Abstract

Much recent work in sociobiology can be understood as designed to demonstrate the sufficiency of selection operating at lower levels of organization by the development of models at the level of the gene or the individual. Higher level units are accordingly viewed as artifacts of selection operating at lower levels. The adequacy of this latter form of argument is dependent upon issues of the complexity of the systems under consideration. A taxonomy is proposed elaborating a series of types, or grades, of hierarchically organized systems. These range from aggregative systems, in which there is no organization relevant to systemic properties, through several graded variations reflecting various degrees of functional interdependence of components, to integrated systems, which manifest component specialization and diversification as well as a subordination of component function to systemic function. It is suggested that the most complex form of organization is plausibly treated as indicative of higher level units of selection.