Neuroethics Symposium: Special Issue on The Biology of Desire by Marc Lewis

Marc Lewis, The Biology of Desire: Why Addiction Is Not a Disease (Public Affairs, 2016)

It is my pleasure to introduce the latest in our series of symposia on papers from the journal Neuroethics.

The focus of the current symposium is a forthcoming special issue of Neuroethics on Marc Lewis‘s book The Biology of Desire: Why Addiction Is Not a Disease (PublicAffairs, 2016). In his book, Lewis challenges the “disease model” of addiction, arguing that this model both misconstrues what addiction is and undermines proper routes to healing.

Our symposium begins with an introduction by the editors of the special issue, Steve Matthews and Anke Snoek. Following this are commentaries by three contributors to the issue: Kent BerridgeHanna Pickard, and Jerome Wakefield. Each of the commentaries discusses the others two authors’ contributions to the symposium in Neuroethics. These contributions are linked below, as is Marc Lewis’s précis to his book.

We are grateful to Neil Levy, editor of Neuroethics, and all the contributors to this symposium for their work in making it possible.


From the Neuroethics special issue:


Contributions to the Brains blog symposium (click on titles to expand):

Introduction by special issue editors Steve Matthews and Anke Snoek:

Welcome to the Brains Blog symposium on Marc Lewis’s developmental account of addiction. In his recent book Lewis argues that addiction is not a disease; rather, it is a destructive habit that grows and self-perpetuates when we repeatedly pursue the same highly attractive goal. In a current special issue of Neuroethics Lewis condenses the argument of his book against the brain disease model of addiction (BDMA) in a target article. The special issue then contains fifteen commentaries which use these sources as a foil to develop some highly interesting defenses of positions that represent a neat snapshot of the field as it currently stands. In this symposium three of the special issue authors – Kent Berridge, Hanna Pickard, and Jerome Wakefield – have very kindly offered to present some comments on each other’s papers. To background this, we summarize Lewis’s position, and why at this point it is so important.

What is addiction, and why does it matter that we get our account of what it is right? In the academy, in clinical practice, and in media representations, we find a lively and unresolved debate in answer to these questions. To boil it (right!) down: on one side of the debate there are those who argue that addicts make deliberate blameworthy choices, and (so) do not deserve our sympathy (the choice/moral model), and there are those who argue that addicted persons are victims of a disease which causes their behavior, and so they need treatment by doctors (the BDMA). Marc Lewis is not happy with either of these models! The first he thinks misunderstands the nature of choice, and overlooks the way addicted persons gradually develop, through impulsive and compulsive stages, where control comes to be eroded as the addiction takes hold. But the second, the BDMA, is utterly mistaken he thinks because the brain changes in addiction are not sufficient to count as diseased. Why? Because those changes are normal, given exposure to objects that generate habits, commitments, emotional connections we wouldn’t dare call disease, such as being ‘addicted’ when in love!

Why does this matter? Lewis is greatly concerned that the BDMA, a view he thinks overly dominant, which attracts a lot of funding, and which has been taken up in public settings in the west in a frankly hegemonic way, is having a damaging effect. For one thing, if it was correct, we might expect rates of addiction to have reduced, but on the contrary this has not occurred. Lewis is disturbed by many things about the BDMA (not just that it hasn’t worked), and one suite of related issues standing out are its failure to properly characterize addiction and recovery, its failure to reduce stigma (something it promised to do), and in particular the way it stifles addicts’ capacities to develop out of addiction – for if addicts view themselves as brain-diseased, this feeds into an addiction identity, an identity that contains within it no blueprint for future possibilities of not being addicted. If my brain is an addicted brain, well what can I do about that? Not much if I view my brain as genetically composed, an organ whose features are largely fixed by my internal makeup. Still, the remarkable thing about Lewis’s approach to the moral vs disease model dichotomy (his view is a genuine third alternative, not a compromise between these two), is that Marc himself is a development neuroscientist, and someone who knows a thing or two about addiction from the inside. In his first book – Memoirs of an addicted brain – he recounts his own addictive experiences through the prism of neuroscience. What this means is that he is able to negotiate his way through the brain science and yet simultaneously recognize that just because addicts’ brains change does not mean those changes are set in stone (or flesh). Brains must change in response to all experience. Its endogenous features are not merely the product of genetic instructions. And indeed, it is this very fact – the plasticity of brains, something recognized since William James – which provides the hopeful cue for those who may be struggling to find a way out of addiction, even when there seems little that can be done.

Set against this backdrop we have three authors in this Blog symposium. Kent Berridge thinks that by and large addiction dysfunction co-occurs with ‘normal brain function’ and worries about the disease attribution because, along with Hanna Pickard, he agrees that using a substance requires choice-making. Still his worries are mixed and he does say that the disease label is not unreasonable, even that it may be tolerated. A related difficulty Berridge mentions is that ‘brain disease’ brings to mind cases like Alzheimer’s, tumors and strokes, where ‘pathological lesions’ or ‘shriveling neurons’ are present. But addicts’ brains are not damaged in this way. Still, despite his worries about the disease attribution to addicted persons, Berridge is leery of moves to eliminate the disease model from our thinking, and that is because there is a danger this could blow up in our faces if the vacuum left by this move led to excessive and regressive blaming of addicts. Abandonment of this runs the risk of ‘fossilization’ of existing treatment methods. Notwithstanding these points, and particularly in relation to the way funding occurs, he thinks that a move to Lewis’s ‘third stage in the governing image of addiction’ is important in the struggle to find new and effective treatment methods and public understanding. (Find Berridge here: Berridge, Kent C. 2017. Is addiction a brain diseaseNeuroethics.)

Hanna Pickard is very keen to avoid the disease label because this has the effect of denuding agency from discussions of addiction and addicted persons. She argues that we can adopt a choice stance to addiction while not falling into moralization if we distinguish responsibility from blame; for in holding someone as the responsible agent this has the effect of building up their agency, and we can do that without at the same time scolding a person for their bad (addictive) behavior. The line of argument here does regard addictive actions as choices, but the idea is to get addicted persons to recognize this fact – use is choice – and in coming to own the behavior the addicted person is in a better position to acknowledge their behavior and then to change it. The tricky part is in separating responsibility from blame so as to avoid falling into moralization of addiction. Lewis is sympathetic to Pickard’s approach, though in his reply to her paper he thinks questions arise around her conception of choice. The sticking point is how this has to be conceived in the context of developing addiction. Lewis says he rejects a simple notion of choice as ‘voluntary choice’, and that to understand it in this context ‘… we must examine how a multitude of cognitive, emotional, environmental, and temporal variables impact on volition in decision making. Then we’ll have begun to see how much of choice is voluntary and how much is not.’ Still, Pickard and Lewis are not too far apart in their positions, especially on the classification question, where Pickard remarks that she accepts along with Lewis that addiction is not a disease, ‘…at least given the typical meaning and implications of that concept.’ She goes on to comment specifically that addicted brains are not so changed such as to warrant the pathological label. Finally, she remarks that…’I believe Lewis is correct to emphasize the central importance of a sense of agency, empowerment, and personal growth and self-understanding, in overcoming addiction.’ (Find Pickard here: Pickard, Hanna. 2017. Responsibility without blame for addictionNeuroethics.)

Jerry Wakefield has two articles in our special issue. His wide approach and the span of his arguments makes a neat and complete summary impossible here, so we mention the broad sweep of his ideas to give a bit of the flavor. Wakefield grants Lewis his assumption that brain changes in addiction are not too different from the changes occurring in many other pursuits such as intense love, but he thinks the difference is in the context. Addiction to heroin for example is harmful once we see how the behavior associated with it is dysfunctional, and in human social contexts the behavior of heroin-addicted persons is indeed dysfunctional. (And, on the whole, addiction to love is not.) More generally, Lewis’s argument is based on the thought that disease attributions require neurological-level breakdown. In addiction we do not see such breakdown. But, Wakefield’s harmful dysfunction account of disease can apply even in cases where there is no such breakdown, and there’s the rub. Wakefield argues that although the brains of addicted people are functioning normally, as Lewis outlines, addiction is nonetheless a medical disorder because the desire/deliberation/choice system is not functioning properly. The disorder occurs because of peremptory desires that are operating ‘…outside of biologically designed parameters and [which] override the system’s usual adaptive workings. It is a dysfunction of the brain to be so sensitized to opiates or alcohol as to be unable to function as biologically designed due to the need for substance intake, even if this dysfunction occurs via a normal learning system.’ Thus, disease attributions may be justified even granting Lewis’s insistence that the brain changes in addiction are within normal range. (Find Wakefield here: Wakefield, Jerome. 2017. Neuroethics addiction and the concept of disorder, part 1 : why addiction is a medical disorderNeuroethics.)

Overall, the articles in the special issue and in this symposium provide a systematic analysis of the challenges the BDMA faces, the merits and shortcomings of Lewis’s developmental learning model, and many fruitful suggestions for a third wave in how we may conceptualize addiction and the practices surrounding it. It is clear from these articles that we are developing increasingly sophisticated and nuanced theories about the mechanisms that underlie addiction, and as a result new approaches for clinical practices and treatment.

Kent Berridge on Pickard and Wakefield:

Addiction: normal deep learning or psychopathology?

Starve us enough, and we’ll all become food addicts: food would be more tempting than ever before, difficult not to think about when absent and difficult to resist if available.  Intense ‘wants’ are a normal feature of human psychology – under particular conditions such as appetites or stress states.  But most of us in modern societies aren’t starved, and it’s only the addicts among us who regularly face such high peaks of temptation.  Incentive-sensitization of brain mesolimbic systems can do to drug addicts (in making them ‘want’ drugs) what starvation would do to the rest of us (in making us ‘want’ foods).  This is the ‘pathological parameter’ side of addiction incentive-sensitization.  It pushes a normal brain mechanism to rare and extreme limits – operating outside the normal range while still in normal environments, and with harmful consequences.

Addiction dysfunction thus overlaps with normal brain function.  Drugs induce addiction, but they aren’t the essence of it: that lies in the hyper-reactive response of brain and mind.  Even without drugs, incentive-sensitization can arise in some people to produce addictions to gambling, food or other nondrug targets.  As Jerome Wakefield puts it in his Commentaries, “You don’t need high tech causes for addiction; novel environments interacting with normal variants in evolved self-regulatory mechanisms will do as well.”

Yet choice & agency remain

Incentive-sensitization has a neural signature of brain mesolimbic hyper-reactivity, and a psychological signature of excessively intense ‘wanting’. Still even in addiction, use always remains a choice.   Hanna Pickard’s Commentary makes an eloquent case for the choice model, which in this is similar to Marc Lewis’s deep learning model.  As Pickard puts it, “In short, the evidence is strong that drug use in addiction is not involuntary: addicts are responsive to incentives and so have choice and a degree of control over their consumption.”  For example, addicts may choose money over drugs in the lab, or in the world may choose to go cold turkey for a while to reduce tolerance and get better highs from future drugs at lower price.    Further,  Pickard points out that any hope for escape rests on addicts making a powerful act of agency.  The very hope presupposes addicts do have a choice: “Agency needs to exist to be mobilized: you can only decide to quit and do what it takes to stop using …if you have some choice and control over your use and your identity.”  On these points, I believe Hanna Pickard, Jerome Wakefield, Marc Lewis and I would all agree.

Disorder as harmful dysfunction, but not involuntary fate

Blaming addicts may not be logically necessary consequence of saying addiction is a choice, as Hanna Pickard notes, and good therapists can distinguish between blame and responsibility.  Still, I worry that choice or learning models alone – if not also qualified by appropriate concepts of disorder can have unfortunate repercussions. Blame was once common attitude toward addiction, and that attitude could surge again today if the disorder model were generally abandoned.

Jerry Wakefield’s writings about mental disorders have been influential for years, and his commentaries provide thoughtful and detailed analyses of Marc Lewis’s thesis.   Wakefield concludes that drug addiction does indeed meet the “harmful-dysfunction” criteria needed to be classed as a disorder.  He suggests “Addiction is a medical disorder in which there is a harmful partial dysfunction of the desire/deliberation/choice system due to peremptory desires that are outside of biologically designed parameters and override the system’s usual adaptive workings”.  Similar to incentive-sensitization, he suggests that addiction “dysfunction is describable in sheerly physicalist terms of neuronal hypersensitization”.   As Marc Lewis points out, we cannot “cure” addicts.  But we can say to them “Addiction is a disorder you can escape only by your own agency.  We will do all we can to help.”

Addiction processes are more important than labels

It’s a reasonable position for Hanna Pickard, Marc Lewis or other addiction specialists to say “I never apply the term ‘brain disease’ to addiction, and here’s why….”.  But it’s less useful to assert “Brain disease should be dropped from conversations about addiction”.   Because then the effort goes into controlling the conversation, and victory will be hinge on herding people into compliance.  But controlling conversation is a victory that wouldn’t actually add any new insight to addiction, and would backfire if it led to more blame.  We all need to stay focused on understanding the underlying processes of addiction, both normal and disordered, and finding better ways to help.

Hanna Pickard on Berrridge and Wakefield:

Let’s distinguish two questions. (1) Is addiction a brain disease? (2) Should we call it a brain disease? Naively, one might suggest that the answer to (2) depends only on the answer to (1): We should call addiction a brain disease if and only if it is one. But our words matter apart from whether or not they tell the truth. Truths can hurt, and falsehoods can do good. All of us who debate whether or not addiction is a disease should acknowledge that we may be motivated not only by truth but by consequences.

Kent Berridge thinks that addiction is a brain disease. “The disease label can fit, and deserves to be tolerated.” He also thinks that calling it a brain disease helps to combat stigma and get people the treatment they need. This is good. Marc Lewis thinks that addiction is not a brain disease. He also thinks that calling it a brain disease makes addicts see themselves as passive victims of disease. This is not good. In order to quit and create a better future, addicts have to see themselves as agents with the power to choose and shape their selves and their lives.  Strikingly, both Kent’s and Marc’s views of what addiction is accord with their views of what it would be good for addiction to be called.

One way to reconcile this disagreement is to relativise claims about consequences to people and contexts. Calling addiction a brain disease gives other people, from policy makers to voters to family and friends, a way to maintain sympathy and avoid blame. We medicalise instead of moralise. But it may not help addicts themselves to call addiction a brain disease, whether in the clinic or elsewhere in their lives, because it disempowers and pathologises.  No deep disagreement. Just a difference in focus which leads to a difference in emphasis.

No one should disagree with Kent’s focus on the need to combat stigma and offer better help. But there is disagreement about whether the brain disease label has in fact delivered on its promise to soften negative social attitudes and improve treatment. For my part, my focus has been shaped by my clinical work. I agree with Marc about the detrimental effect of the brain disease label on addicts. Rather than “tolerate” the brain disease label, I would prefer us to take a long hard look at ourselves, and fight negative social attitudes like stigma and blame directly.

Relativisation might reconcile disagreement in theory. But it won’t work in practice. One reason is that we all live in a shared world. Addicts will hear what’s said about addiction to other people, and other people will hear what’s said about addiction to addicts. Another reason is that owning up to the way our views may be guided by their consequences does not give permission to overlook the truth. An answer to (2) that pays no attention to (1) is not stable. It risks being undercut by the facts.

So what is the answer to (1)? Is addiction a brain disease?

Over the past decades, we’ve learned a great deal about addiction from neurobiology, as well as cognitive and social science. Kent’s body of work on the incentive salience theory represents a great advance. But there is unquestionably much we still don’t know about drugs and addiction. We also don’t have a clear and agreed understanding of what makes a condition “pathological” as opposed to “normal” and how we can know when it is. (I use square quotes here to flag that these words are placeholders until our understanding improves – I refrain from scare quotes in what follows but the point remains). In my view, we simply don’t know enough about addiction and disease to answer (1).

Cautiousness is always boring compared to boldness. So let me risk boldness by nonetheless making some claims and raising some questions.

First, what might we mean by disease? Our folk concept of disease is multi-faceted. On the surface, there are observable symptoms and suffering. But underneath, there is pathology. The underlying pathology is supposedly the cause of the surface-level symptoms and suffering. Symptoms and suffering are therefore involuntary. Our folk concept of disease accordingly positions people who have a disease in what Talcott Parsons famously called “the sick role”. If you’re sick, you should do what the doctor orders to get well, but you’re not responsible for your condition and you’re exempt from many of the duties associated with your other “non-sick” roles. Applying this folk concept of disease to addiction means that, if you’re an addict, you can’t help it. Your drug use and its consequences – the symptoms and the suffering – are not caused by you but by the underlying pathology. This is what makes addiction a disease. What makes it a brain disease in particular is the further claim that the underlying pathology is found in the brain.

When most addiction scientists talk about addiction as a brain disease, they are using this folk concept. This concept can be accurately applied to many devastating conditions. (E.g. consider the way core symptoms of Parkinson’s, like tremor and slow movement, are caused by brain degeneration). But one thing that we can say with certainty is that it doesn’t apply to addiction. The reason is that the evidence is overwhelming that addicts retain choice and control over drug use in a great many circumstances. The core surface-level symptom, drug use, is responsive to incentives. For this reason, whatever the underlying pathology in addiction, it does not override agency. As a result “the sick role” can’t straightforwardly apply and hard questions remain – to understand why addicts make the choices they do, and to decide how we in turn should respond to these choices.

In my work on addiction, I’ve argued that we can make surprisingly good progress answering the question of why addicts make the choices they do simply by appeal to a range of normal features of human cognition, alongside acknowledgement of the adverse psycho-socio-economic circumstances many addicts face. Yet we might wonder whether there is a more scientific concept of disease to be championed in place of our folk concept. Never mind the surface-level symptoms and whether or not they express agency. Let’s just focus on the underlying pathology. With this in mind, we might suggest that an additional explanation for why addicts make the choices they do is that something is wrong with their brains. “Addiction is a brain disease of temptation and choice” writes Kent. “The distinct neural changes in the brain involved in addiction are extreme enough to be viewed as pathological.” Hyper-reactivity creates a “too-high” pulse of dopamine in response to drug cues, causing temptations so “intense” and wanting so “excessive” that the brain counts as diseased. “Hardly anyone else has parameters that extreme.”

Addiction does involve cravings for drugs. When access to drugs is limited, it can be psychologically all-encompassing and distressing. When there is in addition a state of dependence, withdrawal can cause physical pain and suffering. Incentive salience theory offers in principle an elegant explanation of cravings, which has garnered great philosophical interest. But there are yet questions about its correct interpretation and status within the scientific community. My collaborator Serge Ahmed tells me that he knows of no relevant study that has been conducted to directly compare levels of dopamine in addicted subjects caused by drug versus non-drug cues, using either animal or human models.  He also believes that the vast number of studies he has read reporting either dopamine response to drug cues in addicted subjects or dopamine response to non-drug cues (such as sex, food, and even music) in non-addicted subjects indicate these are roughly identical, both in animals and in humans. But, however this empirical question is ultimately resolved, pathology is not the same as atypicality. We need an answer to the question of what makes a pulse of dopamine so high that it is “pathologically-high” or a desire so “intense” and “excessive” that it constitutes a disease. I don’t think we have an answer yet, although Jerry Wakefield has provided one of the most serious and sustained attempts to give us one.

Jerry offers a hybrid account of our concept of disorder – a cognate of disease – as harmful dysfunction. According to Jerry, underlying pathology is not sufficient for a condition to be a disorder. The dysfunction must also cause surface-level harm, as judged by the standards of society. But underlying pathology is nonetheless necessary. Jerry argues that “dysfunction” should be understood as deviation from normal biological function as determined by evolutionary design. Strikingly, he further claims that a brain that is functioning normally in one sense can yet be dysfunctional in another. The reason is that “normal biological function” is an umbrella term, covering many levels of description and explanation. Once we have distinguished between psychological and neurophysiological levels, we can see that dysfunctional mechanisms can be psychological-relative-to-the-environmental-context, rather than having anything to do with the underlying neurophysiology. By analogy with a computer, even if nothing is wrong with the hardware, the software can still malfunction if the input is not what it has been designed to process. Applying this argument to addiction, Jerry suggests that the addicted brain may be neurophysiologically normal yet dysfunctional, because it has not been designed to deal with drug rewards, unlike natural non-drug rewards such as sex and food.

It has become something of a mantra within mainstream addiction research that drugs, unlike sex and food, “hijack” the reward system, supposedly because they were neither present in our ancestral environment nor have fitness benefits. But is this true? I don’t know for sure. But I do know that some evolutionary psychologists argue compellingly that it’s not. Edward Hagen and Roger Sullivan and colleagues point out that the majority of psychoactive drugs are found in plants and, despite their potential toxicity, have many benefits, including not only medicinal properties such as defending against pathogens and parasites, but also social ones, such as facilitating mating and bonding. Herbivores are known to ingest such plant toxins – perhaps we have similarly evolved? The neurotoxin regulation model explains a range of otherwise puzzling data by hypothesizing that we have evolved to take drugs. Are we really so sure drugs are new on the scene and “hijackers” of the reward system?

Caution may not be exciting, but it is advisable. Once we’ve jettisoned the folk concept of disease and with it the commitment to passivity of surface-level symptoms, I think we just don’t know whether addiction is a brain disease. There’s too much left to learn. So where does that leave us with respect to (2)? Should we call addiction a brain disease? Maybe caution’s advisable here too. Indeed in this respect, I think Kent and perhaps even Jerry would at least partly agree. We should hold our tongues for now, and focus on fighting stigma, helping addicts, and doing further research to better understand both addiction and pathology.

Jerome Wakefield on Berridge and Pickard:

I thank the Brains Blog for the opportunity to discuss further some of the issues raised in a special issue of Neuroethics that addressed Marc Lewis’s provocative claim that addiction is not a disorder because the addicted individual’s brain is normal and simply engaged in deep learning of an intense pleasure. Based on my “harmful dysfunction” (HD) analysis of the concept of disorder, I argued to the contrary that addiction is a disorder because the brain was not biologically designed to be reinforced by the kinds of substances and activities that create addiction, and that constitutes a kind of dysfunction. In my comments that follow, I address my fellow bloggers. I apologize ahead of time for any misunderstandings on my part of their claims or arguments. Finally, note that I address the phenomenon of addiction and not the category of people who fall under the supposed “addiction” category of DSM-5 (labeled “substance use disorder” or “behavioral addiction”), which I believe is too broad and mistakenly classifies many non-addicted individuals as addictively disordered (see “Missteps”).

KENT BERRIDGE: Addiction as Incentive Salience

Against all odds, Kent Berridge has found a scholarly sweet spot in the addiction debate. His work is so lucid, thoughtful, and persuasive that all the bloggers found his view to be congenial to their own! In addressing the supposed tension between “disorder” and “choice” models of addiction, Kent hits the nail on the head when he rejects the false dichotomy and says: “Addiction is a brain disease of temptation and of choice itself. Addiction doesn’t replace choice, it distorts choice.” Kent also explains, consistent with my view, that addiction is a disorder because the extreme incentive salience (i.e., salience of cues triggering wanting of drugs) that occurs in addiction is not an expression of any biologically designed function of our mental machinery, so is a harmful evolutionary dysfunction.

Not that Kent steers clear of all trouble. Being a brain scientist, he tries to establish the pathological bona fides of addiction by identifying a mark of pathology of the brain process underlying addiction. Addiction’s “’software pathology’ in craving and behavior” is rooted in “underlying extreme-parameter brain changes that are the ‘hardware pathology’,” consisting of abnormally large dopamine spikes in reaction to drug cues. Such spikes “occur as extreme values of some normal neuronal parameter. It is the extremity of those changes in neural-psychological parameter values that causes problems. Those changes are pathological partly in the sense that hardly anyone else has parameters that extreme.” Thus, addiction incentive-sensitization “pushes a normal brain mechanism to rare and extreme limits – operating outside the normal range” and “amplifies temptation for addicts to a level more intense than most people ever face.” Indeed, “the essence of [addiction] lies in the hyper-reactive response of brain and mind.”

The problem is that the claim that addiction is categorically distinguished by the magnitude of dopamine reactions to cues appears to be in conflict with Kent’s own assertion that the starving individual experiences wanting as intense as the addicted individual’s desire for drugs: “Starve us enough, and we’ll all become food addicts: food would be more tempting than ever before, difficult not to think about when absent and difficult to resist if available. Intense ‘wants’ are a normal feature of human psychology.” Hunger is neither a disorder nor true addiction, so Kent’s example seems to demonstrates that it is NOT hyper-reactivity that defines addictive pathology. Marc Lewis turns this sort of point against the “disorder” model by citing the intensity and cognitive distortions of love to show that addiction cannot be considered a disorder on any of the usual bases.

Hanna also challenges Kent’s “size matters” claim on two grounds. First, in the spirit of the hunger and love counterexamples but based on what little empirical evidence we have that bears on the matter, she suggests that the size of dopamine spikes in addiction are not all that much greater than the spikes during normal-range cravings for food or sex. Second, at a more conceptual level, she argues that even if the spikes are larger and the wanting more intense, a quantitative difference along a dimension doesn’t translate into a qualitative disorder/nondisorder distinction because atypicality is not the same as pathology.

On the basic question of whether pathology must be more intense in circuit firing or phenomenological arousal than normality, the answer is clearly “no.” For example, pathological depression out of the blue may not be more intense than the throes of normal acute grief, and a pathological panic attack while standing in line at the grocery store need not be more intense than the normal anxiety you would feel dangling from a cliff. Beyond Kent’s illuminating empirical observations of dopamine response sizes, there is clearly a piece of the conceptual puzzle of the disorder status of addiction still missing. In my paper, I argued – and Kent also asserted — that it is the inappropriateness relative to the biologically designed triggering and functioning of these responses that makes certain reactions pathology.

Hanna is also, of course, right that atypicality is not itself disorder. However, atypicality can sometimes play an important evidential role in deciding whether a condition is a disorder. This is because certain forms of atypicality can reveal or at least suggest a failure of underlying mechanisms to function as they were biologically designed to function, thus the presence of underlying dysfunction, as in the above depression and panic examples in which emotional responses occur without the triggers that are biologically designed to cause such emotions. In such instances, the magnitude of a reaction is so disproportionate to appropriate contextual triggers that it suggests that something has gone wrong with underlying mechanisms designed to generate proportional emotional reactions.

Kent surely would not disagree upon reflection. Suppose that Marc is correct that passionate sexual love involves dopamine spikes as high and cognitive distortions as dramatic as any that occur in addiction, or Hanna turns out to be right that food and sex yield spikes and distortions comparable to addiction’s, or Kent himself is right that hunger can create addiction-size incentive sensitizations. Even if so, surely Kent would still hold that, say, opioid addiction is a disorder. He implicitly recognizes that there is an evolutionary component to addiction’s disorder status, and I would add that it is the ultimate grounds overriding even the issue of the magnitude of cue reactivity.

Kent (consistent with my own position) points out that addiction works by distorting choice. I think one might suggest two different approaches – I’ll call them the “force” and “field” approaches — to how such distortion works. A force view mimics the dopamine spike idea at the psychological level. It construes the addict’s problem as one of an extremely strong desire (or “want”) intruding on the overall psychological system of relatively stable beliefs and desires. No doubt this is part of the equation. This sort of force metaphor goes back to Plato and of course endured through to Freudian conflictual dynamics.

A “field” view refers to a global transformation of the psychological field’s phenomenology via a transformation of the relative saliences of a great variety of beliefs and desires as well as alterations in the inferential and associational linkages among psychological states. For example, consider Marc’s description of the global alterations in psychological functioning in both love and addiction: “Psychologically, both are characterized by narrowed attention and intense desire, overvaluation of the object of desire, entrapment in the ‘now’ which occludes rational, long-term decision making, impulsive and compulsive phases, and a preoccupation with goals that promote and maintain contact.” Much more than this happens, of course; associative pathways change so that many different things remind one of the beloved, and many cues trigger desire for the beloved. This sort of global alteration of the psychological system occurs transiently in all intense emotions and moods; when you are feeling enraged or terrified or joyously exuberant, some aspects of the world are more salient than usual and you are likely to make markedly different decisions than you usually do. Typical desires also work this way; Kent reminds us that severe hunger causes one to be unable to think of anything other than food, and the world is transformed into a set of potentials for finding food and associations to food. If wanting works by a global transformation of salience and causal links among psychological states, the sheer intensity of specific motives is not the full explanation for what is happening. The difference between these normal emotions and desires versus addiction is that emotions and desires are biologically designed alterations in global cognitive function, whereas addiction interferes with biologically designed functioning.

Allow me a final digression. Kent elsewhere distinguishes liking from wanting, arguing that due to sheer incentive salience one can want something without there being any “liking” driven by expectation of pleasure or reduction of pain. This idea is striking and a little frightening. For example, I incessantly check my iphone during the day, but I don’t feel any distress pushing me to do so and I don’t feel any pleasure or reduction of tension once I do it, it just arises in my mind saliently to do so and I want to do it (i.e., there is a pull to do it), and once I do it I experience a cessation of wanting to do it. In a related vein, I once suggested in a Merleau-Pontian phenomenological spirit (see “Phenomenology”) that action could occur without any motive at all, in some optimizing way in which body related to environment analogous to the formation of a bubble, where a perfect shape emerges without any guiding representation of the final shape just by virtue of the local dynamic interactions of the bits of bubble yielding a complex global result. I imagine pure wanting-based action in response to incentive saliences to be a little like that. An association I have to this is that maybe the wanting versus liking distinction can solve the “paradox of altruism,” that if you act to help another it is always for your own motive and therefore you are gaining some satisfaction out of it and therefore it cannot be fully and genuinely altruistic (see “Altruism”). One way out of this paradox may be to insist that altruism is pure wanting-to-help rather than liking-to-help and thus offers no selfish rewards at all, just action based on altruistic incentive saliences.

HANNA PICKARD: Responsibility Without blame

Hanna Pickard usefully elaborates her appealing “responsibility without blame” framework for therapeutic intervention and for our attitudes towards addicted individuals more generally. Hanna’s approach was inspired by her experiences in a therapeutic community. It has been standard wisdom in the psychotherapy field since its inception that therapists must treat their patients without blaming them. Yet, the patient is seeking help to change something seen as negative that under usual circumstances might warrant blame or at least disapproval. How a therapist negotiates this dilemma is at the heart of many theoretical developments. Use of theoretical notions like “transference” help. Freud in his papers on technique asserted that psychoanalysis is the invention of a new form of human relationship, and by this I think he had in mind the shifting back and forth between what Strawson (see “Strawson”) called reactive and objective attitudes towards the patient evident in the therapeutic encounter. It is, I suppose, a new kind of human relationship to respond to the patient’s heartfelt declaration of love by neither embracing nor rejecting it (or, like Breuer, fleeing the relationship) but rather by analyzing its meaning as a specimen to fuel the therapy process.

Carl Rogers famously insisted on unconditional positive regard for the client, thus precluding blame. Albert Ellis insisted on evaluating only specific behaviors and never blaming a person (he seems to have embraced a Humean argument that over and above specific actions there is no self to blame, thus “self”-esteem is a conceptual error). I tend to resolve this issue by embracing what I call “Socratic interpretation,” accepting Socrates’s view that all human beings are pursuing their vision of the good. Thus, once one understands an individual’s belief system and desires from the inside, one understands why what they do appears to be good from their perspective. This makes it possible to evaluate a person’s beliefs and desires but not to blame them for their actions given their intentional system. (This view gives rise to vigorous dinner-party conversation about Hitler, etc.) Hanna takes on this hornet’s nest of issues in a deeper and more sustained way than others have before her, and her perspective should find application within psychotherapy training.

To sustain the responsibility-without-blame approach, Hanna rejects the pathology view of addiction because she believes that the pathology view entails that addicted individuals are acting involuntarily and cannot make choices. This would paralyze therapeutic efforts that all the bloggers agree often depend above all on the addicted individual making painful choices. Thus, Hanna sees a pathology versus choice dichotomy in understanding addiction. She argues that in a disorder “the underlying pathology is supposedly the cause of the surface-level symptoms” so “symptoms… are therefore involuntary” because they are “not caused by you but by the underlying pathology.” She concludes that “we can say with certainty that [the category of disorder] doesn’t apply to addiction” because “addicts retain choice and control over drug use in a great many circumstances” and are “responsive to incentives” so addiction “does not override agency.”

The problem with this argument, as Kent also points out, is that it is a fallacy to say that because there is choice and agency, therefore there is no underlying pathology or dysfunction. The dysfunction is in the brain processes that distort the preference function on the basis of which agency and choice are exercised. Choice is pathologically biased or restricted in certain ways but not eliminated, so “choice versus pathology” is a false dichotomy. Hanna seems to be thinking of the symptom as choosing to take a drug or engage in a behaviour, whereas in fact the symptom is the set of background perceptions that shape the likelihood of that choice – which is still nonetheless a choice. All the bloggers agree with Hanna that understanding that the addicted individual has the capacity for choice is critical for a sensible approach to treatment that exploits the addicted individual’s ability to make choices. The difference is that Kent and I see no conlict between choice and pathology.

Is there a paradox in holding both that addiction is a pathological impairment of choice and that making difficult choices is an integral part of escaping addiction? Marc and Hanna go to great lengths to evade this supposed paradox. Perhaps a vivid (to me) analogy would help explain why there is no such paradox in using choice in treating a pathological impairment of choice. At the moment, I am suffering from an excruciatingly painful malady called frozen shoulder, in which the range of motion of my shoulder is severely impaired due to inflammation-caused pain whenever I attempt to move my left arm too much in any direction. The orthopedist’s prescription for physical therapy specified the treatment in one elegant but terrifying phrase: “aggressive stretching.” Now, a medical sophist might argue that there is a tension here: how can I treat my condition by undergoing “aggressive stretching” when the stretching movement involved in the treatment is precisely the function that is impaired and exquisitely painful? Common sense answers that, with courage, one can push the shoulder at the threshold where the pain and movement inhibition occurs to gradually increase the range of motion; one can treat motion impairment by systematically using motion to push back the impairment’s boundaries. Analogously, one can think of addiction treatment as painful “aggressive stretching” of an impaired choice function, where, with courage, a choice disorder is treated by systematic and gradual “stretching” of the individual’s choices in ways that have hitherto seemed impossible. Although choice is malfunctioning, its healthy aspects can be exploited for treatment progress.

Hanna’s statement to the contrary notwithstanding, addiction as it standardly occurs fits well the usual requirement of the classic “sick role” that the condition happens to you and is not a voluntary action. Taking drugs or engaging in gambling are voluntary actions, but having your choice function altered and thus becoming addicted is not a voluntary action or even a series of actions. Becoming addicted in its ontological status is analogous to falling asleep; you can place yourself in the circumstances where it is likely to happen and you can even intend to bring it about as a result of placing yourself in those circumstances, but it is still not in itself an action but rather something that happens to you. You cannot will yourself to be addicted any more than you can will yourself not to be addicted, although of course you can will yourself to engage in actions that bring about your being addicted or non-addicted.

As to the fact, emphasized by Hanna, that circumstances (such as poverty) can influence an addicted individual’s choice, that is not in tension with a disorder attribution. Many disorders are context-sensitive with symptoms that emerge only under certain circumstances. (E.g., my brother had terrible asthma in New York and moved to Arizona where his pathology was no longer triggered to yield symptoms.) Circumstances such as poverty do explain drug use, but they do not explain the development of addiction per se. Like Hanna, Marc also tended to put emphasis on social factors like poverty in the genesis of addiction, but — as I discussed in my paper — when he was challenged by someone who cited the opioid epidemic in the U.S., he retreated from that position. This is because the opioid epidemic, an iatrogenic scourge generated by the decision to widely prescribe opioids for noncancer pain, is destroying middle class intact families and communities that have traditionally not had social problems of the sort appealed to by Hanna. The problem followed not an increase in poverty or other such factors but simply the increased availability of opioid pain medicines to the general population of people with chronic noncancer pain (which includes about a third of the U.S. population).

Hanna in her blog comment, in arguing against an evolutionary pathology approach, mentions work by evolutionary psychologists that argues that perhaps some human ingestion of some addictive substances that evolved in plants as defensive neurotoxins is not entirely an evolutionary novelty for humans and that perhaps we used these naturally occurring substances during our evolution, so our brains are actually evolved to cope with them to some extent. This approach might offer an explanation of the pleasure experienced from certain substances that is different from the standard “hijack” account that it is just an accident that these substances stimulate and take over our reward systems. However, even if this speculative theory turns out to be true, the amounts and purity of currently available substances certainly remains an evolutionary novelty that goes way beyond anything our forebears could generally access to an extent that might have led to selection pressures. So, there is no reason to think that this hypothesis would explain modern addiction-related phenomena as a naturally selected response. The theorists Hanna cites themselves make clear that their approach does not address addiction (e.g., “A caveat: neurobiological theory of drug use usually contrasts initial seeking and use with longer-term phenomena such as drug tolerance and addiction. We focus on initial drug seeking” [Hagen et al.]). Thus, this theory of some level of brain adaptation to certain substances remains consistent with conceptualizing addiction as a disorder, even if the “hijack” notion would require some revising.

Both Hanna and I tend to place considerable responsibility on society at large for the problem of addiction, but with different emphases. Hanna sees unfortunate social circumstances such as poverty and lack of life options as largely explanatory of addiction. I agree that such factors enter into the explanation of behaviors that lead to and maintain addiction, but the primary social factor in addiction is the social creation of evolutionarily novel addictive substances and activities and their active promotion, often for profit. The interaction of these novel stimuli with normal genetic and psychological variation creates vulnerability to specific addictions in subsets of the population. An analogy is to the misguided “blaming” of many individuals who are obese for their overeating. There are a few true genetic obesity syndromes that account for a small percentage of cases of obesity. However, most cases of obesity appear to be due to normal variations in genes regulating fat storage that were adaptive during our evolutionary past interacting with the unprecedented amount of high-calorie food not only easily available to people in a contemporary advanced society but marketed to them in ways aimed at maximizing temptation. In many cases, individuals with the same so-called obesity genes would be fine in an evolutionarily expectable environment of relative food scarcity. Similarly, evolutionarily novel addictive substances and activities are made available by our society and interact with normal-range genetic variation to produce addiction in individuals who would be fine in an evolutionarily expectable environment. Thus, society has a heavy responsibility – and perhaps blame! – for much of the problem of addiction. Offering treatment is not just a matter of compassion but also a matter of justice. This principle seems to be recognized, for example, when localities that allow gambling accept the inevitability of some percentage of the population developing gambling addiction as a result and factor support for treatment into the cost of allowing such activities.


Hagen, E. H., Roulette, C. J., & Sullivan, R. J.(2013). Explaining human recreational use of ‘pesticides’: The neurotoxin regulation model of substance use vs. the hijack model and implications for age and sex differences in drug consumption. Frontiers of Psychiatry, 4:142. doi: 10.3389/fpsyt.2013.00142.


Strawson, P.F. (1962). Freedom and resentment. Proceedings of the British Academy, 48, 1–25.

Wakefield, J. C. (1993). Is altruism part of human nature? Toward a theoretical foundation for the helping professions. Social Service Review, 67, 406-458.

Wakefield, J. C. (2014). The biostatistical theory versus the harmful dysfunction analysis, part 1: Is part-dysfunction sufficient for medical disorder? Journal of Medicine and Philosophy, 39(6), 648-682.

Wakefield, Jerome C. (2015). DSM-5 substance use disorder: How conceptual missteps weakened the foundations of the addictive disorders field. Acta Psychiatrica Scandinavica, 132, 327–334.

Wakefield, J. C., & Dreyfus, H. L. (1991). Phenomenology and the intentionality of action. In E. LePore & R. V. Gulick (Eds.), John Searle and his critics (pp. 259-270). Oxford: Basil Blackwell.


Comments are closed.

Back to Top