Introduction

Abstract

A causal explanation of animal behavior by definition would include the formulation of a set of rules enabling us to predict future behavior from knowledge concerning the past. The student of behavior will in fact search for the simplest set of rules which will enable him to predict those aspects of behavior he is interested in. It has been found repeatedly that the most simple type of explanation which treats an animal as a stimulus-response machine, will not do. Thus an animal, when confronted with a standard kind of food in a standardized environment, will sometimes eat, and sometimes it will not eat and moreover when it does eat, the amount of food ingested per unit of time may vary. It may then be attempted to describe the animal as a system which may be in one of various states. The past history is supposed to be reflected in the state the animal is in, and the behavior which an animal is going to perform is supposed to result from present state and present “input”, e.g. the stimuli which the animal at present receives from its environment, (cf. Metz, H. 1974; de Ruiter and Wiepkema, 1969; de Ruiter, Wiepkema, and Veening, 1974; McFarland, 1971). Thus whether or not an animal will eat when confronted with food is said to depend on “how hungry” it is. The amount of hunger is the end result from previous food intake and previous behavior as a function of time. It is possible to manipulate “hunger” by letting the animal fast during a given period of time and/or to compel it to perform a certain amount of physical activity such as running or swimming.