Educating our early career researchers is becoming more complex. The range of doctoral degrees, the fast moving nature of knowledge, internationalisation, the demands of funding bodies and employers are all pulling on PhD supervisors. The untrained PhD supervisor will copy (or avoid copying) the way that they were supervised themselves. Current literature on PhD supervision focuses either on lists of tasks that the supervisor must undertake or on conceptions of research. There is a need for a conceptual approach to research supervision. This review of the literature relating to research supervision identifies six main concepts all of which contribute to our understanding. These are not competing concepts. Supervisors, supervisory teams and co-supervisors might use them to define or illuminate their practice.

It is proposed that the range and depth of concepts that a supervisor holds will dictate how they supervise and the type of researcher who emerges at the end of the process. In an age of supercomplexity, when demands of academic and other employers are unpredictable, the skills of the effective researcher, and thus their supervisor, are likely to become even more important.