When we inquire about the mechanism of a biological effect, we have implicit reference to a picture of how nature is organized and how it should be approached. In one view, the biological system is seen as more than a sum of its parts and it is held that one cannot understand an organism's essential characteristic-life-by studying subsystems below a certain structural level because life does not exist below that level. This idea was precisely stated by Paul Weiss (1):
If a is indispensable for both b and c; b for both a and c; and c for both a and b; no pair of them could exist without the third member of the group, hence any attempt to build up such a system by consecutive additions would break down right at the first step. In other words, a system of this kind can exist only as an entity or not at all.
Thus, for example, even complete knowledge of the properties of a protein solution would not tell us how the protein functioned in vivo; we would not even know whether its in vitro properties had any relevance at all. Under this approach, the proper starting point to study nature is the whole organism in its normal environment. It is recognized that, considering the organism's physiological control processes, not all biological phenomena can be localized to specific tissues in the organism. In contrast to this cybernetic approach is the idea that, ultimately, living things will be describable solely in terms of the physical laws governing inanimate things. Methodologically, this analytical approach consists of the study of increasingly more complex models of the organism's parts, with the goal of explaining the organism's characteristics and behavior in terms of the characteristics and behavior of the models. The amount of whole-animal data presently available is much greater than that involving model systems and, for this reason, the cybernetic approach gives a more general and more useful picture of bioelectrical phenomena. This approach is described below; work that can be considered to have arisen from an analytical approach is described in the following section.