BOOK REVIEWS BOEKRESENSIES

Somewhat lacking in conceptual and methodological imagination

Louwrens Pretorius

University of South Africa

Lucy Holborn, The Long Shadow of Apartheid: Race in South Africa since 1994
South African Institute of Race Relations, Johannesburg, 2010
220 pp
ISBN 978-1-86982-583-6
R160

"This research attempts to uncover some trends in race relations in South Africa" (p 1). Once upon a time... or was it thrice?

The Long Shadow of Apartheid presents information on events and issues concerning race relations in South Africa. It contains sets of summaries of newspaper reports, comments on the reported events and "racial sentiments" reflected in the reports; transcripts of interviews with public notables; and renditions of selected results from opinion surveys. My focus is on part one, Race in the Press. This comprises "a review of expressions of racial sentiment by decision-makers (very loosely defined) and reports on inter-group tensions as reflected in the print media" (pp 1-2).

The format is somewhat reminiscent of the great tradition of the SAIRR's Race Relations Survey. That was for long - and its current incarnation perhaps still is - one of the more valuable reference sources available to academics, journalists (to whom much of the Surveys' contents were owed) and of course, all manner of "opinion leaders" with an interest in South African affairs. Some 66 per cent of The Long Shadow is based on "print media reports". In this, too, the book follows the tradition of the Survey - although the latter also drew information from other sources (notably Hansard). Quite a number of the Survey editions included lists of publications on its wide domain of concern. The Long Shadow contains preciously few signs of awareness that systematic research with relevance to race relations has been conducted outside of the SAIRR. The most visible exceptions are the opinion surveys ("polls") which are cited in chapter 6.

Perhaps seekers of information on race relations may find the book useful as a source for identifying, for example, contentious events (under the rubric of "inter-racial violence" in chapter 1); identifying the names of persons (individual or corporate) who opined on affirmative action and employment equity (under the rubric "business and employment" in chapter 3); or "expressions of racial sentiment" regarding processes of "re-integration [sic] of the education system" (in chapter 4). Unfortunately the value of the book as a first-stop reference source is diminished by its own limited data-base.

The research project's coverage of print media was, by the author's own admission, not "comprehensive". The reasons were "the sheer volume" of reports "with a racial element" since 1994 and "the desire to uncover general trends" (p 2). However, no reasons are given for limiting the "review of expressions of racial sentiment" to "the English-language print media". The effect of excluding Afrikaans papers from the review was noted in "a small seminar held... to discuss the findings of the research" (pp 2, 233). The implications of such neglect for the plausibility of generalisations about "race relations" in this country should have been glaringly obvious at the beginning of the research - especially to the management of an organisation that has been involved in South African politics since 1929. But it is not only the Afrikaans print media that were neglected. African-language papers are also absent. On the basis of my (admittedly very rapid and thus probably not altogether accurate) scan of the reference lists, it would seem, moreover, that at least 90 per cent of the print media clippings that are cited in The Long Shadow came from Johannesburg-based newspapers. I cannot claim to understand the dynamics of current news-sourcing and sharing (which may justify concentrating on specific newspapers), but I do suspect that there is merit in gathering data from regional papers; and particularly in looking at local papers from towns beyond the metropolitan areas. If there is a reason for limiting data sources to English print media from the heart of Gauteng, we should be told what it is. Even first-stop reference sources must be clear about all the limitations of the primary data from which they were constructed.

Unfortunately books are likely to be judged not only on their potential uses but also by their pretensions. Instead of presenting The Long Shadow as a compilation of summarised information, the author claims that "This research attempts to uncover some trends in race relations in South Africa" (p 1, my italics). That is a purpose that imposes some onerous conceptual and methodological burdens.

The extent to which the book reflects achievement of that aim will depend on each reader's understanding of the italicised terms. I read them from the position of a political sociologist with a more than passing interest in the arcane arts known as "theory" and "methodology". Judged from that position, I have to say that The Long Shadow does not reflect awareness of the rich extant literatures on race relations and related themes. Neither does it reflect knowledge of a variety of sophisticated methods for generating data from print media for the purpose of research on the representation of race and race relations in the media. Short book reviews do not allow space for even a few lessons in Methodology 101 (not to mention Sociology of Race 101), but since this book review appears in an academic journal, a few methodological comments are required.

The labelling of social phenomena is complex. Generalisations that are presented as products of "research" must be framed in terms of concepts with meanings that are clarified for the benefit of the intended audience - and because systematic research and persuasive argumentation is impossible in the absence of clarity on the meanings of key concepts. What should a reader expect from a report that begins with the assertion: "The measurement of race relations is always going to be tricky, partly for want of knowing what to measure and how to measure it" (p 1), but then continues regardless to "uncover some trends in race relations"? Not only that, but the report also reflects the assumption that concepts such as "racial sentiment" and "inter-racial violence" are unproblematic; and that there is a possibility that press reports on contentious topics that are called "inter-racial violence" and "affirmative action" in South Africa might just not reflect "racial sentiment" or "racial motives".

Lack of conceptual discipline may work for press reporting and for the internal purposes of hybrid research-lobbying organisations such as the SAIRR, but it cannot pass as authoritative "research". As the author of The Long Shadow clearly knows, if the object cannot be defined, it cannot be "measured". How then, was it possible to assess, for example, "trends" in the "extent" to which "instances of inter-racial violence" were accompanied by "expressions of racial sentiment"? (pp 1, 2) And what are we to make of the label "decision-makers (very loosely defined)"? (p 1)

Even if we all agree that the meanings and manifestations of say, "racial sentiment" are immediately recognisable in all manner of press reports, the formulations of the research purposes imply the need for rules for the categorisation and counting of the reports - and also of the actors, expressions of "racial sentiment", and "incidents" to which the reports refer.

No definitions, measures, counting rules or - indeed - of estimated numbers are to be found in this book; except, that is, for the occasional citation of the number of persons killed or beaten and the number of perpetrators that were involved in specific incidents.¹ In fact, the very meaning of the term "trend" is left to the imagination of the reader. The summaries of issues and events are, I must admit, ordered by type and times. However, readers with some familiarity with the difficulties involved, for example, in categorising types of violence and the associated circumstances, motives and observers' comments, may doubt the value of the superficial nominal classifications that are used in the book.

In the absence of even a modicum of methodological illumination, the conclusions of "research" will lack credibility. The most striking weakness of the book is that it reports no attempts to count (literally) the objects of interest - despite the fact that it contains explicitly quantitative claims. Consequently, the claims cannot be assessed. For example:

Another quotable consequence of not knowing how to measure "trends" - and perhaps not caring about the significance of words such as "significance" and "quantity" - is the following "conclusion":

Then there is the occasional banality: "Farm violence during the period aroused racial sentiment reflected in the print media" (p 20).

I do not hold a particular brief for or against either quantitatively or qualitatively-oriented research. In fact, I believe that race relations research and research on other social issues require close integration of reliable counting with deep probing of people's motives and beliefs. The social sciences possess a large array of sophisticated (and not all of it excessively resource intensive) methods that facilitate the employment of both orientations for the analysis of print and other media. They go under the collective label of content analysis and they have many uses. The uses include, for example, the study of representations of race and racial motives in the media and the generation of data about the dynamics of contentious events from news reports.

Perhaps the SAIRR and other organisations that collect and collate newspaper clippings and want to transform those into published research should take note of such methods and of the extant literature on their themes of choice. Research managers and consultants have a serious responsibility in this regard. The Long Shadow concludes with a report on "some of the issues raised at a small seminar... to discuss the findings of the research" (p 2). The contents of the report indicate that the attendees include at least three of the most experienced and respected social scientists in South Africa (pp 231-237). All three also have more than occasional relationships with the SAIRR. Hence I would have expected a research product that reflects a higher level of conceptual and methodological imagination (not to mention tighter expository writing - which is another reviewable issue) than that which is revealed in this book.

1. There are many percentages in chapter 6 - which summarises findings from opinion surveys. But my concern is with the conclusions from the "review of expressions of racial sentiment... in the print media".