Language in South Africa (LinSA) is a very handsome book, beautifully edited, carefully proofread, and produced on thick paper in elegant fonts. It is in fact the same book, although revised and updated, as Language and social history: Studies in South African sociolinguistics (Mesthrie 1995). Just looking at the two volumes, side by side on my desk, I could write an essay on publishing and face validity. I am happy that this book has found an international publisher, because it deserves wider reading and better promotion (I never saw the first book reviewed or promoted), but the easy conclusion that the book under review is a better book is not necessarily warranted. As the Irish say about their horses, handsome is as handsome does, and both volumes do handsomely indeed.

In fact, I prefer the title of the first volume. The irrefutable fact is that you can understand nothing about the language(s) situation in South Africa (SA) unless you are keenly aware of its social history: The genesis and history of Afrikaans is social history; the relationship, ever-changing, ever-ongoing, of Afrikaans and English and of their speakers is social history; the basically tribal (a politically incorrect term never used in the present book) relationship between Inkatha and the ANC, between Zulu and the other eight official African languages, is social history; and of course, above all, the relationship among white, colored, Indian, and black, the former apartheid official categories that are now history, is social history, without an understanding of which nothing makes sense of the role of language in South Africa. Just how to write these four categories is controversial, and Mesthrie carefully explains the rationale for his choice in the introduction. Not really discussed is that they never made any sense. I was amazed to hear that you could apply for a reclassification which, if granted, was published in a long list each year; or that Japanese were white and so could frequent white restaurants and hotels, while Chinese were not – presumably a decision made on the basis of volume of trade.

None of apartheid made sense, and it was indeed an abomination, but it does not follow that all Afrikaners should be blamed for it. The contribution of Afrikaner linguists in the rebuilding of South Africa is noticeable for their dedication, but I get very little sense of it in this volume; they are mostly ignored, and so is the process and excitement in the rebuilding of SA as a rainbow coalition. Keith Chick, one of the contributors, in another piece (“Constructing a multicultural national identity: South African classrooms as sites of struggle between competing discourses,” 2002), does catch the difficulty and excitement of what is happening about, around, and in language in SA today, at the same time as he makes clear (as he says himself) the tendency that the more it changes, the more it stays the same. This is an enormously significant and problematic time in the social history of SA, and for better but probably worse languages index all the injustices and values of past racial, tribal, ethnic, and ultimately economic relationships and so signal future tensions, especially between Afrikaans and English. Afrikaners are concerned about the future of their beloved language, with English at present taking over all domains, including schools, government, courts, and even the army. Mesthrie knows all this better than I do, but I find no sense of it in his book (as the reader can, for example, in Deprez & du Plessis 2000). I think there are two reasons for this. The first version of this volume was published in 1995, which means it was written in 1993–1994 (or in some cases even earlier), before this struggle of reconstruction had really begun. The more significant reason is probably the genre, which is more that of a “Handbook of languages in SA,” with the focus mostly on linguistic description with attention to social factors rather than on the process and relationship among the languages – an awkward way of saying “among the speakers, white, colored, Indian, and black.”

Therefore, rather than criticize LinSA for what it is not, let me attempt an account of what it is. LinSA consists of an “Introduction” by the editor; 24 chapters in three parts; three indexes (names, languages, and subjects); and wonderfully, many maps.

Part I, “The main language groups,” begins with a very good sociolinguistic overview by Mesthrie, followed by chapters on the Khoesan languages, the Bantu languages, the origin of Afrikaans, South African English, South African Sign Language, Indian languages in SA, and (somewhat out of place) German speakers in SA, a chapter that once again documents shift among an immigrant group and is of not much interest in this connection.

It seems to me rather strange that there should be so much disagreement about the origin of Afrikaans. Paul Roberge gives a brief history with dates and an objective, informative overview of the various positions – the superstratist hypothesis, the variationist/interlectist hypothesis, and the creolist hypothesis – of which he favors a reformulated creolist position. It also seems strange that no one in this entire volume, where terms like “creole,” “semi-creole,” “koine,” and so on are frequently bandied about, mentions the theories of the French creolist Robert Chaudenson for Réunion Creole. Simplified greatly, he holds that the creole began life as an L2 version of French, but as the number of slaves increased and fewer people had access to mother-tongue speakers of French, the L2 began creolizing over several generations – and without a pidgin stage (Mather 2000). The same conditions seem to fit Afrikaans quite well, and Chaudenson should at least receive a nod, especially as his notions seem compatible with Roberge's own.

Totally missing is any consideration of the standardization and the taalstryd, or “language struggle” for the official use of Afrikaans in education and government, which explain much about Afrikaner language attitudes. I am not saying this should have been in Roberge's chapter, but that it should have had a chapter of its own. Another missing topic I would have liked to see included is that of South African Jews, rarely mentioned in the literature. SA has a population of about 90,000 Jews (Theo du Plessis, personal communication, 2003), many of whom are involved in the diamond trade. We know that many of the Jewish members of the diamond trade in New York City and Rotterdam and Amsterdam are ultra-Orthodox and Yiddish speakers, a nice example of ethnic language maintenance as a result of the effectiveness of the double boundary maintenance of language and religion, similar to the case of Pennsylvania Amish. I know that this SA population is English-speaking, but what else? And what is their role in the construction of a multicultural national identity?

Aarons & Akoch's chapter, “South African Sign Language: One language or many,” is a welcome addition. It does not really answer the question of the title, but it is a clear, sound introduction to Deaf culture and the role of language, and it applies equally to American Deaf culture.

The other chapters in Part I are basic language description, exemplifying William Labov's point that you can't adequately do linguistics without considering social factors, and they are more suitable for reference than for understanding the present language situation in SA, but of course a very necessary first step.

Part II, “Language contact,” the major part of the book, is divided into four subparts: “Pidginization, borrowing, switching, and intercultural contact,” “Gender, language change and shift,” “New varieties of English,” and “New urban codes.” It is clear from all these chapters that there is a lot of language “mixing” taking place, and that the processes, functions, meanings, and kinds of this mixing are poorly understood and documented; virtually all the chapters point out the need for a new direction of research, or simply more research. In this context, it is illustrative to cite a passage from Sarah Slabbert & Rosalie Finlayson's chapter, “Code-switching in South African townships,” where they in turn are citing the work of David Brown (1992:71):

This is virtually the only passage in the entire volume that hints at the existence of Critical Theory. Besides a critique of modernity and a commitment to the centrality of discourse in constructing social order, the two characteristics Philip Smith outlines for this school of thought are the belief that “theory is part of a moral and political enterprise” and a “recognition of the perspectives, voices, and cultures of subordinate groups” (2001: 233). To this I would add a frequent romantic idealism concerning specific remedies in specific language situations.

If there is one country in which the language situation lends itself to a critical analysis, it must be SA, but it is remarkable by its absence in this very conservative work. Chick 2002, mentioned above, or a quick perusal of the latest AILA Review, Africa and applied linguistics (Makoni & Meinhof 2003), gives a very different view. I for one would rather have romantic idealism than stoic description of languages that ignores the speakers.

Finlayson's “Women's language of respect: Isihlonipho sabafazi” (granted, in a chapter revised from an article first written in 1984), offers exactly what she criticizes now: a description of exotic language use, women's avoidance of the same syllables as found in their in-laws' personal names. This usage is now disappearing in urban environments – “In this modern world of ours, there appears to be no time for the finer details of customs of respect” (294) – but if you consider the social circumstances for its continuance (“a patriarchal ordering of society uninfluenced by any feminist perspectives or demands” [Dowling 1988:145 cited in Finlayson 294]), I say hurrah and good riddance. I have long questioned linguists' dismay at the disappearance of language use and even languages, considering the social circumstances that speakers must remain in to continue that use.

Robert Herbert's “The political economy of language shift: language and gendered ethnicity in a Thonga community” is one of my favorite chapters, but I am partial both to functional explanations and to women who stand up for themselves. In brief, Thonga men's shift to Zulu – in the area just south of the Mozambique border – follows the norm of shift to the language of prestige and power, while “the women's non-shift to Zulu and their historic maintenance of Thonga has gone largely unexamined” (320). These women have now, under some duress, given up Thonga but speak “very bad Zulu” (according to Thonga men), and Herbert asks, “What are the rewards associated with very bad Zulu?” Part of the theoretical interest of this question lies in the role women play in language maintenance and shift, which is still far from clear. In some cases, women are more conservative and maintain the language the longest, but in other cases they originate a shift. Herbert's explanation is simple and elegant and very convincing. A Zulu woman is a “perpetual minor” who moves at marriage from the control of her father to that of her husband; Thonga women, in contrast, traditionally enjoyed a number of rights and showed a great deal of independence (331). “For women to embrace Zulu identity, ethnicity and custom as their men do would involve a marked diminution of their status and power” (331), so the women maintain their Thonga identity – and independence – and mark this nonconvergence of male and female domains with maintenance of “very bad Zulu.” Three cheers for Thonga women!

The last two chapters in Part II – “An introduction to Flaiitaal (or Tsotsitaal)” by K. D. P. Makhudu, and “Language and language practices in Soweto” by Dumisani Krushchev Ntshangase, both very brief – return to the topics of code-switching, mixed languages, and convergence or harmonization. I am purposefully vague, as are the authors, because it is clear that much basic linguistic description and analysis is needed to clarify the large-scale language change taking place in South African townships (253), even in regard to something that should be fairly obvious but apparently is not: whether Iscamtho is the same language variety as Flaaitaal, a point on which Makhudu and Ntshangase disagree.

On the same topic, Slabbert & Finlayson point out that, in their code-switching data from Botshabelo, it is impossible to distinguish a matrix language based on the norms of Carol Myers-Scotton's (1993) matrix language frame model, and they state that it is difficult to classify the discourse as code-switching between dialects or languages, or as a new interlanguage (253). Makhudu wonders whether Flaaitaal, like Creole languages, will stabilize into a first language. The underlying cause of the fluidity of the language situation has a positive aspect, a cooperative multilingualism: “The important point is that people in the townships are prepared to accommodate each other and believe that it is important to do so because the issue of communication is at stake” (Finlayson et al. 1998: 403).

Part III, “Language planning, policy and education” (LPP), is the topic I find most significant from the viewpoint of reconstructing SA as a rainbow nation, of the “prospects of South Africans being able to construct a truly multicultural national identity” (Chick 2002:476). But in spite of all planning, we find, de facto, what Timothy Reagan elsewhere calls “the overwhelming dominance of English” (2001:63); in spite of all planning for mother-tongue education for all, we find black African insistence on English-medium education. In the educational institutions, however, the administration does have some control over language use in textbooks, test-taking, language proficiency of teachers hired, medium of instruction, and so on. Schools can be quite inefficient in maintaining a language in a shift situation; on the other hand, they can be enormously efficient in providing access to the language of choice, and the black African parents' choice par excellence is English.

Reagan, a frequent writer on LPP and education in SA, has another fine chapter on basic issues, demonstrating his customary sound scholarly judgment in an excellent overview. Sarah Murray's “Language issues in South African education: An overview” is interesting for an update on educational language policies, but it again demonstrates a need for more explicit description of language use in the classroom. She presents no data, but having observed untold bilingual classrooms, I very much doubt that the teachers code-switch. As teachers, they will have a low opinion of Flaaitaal and similar varieties. What bilingual teachers do in the classroom when their students don't understand is to translate into the L1 of the children, which is not the same as code-switching.

What is happening in South Africa today is societal change from a caste-like social stratification to one based on social class. One of the more interesting phenomena is the growing middle class of people of color and the concurrently growing number of private schools. It is exactly the educational institutions that will be one of the major mechanisms for facilitating this new social mobility: “The emerging class story comes out most clearly in the middle-class black youngsters we have at university at present” (Albert Weideman, personal communication, August 2003).

For this reason, for better or worse, the importance of language-in-education planning cannot be exaggerated for the future of SA, for government and the legal system, for trade and economy, for quality of life. Curiously, not one of the three chapters in Part III refers to the growth of the private education sector and its likely role in the transformation of the new SA.