This practice-based PhD is focused on མཛོད་དགེ (Zoige), on the eastern edge of the Tibetan Plateau in southwest China, where pastoral environmental knowledge has been formed and renewed through lived relations with land. Grazing, animal care, seasonal movement, place names, stories, rituals, and household labour are among the practices through which this knowledge is learned and carried across generations. Architecture is understood here not simply as built form, but as the spatial, material and relational arrangement of situations in which pastoral environmental knowledge and relations of care are practised.

The thesis first traces socio-environmental transformations in མཛོད་དགེ (Zoige) in relation to changing state-supported scientific and administrative modes of representation. Across periods, the land was represented as ‘marshland’, ‘wetland’, a ‘biodiversity hotspot’, and a ‘carbon pool’ through surveys, classifications, maps, remote-sensing images, and policy designations. These representations did not transform the land by themselves. They mattered when mobilised through scientific and administrative institutions, policy, and programmes. Through these interventions, patterns of land use, mobility, household life, and environmental conditions were altered, changing how pastoral environmental knowledge could be practised and transmitted.

Against this background, the thesis asks how practice-based research in architecture can support the intergenerational transmission of environmental knowledge through community-based environmental practice developed with pastoral community members. Drawing on an initial engagement in 2021 and sustained fieldwork conducted between 2023 and 2025, it follows how my research expectations, methods, and role were revised through relations with community members. Repeated stays, shared labour, hospitality, and everyday conversation changed the questions and practical possibilities of the research. Mapping, photographs, stories, and other forms of representation remained, but the questions they addressed, their purposes, and the ways they were arranged changed. Community members’ knowledge, concerns, suggestions, and refusals increasingly redirected the research. Through these field relations, the research became increasingly collective while representing with developed as a situated methodological orientation.

Four collective environmental practices took form: collective environmental memory sessions, a letter addressed to the community, collective seed collection, and collective seeding. Collective mapping, photo-tables, stories, letters, and recordings supported recollection, discussion, correction, and circulation, while household work, ritual, seed collection, and seeding carried knowledge through bodily, material, and land-based practice. I argue that these practices created situated conditions in which environmental knowledge could be recalled, questioned, practised, and further developed across generations.

The thesis makes three key contributions to the field of architecture and its associated methodologies. First, it identifies how socio-environmental transformations in མཛོད་དགེ (Zoige) have changed some of the conditions through which pastoral environmental knowledge is practised and transmitted, while tracing how state-supported modes of representation were mobilised through institutions, policy, and programmes. Second, it develops representing with as a situated methodology and clarifies its relation to collective research. Third, it contributes to practice-based research in architecture by approaching architectural practice through the making and sustaining of conditions that support the intergenerational transmission of environmental knowledge.