top of page

The Intoxication Cure: Sickness, Sadness, and the Self-Medication Hypothesis

Updated: Aug 29, 2023

wine laced with marijuana. I'd need a cup of coffee after that.

Wine laced with marijuana. I’d need a cup of coffee after that.

When we use a drug off label because it makes us feel good and we are tired of feeling bad, or calm nerves with a glass of wine, or have an extra shot of espresso to get through a long day, we are self-medicating. “I’d better figure out where to score some pot,” my friend said before beginning her treatment for breast cancer. People self medicate. Obviously.

That’s the first claim of Edward Khantzian’s now almost axiomatic hypothesis about drug addiction. People who are addicted to drugs are self-medicating. Obviously. Why else would they compulsively use drugs in the face of so much damage to themselves and their loved ones?

Self-medication helps make sense of nonsensical behavior, implying a rational attempt to treat what ails a person, offering an order to the chaos. It shushes any concern about what Derrida would call, “pleasure taken in an experience without truth.” Drug addicts, if they are self medicating, are trying to get better, albeit problematically and sometimes dangerously.

The self-medication hypothesis also insists on a certain empathy for the addict, who lies on the same continuum as the rest of us, but for the substances used. It has become an unavoidable cliché in popular books about addiction to note that the author, too, has an addiction of sorts. I have been reading Gabor Mate’s memoir about working with addicts, In the Realm of Hungry Ghosts. Sure enough, Mate confesses to an addiction to buying classical music CDs, trotting out a rather facile argument by analogy. I remember the precocious undergraduate who showed me her manuscript about volunteering at a local syringe exchange. The first chapter culminated in the revelation that her fellow undergraduates were just as drug-ingesting as the people serviced at the exchange. And I will confess, during the dark days of a severe depression, to feeling a painful kinship with a clearly intoxicated woman on the trolley. She and I were both struggling to manage our medications, and, on that day, neither of us were having a lot of success. My drugs, I noted, were covered by health insurance. I was fairly certain hers were not. Drug users are just like us and, more provocatively, we are just like them. We are all just trying to get through the day.

Self-medication enables all of us to stay within the realm of the disease model, even if the user has transitioned from the sick role to the sad role. Drugs are medicines and the key difference is that the addict is acting as his or her own doctor. Self-medication exists comfortably within 12-step recovery models at the same time that it can justify substitution therapy, which is often rejected by those same self-help communities. Self-medication reminds us that many of the illicit drugs people use were once considered medication and, but for some pretty serious risks, some of them are pretty effective.

soothing indeed

soothing indeed!


nervous? unable to sleep? take some cocaine!


feeling blue? nervous? suffering from brain congestion?

This version of Khanzian’s theory has become so widely accepted in the popular imagination that it structures every episode of Intervention. The show’s classic in medias res narrative arc begins with the utter chaos of addiction before flashing back to a promising childhood marred by a terrible event, which leads inexorably to drug addiction. The Self-Medication Hypothesis turns addiction into a murdered body in a crime drama, begging us to go back and back and back to find the culprit. Only in this mystery, the culprit is always suffering.

Trauma lies at the root of the popular understanding of the self-medication hypothesis. Drug addicts suffered deeply even before their addiction; they often struggle with PTSD and often have an accompanying mental illness diagnosis. When self-medication is invoked as the reason for the addiction, we are compelled to bear witness to the realities of deep suffering, even if only indirectly and after the fact. When my high school boyfriend slipped deeper and deeper into drug addiction, his schizophrenia diagnosis almost felt like a relief. Now I knew why he couldn’t seem to walk away from the drugs the rest of us just dabbled in before moving on to bigger and better adventures.

In response to various deployments, criticisms, and studies of this theory, Khanzian has modified this original claim from arguing that addicts self-medicate the symptoms of their mental disorders, to arguing that addicts self-medicate the painful emotions that accompany mental illness, to arguing that addicts self-medicate their inability to self-regulate. They cannot calm themselves down or cheer themselves up. They also cannot act in their long-term best interest and often act counter to their own deepest values. Drugs are a sort of self-regulation tool, muting the pointier edges of painful (and often ‘inappropriate’) emotions.

Khanzian’s hypothesis, however, has a second claim, which is that there is a psychopharmacological specificity to the addict’s substance of choice. Stimulant addicts are medicating their depression while opiate addicts are medicating their anger. As with his first claim, this one has been contested and modified in various studies, and the theory that the drug of addiction can reveal something of the inner emotional workings of an addict remains more a general clinical impression than a documented fact. I won’t go into all the twists and turns of the research, so you’ll have to trust me when I say that this is still very much contested territory, in part because no one seems to agree exactly what this second claim means.

For the sake of this post, the most provocative response to Khanzian’s theory comes from Shane Darke, who noted that most addicts are polydrug users, muddying any claim for pharmacological specificity. (In reply, Khanzian noted that most addicts have a “king” drug, and that this is the one that therapists should note). For Darke, however, the fact of polydrug use suggests that addiction itself is the wrong interpretive frame. Instead, he proposes, we should look to intoxication.


And there we are. Intoxication. The red-headed step sister of addiction. A false consciousness if ever there was one. The stumbling, grandiose, visionary, unproductive, short-lived, ridiculous-from-the-outside, sublime-from-the-inside state most of us have visited at some point. In “The Pointless yet Poignant Crisis of a Co-Ed,” Dar Williams describes a college student who is a member of what she calls a “hemp liberation league.” She writes that there was a problem with the league: “every time the group would meet everyone would light up. It made it difficult to discuss glaucoma and human rights, not to mention chemotherapy.” Intoxication gets in the way of the real issues of drug use, abuse, and addiction. Only after she becomes a horticulturalist is the woman able to write about decriminalizing marijuana to her senator with any sort of authority. Intoxication doesn’t make a person very credible; if you want to talk about drugs and addiction, it’s best to be stone cold sober.

Addiction helps contain intoxication. It becomes the lost innocence of edenic bliss paid for with years of useless suffering. When I was doing drug ethnography, many active users told me that they didn’t even enjoy the drug anymore; they just used to feel ‘normal’ – as if pleasure is a liability to the ethos of addiction. That struck me as sad – that someone who lost so much for a substance did not at least get a bit of pleasure from its use. I hoped they were just telling me what they thought I wanted to hear.

Khantzian’s hypothesis works to obscure any possible pleasure by proposing that addicts are seeking a relief from dysphoria. When my depression started to lift, those moments when my mind became unclouded and I felt ‘normal’ still stand in my memory as moments of great pleasure – much as I remember with almost-nostalgia the feel of cool sheets when I was still young and stupid enough to get really drunk, or the moment when the percocet obliterated the pain of a case of shingles. Sometimes, the cessation of pain is pleasure. If done right, intoxication is a wonderful, if complex, pleasure, and sometimes it is even a great tool, and I’m not ready to throw that out in the dirty bathwater of addiction and recovery.

This attention to intoxication feels important because it exposes what I think might be the tender underbelly of an otherwise attractive hypothesis. Khanzian is careful to apply his self-medication hypothesis only to addicts, not to people who are able to manage their use. The addiction itself, his use of the term implies, is what “medicates” the self. Yet I can only think of a few people who would argue that their addiction itself served a purpose – the example that comes to mind is Ann Marlowe, who, in How to Stop Time: Heroin from A to Z, argued that the daily requirements of heroin addiction helped her manage her fear of death. It requires social or economic privilege to be able to “use” addiction as a form of self-medication and the crisis it mediates is usually of a more existential nature. I suspect that for most people, it’s intoxication that serves as a vehicle of the cure. What then becomes iatrogenic addiction is simply a by-product, a risk to weigh against the benefit an intoxicating cure.

english garden at sissinghurst
french garden at versailles

or there’s the French style — here is part of Versailles, where gardening meets OCD

There are plenty of people who use opiates without significant harm to themselves – estimates are that around 20% of heroin users become addicted, for example. Just as we look for evidence of “grit” in school children, and resilience in trauma survivors, perhaps we should also look to those who self-mediate successfully for lessons in survival.

If use did not spiral out of control, then their experiences with drug use did not fall under Khantzian’s rubric, but maybe they should. Maybe, in other words, it is actually possible to use drugs to self-medicate. When I look at my legitimate pharmacopia, I note that the accompanying information explains that all drugs have risks and benefits and that my physician has decided that the benefits outweigh the risks. A more accurate self-medication hypothesis might recognize that doctors are not the only ones who should weigh risks and benefits.

for the rest of this amazing comic, go here:

for the rest of this amazing comic, go here:

Vietnam veterans and rats stuck in cages self-medicated with opiates. When they returned to relative safety or were set free in Rat Park, most of the people and the rats gave up the opiates because life was no longer terrifying and out of control.

A letter to the editor of The American Journal of Psychiatry in 1999 noted success treating three people who had treatment resistant depression with small doses of opiates. None of them increased their dose as of the time of the letter and they all reported a remission of their depression.

In “The Self-Medication Hypothesis Revised” (2003), Khantzian noted that “the general muting actions of opiates can quiet anxiety, but . . . most such individuals do not become hooked on opiates.”

A 2012 article in Science reports that ketamine can essentially “cure” depression within a few hours. The remission can last up to two weeks.

In a 2014 update on the application of the self-medication hypothesis to cocaine dependence, Khantzian joins other authors in claiming that taking Ritalin can help people recovering from cocaine dependence feel better.

Some people suffering with depression discuss minute doses of suboxone curing their depression. Some claim that ibogaine can cure addiction. Some believe that people with low natural levels of THC self-medicate with marijuana. Some believe that nicotine might reduce symptom severity for schizophrenics.

E.M. Jellinek, author of The Disease Concept of Alcoholism, which is where I found these studies

E.M. Jellinek, author of The Disease Concept of Alcoholism, which is where I found the alcoholism studies I reference here.

Then again, early research into alcoholism theorized that people self-medicated with alcohol because it corrected vitamin or enzyme deficiencies. Confusing cause and effect, other researchers theorized that alcoholics were actually self-medicating their liver disease. Taken too glibly, Khantzian’s second claim about pharmacological specificity can dangerously mix cause and effect.

I don’t want to sound glib here. I just want to point out that linking self-medication unproblematically with addiction leaves out the possibility that sometimes intoxication works. If we begin there, then we might move on to questions about acute self-medication versus episodic self-medication versus maintenance self-medication, or think about what enables one person to self-medicate without becoming addicted when someone else cannot walk that line. Or we might think more about who gets to decide what is recreational, what is self-medication, and what is abuse? Or we might think more carefully about what the goal of self-medication is in the first place.

In 1956, Bill W tried LSD. He had initially been “extremely unthrilled” at the idea of giving any drugs to alcoholics, but Bill W also believed in the necessity of a “spiritual awakening” for recovery from alcoholism. Without that awakening, he feared, alcoholics would be trapped in the shadow of an “icy intellectual mountain.” Three decades had passed since his own spiritual conversion and LSD offered a way back to that moment of transcendental insight. In those intervening decades, Bill W remained convinced that the way to “get” the program was through that burst of insight. After his experiment with LSD, Bill was “enthusiastic about his experience; he felt it helped him eliminate many barriers erected by the self, or ego, that stand in the way of one’s direct experience of the cosmos and of God.”

Bill W’s experiment with LSD suggests a more nuanced way to think about intoxication and self-medication. Rather than limiting the Self-Medication Hypothesis to simple “insert substance A into problem B” or “addiction to substance A indicates problem B” models, why not expand it to include moments when intoxication offers the burst of transcendent insight that enables a person to return, again, to the problem of despair and anxiety, this time with a memory of what could be if only she will try, again, to find a way out.



bottom of page