Degree Name

Doctor of Philosophy


Faculty of Education


Agriculture is the mainstay of Tonga’s economy. It provides approximately 80% of employment opportunities and about 32% of Gross Domestic Product (GDP). The fragile environment and limited resources of Tonga such as small landmass, isolation, proneness to natural hazards, long distances and expensive transportation costs from overseas markets of Australasia, United States, Europe and Asia, create fiscal deficits, high inflation and vulnerable economic growth (Briguglio, 1995; Government of Australia, 1992).

Before 1960, subsistence farming i.e. growing crops mainly for home consumption and festivities was the mode of production in Tonga. During the early 1970’s, local farmers started to apply agrochemicals and fertilisers as the substitution for traditional farming practices such as shifting cultivation, moon planting guides, seed saving and composting when trading with developed countries. Currently, many Tongan people rely on agriculture as their main source of livelihood but the high use of synthetic fertilisers and pesticides to gain better yields is a short-term fix that has the potential to do serious damage to the health of the people of Tonga and their fragile environment.

The theoretical framework for the research was participatory action research (Lewin, 1946). Participatory Action Research (PAR) is an alternative philosophy of social research and often associated with social transformation in the Third World. It is based on “liberation theology” and neo-Marxist approaches to community development in Latin America (Freire, 1968). Three particular attributes are often used to distinguish PAR from conventional research. These are shared ownership of research projects, community-based analysis of social problems and an orientation toward community action. This study was designed to appraise the usefulness of an appropriate educational theory to guide and support a research project that had as its attributes shared ownership, community based analysis of the problems and an orientation toward community action. The methodology framework for practical implementation of the project was participatory rural appraisal (PRA), a framework commonly used in agriculture development programs in less developed countries.

The study had three main purposes: 1. To assess the strengths and limitations of Participatory Rural Appraisal (PRA) as a framework for organising community education practices designed to improve ecologically sustainable methods of crop pest control in Tonga. 2. To identify an educational theory that would be suitable to support the PRA framework and that also addresses the reported limitations of PRA. 3. To use the perspective of a participatory action research framework (PAR) to evaluate the implementation of this theoretical educational model in the context of a two Tongan community groups who were the focus of the data collection. These were a group of women from a town and a group of young farmers from another village.

The literature review suggested that Cambourne’s (1988) conditions of learning would be a suitable educational theory to support the PRA process and to address the limitations with PRA. The study then examined how Cambourne’s (1988) conditions of learning could be activated to support adult learning in the context of Tonga – in particular the following attributes of participatory action research: shared ownership of research projects, community-based analysis of social problems and an orientation toward community action.

A qualitative research methodology was adopted to help understand the richness of the cultural context of the participants (groups of women and young men in villages in Tonga) and to understand how the various conditions of learning, as recommended by the educational theory, were activated during the implementation of the education program. The focus was on the education activities as recorded by the researcher and extensive use was made of field notes, interviews and photographs, as well as other artefacts. These data were used to maintain a consistent record of the changes and interventions that occurred during the learning activities.

The findings showed that the Cambourne’s model could be adapted to effectively guide an adult education program in the context of Tonga and concluded that such a model has the potential to be used for similar education programs in other rural education contexts.