‘Cascading participation’ and the role of teachers in a collaborative HIV and Aids curriculum development project

This paper presents findings of four Grade 6 teachers’ involvement as facilitators of a participatory action research (PAR) project conducted in three South African primary schools. Based on the results of Phase One research which indicated that Grade 6s learn about sexuality, Human Immunodeficiency Virus (HIV) and Acquired Immunodeficiency Syndrome (AIDS) from multiple sources, the Phase Two project designers developed a toolkit to help Life Orientation (LO) teachers consult learners on what they know and how they want to be taught. In each school, a curriculum development group comprising the participating teacher, learners, parents and an HIV and Aids specialist worked to enhance the official HIV and Aids curriculum using the information gathered each week by the teacher. This dialogue between the study participants represents the culmination of what we describe as the project’s ‘cascading participation’ research model, a term denoting the multiple levels of participant involvement in the study. Although theories of participation often depict a binary relationship between those with power and those without it, the implementation of this project shows how the official curriculum, cultural norms and low parent involvement can exert pressure at different levels to diminish teachers’ ability to facilitate social and educational change.


Introduction
Teachers in South African schools are expected to meet two broad sets of duties. They are required to fulfil their institutional duties in, among other things, implementing the official curriculum in the classroom setting, maintaining accurate academic records and complying generally with the school's policies of communication with learners, parents and colleagues (Heeralal, 2014;South African Council for Educators, 2002). Yet politicians and community members often also anticipate that teachers should monitor learners' wellbeing and inspire them to overcome the difficulties of growing up in the midst of violent crime, poverty and high levels of unemployment (Joyce & Mmankoko, 2014;South African Press Association, 2013). In the words of Joyce and Mmankoko (2014:19), "Society hopes that teachers can show learners that their lives and futures can be different". This set of aspirational duties goes above and beyond teachers' institutional obligations.
Referring to teachers' aspirational duties, Ferreira (2013:1), (emphasis added) states that [South African] schools are viewed as nodes of care and support -material support, as well as emotional, social and spiritual guidance. Teachers and school principals are typically valued as key figures in communities, who potentially possess the ability to make a difference in people's lives. Recognising teachers' particular position in society, researchers have begun to contribute to the discussion of what role teachers should play in driving social and educational change in the country (Ferreira, 2013). This paper adds to this important and growing body of literature by presenting the findings from a teacher-led curriculum development research project in three South African primary schools. The participatory study involved collaboration between the researcher, teachers, Grade 6 learners, and parents from three Cape Town neighbourhoods.
Participatory research methodologies, with their emphasis on collaborative knowledge generation among researchers and community members, have already influenced studies that examine teachers as agents of social change. Yet as Ferreira (2013) emphasises in her introduction to a recent Special Edition on Participatory Methodologies and Education Research, there is still work to be done on how researchers can use participatory methodologies to respond to communities' needs. It may be the case that many community members see teachers as "beacons of hope" (Joyce & Mmankoko, 2014:19), but the study we report on indicates that several factors can inhibit teachers' involvement as project facilitators. The restrictiveness of the official curriculum, prohibitive cultural norms and parents' resistance to joining the project exerted a counterforce to teachers' presumed authority and potential to make a difference. In presenting the methodological challenges encountered in this PAR project, this paper emphasises that before teachers can affect social change, they must confront power and resistance from different members of the school community. Teachers are neither wholly empowered nor disempowered and work within a set of relationships -including principals, the official curriculum, learners and parents -that frequently affects whether teachers are successfully able to lead a PAR project and facilitate change. Before describing in more detail the project and the role of the teacher in its cascading participation model, we present a brief history of the contested concept of participation as it appears in 'development' work.
'Participation': a brief history of the concept in the 'development' paradigm During the past 50 years 'participation' has become a mantra of development and has become an idea which has attained "the status of development orthodoxy" (Cornwall, 2006:49). Cornwall (2008) observes that the discourse of participation now charac-terises the work of community development practitioners, local and international nongovernmental organisations (NGOs), and global institutions such as the International Monetary Fund and World Bank. The notion of participation, used in these varied contexts, tends to refer to citizens' or communities' involvement in decision-making processes (Cornwall, 2008). However it has proven also, in Cornwall's (2008:269) words, to be an "infinitely malleable concept", an issue that has led social scientists to develop several typologies of participation.
Useful as analytical tools, typologies of participation presented in academic literature have tended to entrench bifurcated sets of power relations, even as they have tried to examine participation in practice. One example of this work is Arnstein's (1969) well-known typology of participation which consists of an eight-rung ladder, each rung representing varying degrees of citizens' potential to influence collective decision making. The theoretical implication is that for all instances of participation, two distinguishable parties are involved. The nature of participation depends on the degree to which the less powerful group becomes empowered through the project or programme.
Rural agricultural development work and scholarship has also analysed participation in binary terms. For example, Chambers ' (1994a, 1994b) research in East Africa and Asia argues that participation occurs between "uppers" and "lowers" (Chambers, 1994a:1). In his view, human relations are governed by power and weakness, or dominance and subordination. Individual people, or "magnets" (Chambers, 1994a:7), in this field of power relations struggle to effect change because the combined force of the field is strong enough to maintain the status quo.
To a large degree, scholars have tended to theorise forms of participation that assume a dichotomous set of relations between those 'above' -governments and fund-holders -and those 'below' -'the people', 'the poor' and 'the disempowered'. In contrast, PAR is a process involving cycles of action and reflection, in which practical knowledge is produced together with, and used to improve the lives of, individuals and communities that are most affected by the issues being researched (Chambers, 1994a;Ferreira, Ebersöhn & Botha, 2013;Gaventa & Cornwall, 2006;Reason & Bradbury, 2008). In PAR, people who are usually seen as the subjects of research and objectified as part of the researcher's gaze become active, participating co-researchers (Gaventa & Cornwall, 2006;Reason & Bradbury, 2008). It is a particular approach to participation that recognises and encourages a diffuse distribution of power.
Phase Two of the project here under review embedded its methodology in the PAR tradition. Its particular model of participation -cascading participation -contests the 'participation binary' outlined above by highlighting how participatory research in school contexts involves numerous relationships between differentially empowered stakeholders. The results from Phase Two show that a teacher-led participation process in South African primary schools is likely to be complex, as teachers attempt to manage relationships with learners, parents, school principals and education officials, as well as the expectations of each of these groups of people. The tiered process of participation, described below, exemplifies how teachers hold positions of authority even as they struggle to assert this authority within various relationships.

ASKAIDS: A collaborative HIV and Aids curriculum development project
The three-year, multi-country (Ghana, South Africa, Kenya, Tanzania, Swaziland and Botswana) HIV and Aids curriculum development research project -African Sexual Knowledges of AIDS (ASKAIDS) -was initiated and administered by the Centre for Commonwealth Education (CCE) at the University of Cambridge. Our term, 'cascading participation', applies to the way different groups were recruited to the project. In each country, the CCE partnered with a local African researcher who then engaged school personnel and an HIV and Aids expert to act as a consultant in participating primary schools. Teachers subsequently recruited learners and community stakeholders. Participation, therefore, consisted of one group recruiting another group, which in turn sought out the next group of participants. The cascade represents a flow of participation, but it also illustrates that there are several levels of power shaping the project's progress. The outcomes of Phase One that investigated children's knowledge and teachers' pedagogical practices regarding HIV, Aids and sexual relationships are presented elsewhere (McLaughlin, Swartz, Kiragu, Walli & Mohamed, 2012). This paper presents findings from the implementation of the second phase -an interactive curriculum development process -in South Africa during the 2011 school year, with a particular focus on how teachers related to the various participants and how this impacted on their facilitation of the project.

Cascading participation and the South African case study
In South Africa, the first tier of participation comprised a formal partnership between the Human Sciences Research Council (HSRC) and the CCE. The HSRC enlisted an HIV and Aids 'consultant' (a local youth worker with specialist knowledge) -this was the third tier of the cascade -and subsequently approached three school principals to commit to the aims of the study. The recruitment of LO teachers, learners and parents constituted the fifth and sixth tiers of the cascade. This multi-tiered process is what defines the cascading participation model and is what differentiates it from the 'participation binary' that appears so frequently in academic literature. Figure 1 depicts the cascading participation process using the South African project as an example.
In preparing for Phase Two of the ASKAIDS project, the research planners used the findings of the first phase to design a toolkit that researchers envisioned would guide an interactive curriculum development process. The ASKAIDS toolkit is a document which includes information on HIV and Aids, and describes why it is important to take into account young people's existing knowledge in curriculum construction. The toolkit also outlines several techniques for developing a co-constructed HIV and Aids curriculum, such as keeping a journal, role-playing, using suggestion boxes and using cameras to document learner participation during lessons.
Phase Two, which ran between February and September 2011 and incorporated the first three terms of the year, involved the creation of a curriculum development group (CDG) in each school, led by the participating teacher with the aid of the toolkit. Each group included four learners from the teacher's Grade 6 class, two parents, an HIV and Aids consultant, and the researcher, whose role was to observe and record discussion between participants. The PAR programme proposed that through each CDG, the LO teacher, learners and parents would draw on learners' experiences and existing knowledge of HIV, Aids and sexuality to design an HIV and Aids curriculum that could be incorporated into the lesson plans of the participating LO teachers and which could be sustained with minimal external support.
The CCE planned for each CDG to meet twice each term to support the teacher in curriculum design and to provide a platform for meaningful dialogue around issues of HIV and Aids knowledge, whereupon members of each group would inevitably confront others with their assumptions and values. If successful, CDG discussions would lead to the production of a hybrid curriculum that incorporated children's everyday knowledge into the official curriculum (Bernstein, 1996). Since the participation of learners and parents was contingent on the active participation of teachers, each teacher was a crucial intermediary between the CCE, the country-specific research organisation and the school community.
The following section on the methodology of the project expands the description of the Phase Two research. The section covers participant recruitment, data generation, and data analysis.

Methodology
The project planners designed the second phase of the ASKAIDS project so that it was standardised across the six countries involved in the study. They did so with a view eventually to undertaking a comparative analysis of the multi-site curriculum development process. We describe the different aspects of the project methodology below, focusing on participant recruitment, data generation -which involved qualitative and quantitative methods -and data analysis in the South African context only.

Participant recruitment
Having first obtained ethical clearance for the South African project from the HSRC's Research Ethics Committee, during the latter part of January 2011 the researcher invited three primary schools to participate in the ASKAIDS study. Though each of these schools was based in Cape Town in the Western Cape province, the areas in which they were located represented diverse social and educational contexts. The communities were dissimilar in their levels of HIV infections and distribution of poverty, but comparable in the tense interplay between HIV and Aids and religion, culture, and traditional practices. Two of the schools had been involved in the first phase of the ASKAIDS project, meaning that the principals of these schools were already familiar with the principal investigator at the HSRC. They readily agreed to participate in the second phase of the study.
Once the principal at the third school had expressed her interest in the study and recruited an LO teacher to lead the project, learner recruitment -the next tier of participants in the cascading model -began in all schools. LO teachers invited the learners in their Grade 6 class to participate in the project. Learners' eventual involvement in entry and exit questionnaires was determined by whether they gave their signed assent along with a parent's or guardian's formal consent. Each LO teacher also invited four learners -if possible, two male and two female -to volunteer to participate in the CDGs. The researcher suggested that these learners should be representative of the class as a whole and not simply those who performed best in assignments, although there was ultimately no way to 'test' the representativeness of the group. Finally, each LO teacher recruited, or attempted to recruit, two parents or community stakeholders to the CDG. Table 1 summarises the number of participants recruited into the study, distinguished by school and participant type. Table 1 shows that fewer learners in School 1 participated in the initial survey than in the other participating schools. This was because the parents or guardians of nine of the 34 learners approached by the teacher did not return, or declined to sign, the consent forms needed for their child to complete the questionnaire. This was the same school in which the LO teacher stated that he was unable to recruit any parents to the study. We discuss the implications of this inability to involve parents in the project later.
Table 1 also indicates that the teacher at School 3 recruited one extra female learner to the CDG. This differed from the original plans for the CDG, which suggested that teachers should recruit four learners to the discussion group. When the researcher consulted the teacher about the change, she explained that the five learners had volunteered to be part of the process and she felt that she could not turn one of them away. Since having one extra learner in the CDG did not pose a challenge to the structure of the study, the researcher deferred to the teacher's judgement, especially considering she had been entrusted with managing the process.

Data generation
During the course of the eight-month data collection period, the HSRC researcher attended all CDG meetings, each of which was held in a school classroom and lasted approximately 45 minutes. He made detailed observation notes and summarised these, together with classroom observation notes, in four quarterly research reports. In addition to participant observation, the research methodology comprised individual interviews with teachers, stakeholders and CDG learners on a termly basis. The researcher also conducted interviews with the school principals, the HIV and Aids consultant and held focus group discussions with Grade 6 learners at the start and end of the project. The researcher recorded these interviews after gaining informed consent from the participants and from parents or guardians in the case of learners. All of the interviews and focus group discussions were semi-structured in that the researcher followed an interview schedule yet also pursued lines of conversation which deviated from the schedule but tied in with the six overarching research questions (Kvale, 1996;Maxwell, 1996). These research questions provided the framework for the interview schedule for each term, even though the schedule itself differed so that it was relevant to the particular stage of the research. Three examples of the overarching questions are: 1) In building a hybrid curriculum in HIV and Aids and sex education through consultation, what use is made of the young people's and community stakeholders' knowledge? 2) What are the obstacles and facilitators of the dialogue process for teachers, learners and community stakeholders? 3) How does using the toolkit in CDGs and in classroom settings affect the curriculum's content and pedagogy? The remaining questions tried to elucidate the implications of a discussionbased approach to HIV and Aids and sex education in schools, and for teacher education and preparation in South Africa.

Data analysis
The South African research team used qualitative analysis software to code the quarterly research reports and researcher field notes, as well as the transcripts of interviews, focus group discussions and the end-of-year reflective workshop. These codes were used to conduct a thematic analysis (Boyatzis, 1998;Miles & Huberman, 1994), which examined how the different themes functioned in the data and how themes related to one another.
Whilst ensuring complete reliability of a study is almost impossible, certain steps can be taken to ensure that the study is trustworthy and the "quality of its craftsmanship" (Kvale, 1995:26) is evident. In order to achieve credibility in the study we made use of a team approach to analysis and a community of enquiry, or "interpretive communities" (Kvale, 1996:246). As frequently as was possible, researchers from the different countries would gather to comment on findings from the other countries involved in the study. In the South African context we used multiple coders of data and worked as a team to come to analytical conclusions. According to Miles and Huberman (1994:263), such a team approach helps ensure against "sources of potential biases … checking for representativeness and for researcher effects … ruling out spurious relations". The following section presents and discusses the results from the study. We have used general identifiers throughout the paper to refer to schools and participants to make certain of their confidentiality.

Discussion of results: Inhibitors of the cascading participation approach
As we described earlier and depicted in Figure 1, the success of the ASKAIDS project required different stages of participation. One group needed to join the project in order for the next group of people to come on board. Despite the importance of each cascade, we focus on the extent of teacher participation because their involvement impacted the core aims of the project most directly. In this regard, School 2 was most successful. The LO teacher responded positively to the researcher's requests and collaborated with the Grade 6 class teacher to recruit parents and learners to the CDG. In Schools 1 and 3, however, LO teachers struggled to convene CDG meetings, partly because of a lack of parent involvement. This section presents three themes that, to varying degrees, limited teacher involvement and, therefore, the cascading participation model in the three participating South African schools. Each of these findings is described and analysed in turn.

Teachers' commitment to deliver the official syllabus inhibited their participation
One of the most consistent themes to emerge from teacher interviews was that teachers experienced intense pressure to cover the official curriculum in full despite feeling pressed for time. For example, the LO teacher from School 1 commented during her exit interview that "you must have somebody to run this [ASKAIDS] to the schools. Don't use us". She described the ASKAIDS programme as an imposition or "burden" on her schedule and stated she would not in future want to lead the project on her own. The School 1 LO teacher perhaps insisted most strongly that she was too busy for the ASKAIDS project, but the following excerpt from the researcher's first quarterly report reveals that a sense of teacher fatigue was common in each of the three schools, though not necessarily equally pronounced: "The greatest challenge (in this [School 1] and other schools) has been to work around teachers' timetables, so as not to put them off the project completely, while at the same time motivating them to convene meetings and interviews". As this statement suggests, participating teachers tended to prioritise their official teaching commitments ahead of the ASKAIDS project during the early stages. Teachers at Schools 1 and 3 never fully integrated the programme with what they perceived to be their core duties. In effect they seldom consulted the toolkit and only sporadically followed the suggestions for classroom activities. The School 3 teacher expressed her hesitation with using the toolkit to incorporate the ASKAIDS project into her teaching programme in a comment she made during her exit interview: Maybe This sentiment around the inflexibility of the school curriculum surfaced on another occasion. The School 2 LO teacher, who was in fact most comfortable using the toolkit to consult learners on their opinions, stated: "Because this is a school structure, we are governed by certain rules and regulations. So you gotta go according to the syllabus as well. The toolkit and syllabus need to work hand-in-hand, on a Grade 6 level". The evidence from these interview extracts shows how participating teachers felt that they worked within strict organisational and curriculum frameworks. From the teachers' perspectives, it was not a simple matter to integrate the ASKAIDS activities into their lessons. And while the ASKAIDS content fitted well with lessons dealing with HIV and Aids, the Western Cape Education Department's (WCED) HIV and Aids module was only scheduled to be delivered in the third term, placing it outside the teachers' lesson plans for the first two terms. In this sense, the standardisation of the time frames and outputs of the six-country project presented a challenge to the implementation of the South African study. The South African teachers' timetables were set ahead of time, and though teachers in other participating countries were on the whole able to accommodate the research programme, teachers in the three Cape Town schools -Schools 1 and 3 in particular -expressed concern that they did not want to deviate too far from the government-prescribed Grade 6 curriculum.
This section has shown that although LO teachers are well placed to facilitate the co-construction of knowledge of HIV, Aids and sexuality, they felt constrained by the official curriculum to focus on pre-set topics in each lesson. Only the School 2 teacher felt comfortable introducing some of the toolkit activities into his lessons. Two issues contributed to this general sentiment: the first was structural in that the LO curriculum, and the time in which it could be delivered, was set by the WCED. The second issue proved to be individual. The School 2 teacher expressed some concerns with the toolkit and implementation of the project, but on the whole was enthusiastic about using the toolkit suggestions to consult his Grade 6 class. In contrast, teachers in Schools 1 and 3 felt uncomfortable leading the ASKAIDS project and preferred to stick to the curriculum rather than consult learners on their existing knowledge of HIV, Aids and sexuality. The following section elaborates on teachers' personal anxieties around consulting learners, some of which linked to traditional discomfort around adult-child discussion of sexuality.
"It's just not easy": Teachers uncertain how to discuss sexuality with learners Teachers at South African schools in impoverished communities experience the dayto-day challenges of working and, sometimes, living in under-resourced areas. However, they are also often better educated, speak more English and have higher incomes than people in the communities in which they work. Yet despite having a deep understanding of the conditions in which the Grade 6 learners live and possessing a greater skills cache than many other community members, teachers nevertheless did not always feel confident facilitating conversation during the CDG meetings or leading classroom discussions concerning sex. The following interview quote from the School 3 teacher points to how some teachers, particularly those in Schools 1 and 3, did not know how, in the consultant's words, to "transition" from their traditional classroom role to being a facilitator during CDG meetings: "At the beginning it was difficult. The children were not free because I was there. Because I realised that the time I come in then they stop asking questions, you know". Speaking more generally, the School 3 teacher stated that "I can say that because I've been doing it [teaching LO] for such a long time, so at least I'm a bit confident. But not as yet to go into detail, like, sexual education". A factor that contributed to teachers' inability to discuss sex and encourage learners to do the same was the frequently referred to cultural norm that adults should not speak openly to children about sex and sexuality. The following conversation between the researcher and the School 3 teacher illustrates how cultural norms inhibited the role of the teacher as change agent . In School 1, the female LO teacher stated in an exit interview that learners of both sexes "are more free to male teachers than female", which she felt constrained how she was able to interact with learners in the project. Yet the male class teacher at the same school, who worked together with the LO teacher in coordinating the project, stated the exact opposite, namely, that "such things cannot be dealt with by a man, male teachers; such things can be dealt with by a female teacher. So, those are the stereotypes, the barriers". Despite the clear contradiction between these comments, it is clear that teachers' perceptions of cultural norms inhibited the role they believed they could play in CDG meetings and in the classroom setting.
The cultural taboo of talking about sex with children was not as pronounced in School 2, where the teacher was more assured of his role as a facilitator. Nevertheless, the LO teacher described the negative reaction from the school community when he had tried previously to use a school concert to address the "real, darker" issues of sex, teenage pregnancy and drugs in learners' lives. He asserted that "the religious people in the community, they wanted the more perfect side. But you also got to look at the other side as well".
The combination of uncertainty at leading conversations with learners and parents and the uneasiness that accompanies any discussion of sexuality, HIV and Aids led each of the South African teachers to state in interviews that the HIV and Aids consultant was integral to having productive CDG meetings. The School 1 LO teacher described the consultant in the following terms during an exit interview: "He knows what he's doing, he knows about HIV and Aids. He knows about kids, you know. …So he is so broad-minded". Similarly, the School 3 teacher noted that "he's also comfortable, you know, talking about sexual issues, you know… . Because he's very, unlike myself. So they [learners] find that it's very easy, yes, to speak to him". The researcher commented in his observation notes that learners were initially shy to express their views in front of adults, but that the consultant's self-deprecating manner and enthusiastic reception of any information offered by learners eventually led to them becoming more relaxed during meetings and their contributions more substantial. One learner, who stated that he felt "nervous" in his teacher's' presence, remarked that "when he [the consultant] is here, I talk, I don't think, I take him out of my mind. …I talk as if he is not there". In contrast, it was frequently the case that teachers did not manage to comfort learners and encourage them, in learners' terms, to "feel free".
The South African consultant ordinarily worked in the NGO sector helping to mentor learners in Grades 8-12 on developing leadership qualities. His background meant that he was ideally placed to support teachers in generating participation among the next tier of participants, namely parents and learners, and to help teachers develop a collaborative pedagogy. Ultimately, however, the consultant tended to lead the CDG meetings rather than facilitate them. The fact that some of the teachers felt uncomfortable and unprepared to facilitate dialogue on sexual knowledge and practices meant that they gave over much of their CDG responsibility to the consultant. Understandably, this impacted negatively on the cascading participation approach, as the consultant tended to lead rather than facilitate meetings with the result that teacherlearner dialogue was not always evident.

Insufficient teacher support from parents
Parent involvement in school life is something that impoverished schools in many contexts, including South Africa, struggle to attain (Epstein, 1995;Hallgarten, 2000).
Working-class parents have often had negative experiences of schooling or, as Pitt, Luger, Bullen, Phillips and Geiger (2013) observe in their South African-based study, have themselves received very little education at all. Hallgarten (2000) notes that parents in these circumstances are frequently intimidated by the idea of participating in school life. The ASKAIDS project in South Africa experienced this same trend of low parent participation as teachers struggled to recruit parents, or other community stakeholders, to the CDGs. According to the researcher's observation notes, no parents attended CDG meetings in School 1. At School 3 one parent was present at the second of the three meetings. School 2 managed to secure two parents for the first meeting and a single parent at each of the following three meetings.
Researchers have suggested that it is primarily the responsibility of teachers to create policies and practices that increase parent involvement. They contend that teachers are the more powerful group in the teacher-parent relationship (Hallgarten, 2000;Van Wyk & Lemmer, 2009). However, as Van Wyk and Lemmer (2009) observe, many teachers do not have the experience or knowledge of how to encourage parent participation in school life. Indeed, when teachers in Phase Two of the ASKAIDS study expressed frustration at what they believed was a lack of parent participation in school life and the lives of learners, they did not have alternative plans to try and involve parents. The School 2 principal exemplified the teachers' exasperation when he stated in an interview that the school is often left to do what he perceived to be the work of parents: "I think you can read it like this,…the more someone else can do and the less that I can do as parents, that's fine, that suits them. And often we find ourselves going beyond the call of duty for the sake of the kids". The School 1 class teacher echoed this sentiment when he asserted that "parents I would say are not giving enough -100 percent support in dealing with HIV and Aids issue". He later commented that "only one parent…came to me and I explained what is the good part of the project, what are the objectives of the project, and then she signed and said 'big yes'". The greater significance of the difficulty teachers faced in recruiting parents is that it inhibited the cascading participation model of the ASKAIDS study. The 'flow of involvement' was unable to run down from the teachers to parents. Since parents are, in the words of Pitt et al. (2013:3), "critical in the wellbeing and the healthy development of the child, and the child's functioning and progress at school", this finding indicates a possible limit to teachers' authority to mobilise change in the school community and bridge the school-home divide.

Conclusion
School-based studies, like ASKAIDS, often require that teachers act as project facilitators and mediators between programme designers and the ultimate recipients, in this case learners and parents. As such, forms of what we have referred to as cascading participation -processes in which a teacher leads a PAR project and interacts with multiple, differentially empowered participants -are likely to be relatively common in school contexts, although we are not aware of a systematic study of this phenomenon. In presenting evidence of three issues which inhibited teachers' involvement in the South African ASKAIDS programme and ultimately stunted the cascading participation process, we hope to have contributed to a more nuanced understanding of the relationships which teachers have to negotiate when implementing an HIV and Aids PAR project. The factors which we identified as inhibitors of teachers' involvement in the ASKAIDS project and consequently the cascading participation model are as follows: (1) Teachers' commitment to deliver the official school syllabus inhibited their participation in the programme; (2) Teachers were uncertain how to discuss sexuality with learners and also culturally reluctant to do so; and (3) Teachers received insufficient support from parents and were ill-equipped to enlist further help.
The results from Phase Two of the ASKAIDS project show that a teacher-led participation process in South African primary schools is likely to be complex, as teachers attempt to manage simultaneously the expectations that accompany relationships with learners, parents, school principals and education officials. The tiered process of participation illustrates that teachers are positioned within a set of relationships that can limit their ability to successfully lead a PAR project and act as change agents in their community. It also exemplifies how school-based PAR projects that resemble the cascading participation model differ from conceptualisations of 'participation' in the academic development literature, which generally describe binary relations between 'lowers or the poor' and 'uppers or the powerful'.
Although this paper has focused on the challenges to a cascading participation approach to PAR, it has argued that such a methodological and analytical approach to participation reveals, and offers the opportunity to confront, the multiple levels of participant involvement that inevitably arise in school-based projects. Considering the emphasis it places on multiple relationships, the cascading participation approach holds the potential to become a sustainable component of school life, driving whole school development by strengthening a range of relationships that add value to the school as an institution. The challenge for future researchers, practitioners and teachers will be to devise project designs that teachers are able to adapt to particular schools and contexts, thus avoiding the limitations of an inflexible 'one-size-fits-all' approach.