Does Free Will Disappear Because of Addiction?

I have been wondering about what we can learn from the addiction literature about free will. I'd like to hear people's thoughts on this. I am not an expert on the philosophy of drugs (i am experimental social psychologist with expertise in self-regulation and a smattering of other stuff), and i am just going in and reading the literature to see what i can see. I try to have no preferences other than to figure out what's up, and simply to follow the data.

It seems there are two very different positions. One is that addicts lose free will, though only specifically with respect to the addiction, and they retain free will (and moral responsibility) in most or all other respects. The other position is that there is no loss of free will and that maintaining addiction is voluntary behavior.

It looks like addicts themselves and the medical establishment firmly favor the no-free-will position. But then it is self-serving for them, and they do not mostly have large impartial data sets. In contrast, the researchers, who do have these broad data sets, are somewhat more divided, with perhaps more leaning toward the free will side.

Addiction seems somewhat like a really bad itch that you know you shouldn't scratch. You can't help wanting to scratch, though ultimately you do have free will and retain control over whether you scratch or not. The problem is, an occasional scratch won't really matter, but the few lead to the many, and many scratchings are very bad for you. As George Ainslie says, willpower resides in treating each act as an examplar of what you should always do.

A variety of evidence supports the free will position. First off, most addicts do quit. Many quit and then relapse, but many others quit and do not relapse. This empirical pattern was not found and then not appreciated for a long time, especially by the medical establishment, because addicts in professional treatment samples mostly do not quit and then mostly relapse. What's more, those clinical samples of addicts  mostly have multiple psychological problems. All these things are different in nonclinical addict samples, the majority. The thing is, most research studies use the more convenient samples, namely the addicts in treatment, not realizing that they aren't typical. So the literature gets this distorted impression of addicts as deeply pathological, hopeless cases, who either cannot quit or will soon relapse.

    This is the sort of error psychology has made repeatedly. A famous previous example was with homosexuality. Theories of homosexuality first emerged from clinicians who treated many homosexuals, and they observed that homosexuals were neurotic and so forth. Homosexuality itself was long considered one form of mental illness. But eventually it became clear that most homosexuals, especially those not currently going to psychotherapy, were not neurotic, and homosexuality itself is not a mental illness.

The mistake with homosexuality got corrected in part because many gay people came forward to insist that the prevailing views were wrong. Their voices were heard. Unfortunately for the cause of truth, no addicts are standing up to insist that addicts are mostly sane, capable, responsible individuals, or otherwise undeserving of being regarded as mentally ill. Au contraire, many addicts benefit from being diagnosed as mentally ill, and especially as unfree. When celebrities or public figures are disgraced because of misbehavior from drugs or addiction or whatever, they prefer to come forward and say that this was the addiction talking and acting, it robbed me of free will. It is an illness, like diabetes, and I cannot be blamed it or for its effects on me.

Even with heroin, apparently, the most common pattern is to become addicted in one's early twenties and to stop using heroin pretty much for good in one's late twenties. Stanton Peele said "People who have better things to do don't become addicted." This struck me as outrageous when I first read it, but I must say that most of the evidence I have seen is consistent with it. Most heroin addicts approaching thirty realize that if they want to pursue some particular chance of a career or family, they need to stop spending all their time zoned out, and they manage it, withdrawal and all. Addiction starts and more-or-less normal adult life starts.

Likewise, cigarette smokers have quit in large numbers. Back in 1960, when the bad news about smoking began to come out in the mass media, most American men were smokers. Today less than a quarter are. Moreover, many of them have quit multiple times, which shows they can quit.

There seem to be differences of opinion about the physical dimension of addiction. Some say there is no way to tell the difference between a corpse of an addicted alcoholic and a corpse of somebody who drank the same amount without being addicted (even, for ex., somebody imprisoned and forced to drink). Others point to withdrawal symptoms and other things indicative of some physical dependency. Even so, does physical dependency entail loss of control? Or is it just like a really tempting itch?

With smoking, it seems, lots of people quit. At first it is very unpleasant, with many negative symptoms. After about a month, all but two of the symptoms are mostly gone. The remaining two are increased food appetite and weight gain. The most likely explanation is that nicotine suppresses appetite and thereby keeps weight down, and these go back to baseline when one quits, so they are not symptoms of withdrawal but simply the normal bodily condition that was artificially suppressed during the smoking addiction.

Thoughts, anyone?

Comments

[Nicole Vincent]

Hi Roy

I think that there are at least two separate questions here (by "here" I mean "in my mind") which we should disentangle:

(1) Does free will necessarily disappear as a result of drug use?

(2) Does free will necessarily disappear as a result of drug addiction?

As regards question (1), it seems to me that the answer is "No". Consider for instance a person who drinks a glass of alcohol at a party to overcome his shyness. His free will is restricted before the drink sinks in, since he lacks the courage to walk up and talk to anyone. But once the alcohol does its magic and he loosens up, his free will is arguably restored. Other examples for the positive effects of drugs on free will can be found in the many psychopharmacological treatments for various crippling disorders (e.g. OCD) which arguably reduce free will. But I don't think that this view only applies to the use of drugs to overcome arguably pathological conditions like extreme shyness or conditions classified as mental disorders, since someone who wants to do something very brave but lacks the courage to do it – i.e. something which ordinary folks would lack the courage to do as well, and thus something which it is not pathological to lack the courage to do – may also take some "Dutch courage" to expand their free will and do what they believe they ought to do. http://en.wikipedia.org/wiki/Dutch_courage Arguably, this person's free will is enhanced not just restored — i.e. it's not just that the drug doesn't reduce his free will, or that it restores it to normal, but it's rather that the drug actually expands his free will even further beyond the norm. Similar comments could be made about the free will enhancing effects of cognitive enhancement drugs; I've a paper on this topic which I can share if you're interested. So if the question is whether drugs _necessarily_ reduce free will, then I think the answer is "No".

But none of the above entails that drugs can't reduce free will (i.e. I've dropped the "necessarily" clause), nor that specific drugs have no free-will-reducing features. I don't think that we can give a general answer to the question "Does free will disappear as a result of drug use?", because it will vary from drug to drug, from person to person, and from situation to situation (and maybe a range of other contextual factors also come into play).

As regards question (2), it seems to me that when the question is put this way (again with the "necessarily" clause in place) then the answer should be "No". For instance, a person who becomes addicted to caffeine might never the less value its effects, and the fact that she is addicted may be neither here nor there for her. She might say "OK, so I'm addicted; but I'd keep drinking my cups of coffee anyway even if I weren't addicted. I like what caffeine does for me — i.e. the enhanced productivity." The fact that she is addicted seems to only reduce her free will to kick the habit, but why would she care about whether she has or lacks the free will to stop doing something beneficial? Conceptually, I can see that this is a reduction in free will, but practiacally it's not a reduction that I would care about. So I don't think that addiction necessarily reduces or eliminates valuable instances of free will.

But again, drop the "necessarily" clause, and I think the answer will again become more contextualised to the precise drug being investigated, the person taking it, and the situation to which the discussion pertains (and maybe other factors).

***

So, to summarise, what I'm arguing above is that free will is not necessarily reduced by addiction (though it may be reduced in some contexts), and that free will may even be restored or enhanced by some instances of drug use (though again contextual factors must be considered).

Cheers

Nicole

[Neil]

I think the cue dependence of addiction is crucial. In the presence of cues and the availability of the means of consuming the drug, the addict experiences ego depletion (you've heard of that, right?). Ego depletion leads to consumption. Does it lead to a free choice to consume? That depends on the mechanism. I have argued that ego depletion leads to a loss of control over the agent's mental life: she experiences an uncontrollable shift from judging she ought to refrain to judging that she has better reason to consume. Given the cues and the availability of the drug, I think she is not free to resist unless she knows how to resist (and has the ability to put this knowledge into practice): by removing herself from proximity to the cues. So the story is compatible with addicts' giving up.

A harder case, and one about which I am unsure what to say, concerns the freedom of the actions needed to procure the drug. There is evidence that these actions themselves can become conditioned stimuli, but the story that the addicts loses mental control with regard to each stage of the process looks a stretch.

[Eddy Nahmias]

Interesting thoughts, Roy, especially the comparison you make between the limited population of homosexuals and of addicts that the medical profession and psychologists focus(ed) most of their attention on. I'd guess that there is a literature on "functioning addicts" but I don't know about it.

My own view, which you seem to share, is that free will involves capacities for rational deliberation, conscious choice, and self-control, and these capacities can be possessed to varying degrees and exercised to varying degrees. So, different addicts may possess more or less free will regarding their addictions, and they may be able to (have the opportunity to) exercise free will to varying degrees in specific situations (e.g., refraining while at work vs. refraining while surrounded by friends who are partaking).

We certainly don't have the resources (yet) to measure people's free will, but we already make more fine-grained distinctions than the all-or-none picture suggested by both the philosophical literature on free will and the psychological literature on addiction you mention. Addicts, like the rest of us, don't simply have free will or lack it.

[Ann the oracle]

Free Will ( or agency ) is always present and available to the addict and everyone else. Free will is the same for everyone - It is inherent in our divine nature. Clouded in thinking errors, the recognition of your own innate power to overcome and make different choices ( consciously applying free will ) is also affected by the injury to the decision making part of the brain by the misuse, overuse and abuse of substances.

[Peter B. Reiner]

I don't think one can make a strong case for free will disappearing in the addicted brain (A separate discussion relates to whether one can make a strong case for free will existing in any brain, but I digress.) I think that one can make a strong case for the position that pharmacological forms of addiction modify one's neural circuitry in such manner that one's strength of will - in one particular domain - is weakened. [This is related to Neil's comments on cue dependence.]

If addiction caused a generalized weakness of will, then one could suggest that the locus of the neurobiological change was precisely where strength of will was located. But this seems not to be the case.

Using cigarette smokers as an example, it is precisely their ability to resist the pull of nicotine that is compromised, not their ability to resist the attraction of brownies. From this, one can conclude that it is the encoding of the attractiveness of a particular sub-domain that is altered in addiction.

[Josh Shepherd]

Good stuff, Roy.

1a] I've come across some interesting articles and talks on addiction lately. Oneplace is the blog Practical Ethics News, which has some nice papers and talks up: http://www.practicalethicsnews.com/practicalethics/addiction/

1b] Another, which you've probably seen, is the 2008 Behavoiral and Brain Sciences target article by Redish et al, 'Addiction as vulnerabilites in the decision process.' The authors provide a nice synthesis of addiction research which seems friendly to Eddy's view, as laid out above.

2a] Addiction certainly contains a normative component, and I'm not sure how many people address this explicitly. Most agree that one can be addicted to drugs or gambling or philosophy blogs, but most would not agree that infants can be addicted to their mother, or that healthy people can be addicted to food, even if physiological and mental phenomena (i.e., homeostatic and allostatic processes in the body and brain triggering typical reactions in response to stimuli in various conditions and at various times, along with the kind of mental 'itch' wherein one can think of little else than that next meal, or whatever) are similar in these cases.

2b] In addition to the normative component, I think there might be an interesting social component to addiction -- a way in which social environments both enable and attenuate behavioral patterns which might qualify as addiction, on some views of addiction. What this might say about free will, I'm not completely sure.

[Philoponus]

I would like to comment on Nicole’s point about certain drug therapies “expanding” someone’s free will. OCD patients who have had a successful course of treatment—the best SSRI for OCD is probably fluvoxamine (brand name Luvox)—do say thing like the drug has set them free (from the rituals which were dominating much of their waking life). Medtronics has also been in trials for some time with neural implants to treat OCD (and depression). Same reaction from those patients. Note that Frankfurt-style manipulations are now technically possible! Suppose then a very depressed or compulsive individual, who has actually done nothing illegal because of his severe depression or OCD, gets on Luvox or implants and feels liberated—to begin a career of crime. The therapy allows him to act out on his criminal desires, frees him in this sense to do what he wants to do. But has it expanded his free will, which I think we want to associate at least in part with a capacity for deliberate choice and action? Does a drug or implant that removes a disability and lets someone act out his desires thereby increase his free will and responsibility or just compromise it more? I expect any day to hear about the “Medtronics Defense”, when a patient, cured of his depression by implants and committing a serious crime, claims the implant is largely responsible for his criminality.

[Chandra Sripada]

Hi Roy,

Thanks for the very thoughtful post. I wonder if it helps to make a distinction between two questions:

Question 1: Is it the case that most addicts lack free will? (where ‘addict’ is stipulated to mean something like people that meet DSM-IV criteria for substance dependence)

Question 2: Is it the case that at least some addicts, presumably those with quite severe addiction, lack free will? Or is it the case that even in the most severe cases of addiction, the addict ultimately is free with respect to his or her substance-procuring/consuming actions.

These are two very different questions and the answers are quite independent. I am in total agreement with you that with regard to question 1, the medical establishment exaggerates the plausibility of the no free will position, and you nicely summarize some of the relevant evidence in your post. But the second question, in my view, is the one that has most interested philosophers. To answer the second question, I think we have to get clear on whether some desires can be truly irresistible, as opposed to merely being hard or unpleasant to resist. This requires some good philosophical work to try to make sense of the concept of irresistible impulses (which turns out to be a pretty slippery notion to characterize). But scientific evidence is certainly relevant as well – and your own work on self-regulation and ego depletion will surely inform whatever account of irresistible impulses turns out to be correct. In short, I think the evidence you cite in your post is mostly relevant to question 1, and I would be very interested to know your views on question 2.

[Dan Speak]

Let me recommend Gary Watson's work on addiction and the will. With characteristic philosophical skill and psychological sensitivity and realism, Gary argues that an addiction is something like an acquired appetite. This is, I think, fodder for Roy's free will emphasis. But Gary also notes the unruly nature of our ordinary appetites and allows for their being the sorts of things that get in the way of our freedom in some contexts. So, again, tip of the cap to Eddy's graded picture of these issues.

[Tom Clark]

Roy,

It seems to me the first step in this inquiry is to define clearly what you mean by free will. When you say that one view of addiction is “that there is no loss of free will and that maintaining addiction is voluntary behavior,” then on this definition acting freely means acting voluntarily, without coercion, doing what one wants on one’s own recognizance. So to say that the free will position on addiction is supported by the fact that “most addicts do quit” is just to say that eventually that’s what most addicts end up wanting to do, and voluntarily choose to do. Fair enough. But of course it isn’t as if they could have freely chosen to quit *at any time*, that is, it isn’t that *at any time* they could have wanted to quit more than continue but simply failed to muster that desire. No one has that sort of magical self-control. Rather, in each case of quitting, there’s a causal explanation of how the desire to quit grew stronger and eventually expressed itself in the voluntary choice to quit.

The explanation has completely to do with changes in the addict’s brain chemistry and circumstances, such as avoiding the cues and the availability Neil mentions, being forced by parents, peers, and having children to take on adult responsibilities, finding better things to do, “hitting bottom,” discovering AA, getting the right pharmacotherapy and counseling, etc. So to chalk the decision to quit up to the addict’s free will in the sense you mean - voluntary choices and behavior - is shorthand for this causal explanation, which of course implicates the addict’s whole situation, internal and external. Yes, addicts can gain control over the desire for drugs (which is just to say other desires come to predominate), but only if the conditions are right to generate that shift in their motivational economy. That shift can’t and doesn’t magically materialize ex nihilo from within the addict. To think that’s possible is to define free will contra-causally, as *libertarian* free will, which is not what you mean by it.

So to answer your question, no: addiction doesn’t make free will defined as voluntary action disappear, but rather radically alters the balance of rewards in favor of continued choices to use drugs. A recent book that explains addiction in terms of the rather surprising, often self-undermining dynamics of voluntary behavior is Gene Heyman’s Addiction: A Disorder of Choice, reviewed at http://www.naturalism.org/Heyman.htm . Since the disease model of addiction (Heyman’s target) is received wisdom in the treatment community the book hasn’t been received that kindly, but I highly recommend it. I think Heyman’s evidence-based theory (he cites the data you do on quit rates, and more) fits well with your essentially physicalist, naturalistic understanding of self-control as a limited but improvable (given the right conditions!) resource.

[George Graham]

Roy (if I may),

You ask what the addiction literature teaches about free will. On my reading of the literature, one thing it teaches is that addictive behavior patterns are sustained or maintained by an addict’s choices. Various choices are made by an addict, though just what sorts of choice depend upon differences in the forms or types of addiction. Addiction to heroin is not the same as addiction to cocaine or to gambling or to shopping. It also depends upon whether a person, an addict, is motivated to refrain or to quit, and the nature of their failure in relapse, if or when they relapse. As we know, more than one motive can exist at one time. When motives co-occur, impulses and inhibitions may compete. Some addicts resolve this motivational conflict one way, others another way. Some relapse in spite of their best and most sincere efforts to the contrary.

Are the choices made by an addict in an addictive pattern ever free? Some choices may qualify as free depending upon what we mean by ‘free’. If an addict in making a choice is sensitive to positive and negative reasons or incentives in the same sense in which a free choice is sensitive to such reasons, then such a choice is free even if it is severely hampered or constrained by all sorts of personal and contextual factors.

In the context of popular speech, it is common to refer to people as addicted to just about anything – television, pornography, inhalants, chocolate. If we are to make progress in our understanding of the lessons of addiction for freedom of the will, as well as for other topics, we must be sure that we are using a more sound and sensible notion of addiction than that represented in everyday discourse. To that end, Robert West’s Theory of Addiction (Blackwell 2006) is a helpful resource. It asks the question of what makes for a good concept of addiction and offers useful suggestions.

Finally: Jeffrey Poland and I have recently put together a collection of original papers, by others, on addiction and responsibility that The MIT Press will soon be publishing, and a number of the papers in that book address the choice based dynamics of addictive behavior. The editorial introduction to the book surveys the lay of the conceptual land.

[Noah Te Stroete]

Roy,

If your question is specifically, "Does free will in regards to the addicted agent's decision to use drugs disappear?", then the answer is definitely "no". The behavior of actually consuming drugs is much too complicated neurologically for a chemical such as nicotine that interacts with neuroreceptors in the brain to "determine" this complicated behavior. The nicotine may cause an itch, as you suggest, but it does not causally determine the intricate and complicated neurological process that is involved in drug consuming behavior. Now, what does cause this complicated and intricate neurological process of behavior? Well, that is another can of worms althogether. Could it be a combination of my genetics and my personal history along with everything that is occurring in my environment at the moment? Well, that is just too many factors for the human mind to consider. Or is there libertarian free will afterall? Maybe there are just boundaries that I may navigate between but I am ultimately free to choose how I navigate between the boundaries (the boundaries being the laws of nature and my own limitations), and there is nothing causing my decisions except my on agency (whatever that is).

[Roy Baumeister]

Responding to Peter Reiner, I agree that one could make a case for localized loss of control -- but one can make the contrary case as well. All my research has indicated that the powers of self-control and volition all draw on the same resource. If something depletes that resource, then control over everything is compromised. Hence it is not plausible that smokers would lose control over smoking but not over, say, eating. If we preserve the distinction between impulse and restraint, then addiction is all on the impulse side. It means a strong, recurrent impulse. One's powers of restraint remain the same. To be sure, you can use up your powers of restraint fighting the one impulse, but then they would be unavailable for fighting other impulses too.

Responding to Nicole Vincent:
Excellent distinction. You are right, I was dealing with the majority of cases, as social scientists are inclined to do, and I recognize that philosophers may be more interested in the extreme ones, about which I have been cautious. (I tell my graduate students that "social scientists should never say always" but few of them get the joke.) Personally I could believe that there would be some cases in which the person is physically unable to resist particular strong impulses. These would be exceptional, to be sure, and might depend on the physical bodily resources on which volition depends being severely depleted. To be sure, a person in such a state would be a wreck, hardly able to make a sandwich. But it could be possible. Also there were some stories in the news last year about a new drug in South America that supposedly took away the person's free will. These are worth exploring, perhaps.

Responding to Tom Clarke:
Very good points. And Heyman's book is a terrific read, and was the source of some of the information in my blog. Another, more strident source is Schaler, Addiction is a Choice. One small point: I'm not sure addicts can really 'gain control over the desire.' To my way of thinking, the desire itself originates in automatic processes and cannot be controlled. The addict merely controls whether he or she acts on that desire.

[Tom Clark]

Roy, thanks for the feedback. What I see in your model is the idea of a physically-based conscious (non-automatic) control or volitional capacity that's independent of desire:

"the distinction between impulse and restraint"

"...cases in which the person is physically unable to resist particular strong impulses. These would be exceptional, to be sure, and might depend on the physical bodily resources on which volition depends being severely depleted"

"desire itself originates in automatic processes and cannot be controlled. The addict merely controls whether he or she acts on that desire"

In your model, this capacity (dependent on glucose levels if I remember correctly) is limited and can be depleted or augmented, it isn't a situation-independent or biology-independent capacity. So I don't see how cases in which the person is physically unable to resist particular strong impulses have to be exceptional, as you say in response to Nicole. In fact they are (unfortunately) quite common: all addictive behavior can be accounted for on your model as cases in which the physically-based control capacity is insufficient to counteract or control an impulse, also physically based. There is no non-physical standpoint from which the addict could have acted otherwise, given his impulses and control resources. If an impulse is not resisted, this means the control capacities didn't exist to resist it, not that the addict simply didn't choose to exercise them. The failure to exercise control *is* the lack of such control capacities. To think otherwise is to suppose there's something going on independent of one's brain-based motivational tug of war between impulse and control.

Glad you liked Heyman, quite the breath of fresh air into the addiction debate.

[Roy Baumeister]

Reply to Tom Clark’s most recent comment:
Tom, your response is very nicely articulated. I disagree fundamentally, but I admit that your point is plausible. (It is also what addicts would like to have us believe, insofar as it frees them of responsibility.) If I understand you correctly, you are saying that all addictive behavior occurs because the person is unable to resist. In essence you are answering “Yes” to the original question of my blog, as to whether free will disappears in addiction. (I apologize if I have misunderstood you.)

My view is that I do believe in choice. Hence there is a big difference between the (rare) case of someone so depleted he or she could not resist an addictive craving, and the (common) case of someone who could resist the craving but prefers and chooses to indulge it.

In my research on depletion, most depletion effects are conservation effects, as Muraven has shown. People still have the wherewithal to exert control and resist temptation, but often they tend to favor giving in so as to preserve their somewhat depleted energies rather than expend more energy. If there is enough at stake, they will still exert control (and deplete their willpower further).

Anecdotally, I have been asking smokers and ex-smokers what it was like. So far they have all said they always thought they could resist smoking on any particular occasion (and often did choose not to smoke). You may say they are all mistaken – every time they smoked, they could not possibly have resisted. And hard-core determinism, as I understand, says pretty much the same thing, that whatever happened could not have been otherwise. But what they are saying indicates that they had control and thus is fundamentally different from what I described as being so depleted that one was physically powerless to resist the impulse.

Someone who reached the extreme, super-depleted state of being unable to resist a simple urge to smoke a cigarette (or play a video game, watch porn, snort cocaine, or whatever the addiction is) would be incapable of most acts of volition. That person would be unable to do anything that was not pure habit and/or impulse. Making a sandwich, finding a parking space, buying a sweater, drawing a picture, or reasoning logically would all be equally impossible, unless every step was a well-practiced habit. That’s why I suggested this state would be rare.

I am now surprised you like Heyman’s book, because he argues forcefully against the position you espouse, again if I understand you correctly. He says addicts do have the capacity to resist but choose to give in.

[Tom Clark]

Roy,

As a physicalist about self-control, it’s interesting you resist determinism, since, quantum effects aside, there’s no reason to suppose that on a physicalist understanding one’s exercise of control could have been otherwise given the internal and external conditions in play. If you think the addict could have resisted an impulse but didn’t, then you have to explain *why* he didn’t. On your theory, there’s a specific amount of physically-instantiated control capacity available to the addict, which is no doubt true in some sense. But here (“I do believe in choice”) it sounds like you’re sliding into libertarian free will, which is different from free will as voluntary behavior which is what I thought you meant by free will. It sounds as if you think that, in addition to the addict’s control capacity, there’s some sort of capacity to use one’s control capacity or refrain from using it (“a choice to preserve their somewhat depleted energies”). If so, that capacity too has to be included in your theory otherwise you can’t explain why in a given situation an impulse is or isn’t resisted except in cases of total physical depletion.

It seems to me that explaining addictive behavior is a matter of understanding the contest between a neurally-instantiated impulse to use drugs and a countervailing neurally-instantiated control capacity not to. If there’s a further capacity that adjudicates the contest, its influence, its decision *also* has to be grounded ultimately in some sort of physicalist mechanism, otherwise its mode of operation is obscure and shouldn’t play a role in naturalistic explanations of addictive behavior. Why should we believe that a smoker could have resisted the urge to smoke but did not based on his avowal that he could have, but didn’t?

Heyman isn’t a (libertarian) free willist. On page 114 of Addiction: A Disorder of Choice he says “The facts and logic presented in this chapter do not say that it is always the case that someone addicted to drugs can choose to quit. Voluntary behavior is not the same as [libertarian] free will…Thus the conclusion that addiction involves voluntary drug use does not imply that quitting will be easy, and in some situations it says that it may be impossible.” I’m very happy with Heyman’s analysis precisely because it’s completely and transparently causal in showing how the capacity to resist impulses has been undermined by drugs, which explains *why* addicts voluntarily choose to continue to use them.

Admitting the (likely) fully-determined nature of addiction is simply to acknowledge that human behavior is a matter of cause and effect, which if we understand better will give us more control in preventing, controlling and curing addiction. Admitting we don’t have (libertarian, contra-causal) free will doesn’t entail that addicts can’t be held responsible. It only means we should hold them responsible as compassionately and non-punitively as possible, since we understand they couldn’t have done otherwise in their situation, and that we wouldn’t have done any better were we sufficiently in their environmental and biological shoes. See for instance http://www.naturalism.org/addictio.htm#Causality

You might say (and have said on your Psychology Today blog a while ago, url below) that resisting an impulse is a matter of response *probabilities*, not determinism, but that just says there’s a range of uncertainty or non-determinism in 1) how the addict responds in a specific impulse-resisting situation (reproduced exactly in a thought experiment) or 2) the variability of exact situations that an addict responds to, since we can’t reproduce a given situation exactly, or 3) in our inexact *measurement* of that response. But in none of these cases does the probabilistic view of resisting an impulse give the addict further control over what’s available on a deterministic view as far as I can see. http://www.psychologytoday.com/blog/cultural-animal/200902/just-exactly-what-is-determinism-0/comments

[Noah Te Stroete]

Tom,

I am curious to know if you believe that a physicalist position necessitates a deterministic position. If so, I suspect it is because every event has a cause? What if I were to espouse the position that physicalism is true, but causes and effects do not line up in a nice 1 to 1 correlation? Perhaps the closer someone is to omniscience, the more easily one can say that things are determined, not being of a 1 to 1 correlation of causes and effects, but because of the omnicient's knowledge of the entirety of existence with time being an absolute 4th dimension that is just as set in stone as the three dimensions of space, an eternal presence that we as temporal beings can never capture in our temporally bound brains.

Forgive me for being so obscure, but perhaps given our capacity for knowledge, and my belief that every event has perhaps a multitude (perhaps thousands, millions) of "causes", that physicalism can be true and libertarian free will can exist in some sense at the same time. If my decision not to smoke at a given time is determined by my brain state at the moment, which is determined by my genetics and my personal history, in turn determined by the history of my ancestors leading up to my existence and everything that happened to them and the world they lived in, then clearly not every event has a single cause (perhaps the lack of nicotine stimulant withdrawal symptoms). I stress: Given we cannot trace this infinitely complicated web of existence from moment to moment, there is no way to know if I can be an agent of ex nihilo decision-making or if it is all determined. However, it is much too simplistic to simply say that physicalism and determinism imply each other based on that "the (likely) fully-determined nature of addiction is simply to acknowledge that human behavior is a matter of cause and effect, which if we understand better will give us more control in preventing, controlling and curing addiction." I simply have trouble believing that we can ever fully trace that multitude of causes that determine brain states. Perhaps some day with a super-computer about the size of a planet we might make some headway into this baffling enigma.

[Noah Te Stroete]

I really do think it would take a super-computer that is lightyears ahead of anything we have now in order to model our world and the conscious beings in our world. Only after entering in all the known inputs including the TRUE laws of nature (which we are still working on) would we be able to make a determination whether or not everything is determined including our decisions arising from our brain states, or whether, on the other hand, the model suggests that we really do make ex nihilo decisions of our own volition and agency.

Does anyone else here really think we can solve this problem without first having a thoroughly tested and accurate philosophy of mind, a unified theory of physics, and a giant super-computer?

If not, what experiments could we possibly do today to resolve this issue? Thoughts?

[Roy Baumeister]

good point. in practice no computer could ever do this. youd need a computer several times bigger than the universe itself. plus, there is no way to ever know whether we have understood (and programmed into the computer) ALL the laws of nature.

so to me this deterministic assumption is looking more and more like mental masturbation.

[Fritz Warfield]

I'm late to this discussion but want to raise one issue.

Some (examples, Tom Clark and George Graham above) think that to answer whether free will disappears given addiction one first needs to say what one means by "free will". This kind of observation is also often made about how to go about answering whether free will is compatible with responsibility and whether free will....

And I confess that it seems like a plausible observation. But I wonder whether it really is plausible under assessment. Here are some other places where the same kind of observation seems out of place:

1. Epistemolgoy. Does failing to have evidence that entails one's belief preclude one's belief being knowledge? To answer this first define "knowledge". (or "say what you mean by" 'knowledge')

2. Mind. Is non-reductive physicalism consistent with mental causation. To answer this first define "causation" (or "mental causation"). (or "say what you mean by...")

3. Ethics. Does acting from a bad motive preclude acting rightly? To answer this first define "right action".

It's a good thing we don't have to give a full definition of knowledge in order to argue that some things are and are not consistent with having knowledge. And if the call isn't to "define" the term but simply to "say what one means" by it then one should wonder how *that* will be relevant to claims about knowledge in the event that what one says one means by it is an account straightforwardly open to counterexample?

I don't have a verdict to pronounce about this issue -- as I say, saying what one means by "free will" does in some ways sound like a plausible thing to do if one wants to explore whether addiction precludes freedom. But in some other ways it doesn't seem like a plausible thing to do.
Perhaps someone with a better grasp of the meta-philosophy involved here can enlighten all of us?

[Tom Clark]

Fritz,

Since Roy presumably has some idea of free will in mind when thinking about whether addiction overrules it, I think it’s fair to ask him for clarification, not to kick the issue upstairs to meta-philosophers. The ambiguities in Roy’s conception of free will are the focus of papers I mentioned in the last comment on “Engaging the public about free will,” including his target paper and responses, see http://www.springerlink.com/content/1t786q7542800j6x/
http://philpapers.org/rec/NAHWWL and
http://web.mit.edu/holton/www/pubs/Baumeister.commentary.pdf

Unless Roy states clearly what he means by free will in terms of actual human capacities, whether as an independent variable or explanatory target, then it shouldn’t be a focus of research or speculation involving real-world behavior (as opposed to the focus of general philosophical questions similar to those you raised about knowledge, mental causation and right action). Same applies to anyone doing research on free will or beliefs in free will as it affects issues involving things like addiction – what is being refered to? Something contra-causal or not, or what? Roy’s physicalist notion of an addict’s limited self-control capacities strongly suggests that a causal analysis of addiction (for instance using something like Heyman’s model of voluntary choice) should eventually be possible to explain why addicts quit or don’t in a full range of situations. But on the other hand he says that addicts have the additional capacity to quit *at any time* unless their control capacities are completely depleted. He calls that capacity their free will. How does it work, and can it be included in his theory of self-control or not?

Noah,

I don’t believe that physicalism necessitates determinism, but don’t see much evidence for libertarian free will (LFW). Your argument for LFW is an argument from ignorance: since we can’t “ever fully trace that multitude of causes that determine brain states,” for all we know we might be libertarian agents. But we don't have to have perfect knowledge of all the actual causes to make the plausible claim that there were (or are) sufficient causes. Asking for such knowledge in advance of claiming causal regularities would stop science dead in its tracks. The assumption of a near enough, for all practical purposes determinism (that is, *maybe* some inherent non-determinism, and definitely some epistemic uncertainty, but not enough to block eventual explanation, prediction and control of a phenomenon) is the default assumption when investigating addictive behavior. Otherwise one would give up too soon in attempts to ascertain causal regularities and instead chalk things up to LFW, when in fact they are results of a failure to investigate and/or measurement error. Such a failure, and the possibly unwarranted assumption of LFW, impedes understanding the genesis and maintenance of addictive behavior, making addiction harder to prevent, treat and cure. It also, perhaps mistakenly, assigns ultimate responsibility to the addict which helps to justify punitive attitudes and policies. There has to be better evidence of libertarian agency besides our inability to perfectly understand and describe all causal chains and influences.

Roy,

As you explain in your post “Determinism: the good, the bad and the ugly,” the free will you think is worth wanting and worth calling free will is that capacity which, by evading reductionism and predestination, permits the addict to “exert control in the sense of choosing and steering among multiple possibilities” in way such that he’s not merely a “throughput.” But the question for those trying to understand addiction is *why* addicts choose they way they do, why a *particular* possibility became actual. Researchers on addiction don’t (or shouldn’t) prejudge the question by saying the addict can’t be “merely” a throughput, or that reductionism or predestination must be false because certain intuitions about choice and free will might be upset. Our decisions about such matters have to be based on what the evidence shows about human behavior. If your concept of free will, like your concept of self-control resources, can play a role in explaining why addicts behave as they do such that it makes possible better prediction and control of addictive behavior, it will gain empirical support and add theoretical value. If not, which is what I fear given your posts here, it will remain an obscure non-scientific alternative explanation that somehow separates the addict from his biology and environment while offering little or nothing for prevention and treatment. I’ll be happy to be shown wrong about this.

[Noah Te Stroete]

Tom,

I think you misunderstood my post if you think I was making an argument for libertarian free will. This was not an appeal to ignorance, as there was no argument being made for libertarian free will. I was simply making the observation that libertarian free will may never be fully discounted. I was not suggesting because it can never be fully discounted that we should conclude that we have it.

Furthermore, even if addiction is nothing but a really strong itch, that in itself is a contributing factor to addiction that should not be ignored. A really strong itch can still be just as debilitating to someone. We need not conclude that addicts have no control over there addiction in order to feel compassion for them or to help them to get rid of the cause of the itch.

I am not denying causes, I am simply saying that it does not make sense to insist that espousing libertarian free will is in any way contrary to science. That is no where close to anything I believe. Certainly the itch or the irritation or the strong desire to consume drugs is caused by something that science can figure out, but that does not mean that someone is necessarily coerced by determining forces to consume drugs.

Perhaps Dr. Baumeister has it right that we have a certain amount of self-control that can be used up...and if an addict is using up his self-control on other activities throughout the day, then perhaps s/he is not going to care enough to exert that energy toward kicking the habit.

Or do you deny that Dr. Baumeister's studies are science because he espouses free will?

[Tom Clark]

Noah,

I’m glad we agree that not being able to fully discount the existence of LFW isn’t a good reason to believe we have it. And I agree that we can have compassion for addicts even if we suppose (which of course I don’t) that they could have resisted the impulse to use drugs given the exact conditions in which they failed to resist. But I think it makes compassion at least somewhat less likely since under LFW the addict is understood to be a first cause of his failure to resist, hence blameworthy in a particularly pointed, ultimate sense that those wanting justifications to respond punitively to addicts or other sorts of offenders sometimes cite (in my experience).

It’s an interesting question whether espousing LFW is contrary to science. So long as those espousing LFW as something factually true of human agents (so that it isn’t an empirically empty claim) are willing to have their claims be adjudicated by scientific investigation then there’s nothing contrary to science going on. But if they hold to LFW matter what the evidence, or make factual claims about its existence that can’t be tested, then it seems to me they are parting company with science. Some suppose there are good *non-scientific* grounds for believing in LFW, and they might say those grounds don’t conflict with science but are simply orthogonal to science, on a different epistemic plane so to speak, or a pure deliverance of armchair reason. At that point the question arises of the validity of non-empirical justifications for factual claims, something well beyond the scope of Roy’s original question.

I am of course urging Roy to be fully scientific in his approach to free will, which requires explicit, testable definitions of what free will consists in, such that for instance we could tell whether addiction overrides it. But the impression I get is that he thinks free will somehow trumps or obviates or makes impossible any sort of complete causal analysis (tempered by our epistemic limitations) of why a particular impulse was or was not resisted, which seems to me contrary to science, or at least to go beyond or outside science in explaining (very obscurely!) why addicts behave as they do.

[Noah Te Stroete]

Tom,

Thanks for your latest comment. Now that your position has been clarified (maybe that was my fault for not understanding you) I now see that we are not so much in disagreement. I agree that we should first define precisely what we mean by free will before we appeal to it. And if the definition appeals to supernatural or other unscientific explanations, then it should be discarded as being contrary to a scientific approach. I'm also glad that you agree that espousing LFW may not be contrary to science if it is indeed factually true (which is an open question at this point).

Regards,

Noah

[Fritz Warfield]

Tom,
I don't mind asking people what they mean by a phrase like "free will" when they make and defend various claims about it.
As I said, it seems like a natural thing to do.
But as I also said, and illustrated with examples, this seemingly natural call for clarification seems out of place in various other seemingly similar contexts in philosophy (in mind, in ethics, in epistemology).
So I wonder whether it's a good and sensible move here but not there (and if so why) or whether it's not a good move here or whether it is mistakenly thought not to be a good move in those other debates.

As I said, I have no answer on this one, only uncertainty for now.

[Philoponus]

Do addiction researchers use instruments for rating the degree of addiction/dependency? They must, but I’m googling without success. Mental health researchers use a variety of scales including the Hamilton Depression Index for rating major depressions. We hear that there is a strong correlation between (advanced) alcoholism and major depression, though I can’t track down an attested r value right now. My point is that several previous comments have suggested that “free will” needs to be defined in such a way that it can be studied as a testable variable. Yes! Then and only then correlation studies could be pursued relating it to addictions, mental illnesses, etc. Can anyone tell me whether (and where) this kind of work on “free will” is actively being pursued?

[Tom Clark]

Philoponus,

Re diagnosing addiction, there's the Addiction Severity Index, see http://www.tresearch.org/ASI.htm and http://www.densonline.org/DENSASI.pdf . Also see http://www.tresearch.org/resources/pubs/26_DSM-IV.pdf re the DSM-IV diagnostic criteria for dependence. Note that addiction (compulsive drug seeking and use despite serious negative consequences for the addict) is now standardly distinguished from dependence (withdrawal symptoms upon cessation). For the latest on diagnostic instruments, maybe query http://www.caas.brown.edu/ , http://www.JoinTogether.org , or http://www.hazelden.org/ .

I don't know of any scales that measure the extent of one's free will, but of course some in the treatment community suppose it varies inversely with the extent of one's addiction. As for defining free will, Al Mele says in a press release about his Templeton project: "What I want to do is make significant progress on discovering whether we do or don’t have free will.” This suggests his group should have at least some working definitions and hypotheses concerning it, and indeed they cite the usual compatibilist and incompatibilist accounts of what free will (or free action, as they call it) consists in, see http://www.freewillandscience.com/FW_Lexicon.pdf . Since there’s little debate about whether we have compatibilist free will, my take on the project is that Templeton wants to keep hope alive for the libertarian variety, see http://www.naturalism.org/roundup.htm#templeton

[Philoponus]

Thank you so much, Tom. So dependence at least is a testable and measurable property via the ASI. And presumably addiction is also via other instruments. But why aren’t we trying to do this with FW? I’ve been trying to think about Ray’s original question as something we could bring experimental work to bear on. I want to expand “Does FW disappear because of addiction?” into something like “does our capacity to make free choices diminish and eventually extinguish as the severity of our addictions increase?” Good question, yes? And surely some people are already studying how chronic drug and alcohol use cripple our decision making skills: our ability to recognize choices, to frame alternatives, to rationally investigate them, to resist external and emotional pressures, to deliberate and choose and then stick with the choice our decision process identifies as best. Construed in this way, I can understand Roy’s question as an issue on which experimental psychology and neuroscience can and should be brought to bear. What puzzles me is the (apparent) lack of interest in the philosophical debate about trying to nail down FW as something science can investigate. Why does FW have to be some untestable metaphysical construct and not a property that we can test for and measure via scientifically respectable instruments?

[Tom Clark]

Philoponus,

You’re of course right that important work is being done to see how addiction undermines various control capacities, including cortically-based decision-making and impulse control, both of which get compromised significantly in late stage addiction. If you define free will as having a certain minimum of such capacities (as do some compatibilists), then at some point addicts lose their free will, or you could look at it on a continuum and say they have much less of it. Nora Volkow, head of the National Institute on Drug Abuse, talks this way about addiction undermining free will. So on this definition we could perhaps eventually test for and measure free will scientifically if we can get objective measures of the capacities in question.

But of course free will is often conceived of, defined as, something *independent* of brain-based control capacities, something that science can never pin down but which can be deployed by the addict to quit any time. If you’re a naturalist about human agency, this is a very difficult proposition to maintain. But most folks aren’t naturalists, so wouldn’t buy your idea about the variability and measurability of free will. Even some naturalist compatibilist philosophers might resist the idea that free will can be operationalized, not to mention the libertarians, naturalist or non-naturalist. Many folks are attached to the idea that free will is precisely that “untestable metaphysical construct” which permits us to rise above our natures, or nature, or causal laws, if only we would choose to exercise it. So I doubt that free will will be generally conceived of as a scientifically measurable, physically-instantiated set of control capacities any time soon.