Freedom and addiction in four discursive registers: A comparative historical study of values in addiction science

Mainstream addiction science is today widely marked by an antinomy between a neurologically determinist understanding of the human brain ‘hijacked’ by the biochemical allure of intoxicants and a liberal voluntarist conception of drug use as a free exercise of choice. Prominent defenders of both discourses strive, ultimately without complete success, to provide accounts that are both universal and value-neutral. This has resulted in a variety of conceptual problems and has undermined the utility of such research for those who seek to therapeutically care for people presumed to suffer from addictions. This article contrasts these two contemporary discourses to two others that played vital historical roles in initiating both scientific and popular concern for addiction. These are the Puritan and civic republican discourses that dominated scholarly discussions of addiction in the early modern era. In each case, the place of values in these discussions is highlighted. By comparing them to their early modern historical antecedents, this article seeks to reflexively explore and develop more intellectually sound and therapeutically relevant alternatives to the troubled attempts at universality and value-neutrality now fettering debates in mainstream addiction science.

Mainstream addiction science is now widely, though not uniformly, marked by an antinomy between a neurologically determinist understanding of the brain 'hijacked' by the biochemical allure of intoxicants and what I am calling a liberal voluntarist conception of drug use as a free exercise of choice (see Heather and Segal, 2017;Heyman, 2009;Lewis, 2018;Volkow, Koob, and McLellan, 2016). Prominent defenders of both discourses strive, ultimately without complete success, to provide accounts of freedom and addiction that are both universal and value-neutral. This has resulted in a variety of conceptual problems and has undermined the utility of such research for those who seek to care for people presumed to suffer from addictions. In this article, I argue contemporary debate would benefit substantially from a review of the discursive contexts within which the concept of addiction originally gained a degree of intellectual legitimacy and broader cultural traction. These are the Puritan and civic republican discourses that dominated scholarly discussions of addiction in the early modern era. By comparing them to their early historical antecedents, I seek to reflexively explore and develop more intellectually sound and therapeutically relevant alternatives to the troubled attempts at universality and value-neutrality that now fetter debates in mainstream addiction science.
After reviewing prominent positions in the 'brain disease' and liberal voluntarist discourses, I then proceed to discuss prominent positions in the early modern Puritan and civic republican discourses on addiction. In each case, the place of values in these positions is highlighted. I conclude with a statement of some of the more important ramifications that follow from a more historically informed and, thereby, analytically incisive understanding of the real-world vicissitudes of freedom and addiction as they take form 'in the wild', to borrow Edwin Hutchins' (1995) evocative phrase-outside laboratories, and in the more therapeutically relevant contexts of people's everyday lived experiences. These ramifications include a proposed return from the present preoccupation with the dichotomy between freedom and neurological determinism to that which preoccupied early modern theorists between freedom and slavery. They also include a proposal for updating the appreciation, now lost to addiction science, that early modern theorists had for the intersection between judgements of freedom and slavery, on the one hand, and judgements of virtue and vice, on the other. As will be shown, this updating can be achieved by supplementing contemporary liberal theory's now preponderant tendency to equate freedom with the autonomous pursuit of hedonic values (concerning desires) with the pursuit of what Aristotle called eudaimonic values (concerning wellbeing).

A note on theory and methodology
This article has been written with the intention to contribute to current debates in addiction science by recourse to early modern history. With respect, then, to the brain disease discourse and the liberal voluntarist discourse, I take constructively critical positions. These positions are predicated on my sharing a value commitment widely, but not uniformly, exhibited in these discourses to make addiction science more relevant to the work of both caring for addicts and providing broader political and cultural warrant for therapeutic care over punishment. Defenders of the brain disease paradigm have consistently emphasized the importance of defending a medical frame for drug problems not only because they believe the best neurological research suggests the scientific truth of this frame but because they believe it is culturally indispensable that we define addiction as a disease if it is to be met with enlightened societal responses, sympathy rather than scorn (see Leshner, 1997). It is only thus, they often insist, that we can argue addicts should not be blamed but helped. However, as I show, their efforts, and those of many of their most prominent critics, to foster more humane interventions are undermined by the aspirations to make universal and value-neutral scientific claims about addiction. In this article, I explicitly demonstrate these difficulties among both those in the disease camp and those in the liberal voluntarist camp who oppose them and seek to open a path forward through an examination of two major early modern discourses on addiction.
Hence, my analytic take on the second two discursive registers I consider is decidedly different from that I bring to bear on the first two. Rather than positioning myself as constructively critical and aligned in the project of better promoting certain values also promoted in these discourses, I take a more historicist position. In the second two cases, then, my effort is simply to reconstruct and understand the meaning these discourses had for those who participated in them without any effort to assess their validity or to improve upon them by contemporary scientific lights. Instead, the purpose of these analyses is to demonstrate to readers how contemporary addiction science can better fulfil the objectives of therapeutic relevance and providing credible warrant for therapy over punishment through the selective adoption and modification of certain elements of these earlier discourses.
One reviewer of an earlier draft of this article usefully noted that neither determinist nor voluntarist addiction science is homogeneous and that one can, in fact, locate researchers on a continuum between determinism and voluntarism. More specifically, s/he proposed I stress that the following analysis makes use of Weberian ideal types to excavate and analyse how freedom and addiction are variously conceptualized in addiction science. In an important sense s/he is entirely correct, insofar as I do focus attention not on highlighting the diversity of positions found in addiction science but on particular major tendencies. However, as I reacquainted myself with the literature on ideal types and, in particular, the eminent Weber scholar Richard Swedberg's (2018) recent illuminating commentary, I became convinced that my own theoretical and methodological approach diverges from Weber's. For Weber, ideal types are artificial constructs, indeed fictions, with which social reality is compared. My own approach is instead to factually identify important theoretical tendencies in the addiction science literature and their ramifications with respect to our understanding of freedom and addiction. To reiterate, this study emphatically does not exhaustively catalogue or comprehensively assess the neurological or social scientific literatures on addiction but is only a selective analysis of certain central and important trends.
It should also be emphasized that this analysis is focused only on demonstrating and overcoming some significant limitations that follow from the widespread efforts in both the brain disease and choice theoretic literatures to produce theories about addiction that are both universal and value-free. I do not think it can be credibly denied that such research is indeed pervasive in addiction science. But this also means that to the extent contemporary research does not aspire to universality or value-neutrality, it is not a central focus of my analysis. Pertinent to this point, another reviewer of an earlier draft of this article recommended a more thorough engagement with current anthropological research on addiction. It is certainly correct to note a long and fruitful tradition of anthropological research related to drug and alcohol use and an important flowering of ethnographic research on addiction in anthropology and allied disciplines over the last 15 years or so (see Bourgois and Schonberg, 2009;Carr, 2011;Dennis, 2019;Dilkes-Frayne et al., 2017;Duff, 2008;Fraser and Moore, 2011;Garcia, 2010;Gowan and Whetstone, 2012;Raikhel, 2016;Raikhel and Garriott, 2013;Schull, 2012;Weinberg, 2005;Zigon, 2011). This literature has vastly enriched our acquaintance with the phenomenological nuances of drug use, addiction, and the extent to which these are inextricably entwined with broader economic, political, and cultural realities.
However, it only very indirectly addresses the concerns I am raising here. That is, the limits placed on our ability in both the brain disease and liberal voluntarist literatures to produce therapeutically relevant findings or warrants for therapeutic care by the widespread scientific aspirations to produce universally valid and value-neutral theories of addiction. In fact, much of the classic ethnographic literature on drug use is both consistent with and uncritical of the universalistic and value-neutral aspirations of the liberal voluntarist discourse. This literature is largely devoted to describing the locally adopted cultural norms and practices within which unproblematic drinking and drug use occur rather than the sometimes seemingly self-destructive addictive aspects of their use (see Douglas, 1987;Heath, 2012;Room, 1984;Singer, 2012). While these practices are culturally diverse, there is little if any suggestion that they are unfree. And most ethnographic work specifically focused on addiction itself tends to foreground the important roles played by language and social structural deprivation. For example, many ethnographies suggest that addiction discourse be understood as what C. Wright Mills (1940) famously called 'vocabularies of motive' furnished by, for example, addiction treatment clinics (Carr, 2011;Davies, 1992;Garcia, 2010;Weinberg, 2000), or that putative addictions are often practical adaptations to the hardships of social structural deprivation and oppression (Bourgois and Schonberg, 2009;Garcia, 2010;Waterston, 1993;Weinberg, 2005). These studies vividly highlight the structurally and culturally specific conditions under which people make their decisions concerning drug use, addiction, and recovery but rarely, if ever, explicitly consider the question of whether and how addiction reflects a loss of self-control.
These kinds of analysis incisively and, I would argue, crucially identify various elements of social context as integral to the production and reproduction of both addictions and discourses thereof. However, close inspection reveals they are much less explicit about how we might warrant or inform therapeutic or exculpatory orientations to addictions. No doubt, it is correct to insist that social structural dynamics powerfully encourage both putative addicts and others to construe their problems in personal, rather than social structural, terms and to thereby misrecognize the various ways in which those problems are caused by macro-structural regimes like economic exploitation, racism, sexism, or homophobia (see Bourgois and Schonberg, 2009;Gowan and Whetstone, 2012;Weinberg, 2005). But while these insights valuably encourage more attention to how addictions can be remedied through social structural interventions, it is less obvious how they mitigate putative addicts' ethical accountability for their specific responses to the deprivations and oppressions they suffer. Some years ago, Erving Goffman (1961: 86-7; original emphasis) relevantly observed, Although there is . . . an environmental view of crime and counter-revolutionary activity, both freeing the offender from moral responsibility for his [sic] offence, total institutions can little afford this particular kind of determinism. Inmates must be caused to self-direct themselves in a manageable way, and, for this to be promoted, both desired and undesired conduct must be defined . . . as something [they] can do something about.
Goffman's observation in this regard holds not only for what he called 'total institutions' devoted to the management of addictions but, indeed, wherever people wish to promote a specifically therapeutic orientation, in contrast to neurological or critical sociological orientations, to addiction-that is, to promote the emancipation of particular people from their own putative addictions rather than address addiction's general aetiological causes in neurological processes, vocabularies of motive, or social oppression. However, in stark contrast to a criminalizing frame, a therapeutic frame requires that we interpret putative addicts not as culpable but as afflicted and therefore in need of care. Hence, warranting and implementing a therapeutic frame for addiction does not in the first instance require theories that serve only to explain the general causes of putatively addicted behaviour; it specifically requires theories that allow these behaviours to be, at least partially, ethically disowned. It is only by distinguishing the free agency of addicts' particular selves, their self-control, from the causal effects of their specific addictions that people might be simultaneously understood as amenable to therapeutic empowerment or emancipation from their addictions through recovery and somehow also as afflicted by an addiction that justifies and demands such a therapeutic engagement in the first place (Weinberg, 2000).
My argument is that these specifically ethical and experiential aspects of addiction and recovery have been clouded by the universalist and value-neutral tendencies observable in much of the scientific literature on addiction. These tendencies manifest, on the one hand, in neurologically and social structurally deterministic explanations that tend, as Goffman observed, to gloss the complex forms of discretion and ethical accountability that practically arise in our everyday lives and, on the other hand, in liberal voluntarist discourses of 'choice' that tend to gloss the complex ways in which our discretion, and hence our individual rights and responsibilities, are often practically diminished or mitigated in our everyday lives, sometimes encouraging an ethic of care over a liberal ethic of individual rights and responsibilities (see Nussbaum, 2006). 1 As I have noted, universalism and value-neutrality foster a tendency toward scientific and ethical detachment that is inconsistent with a therapeutic and/or ethical engagement with putative addicts' particular lived experiences. As a result, they fetter our capacities to inform or warrant the provision of therapeutic care over punishment. It is in the interest of identifying and beginning to overcome these fetters that this article has been written.

The brain disease discourse
In a widely cited article in the Annual Review of Neuroscience, the distinguished neurologist Steven Hyman and his colleagues wrote, Unlike natural rewards, addictive drugs do not serve any beneficial homeostatic or reproductive purpose but instead often prove detrimental to health and functioning. Much work over several decades has begun to paint a picture of how addictive drugs come to masquerade as, and eventually supplant, natural rewards as highly valued goals. (Hyman, Malenka, and Nestler, 2006: 571) 2 We see here, cast in the presumptively value-neutral register of neurology, an explicit contrast of natural rewards, as experiences whose value is biologically determined to foster behaviour conducive to health and reproduction, with the presumptively unnatural, and health-threatening (or at least health-irrelevant), rewards of drug use. This distinction lies at the heart of the now ascendant scientific model of addiction, what historian David Courtwright (2010) has dubbed the NIDA (National Institute on Drug Abuse) brain disease paradigm. It is precisely by way of this distinction that a scientific case is made for the argument that addictive drugs deprive addicts of their freedom, and precisely thereby create a medical disability warranting medical care (Kalivas and Volkow, 2005;Kelley and Berridge, 2002;Robinson and Berridge, 2003). By unnaturally 'hijacking' the brain's natural proclivities to reward healthy behaviour, addictive drugs deceive, usurp, and enslave the brain's reward circuitry, causing a 'loss of self control' and rendering untold harms to victims themselves and to their societies (Volkow, Koob, and McLellan, 2016: 364). Ironically, there is implicit in this wholly deterministic account of the human brain and its biomechanical responses to addictive chemicals a distinct (if dubious) theory of human freedom. If addiction, or the loss of self-control, flows from our sustained indulgence in, and pathological valuing of, the unnatural rewards of drug use, then our freedom, or the retention of our self-control, must, by implication, consist largely in confining our attention to the value of natural, or biologically healthy, rewards.
The NIDA brain disease paradigm is undoubtedly hegemonic in contemporary addiction science, but it is by no means uncontested. Critics point to the challenges of reconciling the NIDA paradigm with the epidemiological facts that only a small fraction of those who have used addictive drugs fall into detrimental patterns of use (SAMHSA Office of Applied Studies, 2008); that many who do fall into these patterns 'mature out' of them without treatment (Robins, 1993;Winick, 1962); and that it is often social rather than genetic or neurological disadvantages that best predict who is most likely to succumb to addictive drug use and who is least likely to recover from it (Alexander, 2008;Edwards, 2005;Waldorf, Reinarman, and Murphy, 1991). As awkward as these epidemiological facts certainly are for defenders of the brain disease paradigm, a still more fundamental difficulty attends their conspicuous reticence to explicitly develop their implicit orientations to freedom-that is, their views of exactly what addicts actually lose when they lose their self-control. However, we can begin to flesh out these views by attending to their counterparts: ascendant neurological theories of addiction itself. How do proponents of the brain disease paradigm provide for the putative loss of our freedom to addiction?
They do so in primarily two ways. First, incentive sensitization theorists argue that through sustained exposure addicts' desire (want) for drugs is neurologically disjoined from the degree to which they find pleasure in (like) drug use (Robinson and Berridge, 2003). Hence, their desire for drugs is not only 'unnatural' but eventually unjustified by the pleasure users believe they derive from them. This allows brain disease scientists to cast this desire as both pathological and involuntary because it is both unhealthy and inconsistent with orthodox postulates of rational choice. But the equation of freedom with either health maintenance or a generic model of rationality is not scientifically sustainable. In modern liberal societies, not all unhealthy conduct, irrationality, nor even misplaced desire is considered unfree, let alone pathological. In practice, people freely value a wide range of both healthy and unhealthy preferences under widely divergent conditions of (mis)understanding. Beyond valorizing health and rational choice, addiction scientists must tell us how people actually grow estranged from their behaviour enough to warrant the claim that their self-control is genuinely afflicted, and, hence, their freedom attenuated, by addiction. Findings that sometimes people want things more than they like them is not sufficient warrant for such a claim.
The second way addictions are normally said to deprive people of their freedom is by compromising brain processes associated with 'executive functions' (Kalivas and O'Brien, 2008). These functions are not always clearly specified in the brain disease literature, but they include things like attention, response inhibition, planning, problemsolving, and working memory. Like research on incentive sensitization, this research also seeks universal and value-neutral neurological measures of freedom and addiction. But the argument that people universally equate their own or one another's freedom with long-term planning, problem-solving, and impulse control is undeniably false. Not only do we freely throw caution to the wind on occasion, but so too on occasion do we equate our most authentic values with our gut instincts, spontaneous desires, and other emotional impulses and, indeed, equate the kinds of cognitive processes associated with executive function with alienation from our real selves, freedom, and authentic selfcontrol (Alasuutari, 1992;Hochschild, 2012;Turner, 1976). On these accounts, our freedom is undermined rather than facilitated by the inhibitions imposed by executive functions. Indeed, a vivid and pertinent illustration of this is the extensive therapeutic emphasis drug rehabilitation programmes like Alcoholics Anonymous themselves place not on the executive repression of spontaneous emotions but on encouraging their free and open expression in the interest of facilitating people's better self-understanding and self-control (Carr, 2011;Garriott and Raikhel, 2015;Valverde, 1998;Weinberg, 2013;Zigon, 2011). The brain disease literature's tendency to disregard the manifest diversity of empirical forms taken by freedom and addiction, self-control and its loss, stems from an evident inability or unwillingness to breach the boundaries of brain biology in any but the most cursory manner (Campbell, 2010;Vrecko, 2010). While brain disease theorists allow that Pavlovian conditioning may arbitrarily link environmental cues with the unnatural rewards neurologically intrinsic to addictive substances and activities, these are entirely ancillary to the neurological reward circuitry by which our ultimate value preferences are determined. In this way, both freedom (implicitly defined as a rational value preference for rewards conducive to biological health) and addiction (implicitly defined as an irrational value preference for rewards that may threaten health) are conceptualized as fundamentally ahistorical and neurologically determined. Addiction is thus held to afflict our personal freedom only to the extent we equate freedom with an invariant commitment to value biological health above all else. While this research may very well continue to yield scientific dividends in other ways, it will never yield an empirically adequate grasp of the nuanced phenomenology of becoming estranged from one's own behaviour-that is, losing self-control-nor the jointly intrapersonal, interpersonal, and social structural dynamics that render that estrangement so real for people. Because it fails to adequately explain the loss of self-control to addiction, this research fails to fulfil its ethical objective of providing scientific warrant for therapeutic care for addicts. Insofar as it does not fulfil its promise of distinguishing addictions, specifically as afflictions, from the free moral agency of people who suffer from these afflictions, brain disease discourse provides no coherent warrant for medical care over blame and punishment.

The liberal voluntarist discourse
The medicalization of addiction has always been vigorously contested (see Alexander, 2008;Fraser, Moore, and Keane, 2014;Granfeld and Reinarman, 2015;Netherland, 2012;Satel and Lilienfeld, 2013). Indeed, the idea that addicts never in fact lose their freedom of choice remains widespread in both the social sciences and popular culture (Valverde, 1998). Dating back at least to John Stuart Mill's essay On Liberty (2005[1859]), this position has been largely based on the liberal democratic value judgements that all people should be viewed to possess the faculty of self-government and that it is unjust to suppress their free exercise thereof (Foddy and Savulescu, 2010;Szasz, 2003). Perhaps ironically, these value judgements have become widely entrenched in the social sciences via ostensibly value-neutral postures of agnosticism toward elite and/or mainstream norms disparaging historically marginalized groups and a concomitant emphasis of the local rationality of these groups and their own sense of moral legitimacy. This has been part of a more general social scientific tendency to understand social reality as invariably composed of the consensual or conflictual interactions of integrated social groups and to define their members as integrated individuals: rational, selfgoverning agents who affiliate with these groups based on their belief in the group's traditions, interests, or values.
Hence, more specifically, by the mid 20th century putative addicts were often described as members of subcultures with their own distinctive value systems rather than sufferers of intrinsic personal afflictions or deficits of any kind (Finestone, 1957;Hughes, 2007;Preble and Casey, 1969;Stephens, 1991). Concerned to avoid slipping from their presumed value-neutrality into an illiberally biased moralism, social researchers have often overlooked the fact that addiction, understood specifically as a loss of selfcontrol, is an idea putative addicts often themselves take seriously and is not necessarily coercively imposed from without. In short, social scientists have very often held that addiction reflects a notorious but nonetheless voluntary choice to value the use of drugs over matters that others in society consider more important (Weinberg, 2011).
Since roughly the 1980s, a growing collection of self-described 'choice theorists' in sociology, economics, social psychology, philosophy, and even neuroscience itself has sought to develop this position more methodically (see Davies, 1992;Heather and Segal, 2017;Heyman, 2009;Lewis, 2018). These theorists tend to begin with the observation that most behaviour attributed to addictions is 'incentive sensitive' (consistent with costbenefit analysis) and not compulsive in the orthodox neurological sense emphasized by disease theorists. While this is true, we should not make too much of it. Plants are incentive sensitive in the sense that they grow toward resources like light and water, but we would not normally want to conceptualize this as a voluntary choice. Defending the liberal voluntarist understanding of addiction, the philosophers Bennett Foddy and Julian Savulescu (2007: 31) argue that while it may be 'hopelessly romantic' to suggest there are no biological correlates to human preferences, the reduction of freedom to the pursuit of health-relevant natural rewards is untenable. Self-government is quite obviously motivated by a wide range of values, many of which put our biological health at risk. Activities like mountaineering, refusing under torture to divulge state secrets, or even indulging a sweet tooth may very well be highly valued and freely engaged in despite their failure to fulfil anything neurologists would regard as naturally rewarding or serving of a biologically 'beneficial homeostatic or reproductive purpose'.
From these and similar observations, choice theorists conclude that it is scientifically false (and politically illiberal) to insist preferences are freely adopted and natural when they foster health, unfree and unnatural when they might threaten health. Preferences, they argue, are neither natural nor unnatural. They simply reflect predispositions to avoid or relieve experiences we as individuals devalue and pursue experiences we personally do value. And if there is no scientifically sound way to distinguish natural from unnatural preferences, there is no way of objectively distinguishing free from pathological motivations, and we must conclude that putative addicts voluntarily choose to act as they do. According to this argument, the notion that addiction entails a loss of freedom is nothing but a myth with which we improperly exonerate people for their wrongdoings or denigrate, dehumanize, and persecute marginalized members of our societies (Davies, 1992;Heather and Segal, 2017;Heyman, 2009;Lewis, 2018;Szasz, 2003). This argument not only flatly denies the validity of current scientific claims distinguishing addiction from freedom but, according to authors like Foddy and Savulescu (2010), requires that any future scientific claims in this regard must be conceptualized independently of not only neurological conceptions of natural and unnatural rewards, but any substantive claims at all regarding the health or objective value of people's preferences. Citing the 19thcentury diagnosis of 'drapetomania', or the pathological propensity of slaves to run away, they write, In the American South during the mid-nineteenth century, it may have been difficult to believe that a sane slave would wish to escape captivity. Today, it is difficult to believe that a sane person would wish for outcomes that are harmful to their health, simply because normal people prioritize health ahead of pleasure. The case of drapetomania explains why no version of the claim that addiction is a disease should contain substantive normative claims about what a person's preferences should be. (Foddy and Savulscu, 2010: 9) Because we can find historical examples of disease categories like drapetomania that were based on now discredited claims regarding normatively appropriate behaviour, we must confine the use of the concept disease to value-free and asocial accounts of biological dysfunction. While on first blush this appears plausible, there are in fact immense, and probably insurmountable, challenges to defining even physical diseases in ways that altogether avoid normative presumptions of a dispreferred state of affairs (Metzl and Kirkland, 2010). This problem is particularly acute in addiction science, and indeed mental health research more generally, wherein we must contend with a much higher level of dissensus about the natural functions and/or dysfunctions of mental structures or processes (Bolton, 2010;Fulford, 1989). The clinical validity of even uncontroversial psychiatric diagnoses like OCD or schizophrenia is invariably predicated not on generic scientific findings of biological or psychological dysfunction but highly contingent value judgements of personal or social dysfunction furnished by patients and their significant others-that is, friends, family members, colleagues, and others with personal knowledge of the patient's unique biography and social circumstances.
On its face, one would not think this should be a problem for liberal theory. After all, since its inception, liberalism has been vigorously opposed to the idea that people's values should be dictated by any singular authority, be it a king, divine will, biology, or nature more generally. It has insisted instead upon a pluralistic moral universe wherein we decide for ourselves what it means to be free or to pursue our own conceptions of the good life without interference. One might reasonably expect, then, that liberal theory would also hold each of us entitled to decide if and when our freedom had been diminished and by what. But, sadly, what was once a pluralist and egalitarian ethic of moral and epistemological tolerance has for many become a presumptive social ontology wherein human behaviour is axiomatically understood as always free-that is, unless physically compromised-and invariably expressive of the values of its author. A prominent defender of this view, the sociologist Anthony Giddens, has insisted that someone threatened with death for defiance is still making a free choice to do as s/he is told predicated on a personal value judgement that living is better than dying. By these lights, only physical forces can literally determine our actions rather than merely incentivizing us, through appeals to our personal values, to freely adopt them (Giddens and Pierson, 1998: 84). As the writings of Thomas Szasz, and other neoliberals, amply demonstrate, this ontology affords no room for mental disease at all. The price of its adoption is thus a dogmatic refusal to recognize any form of mental affliction that is not demonstrably linked to biological dysfunction. Because the neuro-adaptations identified by brain disease theorists are not invariably dysfunctional (they result from prolonged exposure to all sources of reward), they cannot support the claim that addiction is a disease or anything other than an expression of strong, but nonetheless freely pursued, appetites (see Lewis, 2018).
Hence, like many contributors to the brain disease discourse, many contributors to the liberal voluntarist discourse on addiction tend to define freedom and addiction in universal and value-neutral terms. They do so not by conceptualizing freedom and addiction biologically. Instead, freedom is conceptualized a priori as an axiomatic resource with which to explain history but which is itself an historical constant-a universal feature of human behaviour that neither varies historically nor is ever attenuated by anything other than physical constraints. While this social ontology may continue to yield other theoretical dividends, it, like the brain disease paradigm, cannot adequately describe nor explain the nuanced phenomenology of becoming estranged from one's own behaviour-that is, losing self-control. And, hence, to the extent they aspire to universality and value-neutrality, defenders of the liberal voluntarist discourse have also provided no coherent warrant for therapeutic care as opposed to blaming addicts for their fates.

The early modern Puritan discourse
Taking issue with Harry Gene Levine's (1978) landmark analysis of the discovery of addiction, historian Jessica Warner (1994: 685) argued the earliest exponents of the disease theory were Puritan clergyman, not physicians: It is in the religious oratory of Stuart England that we find the key components of the idea that habitual drunkenness constitutes a progressive disease, the chief symptom of which is a loss of control over drinking behavior. By the same token, the modern conception of addiction was not, as Levine claimed, first formulated in the medical community but had previously been fulminated from the pulpit.
Warner did not take issue with Levine's distinction between addiction as sin and as sickness-moral and medical orientations to addiction-but sought, instead, to insist that Levine's crediting of physicians with originating the medical concept was mistaken. This is a somewhat puzzling claim. In opposition to Roy Porter (1985: 390), who advised efforts to 'understand the making of the idea of alcoholism within its socio-intellectual milieu, in particular in regard to changing conceptions of disease', Warner (1994: 686) argued, 'The noun "disease" appears to have had a fairly constant meaning over the past several centuries'. That this argument is factually incorrect will be demonstrated later. For now, let us confine our attention to its implications for her general thesis. If religious orators in Stuart England understood the concept of disease as orthodox biomedicine does today-as a biomechanical failure of the body, involuntary and therefore morally exempt-how did they reconcile this understanding with the traditional religious view that habitual drunkenness was a sin? Levine (1978: 150-1) also noted cases of pre-industrial clerics describing habitual drunkenness as a kind of 'madness' and an 'incurable' habit but argued their Calvinist theology ultimately blocked them from accepting the idea that some forms of drinking were genuinely beyond the drinker's control: Puritan ministers were the most troubled by habitual drunkenness, and in some scattered phrases and sentences we find evidence of their trying to stretch beyond the ideas of their days. Increase Mather, for example, declared that habitual drunkenness was a kind of madness, and Foxcroft warned moderate drinkers that they were 'in danger of contracting an incurable Habit'. But the ministers were not able to synthesize their observations; they were bound by the categories of their theology and psychology . . . for Puritans. . . . The individual was always viewed as having the freedom to choose to sin or not.
Though for different reasons, Levine's argument here is also puzzling. Given their wellknown doctrinal rigidity, what motivated Puritan ministers to 'stretch beyond the ideas of their days'? How could Increase Mather have reconciled describing habitual drunkenness as a kind of madness with an insistence that it was nonetheless a sinful exercise of free will? Pace Levine, Warner's first attempt to solve these puzzles was to place staunch faith in the self-evidence of experience. Puritan ministers, she argued, were not bound by their theology to interpret people's behaviour as freely chosen. Their sermons were instead directly influenced by first-hand observations. After citing some instances of preachers discussing habitual drunkenness as a loss of self-control, she wrote, It would be wrong to place undue emphasis on the examples just given, or to assume that the notion of addiction was central to earlier definitions of habitual drunkenness. But it would be equally wrong to assume that earlier generations were inherently incapable of describing destructive behavior in an empirical or critical fashion, or to assume that the mind-set of preindustrial society somehow blinded contemporaries to the addicts in their midst. (Warner, 1994: 688) In this passage, Warner seems to have been insisting the difference between addicts and voluntary drunkards was empirically obvious and beyond debate. The Puritan clergy's theological conviction that sinful behaviour invariably hailed from sinful free choices was, then, simply overwhelmed by the brute force of empirical observation. The degree of controversy that even now surrounds the very existence of addiction, let alone the diagnoses of particular cases, casts Warner's seemingly naïve empiricism in considerable doubt (Alexander, 2008;Reinarman, 2005;Reith, 2019). As we have seen, contemporary liberal theory provides robust evidence that those who choose to interpret self-destructive drinking as freely chosen behaviour are quite capable of construing the empirical evidence accordingly. Hence, Warner's claim that Puritan ministers' commitment to the doctrine of free will was decisively trumped by empirical evidence alone cannot be taken seriously. Later in her article, Warner (1994: 690) proposed a second rather more plausible argument: that preachers' use of medical language was often just an ad hoc rhetorical strategy designed not to lay any serious claim to medical expertise but only to influence parishioners: We have already seen that preachers routinely cited medical evidence when exhorting their audiences to abandon the sin of drunkenness. It is perhaps in much the same spirit that friends and family might today tell a chain-smoker that he or she runs a high risk of an early death; they might say so not because they are physicians or are themselves especially familiar with the medical evidence, but because they abhor the habit in question, and knowingly resort to a variety of appeals and rhetorical stratagems in hopes of rectifying an undesirable behavior.
There is a big difference, though, between loosely invoking claims like 'smoking kills' that have already been medically authorized and simply making up one's own health claims without medical authorization. If Warner was suggesting the former, her argument regarding the historical priority of the clerics' formulations is lost. If she was suggesting the later, she would have needed to either explain what justified clerics' claims to medical knowledge independent of medical authorization or accept that their claims were not epistemologically authorized-that they were only rhetorical stratagems rather than literal descriptions. This would then raise the question of why anyone would have eventually taken such rhetoric sufficiently seriously to begin treating it literally, let alone scientifically.
Though largely critical of her analysis, Peter Ferentzy (2001: 387) agreed with Warner that clerical invocations of disease were often just ad hoc exhortations: 'Preachers often used terms such as "disease" and "sickness" loosely, with little or no suggestion that medical issues were involved'. No doubt, Ferentzy was correct to advise caution in interpreting pre-industrial sermons and religious commentaries too literally. It is certainly true that many religious orators had few qualms about taking rhetorical liberties or drawing upon metaphors in their efforts to sway their flocks. That said, the claim that clerical uses of medical language were never literal is empirically unsustainable. Consider the following passage on habitual drunkenness from the pre-eminent Puritan scholar Richard Baxter's A Christian Directory (1825[1673: 410-11): Had God made thee an idiot, or mad and lunatic, thy case had been to be pitied: but to make thy self mad and despise thy manhood, deserveth punishment. It is the saying of Basil; 'Involuntary madness deserveth compassion, but voluntary madness, the sharpest whips.' "Judgments are prepared for scorners, and stripes for the fools back: especially for the voluntary fool": He that will make himself a beast or a madman, should be used by others like a beast or a madman, whether he will or not.
In this passage, Baxter very clearly distinguished divinely imposed from self-imposed madness to make a point regarding moral culpability. It is very hard to see how this could have been anything other than a literal distinction. But there is a second point to be made with respect to this passage. Baxter also very clearly treated habitual drunkenness as a sinful madness-a madness, but one deserving of punishment. This formulation is manifestly at odds with the contemporary biomechanical formulation of disease as morally exculpatory and with the choice theoretic literature that also starkly contrasts immorality and sickness. Without impugning its contemporary ethical merits, there can be no doubt that the now scientifically entrenched categorical dichotomy between immorality and sickness is inconsistent with much of the historical record of how early modern Europeans oriented to habitual drunkenness (Baumohl and Room, 1987;Nicholls, 2009;Valverde, 1998). It also leaves anomalous the fact that religious writers historically preceded physicians in literally asserting that habitual drunkenness was often an enslaving disease. This anomaly can be resolved by recalling the relationship people of the period routinely drew between religion and health.
During the 17th and 18th centuries, most Protestants understood all diseases as products of humanity's fall from grace and, in the case of particular communities or individuals, God's punishment of sin (Starr, 1982;Thomas, 1971). 3 By the 18th century, it was a cultural commonplace that temperate eating and drinking habits, sex lives, and lifestyles more generally were healthy because virtuous and, to some extent, virtuous because healthy. 4 The Enlightenment certainly fostered more optimism about the possibility of proactively overcoming human ills, but this did not supplant so much as transform the putative role of God in creating, and religious virtue in stemming, those ills. God was now understood to reward virtue and punish vice less through the mysterious dictates of His divine discretion than through the divinely authored medium of nature, the laws of which could be discerned and honoured through use of His greatest gift to humanity: our reason. People retained their health by use of this divine gift to better understand and more strictly obey the dictates of God's natural design.
That this fusion of religious and medical thought served as a common-sense backcloth to the work of both ministers and physicians in the 17th and 18th centuries must not be forgotten as we seek to interpret their writings. Against this backcloth we can see that pre-industrial Puritan preachers likely felt little need to explicitly justify the then patently obvious claim that many diseases were in fact the wages of sin-particularly those that resulted from a manifest disregard for God's law. This fact of life was not, as Warner might claim, empirically self-evident and immune to culturally informed presuppositions. It was instead itself a well-established cultural inheritance widely shared by 17thand 18th-century Protestants. For those ensconced in this common culture, symptoms of culpable diseases may not themselves have exhibited 'choices', the free will of the afflicted, but they nonetheless quite certainly bore the stigma of sin and God's disfavour.
Diseases like leprosy and the pox were widely thought to follow from sinful sexual behaviour, dropsy and madness from gluttony, and so on (Thomas, 1971). For many early modern Protestants, these were very real diseases, but diseases that invited contempt rather than sympathy. They were diseases that marked one as an unregenerate sinner, damned and deserving not of care but all the brutality one could anticipate from hell. As Richard Baxter advised, 'Involuntary madness deserveth compassion, but voluntary madness, the sharpest whips'. Hence, the questions that have hitherto framed much of the debate on the medicalization of addiction-when, where, and how was addiction transformed from a sin into a sickness?-do not fully square with the historical record. We should instead be asking when, where, and how the sinful sickness of habitual drunkenness became a cause for compassion rather than contempt. What changed was not the status of habitual drunkenness as a sinful sickness but the status of those perceived to be afflicted, from damned and despised to reformable and deserving of support. Before the middle of the 18th century, habitual drunkards were often described as pathologically enslaved to drink or drunkenness (Nicholls, 2009;Porter, 1985;Warner, 1994). What was largely missing was any expression of empathy for those so enslaved or any faith that they could or should be freed from their madness.

The early modern civic republican discourse
The early modern Puritan discourse on freedom and addiction was predicated on the doctrine of predestination, a doctrine that strictly opposed the idea that our eternal fate was a reward for virtue or punishment of vice. Vicious and virtuous conduct were not held to issue from a provisional character capable of moral corruption or elevation but instead a fixed and eternal character either damned or saved from the outset. Freedom, as opposed to the slavery of habitual drunkenness, was understood as the conduct of one's life in accordance with the dictates of a morally ordered universe. It reflected one's status among the elect and the happy harmony of one's own rational judgement with God's will. Conversely, the sin of habitual drunkenness was understood as a form of slavery to Satan rather than, as liberal theorists would have it today, an independently chosen moral transgression. 5 One was punished as a minion of Satan rather than, as now, an autonomous and reformable wrongdoer. It was within this discursive frame that the logic of brutally forsaking those enslaved by alcohol made moral sense.
The early modern rise of civic republicanism rather dramatically reoriented thinking on these matters (Nicholls, 2009;Schmidt, 1995). In contrast to Puritan thought, this orientation to freedom was predicated not upon abstract scholastic dialogue regarding the fixed characteristics of the immaterial human soul and its worldly tribulations, but upon mundane practical dialogue through which was fluidly shaped one's eminently cultivated and worldly personal character. Civic republican conceptions of freedom and slavery can be traced back to ancient Greek and Roman political thought, but, via the Italian renaissance, they enjoyed a notable comeback in 17thcentury England (Pocock, 1975). For civic republicans, freedom consisted in our use of reason both to tame our unruly and selfish passions and to develop and implement an educated understanding of the public good. This was in its orthodox form a plainly aristocratic ethos whereby men of excellence exhibited their distinction through wise and honourable public service.
However, its revival in early modern England reflected a much more dynamic public culture contested by royal, aristocratic, religious, and commercial actors, among others (Withington, 2007). This social structural context had the effect of shifting early modern orientations to reason and freedom from state-centric aristocratic virtues like courage and honour in the direction of more civic virtues like mundane sensibility, tact, and diplomacy (Knott, 2009;Wood, 1998: x). Slavery consisted in one's capacity to realize these virtues being denied through unreasoned servitude (Pocock, 1975: 229). Crucially for present purposes, the capacity to develop these virtues could be denied not only by the tyranny of others but also by the tyranny of our own passions and desires. By civic republican lights, slavery to habitual drunkenness was less an exhibit of one's fixed status among the damned than a profligate but acquired propensity to debase one's character-an unchecked desire for excessive (and therefore 'unnatural') bodily enjoyment eclipsing one's commitment to act and reason soundly for the public good. Moreover, yielding to temptation was understood not as a mark of one's fixed status as eternally damned but as a disabling process whereby one became progressively more dissipated, enervated, and enslaved the more one continued to yield (Berry, 1994).
During the 18th century, one could increasingly observe invocations of civic republican themes among those concerned that the luxury and licence of the prosperous were pervading society to an extent that threatened the nation's survival (Reith, 2019). For example, in his highly influential tract, An Enquiry Into the Causes of the Late Increase in Robbers, Henry Fielding (1751: 6) wrote, First then, I think, that the vast Torrent of Luxury which of late Years hath poured itself into this Nation, hath greatly contributed to produce, among many others, the Mischief I here complain of. I am not here to satirize the Great, among whom Luxury is probably rather a moral than a political Evil. But Vices no more than Diseases will stop with them; for bad Habits are as infectious by Example, as the Plague itself by Contact. In free Countries, at least, it is a Branch of Liberty claimed by the People to be as wicked and as profligate as their Superiors.
Though his primary concern was the corrupting effects of the spread of luxury and licence to the lower orders of society, Fielding made plain his view that these effects were evident throughout all the strata of English society and that they hailed in the first instance from the highest strata (ibid.). This spread of profligacy was exacerbated by the growth in trade and the material prosperity that came with it (ibid.: xxiv). Wealth was distracting the English people from virtue and thereby from both true freedom and good health. As Fielding's writing illustrates, civic republican arguments provided a timely opportunity for Protestant reformers and others to join forces in opposition to both the lusty indulgences of the crown and court and those of the increasingly ubiquitous men of commerce. It was they who were to blame for enticing the morally and intellectually vulnerable working classes into extravagant and unnatural habits that, while immoral among the better sort, were positively ruinous for the poor, particularly poor women and children. This trend was enervating and enslaving not only individuals but the nation as a whole by making the 'useful' ranks of the population morally, intellectually, and physically unfit. Few of the corrupting effects of luxury received more attention than the abuse of distilled spirits.
By the time Benjamin Rush and Thomas Trotter, widely regarded as the fathers of addiction medicine, lent their authority to the notion that habitual drunkenness was a genuine disease, the general contours of their arguments had become all but commonplace among civic republican patriots (Nicholls, 2009). These arguments were dedicated to stemming the rot of royal and aristocratic indulgence and caprice, the amoral and licentious pursuit of profit through trade, and the growing depravity and disorder they observed in the increasingly overflowing and overwhelmed urban centres. Translating republican concerns for the vicious temptations of luxury and licence into the quasi-medical language of nervous overstimulation and natural and unnatural passions, the arguments of Rush and Trotter reiterated broader anxieties regarding the political, social, moral, and medical dangers of overindulgence and, in particular, the growing temptations in this regard introduced by growing wealth, modern civilization, and urban life (Porter, 1992). And also like their civic republican compatriots, Rush and Trotter located the cure for addiction in the embrace of temperance, personal bonds, modest living, and the refuge of the agrarian countryside. Freedom from addiction, then, both could and should be fostered through the moderation of nervous stimulation and the suppression of desires to satiate the unnatural passions provoked and unleashed by the dazzle of urban life.
While thinkers like Trotter and Rush were plainly concerned about the dark side of their growing consumer societies and the proliferating and increasingly potent temptations to which people were ever more routinely exposed, they were not wholly critical or pessimistic regarding modernization. Rush, in particular, was in much of his writing quite hopeful that post-revolutionary America would become a bastion of republican egalitarianism from which the rest of the world might draw inspiration. Unprecedented religious and political freedoms would enrich and empower the United States while also rendering them havens of Christian and republican virtue (D'Elia, 1974). While some people might require tutelage and encouragement to meet their civic republican obligations, he did not for a moment see this as incompatible with the flourishing of religious, political, and cultural freedoms he believed were the hallmarks of his young nation. For Rush, the promotion of republican virtues was entirely compatible with free inquiry, candour, and a robust tolerance of difference and dissent. Pace predestinarian Puritans, then, he believed recovery from addiction was eminently achievable through temperance. And temperance was emphatically not a matter of slavish devotion to received medical, moral, or political doctrine but of inclusive, sociable, and self-critical dialogue and debate.

Concluding remarks
In early modern Puritan distinctions between living in conformity and living in opposition to the divine dictates of nature and civic republican distinctions between natural and unnatural passions, one can rather easily discern genealogical precursors to the distinction NIDA brain disease boosters now draw between natural and unnatural rewards. However, unlike today's neurologists, early modern commentators had the discursive advantage of grounding their distinctions in a Lockean natural universe as yet still saturated in largely Protestant moral meaning, a natural universe putatively designed providentially with direct respect to humanity's moral freedom, to encourage virtue and good health and to discourage vice and physical enervation. Pressed by contemporary scientific standards to conceive of human nature in universal and value-neutral terms as a wholly biomechanical product of natural selection, brain disease theorists are now left largely bereft of conceptual resources with which to link their biologically deterministic accounts of health and illness to the vicissitudes of virtue and freedom. As we have seen, the effective conflation of freedom with the narrow pursuit of health and reproduction through natural rewards simply does not hold up against the more liberal conceptions of freedom that have come to dominate our contemporary scientific and popular cultural imaginations.
Conversely, though, the presumptively universal and value-neutral liberal ontology of freedom one finds in much addiction science leaves us largely bereft of conceptual resources with which to account for the experiences of attenuated freedom, alienation from our thoughts, feelings, and behaviour, or any manner of genuine mental affliction (and, hence, any convincing intellectual warrant for compassion rather than contempt for putative addicts). Instead, it is widely held as axiomatic that all preferences exhibited by a particular person are also the preferences of that person's self. 6 However, as an empirical matter, we must acknowledge that people do occasionally exhibit behaviour radically inconsistent with the ordinary proclivities they and others normally find characteristic of their particular selves, and that such inconsistencies often serve in both clinical settings and everyday life as empirical grounds for relinquishing the faith that they have freely or deliberately chosen to so behave and, indeed, as warrants for therapeutic care (Weinberg, 2005).
A review of the early modern discursive registers considered above highlights first of all that when it comes to understanding the relationship between freedom and addiction, history does in fact furnish alternatives to the now widely reified antinomy between ontological liberalism and biological determinism. Indeed, I would argue the more fluid and dynamic dichotomy between freedom and slavery that preoccupied early modern thinkers is a far more apt one for understanding addiction than is the much more rigid contemporary dichotomy between freedom and biological determinism. Secondly, a consideration of these discursive registers also highlights the considerable intersection that was once taken for granted between judgements of freedom and judgements of virtue and vice. Whereas we now tend to see virtuous and vicious conduct as equally free, early modern thinkers were much more inclined to see vicious conduct to exhibit some manner of slavery to either Satan, one's baser passions, or both. Perhaps our understanding of the contemporary lived experience of addiction and its relation to freedom would be well served by a reconsideration of the extent to which addicts and/or their significant others might be somehow similarly predisposed and, indeed, what if any revisions to the liberal regard for freedom and addiction this might suggest.
As is well known, modern liberalism owes much to the legacy of the Scottish Enlightenment. In the 18th century, Francis Hutcheson insisted humans were endowed with a moral sense that provided for our learning from, and correcting, our moral errors. David Hume added that it was through the experience of pain in the face of vice and pleasure in the face of virtue that this moral sense was psychologically realized. By way of Adam Smith and, later, utilitarians like Jeremy Bentham and John Stuart Mill, this naturalized and sentimentalized conception of good and evil was fused with a broader conception of personal preference and transformed into a general conception of human rationality predicated on the pursuit of happiness (Levine, 1995). Confluent as it was with socio-historical trends toward religious and ethical pluralism, parliamentary democracy, and free-market capitalism, the utilitarian conception of the self as autonomous happiness maximizer has become far and away the most influential of the modern era and remains fundamental to the liberal creed.
But by these lights there can be no distinction between what we desire and what we consider good because our only measure of goodness is what we desire. As we have seen, this makes distinguishing freedom from the pursuit of what we desire axiomatically impossible and, hence, conceptualizing the idea of enslavement to one's desires-that is, addiction-equally impossible. Moreover, individuals cannot be mistaken in this regard, because for anyone to take issue with the individual's desires is, by definition, an act of oppressive interference with their free and autonomous pursuit of happiness. But this rigid orthodoxy of extreme moral agnosticism is plainly inconsistent with how people, as an empirical matter, normally orient to their own and one another's conduct in their everyday lives. In practice, we routinely evaluate whether our own and one another's desires are or are not good for us or for those around us. And these moral evaluations are contingent upon a vast range of contextual considerations (Weinberg, 2005;Zigon, 2011). Hence, if we are to reconcile liberal theory to the manifest empirical contingency of judgements concerning freedom and addiction, it will be immensely helpful, drawing our cues from early modern thought, to distinguish between contemporary liberal theory's preponderantly hedonic orientation to personal values and what Aristotle called eudaimonic values. 7 Unlike hedonic values, eudaimonic values are not constituted by individual desires alone. Indeed, they are values that may very well clash with our desires, and with which our behaviour may, even chronically, fail to conform. This allows for a liberal, or individualized and multicultural, appreciation for the broad range of values that may orient free and autonomous behaviour without thereby assuming these values are invariably hedonic. It thereby creates a possibility for conceptualizing disjuncture and tension between our desires and our free agency and, hence, addiction, without capitulating to the fraught conceptual antinomy between freedom and biological determinism. Liberal critics of the brain disease paradigm are no doubt correct to note the empirically untenable narrowness of the neurological equation of freedom with the rational pursuit of natural rewards. But, contrary to these liberal critics, perhaps it is more empirically tenable, therapeutically useful, and humane to view the experience of freedom among those who consider themselves addicts as embodied in the work they do themselves and in collaboration with others to live more consistently in line with the eudaimonic values they consider conducive to their personal and social flourishing rather than continuing to adhere to the widely, if not uniformly, held liberal ontological doctrine that freedom is only ever to be found in the pursuit of pleasure.
Returning, then, to the contemporary project of providing scientific warrant for therapeutic care over blame and punishment, we might observe that sometimes, and for a variety of biological, psychological, and social reasons, we may become sufficiently alienated from certain of our habitual desires that we cease to experience them as genuinely our own-but, rather, as afflictions and in some sense enslaving. Whereas much of contemporary addiction science now assumes a more or less unified human subject possessed of habits largely integrated through cost-benefit analysis, research in neurology, psychology, and sociology is beginning to more fully appreciate the reality of subjective fragmentation and disunity. For example, the distinguished neurologist of addiction, and critic of the brain disease discourse, Marc Lewis has written in this regard that habit-learning originates piecemeal in setting specific kinds of ways but that eventually, because brains tend to conserve structure and resources, habits, or acquired 'synaptic networks', often converge and become mutually reinforcing. However, he notes, these processes need not converge on a wholly unified subject, or fully integrated hub for all experience, deliberation, and volition. Instead, 'alternative synaptic networks can compete with each other . . . this is the case when addiction arises in development, but also when it dissipates, replaced by the desire for and belief in alternative outcomes' (Lewis, 2017: 182).
What Lewis does not specify, however, and what prevents him from reconciling his incisive understanding of the neurology of addiction with his hope to produce a warrant for therapeutic care, is the labile and dynamic relationships that occur between the diverse constellations of habitual desire he describes as the neuroplastic synaptic networks to be found in our biomechanical brains, on the one hand, and, on the other hand, our selves specifically as self-controlling, self-discovering, self-actualizing, and ethically accountable. The fact that such selves can be-and, indeed, too often actually are-genuinely alienated from and afflicted by deeply habituated desires perceived and often dreaded as traumatizing, dangerous, or, at the very least, profoundly morally inconsistent with who they are or wish to be provides robust warrant for therapeutic care. And it is precisely the emancipation and empowerment of these selves over their addictions that therapeutic care for addicts often is, and ought to be, designed to foster.

Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/ or publication of this article.

Funding
The author received no financial support for the research, authorship, and/or publication of this article.

Notes
An earlier draft of this article was presented at the Intoxication, Discourse and Practice conference at the Humanities Research Institute at the University of Sheffield, funded by the ESRC and AHRC. I would like to thank Phil Withington for inviting me to participate and Phil Withington, James Nicholls, and David Clemis for particularly helpful comments.
1. While these issues have not received much explicit attention in the ethnography of addiction as such, they have been lucidly analysed by some in the now burgeoning anthropology of ethicssee especially James Laidlaw's (2015) The Subject of Virtue: An Anthropology of Ethics and Freedom. Jarret Zigon's (2007Zigon's ( , 2011Zigon's ( , 2019 work contributes to, and draws upon, this literature in analyses of important aspects of addiction and recovery, but it does not address the central issues raised here: the difficulties of informing and warranting therapeutic care for addicts that arise as a consequence of aspiring to value-neutrality and universality. 2. Among other accolades, Professor Hyman is a former director of the National Institute of Mental Health (1996Health ( -2001, editor of the Annual Review of Neuroscience, a fellow of the American Academy of Arts and Sciences, a fellow of the American Association for the Advancement of Science, and a Distinguished Life Fellow of the American Psychiatric Association. 3. One particularly well-known example from the late 16th century is the close association Shakespeare drew between physical disability and moral corruption in Richard III. 4. The celebrated Georgian physician George Cheyne wrote in 1725, 'The infinitely wise Author of Nature has so contrived Things, that the most remarkable rules of preserving Health are moral duties commanded us, so true it is, that Godliness has the promises of this Life, as well as that to come' (quoted in Rosenberg, 1992: 54, note 59). See also Benjamin Rush in 1799: 'Christianity when believed, and obeyed . . . is more calculated to produce those effects, than any other religion in the world. Such is the salutary operation of its doctrines, and precepts upon health and life, that if its divine authority rested upon no other argument, this alone would be sufficient to recommend it to our belief' (Runes, 1947: 171). 5. It is worth underlining that in both the early modern Puritan and the civic republican discourse, the relevant model for understanding the relationship between freedom and addiction was not free will versus biological determinism but free will versus slavery. Unlike biological determinism, slavery can be understood as a contextually variable form of unfreedom, a matter of degree in the sense that one can be more or less enslaved and a condition that may grow more and less influential through time. It thus tends to preserve a degree of autonomous (if severely attenuated) personal judgement for the afflicted actor that is not so well preserved in discussions of biological determinism. 6. This formulation draws upon Rom Harre's (1987: 42) distinction between the person (understood as the 'human being as a social individual embodied and publicly identifiable') and the self (understood as 'that inner unity to which all personal experience [and, I would add, conduct] belongs as attributes of a subject'). 7. Hedonic values concern our desires and are those that have traditionally preoccupied utilitarian calculations concerning the rational pursuit of happiness. Eudaimonic values, on the other hand, concern virtue as such. This contrast allows us to distinguish judgements of personal desire from judgements of what it would mean to flourish-what, in other words, is conducive to our personal well-being above and beyond the satisfaction of personal desires or, for that matter, above and beyond our devotion to enshrined ethical principles (Ryan and Deci, 2001;Turner, 2018).