The notion of adaptive evolution of social systems as of a real process of selection of the properties of such systems implies group selection. But strong evidences of effective group selection seem impossible, at least in vertebrates. However, understanding the origin of social systems adaptivity based on individual selection is difficult, as well, without analyzing the proximal mechanisms of the formation of such systems. I suppose that social systems change due to changes of individual features that underlie the proximal mechanisms of the system formation. These features are the characteristics of neurophysiological and hormonal regulatory mechanisms. They are strongly associated with intrinsic biochemical processes and are coded in the genome. Thus, the evolution of social systems is the evolution of their proximal mechanisms. At the same time, the specificity of neurophysiological and hormonal regulation determines not only social interactions, but also the individual behaviour of animals. The most important characteristics of life history, such as the regime of activity, foraging strategy, etc., are strongly affected by the same regulatory mechanisms. This view is useful for understanding the relations combining many features into an integrated and adaptive species-specific life form. I suppose that such forms emerged as evolutionary consequences of changes in regulatory mechanisms adaptive to specific environment. Thus, we have as substantial reasons to discuss adaptations of social systems to ecological features as to discuss ecological features adapted to particular social systems. The species-specificity of regulatory mechanisms is probably based on different kinds of evolutionary choice between the rapidity and the perfection of adaptation, between flexibility and stability, and between sensibility and resistibility. I think that this choice depends largely on the predictability of the environment. The less predictable it is, the more it increases the selective value of sensibility, flexibility, and rapidity of evolution. On the contrary, stable and predictable environment stimulates less rapid but more perfect adaptations. Such choices consolidate in the genome during evolution as specific features of neurophysiological and hormonal regulation systems. These specific features, in their turn, determine ecological, behavioural, and physiological species-specificity. From this point of view, evolutionary changes in social systems can be readily perceived as consequences of the selection of individuals, promoting optimal properties under particular conditional features of regulation systems. The boundary condition for this model is the absence of specificity of the characteristics of regulation systems to different forms of stress. This condition needs to be considered closely.