Intraorganismic Versus Interorganismicfactors In The Control Of Behavior

In this section of the text we will be dealing with the description of the activities of the organism from the point of view of behavioral rather than neurologic science. Obviously, many processes going on intraorganismically are relevant to how a man acts. Nevertheless, it is still useful to proceed with a natural scientific analysis of the actions of man from the point of view of behavior rather than neurologic or chemical science. Eventually it will be possible to relate the neurologic activities of the organism to its activities in the external environment. At present, however, the amount of information available about the interrelation between the functioning of the nervous system and the activity of the organism is very small compared to the order of magnitude of the data to be explained. Even when neurologic and physiologic science has advanced further, there will still be advantage in a prior analysis of man's performance in terms of its effect on his environment. A natural-scientific analysis of the behavior of an organism can proceed without reference to its neurologic or physiologic substrates by describing the activities of the organism as they occur in time and as a function of the relevant conditions of the environment supporting them. This behavioral information will be of practical and scientific importance even after the neurologic and other physiologic substrates have been discovered because it specifies manipulable conditions which, when altered, will enable control of the behavior. A natural-scientific analysis of the relation between a man's performance and its effect in the social and physical environment will also provide the focus and define the problems for future research on the role of neurologic and physiologic factors. Before we can discover and describe these intraorganismic factors, however, we must first identify the phenomena and processes to which they may be relevant. Heretofore unknown aspects of human behavior are being discovered continuously. These new kinds of environmental control of behavior raise questions concerning the intraorganismic factors which otherwise could not be asked.

A Functional Analysis Of Behavior

It is not sufficient in itself simply to describe behavior in objective terms. A functional analysis relating some objective description of a man's activity to systematically defined independent conditions in the environment is required. We would have no problem, for example, in describing the behavior of a child crying. We could take a motion picture and record the sound. A more complete description of the child crying, however, requires a functional analysis relating the observed behavior to the antecedent conditions of which it is a function. The child could be crying, for example, because it had just been spanked. It could be crying because it had a toothache. It could also be crying because this behavior leads to parental intervention in the solution of a problem. The form of the behavior is identical in all of these cases, but functionally the performances are very different. Some of the environmental factors which in fact could be responsible for the behavior of the crying child will be discussed later, but for the present it is sufficient to indicate that very different kinds of environmental control all involve the identical behavior: the vocalization we characterize as crying. The major problem in the analysis of behavior is to relate objectively measured performances to the relevant antecedents and effects in the environment. Many other examples of topographic description of behavior that tell us little about the important aspects of man's behavioral repertoire are possible.

Much of the functional analysis of behavior has come from animal experiments, which are used extensively in a manner similar to the use of animal experiments in physiology and pharmacology. The use of animal subjects is almost mandatory in discovering general principles of behavioral control because of the enormous convenience and the ability to manipulate variables and restrict experimental situations. While the rat and pigeon may seem a great phylogenetic distance from the human, behaviorally there are very broad similarities in the basic underlying processes, as the digestive, nervous, and circulatory systems. Important differences, indeed, do exist among the various species, but it is nevertheless true that there are substantial broad similarities. Animal experiments have revealed that many behavioral processes are general over a wide range of species. This correspondence holds true for other processes such as the way in which new events come to be reinforcing stimuli. The subsequent sections will describe in greater detail many processes relevant to important behaviors in man which may be observed in their most general form in many species.

The Reflex And Conditioned Reflex

The discovery by Pavlov of the conditioned reflex at the turn of the century was the first major scientific discovery in which some aspect of an organism's activity relevant to broad problems in psychology was controlled by manipulating the environment in some lawful way (Pavlov, 1927). Pavlov discovered, essentially, that any stimulus present at the time a reflex is elicited would also come to elicit the major portions of the unconditioned response of the reflex. In his early classical experiments, Pavlov sounded a buzzer every time he placed food in a dog's mouth. He discovered that after a sufficient number of pairings, the buzzer alone would cause salivation. He called the process by which the buzzer (unconditioned stimulus) acquires control over the salivatory response (unconditioned response) conditioning, and his discovery of this process represented the first description of a lawful relation between the activity of the organism and its relation to the external environment. Pavlov studied the properties of the conditioned reflex in considerable detail in a classical series of experiments. Much of the work stands today and has provided major emphasis for the further analysis of the activities of the living organism in its environment. Pavlov's discovery of the laws of extinction, generalization, stimulus control, disinhibition, and differentiation of the conditioned reflex provided a basis for many of the significant principles of psychology that have been developed subsequently. While the principles of the reflex are not directly applicable to the bulk of human behavior, e.g., voluntary behavior, they carry the outlines of some of the major principles operating there.

Although Pavlov dealt almost predominantly with the salivation reflex and the flexion reflex, in response to electric shock, the scope of the conditioned reflex is probably as extensive as is the control by the sympathetic and parasympathetic nervous systems. While only limited amounts of systematic and precisely controlled observation are available on the formation of conditioned autonomic responses in many organ systems of the body, it is clear that we may expect to find major and important conditioned reflexes in almost every sphere of organ activity affected by the autonomic nervous system. Conditioning has been demonstrated, mostly by Russian physiologists, in such a large number of organ systems of the body that there is no doubt that reflex conditioning represents one of the major avenues of change in the internal economy of an organism as a result of its interaction with the external environment. Constriction and dilation of the vessels of the arterial system have been conditioned and have shown properties similar to those of the classical conditioned reflex. This is also the case with excretion of urine from the kidney into the bladder, the excretion of bile, heart rate, respiration, gastric secretion, and splenic contraction. Even though the precise effects of environmental changes via conditioning of autonomic effects are only beginning to be understood, they represent a major class of variables which have direct relevance to many medical specialties. In the area of cardiovascular disease, for example, the systematic change in heart rate and arterial diameter during the day-to-day exposure of the individual to his environment via conditioning is of a large order of magnitude and will undoubtedly prove to be a significant factor in etiology, treatment, and prevention of circulatory pathology. A similar situation exists with immunologic responses affected by various hormones, which are in turn controlled by the autonomic nervous system and potentially affected by environmental control via reflex conditioning. In the field of gastric ulcers the evidence that environmental factors are responsible for prolonged pH changes in the gastric fluids is becoming almost compelling (Brady et al., 1958). We do not know exactly how the day-to-day conditioned reflexes of the autonomically innervated organs of the body are related to the development of pathology. Nevertheless, at the present stage of knowledge, it would be surprising if the effects were not of major importance. The great amount of variability of organisms in response to disease could well be due, in large part, to the varied history of the individual and the effect of this history of reflex conditioning on the autonomically innervated structures of the body.

The reflex, on which the Pavlovian conditioned response is based, derives its main features from the phylogenetic history of the organism. In the reflex, we are concerned predominantly with the response of the smooth muscles and glands, and other organs innervated by the autonomic nervous system, to eliciting stimuli from the external environment, for example, the eye-blink elicited by an object moving toward the eye, the salivatory response of the parotid gland in response to food in the mouth, gastric secretion of the stomach to food or the cessation of secretion due to trauma, secretion of the sweat glands with temperature, vasoconstriction, ACTH-hydrocortisone response, and heart rate and heart form.' These examples of activity of an organism can be contrasted with a second broad class of activities called operant behavior involving, predominantly, the striated musculature and the so called voluntary system, which is related to the central nervous system rather than the sympathetic and the parasympathetic systems. The operant repertoire will be the main focus here. The reflex, perhaps, belongs more properly to the realm of physiology than psychology because it is a one-way interaction with the external environment: an eliciting stimulus (unconditioned stimulus) in the external environment produces a specific change in the organism (the unconditioned response) which in turn has importance mainly for the internal economy of the organism rather than for reaffecting the external environment. Hence, the term respondent behavior for reflex behavior. It is the unilateral response of the organism to the environment. The reflex is the integrated sum of the unconditioned (eliciting) stimulus and the unconditioned response. The reflex (respondent behavior) represents involuntary control in the sense that full control of the behavior is in the eliciting stimulus which in turn derives its effect almost completely from the phylogenetic rather than ontogenetic history of the organism. When the unconditioned stimulus is specified, the unconditioned response is almost completely determined. More is known about the neurologic substrate of the reflex than about other activities of the living organism. Nevertheless, the reflex can still be profitably studied behaviorally or environmentally by describing the relationship between the various characteristics of the eliciting stimulus (unconditioned stimulus) and the elicited response.

Magnitude of the unconditioned response is the function of the magnitude of the stimulus. The greater the force with which the patellar tendon is struck, the greater the deflection of the leg.

Refractory Phase: After a reflex is elicited, there is a short period during which further stimulation will not produce the unconditioned response.

Temporal Summation: Prolongation of the stimulus, or repetitive presentation within certain limiting rates, has the same effect as increasing the intensity.

Adaptation: The strength of a reflex declines during repeated elicitation and returns to its former value during subsequent inactivity.

These then are the relations between properties of the elicitingout reference to the inner mechanisms responsible for the performance. Regardless of what neural mechanisms are present, we can explain the reflex in a practical sense when we describe its conditions of elicitation by the environment. We owe the laws of the reflex mainly to the work of Sherrington (1906) in his classical experiments reported in his lectures on "The Integrative Action of the Nervous System." Even though the work is intended to provide a neurophysiologic basis of the reflex, the main findings are a description of the relation of the eliciting stimulus to the correlated response.

It is interesting to note, as with Sherrington, that Pavlov presented his experiments as investigations of the activity of the cerebral hemispheres. Actually, Pavlov's references to the central nervous system were almost entirely inferential as is also true with all of the recent Russian work on conditioning. The central nervous system enters into these experiments only in the sense that the conditioning of a new stimulus-response relationship involves the integrated activity of the organism. The central nervous system is relevant to the extent that it is responsible for the integrated activity of the organism, but the actual technical analysis carried out by Pavlov did not manipulate or measure any factors in the central nervous system.

Operant Reinforcement

While the pioneering experiments of Pavlov on the conditioned reflex have had a great influence in the development of an objective science of psychology and have provided a major impetus for the further analysis of the activities of the living organism in its dealings with the environment, it has become increasingly apparent that the simple Pavlovian reflex involves a very different activity than is characteristic of the bulk of human behavior. Although the principles of reflex conditioning stimulated much research activity in the application of technics of natural scientific inquiry to problems in human behavior, it did not provide answers to major questions of how a man deals with his environment in the very complex ways involving the striated musculature, as for example in walking, talking, solving problems, and constructing machines. The most casual observation of human behavior reveals a rich variety of topographic forms which develop from the relatively simple initial repertoire of the infant. The conditioned reflex, in contrast, deals with rather fixed forms of behavior in which the form of the response which is conditioned is fixed and determined by the phylogenetic history of the organism. The conditioning procedures simply bring an already existing reflex under the control of a new stimulus. The reflex obviously lacks the sensitive interchange between the behavior of the organism and its environment that characterizes most of human and animal behavior. More relevant to the bulk of human psychologic activity is the operant repertoire of the organism.

The operant behavior of the organism is defined as such because it is maintained by its effect on the environment. The operant repertoire consists of those activities of the organism which change the external environment, which in turn alters the subsequent state and behavior of the organism. This is the voluntary behavior of the striated musculature instead of, in most cases, the smooth muscle and autonomic control of reflex behavior. An operant response may also be reinforced by its effect on the same individual's behavior, as for example in self control. The behavioral processes involved, however, are still of the same kind as operant reinforcement. The analysis of this kind of behavior is treated elsewhere (Skinner, 1953; Ferster, Nurnberger and Levitt, 1961). The effect of operant behavior in changing the organism's environment makes possible the tremendous complexity and variety of the voluntary behavior of higher animals. The basic principle by which an organism acquires operant behavior is called operant reinforcement, or the law of effect. The major principles were discovered at the turn of the century, but only recently has sufficient technical development occurred to make possible the full realization of the power of these principles.

Not all of the changes in the environment which occur as a consequence of the operant activities of an organism will affect the organism significantly, however. Changes in the environment may be ineffective, effective only under certain conditions, or have other functional relations to behavior which will be discussed later. Those consequences which increase the frequency of occurrence of the activity they follow are called reinforcers.

The exact nature of the reinforcement is determined by the phylogenetic history of the organism and, as a result, would be very different from species to species, and, perhaps, from individual to individual. A hungry bird, for example, will repeat those activities which produce grain, while for a dog the reinforcing event would have to be, perhaps, meat. Once we identify which events are reinforcing, however, it is possible to describe the characteristics of the processes by which these reinforcing events support the behavior of the organism.

An animal demonstration illustrates the essential details of the process of operant reinforcement. A hungry pigeon is trained to eat grain from a food dispenser, designed to deliver food in small amounts. In practice, approximately 60 reinforcements of approximately 0.25 grams each constitute the bird's entire daily food ration. A prominent light and a sound accompany the presentation of food, and the bird is trained to approach the food magazine whenever the light and sound occur and not to approach it in their absence. It is then possible to increase the frequency of some act in the bird's repertoire by following (reinforcing) it with the stimuli accompanying the food magazine. To increase the frequency with which the bird raises its head, we would wait until the bird stretches his neck and follow it instantly with the operation of the food magazine with its light and sound. The frequency of neck-stretching increases and the bird will continue to stretch its neck so long as the presentation of the food magazine follows within the required temporal limits, and so long as the bird is hungry. When we discontinue the relation between the conditioned activity (neck-stretching or facing the wall) and continuation of reinforcement, the frequency of neck-stretching declines and reaches the low value present before the behavior was conditioned (reinforced). The discontinuation of reinforcement is called extinction and represents a second major principle of behavior. The reinforcement may be used to increase the frequency of almost any performance, so long as it has some minimal strength in the bird's repertoire. We could increase the frequency of lifting a leg, raising a wing, pecking, or lowering the head.

The law of operant reinforcement is the major factor responsible for the development of the new behavior in an organism, answering the question, "Where does behavior come from?" It is important to note that operant behavior actually changes the environment in contrast with conditioned reflexes which change the internal milieu and have little direct effect on the external environment. The operant behavior of shaking a tree, reinforced by the apples which fall, is an example of a performance producing a magnitude of environmental change. The actual physical properties of the tree define the behavior which will be reinforced. Those responses which will make the tree move and the apples fall will increase in strength, while other responses extinguish. Eventually, the behavior that develops will be those responses which deliver apples. The person shaking the tree might also salivate as a conditioned reflex to the sight of the apple, but this response has little direct relevance to strengthening the performances reinforced by the receipt of apples and, of course, no direct effect on the tree. It is this feature of operant behavior that makes possible the sensitive interchange between the organism's behavior and its environment which characterizes so much of human behavior. Operant reinforcement is the basic operation for most environmental control of behavior. The term, environmental control of behavior, does not refer to an over-all influence, as in the effect of climatic conditions, but to the fact that specific kinds of alterations in the functional environment can produce a new form of behavior or increase the frequency of some item in the organism's repertoire. By discontinuing reinforcement, behavior may weaken or disappear, and by resuming reinforcement, the performance in question may be once more strengthened. In this sense, the environment contains the independent variables of which the organism's behavior is a function.

The Successive Approximation Of New Complex Forms Of Responding

It is crucial that the actual reinforcement in the animal experiments just described was not the food, but rather the light and sound which accompanied the delivery of food and which did not occur when food was not delivered. It is this specificity which makes necessary the new technical term, reinforcement, rather than the colloquial "reward" which would seem at first hand to be adequate to describe the situation. The light and sound of the food magazine are the actual reinforcement used to increase the frequency of the specified activity. The presentation of the food is, of course, a necessary event for the continued maintenance of the reinforcing effect of the light and sound. Using the stimuli accompanying the delivery of food (conditioned reinforcement) makes it possible to bring the reinforcement to bear upon very specific performances. We can arbitrarily select almost any aspect of the ongoing behavior of an organism and increase the frequency of occurrence of that specific activity because the conditioned reinforcement can occur almost instantly. The delay is no greater than the few milliseconds necessary to energize a lamp or solenoid. This technical aspect of reinforcement makes it possible to develop complex forms of behavior by a corollary process of reinforcement which can be called, variously, conditioning by successive approximation, shaping, or response differentiation. Once there is a reinforcer which can be made instantly contingent upon a specific activity of an organism, it is possible to begin with some performance already in the repertoire of the organism, strengthen it by following it with the reinforcement, and then gradually develop more and more complex forms by observing the natural variability from instance to instance of the activity, and shifting the reinforcement in the direction of the complex form of behavior that is required. By successively reinforcing small gradations in the direction of the required performance, it is possible to shape the activity of the animal much in the manner that a sculptor shapes clay.

A commonly used classroom demonstration in which a pigeon is trained to peck at an illuminated disc on the wall of the cage illustrates the essential nature of the process. A pigeon will ordinarily have a nearly zero frequency of pecking at the wall of its cage. The first step is to operate the food magazine only as the bird turns its head in the direction of the disc. The effect of one or two such reinforcements in a properly prepared demonstration is an almost instant increase in frequency of turning the head in the direction of the key. Reinforcement is then withheld in favor of a slight nod in the direction of the key or a slight step in the direction of the key. At each stage the response already in the bird's repertoire is reinforced, but the experimenter tends to withhold reinforcement in favor of a response which is the next approximation of the required performance. By a series of gradual steps, the reinforcement contingency, and, hence, the behavior, is shifted gradually until a required form of behavior is achieved. By continuing the process of reinforcing successive approximations, extremely complex response repertoires different from any behavior originally in the bird's repertoire may be built. At any stage, however, responses which are reinforced are already in the bird's repertoire. The change in behavior is produced by the reinforcement of small variations that occur in the form of the response from stage to stage.

Although reinforcement is relevant to almost every behavioral process, it is especially relevant to growth and development of the maturing child where we are primarily concerned with the question, "Where does the rapidly developing complex repertoire come from?" It is in the infant that we see the rapid emergence of new forms of behavior, all evolving from the very limited means of changing the environment that are present in the healthy newborn infant. The two major phenomena are: (1) The development of reinforcing events. As the child matures, events become reinforcers through the development of explicit behavioral processes as well as through the maturation of the organism. (2) The development of complex repertoires through successive approximations of more and more complex performances by the new reinforcers as they are in the child's repertoire.

These processes operate in the context of the physiologically maturing organism. The effect of differential contingencies of reinforcement in developing the increasingly more complex repertoire of the growing child is limited by the state of muscular, neurologic, and biochemical maturation. In the specific case of a limited behavioral repertoire, we may not know how much is due to the incompletely developed musculature and how much to the lack of potential environmental control. In any event, however, even if there is a potential of being sensitively affected by differential reinforcement from the environment, the process cannot be effective if the musculature and its neural control are not sufficiently developed to be capable of an increased complexity of movement. Even though anatomic and behavioral maturation are intimately interrelated, we will deal specifically in this section with the form of the processes involved in behavioral development.

As the individual matures, new kinds of environmental events come to be effective in changing the child's behavior. The effectiveness of some of these reinforcers depends primarily on anatomic maturation. Others depend more on the development of new performances as a result of the same process of successive approximation described above. The ability, for example, to be reinforced by the attention of the parent or punished by his criticisms represents the action of a reinforcing stimulus which derives its effect from very complex behavioral processes. Conversely, direct environmental reinforcers, such as the change in visual stimulation which occurs when the child moves its fingers in front of its eyes, become relatively less reinforcing in the context of the more complex repertoire of the older child. The early behaviors of the infant provide performances which are established by successive approximation. If changes in the field of vision, for example, are potential reinforcers for an infant, the movement of the hands and fingers will be differentially reinforced by its effect on the visual field. Movements which bring the hands into the field of vision are reinforced and others extinguished. Tactual exploration provides a similar example. The differential reinforcement and continuous shaping of movements of the hand which bring the most sensitive parts of the hand over various textures which may be available to the child reinforce the effective forms and weaken the ineffective ones.

Operant reinforcement in the newborn infant often acts in context of behaviors which are originally reflex, as, for example, crying and eating. The newborn infant will begin sucking as a reflex response to the tactual stimulation of the mother's breast. After some experience, however, the operant control resulting from the differential effects of different degrees of sucking "shapes" up the behavior because the performance is one which is potentially in the operant repertoire. The appearance of milk in the infant's mouth as a result of sucking increases the frequency of the specific responses which increase the flow of milk. Mouth and diaphragm movements which are more effective in producing milk become more capable, and less effective movements tend to disappear because of nonreinforcement.

The same differentiation process occurs in the development of the behavior in sucking milk from a bottle. The physical properties of the bottle define the conditions under which reinforcement can occur, and the differentiation process is the interaction between the initial repertoire of the infant and the effects of these performances in producing milk from the bottle. Eventually, the performance becomes even more firmly differentiated as the magnitude of the sucking comes to change as a result of the amount of milk in the stomach and the rate at which it can be swallowed. The difficulty of the infant in releasing the nipple so that air may enter the bottle illustrates the importance of immediate reinforcement in developing new behavior. The infant has no feedback of the immediate consequences of releasing the nipple. The required chain of responses does not usually develop until the end of the first year.

The shaping of behavior by differential positive reinforcement occurs continuously in almost every stage in the development of the infant. Consider, for example, weaning the infant to a cup. Here, a new set of performances, reinforced by the ingestion of milk, is to be developed. At first, the repertoire which the child brings to bear upon drinking from the cup is almost entirely appropriate to the behavior developed previously when it sucked its milk from a bottle or breast or put other objects in its mouth. If the initial performances of the lips, tongue, and diaphragm have any success in producing some milk in the mouth, the behavior will be sustained and conditions will be present for the differential reinforcement and shaping of the behavior in the direction of the complex performances eventually required in drinking. Any variation in the infant's behavior which will produce more milk in the mouth will increase in strength. The required performance is a complex one involving the coordination of the lips, mouth, tongue, diaphragm, and the muscles used for swallowing. Those activities which have no effect will gradually extinguish. The likelihood of developing the new performance depends upon the physiologic development of the organism so that it is ultimately capable of the behavior even under the most favorable psychologic conditions; a current repertoire which contains the minimal elements to produce at least some food in the mouth, and, hence, maintain the behavior necessary to the further shaping of the final drinking performance; and a sufficient level of food deprivation so that the appearance of given quantities of milk in the mouth from the cup will be sufficiently reinforcing to continue to maintain the performance. There is less likelihood, for example, that a child will learn to drink from a cup just after having completed a meal than before completing it. If bottles are readily available during weaning, then the development of the new repertoire will depend upon an initial repertoire in which drinking will produce more milk than sucking.

Environment differentially reinforces almost all performances with continuous shaping behaviors such as picking up objects, eating from a spoon, crawling, walking, sitting erect, focusing the eyes, and manipulating objects with the hand. All of these emerge from performances which initially have only limited effects on the environment and later develop progressively as each change in form leads to a more direct effect on the environment. Each new performance raises the possibility of successive approximation of other behaviors. When, for example, the block slips out of a child's hand slightly out of reach, any movements which produce a translation of the child's body on the floor will be reinforced by the block. As a result, the crawling movements of the child would be successively approximated in the direction of forms maximally effective in bringing the child into physical contact with reinforcers which would be unavailable otherwise.

Why Does a Reinforcer Increase the Frequency of a Response? The reasons why a given effect of an organism's behavior on the environment is reinforcing is ultimately found in the phylogenetic history of the organism. For example, grain may be used as a reinforcer for a pigeon, meat for a dog, or hay for a cow. Other reinforcers may be derived from their function in chains (see Chapter 13) leading to phylogenetically based behaviors. Whatever the reasons for a reinforcer increasing the frequency of a response, it is the cause, in a sense, of the behavior because it is a manipulatable event by which the behavior is created, strengthened, or weakened. While the effect of a reinforcement is altered significantly by its interactions with other behavioral processes, a reinforcer can be identified best by arranging that it no longer follows the behavior. The disappearance of the behavior when it no longer is followed by the reinforcer may then be taken as evidence that it was, in fact, reinforcing behavior. For example, if the question arises as to whether the child's crying is being maintained (reinforced) by the appearance, attention, and fondling of the parent, the experiment to be carried out is to arrange that these consequences no longer follow crying. If, in fact, these consequences were the reinforcers maintaining the behavior, then the frequency of crying should decline continuously at a rate depending upon the previous history of reinforcement.

The ultimate phylogenetic basis of most reinforcers does not imply that all human behavior is based on events which have a homeostatic physiologic effect, such as eating. There is ample evidence that many simple, direct effects on the environment completely unrelated to important physiologic regulatory mechanisms may maintain behavior significantly. Experiments with monkeys and rats show the possibility of maintaining significant amounts of behavior with reinforcers which bear little relation to feeding, sexual behavior, or reduction of aversive stimuli. In actual practice it is very difficult to distinguish between environmental consequences which are innately reinforcing and those which derive their reinforcing effect from some phylogenetically based behavior. It is usually not necessary to make the distinction, however, so long as conditions are present which make a given environmental change reinforcing. We may in most cases analyze the environmental effects of a reinforcer by using it to control behavior without inquiring into its origin. The reinforcing effect of food, for example, depends on a complex chain of gastrointestinal reflexes blending ultimately into the complex chemical processes of digestion. The role of these subsequent events is clearly of interest in explaining some of the reasons why food is reinforcing. From the point of view of a psychologic analysis, if we could assume that the gastrointestinal events subsequent to swallowing occur uniformly, we would begin at the point when the animal eats.