Readers of this blog know, or may remember, that yours truly was, briefly, a drug addict. It was actually a rather interesting experience to undergo, philosophically speaking, and one of the things I did while going through it was to read up on the philosophical and psychological literature on addiction, and to compare what I read there with my own six-month experience of addiction. I have a folder full of journal entries on the subject–at least a hundred pages or so–and some day I’d like to get some of that material out there into “the literature.”

A basic problem with the literature, as I see it, is that very few of the people writing in it either are, or have ever been addicts, and their lack of first-hand experience distorts much of what they write on the subject.* Their definitions of “addiction” are far too narrow to cover the varieties of addiction (even to cover the varieties of specifically pharmacological addiction, setting aside the supposed behavioral varieties, e.g. sex addiction, shopping addiction, etc.). And by my lights, they’re far too timid about considering the possibility that addicts are responsible for having becoming addicts, and are capable of choice as addicts.

But one particularly problematic assumption, ubiquitous in both the philosophical and psychological literature, is the claim that addiction necessarily involves a craving for the addictive substance. The paradigm example of this assumption is the celebrated discussion of addiction in Harry Frankfurt’s famous paper, “Freedom of Will and the Concept of a Person” (originally published in the Journal of Philosophy, 68:1 [Jan. 1971], reprinted in The Importance of What We Care About [1988]). It’s in many ways a very insightful paper, and like a lot of people, I’ve been heavily influenced by it. Reading Frankfurt while I was an addict, however, I couldn’t help thinking that he’d generated a conception of “addiction” designed specifically to clarify the thought-experiments in the essay, regardless of whether any of it bore any relation to the real-world phenomenon of addiction.

Whether it’s explicitly cited or not, the Frankfurtian conception of addiction plays an outsize role in the literature on addiction. And it’s not hard to see why. Suppose that you’ve never been an addict, but are interested in the topic. Suppose that you don’t know any addicts, either. How do you know what it’s like to be one? As it happens, you can’t really get a visualizable “picture” of addiction by reading social scientific or psychiatric studies of addiction in peer reviewed journals, by reading the “substance abuse” chapter of a textbook of abnormal psychology, by consulting the newest version of DSM, or by reading either the philosophical or psychological literature on “addiction science.” Nor will it help to attend lectures of this sort. The preceding sources will give you important facts about addiction, and teach you how to logic-chop some important distinctions. They’ll give you some important vocabulary, as well, and introduce you to the various “models” of addiction. But they won’t tell you what it’s like to be an addict, and like it or not (so to speak), the first-person perspective is crucial for understanding what it is to be one.

Enter Frankfurt: Frankfurt gives his readers a vivid “picture” of what it’s like to be an addict. Though it’s a third-personal account, it’s vivid and detailed enough to enable a non-addict to imagine what it would be like to be a (Frankfurtian) addict from the first-person perspective. And clearly, it would suck: a Frankfurtian addict is someone with an irresistible first-order craving for a pharmacologically-addictive substance. Either he resists this first-order craving at the second-order level, or not, and different implications follow in each case. Frankfurt never mentions by name what addictive substance he has in mind, but I get the impression that he’s discussing a stereotypical case of either heroin or cocaine addiction (or perhaps alcoholism).

As I say, it’s an interesting discussion, but I find the picture it paints of the addict very misleading. In particular, I don’t think there’s good reason to think that cravings are either necessary or sufficient for addiction.

To see this, consider a somewhat stylized, thought-experimental version of my own case of addiction. Imagine a very strict Kantian who goes to the doctor with some medical complaint. Our Kantian takes his doctor to be a reliable authority on medical matters, and regards following his doctor’s orders as a matter of duty to self. Further, our Kantian discharges his duties to self from the motive of duty. In other words, if the doctor tells him to do something, he does it because it’s his duty (to self), whether or not he wants to.

So our Kantian goes to the doctor with some medical complaint, and the doctor gives him strict orders to take a certain medication, X. As it happens, X is an addictive, psychotropic medication. Suppose that our patient has a temperamental hostility to the idea of taking any drug for any reason. So he really doesn’t want to take X. But he feels duty-bound to do so, under the doctor’s orders. So he grudgingly fills out the prescription and grudgingly takes X. Within a few weeks, he becomes addicted to it, but doesn’t know that he is. He might in principle continue like this for years, never grasping that every dose he takes pushes him further and further into addiction.

So here is the situation:

• Our Kantian is ex hypothesi addicted to X;
• He keeps taking X, thereby reinforcing his addiction to X;
• He would suffer intense withdrawal if he stopped taking X;
• Despite not wanting to take X, he continues to take X, but only from the motive of duty.

I take it to be obvious that you cannot have a craving for a substance that you do not want to take, and you cannot have a craving for a substance that you only take from the motive of duty. And yet you can clearly be addicted to such a substance, at least in the pharmacological sense of being physically dependent on it. If that’s right, craving for X is not a necessary condition of addiction to X. You can be addicted to X and not know it, hence not crave it. You can be addicted to X and not want to take it, but take it from the motive of duty–hence not crave it.

Reflecting a bit on my own experience, I’m willing to admit that there’s a slight complication here. (The phenomenology of addiction defies neat philosophical claims.) Even in the case of the Kantian addict, I think it’s possible that though our Kantian doesn’t want to take X, and takes it from the motive of duty, the pharmacological/physiological effects of addiction can alter one’s personality so that he’s in some sense psychologically compelled to take X without craving it.

This is an odd thought (and phenomenon), and I would have dismissed the possibility out of hand had I not experienced it myself. Think of it like this. Suppose that our Kantian takes X from the motive of duty and only for that reason. He doesn’t like taking X, wishes he didn’t have to, doesn’t want to. But dutiful Kantian that he is, he takes it. Suppose he takes it every night at precisely 10 pm. As 10 pm approaches, he might find himself in the grips of some very odd internal states. He might, for instance, develop an anxious compulsion to take X, or an uneasily anxious feeling about the idea of not taking X. He would thus find himself in the odd state of taking X from the motive of duty, not wanting to take it, but anxiously feeling compelled to take it, and averse to the idea of not taking it–all at the same time. I actually felt like that fairly often.

Related is the possibility that if our addict fails to take X promptly at 10 (and is sufficiently addicted to it), he either senses or subconsciously anticipates the onset of withdrawal symptoms, and develops a vague (but powerful) psychological compulsion to hurry up and take it. (“Hurry up, please, it’s time….”) Remember, ex hypothesi  that our Kantian neither knows that he’s addicted nor knows that withdrawal is an issue. My point is that the physiology of withdrawal can to make its presence felt in his appetitive states despite his ignorance.

Some might be tempted to call this physiologically-induced appetitive presence a “craving,” but it doesn’t feel, phenomenologically, like anything I would call a craving. In retrospect, I think of it as a classic case of chronic, pharmacologically-induced anxiety.  I’m inclined to think that in a Kantian, this anxiety would manifest itself as a specifically deontic compulsion: the compulsion to take the drug would not be experienced, phenomenologically, as a “craving” for it, but as a very urgent, anxious imperative to the effect that X must be taken. (“Hurry up, please, it’s time….”) But an imperative or an anxiety is not a craving in the ordinary understanding of that term, even if it produces a compulsion to do something. (I’m not a Kantian, but the picture of the Kantian agent I’ve painted here approximates my own experience of addiction. One feature of addiction is that it alters your personality so that you find yourself doing things that would otherwise be “out of character,” and yet weren’t produced ex nihilo, either.)

I suppose you could reintroduce the idea of craving here by claiming that our Kantian has a craving for the substance under the guise of a “craving” for doing his duty from the motive of duty, but even if that is a coherent thought (I’m not sure it is), it’s so distant either from Frankfurt or from what the literature describes as a “craving” that we’d have to revise our understanding of “craving” to be able to use it this way.

So while I want to insist that cravings are not a necessary condition for addiction, I’m willing to accommodate some version of the phenomenon that the Frankfurtian picture ascribes to addiction: addictions involve compulsive or anxious behavior, but compulsions are not accurately described as “cravings.” (It’s essential to my account that in large part, the compulsion or anxiety has a pharmacological etiology. Of course the pharmacological etiology could itself have a psychological one.)

I think it’s obvious that cravings are not sufficient for addictions. We crave many things, but it’s an abuse of language to say that we’re addicted to them. I crave knowledge, but I can’t be said to be addicted to it in the way that I was addicted to Ambien. I once had a three-year-long craving to listen to a single album (AC/DC’s Black Ice): I listened to it several times a week for three solid years. But that wasn’t an addiction in the relevant sense, either. I’m very skeptical of the extension of the concept of “addiction” to behavioral contexts without a pharmacological component, e.g., sex addiction, porn addiction, shopping addiction, etc. In my view, “addiction” is a specifically pharmacological concept involving the ingestion of a physical substance and a neurobiological mechanism that produces physical dependence on the substance.

A final observation: I get the sense that the addiction literature has not fully taken on board the possibility that prescription drugs are, like “illicit” drugs, highly addictive, psychotropic substances.** The literature, then, seems fixated on addictions to alcohol, heroin, cocaine, cigarettes, and the like, and has much less to say about FDA-approved drugs–neuroleptics, anti-depressants, stimulants (including caffeine), benzodiazepines, SSRIs, and so on. That seems to me a massive omission. If anything, it’s the latter category that needs more sustained philosophical attention than the former. I hope to give it some more attention in future posts here.

*A notable exception to this rule is Owen Flanagan of Duke University. See Flanagan’s “What Is It Like to be an Addict?” in Jeffrey Poland and George Graham, Addiction and Responsibility.

**Flanagan is, once again, an exception to the general rule. See the preceding note.

Postscript, March 2, 2015: A simpler and more obvious counter-example to the “craving conception” of addiction just hit me. Suppose that X is addicted to a psychotropic medication, and simply forgets to take it at the appointed time. Surely forgetting to take X is incompatible with craving X. QED.

Anyone who doubts the supposition (that psychotropic medications are addictive) can either check the Physicians’ Desk Reference or Peter Breggin’s Psychiatric Drug Withdrawal for clinical information, or Robert Whitaker’s Anatomy of an Epidemic for narrative/anecdotal accounts.

Obviously, an even simpler counter-example to the craving conception of addiction is the (to me, obvious) phenomenological fact that people can be addicted to psychotropic drugs, experience no craving for the drug whatsoever, and willfully “go off their meds” when they decide for whatever reason to do so. The example in the post is, after all, just an elaborate way of saying that.

According to Jon Elster, “All addictive behaviors seem to go together with some form of craving. The idea of craving–the most important explanatory concept in the study of addiction–is complex” (Jon Elster, Strong Feelings: Emotion, Addiction, and Human Behavior, p. 62). I agree that the concept of craving is complex, but the rest of Elster’s claim–an axiom of the literature on addiction–seems hopelessly wrong to me. It either ignores the possibility (and reality) of iatrogenically-induced addiction to psychotropic medication, or else consigns it to a different, and ultimately marginal conception of addiction that plays almost no role in the sexiest, most prestigious books and journals. The literature doesn’t yet seem to have taken seriously the possibility that doctors can impose addictions on unwilling and unwitting patients.  The very definition of “addiction” manages to get doctors off the hook, so to speak, and manages to blame the victims.

For another couple of examples of the craving assumption, check out Merle Spriggs’s “Autonomy and Addiction,” (PDF) especially pp. 6-7, along with the reference to Morse (n.42).

Postscript, September 28, 2015: I’ve been in the market for a therapist lately. To find the right one, I made an initial list of seven who seemed suitable, drawn mostly from the overlap between the Psychology Today “Find a Therapist” listing and the one for my insurance carrier. One turned out not to be available, one never responded (not the first time), and the conduct and demeanor of a third struck me as off-putting and unprofessional.

So I made appointments with the remaining four, three of whom turned out to be excellent, but one of whom, a PsyD (for whatever that’s worth), struck me, frankly, as a hack. Within short order, Dr. Hack had driven the intake session down (what seemed to me) an irrelevant byroad, and had decided to conduct an aggressive interrogation designed to uncover my flaws as a person. The “flaws” tumbled out, one after another, all based on inferences that no human being could legitimately have made about a stranger within twenty or thirty minutes of meeting him.

It didn’t take Dr. Hack long to conclude that I was clinically depressed and needed to go on an anti-depressant. My affect, Dr. Hack informed me, was “flat,” and that flatness was an infallible indication of depression. It hadn’t occurred to Dr. Hack that perhaps the “flatness” of my affect was a response to the flatness of his personality. When I protested that I didn’t think I was depressed (at all)–didn’t feel depressed, didn’t meet the clinical criteria of depression–I was abruptly told that that was precisely how depression manifested itself in men (as opposed to women): men denied their depression in bouts of irritation and rage; women “stayed in bed all day.” The latter had become the societal stereotype of depression, Dr. Hack informed me, but since atypical depression is still depression, I’d have to accept a diagnosis of depression, whether I liked it or not. And that meant going on an anti-depressant as a condition of working with Dr. Hack, too. Dr. Hack magnanimously allowed that he wasn’t qualified to tell me precisely which anti-depressant at which dose; that was a job for a psychiatrist. But the bottom line was: no anti-depressant, no therapy.

That made things easy, since I had no intention either of going on an anti-depressant or of working with Dr. Hack. Bottom line: I unloaded my co-pay and got the hell out of there.

I tell the story because I think it tells us something about the therapy profession today as well as about its relationship to psychotropic medications.

For one thing, I think therapists suffer from a real problem of professionalism. Even when they get PsyD’s, a supposedly practical doctorate, some of them don’t seem to learn the basics of professional etiquette. Going back to one of the therapists I called before I met Dr. Hack: it’s not kosher to ignore a legitimate query regarding professional services you’ve advertised. You may not want a certain client, even based on the message they leave on your voice mail, but it’s not legitimate to ignore them as though they’d never called you at all.

Therapists like to think of themselves as “health care practitioners,” but don’t seem to have grasped that behavior like that is flatly unacceptable in a health care profession. Incidentally, for a profession so eager to regulate the rest of the world, it’s amazing how proprietary they can be about their supposed right to refuse service (or refuse to contact potential clients) on the basis of whims and hunches about X’s “sounding like” the proverbial “problem client.” In conversation outside of clinical contexts, I’ve heard therapists tell me, sotto voce, “Oh, I stay the hell away from clients like those.” Fine: you have the right to stay away from a certain kind of client. You don’t have the moral right to delete a legitimate query from an unwanted client without further ado.

A second aspect of the same problem: the rush to clinical judgment. As a rule, no therapist can (legitimately) give a DSM-5 diagnosis within thirty minutes of the first intake session. Maybe there are clinical geniuses out there–and/or sufficiently simple cases–that are exceptions to that rule, but otherwise, it seems to me a pretty clear rule.

A corollary of the rule is that you shouldn’t be reaching for the prescription pad half-way before the first session is done. Yes, there are some obvious exceptions to that rule, but the exceptions don’t find their way that often to the average therapy office.

Further implication: prescription is a medical judgment. That means that if you’re going to prescribe a psychotropic medication, you’d better have done a history and physical on your client/patient in the medical sense. If you don’t know how to do a history/physical–and most therapists don’t–then you have no business talking about prescriptions. By “talking about prescriptions,” I mean: saying anything that asserts or implies that the client needs a prescription for some psychotropic medication. At best, a non-MD has the professional right to refer the client out to an MD, but that’s it. Otherwise, my view is that they should keep their mouths shut on the subject.

One more implication: Given the way graduate programs in psychology are currently structured, no PsyD (qua PsyD) ever has any business talking about prescriptions.  Maybe some day, PsyD’s and Ph.D’s will be educated so as to know what they’re doing when it comes to psycho-pharmacology–my friend Ray Raad has made some interesting arguments for that–but that day hasn’t arrived yet, and won’t arrive anytime soon. Until then, I’d prescribe silence.

The mental health professions have expanded the concepts of “mental illness” and “addiction” far beyond what those terms mean in ordinary discourse. Maybe we ought to consider medicalizing the overprescription of psychotropic medications by mental health care practitioners. I’d be interested to see the profession’s reaction to the proposal that overprescription is itself a mental illness or an addiction. At that point, it seems to me, the old adage “physician heal thyself” would come to have new and revolutionary meaning. A thought for DSM 6.