An important question in multilingual societies is whether the use of a dominant language
threatens the maintenance of language diversity. South Africa offers a particularly interesting
case study to explore this question. In the post-apartheid era, South Africa has adopted a
language policy which gives official status to 11 languages, including English, Afrikaans and
nine Bantu languages. However, English remains the dominant language of business and
politics, and it is increasingly the language of instruction at schools.
An analysis of the Population Census data for South Africa shows that in 1996, English was
reported as the main home language by less than half a percent of Africans1
, who accounted
for 76 percent of the total population (own calculations, South African Population Census
1996). However, a number of qualitative studies in the past decade have documented the
increased use of English among Africans (De Klerk 2000; De Kadt 2002; Kamwangamalu
2003; Deumert 2006); the preference for English language education among both parents and
students (cf. De Kadt 2005; Dalvit and De Klerk 2005; Bangeni and Kapp 2007); and
perceptions of English as an “empowerment device” (Rudwick 2008: 110) and as “the
language of upward mobility and access” (Probyn 2009: 126). These findings are seen as
markers of a language shift towards English, and there is some concern, expressed both in
academic research and in the media, that the future of the Bantu languages spoken in South
Africa could be at risk.
However, the increased use of English does not have to imply language shift, in the sense that
English is replacing the Bantu languages in both public and private arenas (Deumert 2010).
Rather, English may be used alongside an African home language, a bilingualism that can
remain stable over time and that may be the end product of linguistic contact rather than a
precursor to monolingualism. Many of the studies which investigate language shift in post-apartheid South Africa have
been based on qualitative research about relatively small samples of speakers from
geographically specific areas and demographically specific groups. These studies have been
able to explore the complexities of language behavior and attitudes to language use, but given
their scope, they cannot describe national patterns and trends in language demographics.
Quantitative micro-data on the languages which people speak lack the nuances and texture of
qualitative data, but they provide the means to track broad national patterns in language use
and how these change over time.
In post-apartheid South Africa, the possibility of describing trends in language demographics
at the national level has been limited by the nature of micro-data available. Although the
1996 Population Census collected information on the two languages most often spoken at
home, this information was only collected again in the 2011 Population Census, which was
released publically in 2014. The intervening census (conducted in 2001), and all the
nationally representative household surveys undertaken in the post-apartheid period, asked
respondents to report only on the (single) main language spoken at home.
In this paper, we use census data to describe changes in home language reporting among
Africans from 1996 to 2011. We focus on the reporting of English, as both a first and a
second home language, and we explore whether there is evidence of a trend in English
language use which signals a shift away from the Bantu languages. In the next section, we
briefly review language policy in South Africa after the end of apartheid, and we discuss
studies of language shift in this context. In section 3, we outline and evaluate the national
micro-data analysed in the study (collected in the 1996, 2001 and 2011 Population Censuses).
We present these data in section 4, where we describe trends in first home language
reporting, bilingualism, and overall language use, and in section 5, where we probe how
reporting on English as a second home language varies by age and income class. In the final
section, we summarise and review the main findings from the empirical description.