Is Addictive Behavior Compulsive?

The concept of compulsion is ubiquitous in scientific theorizing, professional practice, and lay understandings of addiction, and it forms the cornerstone of the disease view of addiction, including its latest version – the brain disease model of addiction. According to the disease view, it is because addictive behavior is compulsive rather than freely chosen that addicts should not be blamed and punished for their troublesome behavior. This logic is being increasingly transferred from addiction to psychoactive drugs to harmful and repetitive forms of gambling, shopping, Internet use, and other so-called behavioral addictions.

Given the centrality of the concept, it is a surprising fact that it is not at all clear what exactly compulsion is supposed to mean in relation to addiction. A review of the relevant literature soon reveals that different authors and different institutions mean different things by the term. Among the variety of definitions and characterizations, however, it is possible to identify three distinct meanings of compulsion with different implications for the attempt to understand addictive behavior (see Heather, 2017a).

The minimal definition

An articulate rendition of this kind of meaning is provided by Steven E. Hyman, a former Director of the National Institutes of Mental Health  in the USA:

“The term compulsion is imprecise, but at a minimum implies diminished ability to control drug use, even in the face of factors (e.g., illness, failure in life roles, loss of job, arrest) that should motivate cessation of drug use in a rational agent willing and able to exert control over behavior” (Hyman, 2007, p.9).

The idea here is that, because addictive behavior leads to degrees of harm that no-one would willingly inflict upon themselves, people should not be seen as choosing to behave in this way; they must be compelled to do so. It is also implied that it is despite knowing that repeated, persistent drug use causes them (and possibly others) harm that addicts continue to do it. As Hyman says, this is a minimal definition of compulsion because it makes no other assumptions and offers no other insights than what has been stated. To reiterate: repeated, persistent drug use causes harm; addicts know that it causes harm but continue to do it; no rational person would choose to behave that way; therefore, they must be compelled to do it.  

Now, if that is all that is meant by compulsion in addiction, and one suspects that this is often the case, it must be recognized as a legitimate use of the word in this context. Unfortunately, it gets us nowhere for scientific purposes. This is because it is purely descriptive and offers no clue as to why addictive behavior is compulsive in the way described. This minimal definition is merely a restatement of the crucial puzzle of addiction that addicts repeatedly behave in ways they know are bad for them. The danger is of assuming, implicitly at least, that by calling such behavior compulsive something has been explained, that we have somehow gained a better understanding of addiction than we had before we called it compulsive. Of course, any “explanation” of this kind is circular. Why do addicts continue to use despite knowing it to be harmful? Because of compulsion. What do we mean by compulsion? The fact that addicts continue to use despite knowing it to be harmful.

There is also a conflation here of the concepts of compulsion and irrationality. Whether or not it is usefully called compulsive, the behavior described by Hyman is certainly irrational, both from the viewpoint of an outside observer and of the person in question. Unfortunately, human irrational behavior is not confined to what is conventionally called addiction. Ever since the preoccupation by ancient Greek philosophers with the concept of akrasia or “weakness of will” – the fact that all of us sometimes choose to act against our better judgment – we have been scratching our heads to try to figure out why this kind of irrationality occurs (Heather, 2017b). The fact of unwanted, self-destructive consequences of drug use is not inconsistent with assuming that addicts make choices. These choices need not be rational but it remains a mystery in need of explanation why addicts’ irrational choices are so self-destructive. The minimal and commonly encountered definition of compulsion illustrated by Hyman confuses compulsion with irrationality. The conclusion is that, to be useful in trying to explain addiction, the concept of compulsion needs to refer to something more than the mere fact of irrational, self-harmful behavior.

Automaticity: the strong version of explanatory compulsion

There are uses of compulsion in the addiction literature that do qualify as attempted explanations and I shall discuss two broad categories of such uses here, which I call the strong and the weak versions of explanatory compulsion.

Aristotle compared compulsion in human behavior to someone being carried by a strong wind, that is when the cause of the behavior is in the external circumstances and the agent contributes nothing to it.  It is in this sense of compulsion that behavior is “against the will.” It is in the spirit of Aristotle’s understanding to include as compulsion an internal state such as brain damage leading to someone behaving in ways she does not desire. The crucial point is that compulsion takes control of behavior away from the person’s volitional, motivational states (see Stephens & Graham, 2009).

A modern version of Aristotle’s wind that blows someone along against her will is “automaticity,” a topic that has been studied by psychologists for the past 40 years. The basic idea is that human cognition is made up of two types of information processing – automatic and controlled. This has resulted in what is called the dual process theory of human cognition which assumes that observable behavior emerges from the interaction between automatic and controlled cognitive processes. Automatic processes are activated in the absence of ongoing control or attention, are based on highly repetitive learning experiences and characterize most daily activities. By contrast, controlled processes refer to behavior activated under the control and attention of the person.

Aberrant learning theory

The strong sense of compulsion refers to the idea that addictive behavior is the result of automatic processes over which the addict has no or little control. It is called “strong” because automatic processes are considered both necessary and sufficient for the occurrence of addictive behavior. A prime example of a theory of addiction along these lines is Everitt and Robbins’ (2005) aberrant learning theory in which addiction is viewed as the culmination of a series of transitions from voluntary drug-taking, through habitual use, to compulsive use.

Following repeated self-administrations, the user comes to associate certain stimuli (people or settings associated with drug use, drug-using paraphernalia, etc.) with reward and these become conditioned reinforcers that maintain drug-seeking behavior. As a result, drug-seeking becomes largely “automatic” and is carried out independent of the user’s conscious preferences and motivations. Although addictive behavior is not directly equated with procedural skills, like playing the piano or tying one’s shoelace, it is stressed that there is a “constant re-initiation” of automatic behavior of this kind. Behavioral changes are accompanied by transitions at the neural level from control principally by the prefrontal cortex to control by the striatum; at the same time, there is a shift of control in the striatum from ventral to more dorsal regions.

Evidence for and against the strong version

Automaticity may certainly apply to acts of ongoing drug consumption. Alcohol addicts narrow their drinking to a restricted selection of places, times, and beverages; a smoker unthinkingly lights up a cigarette while one is already burning in an ashtray, and the rituals accompanying intravenous drug injection are well documented. If this were all addiction consisted of, then stereotyped and automatic might be appropriate descriptors.

However, it is an account of the drug-seeking, or of seeking opportunities for addictive behavior in general, that is essential to any satisfactory account of addiction. Ethnographic research shows that, while drug users’ lives could be described as highly structured and narrowly focussed, owing mainly to the daily effort to obtain an expensive supply of drugs, the means by which this is accomplished are flexible and varied, even at times highly ingenious (see, e.g., Neale, 2002).  People with addictions are certainly not helpless victims of automatic forces over which they have little or no control. The notion of automatic processes may well apply to the occurrence of urges and cravings in response to drug-related stimuli but such urges and cravings do not automatically lead to drug-seeking behavior and do not determine the form it takes.

Another problem for the strong version of compulsion is that it makes explaining relapse, i.e., the return to an unwanted behavior after a period of intentional cessation, difficult.  It is commonly accepted that addiction is “a relapsing condition,” i.e., it is relatively easy to make an initial change in addictive behavior but much more difficult to maintain that change over time.  It can be argued that, without relapse, there would be no problem and no need for a special term like “addiction”; it is only because of the frequent occurrence of relapses that addictive behavior seems so hard to change and that a special term to describe this difficulty is needed. That being the case, any satisfactory theory of addiction must be able to provide a convincing account of relapse.

It is just about conceivable that relapses take the form of automatic behavior, without the conscious awareness or intention of the relapser; the existence of “absent-minded relapse” has been proposed but what evidence there is suggests that it is a relatively rare phenomenon (Catley et al., 2000).  It seems most unlikely, in my view, that relapse normally occurs in this way rather than by conscious decision and intentional action to resume the behavior, for whatever reason. It is relevant that, in the most influential model of the relapse process, by the late G. Alan Marlatt and his colleagues (e.g., Marlatt & Donovan, 2007), relapse involves cognitive processes in which the resumption of use is due to outcome expectancies for the effects of the substance and to the “rule violation effect.” Both these processes are available to conscious awareness and decision-making, albeit possibly subject to rationalisation and other forms of self-deception.

Perhaps the largest body of evidence against the strong version of compulsion is the veritable mountain of scientific reports showing that addictive behavior is operant behavior, i.e., it is shaped and maintained by its consequences. In other words, it is goal-directed, controlled and voluntary rather than automatic, involuntary behavior. A large number of laboratory experiments in the 1960s and 1970s showed clearly that drinking by the most chronic and severe alcohol addicts obeyed the same general laws that govern normal, goal-directed behavior of any kind (see Heather & Robertson, 1983, Chapter 3).

In the same vein, a demonstration of the ability of alternative reinforcers to modify cocaine use in experienced cocaine smokers was more recently carried out by Carl Hart and his colleagues (see Hart, 2013). Moreover, as shown by numerous controlled trials, the most effective way of changing addictive behavior for treatment purposes is contingency management (CM). Studies of CM programmes have reported remarkable success rates with physicians, airline pilots, and other professional groups but high rates of recovery have also been obtained among the less-privileged clientele.

Verdict on the strong version of compulsion

In view of the large amount of evidence of different kinds against it, the verdict must be that the strong version is of little utility in the explanation of addictive behavior. Indeed, it is surprising that it was ever taken seriously as the cornerstone of an influential theory of addiction. This may reflect an alarming disconnect between the results of experimental studies of nonhuman animals in laboratory settings and the manifestation of addictive behavior among humans in the real world. The evidence on humans overwhelmingly suggests that addictive behavior is voluntary, controlled behavior at the time it is carried out. 

On the other hand, there is no doubt that automatic processes do play a role in addictive behavior.  As we have seen, ongoing drug consumption among long-standing addicts is often characterized by deeply-ingrained habits that can be termed automatic — of the same kind perhaps as habitually switching off a light when one leaves a room. But, in terms of a theoretical account of addiction, this is a relatively trivial matter; once one is aware of them, habits of this kind can be changed without great difficulty.  More importantly, automatic processes result in cue-elicited craving, urges, attentional bias, automatic approach tendencies, implicit memory associations and cognition, and the inclusion of such processes is an indispensable part of the development of a modern, dual-process theory of addiction.  The mistake is to imagine that these automatic processes necessarily override consciously-aware and controlled processes in the production of addictive behavior.

“Irresistible” desires: the weak version of explanatory compulsion

The weak version of compulsion in addiction refers to the effects of powerful desires, urges, cravings, or impulses, etc. The proposal is that such powerfully motivating feelings or sensations cause addicts compulsively to carry out addictive behavior against their wills. To distinguish it from the non-motivational, automatic kind of compulsion in the strong version, we can call the weak version simply ‘”motivational compulsion.” It is dubbed “weak” because, although it may be necessary for addictive behavior to occur, it is not sufficient. It is not sufficient simply because addicts do occasionally resist such motivational forces and refrain from addictive behavior.

Since the beginnings of the disease concept in the early 19th century, temptations to consume substances have been described as “irresistible” or as “overpowering” any resistance the person might put up. An oft-cited passage in the addiction literature comes from Dr. Benjamin Rush, the “father” of American psychiatry, reporting the words of a patient: “Were a keg of rum in one corner of a room, and were a cannon constantly discharging balls between me and it, I could not refrain from passing before that cannon, in order to get at the rum” (Rush, 1812, p. 266). Other dramatic and lurid 19th-century anecdotes about “inebriety” could be quoted and similar tales continue to be part of the folk wisdom on addiction. Whether or not such anecdotes give an accurate and representative portrayal of the temptations associated with addictive behavior, we can ask how much the compulsion said to arise from them contributes to a scientific understanding.

Incentive-sensitization theory

A leading theory that uses this kind of compulsion to explain addictive behavior is Robinson and Berridge’s (1993) incentive-sensitization theory. This theory begins with the observation that a common property of addictive drugs is the ability to increase dopamine transmission in the mesolimbic dopamine system of the brain, which is known to be involved in reward and motivation for natural reinforcers. The principal function of this system is to attribute “incentive salience” to the mental representation of certain kinds of stimuli or events, where incentive salience is the process that makes stimuli attractive and sought-after. Repeated drug use leads to incremental neuro-adaptations in this system, making it increasingly sensitized to drugs and drug-associated stimuli. Sensitization of incentive salience is, therefore, the mechanism that is hypothesized to transform ordinary desires for drug experiences into drug craving; it is also responsible for relapse to drug use, even after extended periods of abstinence and the cessation of withdrawal.

Essential to understanding the role that compulsion plays in Robinson and Berridge’s theory is the distinction between “wanting” and “liking” drug experiences. Sensitization of incentive salience, described above, is called drug “wanting,” but the changes in neural systems that go with it can occur independently of changes in other neural systems — for example, those accompanying the pleasurable effects of drugs. The latter is called drug “liking.” In this way, incentive sensitization can produce compulsive drug-taking and -seeking even when the expectation of drug pleasure, or relief from the pain of withdrawal, is reduced and even when there are strong disincentives to drug use.

Note that incentive sensitization can occur outside the addict’s awareness and might, therefore, be thought to be an automatic process, similar to those in Everitt and Robbins’ theory. However, although often non-conscious, incentive sensitization is part of the person’s motivational make-up and is not the product of learning and habit-formation. Incentive sensitization theory is rightly seen as representing the weak version of compulsion in addiction.

Evidence for and against the weak version

As the most sophisticated and evidence-based illustration of the weak version of compulsion, incentive sensitization theory is appealing as an explanation of addiction in several ways. It seems to explain well some of the more puzzling aspects of addictive behaviour: the excessive preoccupation with the object of addiction; the fact that the behaviour is continued when the addict no longer gains much pleasure from it; and why relapse sometimes occurs after many years of abstinence. The idea of “drug wanting” can account for addicts’ reports of feeling driven to continue drug use without being able to give reasons or understanding their motivation for doing so, together with a sense of painful bewilderment. Most directly relevant is that fact that addicts do experience and report intense desires for the effects of drugs or activities. For all these reasons, some variant of the motivational basis for compulsion seems far more plausible as an explanation than the strong, automatic version and to offer greater chances of theoretical progress.

On the debit side, incentive sensitization theory, and the weak version of compulsion in general, is still embarrassed by the evidence briefly summarised above that addictive behavior is operant behavior,  particularly the demonstration of preferences by addicts for alternative reinforcers, including those that convey apparently quite small rewards, over drug consumption. All this evidence suggests that it is inaccurate to describe addictive behavior as compulsive in any sense at the time it is carried out.   That being the case, is it possible to rescue the weak sense of compulsion as a way of explaining addictive behavior?

One such possibility arises from a conjecture by Robert Noggle of Central Michigan University.  This begins with the idea of “ego depletion,” developed by Roy Baumeister and his colleagues, according to which “willpower” is a limited resource that becomes depleted by use; we can resist the temptation for a while but not indefinitely (see, e.g., Baumeister, 2003). With regard to addiction, craving and urges persist long enough to deplete the addict’s willpower, leading to relapse. Thus, drug cravings can be called irresistible but not in the obvious sense that they cannot be resisted at any one time; rather, the ability to resist is worn down and eventually fails.

What Noggle suggests is that the persistence of the desire to consume drugs is an important form of dysfunction in its own right and makes a significant contribution to what is called compulsivity. Thus, addiction involves dysfunction in a mechanism that normally prevents someone from being tempted to consume drugs “in situations where such consumption poses a grave, obvious, and imminent danger to things that they care about—things like their careers, their family, their health, their freedom, their self-respect” (Noggle,  2016, p. 218). When such desires occur in a non-addict, a “quashing mechanism” either blocks the formation of the motivation to consume or eliminates it soon after it emerges. In contrast to this, the addict’s motivation to consume drugs persists even when it becomes obvious that consumption will lead to disastrous consequences.

Noggle’s ideas are highly relevant to present considerations because they provide an account of compulsivity that concedes that addictive desires are not literally irresistible and that addictive behavior is voluntary at the time it is carried out, while at the same time postulating a mechanism by which relapse becomes more likely to occur over time. If validated by research, this sense of compulsion may have useful application to the understanding of addictive behavior.


A predictable answer to the question that forms the title of this article is that “it all depends what you mean by compulsive.” As noted, there are unfortunately many meanings of compulsion in the literature on addiction and, even more, unfortunately, authors often fail to specify exactly what they mean by the term. We also saw that there exists a large body of evidence of different kinds showing that, at the time it is carried out, addictive behaviour is voluntary behaviour and not compulsive in a straightforward sense of “against the will” of the individual concerned. One conclusion here is that any scientific understanding of addictive behavior that deserves to be taken seriously cannot ignore the evidence in question

Of the many meanings of compulsion to be found in the addiction literature, we were able to isolate three distinct meanings that I claim are those most worthy of close attention. We first encountered a “minimal” definition of compulsion which does qualify as a legitimate usage of the term in the ordinary language. However, this minimal definition seems to be merely a restatement of the central puzzle of addiction that some people repetitively behave in ways they know are bad for them. It does not get us any further in explaining this puzzle for scientific purposes just to label the behavior in question compulsive.

Of the two meanings of compulsion that do have potential explanatory value – the strong and weak versions – we saw first that the strong, “automaticity” version is flatly contradicted by several types of evidence. In addition to their involvement in ongoing consummatory behavior, automatic processes do undoubtedly play a role in addiction, including in relapse, but not in a simplistic way that views them as over-riding voluntary, controlled processes.

On the other hand, the weak, “motivational” version of compulsion does much better at surviving the test of empirical evidence on human addiction and, as represented by Robinson and Berridge’s (1993) theory, seems able to explain several mysterious features of addictive behavior. However, it is still embarrassed by evidence showing that addictive behavior is voluntary at the time it is carried out. A way in which a motivational theory of compulsion in addiction can be modified to take account of this evidence has been suggested by Noggle (2016) and this deserves serious research attention.

The present review of theory and evidence on compulsion in addiction has not considered one other essential aspect of its chosen topic – the implications of “compulsion” for the public understanding of addiction and for  the attempt to engender an informed and compassionate debate about how those we label “addicts” should be viewed and treated by society. This crucial agenda will have to await another occasion.

These findings are described in the article entitled, Is the concept of compulsion useful in the explanation or description of addictive behaviour and experience? recently published in the journal Addictive Behaviors ReportsThis work was conducted by Nick Heather from Northumbria University.


  1. Baumeister, R. (2003). Ego depletion and self-regulation failure: a resource model of self-control. Alcoholism: Clinical & Experimental Research, 27, 1-4.
  2. Catley, D., O’Connell, K., & Shiffman, S. (2000). Absentminded lapses during smoking cessation. Psychology of Addictive Behaviors, 14, 73-76.
  3. Everitt, B., & Robbins, T. (2005). Neural systems of reinforcement for drug addiction: from actions to habits to compulsion. Nature Neuroscience, 8, 1481-1489.
  4. Hart, C. (2013). High Price: Drugs, Neuroscience and Discovering Myself. London: Penguin.
  5. Heather, N. (2017a). Is the concept of compulsion useful in the explanation or description of addictive behaviour and experience? Addictive Behavior Reports, 6, 15-38.
  6. Heather, N. (2017b). Addiction as a form of akrasia. In N. Heather & G. Segal (Eds.), Addiction and Choice: Rethinking the Relationship (pp. 133-150). Oxford, UK: Oxford University Press.
  7. Heather, N., & Robertson, I. (1983). Controlled Drinking (Revised paperback ed.). London: Methuen.
  8. Hyman, S. (2007). The neurobiology of addiction: implications for voluntary control of behavior. American Journal of Bioethics, 7, 8-11.
  9. Marlatt, G., & Donovan, D. (Eds.). (2007). Relapse Prevention: Maintenance Strategies in the Treatment of Addictive Behaviors (2nd ed.). New York NY: Guilford Press.
  10. Neale, J. (2002). Drug Users in Society. Basingstoke: Palgrave.
  11. Noggle, R. (2016). Addiction, compulsion, and persistent temptation. Neuroethics, 9, 213-223.
  12. Robinson, T., & Berridge, K. (1993). The neural basis of drug craving: an incentive-sensitization theory of addiction. Brain Research Reviews, 18, 247-291.
  13. Rush, B. (1812). Medical Inquiries and Observations, Upon the Diseases of the Mind (1st edition). Philadelphia: Kimber & Richardson.
  14. Stephens, G., & Graham, G. (2009). An addictive lesson: A case study in psychiatry as cognitive neuroscience. In M. Broome & L. Bortolotti (Eds.), Psychiatry as Cognitive Neuroscience: Philosophical Perspectives (pp. 203-220 ). Oxford: Oxford University Press.