Behaviour sequences commonly consist of highly variable appetitive phases leading to rather fixed consummatory acts. Action-pattern rigidity is typical of the terminal moments of a reaction chain. This basic fact is all too often obscured by the artificial conditions of behaviour studies. Observations on laboratory or captive animals tend to conceal the degree of variability of the earlier phases of each sequence. The simplicity and sterility of the unnatural environment offered to the animal causes differential damage to its motoric performance, attacking the early stages more and the later stages less. A caged animal will feed, drink, nest and copulate, but it cannot set off on lengthy quests for food, water, nest material or a mate. Notorious laboratory devices such as the Skinner-box have served to eliminate totally any possibility for motoric variability. The emphasis in laboratory studies of this kind has been steadfastly concentrated on the variability in the relationship between simple stimuli and an artificially rigidified response. Although the study of this (SR) relationship is an important aspect of animal psychology, it is extremely misleading to overstress its importance as has been done so often in the past. To equate it with the whole topic of animal behaviour is like claiming that the gaming rooms of Las Vegas reflect the whole of human endeavour.
Appendix 1. Diagnostic Criteria for Psychosomatic Research