First published in The Australian Journal of Politics and History, Vol. XVI, No. 3, December 1970
MOST STUDIES of political factionalism begin with attempts to define the terms 'faction' and 'factionalism'. The resultant definitions are then regarded as having universal validity, irrespective of time, place, or political culture. The use of this procedure, in our opinion, leads either to the unwarranted universalization of causal relationships having limited scope, or to generalizations which are too broad to be of much practical value.
It is true, of course, that definitions are normally generalizations based on a variety of relevant observed phenomena, and purport to capture what is essential to the concept being defined. This involves no great difficulty where a given usage of a term is generally accepted and where those who use it are consistent in their terminology. Where, however, there is wide divergence between the uses commonly made of a term, where, in other words, it is relatively speaking 'undefined', there is considerable danger in beginning from a stated definition. In such a case it is the definition itself that has to be justified and should rather be an end-product of the investigation. Even so, the definition may still be quite justifiably challenged by somebody who simply prefers to use the term in a different way. This is not, of course, to say that argument on the subject is therefore pointless, since there may be substantive points to be made in support of either side, or indeed of both. Much of the argument will be concerned with the usefulness of the respective definitions. The question of whether they 'fit the facts' may well be a secondary one, since there is disagreement about which facts the term itself is supposed to fit.
The term 'faction' lies in this limbo of serious and fundamental semantic disagreement. In part this may be because it is only relatively recently that it has come to the attention of scholars as a potentially useful tool for social and political analysis. 1 It may be argued that the concept has an often unrecognized potential for the unravelling of certain problems of political analysis, and that it can fruitfully be used to illuminate a complex and coherent set of political phenomena which are of considerable practical importance especially in certain types of political systems.