Rhodes University Logo
Rhodes > School of Languages > Latest News

Does the medium affect the message for students?

Date Released: Wed, 31 October 2012 11:19 +0200

Most South African students in higher education are not being educated in their first language. English dominates the higher educational context, including learning material and the circulation and distribution of new knowledge.

The debate about the language of instruction in higher education is usually reduced to the position that English is an “international” and a “common” language and should therefore be the main medium of instruction. In South Africa, other languages are usually only mentioned when proclamations are made that Afrikaans should be used in higher education to maintain the status of the language, while African languages are seldom mentioned.

Surprisingly, very little research exists on the effects of increased exposure to a second language on students, perpetuating myths or pieces of folk wisdom. One such notion is that English has to be the primary medium of instruction at school if one is to succeed at university. Another is that the use of a second language negatively affects one’s first language, and that learning should therefore ideally take place through the medium of the first language. However, based on my research, neither of these two arguments is necessarily true, and the debate should really be about more than just language.

I investigated the effects of increased teaching and learning in English on the Afrikaans of first-language Afrikaans speakers at Stellenbosch University. I was interested in the effects of this increased learning in English on students’ academic literacy (measured in Afrikaans), their academic achievement and the way in which they self-assessed their language proficiency in both tongues. I looked at two groups, each with either “more” or “less” exposure to English. The choice of the degree words “more” and “less” instead of “no exposure” and “exposure” became particularly important, seeing that “no exposure” and “exposure” became indicative of the type of research that does not work in the South African context.

When measuring the effects of one thing on another, a standard approach is to have an experimental group and a control group; in this case, a group with exposure to English and one without any exposure to English would have been ideal. However, it is impossible to find a group of students who are not exposed to English. There are of course non-students who have had no exposure to English, but including them would have made my research findings unreliable.

Had I decided to compare two groups at different universities, the different settings would have further complicated the research design. When I interviewed a selection of students, I also found out how complicated it is to determine who had “more” or “less” exposure to English. Even in rural areas where Afrikaans is the dominant language, students often went to dual-medium English/Afrikaans schools, had neighbours and friends who were first-language speakers of English or they were exposed to English in religious contexts.

My starting point had to be the assumption that all my participants were bilingual and proficient in both languages, and were exposed to a significant amount of English outside of the classroom. I then had to base my categorisation of “more” or “less” exposure to English on the amount of English input students received in the classroom. I selected a group who had more exposure to English in the dual/bilingual medium option at Stellenbosch University, and compared them with a group (from the same faculty) who primarily received their education through Afrikaans. Even this categorisation was problematic: all the students were exposed to a significant amount of English through teaching and learning material, so even the group with “less exposure” to English in the classroom were still exposed to English through textbooks and academic articles.

When comparing these two groups across their first year and at different other points in time in terms of academic literacy and academic achievement, no statistically significant differences could be found. In fact, the groups were extremely similar.


If, for example, one group’s average mark for academic literacy or academic achievement was 66, the other group’s would be 65. This apparent “no result” can be interpreted in a number of ways. While the lack of clear control and exposure groups can be seen as a shortcoming of the research design, it can also be viewed as a strength because it shows the difficulty of trying to disentangle the effects of one language on another (as well as on other things such as academic achievement or cognition) in a complex multilingual context such as South Africa. A simplistic interpretation of the results would be that it does not matter in which language students are educated at tertiary level.

However, this I believe is not accurate. Students who formed part of my study were all top academic performers at high school, were educated in their first language and had high levels of proficiency in their second language. I would thus rather argue that if all these conditions are met – if a solid foundation is laid in the first language, if students have high proficiency in English and if they are academically strong to begin with – then receiving education through English does not seem to affect academic literacy in their first language (in this case, Afrikaans) or academic achievement negatively.

If these conditions are met, then achievement at school is a far stronger predictor of success at university than is the language of teaching and learning.

However, this result is not be interpreted as English having no effect on Afrikaans whatsoever, as the effects of the second language on other language skills and other cognitive tasks were not measured. Students in my study also still had exposure to their first language (in different measures) in teaching and learning, so all of them also had continued support in their first language. In fact, during interviews, many of the students mentioned that they deliberately switched between their two languages as a learning strategy, which indicated that their bilingualism is a resource to them.

South Africa provides an ideal context to investigate multilingualism and much more research is needed. Only then can we move beyond folk wisdom and myths about the effects of the second language on the first in higher education.

Written by: Dr Marcelyn Oostendorp