Article Gender, education and the transition to democracy: research, theory and policy in South Africa C1980-1998 Linda Chisholm and Elaine Unterhalter Education has been a key terrain of struggle in the transition to democracy in South Africa in the last two decades. Much of the debate and the majority of analyses of education in this transition have been primarily concerned with the ways in which apartheid definitions of race marked formal education provision and mobilised extensive opposition. However, the racism of the apartheid education system was gendered in striking and significant ways. In seeking to redress these inequities, the government elected in 1994 noted: Within the education system there are worrying disparities between girls and boys, and many girls and women suffer unfair discrimination and ill-treatment... In many schools and other education institutions, including the most senior, social relations among students, and between staffand students, exhibit sexism and male chauvinism. Sexual harassment of girls and women students and women teachers, as well as acts of violence against women, are common.... (South Africa 1995:46) As part of its policy of redress, the new government committed itself to establishing a Gender Equity Task Team (GETT) within the Ministry of Education to advise on gender imbalances in enrolment, dropout, subject choice, career paths and performance, and the desirability of single sex schools. This Task Team was appointed in late-1996 and reported in 1997. GETT began its work in the context of considerable formal support for gender equity in education in the national Ministry of Education, provincial ministries and in a number of national women's organisations, particularly the ANC Women's League, and the National Women's Coalition, a federation of a broad front of national and regional women's organisations across the spectrum of political positions from right to left. But the research base TRANSFORMATION 39 (1999) ISSN 0258-7696 Linda Chisholm and Elaine Unterhalter regarding gender and education in South Africa that GETT had to draw on was not extensive. There are gaps in analysis and a lack of information in important areas. One of the outcomes of the GETT work will doubtless be the stimulus it gives to further research. Nonetheless, despite these limitations, important work has been done, that is significant not only for South Africa, but also for its resonances with other contexts. The purpose of this paper is first to outline the theoretical and methodological perspectives which have characterised the research carried out on gender and education in South Africa since 1980; second to indicate some of the policies informed by these approaches that are being put into place by national and provincial ministries and NGOs; and, third, to highlight some significant silences in this research and some of the potential problems for new policy that might arise from these silences. While we have attempted to look at a range of published and semi-published material drawing on research, this survey is by no means exhaustive; it is particularly limited with regard to unpublished masters and doctoral research in this area, and has not explored in any depth material published in the press or documentation prepared for NGOs and CBOs. A striking feature of much of the literature on gender and education in South Africa is the extent to which it has been written predominantly by white women and by women located in a handful of universities in South Africa or abroad. This provides an index of some significant features of the context of research and policy making in South Africa. First, it indicates very starkly the ways in which the racism of the apartheid era impeded the entry of black women and men into the research community. Second, it highlights how only a very small handful of historically white South African universities have generated published work in this area. Third, it points to the way in which the small numbers of black women who have written and published on these issues have tended to do so mainly in the context of their work for government, NGOs and CBOs, and the ways in which this writing and thinking has not reached audiences outside a particular policy and implementation cycle (sec Mkwanazi-Twala 1995). The problems of race, location and authorship have been an issue of debate in South Africa for some years; the extent to which it might be a factor in the frameworks of the debate on gender and education outlined in the paper is beyond our scope here, but is an area of continuing concern to both of us. In developing the argument, we will use the distinction now made between Women and Development (WID) and Gender and Development Gender, Education and the Transition to Democracy (GAD) approaches. Concerned primarily with how gender equity has been addressed in international agencies working within a 'development' discourse, recent writers have shown how an earlier stress on WID — equality of opportunity for women which was premised on the belief that the problem lay in discriminatory barriers to women's employment, has now been replaced by GAD — notions of gender equity which 'require the transformation of the basic rules, hierarchies and practices of public institutions' (Kabeer 1994:97; see also Parpart and Marchand 1995; Moser 1993). The difficulty in applying these approaches unprohlematically to South Africa, is that the historical periodisation does not conform to that in South Africa. An early WID approach was not followed by a GAD approach. Historical conditions of apartheid and the nature of the struggle against it particularly after the 1976 Soweto student revolt ensured that both organisationally and in educational research both approaches co-existed, and continue to be in conversation with one another. While a number of recent gender-in-education initiatives in South Africa are strongly informed by liberal modernisation approaches, they also reveal an awareness of institutional and relational issues. A pertinent question here is, however, whether the discourse and institutional practices of policy and development, increasingly prominent in governmental as well as non-governmental agencies, tend themselves to privileging liberal modernisation approaches. A related theoretical and political question concerns the relationship between politics, policy and research or the question of whether policy research can be transformative or not. From a gender perspective, the reduction by policy of complex social and gender relations to simple issues is a major problem. Jane Parpart (1995: 225), writing about the role of development experts in gender and education, notes that 'developmental policies and programs are largely predicated on the assumption that developmental problems can be reduced to technical, ie, solvable problems which involve the transfer of western technical expertise to the developing world'. She indicates that education policy research which reduces deep structural issues of power and inequality to solvable problems is conservative because it depoliticises that which cannot or should not be removed from the realm of debates and actions concerned withjustice, power and inequality. For material feminists such as Weiner and Kenway, the way out is pointed by efforts to resolve the tension between theory and practice. For Gaby Weiner, for example, feminist activism and critical intellectual work took her into 'management and policy projects' which she describes as 'making Linda Chisholm and Elaine Unterhalter change possible and with creating socialjustice or feministpractice' (1994:23). For Jane Kenway, the 'crisis of feminism' manifested in the destabilisation of the professional authority of feminist teachers and researchers in universities does not imply abandoning a voice: What is important is to what end and how feminists practice their authority as intellectuals, whether they make it problematic to themselves and others and whether and how they make themselves accountable and to whom... in various ways (they) can play a powerful supportive role for the non-dominant. (1995a: 14) Taking these two together, it is not so much whether solutions are found to problems that is the issue, as the purposes, methods and accountabilities of researchers. The principal condition in this perspective of a transformative policy research would be its accountability to the non-dominant and non-powerful; however such accountability is not itself without difficulties, particularly with regard to changing organisational forms amongst dominant and dominated groups. Some of the differences in perception between Kenway and Weiner derive from differences in position given differing conditions of feminism and the organised labour movement in the UK and Australia in the late-1980s and the early-1990s. The tendency to dominance of liberal modernisation policy positions in South Africa may partly be an outcome of its history of fragmented feminisms, with women's organisations frequently surviving only because of uneasy alliances with nationalist or religious movements. These concerns - the relationship between policy research and scholar- ship, whether it is possible forpolicyresearch to be transformative, whatthe conditions of such research would be and what the elements of a transformative policy research agenda would be - informs our approach to some of the research we review. For this reason, our review focuses on gender and educational research across disciplines and not necessarily or only focused on policy. Theoretical and methodological approaches to gender and education since the 1980s It is necessary briefly to outline the features of the theoretical approaches we use to examine research and policy on gender and education in South Africa. The first, liberal modernisation, or WID, seeks to analyse the barriers women encounter in accessing and participating in social institutions, including education. This approach has been concerned with identifying ways of overcoming barriers to access and bringing women into modern Gender, Education and the Transition to Democracy institutions which are viewed as unproblcinatically beneficial for women and are often identified as the source of their emancipation (Parpart and Marchand 1995; Moser 1993. For a classic account written within this paradigm see King andHill 1993). The second, GAD, approach draws on feminist critiques of liberal modernisation and poses questions about gender relations within institutions once access to them may have been attained (see also Savage andWitz 1992; Franzway, Court and Connell 1989). An emphasis on women alone is replaced by one analysing shifting boundaries and relationships between women and men and the ways in which the gendering of institutions, or social relations between women and men, shape institutional power relations and unequal outcomes for women. The articulation of educational institutions with other gendered sites like the family, the state and the labour market has been another strand of such research (Moser 1993; Sen and Grown 1987; Elson 1995; Kabeer 1992 and 1994; see also ANC 1994b). Associated with this, and to date the least developed approach in South Africa, are approaches influenced by post-structuralism and attempts to problematise die construction of gender. It looks at changing regional and historical constructions of masculinity and femininity, and poses questions not only about the structural and institutional location of gender inequality, and its changing configurations, but also about language, culture and subjectivity and the ways in which gender is constructed and reconstructed (Parpart and Marchand 1995;Kcnway 1995c; Wolpe 1988; Thome 1995). Linked to each of these theorisations is a tendency to utilise particular methodologies and particular approaches to policy. Thus, the liberal modernisation approach tends to rely on positivist approaches, survey methods and the analysis of government and other statistics. It proceeds by identifying barriers outside education that prohibit the entry of girls and women and advocates policies (often in legal and economic terms) for removing the barriers. The gender relations approach in South Africa has been mainly historical, but is also influenced by structuralist analysis of institutions and by feminist ethnography. Although not always policy-oriented, the policies that can be extrapolated or are advocated tend to focus on processes within and between social institutions. The post-structuralist approach also draws on ethnography but in addition uses biography and autobiography, as well as documentary analysis in seeking to deconstruct discourses and consider the way they shape policy. Its strength, as well as those of earlier feminist social historians, is generally Linda Chisholm and Elaine Unterhalter in highlighting the limits of policy change and the complex and often unintended ways in which changes are received and interpreted by different constituencies. In the next section of the paper we look in more detail at the research carried out within these different theoretical frameworks and the ways in which they have affected current policy. WID/GAD dichotomies in historical perspective Research on gender in education in South Africa has tended to take one of two forms, both of which place a South African perspective on the well known international formulations of the WID/GAD dichotomy. We will first consider literature not focused on education per sc; this will be followed by an analysis of research within education, particularly within the recent period. When we consider literature not focused on education, there is, on the one hand, a considerable body of work conducted over the 1970s and 1980s within a historical materialist frame work.This literature itself divides into those studies in which gender is peripheral and those which focus specifically on women. Although both are materialist, the implications of the former may be seen as lying within the WID paradigm and the latter within the GAD paradigm. In the former, which considers the triad of race class and gender, gender is seen as an intensification of other social divisions; the implication of exclusions based on race, class and gender is inclusion within the political economy on those bases (Walker 1982). Many of the critiques of segregation and apartheid worked with a meta-theory infused by narratives of the modern. Although they were extremely critical of the liberal modernisation theory assumption that capitalism would erode apartheid, they nonetheless shared with liberalism an optimism about the enlightening results of education. Hence, work from this disciplinary base on gender and education could see gender identity and inequity as an inflection of class and/or race divisions and injustices and could see access and removal of barriers as a major priority. Work done in the early-1980s from a feminist materialist perspective tended to prioritise women as opposed to relations between men and women in the analysis of gender relations. They included pioneering studies on the encounter between liberal white missionaries and African women, and the relationship between pre-capitalist and capitalist forms of subordination in the nineteenth century Eastern Cape (Cock 1980). Cock showed how African Gender, Education and the Transition to Democracy women were subordinated through mission education not only to a race and * class-based colonial order, but also to western assumptions about appropriate gender roles for particular classes of people. The conception of education as subordinating implied that education could also be liberating. Gaitskell took up similar themes about the role of education in preparing African women for domestic service in early-twentieth century South Africa, but also showed how the tiny minority that were provided education by female missionaries were trained for occupations such as teaching and nursing considered' appropriate' by white female missionaries and available within the segregated social order (Gaitskell 1990). In a fine social history of fractured female relationships, Marks explored the divisions of race, class and age between three women in the 1940s and 1950s. In the tale that emerges from the correspondence between the women, an academically talented young African girl is thwarted personally and socially; a crucial element in the analysis concerns both the denial of her educational aspirations and her institutional experience of sexual harassment (Marks 1987). These accounts are all concerned, on the one hand, with the social institutions through which particularly African women constructed new gender identities. How women made sense of their racialised and class experience is a crucial dimension of this work. Over time, the emphasis thus shifted from broad structural determinants of women's lives to individual life strategies and consciousness (see Bozzoli 1991). Although it would only be fair to say that none of these studies have education as their main focus, their emphasis on gender relations within the context of broader social relations and institutions of church, family and state, place them firmly within what is now referred to as a GAD paradigm. Feminist materialist work specifically within education (Mager 1992/93; Chisholm 1990a and 1990b; Hughes 1990) was historical and concerned with gender regimes within mission and state educational institutions and gendered constructions of segregated institutional practices, rules and rituals. Here we do find some attempt to try to unpick the multiple layers of gendered experiences of those institutions. The problem is that this work at times collapses gender into race and class. In conclusion, much historical materialist and feminist materialist research on education in the 1980s embodied both WID and GAD assumptions, but GAD assumptions were probably dominant. These were not easily translatable into policy proposals, however, given their primarily historical character. Policy required a different kind of knowledge. Linda Chisholm and Elaine Unterhalter WED/GAD dichotomies in educational perspective: identifying barriers to access Recent research relating to women and gender has been dominated by WID perspectives, although, again, GAD perspectives and approaches are not absent. After the unbanning of political organisations in South Africa in 1990, in the process of preparation for the first democratic election of 1994, new directions were embarked on. Research in education became completely geared to the process of transition. Policy research, which emerged at this time, bifurcated into that which collapsed gender into race and class, and that which dealt specifically with gender issues. Whereas the former became dominant, the latter remained more marginal. The problematic of the former has been WID-informed; the latter has been more sensitive to gendered institutional dynamics, constructions and processes. We will now first consider the former, and then the latter. With the prospect of major change in the forms of subordination of women, research specifically on gender became less historical and more concerned with establishing contemporary patterns of gender inequality with a view to developing policies and strategies to address them. This research relied heavily on the analysis of official statistics to delineate gendered trends regarding access to education and achievement in terms of credentials and labour market position (Unterhalter 1991a and 1991b; Budlender 1991 and 1993;Truscott 1993; BadshaandKotecha 1994; Budlender and Sutherland 1996). Recent work has also used official statistics to indicate women's needs for adult education and for early childhood education (Schindlerl996). While this work established a correlation between race, gender and educational access, it also brought to light the way in which, under apartheid, the numbers of girls and women (of all racialised groups) enrolled in formal education had tended to increase both numerically and proportionately (Unterhalter 1991; Truscott 1993). This work showed the ways in which gendered enrolment patterns in South Africa had more in common with Botswana, Namibia and Zimbabwe than with other countries of Africa. But the reasons for this trend in very different economic and political contexts remained unresearched (Unterhalter 1995; Kann 1995). Many of these new studies were policy-oriented and based on official statistics. They were hampered by existing sources of statistical information on gender and education in South Africa which are extremely weak. Any analysis of enrolments, repetition, drop-out and performance has been and Gender, Education and the Transition to Democracy remains constrained by existing sources of information. Gender- disaggregated information is extremely poor, and there are gaps in the collection of information. Both national and provincial departmental and NGO sources of information, such as the EDUSOURCE Education Atlas (199S), are important but inadequate. Enumeration for the first census conducted under a democratic government was completed in October 1996, and began to provide gender-disaggregated information, although not specifically education-related. By failing to look at the co-constructedness and shifting boundaries and variability of gender across race and class in institutions, much of the policy literature assumes that the formal abolition of legal admission barriers which discriminate on grounds of race and the lifting of economically based barriers like fees, or uniform costs, that discriminate on grounds of class, is the key to a gender policy in education. As such, it will be sufficient to ensure that girls and women gain access to education. A variant of this supposition is found in the 1994 White Paper and the drive to establish a National Qualifications Framework. The belief is that if employers are able to certificate their (assumed-to-be-male) workers' competencies through such a framework, they will support access to education and training (for critical analyses, see Samson and Vally 1996; Samson 1996). While some of this research goes further than a simplistic WID position, pointing to gendered discrimination in the labour market, despite women's educational achievements, the policy outcome tends to be limited and thus not able to do more than establish a formal legal framework, and possibly some fiscal incentives, for equal opportunities. Another approach within the WID paradigm is one that counterposes the modern global rationality and empowerment of schooling against what is portrayed as the static and culture-bound environment of the family (for examples, see King and Hill 1993; for critiques, see Schultz 1993; Parpart and Marchand 1995;Unterhalter 1996). South African work, within this 'family- in-deficit' approach, again tends to be less naive and simplistic than much of the international literature, drawing on analysis which points to the ways in which the family form was constructed in particular ways by the apartheid state (Manicom 1992; Ramphele and Boonzaaier 1988; Segar and White 1992). Nonetheless, assumptions that education acquired outside the family is really useful knowledge, and that gender dynamics within the family prevent girls and women from accessing and utilising education are key notions in the largest study conducted to date in South Africa of gender and Linda Chisholm «nd Elaine Unterhalter educational attainment. This substantial research project was part of a multi-purpose integrated household survey coordinated by the University of Cape Town's Southern Africa Labour and Development Research Unit (SALDRU) and funded by the governments of Denmark, the Netherlands and Norway, with technical assistance provided by the World Bank. The survey covered 9,000 households nationwide and was conducted in late-1993. The survey gathered information on family composition, expenditure, access to basic services, quality of housing, education, health status, income, and employment. One in every six households visited was asked to participate in an assessment of basic literacy skills. A brief examination was given to adolescent and adult members of these households, lasting about 45 minutes. This examination assessed basic language comprehension and numeracy skills. Oral language and reading comprehension were assessed in both English and one of the nine vernacular languages. Although this survey has been criticised for its sampling techniques and the way in which processes undertaken in the data collection phase meant that there was a heavy bias towards surveying urban people and higher socio-economic groups (Sender 1996), it provides insight as to how to overcome some of the difficulties with the analysis based on inadequate official statistics outlined above. It enables some of the national data to be disaggregated and points to a high level of regional diversity and important differentials in attainment. Three studies based on subsequent analysis of the data have produced important new findings. One has shown that although enrolments for girls and women have increased in recent years, literacy levels have declined. Thus, although the percentage of girls enrolled in secondary schools exceeds males in mostprovinccs and although girls, and in particular African girls, on average stay in school one half-year longer than do boys, female literacy is lower than male literacy. However, there are major disparities between the different racial and ethnic groups. For example, although the level of female literacy nationally is on the whole lower than that of males, literacy amongst Tswana-speakers, and especially Tswana-speaking females is remarkably high (Fuller, Liang and Hua 1996). A later study by the same authors tries to provide an explanatory model of the performance of North Sotho girls by reference to decision-making processes within the family economy- Girls' school attainment is linked to social resources and demands placed on girls, the positive influence of 10 Gender, Education and the Transition to Democracy absent fathers, and the recent expansion of schooling in the then Lebowa (Fuller and Liang 1998). A third study, on education and employment, shows that the majority of out-of-school girls live in larger households, and that girls and women in female-headed households are more likely to be unemployed than those in male-headed households. This suggests that they are dependent on income in larger households, and that male-headed households provide access to job networks from which female-headed households tend to be more marginalised. Technical education appears to be rewarded in the labour market, the implication being that girls tend to be excluded, since technical education is predominantly the province of boys and men (Wittenberg and Pcarce 1996). The detail of the large data set allows for a useful picture to be built up which relates household income and employment, division of labour, parental level of education, the length of time girls stay in school, and their level of employment after leaving school. But the assumption that has so far been made in the papers written on the basis of this data is that families are synchronic households, where gender might govern resources. Gender here is a means descriptively to differentiate those who have (usually access to the labour market or schooling) from those who have not; it also becomes a marker of education in terms of opportunity cost. This work provides interesting information on the relationship between family structure, income, education and employment, but is unable to map the much more complex configuration of families as diachronic gendered networks, governing the production and exchange of gendered sexual and cultural identities and the ways in which families make complex investments in education, both in terms of enrolment in formal schooling and in the identification of other locations for education, knowledge and cultural production (see for example, Dclphy and Leonard 1992). Autobiographical writing provides some particularly rich material on these themes (see for example, Kuzwayo 1985; Ntlantla 1992). For a more nuanced analysis, different theoretical assumptions and methodologies would be required. Thus, despite the richness of the data available, this research tends to replicate the WID notion of 'deficit families' which are hindered from taking advantage of formal schooling, particularly because of low levels of parental education and lack of mothers' access to employment. The policy implications of these studies vary. On the one hand, they suggest a gender-affirmative social policy - bursaries for girls, income support for female-headed households and a socio-economic programme of 11 Linda Chisholm and Elaine Unterhalter reconstruction and development which prioritises job creation in which women are targeted as key players. On the other, they promote improved schooling and attention to the education of fathers. A less particular, 'universalist' approach is adopted by the Lund Committee on child and family support which reported in 1996. This committee has not recommended special targeting of women, but the introduction of a flat-rate child support benefit to replace state maintenance grants.The benefit is payable to the primary care-giver of a child and determined according to a simple means test. This approach, although not specifically targeted at women, is nonetheless pro-women, as it deviates from previous comprehensive family preservation policy in favour of recognising the diversity of family forms and dominance of female-headed households in South Africa. The research and policy initiatives described above go considerably beyond the equal opportunity frameworks associated with the first approach within WID. They do look more seriously at relationships between schools, families and the labour market. They are cognisant that educational provision itselfwillnotbe sufficient to address childpoverty, structural unemployment, labour market discrimination, job-recruitment processes or the negative effects of the radical marginalisation of women from the labour market. Nonetheless, they do not address questions of variably gendered relations within educational institutions. Moreover, the job-creation proposals may lead to unintended consequences. Unless they are implemented with insight and sensitivity regarding women's location in families, they can place greater burdens on women. Recent research on the negative impact of income-generating activities on women casts important light on policies which add obligations to work in the formal sector and earn cash for daughters' school fees to the other onerous burdens of household reproduction women carry. Thus policies seeking the possible future emancipation of a generation of young women through such mechanisms might do so by drawing heavily on the exploitation of their mothers (for examples of some of the negative impact of income-generating activities on women, seeElson 1995). The research we have surveyed within this WID paradigm laid important groundwork for the government in its Education White Paper and the Reconstruction and Development Programme to consider gender, race and class and the ways in which they linked with exclusion from education. However, the policies targeted at addressing these issues have tended not to address questions of what happens inside educational institutions, nor 12 Gender, Education and the Transition to Democracy have they been concerned to consider the inter-relationship of education as a gendered site with other sites of subordination. GAD dichotomies: gender, institutional processes and articulations In contrast with the WID approach's focus on bringing women into education, regardless of the processes that take place within the 'black box' of schooling or of girls' complex positioning outside educational institutions, GAD, as applied to education is particularly interested in the gendered processes of educational policy and planning, with teaching, learning and knowledge production as gendered processes, and with questions of who manages the implementation phases, using which approach. It is also interested in the way gendered processes of education articulate with or transform other gendered relationships in society. While WID approaches might be interested in bringing girls and women to the schoolroom door, GAD approaches are concerned both with what happens inside the door, who brings the girls or women to the door and what happens when they leave the schoolroom and talk or plan about what they have learned. Theorisations of gender as a relational concept and closer examination of institutional processes have been present in new educational research that began to be published in the 1990s. In a key article written in 1992, Rob Morrell called for research in South African education to be more focused on gender relations in the classroom as a basis for new policy. At the same time, studies were produced which drew on ethnography and participant observation. These looked at the conditions of work of female teachers and inequalities in educational administration (Kotecha 1994; Sebakwane 1993/ 94 and 1994; Mahlase 1997) and at definitions of'success' and 'failure' in structuring girls' achievement and the way this linked to pedagogy (Ensor 1993/94). Modes of discipline, violence and widespread sexual abuse of girls by fellow male pupils and teachers have been continuously documented (Nzimande and Thusi 1990; Russell 1991; Seekings 1991; Chisholm and Vally 1996b; Wolpe 1995) Sexual harassment is a feature not only of schooling but also of higher education, as path-breaking research at the University of Cape Town and accounts of the student movement revealed (EORP1991; Sutherland 1993). Insider accounts also provided documentation of the difficulties of positionality for black feminist student activists in the late-1980s and early- 1990s and the ways in which the politics of building a broad front against apartheid education led to feminist concerns being seen as divisive (Nkomo 13 Linda Chisholm and Elaine Unterhalter 1993). In addition, work examining the construction of masculinity in South African educational institutions, along with associated forms of violence and aggression, is promising, as it addresses the institutional construction of gender (Morrell 1996 and 1998). Much more work of this kind needs to be done at all levels of education. In 1994, this research found its policy recommendations rather weakly expressed in the ANC Policy Framework for Education and Training, written to guide the policy process for the new government. This document committed the ANC more strongly to correcting the historical marginalisation of women through the education and training system (ANC 1994a). Its clearest statements concerned the need to 'confront gender bias in the curriculum and pedagogy' and ensure representation of women at all levels of the education system. Criticisms of this document came chiefly from within a gender relations approach. They focused on the way the document gave inadequate recognition to the relationship between education, the labour market and family and pointed to how a policy emphasising integration of education and training, and stressing mathematics, science and technical training in curricula generates problems for girls and women 'not directly linked to the labour market' (Wolpe 1994:14). The main concern here was with the way in which the education system, 'through both the overt and hidden curricula, articulates with family formation in a way that reinforces traditional female roles' (Wolpe 1994:12). The implications of Wolpe's work was the need for much more work in understanding the changing relationship between gender relations in families and in schools. In another critique, McClellan argued that the marginalisation of gender issues was largely a consequence of a formal policy of equality which had been outlined 'at the cost of addressing the informal processes of power embedded in social understandings and institutions' (McClellan 1993/94). Internal criticism registered in the ANC's Plan for Gender Equity produced as part of its Implementation Plan for Education and Training (1994b) focused on the way in which dominant universalising theoretical frameworks in use tended to obscure substantive issues of race and gender, and how a focus on access and governance within institutions alone could occur at the expense of empowering women with little or no access to institutional power. Following the production of the Policy Framework, the ANC Education Department produced a comprehensive Implementation Plan for Education and Training (IPET) which included a Plan for Gender Equity and a recommendation for the establishment of a Gender Equity Unit (ANC 1994b). 14 Gender, Education and the Transition to Democracy The proposal for a Gender Equity Unit was with some modification incorporated into the White Paper on Education and Training generated by the Government of National Unity's new national education department. Although its scope was initially considered to be too comprehensive to be practicably implemented (Daniels 1995), its existence within the White Paper nonetheless demonstrated the seriousness of official intent on the matter. The Gender Equity Task Team Report (1997) made recommendations for mainstreaming gender concerns through the establishment of a Gender Equity Unit and an extensive gender equity machinery throughout all departments. Although the Report falls between two stools, being neither a classic policy document nor an academic text, it does provide a major contribution to the theory and practice of gender politics in education in South Africa. It incorporates insights from both GAD and more recent feminist educational interventions. Traversing a wide range of areas, each of the chapters provides new perspectives and approaches on gender practices and commonsense assumptions. And finally, even as the Report recommends the institutionalisation of processes and procedures to take gender matters forward, it is intensely aware of the limitations of strategies based primarily on legislation. Gender equity is being pursued in South Africa through the constitution, the establishment of a Commission on Gender Equality, through various strategies and policies in particular departments and women's groups.The movement towards an approach combining access and institutional emphases has been considerable since 1994. This development at the level of national policy has been accompanied by similar developments at the provincial level and in NGOS. Although we cannot romanticise the degree to which gender is a priority, we cannot but be impressed by the range of activities and initiatives specifically focused on gender and spanning WID and GAD approaches which have emerged in this recent period. They provide some index of the energy currently focused in multiple and heterogeneous ways on gender. They include the banning of all forms of violence, including sexual harassment and corporal punishment from schools in an effort to provide girl-friendly school environments in some provinces; attention to gender-affirmative action within the new bureaucracies; the focus of the South African Democratic Teachers' Union on gender and curriculum; the establishment of a national chapter of the Forum for African Women's Educationalists (FAWE) to 'build gender equity in education through addressing policy making in gender equity in education at all levels, 15 Linda Chisholm and Elaine Unterhalter but particularly in higher education'; and the creation of NGOs such as The Young Women's Network, an NGO 'aimed at providing unemployed young black women with skills and hope' and organisations such as End Racism and Sexism in Education (ERASE). Post-structuralism The link between policy, politics and theory we have tried to explore in relation to writings withinaWID and GAD framework becomes, paradoxically, clearer when the third position - post-structuralism - is explored. Post-structuralism has been influential in a small number of South African education faculties and has inspired a range of critical reflections on the process of educational transformation (for examples, see Muller 1996; Deacon 1996). This work is particularly interested in the knowledge-power couplet, who has produced policy and the history of language and concepts. What is interesting about this work is how little it has been influenced by feminist post-structuralism. Genuflections are made in the direction of gender, but no research on gender or gendered insights are registered. Writing within this framework, even when aware of gender as a variable among respondents interviewed or of gender inequity in a particular region, it fails to consider shifting gender boundaries, identities and discourses (see for example, MacLeod 1995; Deacon and Parker 1993; Muller 1996). Notable exceptions are Betty Govinden's (1998) probing work on identity and difference and her bold application of post-structuralist feminist epistomologies to education, Astrid von Kotze's (1996) analysis of the use of a feminist pedagogy in a writing workshop for adult women learners and Morrell's 1994 work on the social construction of white male masculinities in boys' schools. While there has been comparatively little work done within education using a feminist post-structuralist approach, writing in the broader sphere of cultural studies has raised important questions for debates about curriculum and culture. Deborah James (1994) has analysed women migrant workers' versions of familiar songs and the ways in which themes are reworked and recontextualised given new singers and new audiences. Dorothy Driver (1991) has indicated how language and meanings in official documents carry important implications for feminist politics. The critical studies on South African women writers put together by Cherry Clayton in 1989 raise important issues for thinking about texts as part of the quartet of players (with teachers, learners, and institutional structures) which help frame educational 16 Gender, Education and the Transition to Democracy encounters. Possibly the most influential of feminist work in this field has been Desiree Lewis' critique(1993 and 1994) ofthe silences and assumptions regarding race in much ofthe writing of what remains a predominantly white group of feminist analysts. The questions of race, ethnicity, intellectual production have been difficult and often disabling for South African feminists to confront (see various issues of the journal Agenda). Partly this is the result of political and policy agendas that have been distrustful of identity politics in an era of national liberation; partly also it is an outcome ofthe fragility of feminist scholarship, which has tried to build a consensus across considerable diversity. The very limited work within post-structuralism also seems to indicate how texts and discourses do not appear as the major field of political contestation in the mid-1990s. While this may well change once the government and its policies are well established, concerns ofthe moment are with establishing gender redress within institutions, not with deconstructing institutional discourses. Silences The extensive activity on education and gender redress masks some important silences. We can construct the silences that exist in the literature in terms of what exists in developed countries. Thus we can note how little research is being conducted on important areas such as gender identity, sexuality, citizenship, feminist epistemology and the curriculum, classroom dynamics, assessment and teachers' work. Interesting work in the USA, the UK and Australia looks at the complex ways in which gender intersects with other politically and economically constructed social divisions and identities like race, ethnicity, sexual orientation or disability to shape educational processes. This educational encounter in turn entails a reciprocal reshaping of intersecting unjust divisions and correlative intersecting identities (see for example, Maynard 1994; Mac anGhaill 1994; Epstein etal 1998; Giroux and McLaren 1994; Fine et al 1997; Kenway and Willis 1993; Yates 1998). There is little work on these intersections in South Africa and barely any work critically examining the understandings of gender justice in the transformation project. An important article by Penny Enslin (1993/94) which posed questions about the tension between the notion of education for nation building (with its stress on universalist ontologies) and the potential for educational 17 Linda Chisholm and Elaine Unterhalter institutions to nurture the emergence of a democratic gender politics drew a vociferous response from two black male colleagues (Cross and Leroke 1995). Part of the attack on Enslin's analysis was her use of feminist theoretical frameworks. Enslin's 1995 response tended to focus on the structure of the argument of the critique rather than the political context in which it was made. The bitterness of the exchange highlights the centrality of the issues raised to the transition in South Africa, but unfortunately, to date very little empirical work has been undertaken that could help move the debate beyond clashing theorisations. Furthermore, none of the work we have surveyed on gender and education is concerned with regional or general development issues, either in an uncritical or critical way. Although there has been a concern among some policy makers with the ways globalisation impacts on the South African economy and how knowledge production in South Africa is being harnessed to give the country a competitive edge (Chisholm 1997; Christie 1997a and 1997b), there has been virtually no writing on the gendered dimensions of globalisation. Despite new possibilities opened up for collaboration with the region, comparative research on gender and education is thin on the ground. Finally, the policy discourse has itself displaced history writing which might explore the gendered construction of educational institutions including schools, universities, teachers organisations and unions, the teaching profession itself, andprincipalship and management. South Africa does not, for example, have a history of women teachers or a history of co-educational and single-sex schooling. The gendered character of the policy discourse, the gender implications and impact of new policies, and the gap between (anti-sexist) policy and (sexist) practice are areas that need much more work. Some of these silences indicate the context of research in South Africa, some a lack of organisations with a gendered outlook. Research production in South Africa is historically skewed by apartheid; in education it is moreover thin on the ground. University-based research on gender and education exists in isolated pockets. Policy research within government, universities andNGOs is increasingly being conducted within the framework of powerful multilateral institutions which set the agenda through the control of resources and personnel. Formal acknowledgments of gender as a variable have rarely translated into concrete recognition of these issues in policy formulation. Much of the work on the national qualifications framework, for example, the centrepiece of new policy, has proceeded as if 18 Gender, Education and the Transition to Democracy there were no gendered issues involved in curriculum and assessment. The South African Qualifications Authority has taken this route partly because of a lack of engagement with these issues by women's organisations (Samson 1996; Unterhalter 1998). Nonetheless, Gender and Development and Women in Development approaches are in conversation with one another as the Gender Equity Task Team Report demonstrates.The lack of regional and general analyses of development theory seem to indicate the immediate pressures of putting in places the institutional frameworks for a new government. Conclusion This article has tried to outline different periods, theoretical frameworks and policy agendas that have marked the writings on gender and education over the last 15 years in South Africa. It has argued that even though pioneering research dominated by a Gender and Development perspective was produced in the 1980s, it did not lend itself to policy formulation. The fuller emergence of Women in Development approaches in the 1990s needs to be linked both to the context of transition and attendant processes of institution-building and re-building as well as to the global context which privileges such approaches. While much work, particularly empirically based studies and the refinement of theoretical debates, remains to be done (as well as the development of a cadre of black women intellectuals placed both within and outside the academy) the area of research on gender and education is experiencing its own transition to democracy. References ANC Education Department. 1994a. Policy Frameworkfor Education and Training. Johannesburg. 1994b. Implementation Plan for Education and Training, 'A Plan for Gender Equity', Johannesburg. Badsha, N and P Kotecha. 1994. 'University access: the gender factor', Agenda 21. Bozzoli, B. 1991. Women of Phokeng: consciousness, life strategy and migrancyinSouthAfrica,i900-I983.London:Hemcmaim,}aiY>es Currey. Budlender, D. 1991. 'Women and the economy', Johannesburg: Community Agency for Social Enquiry. 1993. 'Human resource development and gender affirmative action', Johannesburg: Community Agency for Social Enquiry. Budlender, D and C Sutherland. 1996. 'National score card. Indicators of development in education'. South African country report prepared for FA WE 19 Linda Chisholm and Elaine Unterhalter (Forum of African Women Educationalists) South African chapter, University of Cape Town: African Gender Institute. Chisholm,L. 1990a. 'Class.colourandgenderinchildwelfareinSouthAfrica, 1902-1918', South African HistoricalJournal 23. 1990b. 'Genderanddeviance in South African industrial schools and reformatories'. In: C Walker (ed), Women and Gender in Southern Africa to 1945. Johannesburg: Ravan Press. — 1997. 'Restructuring South African education and training in comparative context'. In: P Kallaway et al (eds), Education after Apartheid. University of Cape Town Press. Chisholm, Land SVally. 1996. "The culture oflearning and teaching in Gauteng schools: report of the committee on the culture oflearning and teaching', Johannesburg: Education Policy Unit, University of the Witwatersrand, July. Christie, P. 1997a. 'Globalization and the curriculum: proposals for the integration of education and training in South Africa'. In: P Kallaway et al (eds), Education after Apartheid. University of Cape Town Press. 1997b. 'Globaltrendsinlocalcontext:aSouth African perspective on competence debates', Discourse 18(1). Clayton, C (ed). 1989. Women and Writing in South Africa: a critical anthology. Marshalltown: Heinemann. Cock, J. 1980. Maids and Madams: a study in the politics ofexploitation. Johannesburg: Ravan Press. Cross, M and W Leroke. 1995. 'Mystification of democracy: Penny Enslin's critique of nation-building', Perspectives in Education 16(2). Daniels, G. 1995. 'Gender gaps in education', Agenda 25. Deacon, R. 1996. 'Discourses of discipline in South Africa: rethinking critical pedagogies in postmodernity', Discourse: studies in the cultural politics of education 17(2). Deacon, R and B Parker. 1993. In: N Taylor (ed), Inventing Knowledge: contests in curriculum construction. Cape Town: Maskew Miller Longmann. Delphy, C and D Leonard. 1992. Familiar Exploitation: a new analysis of marriage in western society. Cambridge: Polity Press. Driver, D. 1991. 'The ANC Constitutional Guidelines in process: a feminist reading'. In: S Bazzili (ed), Putting Women on the Agenda. Johannesburg: Ravan Press. Elson, D (ed). 1995. Male Bias in the Development Process (2nd edition). Manchester: Manchester University Press) Enslin, P. 1993/94. 'Education for nation-building: a feminist critique', 20 Gender, Education and the Transition to Democracy Perspectives in Education 15(1). 1995. 'Encounter with a straw man: a rejoinder to Cross and Leroke', Perspectives in Education 16(2). Ensor, P. 1993/94. 'Boundaries at the centre - differentiating pupils in mathematics classrooms', Perspectives in Education 15. Epstein, D, D Ellwood, V Hey, and J Maw (eds). 1998. Failing Boys? Issues in gender and achievement. Buckingham: Open University Press. Equal Opportunity Research Project. 1991. 'Final report of the committee of enquiry into sexual harassment', University of Cape Town, October. Fuller, B, X Liang and H Hua. 1996. 'Did black literacy rise after Soweto: public problems and ethnic archipelagos in South Africa', International Journal of Comparative Sociology XXXVII(l-2). Fuller B and X Liang. 1998. 'Which girls stay in school? The influence of family economy, social demands and ethnicity in South Africa'. In: C Bledsoe, J B Casterline, J A Johnson-Kuhn and J G Haaga (eds), Critical Perspectives on Schooling and Fertility in the Developing World. National Academy Press: Washington. Department of Education. 1995. White Paper on Education and Training, Pretoria, March. 1997. Gender Equity Task Team Report, Pretoria, October. Franzway, S, D Court, and R W Connell. 1989. Staking a Claim. Sydney: Allen and Unwin. Gaitskell, D. 1990. 'Devout domesticity? A century of African women's Christianity in South Africa'. In: C Walker (ed), Women and Gender in Southern Africa to 1945. Cape Town: David Philip. Fine, M, L "Weis, L Powell, and L Mun Wong (eds). 1997. Off White: readings in race, power and society. New York: Routledge. Giroux, H and P MacLaren (eds). 1994. Pedagogy and the Politics of Cultural Studies. New York: Routledge. Govinden, B. 1998 'Woman searching in Southern Africa: questions of identity and difference'. In: M Cross, Z Mkwanazi-Twala and G Klein (eds), Dealing with Diversity in South African Education: A debate on the politics of a national curriculum. Cape Town: Juta. Holdstock, T. 1990.'Violenceinschools:discipline'.In:BMcKendrickandW Hoffman (eds), People and Violence in South Africa. Cape Town: Oxford University Press. Hughes, H. 1990. '"A lighthouse for African womanhood": Inanda Seminary, 1869-1945'. In: C Walker (ed), Women and Gender in Southern Africa to 1945. Cape Town: David Philip. James, D. 1994. 'Basadi ba baeng/Visiting women: female migrant performance from the Northern Transvaal'. In: L Gunner (ed), 21 Linda Chisholm and Elaine Unterhalter Politics and Performance: theatre, poetry andsong in Southern Africa. Johannesburg: University of the Witwatersrand Press. Kabeer, N. 1992. 'Feministperspectives in development: A critical review'. In: H Hinds, A Phoenix and J Stacey (eds), Working Out: new directionsfor women's studies. Brighton: Falmer Press. Kabeer, N. 1994. 'Development from a gender perspective'. In: N Kabeer, Reversed Realities: gender hierarchies in development thought. London: Verso. Kann, U. 1995.'Statisticalindicatorsandtheeducationofgirlsandwomen in Namibia'. Paper presented at the Conference on Changing Education in South Africa, Commonwealth Education Council, London, October. Kenway, J. 1995a. 'Marketing education: mapping the contours of a feminist perspective'. Paper presented at the British Education Research Association Conference, Bath. 1995b. 'Feminist theories of the state: to be or not to be'. In: M Blair, J Holland and Sheldon (eds), Identity and Diversity: gender and the experience of education. Clevedon: Multilingual Matters. 1995c. 'Having a postmodernist turn'. In: P Smith andPWexler (eds), Post-Modernism: education, politics and identity. Brighton: Falmer Press. Kenway, J and S Willis. 1993. Telling Tales: girls and schools changing their ways. Canberra: Department of Employment, Education and Training, Curriculum Division. King, E and M A Hill (eds). 1993. Women's Education in Developing Countries: barriers, benefits andpolicies. Baltimore: World Bank and Johns Hopkins University Press. Kuzwayo, E. 1985. Call Me Woman. London: Women's Press. Kotecha, P. 1994. 'The position of women teachers', Agenda 21. Lewis, D. 1993. 'Feminisms in South Africa', WomensStudies International Forum 16(5). 1994. 'Women and gender in South Africa'. In: VMaphai (ed), South Africa: The challenge of change. Harare: SAPES Books. Mac an Ghaill, M. 1994. TheMakingofMen:masculinities,sexualityandschooling. Buckingham: Open University Press. Macleod, C. 1995. 'Transforming pedagogy in South Africa: the intersection of the teacher as subject', Perspectives in Education 16( 1) Mager, A. 1992/93. '"Girls" wars, mission institutions and reproduction of the educated elite in the Eastern Cape, 1945-1959', Perspectives in Education 14(1), Summer. 22 Gender, Education and the Transition to Democracy Mahlase, S. 1997. The Careers of Women TeachersUnder Apartheid. Harare: SAPES Books. Manicom, L. 1992. 'Ruling relations: rethinking state and gender in South African history', Journal of African History 33. Marchand, M and J Parpart (eds). 1995. Feminism, Postmodernism, Development. Routledge: London. Marks, S. 1987. Not Either An Experimental Doll: the separate worlds of three South African women. University of Natal Press: Durban and Pietermaritzburg. Maynard, M. 1994. '"Race", gender and the concept of'difference" in feminist thought'. In: H Afsahr and M Maynard (eds), The Dynamics of "Race" and Gender: some feminist interventions. London: Taylor and Francis. McCIellan, A. 1993/94. '"And women, too, will play their part": the relevance of gender to equal education in South Africa', Perspectives in Education 15(1), Summer. Mkwanazi-Twala, Z. 1995. 'The participation of black researchers in the education policy-making process in South Africa: Post 1990'. In: G Kruss and H Jacklin (eds), Realising Change: education policy research: Kenton 1994. Cape Town: Juta. Morrell, R. 1992. 'Gender in the transformation ofSouth African education', Perspectives in Education 13(2). 1994. 'Boys, gangs and the making of masculinity in the white, secondary schools of Natal, 1880-1930', Masculinities 2(2). 1998a. 'Of men and boys: masculinity and gender in Southern African studies', Journal of Southern African Studies 24(4). 1998b. 'GenderandeducationUheplaceofmasculinityinSouth African schools', South African Journal of Education 18(4). Moser, C. 1993. Gender Planning and Development. London: Routledge. Muller, J. 1996. 'Dreams of wholeness and loss: critical sociology of education in South Africa', British Journal of Sociology of Education 17(2). Nkomo.S. 1991. 'Organising women in SANSCO: reflections on the experience ofwomen in organisation', Agenda 10. Ntlantla, P. 1992. A Life's Mosaic. Berkeley: University of California Press. Nzimande, B and S Thusi. 1990. 'War and schooling: the impact of political violence on African secondary schooling in Natal', Education Policy Unit: University of Natal, Durban. Parpart, J. 1995. 'Deconstructing the development "expert": gender, development and the "vulnerable groups'". In: M Marchand and 23 Linda Chisholm and Elaine Untcrhalter J Parpart (eds), Feminism, Postmodernism, Development. London: Routledge. Ramphele.MandEBoonzaaier. 1988. "The position ofAfrican women: race and gender in South Africa'. In: E Boonzaaier and J Sharp (eds), South African Keywords: the uses and abuses of political concepts. Cape Town: David Philip. Russell, D. 1991.'Rape and child sexual abuse in Soweto: an interview with community leader Mary Mabaso', South African Sociological Review 3(2). Samson, M. 1996. 'Stripping the NQF down to its human capital jockstrap'. Paper presented at the Kenton at Wilgespruit Conference, Johannesburg. Savage, M and A Witz (eds). 1992. Gender and Bureaucracy. Oxford and Cambridge: Blackwell. Schindler, J. 1997. "The women's budget: education'. In: D Budlender (ed), The Women's Budget. Cape Town: IDASA. Schultz, T P. 1993. 'Theeconomicsofwomen'sschooling'.In: JKerConway and S C Bourque (eds), The Politics of Women's Education. Ann Arbor: University of Michigan Press. Sebakwane.S. 1993/94. 'Gender relations in Lebowa secondary schools', Perspectives in Education 15(1), Summer. 1994. 'Pedi women and colonial legacies', Matlhasedi 13(3), November/December. Seekings, J. 1991. 'Gender ideology and township politics in the 1980s', Agenda 10. Segar, J and C White. 1992. 'Family matters and the state: policy and everyday life', Transformation 20. Sen, G and C Grown. 1987. Development Crises and Alternative Visions: Third World women'sperspectives. London: Earthscan. Sender, J. 1996. ILO Report, Geneva. Sutherland,C. 1991. 'Paying for stolen kisses'. In: S Bazzili (ed), Putting Women on the Agenda. Johannesburg: Ravan Press. Thorne, B. 1995. Gender Play: girls and boys in school. Brunswick, New Jersey: Rutgers University Press. Truscott, K. 1993. 'Gender in education', Education Policy Unit, University of the Witwatersrand, Johannesburg. Unterhalter, E. 1991a. 'Theimpactofapartheidonwomen'seducation in South Africa', Review of African Political Economy48. 1991b. 'Caneducationovercomewomen'ssubordinateposition in the occupation structure?' In: E Unterhalter, H Wolpe and T 24 Gender, Education and the Transition to Democracy Botha (eds), Education in aFuture South Africa: policy issues for transformation. London: Heinemann. 1995. 'Barriers to benefits: policy implications of the analysis of women's education in South Africa'. Paper presented at the conference on Feminism and the Marketisation of Education, Institute of Education, London. 1996. 'States, households and the market in World Bank discourses, 1985-1995: a feminist critique', Discourse: studies in the cultural politics of education 17(3). 1998.'Economicrationalityorsocialjustice?Gender,uieNational Qualifications Framework and educational reform in South Africa, 1989-1996', Cambridge Journal of Education 28(3). Von Kotze, A. 1996.'Thecreakingoftheword:afeministmodel?'In:SWalters and L Manicom (eds), Gender in PopularEducation: methods for empowerment. London: Zed Books. Wittenberg, M with C Pearce. 1996. 'Youth and unemployment: some perspectives from the S ALDRU survey'. In: L Chisholm et al, Out-of-School Youth Report. Education Policy Unit, University of the Witwatersrand, Johannesburg. Walker, C. 1982. Women and Resistance in South Africa. London: Onyx Press. Walker, C (ed). 1990. Women and Gender in Southern Africa to 1945. London: James Currey. Weiner, G. 1994. Feminisms in Education: an introduction. Buckingham, Philadelphia: Oxford University Press. Wolpe, A. 1988. WithinSchool Walls: the role of discipline, sexuality and the curriculum. London: Routledge. 1994.' A framework for gender sensitivity', Agenda 21. 1995. 'Non-sexism: rhetoric or reality?' In: G Kruss and H Jacklin (eds), Realising Change: education policy research: Kenton 1994. Cape Town: Juta. 25