From M.A. Bozarth (1994). Pleasure systems in the brain. In D.M. Warburton (ed.), Pleasure: The politics and the reality (pp. 5-14 + refs). New York: John Wiley & Sons. (Note: Minor typographical errors appearing in the published version have been corrected.)

Michael A. Bozarth
Behavioral Neuroscience Program
Department of Psychology
State University of New York at Buffalo
Buffalo, New York 14260-4110 USA

Neurological research has identified a biological mechanism mediating behavior motivated by events commonly associated with pleasure in humans. These events are termed "rewards" and are viewed as primary factors governing normal behavior. The subjective impact of rewards (e.g., pleasure) can be considered essential (e.g., Young, 1959) or irrelevant (e.g., Skinner, 1953) to their effect on behavior, but the motivational effect of rewards on behavior is universally acknowledged by experimental psychologists.

Motivation can be considered under two general rubrics—appetitive and aversive motivation. Appetitive motivation concerns behavior directed toward goals that are usually associated with positive hedonic processes; food, sex, and wine are three such goal objects. Aversive motivation involves escaping from some hedonically unpleasant condition; the pain from a headache, the chill from a cold winter night are among the list of conditions that give rise to aversive motivation. The notion that hedonic mechanisms might provide direction to behavior can be traced at least to the Greeks (e.g., Epicures); Spencer (1880) formalized this notion into psychological theory and suggested that two fundamental forces governed motivation—pleasure and pain. Troland (1928) suggested that pleasure was associated with beneception, events that contributed to the survival of the organism (or species) and thus 'benefited' the organism from an evolutionary biology perspective; pain was suggested to be associated with nociception, events that had undesirable consequences for the organism. This schema—emphasizing hedonic processes in the regulation of behavior—lost favor with the advance of the Freudian and later behavioristic schools, although variations on this theme have occasionally resurfaced among motivational psychologists (e.g., Bindra, 1969; Young, 1959).

Behaviorism traditionally rejects the notion that subjective experience has a critical role in determining behavior. Specifically, behaviorism describes the relationship between behavior and external factors governing that behavior without reference to internal states, albeit it does help to have a hungry (i.e., food deprived) rat when studying the ability of food to serve as a reward. Behaviorism, or more properly operant conditioning theory, postulates three fundamental principles of behavior—positive reinforcement, negative reinforcement, and punishment. Positive reinforcement describes the situation where presentation of some stimulus event (e.g., food) increases the probability or frequency of the behavior it follows. Negative reinforcement describes the situation where the termination of some stimulus event (e.g., electric shock) increases the probability or frequency of the behavior its termination follows. Both positive and negative reinforcers increase behavioral responses; they differ in the temporal relationship between the behavior and the reinforcing event—positive reinforcers follow the behavior they reinforce, while negative reinforcers precede the behavior they reinforcement. (In colloquial terms, the organism is said to work to receive a positive reinforcer and to work to escape from a negative reinforcer.) Punishment is the third general principle of operant conditioning. Punishment describes the situation where presentation of an aversive stimulus following a behavior decreases the probability or frequency of that behavior. Unlike reinforcers, punishers suppress behavior. Radical behaviorism describes the effects of reinforcement and punishment on behavior devoid of their subjective impact. Indeed, the emotional states associated with reinforcement and punishment are usually viewed as the result of behavioral conditioning and not a cause of behavior.

In general, events that serve as positive reinforcers produce approach behavior defined as appetitive motivation. Events that serve as negative reinforcers or punishers produce withdrawal behavior defined as aversive motivation. Positive reinforcement is usually associated with a pleasant hedonic impact (and hence frequently termed reward connoting this pleasant affective component), while negative reinforcement and punishment are usually associated with an unpleasant hedonic impact. Whether the subjective experience of reward (viz., pleasure) plays an important role in determining behavior is moot for the present discussion. The same principles apply whether the emotional impact of a reward precedes or follows the behavioral response. Furthermore, events that serve as positive reinforcers in humans and other animals are generally described by humans as pleasant; thus, there is an intimate association between reward and pleasure despite controversy regarding the role of the subjective experience of pleasure in determining behavior.

Physiological psychology research has identified separate but interactive neural pathways mediating reward and aversion (i.e., functioning as positive and negative reinforcement systems, respectively). Direct activation of brain reward mechanisms through electrical and chemical stimulation provides a tool for elucidating these neural systems. During the past four decades, considerable knowledge has been gained regarding the anatomical and neurochemical basis of these pathways. This brief presentation addresses only brain mechanisms involved in positive reinforcement because they are closely identified with pleasure in humans and because they underlie the primary process governing much of normal behavior.

Olds and Milner (1954) first identified brain sites where direct electrical stimulation is reinforcing. Laboratory animals will lever press at high rates (> 6,000 times per hour) to obtain brief stimulation pulses to certain brain regions. The reinforcement from direct electrical activation of this reward substrate is more potent than other rewards, such as food or water. The potency of this electrical stimulation is most dramatically illustrated in a classic experiment where the subjects suffered self-imposed starvation when forced to make a choice between obtaining food and water or electrical brain stimulation (Routtenberg & Lindy, 1965). A second distinguishing feature of reward from electrical brain stimulation is the lack of satiation; animals generally respond continuously, taking only brief breaks from lever pressing to obtain the electrical stimulation. These two features (i.e., super-potent reward and lack of satiation) are important characteristics of direct activation of brain reward mechanisms.

Initial work suggested that a number of brain regions could produce rewarding effects, but many of these seemingly diverse stimulation sites were quickly linked through a common neural pathway—the medial forebrain bundle (Olds, 1977). Although it is true that activation of other brain systems can produce rewarding effects, activation of the medial forebrain bundle as it courses through the lateral hypothalamus to the ventral tegmentum produces the most robust rewarding effects. And several neurotransmitters may be involved in the rewarding effects from various electrode placements, but dopamine appears to be the neurotransmitter essential for reward from activation of the medial forebrain bundle system (see Fibiger & Phillips, 1979; Wise, 1978). The neuroanatomical elements of rewarding stimulation have been identified using electrophysiological and neurochemical techniques: electrical stimulation activates a descending component of the medial forebrain bundle which is synaptically coupled at the ventral tegmentum to the ascending mesolimbic dopamine system. Rewarding electrical stimulation thus activates a circuitous reward pathway, first involving a descending medial forebrain bundle component and then involving the ascending mesolimbic dopamine pathway (Bozarth, 1987a; Wise, & Bozarth, 1984). The terms mesolimbic and ventral tegmental dopamine system are used interchangeably in this context, both denoting the same dopamine system involved in reward and motivation.

Research with laboratory animals generally uses an operant conditioning perspective when studying reward processes (viz., without reference to possible subjective effects), but research in human subjects has revealed that comparable electrical brain stimulation is associated with profoundly pleasurable effects (e.g., Heath, 1964). Indeed, some experimental subjects liken the effect of electrical brain stimulation to intense sexual orgasm, and anecdotal reports suggest that human subjects have developed a strong romantic attraction to the researchers performing the experiments. For obvious ethical reasons, research with human subjects has been very limited. But the available data suggest that the principles learned from animal experimentation are valid for human subjects; studies of electrical stimulation of reward pathways in humans provide direct evidence that stimulation that is reinforcing in animals is both reinforcing and intensely pleasurable in humans.

Another approach to studying brain reward systems is to determine the neurochemical coding of these pathways. This can be accomplished by identifying the neurochemical mechanisms whereby various drugs serve as rewards following either systemic or intracranial administration. Essentially, reinforcing drugs can be used as tools for studying brain reward mechanisms in much the same manner as electrical stimulation. Experimental procedures have been developed where animals can lever press to obtain various drug rewards (see Bozarth, 1987b).

Some drugs delivered intravenously can serve as rewards. Most drugs that are self-administered by humans are also self-administered by laboratory animals. The most potent drug rewards include the psychomotor stimulants (e.g., amphetamine, cocaine) and the opiates (heroin, morphine). These drugs are self-administered by laboratory animals that have surgically implanted intravenous catheters. Animals quickly learn to press a lever to intravenously self-administer drugs such as cocaine and heroin. This experimental preparation provides an animal model of human drug-taking behavior and hence a method to study the reinforcing properties of drugs; this reinforcing drug-action forms the basis for drug addiction in humans (see Bozarth, 1987b, 1990). It is important to note that addiction is defined as a behavioral syndrome where a drug seems to exert extreme control over the individual's behavior and is not defined by physiological withdrawal reactions such as those accompanying abstinence from some drugs. Drug use is seen as developing along a continuum, beginning with casual/recreational use where the drug has a modest influence on behavior to the extreme condition (i.e., addiction) where the drug use seems to dominate the individual's behavior (see Bozarth, 1990).

Reward from psychomotor stimulants and from opiates appears to involve activation of the same brain reward system as that activated by electrical stimulation. Dopamine is the neurotransmitter most consistently linked with reward from these drugs, and the ventral tegmental dopamine system has been specifically implicated in psychomotor stimulant and opiate rewards. Other drugs that may serve as reinforcers (e.g., alcohol, barbiturates, caffeine, marijuana, nicotine) also activate the ventral tegmental dopamine system, although the data suggesting this activation is critical for their reinforcing effects are not conclusive. Furthermore, abstinence from cocaine or from morphine after repeated administration may decrease dopamine levels in this brain system (Bozarth, 1989; Rossetti, Hmaidan, & Gessa, 1992); this diminished dopamine function may be related to the intense craving associated with withdrawal in drug dependent humans. The subjective experience of craving is probably related to relapse into drug-taking behavior following abstinence and therefore is an important factor in drug addiction.

Research has progressed to where several distinct rewarding events can be explained by their abilities to activate a common brain reward mechanism: electrical brain stimulation reward, psychomotor stimulant reward, and opiate reward all appear to involve activation of the ventral tegmental dopamine system (Bozarth, 1987a; Wise & Bozarth, 1984). Several other drug rewards, such as alcohol and nicotine, may also involve activation of this brain pathway. This has lead to the assertion that various addictive drugs share the common feature of activating the same brain reward system and this action has been related to their appetitive motivational effects (Wise & Bozarth, 1987). This theoretical perspective deviated sharply from prevailing thought in that (i) it suggested a common neural basis for two distinctively different pharmacological drug classes (i.e., psychomotor stimulants and opiates) and (ii) it suggested that appetitive motivation rather than aversive motivation (such as that associated with physical dependence and overt withdrawal reactions) motivated drug-taking behavior and addiction. From this perspective, addictive drugs are seen to pharmacologically activate brain reward mechanisms involved in the control of normal behavior (see Bozarth, 1990; Wise & Bozarth, 1987). Thus, addictive drugs may be used as tools to study brain mechanisms involved in normal motivational and reward processes.

Other, natural rewards can be modulated by the activity of this system: feeding can be elicited (Hamilton & Bozarth, 1988), sexual behavioral can be aroused (Mitchell & Stewart, 1990), and maternal behavior can be facilitated (Thompson & Kristal, 1992) by opiate activation of this reward system. The origin of the ventral tegmental dopamine system (i.e., ventral tegmentum) appears to provide an important neurochemical interface where exogenous opiates (e.g., heroin, morphine) and endogenous opioid peptides (e.g., endorphins, enkephalins) can activate a brain mechanism involved in appetitive motivation and reward. These and other empirical findings are consistent with the notion that the ventral tegmental dopamine system may serve as an appetitive motivation system for diverse behaviors. This is not to suggest that all motivational effects of these rewards emanate from this single brain system, but rather this dopamine system represents one important mechanism for the control of both normal and pathological behaviors. (For a more technical review, see Bozarth, 1987a, 1991).

The hypothesized activation of the ventral tegmental reward system by endogenous opioid peptides can offer an explanation of seemingly paradoxical behavior—the voluntary self-infliction of stress or pain. Events normally considered stressful and thus aversive may activate the ventral tegmental reward system through the release of endogenous opioid peptides induced by the stressor. (Stress-induced release of endogenous opioid peptides was one of the earliest identified effects for these neuromodulators.) This could explain the attraction some individuals display to seemingly aversive stimulation (e.g., risk-taking behavior, self-infliction of painful stimuli). In some situations the appetitive motivational effect of these behaviors may override the normal aversive motivational effect that usually produces withdrawal behavior; thus in certain pathological conditions, approach behavior indicative of appetitive motivation may be produced by an aversive stimulus normally avoided and described as painful. This is most likely perhaps in situations where the effects of the stress-induced endogenous opioid peptide release out last the abrupt termination of the painful stimulus. Also, cognitive processes may label the stressor as nonthreatening, thereby permitting the pleasurable effects to dominate affective tone.

Activation of brain reward systems can be considered a natural component of normal behavior. Indeed, brain reward systems serve to direct the organism's behavior toward goals that are normally beneficial and promote survival of the individual (e.g., food and water intake) or the species (e.g., reproductive behavior) as suggested by Troland's (1928) concept of beneception. The notion that the brain influences behavior is not particularly radical for twentieth century scientists nor is the notion that many rewards activate such mechanisms through various sense modalities such as taste or touch. The direct chemical activation of these reward pathways does not in itself represent any severe departure from the normal control reward systems exert over behavior. Inhalation of a substance (e.g., nicotine) is no less natural than the ingestion of sugar, although the former has no direct survival value to the organism nor to the species. But both involve activation of brain reward mechanisms and both may be subjectively experienced as pleasurable in humans. So what constitutes the pathological control of behavior termed "addiction"? Certainly not the fact that a substance activates a brain reward system nor the fact that this same system may be involved in the potent reward produced by addictive drugs. Simple activation of brain reward systems does not constitute addiction! Rather, the extreme control of behavior—exemplified by a deterioration in the ability of normal rewards to govern behavior (termed motivational toxicity)—is the distinguishing feature of an addiction. Some drugs quickly and uniformly exert extreme control over behavior (e.g., cocaine, heroin), while other substances exert a much less potent influence on behavior (e.g., moderate alcohol consumption, occasional nicotine use). The fact that a chemical (e.g., nicotine) influences behavior does not constitute addiction any more than the chemical reaction that produces a taste (e.g., sugar-associated sweetness) which influences behavior constitutes addiction.

Motivational toxicity is apparent when rewards normally effective in influencing behavior lose their ability to motivate the organism. This is typically seen in human drug addicts that neglect formerly potent rewards (e.g., career, sex) and focus their behavior on the acquisition and ingestion of drug. The neural mechanisms responsible for this disruption of the motivational hierarchy have not been identified; one potential mechanism involves decreased dopaminergic function following chronic drug use (see Bozarth, 1989). In a reward system with decreased dopaminergic function, natural rewards that activate reward processes much less potently than some drug rewards (e.g., cocaine, heroin) may lose their abilities to engage the organism's behavior. In contradistinction, direct pharmacological activation of a reward system dominates the organism's motivational hierarchy at the expense of other rewards that promote survival. The ensuing motivational toxicity distinguishes drug addiction from simple drug activation of reward mechanisms. Motivational toxicity may develop from neuroadaptive responses to chronic intake of some drugs, but it is not a general property of chemical activation of brain reward mechanisms.

Brain systems involved in what the behaviorist terms positive reinforcement are also involved in the sensation of pleasure in humans. Although radical behaviorism ignores the hedonic impact of positive reinforcers, the subjective experience of pleasure is a usual concomitant of positive reinforcement. Because humans most often describe their own behavior in terms of subjective experience instead of the behavioristic terms of operant conditioning theory (e.g., positive reinforcement), it is appropriate to use reward and pleasure as descriptors of events governing human behavior. Indeed, the phrase introduced by Olds (1956), "pleasure centers in the brain," remains generally descriptive of the neural basis of reward, but the word center (suggesting a single neuroanatomical focus) has been replaced by the word systems (emphasizing multiple neural elements) as additional neural linkages have been identified.

Appetitive motivation is most often associated with goals that have benefited the species from an evolutionary biology perspective. Specialized brain systems have evolved that direct the organism toward historically beneficial goals, and these systems could be termed pleasure systems in colloquial language. Whether the sensation of pleasure is a critical determinant of behavior or a simple concomitant of reward activation remains to be resolved: appetitive motivational is intimately linked with the subjective experience of pleasure.

Much of behavior can be explained by simple processes of approaching pleasant stimuli and avoiding painful stimuli as described by Spencer (1880) in the nineteenth century. The ventral tegmental dopamine system is an important neural substrate for reward, and it has a central role in regulating appetitive motivation: several distinct rewarding events activate this reward system, and activation of this system elicits appetitive motivation. The ventral tegmental dopamine system, along with its various neural inputs and outputs, can be aptly designated a "pleasure system in the brain" with an important role regulating many normal and pathological behaviors.