This dissertation addresses a specific aspect of the broad area of communication systems used among researchers. This research has undertaken to establish a broader view of the communication practices of scholars to understand the motivations behind their publication choices. Open access offers a solution to issues with the scholarly publication system such as delays in publication and restricted visibility of research due to high subscription costs. The principle of open access is to enable maximum access to findings from publicly funded research to maximise social returns on public investments. Despite the apparent benefits of open access, the uptake has been limited. ¶ This thesis research takes a holistic view of the researcher as a communicator to uncover the reasons why researchers are making the publishing decisions they are. In-depth interviews were conducted with 43 researchers in three disciplines at two institutions, the Australian National University and the University of New South Wales. The disciplines, Chemistry, Sociology and Computer Science, were known to have different publication practices, The questions asked about all aspects of researcher communication including researching, authoring, informal communication, article submission, refereeing, mentoring and data storage. ¶ The findings show that traditional arguments for open access are ineffective. The Reward function of scholarly publishing is central to managing academic careers and supports traditional publishing systems. While having work openly accessible increases an academic’s exposure and possibly therefore their citation counts, unless alternative internet-based forms of metrics are adopted, the open access option will not directly appeal to researchers. ¶ Information-seeking behaviour demonstrates how disciplinary differences affect researcher’s interaction with technology. The disciplines showed marked differences in almost all the areas explored, and the behavioural norms expressed in each discipline have direct bearing on the likelihood of members of that discipline embracing open access. The ‘institutional/disciplinary divide’ means that researchers must publish in ways that run counter to their disciplinary norms in order to satisfy institutional and grant funding requirements. ¶ Until governments, and particularly university administrations, recognise the need to consider the discipline and the need to consider the individual and respond to these needs, and until there is a realisation that different disciplines may require radically different approaches, there will not be a large-scale adoption by individual researchers of the current open access tools. Either institutional repositories need to adapt dramatically to offer work practice benefits or the broader academic population will only use institutional repositories under duress, which is not the situation envisaged by open access advocates. The alternative is for communities to develop their own subject-based repositories, a development that again is likely to be highly dependent on communication norms in different disciplines.