Abstract
Investigation of the causes of war requires analysis of the characteristics and behavior of only those dyads of countries that are potential belligerents. Several scholars have offered rules for delineating such “relevant dyads”. One common element of such rules is contiguity. A second common element is major power status. Any dyad involving either contiguous states or a major power is defined as relevant. Such definitions of relevant dyads are simple and useful. Nevertheless, I contend that some contiguous dyads are not relevant to study of the causes of war, while some non‐contiguous dyads are relevant. For example, Israel and Iraq are neither contiguous, nor major powers. With existing definitions this dyad is not deemed relevant. I offer an operational definition of relevant dyads that delineates which dyads are proximate enough in terms of distance and terrain to be potential war fighters, regardless of major or minor power status. Adapting existing work on the loss‐of‐strength gradient, I argue that each member of the international system has an area of the globe within which it can act militarily. This area is the relevant neighborhood of that country. Relevant‐dyads are found where relevant neighborhoods overlap.