Introduction

Would you choose a pneumatic drill to clean a window? Probably not, but the important question is: why not? You would not choose a pneumatic drill to clean a window because you would almost certainly, if implicitly, consider the nature of glass and conclude that a pneumatic drill would be an inappropriate tool to be applied to this material. In other words you would start with an enquiry into the nature of the phenomena under investigation. That is, you would start with ontology. You would then choose a tool appropriate for dealing with the phenomena in question; to clean a window you might choose a damp cloth. This simple example suggests that we should start with ontology and then choose or design a ‘toolkit’ appropriate to the phenomena under investigation. The term ‘toolkit’ is, of course, metaphoric, and whilst it can refer to the physical world, we are more concerned with the social world, and so our ‘toolkit’ will be conceptual, in this context, meta-theoretical.

Empirical researchers working on the HRM–P link do not, however, start with ontology. Instead, they start with an ‘off the shelf toolkit’ in the guise of scientism, the deductive method, and its quantitative, empirical and statistical research techniques and apply this ‘toolkit’, without questioning its appropriateness. Purcell and Kinnie, whilst perhaps not being as thoroughgoing in their critique as us, nonetheless notice something similar: