Blog Coordinator

« Workshop on Free Will, Metaphysics, and Science | Main | The Free Will Inventory »



Feed You can follow this conversation by subscribing to the comment feed for this post.

Hi Hanna,

Great post. Really interesting stuff. I'm inclined to agree with you that addicts typically have the ability to do otherwise. Nevertheless, I wonder what theoretical conclusions we can draw from this. In particular, it's not clear to me that all this shows that theories of action that link agency with the ability to do otherwise have nothing to worry about from considerations having to do with addicts. You write:

"Addiction doesn’t help us make sense of the idea of action in absence of the ability to do otherwise, for addicts have this ability, however hard it is for them to exercise it and however many reasons they may have not to, as much as any of us do. Hence one source of pressure on any analysis of action that links it fundamentally to choice between alternatives and a corresponding notion of control is relieved."

I guess I'm failing to see how this follows from what you've said about addiction, even granted that what you say is true. After all, you also say that "the attribution of any ability to any person needs to allow that there can be extreme circumstances or particular occasions when it cannot be exercised. This is as true of addicts as the rest of us: I am not claiming that every addict is able to abstain on every single occasion of use."

But if we admit that it's possible for someone to act (by, say, taking a drug or a drink) yet lack the ability on that occasion to do otherwise, doesn't this suffice to show that we can make sense of the idea of action without the ability to do otherwise?

Thanks so much for that excellent post! Very clear and informative. I had similar experiences when working with adolescents that had to overcome addiction.

I think you are absolutely right regarding addiction not providing a helpful place to look if one is searching for an example where one acts but does so without free will. Harry Frankfurt's "Freedom of the Will and a Concept of a Person" comes directly to mind. I think similar things can be said regarding the psychopath and the way the philosophical literature (for the most part) tends to mitigate their moral responsibility for similar reasons. I did had a brief question though.

I have been thinking that moral responsibility might just be a binary concept and not a thresh-hold concept that comes in degrees. Most would likely disagree, but I was wondering what you think. Most philosophers would reject the claim that addicts are the same regarding how free their wills might be, many might claim that they have "less" free will which is why they are less blameworthy. Being harder to stop makes overcoming their addiction more difficult and possibly more praiseworthy than those of us who refrain from taking the drug after only one use. Would you agree with those philosophers who take addicts to have less moral responsibility than non-addicts, or do you think that moral responsibility is a binary concept and that both addicts and non-addicts either have the free will required to ground moral responsibility or they don't?

I ask because when you wrote "We stand or fall on this score together" regarding the addict having the same abilities as non-addicts, you seemed to suggest the binary understanding of moral responsibility. Or, am I searching too hard for support for the binary position (which seems to be the minority view) that I am suggesting?

Hi Hannah,

Great topic. I'm wondering if you could clarify what you mean by a "disease of compulsion." It sounds like you're saying that having a disease of compulsion entails that one lacks the ability to do otherwise, with regard to the object of their compulsion. My worry is that this standard is much too strong, such that nothing would satisfy it. For example, surely the obsessive-compulsive isn't incapable of doing otherwise in the modal sense you have in mind. Rather, like the addict, it's that there is extraordinary, seemingly overwhelming pressure in favor of the compulsion.

Sufferers of OCD don't lack the ability to do otherwise in the modal sense you seem to have in mind, but I would think that OCD is a disease of compulsion if anything is. My concern here, then, is that if nothing counts as a disease of compulsion, I'm not sure how useful it is to deny this status to addiction. Maybe the moral of the story here is ultimately that there are no psychological conditions that give us an example of action without the ability to do otherwise. But the alternative, which also seems plausible to me, is that we ought not interpret ability in the modal sense you are utilizing. Just curious what you think about this.

Hi Hanna,
Thanks very much for the post. It was very helpful and informative.

I do, however, want to press back on your solution to the puzzle of drug use in the face of terrible consequences. While I think the story about putting addiction in context is helpful (and quite plausible so far as it goes) I'm not sure it goes far enough. Consider the following case:

A close friend of mine has a family member who is addicted to smoking. She comes from a middle class family. Has several loving children and a healthy marriage. However, she has recently endured a series of strokes which her doctors attribute largely to her smoking for the last 25 years. Her condition is not severe enough that she cannot recover to a large degree, but doctors tell her that she will continue to be at risk so long as she keeps smoking. She has no history of mental illness, and her alternative goods seem to me quite compelling (relief from her physical ailments, better familial relationships, living long enough to spend time with her grandchildren, etc.). Yet she continues to smoke, and her prospects for recovery continue to stagnate.

This strikes me as a case where context doesn't offer much of an explanation for the intransigent drug use. It might be possible to explain this as a sort of epistemic failure (perhaps she simply doesn't believe that her health problems are caused by the drug or could be alleviated by quitting), but surely we can think of other real world cases where this explanation is not available. I'm inclined to agree with you that addicts do have the ability to do otherwise, but I'm not sure the solution to the puzzle that you offer is able to stem the worry in all cases.

Hi Justin (Capes),

Thanks for yet another great question! Here’s how I reconcile the claims: I’ve not here claimed that we can’t make sense of the idea of action without the ability to do otherwise. I’ve claimed only that addiction does not offer an example that helps us. People typically think it does because consumption looks like normal action that is caused by beliefs and desires – but because the desires are ‘irresistible’ there is no choice. The possibility that addicts, like the rest of us, sometimes cannot exercise the capacity to do otherwise does not in my view rest on the idea that, in those contexts, their desire is irresistible. Rather it rests on the kinds of circumstances that affect us all: e.g. (in addition to circumstances such as sleep, being forcibly constrained, etc.) certain psychological and physical states, such as extreme panic, fear, rage, stress, and exhaustion, may temporarily remove executive function or general cognitive capacity, so that the resulting behaviour is on automatic pilot, typically looking like a kind of short-lived frenzy or alternatively paralysis. This can happen to addicts as much as to anyone - indeed, there is some reason to think that on the view I’ve put forward of addiction, where drugs and alcohol are a form of self-medication, abstinence may increase the likelihood of it occurring because of the increase in un-medicated psychological distress addicts must then tolerate. However, whether or not this last point is true, the kind of behaviour that results from these states is neither typical of consumption nor is it clearly ‘action’ – it seems more to approximate a kind of involuntary automatic reflex. So, on the one hand, this is not the kind of behaviour that looks like ‘action’ on the outside – it does not look like it is straightforwardly caused by beliefs and desires. But, even if I were to concede this, it is still not addiction or irresistible addictive desires per se that is doing the work here – but rather the kind of psychological or physical state that affects executive function or cognitive capacity quite generally.

For the record, I do think we struggle to make sense of action apart from the ability to do otherwise – I’m sympathetic to the way both Maria Alvarez and Helen Steward have argued for this claim. But the point of this blog was only to argue that addiction doesn’t help. I hope that speaks to the tension you pointed out - thanks again for the question.

Hi Justin (Cauotte),

Thanks for this sympathetic comment and question which, if I’ve understood it right, I want to answer both yes and no. I’ve always loved this quote from Bernard Williams (from his essay ‘How free does the will need to be?’) which I think explains why this is in fact the right answer: ‘Why does freewill, unlike freedom, not come in degrees? Presumably it is because its assertion consists only of an existence claim. How exactly that claim should be expressed is notoriously disputed, but it is something to the effect that agents sometimes act voluntarily, and that when they do so they have a real choice between more than one course of action; or more than one course of action is open to them; or it is up to them which of several actions they perform… [this] merely requires that there be, in the appropriate sense, alternatives for the agent, and that it is indifference to their number, their cost, and so forth. That is why the freewill that it introduces is different from the freedom that comes in degrees and is opposed to constraint’. Free will is binary in my view. It depends on having a real choice, which amounts minimally to this: you can act or you can refrain from acting. That is compatible with there being literally nothing else you can do but opt for one of these two alternatives. Typically however there are more than two alternatives: you can act one way or not, act another way or not, etc. How many alternatives, and how good they are, is relevant to how free you are and that does come in degrees. So, addicts, for instance, typically have limited and poor alternatives to drug and alcohol consumption, even if they could not use and so nonetheless have free will. They have free will like us but are less free than many of us.

So let’s turn now to responsibility – does it come in degrees? Yes, in my view, because it tracks not just free will but in addition freedom – questions of e.g. limited and poor alternatives, mitigating circumstances, and the effort and cost needed for control (among other things) can all affect degree of responsibility. Yet free will is a necessary condition, and that does not come in degrees. So yes and no: you either are responsible or not, but if you are, there is yet a question of how much.

Hi Eli,

Thanks very much for this comment. I think I can answer it straightforwardly but come back to me if I’ve not. I think we need to treat psychiatric disorders case by case and not generalize from one to another, but I absolutely agree with you about OCD and indeed all the disorders of agency I have some clinical experience and empirical knowledge of: there is no inability to do otherwise. But, that said, I also think that some pockets of clinical culture and the scientific literature, much of popular culture, and a great deal of philosophical literature, does consider addiction and other forms of psychopathology a “disease of compulsion” in the modal sense, claiming, perhaps for effect or carelessly at times, but nonetheless in other places absolutely consistently and clearly, that addicts and other patients lack the ability to do otherwise in this strong sense. (For some philosophical examples, you can look at the first paragraph of my paper ‘Psychopathology and the ability to do otherwise which is here ). So, yes, I think there probably aren’t any diseases of compulsion in this sense, but then it’s important that everyone stop claiming that there are! (And for me, this is not just a theoretical issue: encouraging the idea that patients lack the ability to do otherwise may adversely affect their prospects of recovery, as it puts them into a 'sick role' which is both disempowering and a bar to the rational formation of intentions to do things differently - as you can't rationally intend to do that which you believe you are unable to do.)

I don’t quite see how the conditional or dispositional analysis is going to help make sense of the way in which agency is impaired but not extinguished in addiction and other forms of psychopathology – but do say more if you’d like or I’ve misunderstood what you meant to be implying.

Wonderful post, Hanna! Same with the comments so far and the responses.

This is totally off the wall but I wonder what impact your observations have about xphi research on the concepts of "free will" and the "ability to do otherwise." For instance, you write:

"Additionally, when all patients who suffer from behavioural problems, including addicts, say [they] can’t do something, often what they mean is not that it’s impossible, but just that it’s hard, or that they really don’t want to, or that they have really good reasons not to. So, as important as first-person testimony is to understanding how individual addicts feel and what their experience of living with addiction is like, you can’t read irresistibility of desire or impossibility of doing otherwise off of the way addicts use 'can’t'."

Likely something similar is true of all of us, as you indicate. This has got to have some impact on xphi studies on these matters – for you’re essentially saying that we often incorrectly self-report our assessments of our ability to do otherwise – and I wonder if you have anything to say about that.

Second, WRT why addicts don’t seek withdrawal there are other matters that you are not considering here – such as social stigma. Playing devil’s advocate, why can’t the social pressures that come with revealing to others that you are an addict inhibit the addict from seeking treatment? And if that is case, why can’t we say – in a sense – that the addict couldn’t have done otherwise?


You have been an exemplary blogger for asking not just hard questions but harder ones. I have to say I am quite convinced that most cases of addiction may be beside the question of possessing free will.

Is there anything like compulsion beyond addiction? I have in mind cases like Jeffrey Dahmer's, our own Wisconsin home-grown brand of (psychotic?) serial killer (17 victims). I probably need not remind you of the unthinkably abominable acts he committed that exceeded sexual assault, murder, mayhem of corpses, necrophilia, cannibalism, and much besides. He was judged in a 2nd phase mens rea trial on two questions--1, did he have the substantial capacity to appreciate the wrongfulness of his actions, and 2, did he have the substantial capacity to conform his behavior to the law? These questions were assessed by the standard of preponderance in a double sense by a 12 person jury, each juror thus assessing, but the jury itself conveying a verdict by preponderance, each level therefore using the lowest rational standard of proof. The jury unanimously ruled he attempted (though badly) to conceal his crimes, and so the jury easily assessed him guilty on 1. But the jury split on 2, though not sufficiently to declare him not guilty. Still the minority jurors could not believe, preponderantly, that Dahmer could have done otherwise than he did. If one uses rational choice theory to assess gains versus losses as a measure of free will alternatives, Dahmer's payoff of doing those horrible things for merely the sake of doing them is not preponderantly believable on the assumption of rational choice. So his actions were not ones of choice, but, as the minority judged, compelled. (So the verdict of guilty was wrong by the letter of law by that line of reasoning.) I agree with the minority, given the standard of evidence.

Are there thus any instances of what used to be called "irresistible impulse" (and even still assessed as such in jurisdictions like Texas)? At the extreme, should anyone be exculpated by lack of alternative choice?


Ah ha! I've got it now. Thanks. Moreover, I think I agree. However, I am still inclined to say that addicts might provide a slightly different challenge to the idea that we can't make sense of agency without the ability to do otherwise. How? By providing especially clear examples of a general phenomenon, namely, cases in which a person is strongly motivated to X, has no contrary motivation, and, in the circumstances, is incapable of acquiring contrary motivation. In such cases, I'm inclined (as you know) to say that if the person succeeds in doing X, he will have performed an action, but that he lacked the "all in" ability to do otherwise. My thought is that addicts might provide especially vivid examples of people in this sort of situation, though it is a type of situation that even non-addicts might occasionally find themselves.

from addicts to serial killers. extreme cases sometimes can be enlightening but let's not make the problem too distant from where we are [i am assuming none of us are serial killers ]. as hannah has clearly articulated, the notion of the disease/compulsion model of addiction is quite flawed. but isn't the essential point that the addicts we are speaking of are not so different from the rest of us, really? even in the best of circumstances our freedom is conditioned, and our unfreedoms and compulsions often intertwined with our tacit consent? we have desires and feelings, reactions and cravings, and inclinations of varying intensity.... and we go along or resist. the more we allow ourselves to go with the flow of our impulses the more difficult resistence becomes. some things seem obviously negative to us [looking from the outside] and we become disturbed by how much they can get under our skin. the addictions are of this sort, especially opiates. but, at no point is an addict compelled and unfree to the same extent as i am compelled by gravity to fall when i trip over something. the addict has by steps consented to go along for the ride and has chosen to not resist. they are culpable for it, even if we can see the pattern and feel compassion. just as we are culpable for our character flaws and personality quirks even if we have found ourselves being some way before chosing to be that way. we can always choose to restrain ourselves or even change, however difficult. Serial killers, if we must compare them, are only extreme cases here.... still culpable no matter what...

Hi Joe,

Thanks! Some initial thoughts:

I hadn’t thought about a particular link with xphi before. But did you think there was a particular concern here about self-reports of ability to do otherwise, as opposed to more general concerns about self-reports of all psychological states? I guess I think all experimental methodology that employs self-reports establishes only how people think of themselves and their states – and there’s almost always in principle and often in reality a real gap between that and the reality of their self and their states.

Second, yes, there are lots of factors that affect these kinds of decisions, social stigma being one ... social status being another, as in some sub-cultures drug and alcohol use is an indicator of power and in-group status ... no doubt many other things can factor in to decisions to abstain or not, too. I’m of course happy to say that an addict “in a sense” as you put it couldn’t do otherwise than use ...the sense here being cashed out as follows: they couldn’t do otherwise than use without risking, say, stigma or status. That specifies a cost which may act as a disincentive. But that leaves many other senses of ability to do otherwise intact, modal but also non-modal. Does that answer the devil?

Hi Justin,

Good! I think we do agree up to a point. That point comes with respect the meaning or truth of saying that, in given circumstances, someone is incapable of acquiring contrary motivation. I suspect that just pushes up the “doesn’t” versus “can’t” debate from the realm of action to motivation – but that the contours of the disagreement will be roughly the same.

Hi Alan and Leslie,

Thanks Alan for pushing these hard questions and Leslie for starting to respond to them. I’m going to approach these topics somewhat from the side in this response, and see how far that gets us. So, bear with me ...

In general, I have a policy of trying not to write about kinds of patients and disorders where I have no real clinical experience or first-hand knowledge. I’ve visited high-secure hospitals and prisons, and worked with some patients in the community who have some psychopathic tendencies and have committed violent crimes. But I don’t feel that is sufficient to write in a way I am comfortable with about psychopathy. When I read moral philosophy that appeals to “the psychopath” I often feel sceptical that this person bears much resemblance to real men and women in hospitals and prisons – it seems a bit like an imagined persona, like the way “zombies” are conceived of and used within philosophy of mind. That is OK so far as it goes but not if it is taken to tell us anything about real people and what to think and do with them.

So with that in mind, let’s turn to Alan’s question about whether Jeffrey Dahmer was a victim of “irresistible desires”. I’m going to put aside Alan’s parenthetical remark that Dahmer might be psychotic, for if he is that completely changes the landscape of mental disorder as well as criminal culpability.

I’m not sure quite what the importance of the standard of preponderance is for you Alan in your comment, but I do want to say something about the idea that it’s not believable that anyone could do these things on the assumptions of rational choice. I think that’s exactly the idea lying behind the view of addiction as a disease of compulsion: this behaviour is so puzzling, it seems all costs and no benefits, so how could it possibly be explained by choice? If there was an alternative, the addict would choose otherwise, so it must be they are driven by irresistible desires. My post was, in a sense, intended to draw on my own clinical experience as well as empirical data, to paint a picture which showed why there are benefits, not just costs, to addiction – in essence, to make the choice to use understandable, if not rational. And, as Leslie points out, that makes addicts like us.

The first question I’d want to ask about Dahmer – and one which I’m not in a position to answer, not just because I don’t know enough about him as an individual, but also because I don’t have enough clinical experience and empirical knowledge of real psychopathy – is whether there is a picture to be painted that makes his acts more understandable to us, if not rational. What I can say is that what experience I do have suggests to me that people who have suffered terrible physical and sexual violence, torture, cruelty, abuse and neglect, by no means inevitably but sometimes do go on to be perpetrators of these very things, in a way which is connected to their past experience as victims. Some of the most terrible examples of this come from children, who, having been sexually abused, then, while still themselves children, start to sexually abuse other, younger children. There really can be such cycles of violence, neglect, abuse.

I’m not here suggesting that when adults perpetrate terrible wrongs and harms, their past experience as victims lessens responsibility or reduces culpability – absolutely not (although obviously we may want to think differently about children who are perpetrators, or at least leave open that possibility). What I’m suggesting is that there are a number of ways that we may be able to make sense of the perpetration of terrible wrongs and harms if we can face up to looking at how it may develop in response to the psychological trauma of being a victim and hence have a certain kind of psychological function for the perpetrator – for instance, and here I'm really just speculating, it may offer not just pleasure in the act, but revenge, re-enactment of past abuse, and possibly a defence against one’s own earlier powerlessness.

I cannot emphasise enough that I make no claims here about psychopathy in general or Dahmer in particular. What I’m suggesting is just that I don’t think that our – or the minority of the jury who held that Dahmer was subject to irresistible desires – inability to understand and make sense is an indication that there is no understanding or sense to be had.
So I am inclined on balance to agree with Leslie that, if Dahmer acted, in a deliberate, diachronic, planned fashion, so we appear to have action (rather than the sort of automatic behaviour, such as frenzy or paralysis, I alluded to in my first reply to Justin Capes) then he could have not done it. I do think we struggle to make sense of the very idea of “irresistible desire”. But in ways more importantly, at least given my interests, I think we just don’t know what to think about Dahmer and how to understand the human capacity for such terrible acts. It’s natural to hope that perpetrators like Dahmer “can’t help it” or to think that “evil” is literally a kind of madness. But I don’t know if that’s right. And maybe we hope this in part because it's terrifying to think that such acts are ultimately psychologically understandable - because if we can explain them on the assumptions of rational choice or normal human psychology, that places the acts and the perpetrators of them more on a continuum with the rest of us.

Anyhow thanks for the hard question, this is one of the areas I hope to think about more in future.

And thank you Hanna for a reflective and measured response. The case of Dahmer concerns me in two ways--that he may indeed be a case of a broken and driven human being, and that possibility was too easily passed over by the jury, especially given the low standard of evidence (so that is a mere legal concern). There is no doubt that he managed to conceal his crimes--and he clearly thought them as crimes as perceived by others, if not himself--and that is a clear bit of evidence in favor of some culpability. But I think you see that I was pressing whether anyone can fail to satisfy the Model Code's second concern about a substantial capacity to conform behavior to the law. It seems quite conceivable that at least some people can fail this test, and yet still have some measure of rational capacity otherwise. (E.g., OCD theft has been offered as such a case.)

In any case, thanks again for taking my inquiry seriously and hearty congratulations on your new half-time appointment at Birmingham!

Hi Hanna,

Your view of addiction is shared by Gene Heyman in his book Addiction: A Disorder of Choice, perhaps you’ve seen it. He emphasizes the voluntary component of addictive behavior and questions both the disease model and the idea of compulsion (I’ve reviewed it at ). But even though addictive behavior is sensitive to anticipated contingencies, and thus partially voluntary and thus not accurately characterized as a disease, I think it’s fair to characterize it as compulsive, at least in its advanced stages, given the overwhelming motivational power of drug cravings when compared to other motivations in the addict’s behavioral economy. That is, for someone who’s really addicted, it’s very likely that she will continue to use drugs even when she foresees the very serious adverse consequences of continued use (e.g., losing the trust of those she’s manipulated to get drug money). That said, as you point out, in most cases it’s possible to eventually arrange contingencies, sometimes (in the case of opiate addiction) with the help of medications like buprenorphine and methadone, to alter the motivational balance so that use is discontinued. But this doesn’t change the fact that, for someone with a significant disorder of voluntary choice in their current circumstances, drug use is for all practical purposes compelled by the drug-induced changes in the brain’s reward system (changes which can be hard to completely reverse). And the term compulsion accurately captures the phenomenology of drug use as reported by addicts.

When an addict fails to resist an impulse to use, that failure of course is a deterministic function of the power of the impulse and the situation. In *that* situation, *that* impulse could not have been resisted. Had the situation or impulse been somewhat different, then the impulse might have been resisted. This resolutely causal view of addictive behavior (even when many of the details of particular cases aren’t known) seems to me the basic starting point when it comes to devising effective treatment such as contingency management. It also might help to change punitive attitudes about addicts, which are sometimes premised on the idea that they could have resisted using drugs in a particular situation, but just chose not to exercise their capacity to resist.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan