Education for rural development: forty years on

Education has long been considered a force for social transformation, influencing teaching-learning approaches and policy, including the 2030 Sustainable Development Agenda. This article sets out to problematise the relationship between education and social change, proposing a theoretical framework around social transformation as an alternative to the usual development lens. A historical analysis of an education for rural development project implemented in Nepal in the 1980s reveals that curriculum and training approaches were strongly influenced by the assumption that education could initiate economic and cultural change. The paper argues that this focus on educational interventions as facilitating planned development outcomes limited analysis not only of the project ’ s impact, but also of the broader changes that have taken place in these communities over the past four decades. By contrast, a social transformation lens can help shift attention from formal educational providers to investigate political and commercial actors, amongst others. Such an approach can offer rich insights into informal learning spaces and new communicative practices which have transformed people ’ s lives.


Introduction
'There is no more powerful transformative force than educationto promote human rights and dignity, to eradicate poverty and deepen sustainability' (UNESCO, 2015: 4).
In 1981, an innovative educational programme was launched in the Far Western region of Nepal. The Education for Rural Development Project (ERDP) in Seti Zoneknown locally as 'the Seti Project' -was underpinned by beliefs similar to those expressed by Irina Bokova in the above quotation. In an area with limited infrastructure, poor communications and huge social inequalities, education appeared to have the potential to improve communities' wellbeing and livelihoods, and change attitudes, particularly towards women and low caste communities. Seen as a pilot for the whole country, the Seti project was designed to 'convert the educational system into a powerful force for rural development' (UNDP, 1981: 1). Education was intended to kickstart development in a region viewed as 'one of the most backward zones in the country' (ibid: 5).
When I began work on the Seti project in 1985, based in Bajhang district as a VSO teacher trainer from the UK, it was with a certainty that a new kind of education could or would lead to rural development. In March 2022, I returned to Bajhang for the first time since then to talk to former students, colleagues and educationalists about the changes they have seen, and their views on the purpose and challenges in the education sector today. Walking through Chainpur, the district headquarters, down streets of large concrete houses and shopsall with imported goods from the Terai 1 and beyondit was difficult to visualise the old town where the main commercial activity was local villagers coming in with dokos 2 to sit under the peeple tree and sell bidis (locally made cigarettes) or even selling ice from a nearby mountain peak. Even more striking was the social life nowfashionably dressed young girls walking down the bazaar, chatting, buying home-made ice lollies and stopping to ask if I would 'do a Tik Tok' with them on their phones! Forty years ago, these girls would have been working at home or out collecting fodder for the cattleand it was very rare to find a girl who had been allowed to attend high school.
So, at last development appears to have arrived in this part of the worldat least in the district headquarters. But where does education and the Seti project come into this picture? It would be impossible to attempt to evaluate the impact of a specific educational intervention forty years on, and this paper does not set out to do this. It is also beyond the scope of this research to identify which specific factors have contributed to social change in Nepal. However, my historical analysis of the Seti project ideology and practices can provide a valuable lens on education and development approaches that still continue in Nepal and elsewhere in the world today. Through taking a comparative historical perspective on education in this area of Nepal, this article sets out to develop an alternative theoretical framework and to propose different ways of understanding the relationship between education and social change. 'Education for sustainable development' has now replaced the 1980s' mantra of 'education for rural development' -but many of the earlier assumptions around the 'transformative force' of education are similar. My starting point in this article is that as developers, our focus on education may have led us to downplay the 'transformative force' of other playersparticularly commercial actorsand to consider the ways in which technology, social media and political movements can change social attitudes and facilitate informal learning.
I will begin by problematising the concept of 'education for rural development/sustainable development', through reviewing the literature on social change and social transformation as an alternative to using 'development' as a frame of analysis. My aim is to propose a theoretical framework that can help to capture unintended and unplanned learning and change, rather than focusing exclusively on formal educational and development interventions.

From development to social transformation
The relationship between education and rural development has often been seen unproblematically as linear and with only positive outcomes, supported by evidence on the impact of enhanced literacy rates or schooling outcomes (particularly for women and girls) on indicators like maternal child health (Robinson-Pant, 2004). This body of research literature (see for instance, Caldwell, 1979;Cochrane, 1979;LeVine et al., 1991) strongly influenced policy and programmes like the Seti project, not only in terms of their aims but also their approach to teaching and learning which set out to initiate behavioural change. For instance, functional adult literacy curricula are often linked directly to specific development skills and messages, such as didactic stories on the benefits of small families (Acharya and Robinson-Pant, 2019). To develop a more holistic approach on complex processes of social change and learning, I turn here to the literature on social transformation as an alternative to a development lens.
Comparing the concepts of development and social transformation, Castles (2001: 15) suggests: 'Development implies a teleological belief in progression towards a predetermined goal: usually the type of economy and society to be found in the 'highly developed' western countries. Social transformation, by contrast, does not imply any predetermined outcome, nor that the process is essentially a positive one'. He contrasts the 'predetermined' and hierarchical goals of development (regarding the assumption that countries move from being 'low' 'to 'highly' developed) with the unpredictability of social transformation. There is also the recognition that there can be both negative and positive consequences -of development and of social transformation -for communities. Castles (2001: 15) ) goes on to say that 'development is a question of instilling the "right" orientations, values and norms into the culture of the non-Western world'. His statement could be challenged from the perspective of participatory developmentwhich sets out to respond to people's aspirations and build on their existing cultural practices. However, dominant development discourses have tended to promote not only similar values to the 'Western' world, but also similar practices, whether around gender equality training, income generation or schooling. This relates partly to recent policy debates around promoting 'universal values': Bourn (2022: 26) comments that 'Universal values may be well intentioned, and those such as justice, peace and sustainability are normatively considered a common good'. Yet he points out the dangers if specific national and cultural voices are not recognised and 'universal values' are imposed on such groups.
Unlike development, social transformation does not set an agenda for change, nor assume a certain endpoint 'to which societies evolvesocial change is everywhere and constant' (Haas et al., 2020: 39). Discussing the rapid growth in transnational linkages in the last quarter of the 20th century, Castles (ibid: 18) sees social transformation and globalisation as intrinsically linked and observes that globalisation is 'leading to new forms of social differentiation at the international and national levels'. Arguing that 'for most people, the pre-eminent level for experiencing social transformation is the local' (ibid: 25), he points to the importance of researchers exploring the 'different ways in which globalising forces affect local communities ' (ibid: 18). This multidirectional lens offers great potential for capturing the dynamism of social change in rural communitiesand moves away from the more static input-output model of development discourses which assume that education is the main vehicle for bringing in new ideas and practices. Taking the lens of social transformation does not imply ignoring developmentas development ideologies and practices are also part of the bigger picture of social change.
There is much discussion in the literature on the difference between social change and social transformation, with consensus on the two being intrinsically linked and in simple terms, that social transformation is 'big change' (Haas et al., 2020: 12). Castles (2015 : 4) describes social transformation as 'a shift in social relationships so profound that it affects virtually all forms of social interaction, and all individuals and communities simultaneously'. Similarly, Haas et al. (2020: 16) refer to social transformation as 'macro level fundamental change in the deep structures and organisation of society, affecting all dimensions of social life', as compared to social change which 'refers to the micro-and meso-level, day-to-day and cyclical changes that occur all the time'. In their conceptual framework, they also bring in 'events', which can be 'either manifestations of deeper social transformations… or causes of transformation in themselves' (ibid: 17).
In my own analysis, I am less concerned about when social change counts as social transformation, and more interested in what this tells us about how to analyse processes of change. This is where Yadav's (2016:7) plea for a 'closer inspection of unnoticed, silent changes' is particularly valuable. In her research with women in Nepal, she focuses on those moments when people begin to question previously taken-for-granted practices or beliefssuch as 'the liberation of widows wearing the white sari' (ibid: 125) as a symbol of purity and a life of celibacy. To analyse how or why such changes in attitude take place, she draws on Adkins (2003:1): 'It [social change] is possible when there is a critical reflexive stance towards formerly normalisedor at least taken-for-grantedsocial conditions.' In a different cultural context, Pedwell (2021 : xviii) suggests that we often 'understand social change in the "major" key' and points to our tendency to focus on 'turning points' when we tell stories of socio-political transformation (ibid: 11). A key concept within her analysis is 'habit' and she emphasises the need to 'understand habit and its relation to social change across diverse situations, relationships and contexts' (ibid: xviii). This literature points to ways in which we can analyse how and why social change takes place, including how people engage with and understand change in their own lives.
I argue that this move from the lens of 'development' to social transformation and social change offers the potential for a holistic investigation which aims to make values and ideologies (including development ideologies) more visible within processes of change. The recognition of transnational and globalising forces and interconnections between local and global can help to make explicit relationships that often remained implicit within earlier development-focused research in education.

Education for social change
Drawing on the above analysis of the differences between development and social transformation theories, an important consequence of the latter (social transformation) is that education is not necessarily 'for' social change (i.e. a predetermined goal). Yet, educationin the broadest sense of informal learning and engagement with knowledge -is clearly integrally linked to and embedded in social change. As Sharma and Monteiro (2016: 72) observe 'social change leads to transformation in thinking which then influences behaviour patterns in society'. What is unusual in their statement is the focus on learning ('transformation in thinking') through social change, rather than the other way round (education initiating transformation in thinking).
The idea that education has a central role to play in social change has informed radical critical pedagogies dating from Freire (1970), but also lies behind many state-controlled interventions in school curricula which aim to promote more conservative ideologies (for instance, see Durrani, 2008). Within these diverse approaches, the assumption is that a certain kind of education will lead to certain kinds of social, economic or political change. Rather than arguing whether education or social change comes first, I am interested in the kinds of learning that are facilitated for and through social change. Bourn (2022: 5) points to the risk that 'theories of change can often result in a linear process of learning' and questions whether education should be 'interventionist with explicit learning outcomes related to forms of social action or whether the power of education is important in itself' (ibid: 7). I would add here that for many marginalised communities, it is also the belief (of students, parents, teachers) in the power of education or the symbolic power of education that is important (Robinson-Pant, 2016). Discussing UNESCO's recent initiatives to promote a rights-based approach to education within the sustainable development agenda, Bourn (2022 : 16) warns that this 'could lead to a neglect of key questions such as the aims and purposes of education, the content of the curriculum and approaches towards teaching and learning' -and importantly, could ignore the role of teachers. Analysed through the lens of social transformation, the rights-based approach to education could be seen as promoting a specific political agendaand links to my earlier discussion of 'universal values'.
'Education' is commonly used synonymously with 'schooling', thus limiting how we see learning in relation to social change. The concept of lifelong learning has become more prominent within policy discourses since the launch of the sustainable development agendaand helped to shift attention onto adult, youth and pre-school learning. However, there is still a tendency by governments and other providers to see lifelong learning in terms of 'programmes' and planned interventions, whether within formal institutions or non-formal skills development programmes. A social transformation lens widens the focus to include spontaneous and unexpected learningand here theories of informal learning are valuable.
Informal learning is often considered simply as the opposite of formal learning, having no defined structures, timing, place or curriculum objectives. Rogers' (2013) distinction between different types of informal learning is useful here to begin to analyse how learning takes place through social transformation. He identified informal learning as: incidental, task-conscious, self-directed and/or unintentional, and went on to bring together ideas about how adults learn informally, including: 'natural, like breathing', a process of osmosis or assimilation, situated learning, task-oriented learning (e.g. learning to use a new camera), social learning (through informal scaffolding by peers) and individual learning (ibid: 18). These examples are useful to begin to expand the discussion beyond the common informal-formal binary, and inform my comparative analysis below.

Methodology
I set out initially to develop a retrospective analysis of the Seti project through combining a short period (ten days) of ethnographic-style observation 3 and informal discussions with educationalists in the Far Western province, with analysis of historical documentation and informal interviews with three former project staff in Kathmandu. However, when I visited Bajhang in March/April 2022, I realised that this approach could put limited parameters on the research through my proposed focus on conversations with educational practitioners and visits to schools. In order to develop a wider lens (i.e. social transformation as opposed to 'development'), I opened up the research investigation beyond formal educational institutions and talked informally with around 40 people (teachers, students, political activists, shopkeepers, educational officials, NGO workers, journalists, manual workers). I decided to develop a historical analysis of education and rural development across forty years, through combining my own fieldnotes, diaries and project documents from the 1980s with ethnographic-style research in a wide range of contexts, guided by the question 'how do people in Bajhang view education today, as compared to the 1980s?' However, given the time limitations, I saw this ethnographic-style research as revealing only the 'tip of the iceberg' and am hesitant to put much weight on my initial findings into social change in this area. My main aim was to explore what insights might be offered through taking a social transformation lens on this contextas opposed to a development framework, with a view to conducting longer-term ethnographic research in the near future.
I was not expecting to be able to trace colleagues and students with whom I had worked in the 1980s, but the news of my arrival in Chainpur spread quickly, mainly by word of mouth. I had met beforehand with a former Seti Project colleague in Kathmandu who provided me with phone numbers for retired teachers in Bajhang and briefed me with updates on schools and development programmes in the area. When we worked in this region in the 1980s, the only form of communication with Kathmandu was the project radio -dependent on solar batteries and not reliable in the monsoon. Nearly everyone now has a mobile phone, and I was soon receiving calls from former students, teachers and neighbours who still lived in Chainpur, as well as from people outside the area. I was delighted to meet up again with so many friends and colleagues in an intense weekthough this also raised ethical issues in relation to the boundaries between research and 'catching up'. I had received ethical clearance from my university in advance, and ensured that informed consent was given if conversations were to be used as data. But it was more a question of understanding people's expectations and exploring how I could reciprocate for their time and generosity in such a short time. I had taken large hard copies of photos from thirty-five years ago, including of school staff and pupils, and these proved a popular talking point with both young and old. Many people stopped me in the street to take photos of the photos and to talk about what had happened to the spaces, people and buildings since then. As I had been one of the few people with a camera in the 1980s, it was an unusual opportunity to share memories and I later deposited the photos in the town library. A local journalist also added the e-photos to his Facebook page 'getting to know Bajhang'.
For the historical dimension of the research, I reviewed project evaluation reports and teaching materials (all available in Kathmandu), my own fieldnotes from classroom observations, notes and articles shared by colleagues and my diaries. I also talked to retired Seti project colleagues about their memories of that time and their reflections since thenmany of which are captured in a recent collection published in Nepali by a high school in Bajura (Janaprakash Secondary School, 2022). All these sources provided rich insights into educational 3 This distinction between 'ethnographic-style' and 'ethnographic' research taken from Green and Bloome (1997) was explained by Brian Street: 'an ethnographic style goes much further than just using some ethnographic tools. It means adopting an ethnographic perspective, a de-centring, removing power from the researcher.' (Openjuru et al., 2016: 23). experiences in this area in the 1980s. The recent ethnographic-style research gives an idea of some changes that have taken place sincebut it is important to note that this activity should not be regarded as a 'tracer' study or an impact evaluation of the Seti project. As indicated earlier, the main aim of my study was to take a different lens (that of social transformation) on a project which had been strongly influenced by education and rural development discourses in the 1980s.

Introducing the Seti project aims, activities and curriculum
The Education for Rural Development Project was established in 1981, funded by His Majesty's Government of Nepal, UNESCO, UNDP and UNICEF. The aim was to 'develop a system of basic education that will serve to promote rural development in Nepal' (UNDP, 1981:1) by piloting the approach in the Seti zone, a region known for its poor communication facilities and 'urgent need for development' (ibid: 6). At that time, female literacy rate (over age 8 years) was only 1.9% in the region and only 1.8% of secondary school enrolments in Bajhang were female (ibid: 5). An evaluation report (CERID, 1986: 2) observed that alongside poor transportation, health, agriculture, water supply and no electricity, 'the people are steeped in poverty, ignorance and superstition and are characterised by social diversity and caste distinctions. Women work for more hours than men but have little say in household decisions.' The idea was to pilot a new model of teacher training, curriculum development and school building which could eventually be taken up throughout the whole country. This did happen to some extent when the Basic and Primary Education Project was launched in the 1990s, though Seti project staff reflected that some of the core principles of using local materials and resources were not adopted in the nationwide programme (from personal comments).
The objective of localising the curriculum and strengthening local teaching capacity underpinned all strands of the Seti Project activityand is particularly relevant today at a time when federalisation has led to the current Government's requirement for local governments to produce their own localised curricula (see National Framework on Education 2076) The project challenged the centralised national curriculum at the time, stating: 'curriculum and textbooks had largely failed to adequately reflect either the current or future needs of the country, having been designed in Kathmandu and based extensively on imported models' (Crowley, 1990: 6). It is worth noting in relation to the criticism of 'imported models' that (ironically) the Seti project had been largely conceived and designed originally by an expat educational specialist, Nicholas Bennett. Nonformal educational materials were produced using local paper and with illustrations by local artists to reflect the 'true realities of the area, including local architectural styles, local dress, farming and household implements and jewellery' (ibid: 35). Although the standard centrally-produced school textbooks could not be changed or substituted at that time, supplementary readers were developed to complement them by reflecting local practices and everyday life. As a former project colleague commented, 'We were already drawing on indigenous knowledge and local stories then' [from personal communication] -though the texts had to be written in standard Nepali rather than including local dialect. The focus on producing materials locally (including making blackboards, chalk and teaching aids too) made sense practically tooin an area where it was extremely difficult to transport books, the roadhead being several days walk away and flights being unpredictable and expensive.
The project had several interconnecting strands, intended to be mutually supportive: 'Adult classes encourage parents to send their daughters to the non-formal cheli beti 4 course, which in turn prepares young girls for entry into the formal education system… An improved learning environment benefits from development inputs which vary from the production of supplementary readers for primary schools to the distribution of vegetable seeds and fruit tree saplings… provision of training in all programme areas is an integral component of the project' (ibid: 2). The focus on integrating formal and nonformal programmes, education and development inputs needs to be seen in the context of this remote area at the time. There were no NGOs working in the area and no other externally-funded projects until the 1990 s and few agricultural inputs available commercially either, so the project set out to be selfcontained and self-reliant. Project documents from this time emphasised 'integration' with other sectors and within education as key to sustaining the benefits, mentioning that seven integrated rural development projects were currently underway in Nepal (UNDP 1981: 9).
Three main instruments were identified as key to successful implementation: the administrative system, the teacher as an agent of change and the school as a community institution (ibid: 11). The teacher was positioned as community development worker, with a much wider remit than simply teaching children reading and writing. As the school was to serve the community, the training programme for headmasters included, for instance, how to build a smokeless stove, how to construct a latrine and how to develop a school vegetable garden. The restructuring of the administrative system meant that a high school became a 'resource centre' with several satellite primary schools, adult literacy and girls' literacy classes to supervise and support. The introduction of a 'payment by results' system applied to all project activities, whether school buildings or running a literacy class, and was considered revolutionary at the time, as an evaluation noted: 'traditional administrative styles have been discarded, in much the same way as traditional approaches to community development and education' (ibid: 12).
The most ambitious aim of the project was to facilitate new beliefs and values in communities where relationships and roles were strongly shaped by caste and gender hierarchies. The project proposal noted that previously 'Attitudes and values such as cooperation, a willingness to help those less fortunate than oneself, an involvement in the affairs of the community, a willingness to work with one's hands, etc are not put across in school' (UNDP 1981: 12). The project ran literacy classes in the early morning each day for out-of-school girls (the vast majority) and set out 'to foster increased self-awareness and self-respect among girls and women, along with a heightened realisation of the contribution they can make towards effecting social change' (Crowley, 1990: 32). This objective of education directly influencing attitudes towards gender and caste informed the curriculum too. The cheli beti materials were centred around the life of Kamali, a typical village girl, and only in the last lesson did it emerge that she was lower caste, 'bringing home the fact that caste has little to do with individual worth or capability' (ibid: 33). Like the adult literacy course which aimed at 'the promotion of social change, as on literacy and numeracy', the course disseminated messages about new practices (in health, agriculture) and challenged traditions such as child marriage and caste discrimination. The above statement by the Chief Technical Advisor that caste does not reflect 'capability' contrasted with the perspectives of literacy facilitators who themselves were living in communities where caste determined everything in life. Even persuading their students to build and use a latrine was an uphill struggle, let alone challenging caste and gender discrimination.
The pedagogical approach outlined above reveals an assumption that education could initiate social change through providing information about different ways of behaving and living. In the functional literacy class, people read about building a latrine and then they made one at home for their family. They read about lower castes being equal and then they were supposed to change their behaviour within their community. This approach connects strongly with the 'development' lens which I outlined in Section 2.i. abovethe idea that educational interventions (whether formal or nonformal) could be the main vehicle for changing social attitudes. An analysis of the teaching and learning materials developed reveals a strong belief in development as instilling what Castles referred to as the 'right orientations', whether adoption of improved sanitation or treating lower caste people as equals. However, alongside what might be seen as 'preaching' development messages (self-reference), the project also attempted to promote equality and challenge taboos through the project structures and practicesan approach referred to as 'positive embodied messages' (UNDP 1981: 4). In the cheli beti programme, young girls (often only class 6 level) were trained as literacy teachers in a three-week residential course. Everyone ate together, regardless of caste statuswhereas normally low caste people would have to eat outside and apart from others. The girls were also told that when they were menstruating, they did not need to follow the usual practice of staying outside and could take part in all activities. Joining this unusual communityif only for three weekswas life-changing, and the young cheli beti teachers talked afterwards about how different life could be. Looking back at my theoretical discussion in Section 2, these women could be seen to take 'a critical reflexive stance' (Adkins, 2003:1) on previously taken-for-granted habits.
The above account suggests that the Seti project had predetermined outcomes and adopted a linear model of social change, assuming that educational interventionswhether for adults or children, formal or nonformalwould facilitate social transformation. Taking the 'development' lens outlined in Section 2 above, the project can be seen to promote new values shared by the development community. In the next section, I will take a wider perspective on the projectthrough my fieldnotes and diaries at the timeto explore what happened in practice, and to look at this development intervention through the lens of social transformation.

From aims to implementation: learning perspectives on social change
I am called out of the class to interview two teenage girls. They have come to try for selection as teachers for the cheli beti literacy course. There are very few girls who attend school in this area, so it is a problem to find women educated enough to become teachers for the scheme. These two girls are both studying in sixth grade at the next village, two hours' walk away. They have dressed up in saris for the interview and smeared the red tikka powder (used for worship in temples) like lipstick and blusher around their faces. They are both bright and I feel they will be able to run classes in the morning and still continue with their own education. After a month's training they will be able to run a class in the early morning for primary-age girls, the only time of the day when they will be free from domestic duties. We walk back to their village together to meet their fathers and see if they agree. The walk takes two hours, climbing steeply high above the river and looking down through the pine forests to the spectacular waterfalls below. We cross the bright paddy fields to their village. They want us to stay with them, but we have to refuse in case our selection process is affectedfood is in such short supply here, a glass of milk could seriously influence the results! This extract from my diary gives an idea of the challenges of implementing the Seti project objectives. I was part of a team of teacher trainers walking from village to villageoften for several weeks at a time -recruiting young women who had enough schooling to become a literacy class facilitator. As a project report stated, this was not just about academic performance -'interviews gage their literacy levels and attitudes towards persons of lower caste than themselves' (Crowley, 1990: 40). In Bajura, the neighbouring district to Bajhang, I noted that only six girls had ever completed high school in the whole district (fieldnotes, 5/5/87): I interviewed three girls for literacy teacher jobs today in Martadi. One was only 13 years old and was not able to read the cheli beti lessons herself. The classes usually took place outside, often in a field and with considerable interruptions as I notedthe poor cheli beti teacher had to cope with a strange man standing behind them playing a flute and then a herd of cows walking through the class (18/1/86). The success of the cheli beti programme was assessed in terms of how many girls attended the classes and then went on to join formal schools afterwards. Talking recently to two former cheli beti teachers, they both noted that all their pupils had joined school after this accelerated literacy course and one girl had even gone on to complete a BA. The nonformal curriculum had certainly opened up opportunities for more girls to go to school.
However, looking back at my diaries, I realise that it was not only the provision of a more flexible and relevant nonformal class that changed attitudes about sending girls to school. The project had influenced attitudes towards girls' education through bringing new employment opportunities (including for women as literacy facilitators) and encounters with outsiders such as myself and the trainers from Kathmandu. As one father commented to me, 'if my daughter went to school, she could be like you and earn some money'. As word got round about how much pay teachers would receive for running classes, plus the three-week free residential training, we were faced with queues of girls brought by their fathers for the interviews. Later, when I was visiting cheli beti teachers to give ongoing support at their classes, I commented about a woman who appeared to be forced into this role: It's a difficult situation as I'm pretty sure she is very ill, propped up against the blackboard and unable to teach. But the husband is making her do it for the money. For a low caste family like this, the cheli beti income was considerable. There were few opportunities to earn cash, and many men, particularly from the low caste communities, went to India to work as watchmen, cooks or take labouring jobs. The opportunity for women to earn money running literacy classes in their own villages was welcomed by men as well as women. Although these women could be considered 'empowered' in terms of earning an independent income, many did not have control over how the money was spent. There were exceptions though -Basanti Kumari Singh Shahi, a former cheli beti teacher, recalls in a recent article (Shahi, 2022) how she was able to use her earnings to buy a new sari in the one shop in Chainpur to go to the mela(fair). In another diary entry (18/1/86), I mentioned visiting a family where four out of their eight daughters were cheli beti teachersas the project expanded, more and more young women were getting employment opportunities. This led to a later focus on training women primary school teachersselected from the cheli beti facilitatorsthrough establishing a teacher training centre with a ten-month residential course.
What was striking was how the young cheli beti teachers grew in confidence. This was partly from gaining greater physical mobility in being allowed to attend the residential training course far from their homes, but also through living for three weeks in a self-contained community of young women: We [female trainers] are sleeping in the same high school classroom with the girls, as the headmaster is insistent we have to be with them 24 hours a day in case the boys from the hostel here get near them! As for the actual training, we had a fascinating discussion today about women's education. Half the girls are still at schoolthe other half are school drop-outs. The half who still study were interrogating the others as to why they didn't go to school. I find it incredible that they do manage to go on to high school as many of them walk two hours to get there, crossing deep rivers and then find they are the only girl in their class or even school. It makes you realise how strong the girls are to go through all that just to be educated. They were also describing how they had to fight at home to be allowed to continue (16/9/87).
To use Freire's words, 'conscientisation' was an important part of the training programme, involving much informal and facilitated discussion about women's situations. Through the formal setting in a school, the girls were also introduced to unfamiliar cultural practices, such as using a latrine and eating at tables rather than sitting on the floor. The concerns of parents and teachers about girls mixing with boys before marriage were also a constant presence and threat to girls continuing their education. In the high school in Chainpur which had higher enrolment of girls than elsewhere, I noted that: yesterday all the girls went on strike as some boys wrote 'bad words' about them. There has been a huge detective job finding out who wrote the bad words, with the teachers trying to match the handwriting from 300 exercise books. Then a public beating of the culprit (4/2/86).
When I later visited the cheli beti teachers on supervision toursto monitor their teaching progress as well as note how many latrines etc they had builtthey often asked me to take letters to their friends from the course and wondered if they would ever be able to meet up again. They were very isolated as 'change agents' once back in their home village and missed the company of other educated women on the training course. Many of these girls were still at school themselves so would teach the cheli beti class at 7 am, return home to cook for the family, then leave for school. I reflected in my diary whether for this generation of women, 'empowerment' was problematicthey had a glimpse of what it was to be independent and more equal to men, but this seemed a temporary stage until their parents arranged their marriage, often far from home. However, there were some women who did go on to have paid jobs after marriage -I have since met a former cheli beti teacher running her own shop and another who later trained as an Auxiliary Nurse Midwife and worked in her village health post ever since. Basanti Shahi is now a member of the Karnali Province Public Service Commission and Managing Director of an NGO in Humla.
Looking at the cheli beti programme through the lens of social change, it is useful to distinguish between the impact of the literacy classesreading stories about the low caste girl Kamaliand the ways in which the project was influencing people's ideas about gender equality through providing new employment opportunities for women. Through the residential training, the project was also providing a new space for young women to meet with people beyond their own community. These latter processes of social change can be seen in terms of informal learningand seemed to lie beyond the stated objectives of the project.
Another important aspect of the project was the goal of improving the quality of primary education through teacher training. As with the cheli beti programme, a team of project trainers would trek from village to village giving support and ideas on how to make teaching learning materials, introduce games and discuss how to connect the textbooks to children's experiences. Sometimes this would take place in a school, but was often less formal, as my diary entry suggests: The village headmaster turns up. He has heard of our arrival and asks to see the things we demonstrated at the meeting earlier. We turn the teashop into a workshop, making bamboo pens using the grass-cutting women's scythes and soon everyone is practising writing with them. The headmaster is very excited and begs us to come and see his school up the hill. We get there before dark. He is clearly very dedicated and had convinced the community to work together to cut a flat playground out of the steep hillside. Inside he has made little benches out of stone and mud. He eagerly shows us everything as he has never had visitors before weighing scales made from wood and grass ropes, clay models.
As well as this kind of spontaneous training session, the project ran intensive courses held in a high school resource centre, where headteachers would be introduced to child-friendly teaching methods and practical development skills, such as building a chimney for a stove or making a blackboard. I noted (14/1/86) that one of the project staff tried to get high and low castes to mix at these trainings through modelling different behaviour, serving food to the lower caste teachers first. However, it was much more difficult to discuss social issues and transform attitudes within this older group of teachers. Several of the higher caste headteachers refused to join in the physical work, like digging a pit for a latrine, and it was widely known that they did not allow their own daughters to go to school. The teachers were also constrained by the requirements of the centralised curriculum, particularly the need to ensure children passed the exam at the end of the yearand rote learning was probably the most efficient way to memorise so much material.
As Lekhak (2022: 307) reflects, the Seti Project 'acted as a game changer in the educational sector' and had a strong influence on national policy initiatives. The short teacher training courses developed by the Seti project were later taken up by the Ministry of Education for other districts of Nepal and developed into a diploma in education course offered by Tribhuvan University. The teacher training centre established as a 'laboratory for the development of long-term teacher training programmes' (ibid: 569) continues to play an important role in professional development in this region. The sets of teaching learning materials provided by the projectconsisting of alphabet cubes, weighing scales, maps and gardening toolstended to be kept locked in a cupboard in the staffroom from fear of losing them and there was more interest in making teaching aids from local materials. I noted that the Regional Education Director was very enthusiastic about the approach as the cost of our materials for the whole primary course is only 200 rupees (around £6). We've been really strict about only using bamboo pens and ink, instead of bringing felt pens from Kathmandu, and making multipurpose aids (diary, 26/1/87).
This account gives an insight into some of the issues that the Seti project tried to address directlyincluding caste and gender discrimination, nutrition, health and hygiene, and of course, improving the quality and access to primary education. Through training young women as cheli beti teachers, the project succeeded in providing women with paid employment, greater mobility and status within their communities. However, as my analysis has shown, this process did not necessarily challenge gendered relations and roles within households, meaning most women experienced what Elson (1995) referred to as the 'double burden'. The project set out to initiate social change through the formal curriculum and pedagogy through explicit goals related to development outcomes. However, it was also clear that the project was itself providing new forms of employment and introducing unfamiliar cultural values and practices through outsiders and external funding coming into the area. It is therefore difficult to distinguish between the effects of the improved educational facilities (including the physical infrastructure with the construction of new schools), training and curriculum and the unintended consequences of a project like this stimulating employment in a region that had been largely overlooked by Government and NGOs. I am aware that my diaries and fieldnotes were closely focused on evaluating the educational impact of the project then (i.e. a 'development' lens), so to look at the wider processes of social transformation, I turn now to my recent visit to Bajhang.

So, what has changed today?
In the three decades since the Seti Project, dramatic political, economic and social changes have taken place in Nepal, many of which have had a particular impact in this region. During the 1980s, the Panchayat system meant that political parties were banned, Nepal was a Hindu state and only the official language, Nepali, was used in schools and government institutions (the first language of only 50% of the population). The restoration of multi-party democracy in 1990 was followed by many years of armed conflict between the government's Royal Nepalese Army (RNA) and the People's Liberation Army (PLA) led by the Communist Party of Nepalusually referred to as the Maoists' People's War. Many commentators saw the reasons for the increasing conflict as rooted in the 'continued exploitation of people of specific ethnicity, faith, region and castes', as well as growing unemployment and 'youth apathy' (Panday, 2011: 373). Certainly, in the Far Western region where caste hierarchies were incredibly rigid, the People's War was associated with a time of enormous social upheaval, not least because local young women (some of whom had attended Seti Project literacy classes) were also mobilised into armed conflict. During this period across Nepal, grassroots development workers were in a particularly difficult position, as Adhikari (2008: 321) observed near Pokhara: 'development workers operate under the suspicion of both sides' (ie from Maoists and the Government Army).
The number of NGOs (both international and national) had increased rapidly from the 1990 s in Nepal as a whole (and many were the subject of Maoist attacks). However, it was only relatively recently -after the end of the conflict in 2006 -that international donors gave priority to districts in the Far Western hills. This change may have been related to the strong Maoist activity in Seti zone, which subsequently drew wider external attention to the severe poverty and inequalities in this part of the country. By 2013, the NGO Federation of Nepal had registered 736 member organisations working in the Far Western region (Open Nepal, 2022). Many of these NGOs were working with low caste communities and with womena strong emphasis being campaigns against chhaupadi 5 as a wide-spread practice in this area (Amatya et al., 2018). The limited research literature on this area of Nepal also reflects this focus (see Joshi's, 2022 review).
The transition from a centralised to a federal governance system in 2015 opened up new spaces for people at local level to engage in decision making and formal recognition of indigenous identities, languages and rights. Within the federal system, more budget is allocated directly to local palikas and municipalities, meaning greater access to resources and opportunities for employment. These major policy shifts and political movementsalong with continued increasing migration for work overseas -have facilitated gradual change in social attitudes and family structures, influencing the everyday experiences of everyone in these communities.
I turn now to the more visible signs of change observed during my recent return to the region after more than thirty years. The building of a motor road into Chainpur in 2005 had dramatically transformed the landscape, communications and economic activities. Construction work whether roads, shops, hotels, housesincreasingly provides a welcome source of cash income. I met many women breaking stones on the road today who explained they could come and do this work whenever they had time and they were paid by the sack. A couple had come for a year from their village to build retaining walls alongside the roads. The 1000 rupees a day pay was sufficient for them to rent a room and send their children to the school here in the town and they now returned home only for the annual Desain festival. Young men had found paid employment driving the large diggers hired out for road construction and clearing land. They learned to drive these (and local buses) by sitting in the cab watching the driver and eventually getting a go at the wheel. This can be seen as task-orientated and situated informal learning, which was also social and scaffolded by peers (Rogers, 2013). The shops were piled with vegetables and fruit from the Terai, India and beyond, with no local produce to be found. At the time of my visit in April 2022, the local elections were coming up and outside the high school there were long queues of people waiting to be interviewed for temporary jobs at the polling stations. A former pupil told me he had been successful after doing the running, fitness and writing tests required for selection. He would now be able to earn 42k rupees for the 40 days work (getting people to line up properly and making tea).
It is still a difficult journey into Chainpur, a minimum of 12 hours very bumpy bus ride to the Terai or a flight which is unpredictable and may be delayed for days rather than hours. However, many people are now coming into the area to do businesswhether selling goods, managing hotels or teaching in the various private schools in the bazaar. The yarsagumba -caterpillar fungus with medicinal properties collected in the mountains -trade with China has flourished since the 1990s, bringing huge amounts of wealth into Chainpur. While men continue to migrate to India for low-paid manual work, people are also going further afield to the Gulf countries. The new sources of income from employment outside subsistence agriculture were reflected in people's dress -I did not see anyone without shoes or in ragged clothes (which was common in the 1980s). Mobile phones have transformed not only business, but marriage practices too. A local NGO worker said that whereas in the past girls were being married off early by their parents, now they are arranging their own marriages through Facebookhe had met a 15-year-old girl recently with a baby who had married at 13. He thought more people had 'Facebook marriages' in the villages than here in the bazaar. Even the poorest households had a mobile phone -'they would buy it even if it meant going without food'. A woman in her twenties who worked in a hardware shop explained that local men were 'no good' and she was keen to find a husband from another area of Nepal on Facebook. Talking to older women who had come from India years ago, travelling long distances for marriage is nothing newbut the difference now appeared to be young people's agency to bypass their parents, and to marry across castes in some cases.
The move to federalism with the three tiers of government (local, provincial and federal) established by the 2015 Constitution has led to more political opportunities at the local level, including for women. There is now a requirement of at least 33% representation of women candidates in elections. I met a former cheli beti teacher hoping to stand as vice mayor in the forthcoming election, and a woman walking to the election office to see if she could stand as ward representative. Another former cheli beti teacher now serves on the Karnali Province Public Service Commission. Alongside these political developments, there is an increasing diversity of NGOs and political organisations in Bajhang. At an NGO set up to empower Dalits (lower caste communities), I found that of the 23 staff, all local and educated to at least ISc 6 level, at least half were women. An older woman, who participates in the NGO's activities, talked about dalit rights and said that the situation has improved, though there is still discrimination between groups within the dalit community. She said she had never studied -though I noticed she was able to read texts on her phone as we talkedand she has been to many training courses organised by the NGO, including vegetable growing and confidence building. I also visited a vibrant savings and credit union for women, run by women. With 1800 members, around 300 women a day come to deposit money or to request a loanunlike the formal bank, no collateral is needed. Through these institutions, many women appeared to have gained more agency and voiceincluding low caste and unschooled women. However, although there was far more representation of women in formal institutions, the gendered roles at household level seemed similar to thirty years ago. Women were still carrying huge doko of leaves to feed the cattle or busy with cooking and looking after children, but the difference was that they were likely to be listening to music or chatting on their mobile phone at the same time.
Turning to the education sector, the most visible change (and that reported first by all the teachers I spoke to) is that girls now outnumber boys in all the local schools and the campus in Chainpur. A former headmaster remembered how parents used to be so reluctant to send their daughters to his village school but said 'now there is no one left at home'. Today there are 1000 students in grades 1 to 12 in his school, and more girls than boys. He thought the reason was demography as well as boys being more likely to be sent to private schools on the Terai or Kathmandu. The campus chief recalled how when he himself was a student and was taking the School Leaving examination, there was only one girl in his class, but now at BA level, the female to male ratio is 70 to 30. Previously, girls' attendance had been irregular due to the strong cultural practices around having to stay apart from others during menstruation. Sitting in a school staffroom, I noticed a flip chart on 'menstruation management' and the teachers explained that schools now distribute sanitary pads free. About three years ago, they had stopped the practice of girls staying apart in this school, but the teachers thought it was unlikely such changes had taken place at home. Similarly, 5 Joshi (2002) explains this as 'a cultural practice where women are considered untouchable and forced to stay in separate huts during menstruation and child-birth', 6 Intermediate of Science, equivalent to A level discriminatory practices towards low caste students had been banned in school but still persisted outside, as a male teacher explained: 'they sit together and eat snacks together in school. But it is deeply rooted and they haven't changed inside'. Despite girls' enrolment increasing dramatically, the number of women teachers had hardly changed and out of 40 teachers in a local high school only seven are women. Reasons were given as teaching being 'a competitive job' and that though there is now a quota for women, 'they still have to compete'. Teachers also said that although girls were studying now, the main motivation was that it was easier to marry after grade 12. Shanta, a woman teacher in a primary school half an hour from Chainpur, told me she was one of only five girls who passed the School Leaving Certificate in her home district. She had gone on to do a 10-month teacher training course locally and by radio, then got married and moved to Bajhang. She said she was unusual in continuing to teach after marriage. Other teachers also suggested that girls were not interested in getting a jobfor instance, though there are many girls studying agriculture, 'they tend to use this knowledge at home only'. The campus chief had faced a lot of problems with recruitment of teachersfor male and femaleand had to teach the management course himself. He said so few teachers were reluctant to come and work in a remote area, as there are no special incentives and poor facilities.
Talking to teachers and education officials in the area about the challenges they now saw in the education sector, they emphasised that access has now been addressed but that 'quality' was the issue. The schools had just reopened after two years' complete closure due to the Covid pandemic (with no online or alternative provision) and several teachers commented that students had been promoted to the next grade without achieving the required standard. An education official said that 'the students are weak, the curriculum is OK, but discipline is weak. They need to learn not to challenge the teacher and to do what the teachers says'. He went on to explain that there were also problems with teacher motivation and quality too: '98.2% of primary teachers have now had training but the issue is about implementation and reflection'. A teacher in the local high school felt that the cause of declining quality was 'the liberal system. Because education is free, students don't try. If there were only a few seats, they would have the habit of working. The problem is the system. Now everyone is promoted, everyone passes.' In my brief visits to six schools in the area, I was struck by the diversity in provisioninterestingly, one of the Seti project schools which was considered a model of good practice in the 1980s (three hour walk from Chainpur) still has the most impressive facilities, with a computer lab (solar powered as no electricity), sophisticated science and gardening equipment (including a tractor) and a dedicated special needs unit which also provides residential facilities. By contrast, other schools were often extremely cramped with the increasing numbers of students and had few facilities apart from benches, desks and blackboard. Due to a recent government policy to reduce teacher absenteeism, all the schools have an electronic registration system where teachers had to touch a thumb pad on arrival.
This glimpse into life in Bajhang today indicates the enormous changes that have taken placeboth within the education sector and outside. Taking a 'development' lens on the changes, many of the Seti project goals have been achieved, including: to increase girls' and lower caste communities' access to education, provide teacher training, break down caste and gender inequalities within schools and improve the educational infrastructure and accountability. However, taking a broader social transformation lens, it was clear that schools had only played a small role in initiating changes in attitudes, livelihoods and the economy. The private sectorparticularly in transport and construction has created new opportunities, not only for paid work but also for informal 'on-the-job training' with friends. By contrast, formal education still seemed to be seen as a way out of the areathrough university education in Kathmandu or on the Terai. The Seti project's idea of schools as a centre for community mobilisation and teachers as 'change agents' seems to have been left behind once new institutions moved into the areasuch as the NGO working with low caste communities, saving and credit unions and above all, political parties and local activists. These organisations outside education were active in stimulating informal learning through debates around identity, voice, equality and rightsevident in the confidence which so many 'uneducated' women demonstrated. Alongside these institutions, mobile phones have had a huge influence on access to knowledge, communication and informal learningtransforming and challenging gendered relationships and individual agency within households.

From education for rural development to education for sustainable development
The term 'rural development' meant improved sanitation, health knowledge and girls' schooling for developers like myself in the 1980s. But, as this account has shown, rural development had a much wider meaning for this community in the Far West of Nepalleading to roads, internet connectivity and a vastly expanded local economy. The current international policy goal of sustainable development introduces yet another perspective on the changes that have taken place, raising questions around the environmental impact of the rapid and unregulated growth of building and roads. Already, the Seti riverbanks are overflowing with plastic waste.
In this paper, I have argued that the need to consider education within broader social, economic and political transformation, rather an 'education first' approach. Though an account of the enthusiasm with which people have taken up digital technology, new employment opportunities and entered political spaces, I have pointed to the importance of informal learningwhether learning how to drive a digger with a friend, or how to make relationships online. What emerged was a huge gap between the kind of learning promoted in schools and formal educational institutions, and processes of informal learning and literacies in everyday life. Teachers talked about how 'online learning has no meaning here' in relation to the closure of schools during the pandemic, yet all through this period children and adults were constantly engaging with new knowledge on their phones.
Many of the infrastructure projects that are taking place in Bajhang are due to the decentralisation of budgets and decision making to local governments through the recent federalisation restructuring. As part of this move, the government has recognised the limitations of the centralised school curriculum and now made provision for local governments (of which there are 12 in the Far West) to develop their own local curriculum. The first orientation on developing a local curriculum for grades 1 to 3 took place recently for district education officers. This could be an exciting opportunity to reflect on the ways in which primary schooling can better respond to the rapid changes taking place in remote districts like Bajhang. Rather than simply inserting local pictures and stories into textbooksas we did in the Seti project supplementary readerslocal curriculum developers could start from young peoples' current aspirations and experiences as the basis for exploring what kind of education could enable them to make the most of current informal learning and livelihood opportunities in the area. Whilst 'sustainable development' can be seen as an imposed objective from an international policy agenda, students could be introduced to a historical perspective on development in their area in order to discuss and reflect on what kind of future they would like for their children. As I found when sharing photos of the 'old' Chainpur with young and older people living in the community, intergenerational interaction can be a rich source of learning. The proposed localised curriculum has the potential to introduce intergenerational learning, alongside ensuring that the more marginalised communities also have an opportunity to take advantage of the growing opportunities in this area.

Conclusion
I began this paper with the aim of exploring a different way of understanding education and social change. I argue that the dominant discourse equates education with schooling, and social change with planned development interventions. The related assumption of a linear relationship between development and education emphasises inputs and intended outputs, thereby limiting the potential for analysing wider processes of social change. I interrogated the historical account of the Seti project through adopting both a 'development' and a 'social transformation' lens on change. Firstly, the documentary analysis offers insights into an educational project framed within the ideology of education for rural development (the 'development' lens), demonstrating an 'education first' approach which is still common in many areas of the world. Secondly, my first-hand experiences at the time and retrospective reflections from colleagues complement the formal evaluation reports of the Seti Project (which assessed outputs in relation to inputs). This qualitative data lends itself to a 'social transformation' lens, through giving insights into some of the spontaneous learning and cultural changes taking place at that time.
Taking this perspective retrospectively on the Seti project drew attention to how limited the 'development' lens on education had been, particularly in terms of understanding changes in gender relations. As my account reveals, the cheli beti programme not only provided young girls with access to primary education, and trained teachers as facilitators, but also influenced attitudes around the kind of paid work that women could do in this remote area. I suggest that this process was not always empowering for the women themselvesfor instance, if a father insisted that they took up this role in order to support the household. Alongside the formal training, the informal, often spontaneous, learning that the cheli beti facilitators engaged in was often life-changingsharing experiences with girls from different caste backgrounds and making friendships beyond their own village. These findings raise questions about how far the influence of the Seti project on gender roles and relations was related to the educational componentor would any externally-funded intensive project providing new employment opportunities for women similarly change attitudes, values and practices?
Finally, by taking a 'social transformation' lens on ethnographicstyle data in present-day Bajhang, I set out to investigate what we might learn about education and social change through putting everyday life and informal learning -rather than educational institutions and programmes -at the centre of a study. This approach involves considering a much broader range of actors as 'educational', in that they are introducing ideas for change informally. As my account revealed, such actors might be political activists, traders or truck driversnot only teachers and NGO workers. I have also argued the need to consider the context for social change in a more dynamic sense, rather than focusing mainly on the institutions or spaces where formal learning is provided. This implies taking a broader perspective on what is already going on in a community, to analyse how learning is emerging from processes of social change, whether through new destinations for migration or increasing digital communication. Unfortunately for policy makers, this approach may raise more questions than answers about the relationship between education and development. Adopting a social transformation lens implies an acknowledgement that change is unpredictable, gradual and impossible to attribute to literacy or education alone.

Author statement
I confirm that I have no competing interests or conflict of interests.