Make your own free website on

The Self-Medication Hypothesis

Home | Review | Khantzian's Model | Duncan's Model | Newsletter | Archive | Links


by Brian Johnson, PhD

Three perspectives on addiction promulgated during the 1990s are reviewed, along with many earlier contributions to the understanding of addictive illness. It is suggested that these distinct yet overlapping formulations of the dynamics of addiction form a hierarchy for each patient suffering from an addiction. Assessment of a patients ego strength, and of the relative importance of addictive behaviors in overall character structure, allows referral to various types of treatment, including psychoanalytic therapy. Case examples are presented, including material from the psychoanalysis of a woman addicted to heroin, methadone, cocaine, amphetamines, nicotine, alcohol, and shopping.

In the field of addiction treatment there has been a tendency to eschew dynamic understanding for simple descriptive diagnosis based on verifiable criteria. This tendency has been accompanied by a focus on behavioral treatments that can be reliably evaluated using objective outcome measures and by an immense research effort to understand the biology of addiction. All this leaves the thoughtful clinician with the unanswered question, How does one understand these behaviors empathically? It leaves the insight-oriented clinician with the question, Does a patients capacity for self-observation contribute anything to the treatment of addiction? It can leave the clinician asking, Is addiction treatment a constant process of identifying an addiction and referring the patient away to physicians who prescribe medications such as disulfiram and naltrexone, to Twelve Step programs where mysterious events somehow keep the patient sober, and to relapse prevention specialists who lecture on how to "identify your triggers"? The psychodynamically oriented clinician can feel that addictive disorders somehow fall out-side his or her purview.

Nonetheless, addiction has been an important issue in psychoanalysis since its inception. In 1908, for example, Abraham published "The Psychological Relations between Sexuality and Alcoholism," in which he suggested a number of possible dynamics. In the mid-1960s the early psychoanalytic literature was nicely summarized by Rosenfeld (1965), and since then psychoanalysis has benefited from the work of a number of practitioners specializing in addictive disorders. Their contributions have dealt specifically with addictive dynamics, providing a range of views perhaps regarded as "perspectives" (rather than mutually exclusive schools of theory), useful guides to understanding and interpretation (see Spezzano 1998).

Three perspectives on addiction recur in the psychoanalytic and addiction literature: addiction as a biologically mediated disease, addiction as a response to inability to tolerate affect, and addiction as an object or transitional object equivalent. These themes will be presented with reference to the literature and their usefulness explored. In a field that has been accumulating knowledge for over a century, none of the authors chosen here as exemplars has a completely original idea, and all of them have carefully reviewed the many pathways leading to their specific formulation. In what follows I have resorted to simplification in the service of making these perspectives more salient.


Addictive drugs seem to become entrained into the same drive system that motivates animals to seek food, water, and sex (Miller and Gold 1993; Volkow quoted in Swan 1998). One perspective, articulated by Robinson and Berridge (1993; Berridge and Robinson, 1998), describes the progression of interest in drugs from incidental to driven. This theory might be described as an attempt to localize drive within mesotelencephalic dopamine pathways of the brain. However, the authors combine introspective exercises and a discussion of social factors involved in addiction to produce a complex and comprehensive way of thinking about addiction. Robinson and Berridge begin with three key questions concerning the nature of addiction: (1) Why do addicted persons crave drugs? (2) Why does drug craving persist even after long abstinence? (3) Is "wanting" drugs the same as "liking" drugs? Their answer has four main points.

1. Addictive drugs share the ability to enhance mesotelencephalic dopamine neurotransmission.

2. One psychological function of this neural system is to attribute "incentive salience" to the perception and mental representation of events associated with activation of the system. Incentive salience is a psychological process that transforms the perception of stimuli, imbuing them with salience, making them attractive, "wanted" incentive stimuli.

3. In some individuals repeated use of addictive drugs produces incremental adaptations in this neural system, rendering it increasingly and perhaps permanently hypersensitive to drugs and drug-associated stimuli. The sensitization of dopamine systems is gated by associative learning, which causes excessive incentive salience to be attributed to the act of drug taking and to stimuli associated with drug taking. Sensitization of incentive salience transforms ordinary wanting into "craving."

4. Sensitization of neural systems responsible for incentive salience (wanting) can occur independently of changes in neural systems that mediate subjective pleasurable effects of drugs (liking) and neural systems involved in withdrawal. After sufficient exposure, the pleasure of addictive behaviors becomes irrelevant because this neural system is built in to stimulate the organism to pursue a goal. Compulsive drug taking ensues, despite lack of pleasure and despite strong disincentivesloss of job, homelessness, the agony of withdrawal.

The neurobiological concept of addiction (of which this is only one particularly well articulated example) includes as a strong central tenet that of all the plant-derived chemicals humans have ingested, a small number have been found to mimic in some fashion a natural process affecting parts of the brain. Natural incentives such as food, water, or desirable sexual partners are endowed by evolution to condition pleasure and incentive salience under conditions such as those created by hormones or thirst. If one is drawn to a desirable sexual partner, one may modify the impulse if one notices a wedding band. In the same way, one may be drawn to a drink but have the impulse modified by last nights meeting of Alcoholics Anonymous.

The process of endowment of a stimulus with salience has three steps: (1) Pleasure is a consequence of a particular event or act. (2) Pleasure is associated with a mental representation of the object, act, event, or place in which the pleasure occurred via classical (associational) conditioning. (3) Incentive salience is attributed to subsequent perceptions and mental representations of the associated object, event, act, or place, which causes them to be "wanted." Stimuli that signal the availability of the incentive become attractive. Acts that led to the situation in the past are likely to be repeated.

Robinson and Berridge suggest the possibility that this whole process, or part of this process, can occur unconsciously. There is no need to know one is being influenced by craving in order to want something. For example, when subclinical doses of amphetamine are administered to subjects who cannot distinguish the effects from placebo, and who have no measurable electrophysiological response to drug injection, these subjects choose the drug lever at higher than chance incidence, all the while insisting that there is no difference in the effect caused by either lever, and that their choices are random.

The incentive sensitization theory nicely explains the common clinical phenomenon that patients say they do not "like" smoking cigarettes, or using cocaine, and yet have intense cravings that seemingly can be responded to only by using the drug.

Recent work (e.g., Sora et al. 1998) suggests that regarding dopamine as the sole neurotransmitter system mediating these phenomena is simplistic. However, while our understanding of the underlying biology of the incentive sensitization model may be modified, its basic conceptualization remains an important perspective from which to understand the driven "it," ego-dystonic quality of addiction.


The self-medication hypothesis, first articulated by Khantzian (1985, 1997) states simply that drugs relieve psychological suffering and that preference for a particular drug involves some degree of psychopharmacological specificity. Khantzian believes that opiates attenuate feelings of rage or violence, that CNS depressants such as alcohol relieve feelings of isolation, emptiness, and anxiety, and that stimulants can augment hypomania, relieve depression, or counteract hyperactivity and attention deficit.

Khantzian sees his work as expanding on the work of self psychologists, especially Kohut (1971, 1977). Khantzian (1995) traces the origins of the inability to regulate affects to early life, and to a failure to internalize self-care from parents: "Self-care is a psychological capacity related to certain ego functions and reactions. This capacity protects against harm and assures survival, and involves reality testing, judgment, control, signal anxiety, and the ability to draw cause-consequence conclusions. The self-care capacity develops out of the nurturance, ministrations, and protective roles provided by the parents from early infancy, and subsequently, out of child-parent interactions" (p. 30). Because they lack these internalizations, addicted persons cannot regulate self-esteem or relationships, or provide themselves with caring.

This emphasis on affect intolerance related to early developmental failures is similar to that of Zinberg (1975) and Krystal (1988, 1995; Krystal and Raskin 1981) . However, there is an important difference between Khantzian and Krystal. Khantzian views lack of self-care or self-governance as an ego defect, as a function that never developed, whereas Krystal views self-care as having been prohibited by an overcontrolling parent. In Krystals view, addicted individuals are entirely capable of self-care but "believed that if they took over the control of their vital or affective functions, which they believed to belong to mother, that would be a Promethean transgression, punishable by a fate worse than death" (1995, p. 85).

The self-medication hypothesis is constantly confirmed by listening to patients reports of responding to intolerable affective states by using drugs. A man in a rage sniffs a bag of heroin rather than kill his girlfriend. The heroin allows a pleasant interaction with her. A teenager finds that after a few beers she can enjoy a party rather than be trapped by anxiety. A mans depression can be overcome with cocaine sufficiently to allow a social interaction. A woman who has been abused and molested can engage in sexual relationships after premedication with alcohol or heroin.

While the most important evidence for the self-medication hypothesis is found in the reports of patients, Khantzian carefully examines more quantitative studies in the addiction literature. He finds that psychoanalytically informed experience becomes a check on some of the conclusions offered by researchers who employ more operationalized methods in the attempt to understand and describe addictive behaviors. For example, he considers the possibility that some longitudinal investigatorse.g., Schukit 1986; Vaillant (1983, 1996)find affective disorders a consequence rather than a precursor of addiction, because of their failure to detect earlier subclinical conditions that subjects are already medicating by the time they are diagnostically apparent. Khantzian suggests that relatively infrequent interviews and the requirement that subjects meet diagnostic criteria for relatively severe affective disorders runs against the reality that some people go into action with drugs early in the course of these disorders to alter affective states that are experienced as unbearable. Khantzians view is confirmed by a recent study by Kushnet, Sher, and Erickson (1999) that demonstrates a reciprocal causal relation over time, with anxiety disorders leading to alcohol dependence and vice versa.

The subjective stance provides support for other objective findings. Khantzian suggests that nicotine use is self-medication. He cites a study by Breslau, Kilbey, and Andreski (1993) showing that 1,007 subjects with nicotine dependence were higher on rating scales for negative affect, hopelessness, neuroticism, and general emotional distress than were nondependent smokers. He also cites a study (Anda et al. 1990) in which the quit rate for depressed smokers was found to be 10 percent, as against 18 percent for nondepressed.

Dodes (1990, 1996) suggests that addicted persons have a narcissistic vulnerability to feeling overwhelmed by experiences of helplessness. The centrality of helplessness in the creation of psychic trauma is cited by Freud (1926, pp.166167), and helplessness as a central addictive dynamic created by overwhelming shame by Wurmser (1978). Dodes believes that the enactment of addictive behavior functions to restore a sense of potency against helplessness. He states that the intense aggressive drive to restore this potency, which arises from the narcissistic injury of helplessness, is identical with narcissistic rage. Finally, he notes that the major symptoms of addiction, as well as its intensity and unrelenting, boundless quality, can themselves be explained by the presence in addiction of this narcissistic rage.

Dodes also suggests that addictions can be shown to be compromise formations identical with compulsions. He gives case examples that demonstrate restoration of a sense of power via addictive behavior as a displacement from actual reassertion of power in the real world. For example, a patient who is enraged with his son for embezzling from the patients company goes on a drinking binge. The man feels that it would be wrong to fire his son, so he is rendered helpless to act. Drinking makes him feel better because it is an action he can take; he doesnt feel helpless anymore.

The empathic understanding that patients have been traumatized by helplessness, and are responding in an aggressive but displaced manner, allows the clinician to make interventions that appreciate the drive without encouraging the behavior: (1) the aggressive drive for control of ones existence with integrity is nothing to be ashamed of; (2) the patient needs to struggle to be conscious of what he or she really wants, rather than settle for addictive responses; (3) conflicts and vulnerabilities regarding self-assertion, and difficulty tolerating helplessness when necessary, have their origin in childhood experiences that need to be remembered and worked through in treatment.


The object-quality of addictive behavior is central to many theories of addiction. Winnicotts original formulation of the transitional object (1951) described it as an addiction. Kernberg (1975) describes several object-related dynamics of addiction: it may replace a parental imago in depression or an all-good mother in borderline personality, or may refuel a grandiose self in narcissism. Wurmser (1995) describes the terror of being separated and sees the intense shame and rage manifested in addictive behaviors as in part an attempt to maintain a connection with objects. Wurmsers important contributions (e.g., 1974, 1978, 1981) include, as one dynamic, the difficulty of internalizing interactions with parents into effective superego functioning, and the resulting alternation of slavish submission to unreasonable internal prohibitions with completely unregulated rebellious addictive behaviors. Meyers (1994, 1995) shows, in the psychoanalytic therapy of patients addicted to compulsive sexual behaviors, that these behaviors can resolve as patients begin to rely on self or others as a nurturant object.

In a recent contribution (Johnson 1993) I presented an object model that employed a unique definition of addiction: "An addiction is an ostensibly pleasurable activity which causes repeated harm because a person involuntarily and unintentionally acquires an inability to regulate the activity, and has a persistent urge to engage in the activity. A psychological system, referred to as denial, is created around the harmful behavior. Denial allows the addicted individual to continue this activity despite its detrimental effects" (p. 25).

The function of the denial system of an addiction is to protect the relationship with the addiction. It is made clear with case examples that if there is no denial, then there is no addiction. Denial is part of the pathophysiology of the disease (Johnson and Clark 1989). The definition is psychological in a way that is true to the phenomenon of addiction. While genetic, biological, or social aspects may contribute to the course of the illness, they do not define its essence. This definition both allows addiction to fall into the mainstream of psychoanalytic consideration, and adds the characterological response of the individual to drug-effects as an important consideration in assessing the impact of drugs (see, e.g., Kernberg 1975). It allows the psychology of the relationship of each individual with his or her addiction to be articulated and elaborated by the dynamically oriented clinician (see also Kaufman 1994).

This definition is used to link a number of pleasurable activities that are addictive only if they become compulsive behaviors: drinking, gambling, stimulant use, exercise/endorphin release or opiate (heroin) use, eating, making love, shopping, working, or being slim. These activities are required to be compulsive for characterological reasons because they provide a constant sense of being accompanied. Addicted individuals are unable to have their dependency needs effectively met in human relationships and are unable to tolerate being alone; their need for object constancy is provided by whatever compulsive activity is chosen. A particular addiction is chosen then as a function of environment and gender, and can be shifted with changing environmental conditions. For example, a bingeing/purging food-addicted woman may shift to cocaine dependence because it keeps her weight down, and may subsequently become preoccupied for a time simply with obtaining, using, and recovering from cocaine. A man who is in trouble because of his drinking may shift to compulsive gambling because, at least for a while, he can better get away with this compulsive behavior. When pursuing heroin becomes too much trouble as people grow older, they may shift to alcohol dependence.

I have suggested (Johnson 1993) that a defect that occurs during preoedipal development becomes manifest as an addiction during adolescence because the teenager needs to leave the parents yet lacks the internal development to survive without them. The adolescent does not have a confident sense of object constancy. This property of recall memory is internalized by most toddlers during the period between one and three years of age via a process of separating and returning to the facilitating parents, the "emotional refueling" of Mahler, Pine, and Bergman (1975). The developing child gradually begins to carry an internal sense of being accompanied by the parents ideationally, without needing the concrete parent as a constant reassurance of protective presence. Especially during the rapprochement period, from sixteen to twenty-five months, the child is beset with rageful fantasies of parental destruction because of omnipotent wishes to have the world conform to ones desires. The facilitating adults must help the child hold aggressive urges in safety. The rules and prohibitions of the parental adults are internalized as a superegoan internalized sense of which behaviors are permissible and which behaviors must be held in check. It may well be that inability to negotiate this step has much to do with the environment created for the child by caregivers (Lyons-Ruth 1991); premorbid parent-child interactions that predispose to addiction are described by Shedler and Block (1990). I hypothesized that children who will go on to suffer from addictions do not internalize object constancy during the preoedipal period, and have a specific fear that their aggressive urges may destroy their relied-upon objects. Inability to effectively use superego prohibitions makes their aggressive urges frightening. Years later children are faced with the need to separate from their family of origin and respond by adopting an addiction. The adolescent who has newly adopted an addiction is extremely content. The annihilation anxiety previously experienced has given way to an idealized relationship with the addictive behavior.

The use of an addiction, then, is akin to Kernbergs borderline and narcissistic personality disorders (1975). In narcissism the inability to tolerate being alone is solved by reliance on an organized inner set of idealized fantasies that allows the individual to be indifferent to the comings and goings of real relationships (Volkan 1973). In an individual with borderline personality, affective instability is responded to via a constant, desperate need for reassurance by an idealized person. In addiction, the relationship is neither with idealized internal fantasies nor with idealized persons, but rather with an idealized addictive behavior.

According to the developmental model, some addictions represent a regression under stress, rather than a true adaptation to absence of object constancy. Individuals who have a regressive addiction find that they can give it up relatively easily. In these cases, the addiction is a neurotic behavior that lacks the destructive forcefulness of addictive behaviors clung to as a means of preventing the inner experience of abandonment.

Using this model, I suggested that some patients who have their underlying conflicts analyzed may return to recreational use of alcohol; I cited the liver enzyme results of a patient whose hepatitis resolved during three-times-a-week treatment despite her continued use of alcohol (Johnson 1993). The Twelve Steps of Alcoholics Anonymous, I suggested, involve relinquishing the object constancy delivered by the addictive behavior; adopting, through a "leap of faith," the belief that reliable human objects exist; reworking the superego; and extending this remedial work to the ego ideal, or internalized social values (see also Dodes 1988; Khantzian 1994). AA encourages members to "rely on people, not alcohol (drugs)" and to be carried by an inner "higher power" that provides a sense of purpose and of being accompanied at all times. In an earlier paper (Johnson 1992), I presented the psychoanalysis of a man with active alcoholism and showed the resolution of his addictive drinking as the highly conflictual dependence entered the transference neurosis.


Authors writing from all three perspectives regard as completely erroneous any suggestion that addiction is driven by a desire for pleasure. The incentive sensitization model suggests that the mesotelencephalic pathway carries wanting/craving and that pleasure soon becomes an irrelevant factor in addictive drug use. The affect intolerance perspective posits that chronic inability to master feeling states results in recurrent flight into drug-altered states. According to the addiction-as-substitute-object model, lack of capacity to use relationships, internal or external, results in the constant need for addictive behaviors as a transitional object.

Typical features of addictive disorders are explained in each of the three perspectives in ways that are complementary. For example, the mesotelencephalic pathway is responsible for activation of the animal to seek gratification. When this pathway is cut in rats, the animal does not bother to eat, even though eating (as shown by rat facial expression) is still a pleasurable activity. If a light signals that food is about to appear, the light causes discharge in the mesotelencephalic pathway. Dopamine transmission is triggered not solely during the gratifying activity, but by the simple assurance that gratification is impending. Patients with cocaine dependence often remark on this phenomenon. Their bodies react to the certainty of gratification; symptoms of the gut motility are initiated by the mere intention to purchase cocaine after a period of abstinence. For instance, they will pass wind on the way to the dealer. Dodes notes the same phenomenon, dubbing it "signal satisfaction, akin to signal anxiety," and suggests that the ability to create a satisfying situation gives the individual a sense of power. For a sober person with alcoholism, Dodes says, the very act of ordering a drink at a bar relieves a sense of helplessness.

Helplessness is considered a key affective state in the three psychoanalytic models. Dodes (1990) has suggested that his view helps us understand the role of drugs in avoiding certain affects, as described in Khantzians self-medication model. He suggested that intolerable helplessness is the result of the psychic trauma of being overwhelmed by whichever affective state each individual person finds the most troublesome. Imyself have traced the history of helplessness to early experiences in which the child is unable to master aggression without parental assistance. The addicted person is left with a choice of helpless submission to inner and outer authority, or defiant rebellion against it. (This position is identical to that of Wurmser.) Dodes (1988, 1990), Khantzian (1994), and I (Johnson 1993) all suggest that this experience is reflected in Step One of Alcoholics Anonymous, which begins, "We admitted we were powerless. . . ." By contrast, Robinson and Berridge might take the position that the organism is powerless against a biologically driven demand for drug seeking. The ventral tegmental pathway demands that the animal take action to secure water, food, sex, or drugs.

In summary, these three perspectives on addictive behaviors offer overlapping and complementary explanations. At one or another time, one of these dynamics may appear most prominent as a motivating force. Taken together, they represent a substantial framework from which to listen to patients, to empathically understand their associations and behaviors, and to guide interventions that help them move toward safety and toward more effective ways of living.


This section will begin with examples drawn from patient encounters in which only one of our three perspectives is appropriate. Combined use of the models will be shown later.

Use of the Neurobiological Perspective

Case 1. A forty-year-old man with schizophrenia is referred from a psychiatric hospital to a substance abuse outpatient clinic because of persistent use of cocaine. Auditory hallucinations and paranoid delusions are in remission as a result of fluphenazine decanoate injections administered every two weeks. The patient is eager to become sober, but finds that on the first of each month, despite having a payee for his Social Security checks, he cant help but use any money he can find to buy crack. This results in loss of his housing when he cant pay his rent, and rehospitalization.

The incentive sensitization theory is used by caregivers to understand that money in his pocket is the element that turns on craving. This patient lacks the relatedness required to use either Twelve Step groups or psychotherapy. Lack of adequate supervision during vulnerable periods will result in continued cocaine dependence. Placement in a staffed residence and tighter control over his money result in a remission of use.

Case 2. A substance abuse counselor, sober six years and active in AA, travels to an old haunt in order to help his mother sell her house. He notices unexpected powerful urges to pick up a prostitute, drink, and buy crackall associated activities during his years of drug dependence. He realizes that helping his mother sell her house is a kind of help he is not capable of providing in safety. Simple avoidance of the old neighborhood resulted in complete resolution of the urges to return to addictive behaviors.

Case 3. An international businessman presents for treatment for active heroin use, complaining that he has injected more than a million dollars worth of heroin into his veins over the last fifteen years. He has also supplied his wife with a second million dollars worth. The expenditure is undermining the capitalization of his business. The initiation of heroin use is understood as a consequence of intense stress during immigration from Lebanon to the United States, and the beginning of a business from the back of his car. This stress is now in the past, no longer a factor contributing to ongoing use. The cause of continued use is the unstoppable craving.

For the first two months of weekly meetings, the businessman injects eight bags of heroin before each morning psychotherapy hour. The history of intense craving after detoxifications results in a plan to switch over to methadone detoxification and accomplish a six-month taper. Supportive psychotherapy focuses on tolerating craving. When the patient is down to 5 mg of methadone, he believes that his intense craving will undermine his attempt to decrease the dose to nothing, and he flies to England to obtain his own supply of methadone. He tapers from 5 mg to abstinence over the next three months. He realizes that his wife has a more complicated addiction and separates from her. Supportive psychotherapy is terminated when he is abstinent from opiates for three months. Alcohol and marijuana use do not cause any symptoms. On two-year follow-up he is abstinent from opiates, except for a single use of heroin, which he thinks of as having been "stupid." At that point he does not meet the DSM-IV criteria for any disorder except for "opiate dependence, in long-term remission."

Case 4. A physician was raised in a culture where cigarette smoking was the norm. When in his early thirties he was diagnosed with gingivitis by his dentist, he immediately recognized this as a medical complication of smoking. He has a clear memory of throwing his pack of cigarettes out the car window as he left his dentists parking lot. He had minimal craving when first abstinent, and has not smoked a cigarette in twenty-five years.

One might speculate that for some individuals, such as this physician, the neurobiological aspect of addiction is insufficient to sustain the behavior. Because of relatively healthy ego functioning, denial is easily disrupted, resulting in "spontaneous" long-term remission (see DSM-IV, "Substance Related Disorders" section, p. 189; Shaffer and Jones 1989).

Use of the Affect Intolerance Perspective

Case 1. A thirty-three-year-old woman has a ten-year history of alcohol and cocaine dependence. She has been through twenty detoxifications, has never been sober two weeks, and complains that counselors tell her to go to AA when she has emotional issues to deal with. She tells a horrific set of stories of abuse and neglect, starting when she was seven. Her father was drunk and slapped her mother. Her eighteen-year-old sister went to the kitchen and stabbed her father in the chest. Her father refused to move, stood there for twenty minutes, then collapsed. Her mother reached down, felt for the fathers pulse, said he was alive, and ran off to hide the sister, leaving the patient with the dying father in a spreading pool of blood. Although the family myth is that the father fell on the knife, the sister did prison time for manslaughter. No one has spoken to her regarding this event since. She says that as she sits in the detox interview room, she can see it all as if it just happened.

Two other traumata, including a rape, have left her with a full syndrome of posttraumatic stress disorder. Cognitive function is entirely intact, suggesting that psychotherapy would be of help. A Hamilton Rating Scale for Depression score is 28, suggesting a major nonpsychotic depression. The patient, who is one day sober in detox, begs for medication to help her sleep because she becomes terrified as she falls asleep, with hypnogogic illusions of being touched sexually.

The self-medication hypothesis is invoked as the most relevant paradigm for this particular patient. Fully appreciating that a depression cannot reliably be diagnosed only one day away from alcohol and cocaine, the posttraumatic stress disorder is rated as the most pressing diagnosis, despite presentation at a detoxification center. Trazodone, because of its sedative side effect, is selected as the antidepressant of choice and is given in gradually increasing doses to 300 mg. The patient is referred to a womens halfway house that specializes in victims of sexual violence, and is seen in psychotherapy focusing on memory reexposure and grief work.

Case 2. A thirty-five-year-old woman is unable to stop smoking, despite several attempts. She complains of intense dysphoria when off cigarettes, and of an experience of white-knuckle emotional pain until she resumes smoking. Her addiction to cigarettes is understood as self-medication of an underlying depression. She receives a course of twelve weeks of weekly psychotherapy abetted by sustained release buproprion, 150 mg twice a day. She chooses a quit date five weeks into the psychotherapy, and is amazed at how much easier it is this time.

Use of the Object Perspective

A forty-five-year-old professional woman has been addicted since her teenage years to heroin, methadone, alcohol, cocaine, and nicotine. She has been unable to remain reliably sober, despite a considerable investment in Alcoholics Anonymous. Psychotherapy is begun immediately following discharge from heroin detoxification, and eight months later, when she is sober from all the drugs listed above, she begins four-days-a-week psychoanalysis. The transference is difficult to manage because her alcohol and opiate dependent mother was hateful and manipulative, and has never expressed any interest in sobriety. Her father was himself addicted to work and did nothing to protect her from her mother, from her abusive stepmother, or from her addiction.

The patient attends AA meetings and an AA womens group, and has a sponsor. However, splitting is evident from the beginning of the treatment. Her sponsor is alternately idealized and devalued. Comments describing her sponsor as unavailable, uncaring, or even worthless are interpreted as an expression of the paternal transference toward the sponsorthe patient fears the sponsor is not available and attending to her needs. The maternal transference is active directly in the relationship with the analyst. As one dynamic, the patient expects the analyst to "catch" her with feelings that will be used to humiliate her.

Seven months into the analysis, the intense negative transference seems to have settled down. There are associations, with references over several weeks, to substantial purchases. When the analyst hears that the patients teenage son has been bought a new car, and that the patient is contemplating a new dining room set, he has an awful realization about why the transference has lessened. He asks directly about credit card use, and learns that during the course of treatment the patient has been involved in an escalating debt that is now $66,000. The debt involves twenty credit cards, including some that have been fraudulently obtained in the name of an incapacitated relative. The analyst interprets spending addiction as a resistance to the further deepening of the transference relationship, and gives direct advice about the necessity of immediate cessation of credit card use and of consulting a lawyer regarding bankruptcy. The patient confesses a fantasy that two or three months hence the analysis would have to end because she could no longer afford treatment and credit card interest payments, and that this would be just like the period of transition from feeling in control of a gradually escalating heroin habit to becoming desperate and realizing that she needs to go to detox. She notes that never before has she had any trouble with credit cards or spending.

This sequence of feeling and action enacted in the psychoanalytic relationship is understood as a repeat of the patients experience, as a teenager, that she could no longer endure the intensity of her feelings, especially of anger and humiliation with regard to her mother, in the context that she was not protected by the father, and that she adopted an addiction to enable her to tolerate the continued relationship with both parents. However, the relationship with the addiction supplanted the parental relationships. Despite the relationship with the psychoanalyst, this patient is using addiction as her only reliable object. The analyst recognizes and interprets that active addiction is incompatible with psychoanalysis. The patient continues in psychoanalysis three years after this intervention and is tolerating the transference in part because of the alliance generated by this transaction.

Combined Use of the Three Perspectives

During an earlier hour from this womans psychoanalysis, the patient began by noting that she had arrived an hour early, realized her mistake, and gone for a frantic hour of shopping. Her next set of associations included dread at meeting a new internist, an addiction specialist to whom her analyst had referred her as a replacement for a physician who ordered any medications she requested. Her fear of seeing a physician who "knows about" her"Its hard to trust them when I dont know their thoughts"had been interpreted as a fear that the internist would humiliate her as her mother had. She thought of the abuse her son was suffering from his stepmother, and remembered being nine: "We went out to eat. I was whining that I didnt like anything on the menu. My mother said with a smile, Sweetie, would you want to come in the bathroom with Mommie? I said okay. When we walked in she slapped me so hard it almost knocked my head off. She told me that I was to order something on the menu, eat it without talking, and never mention what had happened in the bathroom to my father. It wasnt the pain that was the worst, it was the surprise."

Rather than accept that this was only a past memory, the analyst interpreted current concerns about how her son was being treated. The patient responded, "Ive been on this mission for a week. Im cleaning and cleaning. I was thinking, When did I use to behave like this? I worry I have to get prepared. Even when I went to detox the last time I brought my taxes. I used to do this when I was twenty-seven. Ralph, the bum who lived with me, said, For god sakes, did you wash the floor again?"

The analyst asked what feeling she had at that time. The patient replied, "I dunno. I just used to clean. Are you supposed to have a feeling? I remember when our dog Spotty died, the one we had for years, my mother said in a sad voice, Spotty died." The patient laughed. "I didnt care a goddamn thing about Spotty. I thought, Im supposed to be sad. Like when my father told me he was divorcing my mother I thought, Am I supposed to act sad?" The patient laughed. "I didnt care. But I think I feel good now. I think I enjoy that running around shopping. You shop, you clean, you fix, you shop, you clean, you fix."

The analyst asked, "Fix?"

The patient answered, "Fix things, make them right."

The analyst said, "Of course, fix has another meaning."

The patient answered, "I dont know why I said that. I meant, you straighten things."

Noting the theme of helpless anger, the analyst said, "I wonder if you dont feel furious." The patient responded:

On and off since Ive been coming here, Id be walking down the street, or in my car, and Id feel I was dying. I didnt feel bad about it. I just felt I was dying. Then I thought, "Thats what my mother is doing, slowly dying." Then I thought, "I hope thats not an identification."

I looked at my hands the other day. I didnt like what they looked like. Im a hand person. I can remember everybodys hands. I may not remember what guys were like in bed, but I remember their hands. I remember what your hand felt like when I shook it the first time I met you.

I bite my nails, even when I put acrylic nails on top. I get frantic. I just have to bite. I want short nails. I want mans hands. My nails are red now. I dont like the color red. My hands look old. Do my mothers hands look old? No. I put on the acrylic nails because once I start biting them, Im on a mission, I bite them until they bleed. . . .

My mother has long nails. My stepmother had long red nails. My fifth-grade teacher moved my desk near her. When she didnt like what I did shed dig her nails into my arm to shut me up. My mother used to grab me with her nails too. . . .

My mother used to bite her nails. Shed wear bandages. Id bite my nails. Shed scream, "Stop it!" Id stop for five seconds, and do it again. I couldnt stop.

The analyst asked, "Do you see how the anger and the compulsion go together?"

The patient asked, "What compulsion?"

The analyst said, "Biting, cleaning, shopping, fixing."

The patient answered, "Other people clean their houses."

The analyst asked, "What do you think about coming early today?"

The patient answered, "I was pissed at myself. Now I try to ignore it, laugh it off. Id be beating myself up all the time if I continually thought about what I do."

She next associated to a constant need to drive past her ex-husbands house when she had to drop her son off there, with her son complaining each time that "the house is back there." The analyst suggested she was angry about having to leave him.

Her next associations were about whether she would cancel her next hour to go skiing. When the analyst took this as a violation of the contract to meet four times a week, the patient seemed reassured. She then realized that she was also scheduled to speak at a commitment of Alcoholics Anonymous that evening. She associated to the need of a friend to keep seeing his psychiatrist, and how his physician-father was known in the addict community as a "croaker" who would sell benzodiazepine prescriptions unethically.

The analyst interpreted the story as a displacement to the transference experience of the patient. She had undone her anger at the analyst by coming early instead of making him wait; she was angry about the loss of time with her son because of the divorce; her associations about compulsive behaviors and compulsive drug use represented associations about how angry she was; and she tended to contain her aggression either by undoing it as a compulsion or by displacing it into addictive behavior. The story of the doctor/father who had injured his son indicated that she was feeling unsafe with the analyst. When she whined to her parents about ordering at the restaurant when she was nine, she had been humiliated for expressing her feelings. The urge to cancel an hour was another means of expressing her anger.

The patient called the interpretation "far-fetched," associated to forgetting to set her two alarm clocks so that she could pick up her son at his fathers and drive him to school, and then suddenly looked at her watch and told the analyst that the hour was over. She was exactly right.

The patient is using her treatment to explore her experiences of addiction and compulsion, including her compulsive shopping. Because of her experience growing up with her parents, the analysts expectation that she associate triggers intense anger and a feeling of helplessness and shame. Dodess assertion that addiction and compulsion overlap, and might be considered identical, is nicely illustrated. Is she shopping compulsively, or does she have a shopping addiction? Both are true.

All three perspectives could be invoked to understand this material. The patient was active with her shopping addiction at the time of this hour. The analyst was completely unaware of this, and the patient may well have not been conscious of the addictive nature of her behavior at that time. Robinson and Berridge might say that the anger she was feeling might have been an associational trigger of her compulsive shopping. Use of their model suggests that there is no possibility of empathic understanding of the behavior, because it is driven by a subcortical pathway. The patient might have the experience that she was "just shopping," and might then construct a denial system to explain or excuse the behavior.

The developmental model I have advanced might be used, as in the earlier example, to explain that the patient was already experiencing abandonment by the analyst, and had taken up shopping as an alternative dependence-gratifying relationship. "Is the time up yet?" might be heard as the experience of a person who knows that shopping never lets her down, while her psychoanalyst-as-mother has repeatedly crossed the line from caring to intrusive, hateful probing, and therefore has already been dismissed as a person who can be there for her.

Khantzian would undoubtedly point to opiates as the drug of choice and would suggest that this patient is unable to tolerate anger/rage as an underlying cause. The patient has no idea how to take care of herself when she is overwhelmed by the experience of anger and humiliation at the inquiring stance of the psychoanalyst, which she finds so reminiscent of her mothers sarcastic "Are you having a feeling, sweetie?"

However, in this particular hour, it seems that the most helpful interpretations for the patient are of aggression directed toward the analyst as a defense against the transference experience of being helpless against a figure who would humiliate her for her feelings. Her compulsive behaviors might be described by Dodes (1996) as a displacement. She is shopping, cleaning, and fixing rather than articulating her angry feelings toward her analyst.

The treating clinician who uses the three models is thereby in a position to evaluate each patient who presents for treatment in the context of the level of ego functioning he or she displays. A healthy patient who has regressed to the use of an addiction as a defense during a stressful period may easily be able to tolerate the craving that comes with cessation of drug use. For example, a relatively healthy patient might be sent to a smoking cessation program. The heroin-dependent patient described in the first clinical section needed only some attention to his experience, some factual explanations, and methadone detoxification. Beyond these relatively simple interventions lies the need to correct the underlying dynamics of the addictive process. Sometimes this can be accomplished by brief behavioral interventions, or by attendance at Twelve Step recovery programs.

Clinicians need to be aware of the tendency to shift from one addiction to another. Researchers investigating the outcome of treatment of addictive diseases need to employ more sophisticated models of recovery. For example, if a patient stops smoking cigarettes and gains a hundred pounds, should this individual be counted as a success, or as someone whose addiction has shifted to a less scrutinized substitute? For many patients, substitution of one addiction for another must be counted a relative therapeutic triumph. For example, Bill Wilson, the founder of Alcoholics Anonymous, died of his nicotine addiction, but many productive years after he became sober from alcohol. In other cases, substitution needs to be taken into account as a sign of inability to attain a stable recovery. Studies of methadone maintenance, for instance, tend to use abstinence from opiates as an outcome measure, when many patients continue to use alcohol, cocaine, or benzodiazepines in an addictive, self-destructive manner (Miller and Gold 1993; Condelli et al. 1991).

For the dynamically oriented clinician, the three perspectives presented here orient the treatment according to whether simple craving is driving the constant impulse toward drugs, or whether more complex dynamics must be taken into account. In this context, diagnosis, relapse prevention coaching, Twelve Step meetings, and supportive or expressive psychotherapy, either alone or in combination, all become understandable, applicable modalities for helping the patient.

The question of whether to treat the patient with psychoanalytic therapy rests not on the diagnosis but on the patients behavior. As described by Dodes (1984), patients who frequently miss treatment hours, whether because of hostility or drug use, or who are physically present but impaired by chemical use, cannot be treated in an outpatient psychotherapy practice. However, patients who come and work, regardless of their involvement with addictive substances or compulsive behaviors, can be helped. Addictive behaviors might have to be addressed directly in order to preserve the treatment relationship. For example, in the psychoanalysis described above, the analyst had to suggest directly that the patient cut up her credit cards, stop her illegal use of a relatives card, and see a bankruptcy lawyer as a condition of her continuing analysis. Flexible use of the three perspectives abrogates any artificial distinction between patients appropriately treated with behavioral, Twelve Step, or psychoanalytic therapies. Any or all are applicable to specific patients, depending on the particular nature of their addictive dynamics.

There can be a tendency to either ignore the addictive process or to elevate it to the only concern. For example, patients who compulsively eat, and who present to psychotherapy, might not address their food addiction directly in their associations. They might have a denial system the clinician complies with, so that a central symptom is left out of the treatment. Many members of Alcoholics Anonymous have complained that they were in psychotherapy or psychoanalysis, yet their drinking was either ignored or relegated to a subordinate position. However, by a strange twist, psychological treatment is often withheld until a patient becomes sober. Some addiction clinicians will tell patients to stay sober for a year using nonexploratory treatments and only then return for uncovering psychotherapy. The case presented above of the man who used eight bags of heroin before his psychotherapy hours, and who continued to use alcohol and marijuana after the treatment, might be regarded as an anathema by some clinicians, despite the excellent outcome in terms of functioning. Using the three perspectives allows a flexibility in treatment that is true to the individual being helped. It allows for the centrality of abstinence in some treatments, yet eschews an insistence on abstinence as a prerequisite for treatment in all cases. At the same time, there is a focus on the use of any compulsive behavior as a sign of distress, whether or not a chemical substance is used in the behavior.

Finally, use of the three perspectives eliminates the strange-ness of addiction, which tends to frighten some clinicians away from engaging with patients who are actively addicted. Addiction is not viewed as a bizarre, awful, or degrading behavior that suggests that the patient be sent elsewhere for treatment. Rather, it is seen as one of the most common character adaptations seen in everyday clinical practice.


No claim is made that any of the three models of addiction presented above are the "truth." They are models in the engineering sensethey exist to help get a real job done. Based on some science, some observation, they are directed toward trying to be effective in the real world. It is likely that the models described will be revised and superseded over time.

There are two advantages to the use of three perspectives on addiction. The first is that it removes the need to find any one magical solution to addiction. Instead, we are content to use the particular way of thinking of an addiction that fits a particular patient at a particular time. The second is that there is a correspondence between more general psychoanalytic psychology and the ways of understanding addiction presented above. This allows for further investigation and elaboration of addictive dynamics. However, no attempt is made here to rival other psychologies or treatment methods. On the contrary, inclusivity is recommended, as is the attempt to understand the relationship of psychoanalytic theory and other psychologies (for example, the way mesotelencephalic activation and associated learning contribute to the drive to obtain chemicals).

Within the context of the three perspectives on addiction, it is hoped that clinicians find empathy for the compulsive behaviors of addiction, because empathy is the essential first step in any attempt to be of service.



Abraham, K. (1908). The psychological relations between sexuality and alcoholism. In Selected Papers of Karl Abraham. New York: Basic Books, pp. .

Anda, R.F., Williamson, D.F., Escobedo, L.G., Mast, E.E., Giovino, G.A., & Remington, P.L. (1990). Depression and the dynamics of smoking: A national perspective. Journal of the American Medical Association 264:15411545.

Berridge, K.C., & Robinson, T.E. (1998). What is the role of dopamine in reward: Hedonic impact, reward learning, or incentive salience? Brain Research Reviews 28:309369.

Breslau, N., Kilbey, M.M., & Andreski, P. (1993). Vulnerability to psychopathology in nicotine dependent smokers: An epidemiologic study of young adults. American Journal of Psychiatry 150:941946.

Condelli, W.S., Fairbank, J.A., Dennis, M.L., & Rachal, J.V. (1991). Cocaine use by clients in methadone programs: Significance, scope, and behavioral interventions. Journal of Substance Abuse Treatment 8:203212.

Dodes, L.M. (1984). Abstinence from alcohol in long-term individual psychotherapy with alcoholics. American Journal of Psychotherapy 38:284256.

(1988). The psychology of combining dynamic psychotherapy and Alcoholics Anonymous. Bulletin of the Menninger Clinic 52:283293.

(1990). Addiction, helplessness, and narcissistic rage. Psycho-analytic Quarterly 59:398419.

(1996). Compulsion and addiction. Journal of the American Psychoanalytic Association 44:815835.

Freud, S. (1926). Inhibitions, symptoms and anxiety. Standard Edition 20:75175.

Johnson, B. (1992). Psychoanalysis of a man with active alcoholism. Journal of Substance Abuse Treatment 9:111123.

(1993). A developmental model of addictions and its relationship to the twelve step program of Alcoholics Anonymous. Journal of Substance Abuse Treatment 10:2334.

& Clark, W. (1989). Alcoholism: A challenging physician-patient encounter. Journal of General Internal Medicine 4:445452.

Kaufman, E. (1994). Psychotherapy of Addicted Persons. New York: Guilford Press.

Kernberg, O. (1975). Borderline Conditions and Pathological Narcissism. New York: Aronson.

Khantzian, E.J. (1985). The self-medication hypothesis of addictive disorders. American Journal of Psychiatry 142:12591264.

(1994). How AA works and why its important for clinicians to understand. Journal of Substance Abuse Treatment 11:7792.

(1995). Self-regulation vulnerabilities in substance abusers: Treatment implications. In The Psychology and Treatment of Addictive Behavior, ed. S. Dowling. Madison, CT: International Universities Press, pp. 1742.

(1997). The self-medication hypothesis of substance use disorders: A reconsideration and recent applications. Harvard Review of Psychiatry 4:231244.

Kohut, H. (1971). The Analysis of the Self. New York: International Universities Press.

(1977). The Restoration of the Self. New York: International Universities Press.

Krystal H. (1988). Integration and Self-Healing: Affect, Trauma, Alexithymia. Hillsdale, NJ: Analytic Press.

(1995). Disorders of emotional development in addictive behavior. In The Psychology and Treatment of Addictive Behavior, ed. S. Dowling. Madison, CT: International Universities Press, pp. 65100.

& Raskin, H.A. (1981). Drug dependence: Aspects of ego function. In Classic Contributions in the Addictions, ed. H. Schaffer & M.E. Burglass. New York: Brunnel-Mazel, 1981, pp.161172.

Kushner, M.G. Sher, K.J., & Erickson, D.J. (1999). Prospective analysis of the relation between DSM-III anxiety disorders and alcohol use disorders. American Journal of Psychiatry 156:723732.

Lyons-Ruth, K. (1991). Rapprochement or approchement: Mahlers theory reconsidered from the vantage point of recent research on early attachment relationships. Psychoanalytic Psychology 8:123.

Mahler, M.S., Pine, F., & Bergman, A. (1975). The Psychological Birth of the Human Infant: Symbiosis and Individuation. New York: Basic Books.

Meyers, W.A. (1994). Addictive sexual behavior. Journal of the American Psychoanalytic Association 42:11591182.

 (1995). Sexual addiction. In The Psychology and Treatment of Addictive Behavior. ed. S. Dowling. Madison, CT: International Universities Press. pp.115132.

Miller, N.S., & Gold, M.S. (1993). A neurochemical basis for alcohol and other drug addiction. Journal of Psychoactive Drugs 25:121128.

Robinson, T.E., & Berridge, K.C. (1993). The neural basis of drug craving: An incentive-sensitization theory of addiction. Brain Research Reviews 18:247291.

Rosenfeld, H.A. (1965). On drug addiction. In Psychotic States: A Psychoanalytic Approach, ed. H.A. Rosenfeld. New York: International Universities Press, pp.128143.

Schukit, M.A. (1988). Genetic and clinical implications of alcoholism and affective disorders. American Journal of Psychiatry 143:140147.

 Irwin, M. & Brown, S.A. (1990). The history of anxiety symptoms among 171 primary alcoholics. Journal of Studies on Alcohol 51:3441.

Shaffer, H.J., & Jones, S.B. (1989). Quitting Cocaine: The Struggle against Impulse. Lexington, MA: Lexington Books.

Shedler, J., & Block, J. (1990). Adolescent drug use and psychological health. American Pscyhologist 45:612630.

Sora, I., Wichems, C., Takahashi, N., Li, X.-F., Zeng, A., Revay, R., Lesch, K.-P., Murphy, D.L., & Uhl, G.R. (1998). Cocaine-reward models: Conditioned place preference can be established in dopamine and in serotonin-transporter knockout mice. Proceedings of the National Academy of Science 95:76997704.

Spezzano, C. (1998). The triangle of clinical judgment. Journal of the American Psychoanalytic Association 46:365388.

Swan, N. (1998). Brain scans open window to view cocaines effects on the brain. NIDA Notes 13(2):12.

Vaillant, G.E. (1983). The Natrual History of Alcoholism. Cambridge; Harvard University Press.

(1996). A long-term follow-up of male alcohol abuse. Archives of General Psychiatry 53:243249.

Volkan, V.D. (1973). Transitional fantasies in the analysis of a narcissistic personality. Journal of the American Psychoanalytic Association 21:351376.

Winnicott, D.W. (1951). Transitional objects and transitional phenomena. In Collected Papers. New York: Basic Books, 1958, pp. 229242.

Wurmser, L. (1974). Psychoanalytic considerations of the etiology of compulsive drug use. Journal of the American Psychoanalytic Association 22:820843.

(1978). The Hidden Dimension: Psychodynamics of Compulsive Drug Use. New York: Aronson.

(1981). The Mask of Shame. Baltimore: Johns Hopkins Press.

 (1995). Compulsiveness and conflict: The distinction between description and explanation in the treatment of addictive behavior. In The Psychology and Treatment of Addictive Behavior, ed. S. Dowling. Madison, CT: International Universities Press, pp. 4364.

Zinberg, N.E. (1975). Addiction and ego function. Psychoanalytic Study of the Child 30:567588.

5 Park Place
Newtonville, MA 02160

Enter supporting content here