publisher colophon

It is argued that alcoholism, and substance addiction generally, is a disease. It is not of its nature chronic or progressive, although it is in serious cases. It is better viewed as a psychological disease than a neurological one. It is argued that each time an alcoholic takes a drink, this is the result of choice; however, in cases of serious affliction, such choices are compulsive and may be called ‘involuntary’ in that they are made against the subject’s will, motivated by an overwhelmingly powerful desire that he wishes he did not have and not to act on. Alternative accounts in terms social learning theory and behavioral economics are critiqued. The conception of alcoholism as a tripartite disease composed of a ‘physical allergy,’ a mental obsession, and a ‘spiritual malady’ is defended from a contemporary scientific point of view.


Alcoholism, addiction, disease, dopamine, spirituality, stoicism, Alcoholics Anonymous

Authoritative sources state that alcoholism is a disease. These include the American National Institute of Alcohol Abuse and Alcoholism (NIAAA), the American National Institute of Drug Abuse (NIDA), and the American Medical Association. The British Medical Association and the World Health Organization recognize what they call “alcohol dependence syndrome.” The American Medical Association tells us that alcoholism is “a primary, chronic disease with genetic, psychosocial, and environmental factors influencing its development and manifestations. The disease is often progressive and fatal. It is characterized by impaired control over drinking, preoccupation with the drug alcohol, use of alcohol despite adverse consequences, and distortions in thinking, most notably denial. Each of these symptoms may be continuous or periodic” (Morse and Flavin 1992, 1012). The NIDA describes alcoholism as a “chronic, relapsing brain disease characterized by compulsive alcohol seeking and use despite often devastating consequences” (Volkow 2010, 5).

However, many researchers strongly disagree. For example:

[t]he disease theory [of alcoholism] has outlived its usefulness.”

[W]hy do heavy drinkers persist in their behavior even when prudence, common sense, and moral duty call for restraint? That is the central question in debates about alcohol abuse. In the United States … the standard answer is to call the behavior a disease—‘alcoholism’—whose key symptom is a pattern of uncontrollable drinking. This myth, now widely advertised and widely accepted, is neither helpfully compassionate nor scientifically valid. It promotes false beliefs and inappropriate attitudes, as well as harmful, wasteful, and ineffective social policies.

“[E]very major tenet of the ‘disease’ view of addiction is refuted both by scientific research and by everyday observation.

To say something is a disease is not to say anything precise, deep, or very committal. Freedictionary. com gives us: “any deviation from or interruption of the normal structure or function of any body part, organ, or system that is manifested by a characteristic set of symptoms and signs and whose [End Page 297] etiology, pathology, and prognosis may be known or unknown”; the Oxford English Dictionary has “[a]n unhealthy condition of the body or mind” (‘health’ being defined in terms of well-being); and The Oxford Medical Dictionary has “a disorder with a specific cause and recognizable signs and symptoms; any bodily abnormality or failure to function properly, except that directly resulting from physical injury.” I shall take a disease to be an “unhealthy abnormality or disorder of any mental or physiological organ or system.” I shall argue that alcoholism is a disease in this sense.1

In Section 1, I specify three features that I take to be definitive of alcoholism, and then dissociate the disease view from a number of claims that are often associated with it, but should not be. In Section 2, I discuss two leading alternatives to the disease view. These offer accounts of problem drinking, roughly speaking, in terms of choice: Alcoholics drink the way they do because too often they choose to drink too much. The first formulates the idea in the terms of social learning theory, the second in terms of behavioral economics.2 I argue that these views are inadequate. In Section 3, I describe a promising cognitive-neurological theory of alcoholism and argue that, if it is correct, then alcoholism is pathological. In Section 4, I respond to an objection to the disease view.3

Section 1

The Defining Features of Alcoholism


I take it as relatively uncontroversial that many problem drinkers experience cravings. There are numerous discussions of this in the literature: See, for example, Lende (2008), Schroeder (2010), and Goldstein and Volkow (2011). Cravings tend to be triggered by external or internal cues associated with rewarding use (pleasure or relief), such as the sight of bottle of gin, a bout of anxiety, or any symptom of withdrawal. Often, when a relapse occurs, it begins with a craving (although relapses are often triggered by something other than a craving). Cravings usually persist during periods of abstinence. The mode pattern is that they gradually decrease over a period of months, and eventually disappear. But other patterns occur. Sometimes abstinence is almost free of cravings. Other times it is plagued by cravings every day. Sometimes a severe craving can occur after trouble-free months or years. Cravings are not totally irresistible: Alcoholic drinking is not like a reflex. However, cravings are powerful and often cause problem drinkers to act against what they would consider to be their own best interests.

Attempts to formulate a refined, scientific notion of a craving have not been particularly fruitful (West 2006). But the notion is clear enough: A craving is a felt need, strong desire, or urge. I take cravings to be conscious. They are, in a rough and ready way, detectable and measurable by subjects’ reports.

Heather and Robertson (1997, 182) dislike the use of ‘craving’ on the grounds that it is tautologous and unexplanatory. There might be an element of tautology in the notion of craving: A craving to drink is partly defined in terms of drinking. But that does not undermine its explanatory power. Cravings to drink are identifiable independently of actual drinking. For example, someone might say. ‘I have a craving for a drink.’ Cravings precede drinking. And they can occur without drinking. Cravings are among the things that can cause a person to drink. The fact that cravings are partly defined in terms of the effects they cause does not mean that they do not cause those effects.


Equally, I take it as uncontroversial that many problem drinkers tend to spend a lot of time thinking about drinking. They become preoccupied. If the problem progresses, the amount of time increases to the point of genuine obsession.

Impaired Control

Disease theorists tend to emphasize the extent to which alcoholics lack control over drinking. Choice theorists tend to emphasize the extent to which alcoholics can control their drinking. Views at both extremes are held. Thus, for example, Glen Hanson, representing the NIDA (in a lecture given at Brigham Young University, March 19, 20104) claimed that “[a]n addict can no more stop their behavior than a Parkinson’s patient can stop their [End Page 298] shaking.” On the other hand, Herbert Fingarette says “[a]lcoholics do not ‘lack control’ in the ordinary sense of those words” (1990, 50). Most theorists on both sides of the disease issue hold a more moderate view, somewhere in between, agreeing that alcoholics’ control over their drinking is impaired in some way or other.

Alcoholic drinking is nothing like the shaking of a Parkinson’s patient. Nor is it like sleepwalking or behavior guided by post-hypnotic suggestion. If you were to point a gun at an alcoholic’s loved one and threaten to shoot if the alcoholic drank, he would refrain from drinking. Alcoholic drinking is certainly intentional action rather than reflex or robotic motion.

Problem drinkers clearly exert some control over their drinking. Experiments have shown that people diagnosed as alcoholics will not always drink all the alcohol available to them; that the amount subjects drink varies with the cost of drink (the more work required to obtain alcohol, the less alcohol they drink); and that the amount subjects drink varies with associated benefits and deprivations (Heather and Robertson 1997, 81; Fingarette 1988, 39; Heyman 2009, 85–7).

Equally clearly, many problem drinkers have serious problems controlling their drinking. Problem drinkers who try to control their drinking often fail. They do not have enough control over it to drink within limits they set for themselves. When I stayed at a rehabilitation clinic, each client was given the task of specifying three occasions on which they had tried to control their drinking and failed. Of the twenty or so clients I observed, every one completed the task. Serious, concerted attempts to quit often fail. One study (Polich, Armor and Braiker 1981) found that approximately 90% of alcoholics are likely to experience at least one relapse over the 4-year period after treatment. Indeed, problem drinkers often try over and over again, and fail over and over again. This occurs short term (‘I will only drink two pints this evening’), long term (‘I will cut down two units per week over the next 6 months, then drink no more than thirty-five units a week for the rest of my life’), and everywhere in between.

Serious problem drinkers can have extremely severe problems controlling their drinking, so severe that they struggle against it for years, lose the struggle, and end up drinking themselves to death. Fingarette’s claim that they do not lack control in the ordinary sense of these words is as far wide of the mark as is the no-control-at-all view. ‘Lacking control’ is a description often used by ordinary problem drinkers, their friends and family, and their counselors, people who are not lacking in ability to use ordinary language, or in the grip of a misguided theory.

Problem drinkers’ alcohol seeking and use is sometimes compulsive in one standard psychiatric sense of the term, whereby a compulsion is a recurrent, unwanted, and distressing (ego-dystonic) urge to perform an act, or an inner drive that causes a person to perform actions against his or her will. In such cases, drinking can also be called ‘involuntary,’ in the sense of ‘acting against one’s will,’ although not in the senses of ‘unintentional’ or ‘not of one’s choice.’ (I return to this matter below.)

However, not all of those diagnosed with alcoholism have such severe control problems. The standard diagnostic interpretation of impaired control is that alcohol is often taken in larger amounts or over a longer period than was intended (e.g., Diagnostic and Statistical Manual of Mental Disorders [DSM-IV]). Your control is impaired if and only if you often drink more than you intend to. Or, if you are a willing alcoholic and do not care how much you drink, you would drink more than you intend to, were you to try to set a limit. Many problem drinkers’ control is impaired, in this sense (NIAAA 2011). The extent to which a subject’s control over drinking is impaired can be measured in various operational ways (Heather, Booth, and Luce 1998).

Hypothesis: Alcoholism is a disease characterized by cravings, preoccupation, and impaired control over drinking. You suffer from active alcoholism if and only if you are subject to all three symptoms. And you suffer from alcoholism if and only if certain stimuli, such as ingestion of alcohol, a belief that you have ingested alcohol, or a high level of stress would trigger active alcoholism. (My criteria pick out roughly the same people as DSM-V, under their diagnostics for alcohol use disorder. I note that, although DSM-IV did [End Page 299] not include cravings in its diagnostic criteria, it is reported that DSM-V does [National Institutes of Health 2013, 1]).

Someone suffering from active alcoholism, who happens to be sober, is in a significantly different state from an abstinent alcoholic whose abstinence is going well. The active alcoholic is much more sensitive to stimuli associated with drinking. For example, if he is used to drinking on Friday evening, then Friday evening will set off an urge to drink, and the chances are he will have the drink. The drink is likely to precipitate a binge, a period of drinking more than intended. The subject whose abstinence is going well will not be nearly so sensitive and will be able to walk past the pub on a Friday evening without any bother. A serious alcoholic in a culture in which drinking is not regimented may well be used to drinking in a very wide variety of circumstances, and will probably be sensitive to a very wide variety of triggers: Friday evenings, pubs, pleasant memories of drinking (‘euphoric recall’), mince pies, cricket matches, television programs, trains, boredom, annoyance, sadness, happiness, and the like. Any of these is likely to precipitate an urge to drink. The alcoholic whose abstinence is going well is not very sensitive to his triggers by association. He is, however, sensitive to drink and to high levels of stress. These are likely to trigger a drink, which is likely to begin a binge. After the binge, when he sobers up, he is likely to be back in active alcoholism.

If my hypothesis is right, then alcoholism is a discrete condition in that either you have it or you do not: “[T]he basic assumption is that the drinking behavior and alcohol-related experiences of people suffering from the disease of alcoholism are qualitatively distinct from the behavior and experiences of those not suffering from the disease” (Heather and Robertson 1997, 63) The alcoholic’s experience is qualitatively distinct from the non-alcoholic’s. He experiences cravings, preoccupation, and sometimes drinks more than he intends to. And although his behavior might not look different from that of a non-alcoholic’s, it would still be distinctive, by being, to a greater or lesser extent, out of his control.

Although it is a discrete condition, it comes in degrees. There is a smooth progression from those who have relatively few and relatively mild cravings, relatively little difficulty in controlling their drinking and only a mild preoccupation with alcohol, to those who suffer frequent severe cravings, have severe control problems, and a true obsession with drinking. But they all suffer from alcoholism.

Heather and Robertson (1997, 88) note that the curve representing how many people drink how much, in a given population, does not have a bump representing people who drink a lot, the alcoholics, nor a bump representing people who drink normal amounts, the non-alcoholics. (Contrast the graph for smoking, with numbers of people on the y axis and amounts smoked on the x axis: It has a sharp, high peak representing people who do not smoke at all, and a small bump with a top representing those who smoke about twenty cigarettes a day.) The alcohol consumption graph gives the lie to a crude disease model that would predict a large group of non-alcoholic, moderate drinkers and a small group of pathological alcoholics who drink lots. My disease model does not make that prediction.

Alcoholism and heavy drinking are different. In principle you could be a heavy drinker without being an alcoholic. That category may be large: One study found that “even in people who have 5 or more drinks a day (the equivalent of a bottle of wine) the rate of developing dependence is less than 7 percent per year” (NIAAA 2011, 1). Equally, in principle, you could be a serious alcoholic and drink nothing. The extension of that category is large: The abstinent alcoholics. And, in principle you could be a serious alcoholic while drinking only modest amounts. I expect that the extension of that category is very small. People with serious alcoholism find that drinking modest amounts causes extensive and highly distressing cravings and preoccupation, which is why they typically prefer to quit. But, more realistically, you could be a mild alcoholic and drink modestly, less, even, than a heavy-drinking non-alcoholic.

I now consider some features of the condition that have been associated with the thesis that alcoholism [End Page 300] is a disease by its proponents or opponents, but that should be firmly dissociated from it.

Alcoholism Is Chronic and Irreversible

For an alcoholic, even one who has abstained for a long period, one or a few drinks are likely to precipitate a relapse. This seems not to be not true. About two thirds of those diagnosed with alcoholism return to normal drinking. However about one third do not, and fit the classic conception of ‘once an alcoholic, always an alcoholic’ (Willenbring et al. 2009).

One might claim that those problem drinkers who later revert to controlled drinking are not real alcoholics. These are people who drink too much for periods and exhibit signs similar to those of alcoholism, but who are not real alcoholics. Or one might distinguish two types of alcoholism: Ephemeral and irreversible. Maybe there is some theoretically interesting neurological or psychological property or cluster of properties that would justify such a distinction. But, as yet, such has not been found. It looks like the same condition occurring in ephemeral and irreversible forms, just as it can occur in mild or severe forms.

If we take alcoholism to consist in cravings, preoccupation, and impaired control, there is no apparent reason to suppose that is always irreversible. It seems likely that many of those diagnosed with alcoholism under standard diagnostic criteria, and who returned to controlled drinking, were alcoholics so-understood.

I conclude that alcoholism is not by its nature irreversible, although it is irreversible in a very substantial number of cases.

Alcoholism Is Progressive

E. Morton Jellenik (1952), a leading early proponent of the disease theory, attempted to detail a course of progression of the disease, specifying a series of distinctive phases. There is some evidence that in cases where the condition does worsen over time, the development follows a pattern much like the one he outlined (Kumar et al. 2005). Some of the features of Jellenik’s progression are distinctive to alcoholism (e.g., blackouts, tremors). It is important to distinguish features of the disease—addiction—from features of use of a particular substance or behavior. I suggest that increasing cravings, preoccupation, and loss of control is the basic general course. The developmental pattern of more specific symptoms depends largely on the specific type of addiction, as well as extraneous factors, such as the social environment of the subject. What is not true is that the majority of subjects at a given stage progress to the subsequent stages (Heather and Robertson 1997, 78; Vaillant 2003, 1049–9).

There exist lifelong alcohol-dependent subjects who have been drinking too much on a regular basis for decades, but who have always drunk about the same amount (say 75 units a week, just over a bottle of wine a day). There are also lifelong drinkers whose drinking gradually grew over time from excessive amounts in their teens to say 350 units a week (two bottles of gin a day) in later life. Thus, alcoholism may progress or it may not. It has not progressed in most problem drinkers. But it has progressed in many.

Alcoholism Is a Brain Disease

Alcoholism might be a brain disease. But more likely it is a psychological disease. Minds are not immaterial entities that can cause a subject’s hand to reach out for a drink. The subject’s brain causes the movement. There is no room for an immaterial entity to affect the firing patterns of neurons that send signals to the arm. Mental processes are brain processes. But that does not mean that mental conditions and properties are neural conditions and properties. The mind stands to the brain as the programs on a computer stand to its physical base. A program is not an immaterial entity that affects the workings of silicon chips. It is implemented by the arrangement of those chips. It is their pattern. A single program can be implemented in countless different kinds of hardware. The rules of the program would be the same, but the physical structures underlying it would be different.

If a computer is not working properly, this may be owing to a hardware problem or a software bug. If there is a bug in a chess program, this could be realized by different physical structures in different machines. A complete theory of what’s wrong would then have to be formulated in terms [End Page 301] of software. No theory framed in purely physical terms would be complete. Alcoholism might be or be like a software bug, rather than a hardware problem. It would then be misleading to call it a brain disease.

There has been a lot of successful neurological research on alcoholism, detailing differences between an alcoholic’s brain and a non-alcoholic’s brain. The relevant features of the alcoholic’s brain might be those implementing a software bug (features of the physical structures encoding a program), or they might be those constituting a mere hardware problem. Alternatively, if alcoholism is a mental but non-cognitive problem, say one to do with feelings, then the relevant neural features might those underlying the relevant properties of feelings. In fact, as far as I know, all existing theories of alcoholism treat it as psychological, rather than neurological.5

I offer one such below.

Alcoholics Are Physically Dependent on Alcohol

Physical dependence should not be counted as an essential feature of alcoholism. It is just an occasional result of excessive drinking. One can be physically dependent on something without being addicted to it. Coffee, for example, gives rise to physical dependence, but not to uncontrollability, preoccupation, or serious cravings (Wilson and Kuhn 2005). And most alcoholics are not physically dependent on alcohol most of the time.

Alcoholics Are Not Responsible for Their Drinking or Its Consequences

Alcoholics are no more responsible legally or morally for their drinking and its consequences than epileptics are responsible for the consequences of their movements during seizures.

(Attributed to the disease theory by Fingarette 1990)

The disease theory arose in part in opposition the view that alcoholism is a moral failing. The issue is still sometimes framed in terms of the dichotomy: Disease or moral failing. And this has generated much heated debate. If it is a disease, the idea goes, then the alcoholic is not responsible for his drinking. If it is not, then he is. But this is a false dichotomy.

The notion of responsibility is so fraught with philosophical difficulty that it might be best not to lumber proponents of a disease theory with views about it. It would not follow from the fact that it is a disease that sufferers are not responsible for their actions. If it is a disease, it is one that affects the mind, and the mind is the source of actions. The question would remain whether the mind’s being affected by the disease absolves the sufferer from some or all responsibility. Moreover, someone aware of the risks of becoming an alcoholic might yet drink wantonly and become addicted. Perhaps he could be regarded as responsible for his alcoholism and the excessive drinking it causes. Conversely, it would not follow from the fact that alcoholism is not a disease that problem drinkers are fully responsible for their excessive drinking. Suppose that there is no such thing as alcoholism. It would remain true that there are features of consumption that can lead someone down a path from moderate drinking to repeated excessive drinking. Someone unaware of these features could be led unwittingly down the path. Drinking can cause cravings, tolerance, reverse-tolerance (desires increasing over time; Robinson & Berridge 1993), and impairments to thinking and decision making. Someone who does not know of these matters and ends up drinking too much might not be held fully responsible for their excessive drinking. I return to these matters below.

Alcoholism Is Genetically Determined

A claim of this sort is sometimes attributed to disease theories. In fact, twin studies indicate that genetic factors account for 50% to 60% of the risk for alcoholism (McGue 1999). This is interesting, but irrelevant to the present issue.

Alcoholics Cannot Recover Without Treatment or a Recovery Program

Many problem drinkers have spent years trying to control their intake or quit, failed to do so and died as a result. They could have done with help. Still the majority of diagnosed alcoholics who recover, do so without treatment or a recovery program (Pickard 2012; NIAAA 2011). [End Page 302]

I turn now to attempts to account for problem drinking that do not posit a disease of alcoholism.

Section 2

Social Learning Theory

Heather and Robertson (1997) offer an account of problem drinking framed in terms of social learning theory. Their account is divided into two parts, discussing first conditioning, then higher cognitive processes. I summarize the account, then respond.


If someone drinks in a wide variety of different circumstances, then the number and pervasiveness of cues for drinking will be greater than they are for someone whose drinking is relatively confined.

Some drinkers learn that drink relieves the symptoms of withdrawal, and some do not. As the subject becomes more used to using drink to bring relief, the link becomes reinforced and the problem increases.

The rewards that drinking brings are immediate. Humans have a general tendency to overvalue immediate rewards. As we mature and learn to act more rationally, we learn to overcome this tendency. However, this maturity of decision making is overcome by alcohol itself: A drunk person tends not to think too far ahead.

Someone in a socially impoverished condition has few rewarding activities available other than drinking. So he will tend to drink more than those more fortunate. Further, a problem drinker tends to lose interest in hobbies and to become alienated from friends and family. So heavy drinking is maintained owing to lack of rewarding alternatives.

Higher Cognitive Processes

Some of our behavior patterns are acquired through modeling—copying others. So, for example, a child growing up in a household of problem drinkers may model this behavior and become a drunk. Similarly, teenagers and adults model peer group behavior, so if someone associates with problem drinkers, he may conform.

In general, our behavior is controlled by various self-regulation processes. These do not always function properly, and failures of self-regulation can lead to problem drinking. Thus, for example, problem drinkers often fail correctly to monitor their intake and underestimate the amount they are drinking. Further, once a pattern of problem drinking has been set up, the subject’s capacity to set and maintain healthy standards for himself becomes diminished. The drinker then ceases to spend enough time and energy on healthy pursuits other than drinking, and devotes more time to his habit.

There is evidence suggesting that a major factor contributing to heavy drinking is high self-awareness: People may drink to escape the stresses and anxieties that arise from it.

Heavy drinking may be linked with aspects of self-image. For example, one study showed that some young Scottish subjects were to some extent motivated to drink to avoid the stigma associated with abstinence. Generally, if one sees drinking as being or reflecting part of the kind of person one is, one will be motivated to drink.

Heavy drinking may be linked with outcome expectations: Owing to cultural or other influences, a heavy drinker may lack faith in his ability to control his drinking and he may underestimate the beneficial effects of so doing. One aspect of this may be learned helplessness: Once a drinker has experienced control problems, he may come to believe that he himself cannot affect what happens in relation to his drinking and a syndrome of passivity, low self-esteem, and difficulty with relearning is likely to come into play.

A problem drinker may attribute his heavy drinking to global, stable, and internally located factors, and this may cause pessimism, passivity, and listlessness. This may happen if the subject believes that he is suffering from an irreversible disease, as the classical disease theorist maintains.

A problem drinker’s tendency to relapse may be caused by the Abstinence Violation Effect (Marlatt and Gordon 1985). This involves two factors: A slip from abstinence is attributed to one’s own personal weakness, and a disparity between one’s self-image as an abstainer and one’s current behavior is observed. The self-image alters, the subject sees himself as a heavy drinker, and learned helplessness kicks in. [End Page 303]

The Abstinence Violation Effect involves cognitive dissonance, the tendency for psychic conflict to cause warped beliefs (Festinger 1975): When someone finds himself behaving in a way that is inconsistent with his prior beliefs, he may adjust those beliefs, bringing them into line with the behavior. Cognitive dissonance may preserve heavy drinking habits in various ways: For example, a subject may come to believe that drinking is a great thing to do because he spends so much of his time doing it.

Problem drinkers may, in various different ways, be poor at problem solving and coping. The subject may be impulsive. He may have excessively high expectations of himself, and when he lets himself down, turn to drink to escape the perceived failure. He may overgeneralize, he may catastrophize, and he may be prone to jumping to conclusions. When drunk or in early recovery, his cognitive capacities are impaired, his memory is poor, his concentration is low, and his thinking is cloudy. In general, his capacity to detach himself from his problems, define those problems clearly, think of possible solutions, make decisions, and assess the effectiveness of decision outcomes is impaired.

Put enough of these factors together and problem drinking results. There is no need to posit a specific disease of alcoholism to account for the phenomena.


Probably some these factors causally contribute to all, or nearly all, problem drinking, and all of the factors contribute to some problem drinking. However, I do not think that they can explain the true nature of the phenomenon. I illustrate with a case history of an acquaintance of mine. To protect anonymity, I shall call him ‘Alf.’ Heather and Robertson’s account does not begin to explain people like Alf or many other high-functioning problem drinkers like him who simply do not drink excessively for the sorts of reasons Heather and Robinson describe.

Alf began to show definite signs of problem drinking in his early 20s while studying at university. Although his early role models drank only moderately, at this point he began to spend his time with heavy drinkers. During term time he would go to the pub with his friends nearly every night. During holidays he would socialize with school friends, all of whom drank a lot. It is not clear how much he was modeling his behavior on that of his peers. He was the prime mover toward drink in his social groups, the one who typically said ‘let’s go to the pub’ and ‘let’s get a few bottles to take home.’ He often drank a little more than his friends while socializing with them. And when he was not in the pub in the evenings but studying in his room, he would have with him a quarter bottle of vodka, which he drank. This behavior was not modeled on anyone else’s. However, there is little doubt that his heavy-drinking social milieu was a causal factor in his development of alcoholism.6

His drinking gradually increased over the years, and by age 40, he was drinking very excessively on a regular basis. He drank at least ten units a day. At weekends and on holidays he drank a great deal more. He spent much of his working day looking forward to his evening drinks. He often dwelt on forthcoming boozy weekends and holidays. When hung over or deprived of drink, he would suffer cravings.

When in his 40s’, he was financially well-off. He had a good job. He was highly functional at work, well able to solve problems and cope with difficulties arising in his work environment. He was a highly self-aware individual, but this did not bring any great stress. He was not generally impulsive. When sober, his thinking in relation to matters other than drink was clear, long-sighted, and detached. He played bridge regularly. He loved food, and devoted time to cooking. He had a rich social life with non-alcoholic friends. His capacity to set and maintain healthy standards for himself was good, except when it came to alcohol. Far from believing that his drinking was the result of permanent character traits, he did not believe that alcoholism was a disease and thought that he would be able to control his drinking if he tried harder. Heather and Robertson’s collection do not begin to account for Alf’s problem drinking. He simply does not fit their profile of the problem drinker. There are many highly functional, serious alcoholics who are very like Alf. [End Page 304]

I recount more of Alf’s history to help illustrate the extreme extent of cognitive dissonance and alcoholic denial and distortions of thinking, and how excessive faith in a subject’s control over drink can be a greater problem than the his lack of faith in it, the Abstinence Violation Effect. This will also help to bring out the pathological nature of alcoholism.

At about 44 years of age, Alf was afflicted by some minor health problems and went to his doctor. The doctor enquired about his alcohol intake, and Alf confessed to drinking about three quarters of the amount he in fact drank. The doctor voiced serious concern and told him that he must cut down substantially and quickly; otherwise, he would inevitably suffer serious health problems and might go on to die from alcoholism. Alf saw a counselor for 6 months. The counselor deployed a combination of cognitive–behavioral therapy and motivational interviewing. After 6 months Alf was drinking fewer than 50 units per week. He planned to continue this regime for the rest of his life. This plan failed completely, and a few months after the end of his counseling he was drinking more than ever before. His drinking was out of control. He continued to try to control his drinking for about 4 years. During this period, he gradually drank more and more, until he was on about 200 units a week. At most times during this period, he believed that he would very shortly start cutting down and return to a regime of controlled drinking.

At the of this 4-year period, he had in his mind convincing reasons to believe that he could not control his drinking: He had been trying his damnedest for years and got precisely nowhere. He also had in his mind convincing reasons to believe that the consequences of continued drinking were very likely to include serious health problems, as well as relationship problems and loss of his job. Yet all of this counted for nothing. He continued to drink. Moreover, he had even ceased to enjoy drinking. He no longer liked getting drunk. Yet he continued to drink.

Alf’s control problems took a variety of forms, falling under the general heading of what the American Medical Association referred to above as ‘distortions in thinking, most notably denial.’

This sort of denial is part of the ‘insanity’ referred to in step two of the AA Twelve Step program (of which more below). The member ‘comes to believe that a power greater than himself can restore him to sanity.’ One can only believe that one can be restored to sanity if one believes that one is insane. AA currently has a membership of around 2,000,000. Many of them have done step two. So, many alcoholics believe, based on their own experiences, that they themselves are afflicted by this kind of insanity. (The step also requires that one has once been sane, which requirement may not be appropriate for all problem drinkers.) As part of the AA program of recovery, every member has to come to believe that he is insane (unless he already believes it). Alf has frequently had a drink, or two, or three, in the belief that he would stop for the evening after the allotted amount. He has frequently had the allotted amount, and thought that he would have just one more and then stop for the night. Some days he just let himself go and drank freely, believing that he would cut down significantly the next day. Once he switched from vodka and wine, his favored drinks, to beer, in the belief that he would then control his intake.

Each time his initial belief proved completely wrong. But the beliefs kept kicking in justifying those one, two, or three drinks, that just one more drink, that just one more day of boozing ad lib, that period of drinking only beer. These recurrent, insane beliefs cannot be accounted for in terms of normal learning, because they fly in its face.

The insane beliefs are beliefs to the effect that one will stop or cut down one’s intake after having had a certain amount to drink, and are formed in the face of very substantial evidence to the contrary. They can be extremely irrational. AA’s choice of the word ‘insanity’ is appropriate and not hyperbolic.

Such beliefs are a significant factor in relapse. After a detox and some therapy, Alf remained abstinent for a few weeks. Then one day he suffered a flare-up of anxiety. This was followed by an insane thought to the effect that he could have half a bottle of vodka to bring temporary relief, then quit again. He had the half bottle. The next day he woke up feeling more anxious, suffered a craving, and bought another half bottle. A couple [End Page 305] of hours after he had finished it, he was still anxious. He noted that he had nothing to do for the rest of the day, and had two days free after that. So, he reasoned, he could drink another bottle that day and he would have plenty of time to sober up and return to abstinence afterward. The return to abstinence did not follow as planned, and he stayed drunk until the next detox, 2 weeks later. In this case Alf managed to forget that he was an alcoholic.

Alcoholics Anonymous (AA) describes this sort of phenomenon (p. 24): “[M]ost alcoholics have lost the power of choice in drink. Our so-called will power becomes practically nonexistent. We are unable at certain times to bring into consciousness with sufficient force the memory of the suffering and humiliation of even a week or a month ago.” This is not an endorsement of the radical, no-choice doctrine discussed. ‘Powerlessness’ in AA refers to the various ways in which alcoholic mentality shatters rational thought and allows the urge to drink totally to dominate the alcoholic’s life. It refers to the fact that, in active alcoholism, the alcoholic always makes the wrong choice, against his own better judgments, against his own will. It does not mean that the alcoholic does not drink because he chooses to.

When a severe alcoholic’s memory is functioning, absurd rationalizations for drinking sometimes kick in. For example, one alcoholic told me: “[I] drank 2 1/2 weeks after an esophageal hemorrhage. Couldn’t bring that into my mind with sufficient force because I had already convinced myself that scar tissue is stronger than normal tissue (they cauterized the wound). Even brought photos home they took with the endoscope to remind myself. I now had a super esophagus and was good to go, so to speak.” This took place after his doctor had told him that if were to continue drinking, he would die.

Alf also sometimes suffered from the cognitive impairment, mentioned by Heather and Robertson, of underestimating the amount he had in fact drunk on a particular occasion. This cannot be accounted for in terms of the general inaccuracy of the drunk’s impaired ability to estimate or keep track of quantities. That would account for random under- and overestimations. But it fails to account for systematic underestimation specific to alcohol consumption.

Cognitive impairments also afflicted Alf in relation to the likely consequences of drinking. His doctor had told him that serious health problems, possibly even death, were inevitable if he keeps drinking. He trusted his doctor, and he believed on other grounds that chronic alcoholism is very dangerous. But the bleak prospect in his case seemed somehow unreal or very distant. Alf also often underestimated the extent of the negative value of the consequences, should they occur. Yes, he might go on to suffer and die. But somehow that did not seem to matter too much.7 As he crept downstairs while his wife was asleep and treated himself to a swift, large shot of vodka, his focus was on the fact that his wife and his doctor would not know about this drink. He reasoned as if their knowledge of the drink was a crucial factor in the value of the outcome of his action. Loss of health and livelihood did not figure in the calculation at all.

The recurrent insane beliefs, the miscalculating of drinks drunk, and the underestimation of the likelihood and gravity of the consequences of drinking are striking cognitive impairments that lead to extreme irrationality. There are people who carry on drinking after they have been hospitalized for life-threatening, drink-related reasons and told that they would die if they carried on. There are people who have destroyed their own liver, then gone on to destroy another one after a transplant. All of these people had ample reasons to believe that these were likely outcomes of continued drinking. But these reasons did not have the impact on their cognitive economies that they would in any normal person. The alcoholic’s extreme cognitive dissonance leads to this distinctive kind of insanity, which in turn leads so many to a pattern of intentional drinking behavior by which they (unintentionally) drink themselves to death.

I move on now to another attempt to account for problem drinking in terms of nonpathological psychological functioning.

Behavioral Economics

Gene Heyman (2009) opposes the disease view of alcoholism and offers in its place a behavioral-economics model based on the idea that addiction [End Page 306] result from nonpathological, although sometimes disordered and irrational, choice making.

Each time someone faces a choice between having a drink and doing something else, he needs to evaluate the options. Problem drinking consists in a very large number of particular events of choosing to have one drink rather than doing something else. Thus, problem drinking might result from features of such particular choices. Almost no one chooses to be an alcoholic. However, tactical errors in choice making may still account for problem drinking.

Prima facie, this is not too implausible. For someone who particularly likes drinking, or someone suffering from anxiety and who is keen for relief, the value of the next drink may be high. The downside of the drink may be minimal. This one drink will not lead to a hangover, and its contribution to a possible loss of job, family, and health is negligible. So it may seem perfectly reasonable to have the next drink. The problem is that if one repeatedly makes this choice, everything goes pear-shaped: The hangover occurs next day, and eventually the big losses kick in. The value of having a drink now may be greater than the value of the alternative, whereas the value of a course of drinking, over time, may be lower than the value of an alternative course. Local versus global choice-making strategies might thus account for the difference between problem drinking and sobriety.

Heyman argues that there is a particular feature of drinking and other addictive behaviors that predicts exactly that. When you have a drink, this affects the values of both later drinks and later alternatives. The drink increases tolerance, so the value of a future drink goes down. But the drink also decreases the value of future alternatives. The way things work out is that, on each occasion, the value of the next drink is greater than the value of the alternative. However, the overall value of a series of drinks is lower than the overall value of a series of alternatives.

Why does drinking lower the value of future non-drinking? According to Heyman (2009, 124):

The DSM account of addiction. … states that the essential feature of addiction is the continued use of drugs despite “significant substance-related problems.” These problems necessarily occur in the course of non-drug activities. For example, withdrawal interferes with doing well at work and intoxication torpedoes conventional interactions.

In addition to withdrawal and intoxication, there are all kinds of long-term effects that kick in. Suppose, for example, that I face the choice of having a drink or going for a walk with my wife. The values of both are high and I choose the drink. Now consider what happens when I face the same choice in a year’s time. The value of the drink has gone down, because of tolerance. But the value of the alternative has gone down as well, because I have been drinking all year. I am tired and unhealthy and do not fancy walking. Moreover, I no longer get on too well my wife. The drink wins out again.

Heyman (2009, 159) provides some evidence that addicts tend to adopt local rather than global choice-making strategies more than non-addicts do, when faced with choices that have nothing to do with their substance. So this local/global distinction may well be a factor in addiction.


Heyman’s characterization of addiction is not quite right. It is true that the DSM and other sources often characterize addiction that way. But the characterization fails to capture the essence of the condition. Addicts exhibit thoroughly addictive behaviors long before these behaviors get them into any real trouble. A characterization like mine, or in terms of standard diagnostic criteria like those of DSM-IV, much better capture the nature of the problem. Thus, the chief explanandum is not, as Heyman suggests, that addicts often continue to drink or drug themselves into difficulties. Setting this up as the problem gives Heyman’s suggestion that a failure to take long-term considerations into account is the root of an addict’s predicament some plausibility. However, the real problem is lack of control. And Heyman’s theory does not address that. The lack of control is explained by the extreme persistence, frequency, and power of the urge to drink. This causes cognitive dissonance and denial. These cause control problems. The control problems lead to the life problems. [End Page 307]

What Heyman’s account chiefly misses is the extraordinarily high value that an alcoholic sometimes assigns to drink from a local point of view. For example, it sometimes happens that an alcoholic who normally values honesty and respect for the law, and who is afraid of being caught, will steal alcohol. It sometimes happens that an alcoholic will choose to drink before collecting his children from school in a car, knowing that this risks social opprobrium, loss of driving license, and death of the children. I met one alcoholic who tried to force herself to drink even though she had damaged her throat, and drinking caused considerable pain. Disulfiram is a medicine that causes very unpleasant symptoms, such as dizziness, nausea, and vomiting, about 10 minutes after the ingestion of alcohol. About 50% of alcoholics with disulfiram implants still drink: “[T]hey will get drunk, but they will also get very sick and sometimes even collapse,” says Dr. Kris Zakrzewski, a private consultant who has fitted hundreds of patients with disulfiram implants (The Guardian, Thursday July 17, 2003). I met one alcoholic who died from alcohol poisoning and was brought back to life. He continued to drink knowing that the drink might kill him. One might say that in a case like that, the alcoholic must have subconsciously wanted to die. Perhaps. But the obvious generalization here is that, ceteris paribus and to a rough approximation, a severe alcoholic tends to choose to drink over any other option, except when the downside of drinking is both very grave and very salient.

Drinking can be a matter of choice, while being (in a sense) involuntary and compulsive. Sometimes a drinker who is not subject to cognitive impairments like Alf’s, and who is aware that his drinking threatens his loves, livelihood, liver, and life, and who does not even expect to enjoy his next drink, will still choose to take it. He chooses to take it because that is what he most wants to do. His choice is governed by a particularly strong desire to drink. This desire is one that he very much does not wish to have, nor to act upon. He desires not to drink and this latter desire is the one that is in line with his considered judgment about the best course of action and with his other desires. It is the one that he identifies with and that he would like to act upon. Hence it is the one that represents his will.8 But, for all that, the desire to drink is stronger than the desire not to. So he drinks and he drinks against his will, hence compulsively and involuntarily. In the extreme cases discussed, the choice making seems to be highly abnormal and disordered, and hence pathological.

In support of his view that addictive behavior is best understood in terms of voluntary choice, Heyman (2009, vii) writes that “everyone, including those who are called addicts, stops using drugs when the costs of continuing become too great.” He provides no evidence in support of this claim. I suggest that the millions of alcoholics and drug addicts who have died trying to control their use and failing, give the lie to his assertion.

The next section looks at the idea of pathological choice making from the point of view of cognitive and affective neuroscience.

Section 3

Addiction as Aberrant ‘Wanting’

I now consider a cognitive-neurological account of alcoholism, and addiction generally, that is billed as a disease theory by its proponents, and I defend that billing against an objection.

The account centers on a feature of the dopa-mine system in the brain. The dopamine system is a reward system. When a human performs an action that brings a greater reward (more pleasure) than expected, dopamine is released, the system’s state changes, and the motivation to repeat the action given the same stimulus is increased.

If the reward is less than expected, the reverse happens, and the motivation to do the same thing again is decreased.

All addictive drugs and alcohol affect the dopa-mine system by more-or-less direct physiological processes. This means that the system acts as if the act of ingestion has had a certain positive value, irrespective of how rewarding the action actually is. Alcohol and drugs are thus treated routinely as though they are more rewarding than they in fact are. To use the vogue term for this, they ‘hijack’ the system. It is possible that the system is also sensitive to food, drink, and sex in the same way, although this has not been conclusively established [End Page 308] (but see Rada, Avena, and Hoebel 2005 and Avena, Rada, and Hoebel 2007, on sugar). Therefore, it is plausible to suppose that the system evolved to encourage activities promoting life and reproduction.

It is likely that this system is oversensitive in addicts. During the process of becoming addicted, the system routinely treats each ingestion as more rewarding than expected, thus increasing the motivation to ingest in the future: This accounts for reverse tolerance. Thus, the addict’s system consistently treats ingestion as considerably more rewarding than it in fact is and the motivation to repeat actions of ingesting alcohol or drugs is highly enhanced (Berridge and Robinson 2003; Hyman 2005).

Berridge and Robinson (2003) call the reward-value assigned by the system to a stimulus ‘incentive salience’ and the motivation ‘wanting.’ Incentive salience causes stimuli to be salient and attractive; it motivates ingestion and it affects higher cognitive systems so that the subject can plan to get more alcohol or drugs. ‘Wanting’ is itself unconscious, hence unmodulated by conscious reasoning. But it can lead to conscious cravings.

The theory is eminently plausible. It explains why addicts tend to use even when they do not expect to enjoy it. It explains the irrationality of the drive toward using: Using is treated as far more valuable than it in fact is. It explains why addicts sometimes use against their will, or against what Berridge and Robinson (2011) call their ‘cognitive desires,’ desires that arise as a result of reflective judgment about what would be best.9

Moreover, it promises to explain cross-addiction: Sensitivity to one addictive substance has the same physical basis as sensitivity to the others (Nestler 2005).

There is, however, a significant feature of relapse still to be explained by the theory. The brain changes induced by drugs are long-lasting, so an abstinent addict remains vulnerable to relapse for a long time. And unintended binges may be explained: An oversensitized system is particularly responsive to ingestion of alcohol, so when the alcoholic has one drink, the desire to drink gets a big boost and a binge results. But that is not the whole story. Relapse does not consist in a binge, but a move from abstinence to active alcoholism. Unlike the initial onset of addiction, which is a slow process of gradually increasing control problems, the change from relatively trouble-free abstinence to active addiction can be sudden and extreme. So perhaps there is some kind of alcohol-sensitive switch in the addict’s brain that precipitates active addiction. But such a switch has not yet been found. Or perhaps what goes wrong has to do with a breakdown of feeling processing, which precipitates the first drink and results in persistent overstimulation of the dopamine system. Either way, the phenomenon remains unexplained.

Nor does it by itself fully explain cravings (there is a gap between unconscious urges and the conscious feelings they sometimes cause), nor alcoholic denial and distortions of thought.

However, it does go a long way toward explaining control problems and preoccupation (we often get preoccupied when we want something a lot) and, overall, it gets enough right to warrant tentative endorsement.

If the incentive-sensitization theory is true, does this entail that the addict’s drive toward using is genuinely pathological? Foddy (2010) argues not. It is reasonable to regard the abnormality attributed to addicts, aberrant ‘wanting,’ as involving the content of a desire. But, Foddy says, the contents of our mental states cannot be candidates for diseases:

[i]f we wish to continue using the concept of disease in the way that we currently do, to determine who receives subsidized treatment, who is released from their workplace obligations, and so on—that is, if we want the concept to do any work at all—then we cannot use a definition of disease that is overinclusive. If we want to avoid overinclusiveness then we cannot appeal to the contents of desires that are acquired in the normal way, when we determine whether or not someone has a disease.

(2010, 85–6)

I object. The desires are not acquired in a normal way: Most people who indulge in addictive substances do not develop oversensitive dopa-mine systems. And having a representation with a certain content in one’s reward system can be unhealthy, if it is abnormally powerful and persistent. (Compare methyphobia, which is defined as a pathological fear of alcohol.) If I persistently [End Page 309] really, really want to drink, and this desire trumps significant countervailing considerations, then it can cause me to lose my loves, my livelihood, my liver, and my life. If it withstands all my efforts to alter and resist it, then I should receive subsidized treatment (and, in Britain, I would) and I should be released from workplace obligations (and in Britain, I would).

The addict’s reward system is functioning abnormally and in an unhealthy way. From a computational point of view, it is also disordered, in that it implements a software bug: It systematically misrepresents the value of ingestion (Berridge 2012).

Section 4

An Objection to the Disease View

Heyman (2009, 169–71) argues that alcoholism is not a disease on the grounds that AA works, but the AA program does not look like medicine or treatment: “[A]lthough it is reasonable to suppose that many chronic disease sufferers would benefit from socializing with others who suffer from similar ailments, it is not reasonable to suppose that such meetings would also prove effective treatments.”


I concur that AA works. Various studies have been designed to show something about the relationship between membership of AA and recovery that take some account of meeting attendance. Summarizing the studies that have been done over the last 20 years, Kelly, Hoeppner et al. (2011, 290) report that the “body of work has indicated that AA confers short- and long-term therapeutic benefit on a par with professional interventions.” Kaskutas (2009), after surveying recent studies, reports that rates of abstinence are about twice as high among those who attend AA than among untreated subjects, that abstinence rates vary closely with level of meeting attendance and that prior attendance of AA predicts subsequent recovery. Among those who join AA about 40% to 50% who stay and attend meetings are abstinent after a year or two, with very good prospects of long-term sobriety (Moos and Moos 2006; Vaillant 1983).

In fact, meeting attendance (‘socializing with others’) is but an incidental part of AA’s program of recovery. The main purpose of meetings is simply to attract newcomers. The therapeutic aspects of meetings are discussed in Teresi and Haroutouian (2011, 247). Kelly, Stout, and Magill (2011) present more data on these therapeutic effects of the social aspects of AA. And, of course, meetings serve as a constant reminder of the risk and dangers of denial and help to combat alcoholic insanity. (This is part of how AA as a power greater than the alcoholic can restore him to sanity, as step two describes).

But the actual recovery program involves doing AA’s twelve steps (something that Heyman does not mention or consider). No studies have yet addressed the relationship between completing the twelve steps and recovery. Because many who attend meetings do not do all the steps, it is reasonable to suppose that recovery rates will be even better among step completers than among meeting attenders. AA’s main constituency is early-onset, severely dependent, chronic alcoholics (Willenbring et al. 2009), that is, those with the most severe form of the disease. And it works well for that constituency.

Now it is far from obvious how membership in AA would help one to adopt global rather than local choice-making strategies, so its efficacy does not seem to be consistent with Heyman’s own proposal. However, it is well explained by the hypothesis that alcoholism is a disease.

AA itself was the first body to make a splash with the idea the disease hypothesis (Alcoholics Anonymous 1939). AA regards alcoholism as a threefold disease consisting of a physical allergy, a mental obsession, and a spiritual malady. The physical allergy is what now seems to be the hypersensitivity of the dopamine system. The mental obsession is the mental obsession. And the spiritual malady is a kind of self-centeredness that leads to an alcoholic to all kinds of problems managing his own emotions and his own life. The program of AA addresses these problems in depth and detail. AA has never offered a cure for the physical allergy, which seems to remain in chronic alcoholics. It does, however, address the problems caused by the ‘spiritual malady.’ (For a book-length account [End Page 310] of this from a neuroscientific point of view, see Teresi and Haratounin 2011).10

My account of the underlying nature of the alcoholic’s mind neatly explains how AA works so effectively. The basic structure of the alcoholic’s mind is this. He has a primitive desire or urge to drink. He has a reflective ‘cognitive’ desire not to drink. The conflict leads to cognitive dissonance, which leads to denial. Denial undermines the reasons for not drinking and so reduces the desire not to drink. The alcoholic’s desire to drink wins. The way to fix the problem is, first, to attack denial, so as to keep the desire not to drink strong. And, second to attack the leading causes of the desire to drink. These are two. The first is alcohol itself. Once an alcoholic has one drink, this sets off cravings and the urge to drink becomes much stronger. The second main booster to the desire to drink is stress. It is well-attested that stress tends to increase drinking or drug taking in most populations (see Teresi and Haroutounian 2011, 133–5, for references to numerous studies showing this). And there is evidence that stress has the same kind of agitating effect on the dopamine system as drugs or alcohol themselves (Berridge and Robinson 2003; Marinelli and Piazza 2002; *Teresi and Hartounin 2011, 139). Thus, stress may fuel the urge for alcohol even as much as alcohol itself does. (This does not imply that stress is always a factor in the development of alcoholism in all individuals. It is possible that some people develop patterns of use that establish addiction owing, for example, to that of modelling their behavior on family and friends, or because they like drinking a lot.)

In summary, AA attacks denial. It attacks the first drink (promoting abstinence) and it attacks stress. Thus, the desire not to drink is kept strong, the desire to drink is kept sleeping, and a healthy mental balance is maintained.

Alcoholics Anonymous attacks denial by describing its various manifestations in ways that alcoholics find compelling. Step one forces them to face up to it in all its insanity. And regular meetings provide constant reminders to the member of the reality of its risks.

Many aspects of the program are focused on stress management: Building a realistic self-image, boosting self-esteem, and avoiding guilt. The approach echoes that of the Stoics. AA makes much of the serenity prayer: “God grant me the serenity to accept the things I cannot change, courage to change the things I can, and wisdom to know the difference.” It is often remarked that the prayer echoes a central theme in the work of such thinkers as Epictetus, Seneca, and Marcus Aurelius. The Stoics are thought by some to have originated cognitive therapy. Both the Stoics and AA hold that problematic emotions can be dealt with, in part, by the application of intellectual analysis. In the case of anger management for example, AA bids a member to consider how he himself contributed to a situation in which he became angered, and what it is about himself that fuelled his angry reaction: Was it, for example, wounded pride? (Alcoholics Anonymous 1939, 65). Seneca bids us not to dwell on a perceived wrong that has been done to us, and not to focus on a justification for revenge. Both Seneca and AA hold that anger is wholly bad and can and should be wholly eliminated. But they differ in that AA accepts that in general a certain amount of emotion is natural, inevitable, and a good thing, whereas the Stoics held that, with work, emotions could be entirely eliminated by intellectual means. Both are oppose to anger for reasons that relate in obvious ways to AA as a fellowship. Seneca writes (letter to Novatus, Inwood 2007, cp Twelve Steps and Twelve Traditions, 48): “human life is founded on kindness and concord, and is bound into an alliance for common help, not by terror, but by mutual love.” AA writes (Alcoholics Anonymous 1939, 65): “Courtesy, kindness, justice and love are the keynotes by which we may come into harmony with practically anybody.”

A lot of AA step work concerns what clinicians call ‘emotion processing.’ Early-onset, heavy-drinking, chronic alcoholics have learned to use alcohol to suppress their emotions and have never learned to process them properly: To bring them to the surface, to understand and label them, to put them in their proper place, or to allow them naturally to dissipate or become their friends. If not processed properly, guilt, shame, anger, resentment, and fear provide powerful, stressing stimuli for a sensitized reward system. [End Page 311]

AA’s steps four and five require a member to construct a ‘searching and fearless’ moral inventory that allows him to process emotions relating to his entire past life. Step ten requires him to keep a daily inventory and share it with another. He can also share aspects of his inventory at meetings. He thus learns to deal with his emotions with the help of his sponsor and other members of the group. A greater power than himself helps to maintain his sanity.

AA’s stress relief program alleviates the mental obsession and cravings caused by the physical allergy. It thus provides excellent therapy for the disease of alcoholism, under the sophisticated understanding of the nature of the disease that AA itself provides.


There are two extreme pictures of problem drinking that one might come across. One has it that problem drinking is simply a matter of choice: The subject just makes bad decisions. This may be owing to some combination of conditioning, stresses and strains, depression, childhood trauma, low self-esteem, poor choice-making strategies, hedonism, self-indulgence, or whatever. But, at the end of the day, he simply chooses to drink too much, too often. Problem drinkers can control their drinking, if they choose to, and do control their drinking when the incentives are large and salient enough. Problem drinkers are fully responsible for their behavior.

The other picture has it that many problem drinkers are afflicted by a disease: Alcoholism. Alcoholism causes absolutely irresistible urges. The alcoholic has absolutely no control over his drinking. The disease is progressive. The urges increase in frequency over time and, because of tolerance and other factors, the subject has to drink more and more. The condition is chronic and irreversible. An alcoholic cannot recover without treatment or a recovery program. With such help, an alcoholic can quit drinking. But the underlying condition is always there and a few drinks will probably precipitate a relapse. The disease is allor-nothing: Either you have it or you do not. If you do not, you can drink safely. If you do, you must abstain forever or drink a lot and die in your cups. The disease is physiological. The alcoholic is not responsible for his condition or for the excessive drinking that it causes.

If what I have argued is along the right lines, then neither picture is quite right, although both contain important elements of truth missed by the other. Alcoholics’ drinking is governed by their own choices. Yet alcoholism is a disease. The alcoholic suffers cravings, is preoccupied with drinking, and suffers from impaired control over it. Alcoholism is driven by a pathological, recurrent, excessive desire to drink, which impairs the subject’s ability to make the choices he would like to make and that he struggles to make. Alcoholism is often, but not always, accompanied by recurrent pathological insane beliefs and other alcohol-centered misjudgments that further impair the subject’s ability to control his decision making. The excessive desire causes the cravings and preoccupation. Alcoholism is a discrete condition in that either you have it or you do not. But it comes in degrees. It can be relatively benign and easy to manage, malignant and almost impossible to manage, or in-between. It is not, of its nature, progressive, although it does progress in many subjects. Some alcoholics can recover without help, others cannot. Sometimes alcoholism is reversible, sometimes it is not. Alcoholism is a psychological disease, a disorder of desire- and, sometimes, belief-forming mechanisms. Whether it also involves a purely neurological element of pathology is not yet known. In view of impaired control, it seems reasonable to suggest that the alcoholic’s responsibility for his excessive drinking and even his drunken actions is diminished under certain conditions, for example, if his drinking is involuntary and his capacities for rational judgment and self-control are substantially impaired. This view is roughly in line with current English law (Galappathie and Jethwa 2010).

Related Articles

Feature Article: Alcoholism, Disease, and Insanity

Commentary: Alcoholism in Theory

Commentary: Is Alcohol Addiction Usefully Called a Disease?

Response: Common Sense, Science, and ‘Spirituality’

Gabriel M. A. Segal

Gabriel Segal is a philosopher by training, but has always had a keen interest in the sciences, particularly psychology and linguistics. He has written two books: Knowledge of Meaning, with Richard Larson, MIT Press (1995) and A Slim Book About Narrow Content, MIT Press 2000. He can be contact via e-mail at


The author thanks Peter Adamson, Kent Ber-ridge, Daniel Friesner, George Graham, Nick Heather, Jim Hopkins, John Kelly, M.M. McCabe, David Papineau, Hanna Pickard, Richard Sorabji, and Steve T for input. [End Page 312]


1. For an history of the topic of alcoholism as a disease, see Heather and Robertson (1997). A great deal that matters has often been discussed in terms of the disease view: Issues relating to treatment/management, law, government policy, responsibility, an alcoholic’s self-conception, and the way their families and others think of alcoholics. The notion of disease is probably too flimsy to dictate answers to any of these important issues, except possibly the question of treatment on the national health services and coverage under medical insurance. Nevertheless, the question of whether alcoholism is a disease remains one of heated debate. I touch on many of the issues that matter to alcoholics’ lives in the body of the paper.

2. Hanna Pickard defends a version of a choice theory in Pickard (2012). I respond in Segal (2012).

3. I assume that alcoholism is just one manifestation of a more general disease of addiction, which includes at least addictions to drugs and sugar, and possibly other food items, gambling, and sex. I focus on alcoholism because that has been most studied. But most of what I say is intended to apply to addiction in general.

5. This includes many of those from theorists (e.g., representing NIDA) who claim to be offering a theory in terms of the brain: They are in fact explicitly trying to account for the hardware features that underlie affect and cognition.

6. Modelling is probably not always a factor in human alcoholism, because there are alcoholic rats and mice.

7. Problem drinkers sometimes talk of the “fuck-it button”: “Yeah, maybe I’ll get ill, lose everything and die a slow and painful death, but fuck it, I’m going to have a drink!”

8. Compare Frankfurt (1971). This is also consonant with Donald Davidson’s (1980) account of acting against one’s will. For discussion, see Heather and Segal (forthcoming). Heather and Segal discuss how an addict’s experience of feeling compelled to use is explained by his own inability to explain his actions. I would add that part of this story is that alcoholics sometimes drink when they do not expect to enjoy it. This makes their behavior even more baffling to themselves than a case of normal weakness of will, when a subject can at least say: ‘I did it because I thought I would enjoy it.’

9. Stephens and Graham (2009) point out that sensitized ‘wantings’ are not automatically compulsions: They are merely more powerful than unsensitized ‘wantings.’ That’s right. And, indeed, the alcoholic’s motivation to drink is not always compulsive. The sensitization model does show how ‘wantings’ can be compulsive: Sensitized ‘wantings’ can represent drinking as far more valuable than the agent reflectively judges it to be (see Berridge and Robinson [2011] for discussion). Reflective judgments can manifest themselves in a desire not to drink, and a wish neither to have nor to act upon the desire to drink, whereas the sensitized ‘wanting’ (which is the desire to drink) still succeeds in motivating behavior. In such cases, ‘wanting’ is compulsive. See Berridge and Robinson (2011) for discussion.

10. The founders’ view was that the spiritual malady required a spiritual solution, which in turn required conscious contact with God. The 1955 edition of Alcoholics Anonymous (AA) tells the story of a member who recovered without believing in God. And subsequently AA have published a couple of documents recognizing this possibility of nontheist recovery (e.g., ‘Do you think you are different?’ at provides all AA’s written material, hence everything the AA ‘thinks’ or ‘says’). One study (Kelly, Hoeppner et al. 2011) did find that a move toward more religious views correlated with greater chances of recovery, but only among ‘more impaired’ subjects. Otherwise, atheists do just as well as theists in AA.


Alcoholics Anonymous. 1939. New York: Alcoholics
Avena, N., P. Rada, and B. Hoebel. 2007. Evidence for sugar addiction: Behavioral and neurochemical effects of intermittent, excessive sugar intake. Neuroscience Biobehavioral Review 32:20–39.
Berridge, K. C. 2012. From prediction error to incentive salience: Mesolimbic computation of reward motivation. European Journal of Neuroscience 35:1124–1143.
Berridge, K. C., and T. E. Robinson. 2003. Addiction. Annual Review of Psychology 54:25–53.
———. 2011. Drug addiction as incentive sensitization. In Addiction and responsibility, ed. J. Poland, and G. Graham, 21–54. Cambridge, MA: MIT Press.
Davidson, D. 1980. How is weakness of the will possible? In Essays on actions and events, ed. D. Davidson, 21–42. Oxford: Oxford University Press.
Festinger, L. 1957. A theory of cognitive dissonance. Stanford, CA: Stanford University Press.
Fingarette, H. 1988. Heavy drinking: The myth of alcoholism as a disease. Berkeley: University of California Press.
———. 1990. Why we should reject the disease concept of alcoholism. In Controversies in the addictions field, ed. R. C. Engs, 48–54. Dubuque, IA: Kendall-Hunt.
Foddy, B. 2010. Addicted to food, hungry for drugs. Neuroethics 4:79–98. [End Page 313]
Frankfurt, H. G. 1971. Freedom of the will and the concept of a person. Journal of Philosophy 68: 5–20.
Galappathie, G., and K. Jethwa. 2010. Diminished responsibility and alcohol. Advances in Psychiatric Treatment 16:193–8.
Goldstein, R. Z., and N. D. Volkow. 2011. Dysfunction of the prefrontal cortex in addiction: Neuroimaging findings and clinical implications. Nature Reviews Neuroscience 12:652–69.
Heather, N., P. Booth, and A. Luce. 1998. Impaired control scale: Cross-validation and relationship with treatment outcome. Addiction 93, no. 5:761–71.
Heather, N., and I. Robertson. 1997.Problem drinking. Oxford: Oxford University Press.
Heather, N., and G. Segal. Forthcoming. The experience of addiction: Donald Davidson and the problem of akrasia. Addiction Research & Theory.
Heyman, G. M. 2009. Addiction: A disorder of choice. Cambridge, MA: Harvard University Press.
Hyman, S. E. 2005. Addiction: A disease of learning and memory. American Journal of Psychiatry 162:1414–22.
Inwood, B. trans. 2007. Seneca: Selected philosophical letters. Oxford: Oxford University Press.
Jellenik, E. M. 1952. Phases of alcohol addiction. Quarterly Journal of Studies on Alcohol 13:673–84.
Kaskutas, L. A. 2009. Alcoholics Anonymous effectiveness: Faith meets science. Journal of Addictive Diseases 28:145–57.
Kelly, J. F., B. Hoeppner, R. L. Stout, and M. Pagano. 2011. Determining the relative importance of the mechanisms of behavior change within Alcoholics Anonymous: A multiple mediator analysis.Addiction 107: 289–99.
Kelly, J. F., R. L. Stout, and M. Magill. 2011. The role of Alcoholics Anonymous in mobilizing adaptive social network changes: A prospective lagged mediational analysis. Drug and Alcohol Dependence 1142, no. 2–3, 119–26.
Kumar, S., R. K. Singh, U. Goswami, and U. Khastgir. 2005. A study of the temporal course of phenomenonology of alcohol dependence. American Journal on Addictions 14:213–22.
Lende, D., 2008. Getting hooked on sin.Scientific American. Accessible at:
Marinelli, M., and P. V. Piazza. 2002. Interaction between glucocorticoid hormones, stress and psycho-stimulant drugs. European Journal of Neuroscience 16:387–94.
Marlatt, G. A., and J. R. Gordon. 1985. Relapse prevention. New York: Guildford Press.
McGue, M. 1999. Behavioral genetic models of alcoholism and drinking. In Psychological theories of drinking and alcoholism, 2nd ed., ed. K. E. Leonard, and H. T. Blane, 372–421. New York: Guildford Press.
Moos, R. H., and B. S. Moos. 2006. Participation in treatment and Alcoholics Anonymous: A 16-year follow-up of initially untreated individuals. Journal of Clinical Psychology 62:735–50.
Morse, R. M., and D. K. Flavin. 1992. The definition of alcoholism. Journal of the American Medical Association 268, no. 8:1012–4.
National Institute of Alcohol Abuse and Alcoholism (NIAAA). 2011. Alcoholism isn’t what it used to be. Spectrum 3. Bethesda, MD: Author.
National Institutes of Health (NIH). 2013. Alcohol use disorder: A comparison between DSM–IV and DSM–5. Bethesda, MD: Author.
Nestler, E. J. 2005. Is there a common molecular pathway for addiction? Perspective 8, no. 11:1445–9.
Peele, S., A. Brodsky, and M. Arnold. 1992. The truth about addiction and recovery. New York: Simon & Schuster, Inc.
Pickard, H. 2012. The purpose in chronic addiction. AJOB Neuroscience Journal 3, no. 2:40–9.
Polich, J. M., D. J. Armor and H. B. Braiker. 1981. Stability and change in drinking patterns. In The course of alcoholism: Four years after treatment, 159–200. New York: John Wiley & Sons.
Rada, P., N. M. Avena, and B. G. Hoebel. 2005. Daily bingeing on sugar repeatedly releases dopamine in the accumbens shell. Neuroscience 134:737–44.
Robinson, T. E., and K. C. Berridge. 1993. The neural basis of drug craving: An incentive-sensitization theory of addiction. Brain Research Reviews 3:247–91.
Schroeder. T. 2010. Irrational action and addiction. In What is addiction?, ed. D. Ross, H. Kincaid, D. Spur-rett, and P. Collins, 391–407. Cambridge: MIT Press.
Segal, G. 2012. Commentary on Pickard, ‘The purpose in chronic addiction.’ AJOB Neuroscience Journal 3, no. 2: 63–4
Stephens, L., and G. Graham. 2009. An addictive lesson: A case study in psychiatry as cognitive neuroscience. In Psychiatry as cognitive neuroscience: Philosophical perspectives, ed. M. Broome, and L. Bortolotti, 203–22. Oxford: Oxford University Press.
Teresi, L., and H. Haratounin. 2011. Hijacking the brain: How drug and alcohol addiction hijacks our brains: The science behind twelve-step recovery. Bloomington, IN: AuthorHouse.
Twelve Steps and Twelve Traditions. 1952. New York: Alcoholics Anonymous World Services Inc.: New York.
Vaillant, G. 2003. A sixty-year follow-up of alcoholic men. Addiction 8:1043–51.
———. 1983. A 60–year follow-up of alcoholic men. Addiction 98:1043–51. [End Page 314]
Volkow, N. 2010. Drugs, brains and behavior: The science of addiction. National Institute of Drug Abuse NIH Pub 10–5607. Bethesda, MD: National Institutes of Health.
West, R. 2006. Use and misuse of craving. British Journal of Addiction 82, no. 1:39–41.
Willenbring, M., S. Massey, and M. Gardner. 2009. Helping patients who drink too much: An evidence-based guide for primary care physicians. Journal of American Family Physicians 80, no. 1:44–50.
Wilson, W. A., and C. M. Kuhn. 2005. How addiction hijacks our reward system. Cerebrum: The Dana Forum on Brain Science 7:53:66. [End Page 315]

Next Article

Alcoholism in Theory

Additional Information

Print ISSN
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.