This thesis examines the multiple and shifting boundaries between alternative and conventional medicine, demonstrating how 'boundary work' is a fundamental process in the (re)shaping of networks of human and nonhuman actors. Social constructionism and actor-network theory are used to explore some of the strategies used to maintain and contest the boundaries between 'alternative' and 'conventional' cancer treatments.

Media reporting of the Liam Williams-Holloway case is used to illustrate the way facets of the print media construct oppositional representations of ‘alternative’ and 'conventional' medicine. It is argued that these accounts tend to naturalise differences between 'alternative' and 'conventional' medicine and consolidate oppositions such as scientific/unscientific, symptomatic/holistic, invasive/natural. Analysis of interviews with selected 'alternative' and 'conventional' health practitioners in Christchurch indicates how these 'natural' distinctions are disrupted in the rhetorics of health practitioners who are involved in strategically consolidating and contesting these oppositions constructed by the media. It is argued that their 'talk' illustrates the complexity, connectivity and contradictory features of their professions. This is evident in the use of the discursive strategies of the 'other' to legitimise their practices.

The thesis focuses on illustrating that distinctions between 'alternative' and 'conventional' medicine are the outcome of particular representations and are variable with respect to time and location. Moreover, 'alternative' and 'conventional' medicine are continually produced by both human (e.g. practitioners, patients, journalists, politicians, lobbyists) and non-human actors (e.g. imaging technologies, diseases, medications). It is argued that institutional mechanisms such as medical journals, ethics committees, pharmaceutical companies, and medical councils actively regulate the ongoing consolidation of the line between what treatments are considered valid/invalid, effective/ineffective, scientific/non-scientific, and ensure the dominance of 'scientific' method and evidence in the medical encounter. This thesis argues that classification and representation of health care practices is a result of past and present human activity, a result of boundary work. It illustrates how 'alternative' and 'conventional' in medicine are a product of both media representations and the discursive practices of practitioners.