used in schools elsewhere.
performance and also the punishment they would receive at home if Example 3: Content-question
they refused going to school. This is seldom acted upon on the
Mw: Tunaandika barua aina T: We write three types of
schools’ part. (For other types of safety strategies, see Wedin, 2004).
tatu. Tunaandika barua letters. How many types
One type of safety strategies that I discovered was a type of
aina ngapi? of letters do we write?
interactional pattern that guided students to give the expected
answers without actually understanding. When I observed I found W: Tatu! Ps: Three!
a chanting-like chorusing being used. This chorusing resembles the Mw: Tatu T: Three
types of interaction observed in classrooms in poor settings
elsewhere (Aziz, 1981; Watson-Gegeo, 1992; Ndayipfukamiye,
1994; Arthur, 1994, 2001; Chick, 1996; van Lier, 1996; Schieffelin, 1
Similar interaction patterns, that resemble what has been described here, have
1996; Heller and Martin-Jones, 2001). This chanting type of been observed among Afro-American priests and politicians in sermons and
interaction soon becomes predictable to students and becomes speeches (Smitherman, 1977).
 Author's personal copy

148 Å. Wedin / International Journal of Educational Development 30 (2010) 145–150

As we see, the answer is given prior to the question. In this case Mw: (. . .) ya pili inasema T: (. . .) and second it says
the teacher repeats the answer in low tone, which is also pamoja na kupewa zawadi, that together with being
frequently done in these classrooms. This gives the interaction amepiga na nini? given gifts, what has she taken?
pattern its chanting character. It is also common that teachers W: Zawad S: Gif
leave the last part out like in the following example.
Mw: A a, amepiga na nini? T: No no, what has she taken?
Example 4: Completion
W: Mapicha S: Pictureses2
Mw: Nyumbani tunaweka maji T: At home we keep Mw: Ee? T: What?
safi. Nyumbani tunaweka maji . . . water safely. At home W: Mapicha S: Pictureses
we keep water . . . Mw: Picha ni picha, amepiga? T: Pictures it’s pictures,
W: Safi! Ps: Safely! she has taken?
Mw: Safi T: Safely W: Picha T: Pictures

This last type, completion, is more abstract and is more often

used in higher classes. In the following example from a classroom
in standard 7 only the last syllable is left out: In this example, where the answer is not clearly given in
Example 5: advance by the teacher, we see that students make mistakes. In the
Mw: Hapa unaandika shilingi, T: Here you write shillings, first mistake (zawadi) they seem to use the same strategy as in
hapa unaandika se . . . here you write ce . . . example 6, repeating the last word said by the teacher in the
W: Senti Ps: Cents penultimate sentence. It shows that they do not understand what
the teacher is saying as the verb—piga cannot be followed by a
At first, the call–response interaction may seem to have the noun such as zawadi. Then we see that they elaborate on their
function of assessing pupils’ knowledge, to help teachers make Swahili as they put a Swahili-plural prefix on the word picha
sure the pupils have understood. However, after a closer look, we (pictures). However, in standard-Swahili loan-words, as this, do
find that pupils usually do not need to know the content to answer not take a plural-prefix.
the questions. In the second type, confirmations, pupils would not I agree with Hornberger and Chick (2001), that one function of
answer no and in the other types the answer is usually given. What this type of interaction is to create safety by saving the face of
pupils need to understand is the interaction pattern. Thus I teachers and pupils and that while this interaction hides the fact
conclude, in agreement with Hornberger and Chick (2001), that the that nearly no academic learning is taking place, in fact, at the same
function of this type of interaction is mainly social, not cognitive. It time it prevents academic learning. As the interaction pattern
gives teachers a chance to manage a seemingly impossible serves to hide lack of knowledge, neither teachers nor students get
situation by giving pupils turns, camouflaged as test questions, to know the actual knowledge of the students. Thus the level of
testing pupils’ knowledge, but in reality filling a social function. what is taught is likely to be far above the level that students can
This is done in co-operation with the pupils. My studies confirm understand.
that students actively learn this skill. In pre-school and standard 1 Hornberger and Chick claimed that the main reason for the
the pattern was explicitly exercised. The following example from a development of such patterns is that a language that neither
pre-school classroom shows this. students nor teachers master is used as medium of instruction in
Example 6: classroom (2001). This is, however, not in accordance with the case
of Karagwe, as the teachers actually master Swahili, the language
Mw: Safi? T: OK? of the classroom. Only students in the lower classes do not master
it. In this case I see the gap between expectations and outcome as
W: Kabisa! Ps: Absolutely!
one causal factor, linguistic problems another and the hierarchical
Mw: Safi? T: OK? system with frequent use of punishment as a third.
W: Kabisa! Ps: Absolutely!
Mw: Sasa mtakwenda T: Now you will go out and 2.5. Call–response in the society
nje mucheze kidogo, play a little, then you shall
halafu murudi darasani return to the classroom and However, since I did not end my study at the classroom door, but
nitatoa hadithi, nitatoa nini? I will tell you a story, also studied language use in society, I found that this interactional
what will I tell? pattern, with high and low tunes used in exchange between speaker
and listener, is also actually used outside school, but in these cases
W: Hadithi Ps: A story
not to hide failure. What I term call–response interaction is one
(. . .) (. . .) feature very often used in some types of speech in Karagwe, such as
Mw: Safi? T: OK? narration, speeches and preaching, in ways that resemble the
W: Kabisa! Ps: Absolutely! interactional patterns in classrooms. In Tanzania, talk-turns are
generally longer than in, for example, Sweden. The speaker may
Mw: Sasa tutasikiliza hadithi. T: Now we will listen to a
‘‘hold the floor’’ for a long time. Also, skills in giving speeches are well
Nitatoa hadithi munisikilize story. I will tell a story and
developed among Tanzanians. In many families, children are
halafu mtatoa, sawa? then you will tell, OK?
explicitly trained to give short speeches. This includes using
W: Ndiyo! Ps: Yes! techniques to activate the listener among which call–response is one.
Mw: Sawa? T: OK? In these contexts, the speaker also calls for the listener’s
W: Ndiyo! Ps: Yes! feedback by using high tone and stress on the last syllable of what
is said. In call–response the speaker calls for the listener to give, for
example, some information in the narration or to give part of a
In the following example it is clearly shown that pupils in lower
classes have difficulties with language. 2
Pictureses is spelled this way to show the double plural constructed by students
Example 7: through mapicha.
 Author's personal copy

Å. Wedin / International Journal of Educational Development 30 (2010) 145–150 149

word, as in the following examples. The first example is from a high expectations and small resources and the hierarchical system
narration. A person is re-telling something that had happened to has resulted in the use of safety strategies among teachers and
him the day before. (A is the narrator and B is the listener). students, in pupils being sorted according to their language
Example 8: background and in low educational outcome overall. The chorusing
that I call call–response, resembles the patterns found elsewhere in
A: Nilipofika niliona nini? A: What did I see when I arrived? poor educational settings. As in the case studied by Hornberger and
Chick (2001), call–response interaction serves to hide school
B: Mbwa B: The dog
failure and at the same time constitutes a hindrance for pupils’
learning as it hides their real level of knowledge. However, I argue
In this example, contrary to the examples from the classrooms, that in the case of Karagwe this interactional pattern at the same
the answer was not given immediately before but is implied in the time could become a potential for education as it helps students
narration. Thus, the listener is required to listen carefully and focus on what is taught, to concentrate, and to memorize what is
concentrate on what is said, to be able to give the right answer. The taught. It could thus provide teachers with tools for handling
following example is from the same narration and here we see an difficult classroom situations if the problem with hiding failure
example where only the last syllable is left out. could be dealt with. The fact that call–response interaction in
Example 9: classrooms in the case of Karagwe builds on interactional patterns
used in the society, gives this type of interactional pattern a
A: Nilimwo A: I saw potential for future development. As children have already met the
B: na B: him pattern outside school, they are likely to gain from its use in
classrooms. However, this demands that other ways to use it,
In this example Nilimwona is one word in Swahili and the part left without hiding students’ real knowledge level, are developed.
out is part of the verb-stem. Here it is clear that in these cases the Research carried out in Australia (Gibbons, 2006, 2008), USA
listener is encouraged to actively take part in the narration. The same (Schleppegrell, 2004) and Canada (Cummins, 2000, 2007), on
pattern is also used in formal religious activities, such as sermons, education in multilingual settings, could be a base for this. These
Bible studies and Sunday school. The evangelist or the priest often studies have shown different ways of enhancing learning in
makes the congregation participate in different ways. In the schools by supporting both development of students’ first
following example, from a Pentecostal church, the priest uses language and the second language that is used in school.
call–response interaction in ways similar to narration in homes. It is a considerable waste of resources that linguistic back-
Example 10: ground is made to constitute a sorting factor for schooled
education, and it is at the same time undemocratic. Large amounts
of resources are ineffectively used in education in Karagwe as so
Mchungaji: Nani anataka Priest: Who wants to cheat us?
much teaching is carried out in a language that students do not
kutudanganya?
understand (Rubagumya, 1994; Blommaert, 1996; Brock-Utne,
Msharika: Shetani Member in the congregation: Satan 2000; Wedin, 2004). Furthermore large numbers of children are
Mchungaji: Shetani Priest: Satan not given a fair chance to learn what they should, as they actually
do not understand what is said during a great part of primary
In Karagwe it is common that local leaders hold seminars for school because the language used is one that they do not
women’s groups or groups of peasants. Here is an example of an understand. When education is regarded as highly important,
experienced local female leader, Ma Eliza, discussing with women when people put so much effort into education, so many teachers
in a rural seminar. One woman has asked what she should do with work extremely hard, so many children spend lot of time in school
a young man she has to support at home. with hard work and so many parents spend more money that they
Example 11: can afford to educate their children, as is the case in Karagwe, then
one should demand that these resources be used in ways that are
Ma Eliza: Mvulana wa miaka Ma Eliza: A young boy of effective and relevant.
kumi na nane ni mtu mzima, eighteen years is an adult, If policy makers would accept the fact that most pupils have
anaweza kufanya nini? what can he do? another language than Swahili as their mother tongue, then the
language question in primary school in Tanzania could also be
Akina mama: Kuoa Women: Marry
open to discussion, and findings from relevant research on second
Ma Eliza: Kuoa Ma Eliza: Marry language education for children in primary school could be taken
into account (Gibbons, 2006, 2008; Schleppegrell, 2004; Cummins,
In this example Ma Eliza uses call–response interaction to make 2000, 2007). Examples of successful teaching, both of a second
the women themselves provide the answer to the question. The use language and academic content in situations similar to the one in
of call–response in all of these examples helps the listener to focus Karagwe, and examples of how children have been successfully
on what is said. It is an effective tool for speakers to make sure that taught in a second language, which Swahili is for children in
the listener concentrates on the topic. Thus, this interactional Karagwe, are many. Findings could then be used as a base for
pattern helps the speaker to get his or her message through and developing language policies, as has been suggested earlier for
also activates the listener by helping him or her focus more on secondary school (Rubagumya, 1990, 1994).
what is said. The main difference between the use of call–response Prevailing interaction patterns that constitute an obstacle for
interaction in school and outside is that it serves to hide failure in efficient learning, could be changed by adapting them to
schools while it serves the function of involving the listener in interactional patterns used in society and the use of the language
what is said outside school. that students master, Runyambo, at least during the early years of
schooling could make learning much easier. As these patterns in
3. Discussion society are used to make listeners pay more attention and focus on
what is said, they constitute a potential for the development of
I conclude that the current policy of Swahili only in primary more relevant and efficient interaction patterns in classrooms. To
schools in Karagwe, together with the considerable gap between help listeners focus on what is said is also very relevant for
 Author's personal copy

150 Å. Wedin / International Journal of Educational Development 30 (2010) 145–150

classrooms, particularly in poor settings where teacher talk is the Cummins, J., 2007. Identity texts. The imaginative construction of self through
multiliteracies pedagogy. In: Garcı́a, O., Skutnabb-Kangas, T., Torres-Guzmán,
main teaching tool. If interaction patterns in classrooms could be M.E. (Eds.), Imaging Multilingual Schools. Languages in Education and Gloca-
adapted so that they still make pupils listen but without hiding lization. Multilingual Matters, Clevedon.
failure and thus preventing teachers and pupils from under- Ericksson, F., 1986. Qualitative methods in research on teaching. In: Wittrock, M.C.
(Ed.), Handbook of Research on Teaching: A Project of the American Educational
standing pupils’ real knowledge level, then teaching could become Research Association. 3rd ed. MacMillan, New York, pp. 119–161.
more relevant and learning could increase. Using Runyambo in the Gibbons, P., 2002. Scaffolding Language Scaffolding Learning. Teaching Second
early years of schooling, parallel to Swahili, could simplify teaching Language Learners in a Mainstream Classroom. Heinemann, Portsmouth, NH.
Gibbons, P., 2006. Bridging Discourses in the ESL Classroom. Students, Teachers and
and learning and thus decrease the gap between expected and real Researchers. Continuum, London.
outcome. Gibbons, P., 2008. ‘‘It was taught good and I learned a lot’’: intellectual practices and
However, I also want to stress that to make this successful, the ESL learners in the middle years. Australian Journal of Language and Literacy 31
(2), 155–173.
developing of new teaching methods should be carried out by
Heath, S.B., 1983. Ways with Words: Language Life and Work in Communities and
policy makers, teachers, teacher educators and linguists in co- Classrooms. Cambridge University Press in Social Anthropology, Cambridge, p. 43.
operation. To copy examples from other contexts uncritically, such Heller, M., Martin-Jones, M. (Eds.), 2001. Voices of Authority. Education and
as education in western schools, should be avoided. To change Linguistic Difference. Ablex Publishing, USA.
Hornberger, N., Chick, K., 2001. Co-constructing school safetime. Safetalk practices
interaction patterns in classrooms is difficult, as they are often in Peruvian and South African classrooms. In: Heller, M., Martin-Jones, M.
unconscious. The problems with the gap between expected (Eds.), Voices of Authority. Education and Linguistic Difference. Ablex Publish-
outcome and real outcome, and the problems with punishment ing, USA.
Lwaitama, A.F., Rugemalira, J.M., 1990. The English language support project in
for perceived failure, should be considered in this context. As long Tanzania. In: Rubagumya, C.M. (Ed.), Language and Education in Africa: A
as teachers and students experience a need to hide failure they will Tanzanian Perspective. Multilingual Matters, Clevedon, Philadelphia.
continue to do that, no matter what interactional patterns they will Mehan, H., 1992. Understanding inequality in schools: the contribution of inter-
pretive studies. Sociology of Education 65, 1–20.
be taught. This means that the first step should be that policy Ndayipfukamiye, L., 1994. Code-switching in Burundi primary classrooms. In:
makers and those in power be made aware of the problem to Rubagumya, C.M. (Ed.), Language and Education in Africa: A Tanzanian Per-
enable a positive change. spective. Multilingual Matters, Clevedon, Philadelphia.
Nyerere, J.K., 1968. Freedom and Socialism/Uhuru na Ujamaa—A Selection from
Writings and Speeches 1965–1967. Oxford University Press, Dar-es-Salaam.
Acknowledgements Rubagumya, C.M. (Ed.), 1990. Language in Education in Africa: A Tanzanian
Perspective. Multilingual Matters, Clevedon, Philadelphia.
Rubagumya, C.M., 1994. Teaching and Researching Language in African Classrooms.
This study was made possible by the support of Högskolan
Multilingual Matters Ltd., Clevedon.
Dalarna, Sweden, and by a generous research grant from SIDA/ Säljö, R., 2000. Lärande i praktiken—ett sociokulturellt perspektiv. Prisma, Stock-
SAREC. I also want to thank two anonymous reviewers for valuable holm.
comments on an earlier version of the article. Schieffelin, B.B., 1996. Making sense of written words in Bosavi. In: Ochs, E., S-
chlegloff, E.A., Thompson, S.A. (Eds.), Interaction and Grammar. Oxford Uni-
versity Press, Oxford.
References Schleppegrell, M., 2004. The Language of Schooling. Lawrence Erlbaum, London.
Smitherman, G., 1977. Talkin and Testifyin, The language of Black America.
Arthur, J., 1994. English in Botswana primary classrooms: functions and constraints. Houghton Mifflin Company, Boston.
In: Rubagumya, C.M. (Ed.), Teaching and Researching Language in African Thomas, W., Collier, V., 1997. School Effectiveness for Language Minority Students.
Classrooms. Multilingual Matters Ltd., Clevedon. NCBE Resource Collection Series, No. 9, George Washington University. http://
Arthur, J., 2001. Code-switching and collusion: classroom interaction in Botswana www.ncbe.gwu.edu/ncbepubs/resource/effectiveness.
primary schools. In: Heller, M., Martin-Jones, M. (Eds.), Voices of Authority. Thomas, W., Collier, V., 2002. Reforming educational policies for English language
Education and Linguistic Difference. Ablex Publishing, USA. learners means better schools for all. The States Educational Standard 3 (1) (The
Au, K., 1980. Participation structures in a reading lesson with Hawaiian children. quarterly journal of the National association of State Boards of Education,
Anthropology and Education Quarterly 11, 91–115. Alexandria, Virginia).
Aziz, A.A., 1981. Strategies for Communication between Teachers and Pupils in a Trappes-Lomax, H.R., 1990. Can a foreign language be a national medium? In:
Rural Malaysian School. Singapore University Press, Singapore. Rubagumya, C.M. (Ed.), Language in Education in Africa: A Tanzanian Perspec-
Blommaert, 1996. State Ideology and Language in Tanzania. Rüdiger Köppe Verlag, tive. Multilingual Matters, Clevedon, Philadelphia.
Köln. van Lier, L., 1996. Conflicting voices: language, classrooms and bilingual education
Brock-Utne, B., 2000. Whose Education for All?: The Recolonisation of the African in Puno. In: Bailey, K.M., Nunan, D. (Eds.), Voices from the Language Classroom:
Mind. Falmer Press, London. Qualitative Research in Second Language Education. Cambridge University
Buchert, L., 1994. Education in the Development of Tanzania 1919–1990. Mkuki na Press, Cambridge, pp. 367–387.
Nyota Publishers, Dar-es-Salaam. Watson-Gegeo, K.-A., 1992. Thick explanation in the ethnographic study of child
Cazden, C., 2001. Classroom Discourse. The Language of Teaching and Learning. socialization: a longitudinal study of the problem of schooling for Kwara’ae
Heinemann, Portsmouth. (Solomon Islands) children. New Directions for Child Development 58, 51–66.
Chick, J.K., 1996. Safetalk: collusion in apartheid education. In: Coleman, H. (Ed.), Wedin, Å., 2004. Literacy Practices In and Out of School in Karagwe—The Case of
Society and the Language Classroom. Cambridge University Press, Cambridge. Primary School Literacy Education in Rural Tanzania. Centre for research on
Cummins, J., 1996. Negotiating Identities: education for empowerment in a Diverse Bilingualism, Stockholm University, Stockholm.
Society. California Association for Bilingual Education, Ontario. Wedin, Å., 2005. Language ideologies and schooled education in rural Tanzania—the
Cummins, J., 2000. Language, Power and Pedagogy. Bilingual Children in the Cross- case of Karagwe. International Journal of Bilingualism and Bilingual Education 8
fire. Multilingual Matters, Clevedon. (6), 568–588.

View publication stats