- ISSN: 17427150
- DOI: 10.1177/1742715009337770
In this article I argue for a revised unit of analysis in leadership. I refer to this unit as a configuration. The need for this revision arises out of a reconsideration of `distributed' as a valid and accurate means of representing leadership. While aggregated, rather than holistic, understandings of distributed leadership have assumed prominence in educational circles and the public sector, aggregation is an indiscriminate approach to demonopolizing the idea of solo leadership and decentring `the' leader. By treating pluralities of leaders as numerically equivalent or all-of-a-piece, for example, an aggregated understanding makes little allowance for different levels of leadership and for qualitative differences among leading units. In a number of empirical accounts of distributed leadership, however, individual leaders still figure prominently as agents of influence, although they frequently do so in company with a variety of emergent `small number' formations. For this reason, the totality of such arrangements represents a time-, space-, context- and membership-bound configuration of influence-based relationships, the dynamics of which, due to the mixed patterning of the formations, are most accurately characterized as `hybrid'. Some suggestions are made concerning the significance of this proposed unit revision along with their implications for research into leadership practice.