A careful analysis of the basic definitions of ``system'' as it is used in systems research reveals a fundamental distinction between two kinds of variety. What I will ``cardinal variety'' can be regarded as a measure of the number of possible states which a system can achieve; and what I will call ``dimensional variety'' can be regarded as a measure of the number of different aspects, components, or parts from which the overall system is composed. A proper system must have both forms of variety: it must have a variety of parts, and each part must have a variety of possible states.
In this paper I will analyze the nature and presence of both cardinal and dimensional variety in the context of both the traditional, structuralist and the constructivist, functionalist views of system. The recognition of these two forms of variety allows a synthetic understanding of system, bridging the gap between the structuralist and functionalist views. I will then discuss the relation between these two kinds of variety, the different special cases obtained by considering the presence or absence of either one or the other, and the concepts of ``distinction'' and ``negation'' in the context of these two forms of variety.