In and out of actornetwork theory a necessary but insufficient journey

Purpose This paper seeks to offer a retrospective look at an intellectual journey in and out of using actornetwork theory, which the author drew on to carry out an indepth case study of the troubled implementation of a computerised reservation system in a major transport company. The application of some key ANT concepts, i.e. human and nonhuman actors, symmetry and translation, is reflected upon, highlighting their benefits and limitations. Designmethodologyapproach The paper's aims are accomplished through a confessional account of how it was done, rather than a normative post hoc justification. Some empirical evidence is provided to illustrate the difficulties and problems encountered in travelling back and forth between theory, methodology and data. Findings In particular, ANT was very useful in focusing the paper on how to look at IS success and failure symmetrically and how social and technical distinctions are socially constructed, for instance in the conception and application of yield management software. Formulating a series of translations to encompass a large number of actors is shown to have provided some explanatory capacity. But a limitation is how to relate local and global actors, which is also a matter of power relations and politics. Originalityvalue The paper explains why, as a late and unplanned reaction to this, but also in contradiction with ANT principles, it ended up complementing ANT with Clegg's theory of power to bridge that gap. The paper concludes with a discussion of where the difficulties lie in using ANT and how it can be misused in IS research the author argues that this is due to a lack of exposure to poststructuralism in IS research, compared with other managementrelated disciplines and that recent efforts by scholars in the science and technology studies field to combine constructivist approaches such as ANT with critical social analysis should be considered.