Accounting for International War: The State of the Discipline

Abstract

In studies of war it is important to observe that the processes leading to so frequent an event as conflict are not necessarily those that lead to so infrequent an event as war. Also, many models fail to recognize that a phenomenon irregularly distributed in time and space, such as war, cannot be explained on the basis of relatively invariant phenomena. Much research on periodicity in the occurrence of war has yielded little result, suggesting that the direction should now be to focus on such variables as diffusion and contagion. Structural variables, such as bipolarity, show contradictory results with some clear inter-century differences. Bipolarity, some results suggest, might have different effects on different social entities. A considerable number of studies analysing dyadic variables show a clear connection between equal capabilities among contending nations and escalation of conflict into war. Finally, research into national attributes often points to strength and geographical location as important variables. In general, the article concludes, there is room for modest optimism, as research into the question of war is no longer moving in non-cumulative circles. Systematic research is producing results and there is even a discernible tendency of convergence, in spite of a great diversity in theoretical orientations.