Mother-tongue education is best

Most African countries have embarked on new education policies and curriculum changes after UNESCO?s 1990 World Conference on Education For All. Yet education success for the majority of children remains elusive.

A draft report of a study on language education research was recently debated at the Conference on Bilingual Education and the Use of Local Languages in Windhoek, attended by education officials and experts from 19 African and 5 European countries. The draft report was compiled by organisations such as UNESCO Institute for Education, the Association for the Development of Education in Africa, and the Deutsche Gesellschaft f?r Technische Zusammenarbeit.

A final report will be submitted to the Association for the Development of Education in Africa Biennale in Gabon in March 2006.

The study shows that although language education policies diverged across sub-Saharan Africa during the colonial period, post-colonial developments are showing a remarkable convergence towards one language model. Tanzania, Ethiopia and Eritrea are the only countries with other models.

The former French, Portuguese and Spanish colonies used only the colonial language in education. Post-independence shows increasing movement from zero mother-tongue education to between one and three years of learning in the mother tongue, followed by a transition to French and, in Mozambique, Portuguese.

In contrast, African languages were transcribed by missionaries in the former British colonies and used as the medium of instruction in education for four to six years. Since independence these countries have tended to whittle away at mother- ongue education (MTE) and either eliminate it or reduce it to a maximum of three years.

Today, with a few exceptions, there is a convergence towards similar language education models across sub-Saharan Africa. In most cases children receive up to three years of MTE, followed by a switch to education in the former colonial language.

But research in psycholinguistics and secondlanguage acquisition shows an inherent design flaw in such models. Children are expected to learn through a second language before they have developed sufficient proficiency or competence in this language to make this possible.

One of the reasons for the design flaw is that language education models used in African settings have their origins in second-language programmes designed in Europe to teach students a second language, such as conversational skills, writing tasks and some literature ? not to prepare them to learn mathematics, science, geography or history through the second language.

A second reason is that the language programme designers have not kept up to date with contemporary research between the cognitive development of children and language learning, and how children use language to learn all areas of the curriculum. And lastly, because these programme designs do not originate in African settings, they do not accommodate the multilingual nature of the continent.

It is almost impossible for pupils to learn enough of the second language in three years to switch to a second-language medium of instruction by Grade 4. In countries where there are well-trained teachers and sufficient classrooms and schoolbooks, children usually need between six and eight years to learn a second language before they can use it as a medium. This means under optimal conditions they should not switch language medium before Grade 7.

In less well-resourced schools, the research evidence shows that it may be possible to switch medium in Grade 9. Switching medium several years earlier results in educational failure, as countless studies demonstrate.

The draft report sets out an analysis of a vast set of apparently incontrovertible data that are likely to shock governments across the continent. Studies from Africa range from the earliest internationally significant study of bilingual education which took place in South Africa in the late 1930s, to a Nigerian study in the 1970s, an HSRC study headed by Carol Macdonald in 1990 and more recent studies in Niger, Mali and Ethiopia. Longitudinal studies of bilingual and English secondlanguage programmes in Nigeria and North America were cross-checked against many shorter-term studies to establish what education planners can expect from different types of language education models.

Insignificant differences in literacy development among different language models are evident between Grades 1 and 3. Any model that is implemented by well-trained teachers and accompanied by good teaching materials will do better than a model that lacks these resources.

But when children are tracked over a long period from Grade 4 onwards, significant gaps begin to appear between children who continue MTE and those who have switched to a secondlanguage medium. We can now predict that most students who switch from mother-tongue medium to another language by Grade 4 are likely to achieve only between 30% and 40% in their second language by Grade 12, even though they seem to have had longer exposure to this language.

In contrast, students who have learned in the mother-tongue medium for at least six years are likely to reach 50% or more in the second language. Those who have MTE throughout (Grades 1 to 12) plus the second language taught as a subject by a teacher who is proficient in the language are likely to achieve 60%.

So, despite popular wisdom, the longer students have MTE plus well-resourced second language as a subject, the better they will perform in this language, and are more likely to achieve well in mathematics, science and their own home language than those learners in models with an early transition to an international language.

The report also dispels the notion that MTE is more expensive than second-language education.

Early transition to the second language is the most expensive to resource, since it requires all teachers to develop native-like proficiency in the second language to teach well in it.

In South Africa, it would mean about 200 hours of English second-language development tuition for 200 000 teachers. If a country were to opt for this, the return would still not be worth the investment. Nowhere has it been demonstrated that a mainstream education system can be successful if it is based on the second language, particularly when this language may in fact be the child?s third, fourth or even fifth language.

The most economical scenario is to equip a corps of teachers with what is known as ?native or near native-like proficiency? in English to teach English as subject language specialists. This corps would comprise about 15% of the teachers from Grades 4 to 12. The return on investment for this model promises far greater rewards, both economically and in terms ofeducational outcomes for students.

A compromise would be to have MTE in Grades 6 to 8, followed by the transition to English. The returns are high but the costs are greater.

The draft report can be traced electronically as follows: http://www.adeanet.org/meetings/en_Aug-locallanguages-2005.htm

Dr Kathleen Heugh is a Chief Research Specialist in the Assessment, Technology and Education Evaluation Research Programme. She was one of the principal researchers of the study on language education.