South African History textbook research – a review of the scholarly literature

School history textbooks are seen to embody ideological messages about whose history is important, as they aim both to develop an ‘ideal’ citizen and teach the subject of history. Since the 1940s, when the first study was done, there have been studies of South African history textbooks that have analysed different aspects of textbooks. These studies often happen at a time of political change (for example, after South Africa became a republic in 1961 or post-apartheid) which often coincides with a time of curriculum change. This article provides an overview of all the studies of South African history textbooks since the 1940s. We compiled a data base of all studies conducted on history textbooks, including post graduate dissertations, published journal articles, books and book chapters. This article firstly provides a broad overview of all the peer-reviewed studies, noting in particular how the number of studies has increased since 2000. The second section then engages in a more detailed analysis of the studies that did content analysis of textbooks. We compare how each study has engaged with the following issues: the object of study, the methodological approach, the sample of textbooks and the theoretical or philosophical orientation. The aim is to provide a broad picture of the state of textbook analysis studies over the past 75 years, and to build up a database of these studies so as to provide an overview of the nature of history textbook research in South Africa.


Background and introduction
The production of textbooks, from conception to distribution to use, is a politically and educationally contentious activity.In light of the above, "the politics of the textbook" is never far removed from the public sphere as three cameos from the South African context will reveal: The Inkatha Freedom Party publically burned copies of an Oxford Grade 12 History textbook arguing that their leader, Mangosuthu Buthelezi, was portrayed in an unfavourable light (Wassermann, 2009); in his 2011 state of the nation address President Jacob Zuma emphasized the key role of textbooks when he foregrounded them amongst the three Ts (teachers, textbooks, and time) needed to be prioritised in uplifting the education system (SA News, 2011); the so-called "Limpopo textbook scandal" whereby some schools in the province did not receive textbooks caused a public outcry and the angry public response that followed forced the government to act, not only in Limpopo, but also in the Eastern Cape (Chisholm, 2013).
In the context of the above, textbook research in general, and History textbook research specifically, is understandably not a recent phenomenon.In fact, since the First World War (1914War ( -1919)), History textbooks have been studied as powerful sources of educational media with the ability to shape the views and consciousness of generations of learners.Research in this regard was not only conducted by individual authors but also driven by international organisations such as the League of Nations and UNESCO.Most of the subsequent research into History textbooks has been centred in Europe and North America but pockets of investigation can also be found in the Far East, Australasia and South Africa.As South African academics involved in textbook research, and the supervisors of post-graduate students studying the genre, we have read much of the local literature as well as that emanating from other parts of the world.We began asking a range of questions about the growing field of History textbook research in South Africa, including: What research has been done in this regard?Who has been doing the research?When was it done?What was the focus of the research?Which methodologies were employed?These questions were the catalyst and motivation, both on a professional and a scholarly level, for us to do three things in this article.
Firstly, we conducted a search of the literature in order to map the completed scholarly research into History textbooks in South Africa in a database.In the process we hoped that the "big picture" of scholarly work done on History textbooks in South Africa would emerge.We deemed a literature search to be a necessary initial step since, to our knowledge, no study exists that provides an overview of the studies already done on History textbooks.The database was then to be subjected, for the purpose of this article, to two levels of analysis.To begin with we answered the following broad questions -who did what (author), when (time), where (location), about what (focus), how did they do it (methodology) and why did they do it (motivation)?Secondly, based on the trends and patterns that emerged from the broad overview, we extracted the studies which had engaged with analysis of History textbooks and drilled deeply into these works by examining the sample size, object of study, the methodologies and the theoretical/philosophical approach.Thirdly, the review yielded a further outcome, namely a database of research conducted on History textbooks in South Africa.We have included this database as Appendix A at the end of this article to allow other researchers access to what we deem to be a workable database which they can challenge, alter, use and expand.
In what follows we will unpack the research methodology employed in doing this study, the quantitative "big picture" analysis and the qualitative deep drilling into articles that actually analysed History textbooks.The article will then be concluded with a discussion of the findings and the database contained in Appendix A.

The research methodology of reviewing the literature on completed studies on South Africa History textbooks
According to Boote and Beile (2005), a literature review transcends the simple search for information as it is a consideration of scholarly works relating to a specific study -in the case of this article, to completed scholarly research on South African History textbooks.The extensive literature search we conducted was thus a means of building on previous knowledge by using existing knowledge (Creswell, 2009) and in so doing reaching an understanding of what came before.What we did, therefore was a "scholarship review" (Mouton, 2001:87) starting as far back in time as we could go.
We started the literature search by drawing on the professional knowledge we had on scholarly work already completed on South African History textbooks.This we tabulated under the following headings: author; year published; title of publication; place of publication and nature of publication.We then employed a student assistant to do a thorough desktop search of all databases subscribed to by the University of KwaZulu-Natal library system as well as a general Internet search.The key words used for the search were "history textbooks" and "South Africa".However, in cases of uncertainty where it was felt that the key words were not serving us well, literature was skim read to ascertain its appropriateness to the study.In a cross-checking exercise we also searched the available databases ourselves.This served for example, to capture theses and dissertations written in Afrikaans that were previously overlooked.Additionally, being aware of the vagaries of desktop searching, we approached Rob Siebörger, Katalin Morgan and Elize van Eeden, all fellow academics involved in some way or another in History textbook research, to scrutinise our evolving database.Their critique of and additions to the database proved invaluable.
We grappled with the questions of what to include and what to exclude and how to justify such choices.After some debate we settled on the principle that literature reviews are about peer reviewed scholarly work -that is works vetted by fellow academics for their suitability and adherence to the expected standards of the discipline.This principle was married to our key search terms "history textbooks" and "South Africa".
Using the criteria of peer reviewed scholarly works on South African History textbooks meant certain publications were excluded.These included newspapers, popular and professional articles 1 , conference presentations and conference proceedings and Honours projects. 2.This was based on the fact that the Department of Higher Education and Training views Masters and Doctorates but not Honours projects as research publications.Conference proceedings were excluded quite simply because it proved difficult to ascertain which ones were truly peer reviewed and thus of scholarly standing and which were not. 3 Furthermore, since we have focussed somewhat pedantically on research which contained the terms "history textbooks" and "South Africa", scholarly works which did not directly use these keywords were also not included as we regarded them as peripheral to our focus. 4The data base thus comprised postgraduate scholarship in the form of dissertations and theses, journal articles, books and book chapters.These three genres were used to organise our database in a chronological manner starting with the earliest works.In the process the database kept on evolving as we became aware, sometimes by chance, of other scholarly works on South African History textbooks.Within these parameters Appendix A -"Database of Studies on South African School History Textbooks" was developed.
The next methodological step was the quantitative analysis of the literature as captured in the database.This was done by engaging with who did what, when, where, about what, how did they do it and why did they do it?The when was done by tracking the dates of publication/completion of a scholarly work included in Appendix A. This was supported by the what (genre of research) which tied in with the time (when) and spatial (where) frames.This was followed by engaging with who undertook the research.As part of this first level of analysis we tried to theorise why research on South African History textbooks as part of the "big picture" happened.For the most part the answers to the above research questions are presented graphically, statistically and discursively in a blended manner.
The second level of analysis consisted of extracting the South African History textbook studies from Appendix A which had engaged with content analysis of History textbooks.We then examined the object of study, sample size, the methodologies and the philosophical/ theoretical orientations of these studies.The abstracts/summaries of the whole book, dissertation/theses or journal article were analysed.These were usually excellent units of analysis although we did come to accept that this method had its flaws as some abstracts/summaries were flimsy and revealed little.The results of the second level of analysis are presented analytically in a narrative style.

The BIG picture of South African History Textbook research -1944-2015
According to our literature search a total of 65 peer-reviewed academic research works on South African School History textbooks have been completed thus far (Appendix A).Of these 25 were postgraduate degrees leading to higher academic qualifications; 19 were Masters and six Doctoral degrees.A further 10 were books or book chapters and 30 were peer-reviewed academic articles aimed exclusively at enhancing the existing knowledge base.However, for a more nuanced understanding beyond the mere figures it is necessary to view the 65 academic research works in a temporal context as outlined in Tables 1 and 2 below.Scholarly research into History textbooks in South Africa started in the 1940s with a Masters dissertation which was the first ever such research endeavour in 1944.This was followed by an article and two postgraduate studies in the 1960s one of which, The power of prejudice in South African education: an enquiry into history textbooks and syllabuses in the Transvaal high schools of South Africa, by FE Auerbach, was published as a book in the same decade.This sudden mini-boom in History textbook research in the early 1960s coincided with South Africa becoming a Republic outside of the British Empire and the subsequent stronger emphasis of an Afrikaner Nationalist historiography in school history.After this, the field of History textbook research returned to its static state and during the 1970s and 1980s, only one dissertation, two books and three academic articles saw the light of day.Thus, during the oppressive heyday of Apartheid, in which History textbooks were dominated by an Afrikaner Nationalist historiography, little reason seemed to exist to engage in scholarly research in History textbooks.This was so because the Afrikaner Nationalist agenda allowed little critique or critical engagement with the texts that learners studied and teachers used to teach.
The dawning of democracy in the 1990s also left its mark on History textbook research and the greater openness birthed ten studies -one less than was completed during the previous five decades.The end of Apartheid created a belief and optimism that school History could be reimagined and remade (Siebörger, 1994(Siebörger, , 1995;;Bam & Visser, 1997).There was a growing realisation of the power of History textbooks as educational media that could profoundly influence society.The research momentum into History textbooks continued into the 2000s as can be gleaned from Tables 1 and 2. During this decade (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009), 19 studies, two less than the combined efforts of the previous 60 years, were completed.These 19 studies also serve to mirror the strengthening grip that the knowledge economy was starting to exercise on academics, for nine of the 19 studies were now peer reviewed academic articles -almost twice as many as during the previous six decades.The growth trend in School History textbook research continued into the 2010s, a decade that is but six years old.During these six years, 25 peer reviewed scholarly works, 14 of them academic articles, were published.Particularly noticeable in this time period (see Table 1) is the escalation in the production of academic articles to the detriment of books and book chapters.This escalation, alongside that in postgraduate work with ten Masters Degrees being awarded for studies on History textbooks since 2010, is indicative of what is expected from academics in the current context -greater research and supervision outputs.What is particularly noticeable since 2000 is the decline of books on History textbooks in South Africa with the last book addressing the topic appearing in 2011 which is Nishino's reworked 2006 Ph.D. being published.This tendency is to a certain extent a fall-out of the growing neo-liberal knowledge economy in South African higher education whereby financial incentives and rewards in terms of subsidies for books are similar to those for accredited journal articles.One possible result of this post-1994 was an upward turn in the output of academic articles in South African History textbooks as the dominant research genre in this field.In short the post-Apartheid societal change heralded an increase in the research into History textbooks with academic articles predominating.However, what must be pointed out is that a substantial number of the academic articles that appear during this time, probably as many as twelve out of the 30 published articles (40%) having roots in dissertations and theses.This is not only indicative of the requirement for academics to publish journal articles but also the relatively strong relationship between published journal articles and postgraduate studies.
The 30 academic articles referred to above, bar two, by Carpentier (2000) and Lieven (2000), were authored by South African-based academics.The journals that attracted the largest number of these articles (five) were the South African Historical Journal, a publication that specialises in History followed by Yesterday & Today, a History Education journal, with four articles.Overall, the 30 articles on History textbooks generated since 1962 appeared in 17 different journals of which eight are international in origin and nine South African.These journals covered fields ranging from Education and History to Archaeology and Qualitative Research.The above provides a sense of interdisciplinarity in South African History textbook research.
English dominated as the language of publication with only two articles appearing in other languages -an Afrikaans article by De Wet (2001) and one in French by Carpentier (2000).All books and book chapters were also published in English.The dominance of English as a research language was also mirrored in the completed dissertations and theses with only three authors, Raubenheimer (1944), Du Plooy (1965) and Schutte (1990) completing their studies in a language other than English, namely through Afrikaans.
It would, however, be myopic to attribute the change in research patterns into History textbooks to socio-political forces only as individual researchers also played a significant role in increasing the research post-1994.In this regard the work of two scholars stands out, namely Siebörger (1994Siebörger ( , 1995Siebörger ( , and 2006) ) and Morgan (2010Morgan ( , 2011Morgan ( , 2012Morgan ( , 2013Morgan ( , and 2014)).The former had contributed to three books or book chapters and two accredited articles and the latter a Ph.D. (2011) and nine accredited articles.Thus between them they have had an intellectual hand in 15 or (23%) of the studies contained in Appendix A. The remaining books, book chapters and accredited journal articles were shared amongst a total of 28 authors all of which, except for Dean, Hartman, Katzen (1983), Carpentier (2000) and Lieven (2000) and Nishino (2006Nishino ( , 2011) ) are South African based.Thus the scholarship on South African History textbooks is generally dominated by South Africans, except for books and book chapters.
In terms of the institutional affiliation of those who had completed dissertations and theses on South African History textbooks, no clear patterns emerged from our analysis of the compiled database.In total the postgraduate scholars came from nine different South African universities and the University of Western Australia.However, eight postgraduates completed their studies at the University of KwaZulu-Natal since 2009, which points to an emerging institutional leadership in scholarship on History textbooks.This was brought about by several factors including an interest in textbooks spurred on by engagement with the Georg Eckert Institute for textbook research in Braunschweig, Germany and increasing administrative demands around securing ethical clearance for research involving human subjects.
An overview of the literature data base shows that there has been an increase in scholarly publications in the last fifteen years.From the 1940s up to 1999, there were 21 publications on history textbooks, which increased to 44 publications between 2000 and 2015.We attribute this to the postapartheid social and political shifts, as well as to a growing neo-liberal hold on universities which emphasises academics' measurable accountability in terms of publications and student graduations.

Detailed analysis of studies which analyse history textbooks
The next section of the article presents a more detailed analysis of the studies that analysed textbooks, and does not include publications which describe the role of the textbook in schools, or the making, selection and distribution of textbooks, or theoretical and methodological issues (these publications are shaded in the database in Appendix A).The table below shows the number of publications analysed for each section of this article.The object of study In most of the textbook analysis studies, the object (or phenomenon) of study is overwhelmingly that of nationalism (specifically African and Afrikaner nationalism) and race, which are studied in thirteen of the twenty journal articles.Within the post graduate dissertations, eight of the twenty three studies deal with these phenomena.Sometimes, these two concepts of nationalism and race are inter-twined in the same study.These phenomena are linked with broader concerns of prejudice, bias, ideology and master symbols.This is unsurprising given the role that history plays in carrying official messages about the story of the nation.Pingel notes that "History and Geography textbooks attempt to explain our roots, how and why we happen to be living in a certain place and how that place can be described and characterised -in other words, who we really are" (Pingel, 2010:7).
The study by Auerbach (1964) was prompted in part by the political change brought about by the new Republic in 1961.He was interested in the ways in which the messages conveyed in Afrikaans and English textbooks were the same.His findings showed that Afrikaans textbooks that he analysed in the early 1960s show a strong Afrikaans ethnocentrism, and the use of the word "volk" excludes all other population groups in South Africa, which mitigates against the "fostering of national unity and racial co-operation" as proposed by the Transvaal Education Ordinance of 1953.
During apartheid, studies on nationalism were concerned with the power of Afrikaner nationalism.After 1994, the emphasis of some studies shifts to look at African Nationalism, with both Pretorius (2007) and Engelbrecht (2008) arguing that there has been a role reversal and that the ideology of Afrikaans nationalism which was so strong during apartheid has been replaced by African nationalism.
Another new focus post-1994 is reflected in the studies by Chisholm (2008) and Maposa ( 2014) which look at how other countries in Africa are represented in South African textbooks.In total, 22 of all the studies that did a content analysis of textbooks focus on nationalism, race or ideology as the object of study.Some studies take particular events or people, and the way in which these are represented in the textbooks as their object of study.For example, van Niekerk (2014) analyses three textbooks for how they represent the person of Nelson Mandela.Koekemoer (2012) focuses particularly on dominant discourses about the holocaust, Morgan (2011) analyses how the topic of Nazi Germany is represented, and da Cruz (2005) examines how the person of Shaka is represented and the discourse about the "empty land".Both Smith (1983) and Mazel andStewart (1987, 1989) focus on how textbooks represent the San hunter-gatherers and what Smith terms "aboriginal peoples".Nishino's (2008) chapter is one of the few studies that engages specifically with historiography.His study analysed 16 Standard 6 textbooks' treatment of the fifth "frontier" war of 1818-19, to ascertain the influence of the pro-settler historiography in the description of the war.There are four studies which focus on gender, one which engages with the representation of "big men" in textbooks (Naidoo, 2014), and three which describe the representation of women (Nene, 2014; Schoeman, 2009; Fardon & Schoeman 2010).
There are far fewer studies that focus on the procedural aspect of history, or focus on historical literacy as the object of study.These kinds of studies emerged after 1994, when the South African history curriculum embraced an understanding of school history as preparing learners to think historically and to "do history" as historians do (Bertram, 2006).Waller's (2009) study focused on how historical literacy manifests itself in grade 10 history textbooks, and Morgan's (2014a) article focuses specifically on the learning of empathy (which is seen as one of the aspects of historical thinking) through textbooks.
There are three dissertations (Bharath, 2009;Firth, 2013;Job, 2015) which take as the object of study the way in which knowledge is represented in the textbooks, drawing on the concepts of Bernstein on knowledge that is specialised to the discipline (strongly classified) or knowledge that is more integrated and everyday (weakly classified).Methodology of the studies Nicholls (2003) argues that methods for textbook research are fundamentally underdeveloped.UNESCO has recently produced a second edition of their methodological guidebook (Pingel, 2010).Pingel describes both quantitative and qualitative techniques and stresses that these are complimentary methods.He describes a range of qualitative methods, such as hermeneutic analysis, linguistic analysis and discourse analysis.Nicholls notes that there are other methods that Pingel does not mention, such as disciplinary or historiographical analysis, visual analysis, question analysis, critical analysis, structural analysis and semiotic analysis.
Many of the journal articles in our database do not provide a description of the methodology used for the analysis, while the post graduate dissertations and theses generally do so.The two methods of analysis used most often are discourse analysis and content analysis.Within the studies which use content analysis, some use an inductive, open set of coding (such as van Niekerk 2014) and others use deductive categories for coding.At least three of the dissertations (Auerbach, 1963;Evans, 1991;Chernis, 1990) use the indicators and criteria developed by the Georg Eckert Institute and the UNESCO report to analyse textbooks.Three studies (Bharath, 2009;Firth, 2013;Job, 2015) use the concepts of classification and framing of knowledge which are drawn from the work of sociologist Basil Bernstein.
Otherwise there is a huge variability in the criteria used to analyse the textbooks.There is little indication that the studies build on work done previously with regard to the kinds of deductive categories that are used to code the texts.Two exceptions are Firth (2013) who develops the coding categories used by Bharath (2009), which were based on Bernstein's classification of specialised or everyday knowledge and Mackenzie (2014), who draws on the model developed and described by Morgan and Henning (2013).
Many of the studies published before the 2000s do not make explicit their methodology or tools of analysis.Over the past 15 years, there is a growing focus on making methodology explicit and Morgan's (2011) Ph.D. developed an interdisciplinary model for text analysis which draws from hermeneutic analysis, discourse analysis, visual analysis, question analysis, critical analysis and semiotic analysis.

Sample size
Of the journal articles, the biggest study analysed 70 books (De Wet, 2001), and the next largest sample was 27 (Engelbrecht, 2005), and then twenty textbooks (Carpentier, 2000).A similar trend is seen in the post graduate dissertations, where three Ph.D. studies have large samples.Du Plooy (1964) analysed 52 textbooks which were published between 1938 and 1963 and were used in Transvaal schools (a period of 25 years), Chernis (1990) analysed textbooks from 1839-1990 (a period of 160 years) that were used in all four provinces and Nishino's ( 2006) study focused on 60 textbooks used in South Africa and Japan between 1945 and 1995 (a period of 50 years).These three studies are unusual amongst dissertations and theses in that they have both a sample size of more than 50 books, and they analyse textbooks published over a number of years.
Aside from the examples mentioned above, the sample of textbooks analysed in the studies is generally small.Nine of the studies published in journals, and 14 of the dissertations present data from the analysis of one to four textbooks.Since most postgraduate studies are at the Masters' level, the tendency is for Yesterday&Today, No. 14, December 2015 students to analyse between two and four books.The predominance of small scale qualitative studies makes sense for a post graduate student, and for the methodology of the studies which tend to be in-depth, qualitative analyses.This mirrors an audit of education research commissioned by the National Research Foundation which found that 94% of education research in South Africa was small scale research (Deacon, Osman, & Buchler, 2009).However, it does mean that findings cannot be generalised beyond the particular sample of books that were analysed.In line with the small sample sizes, the vast majority of studies tend to compare three or four textbooks written for the same grade, published at a similar time.There are a handful of large-scale diachronic studies, such as Du Plooy (1964), who analysed books written between 1938and 1963, Nishino (2006, 2011) who compared Japanese and South African textbooks which were published between 1945and 1995, Smith (1983) who analysed textbooks published between 1914to the 1980s, De Wet (2001) who analysed books published from the 1940s to the 1990s and Chernis (1990) who analysed books published from 1839-1990.More recently there have been some smaller scale studies which compared textbooks used during apartheid with those published post-apartheid (after 1994).For example, Carpentier (2000), Bharath (2009) and Naidoo (2014) use this selection for their sample.
It is also noticeable that all the studies, save for Nishino, analyse only South African textbooks.There appear to be no other studies that compare South African textbooks with those from other countries.
The other noticeable pattern regarding sampling is that most of the studies analyse high school textbooks (that is, what is currently named grade 10-12).Eight of the studies published in journals focus on high school textbooks, three on primary school books and two studies include both.Of the dissertations and theses, sixteen analyse high school textbooks (with a particular emphasis on Grades 10-12), two studies analyse both primary and high school books, and only one study focuses on primary school books.Thus there is a strong research focus on high school texts, particularly Grades 10-12, and very little attention is paid to primary school textbooks.This reflects the strong emphasis in South Africa on the final years of schooling, which is somewhat ironic, given the statistics that show that only approximately half the million learners who start primary school will in fact write the Grade 12 National Senior Certificate exam (Spaull, 2013).Only one fifth of high school learners choose to take History as a Grade 12 subject (based on the figures from the 2013 National Diagnostic Report that 109 046 wrote the Grade 12 History examinations in 2013 of a total of 562 112 full time candidates), thus it may be important to shift our focus to the compulsory phase of History learning in the grades up to Grade 9.

Theoretical / philosophical orientations
Nicholls notes that there are four traditional philosophical perspectives which relate to textbook research which are positivism, critical theory, postmodernism and hermeneutics (Nicholls, 2005).Just as there are a range of analysis tools used in the South African studies, there is a broad variation in the philosophical orientations and theory that is brought to the studies.
Although the term post-positivist is often used now, rather than positivist, the ontological perspective that under pins this tradition is that of realism, of understanding that objects and subjects exist independently of one another.Thus using the scientific method can make it possible to discover patterns and relationships between objects (Nicholls, 2005:25).Earlier dissertations such as Raubenheimer (1943) andDu Plooy (1964) would probably fall into the positivist paradigm where the focus is primarily on quantitative analysis, and there is little engagement with the context in which the textbooks are written.Nicholls (2005) explains that critical theorists see subjects as existing in a world of unequal and oppressive social relations and it is only the raising of consciousness that will overcome this alienation and oppression.In contrast to positivism, the socio-economic, cultural, historical and ideological context is vital, but similar to positivism, critical theorists argue that the objects in the world are governed by universal laws which can be known (p.26).The studies published in the 1980s, such as Chisholm (1981), Smith (1983), Mazel andStewart (1987), andDu Preez (1983), would fall broadly into a critical theory framework, where issues of unequal social relations and powerful ideologies are foregrounded.This reflects the critical turn in sociology in white English universities at the time, when theorists such as Althusser and Gramsci were in ascendance.
The hermeneutic tradition is concerned with the relationships between language, meaning and interpretation (Nicholls, 2005:29).Researchers in this tradition recognise that all interpretations of reality have a subjective dimension.Ontologically, hermeneutic approaches do not believe that text represents an objective world, but that text is subjective and is socially constructed.A constructionist perspective is not only about language or text, but similarly argues that knowledge is relative and is defined through the perspective of the knower (Wheelahan, 2007).
Philosophical orientations to research are loosely informed by the university tradition from which the researcher comes (Hoadley, 2010).Many of the dissertations produced by students at the University of KwaZulu-Natal, School of Education tend to fall within the hermeneutic and social constructionist frame and use discourse analysis as the analysis lens.Examples are Maposa's (2014) study on the construction of African consciousness, Koekemoer's (2012) analysis of the dominant discourses of the holocaust and Naidoo's (2014) analysis of "big men".Morgan's (2011) study on race discourses in Grade 11 textbooks would also fall into this tradition, as would the studies by Schoeman (2009) and Fardon and Schoeman (2010) on the representation of women and gender bias.
The fourth philosophical tradition described by Nicholls (2005) is that of post modernism, whose adherents believe that there are no absolute facts and have no faith in science progress, rationality or the hierarchical structuring of knowledge (p.28).None of the studies in this analysis are located within this tradition.
There is another research tradition that is reflected in a few of the studies here, and that is social realism.A social realist approach acknowledges that knowledge is a social product, and is thus fallible, but that an objective reality does exist and our job as researchers is to understand that reality, despite our knowledge being impartial and socially mediated (Wheelahan, 2007).The studies here that are informed by a sociology of knowledge perspective which focus on structuring and the specialisation of knowledge (Bharath, 2009;Firth, 2013;Job, 2015) would fall into the social realist approach.

Conclusion: What does this mean for the future of textbook research?
In this article we have created a database by means of a literature search of the research conducted globally on South African History textbooks.This database (Appendix A), which we do not claim to be incontestable, consists of 65 dissertations and theses, books and book chapters and peer reviewed academic articles.The analysis shows that we have a sizable body of research in the field of history textbook research, and forthcoming studies need to build on what we already know.However, what is not understood is the position of History textbook research in relation to other research conducted into textbooks in South Africa -a worthy future research endeavour.
One way of building a more coherent body of research into History textbooks is to use methodological and analysis tools that have been used previously in order to make these tools more robust and trustworthy, as well as to make philosophical paradigms more explicit.There is clearly a silence in the completed research on how teachers and learners actually use textbooks in their classrooms.Researchers seem to have a strong assumption that textbooks carry powerful messages, but we have little empirical work to show how teachers mediate these messages in their classrooms and in fact what learners actually learn about History as they use these textbooks.Another gap is that there is very little work on the historiography of the knowledge that appears in the textbooks.The research community does not seem to be asking: "Where do the textbook writers/producers find their sources?"and "Who is writing/ producing textbooks"?Nishino's chapter (2008) on the influence of Theal's settler historiography actually names textbook authors, while the majority of studies use the name of the publisher to identify a book.In this way, authors become invisible, and the role of the textbook can be reduced to simply how compliant it is with the current curriculum requirements.In conclusion, we suggest that the field of History textbook research is healthy in South Africa, but that forthcoming research should strengthen and consolidate what knowledge has already been built in the field.Hopefully this article can contribute to this strengthening.

Database of Studies on South African School History Textbooks
Note: The shaded rows in the table represent studies which focus on the role of textbooks in school, their design, selection or distribution or on theoretical and methodological issues.textbooks in comparison (1945-1995).

Image 1 :
South African School History textbook research by decade and total

1
See for example: GJ Thirion, "Die geskiedenishandboek -'n onmisbare ergernis",Yesterday & Today, 19, 1990,  pp. 17-19.What must be pointed out is that the Yesterday & Today has substantially changed since then and is now an accredited journal. 2 For excluded Honours projects see for example: LK von den Steinen, Three decades of SA school history textbooks: to post-apartheid: Ideological change revealed by racial stereotyping (Paper presented at the "Education, Social Cohesion and Diversity Conference", Washington DC

South African School History textbooks 7
Changing histories: Japanese and South African  textbooks in comparison, 1945-1995 (Göttingen: V&R Unipress, 2011).Lieven, M 2000 Bias in school history textbooks: Representations of the British invasion of Zululand.
7 Nishino's thesis was reworked published as a book: R Nishino,