1. Introduction
Imagine that Diego invited his new partner, Juan, over for dinner to meet his parents. During the afternoon, Juan gets heavily intoxicated, shows up at Diego’s parents’ house in a bad shape and behaves in inconvenient ways. As a result, Diego is hurt and disappointed. Prior to learning further details, it seems fair to assume that Juan is blameworthy for this. Now suppose that Juan suffers from severe addiction.1 Does this mitigate his blameworthiness?
From the perspective of the person with the right standing to blame, the question raises two quite different sets of issues. The first concerns what I will call the principle problem: Is the fact that Juan suffers from addiction a consideration of moral import in assessing his degree of blameworthiness? Is addiction the sort of consideration that might, under appropriate conditions, mitigate moral blameworthiness? Most discussions of addiction and moral responsibility have focused on such questions, and many have argued for the view that addiction excuses. Here I articulate what I take to be the best argument for this view. The key consideration concerns the substantial difficulty that people with severe addiction experience in controlling drug-related behavior (section 3). This, I argue, may in some cases be sufficient to ground a mitigating excuse, given the way in which addiction undermines agents’ responsiveness to relevant moral reasons to do otherwise (section 4).
Much less attention has been devoted to a second set of issues that crucially affect the possibility of applying this general principle to particular cases, which I will refer to as the practical problem (section 5). For the general principle that addiction excuses to have any bearing on the situation at hand, the person with the right standing to blame must make a judgment about the extent to which the agent possessed certain morally relevant capacities at the time of the act. Discussions of moral responsibility in the context of addiction have for the most part neglected the practical significance of the difficulties posed by the ambivalent nature of agential control in this context. Juan might be eligible for a mitigating excuse if his ability to control his behavior was sufficiently impaired in a way that was relevant to the nature of the moral situation he faced. In practice, however, this will prove tremendously difficult to assess, given that the ambivalent nature of agential control makes for an evidentially underdetermined situation. From the point of view of the person with the right standing to blame who has to decide on a fitting response to moral harm, there is often no fully satisfactory way to navigate the intricacies of this situation. Diego must walk a narrow path between the risk of unfairly over-blaming and the risk of condescendingly under-blaming, with no definite guide to arriving at an appropriate response.
The principle problem is the natural focus for theories of moral responsibility. But it does not fully reflect the nature of the ethical challenge faced by affected parties that seek a fair and non-condescending way to respond to addiction-related moral harms. Diminished control may mitigate moral blameworthiness, but this provides only a rough general guide for resolving questions of moral responsibility in particular cases. For the person with the right standing to blame, the challenge of deciding on a fitting response to moral harm is fundamentally about making a judgment about matters that are underdetermined by the available evidence.
2. The principle problem: some preliminaries
It seems natural, to some extent, to think of people with addiction as morally responsible agents. Even in severe cases, addictive drug use remains an intentional action in a recognizable sense of the word. It is, or appears to be, explained in terms of motivation and decision-making, and it is typically performed with a reasonably adequate level of understanding of its consequences. Thus, it seems intuitively unlike paradigmatic cases where a full exemption or a full excuse is warranted.2
There is, however, another way for agents to be less than fully responsible for their actions, which involves mitigation. This obtains when there are grounds for partial rather than full exculpation. I submit that the most intuitively appealing view on the principle problem is that the way in which addiction undermines agency may, in some cases, be sufficient to mitigate moral responsibility without fully exculpating agents from addiction-related moral faults. It speaks to its prima facie plausibility that many scholars have defended claims in the vicinity of such a view in the past.3 And there is also some experimental evidence that folk intuition supports the view to some extent.4 In connection with the principle problem, my aim will be to articulate a justification for this intuition.
(One tricky issue I will leave open along the way is whether this mitigation of blameworthiness is based on a localized imperfect fulfillment of the conditions for being a morally responsible agent (i.e., a mitigating local exemption), or on a localized difficulty encountered in the way of responding to relevant moral demands (i.e., a mitigating excuse). For the sake of brevity and simplicity, I will for the most part resort to the language of excuses, but the argument I develop in later sections is consistent with both possibilities. Deciding between them would require grappling with a difficult issue, namely, whether the source of the difficulty in controlling drug-related behavior experienced by people with addiction is more plausibly located in the agent’s abilities or in the circumstances in which she acts. This is an issue I will not attempt to resolve here).
On what I take to be the intuitively appealing view, when we learn that Juan suffers from severe addiction, we see him, on that account, as less blameworthy than he might otherwise have been, even if we still think he is accountable for his behavior. To illustrate, consider two variations of the case. In both variations, every circumstance and aspect of the situation is exactly the same, except that in one Juan suffers from a severe addiction, while in the other, Twin Juan does not. My contention is that the intuitive view of the case is that Juan is a fitting target for blaming responses, even though he is, on account of his addiction, less blameworthy than non-addicted Twin Juan.
Cashing out this intuition requires producing an excusing argument. There are two basic requirements that such an argument must meet: it must be based on an empirically defensible picture of addictive agency, and it must appeal to a sufficiently plausible theory of moral responsibility. To set the stage for the argument I present in the following sections, consider two ways of arguing for the addiction excuse that fail on these grounds.
The first is built on an analogy between addiction and duress (Husak 1999; Watson 1999). In this picture, a person with addiction may be acting under a sort of internal threat of harm in the form of withdrawal symptoms. If the pains contingent upon not using are severe enough, then—the argument goes—it would be unfair to demand from an agent that she suffers such pain, and so this may provide a (partial) excuse for moral wrongdoing suitably connected with decisions to use.
The analogy between addiction and duress is imperfect on several accounts.5 But withdrawal cannot be what we are getting at if we think that addiction in general excuses—although it can certainly be relevant to morally appraise the actions of people who are experiencing such symptoms. One reason is that there are types of addiction that involve only mild withdrawal symptoms, and some that involve none at all (Emmelkamp and Vedel 2006, 4). And while withdrawal symptoms can be painful and extremely hard to endure in some cases, they are usually relatively short-lived. After some time, they begin to subside and eventually cease to be experienced (Emmelkamp and Vedel 2006, 5). However, addiction continues to have the potential to undermine agency in morally relevant ways, and thus to ground an excuse, long after withdrawal symptoms have ceased to be an issue. Furthermore, the argument portrays the avoidance of withdrawal pain as the primary reason why people with addiction choose to use. This may be true in some cases, but it is surely incorrect as a general explanation of addictive drug use. People with addiction may decide to use for a number of reasons, including but not limited to the need to avoid withdrawal. Other relevant reasons to use include seeking pleasurable experiences, coping with stress or other sources of psychological discomfort, or because it coheres with established self-narratives, among many other possibilities.
Now consider another popular idea: the view of addiction as a disease. It has sometimes been suggested that the exculpatory implications of such a view are one of the reasons for endorsing it. People with addiction are often burdened with feelings of shame and regret, as well as the targets of third-personal resentment and anger. Viewing addiction as a disease, it is argued, can do them a service by undermining such feelings (e.g., Volkow, Koob, and McLellan 2016, 368).
There is some appeal to the idea that someone can be excused for certain behaviors on account of suffering from a disease. For example, it may be that we are under a general obligation to show compassion to people who are unfortunate or suffering, and people who have a disease can fit that description. But it seems that the main consideration when it comes to moral responsibility has to do with agential capacities, and the consideration that someone has a disease serves at best as an imperfect indicator that some of their morally relevant capacities may be affected6—imperfect because some diseases do not seem to affect morally relevant capacities in a significant way. Furthermore, insufficient capacity need not issue from a disease-like cause to ground an exemption or an excuse—think, for instance, of standard cases of immaturity. In response, it may be argued that calling addiction a disease implies that addictive behavior is the result of mechanistic dysfunction, and thus indirectly speaks to the impairment of morally relevant capacities (Sisti and Caplan 2016; Wakefield 1992). But the disease view of addiction can be controversial in its own right, and some have found reason to doubt that it is correct (Field et al. 2019; Heather 2013; Lewis 2017; Pickard 2022; see Burdman 2024a, for an overview of the debate). Luckily, the fate of the addiction excuse does not hang on this controversy, and we need not resolve it here. The most promising place to look when thinking about the addiction excuse is the way the condition affects morally relevant capacities, whether or not it is properly called a disease.
3. Partially impaired behavioral control
By most scientific definitions, addiction involves an element of impairment of behavioral control over drug use—the sort of thing sometimes called ‘compulsion’.7 This is the obvious candidate for an impairment of ability that could ground an excuse, since it directly concerns the volitional condition for moral responsibility. Not coincidentally, many classical pieces in the moral responsibility literature cite addiction as an example of an exempting/excusing condition (e.g., Fischer and Ravizza 1998, 35; Frankfurt 1971; Watson 1975, 325).
If addictive behavior were completely or literally compelled, this would allow for an easy solution to the principle problem. However, there is forceful evidence against the view of addiction as a condition that literally renders agents unable to refrain from drug use. The main challenge in producing an answer to the principle problem is to frame the basic insight that compulsion is incompatible with responsibility in terms of an empirically defensible view of addictive agency.
I will not rehearse here the full case against the view of addictive behavior as purely compulsive (for a summary of the evidence, see Pickard 2015, 2018; Sripada 2018; Heyman 2009). For present purposes, a few basic observations will suffice. The most important relates to the fact that people suffering from addiction are generally able to regulate drug use in a way that is responsive to relevant circumstances and conditions. Given the right kind of incentive structure, even severely addicted people can choose not to use, as both experimental (Hart et al. 2000) and clinical evidence (Petry et al. 2017) suggests. Also suggestive is the fact that many people who are correctly diagnosed with addiction at some point in their lives according to extant diagnostic criteria go on to recover without medical treatment (Sobell, Ellingstad, and Sobell 2000; Heyman 2009). Indeed, a survey of expert opinion on this issue, targeting both addiction therapists and experimental researchers, found that the view of addiction as a condition that makes people simply unable to abstain from using has little support among those who work in close contact with people with addiction (Carter et al. 2014).
The implication is not that addiction does not compromise agency at all. Rather, it is that the way in which addiction compromises agency needs to be understood in a different light than as a literal inability to abstain. Refraining from use remains an open possibility, even in severe cases. Using drugs is not a reflex-like occurrence that bypasses the agent’s will; it is intentional behavior explained in terms of motivation and decision-making. To frame this as a literal inability to do otherwise is simply to misdescribe the nature of addictive agency.
To be clear, there is also compelling support for the claim that addiction compromises agency in relevant ways. This is reflected in the well-known fact that addiction may be extremely difficult to overcome. For those suffering from severe addiction, quitting is far from a simple matter, and many find themselves in the difficult position of continuing to use drugs despite being aware of significant harmful consequences of doing so. An indication of how difficult it can be to refrain from using is the fact that some people suffering from severe cases of alcoholism resort to medications that cause severe sickness when alcohol is consumed, as a self-imposed penalty to discourage future consumption (Banys 1988). Even knowing that such unpleasant consequences are guaranteed, many fail to abstain from drinking.
In sum, addictive drug use is not literally compelled, but neither is it the result of purely ordinary decision-making processes. The ability to control drug use is plausibly portrayed as partially impaired by addiction: it is undermined to some extent, without rendering people with addiction literally incapable of doing otherwise.
Another crucial consideration about addiction is that it is in many ways a highly heterogeneous condition (Pickard 2022). There are significant differences between the patterns of use associated with different substances, as well as between the individual characteristics of people suffering from it and their life circumstances. While a reduced ability to control drug use is a common feature of all cases of addiction, the precise nature of the control-undermining factors at play appears to be variable (Burdman 2022). Potential control-undermining factors include psychological anomalies, situational pressures, and challenging social-environmental conditions, with some of these playing a more prominent role in some cases than in others.
Consider social-environmental conditions first. A social context that offers very limited opportunities to pursue alternative drug-free life trajectories may negatively affect a person’s ability to control their drug use (Hart 2013). An environment that provides support, strong incentives, and realistically available alternatives to a drug-focused lifestyle enhances a persons’ ability to refrain from using. On the contrary, attempts to quit by people struggling with unemployment or housing instability are significantly less likely to succeed (Saloner and Cook 2013).
Situational factors also play a role. The degree to which people are sensitive to considerations relevant to their actions is a variable feature of agents that can be positively or negatively influenced by immediate situational pressures. It is, for instance, much more difficult for someone with addiction to refrain from using in settings rich in drug-related cues and opportunities for use, especially in the presence of drug-using companions (J. R. Schroeder et al. 2001).
In addition to these types of agent-external conditions, the explanation of addictive behavior typically includes a variety of different psychological factors, including anomalies in motivation, cognition, and decision-making processes. Addictive desires may be anomalous in some respects, persisting in a way that is unresponsive to desire-incongruent evaluative judgments and aversive past experiences (Burdman 2024b; Holton and Berridge 2013; Wallace 1999). Drug-related cognition may also be compromised in subtle ways. Evaluative judgments about drug use may become unstable, shifting over time without the acquisition of new evidence (Levy 2014), and drug-related belief formation may be biased toward use-congruent interpretations (Pickard 2016; Segal 2013), or otherwise distorted (Sripada 2022). In addition, addiction significantly affects the allocation of attention. This occurs both at the perceptual level, where drug-related perceptually available items tend to capture attention through bottom-up influences, and in the context of deliberation, where use-congruent considerations are more likely to remain within attentional focus (Cox, Klinger, and Fadardi 2016). In some cases, decision-making processes may be more generally skewed toward the pursuit of rewards that can be obtained sooner, leading to difficulties in appropriately weighing the value of rewards that are more distant in time (Ainslie 2000; Bickel et al. 2014; Bechara 2005).
The interpretation of available evidence is open to dispute and scientific knowledge is always subject to revision. For now, however, the tentative picture that emerges from the current state of knowledge is roughly as follows: people with addiction experience powerful motivation to use, they may have difficulty bearing in mind and appropriately weighing considerations that speak against drug use, and their attentional and belief formation processes may be tilted towards use-congruent outcomes. In some cases, these traits interact problematically with situational and social-environmental factors that contribute to undermining control. Crucially, all of these features are matters of degree. In important respects, addictive behavior remains voluntary and intentional; it is not necessitated. It is no accident, however, that many agents who fit the above description continue to use systematically, find it so hard to quit while they are users, and are so likely to relapse while in recovery.
4. From diminished ability to reduced blameworthiness
The idea that compulsive behavior precludes blameworthiness is treated as a data point by classical theories of moral responsibility. Incompatibilists of various stripes argue that any causal determination undermines moral responsibility, while compatibilists typically rely on intuitions about how compulsive behavior differs from ordinary cases of deterministic causation to argue for the conclusion that it is the former, not the latter, that is incompatible with responsibility. One thing on which all parties to the classical debate seem to agree is that addiction is a prime candidate for a condition that makes agents unfitting targets of responsibility demands. However, they usually do so by assuming that compulsion means that the agent is literally unable to do otherwise. Once we think of impaired control as a matter of degree, things look a little different.
For the purposes of this discussion, I will adopt what I consider to be the theory of moral responsibility best suited to graded distinctions, namely, a capacitarian account.8 On this view, the basic requirement for morally responsible agency is the possession of certain morally relevant capacities (Fischer and Ravizza 1998; Vargas 2013; Brink 2021; Nelkin 2011; Sartorio 2016; Wallace 1994; Mckenna 2013). A useful way of unpacking this proposal is this: for an agent to be aptly held morally responsible for her actions, she must behave in a way that reflects a sufficient capacity to respond to relevant moral reasons pertaining to the situation at hand.9 If an agent does not possess this capacity to a sufficient degree, she is not a fitting target of moral demands.
Reasons-responsiveness is not a have-it-or-don’t property of agents, but a scalar property falling along a continuous spectrum. Following Fischer and Ravizza, theorists of reasons-responsiveness typically think of responsibility as a threshold concept, meaning that there are minimum conditions that an agent must meet in order to be within the domain of morally responsible agency at all—there is some point along this gradient that determines the minimum degree of reasons-responsiveness that makes an agent an apt target of moral demands. Nonhuman animals and small infants are often cited as examples of agents that do not meet such minimum conditions. Their behavior is flexible enough to be modified by environmental circumstances, but it is not sufficiently responsive to the presence of moral reasons for them to be aptly held responsible when those reasons are overlooked.10
A crucial consideration is that the ability of morally responsible agents to track relevant reasons and to successfully respond to them comes in many shades, varying from person to person and within the same agent at different times or under different circumstances. The scalar nature of reasons-responsiveness implies that there will still be significant differences between agents who meet the relevant minimum requirements, i.e., those within the domain of morally responsible agency. This picture of degrees of reasons-responsiveness thus sits well with the intuition that moral responsibility is not an all-or-nothing affair. If we take reasons-responsiveness to be the agential capacity that grounds fitting ascriptions of moral responsibility, then it stands to reason that partial ability will lead to partial responsibility, provided that the minimum threshold conditions for morally responsible agency are met.
A natural development of the theory is then to think of degrees of moral responsibility as more or less directly tracking degrees of reasons-responsiveness (Coates and Swenson 2013; Nelkin 2016). This sort of approach can make sense of some intuitive cases. For instance, we tend to think of older children and adolescents as having an ambivalent standing when it comes to moral responsibility, with some demands on them seeming appropriate while others do not. The extent to which ascriptions of moral responsibility are appropriate seems to be plausibly captured by the extent to which we see maturing agents as having the capacity to suitably respond to the relevant moral reasons.
Now, consider again Juan’s case. Diego reasonably expected him to show up in good shape when meeting his parents for the first time. Thus, there were reasons that, in those particular circumstances, spoke against the decision to use drugs at that time, as it was incompatible with the commitment he had made. The fact that Juan suffers from severe addiction is relevant for the assessment of his responsibility in failing to refrain, as it speaks to a partially undermined ability to respond to relevant moral reasons when decisions to use drugs are at issue. The fact that Juan suffers from severe addiction makes it much more difficult for him to respond to the presence of the relevant moral reasons, insofar as he experiences a substantial difficulty refraining from drug use. His degree of reasons-responsiveness seems sufficient for him to be aptly held responsible, i.e., he meets the minimum threshold conditions for moral responsibility. However, the fact that he enjoys the relevant ability to a lesser degree than non-addicted Twin Juan makes it the case that, all else being equal, he is less blameworthy for his behavior than Twin Juan.11
5. The practical problem
If the foregoing argument is correct, addiction excuses to the extent that it undermines agents’ ability to respond to relevant moral reasons pertaining to the situation at hand. Thus, assessing the extent to which someone suffering from addiction is responsible for her behavior in a particular case involves making a judgment about the extent to which the agent enjoyed the relevant capacities. But making such judgment with any confidence will often prove to be an extremely difficult task. This is at the heart of the practical problem.
Of course, this problem is not unique to the addiction excuse. Moral theory is often concerned with general principles whose applicability in particular circumstances depends on further judgments about the nature of the case, including both matters of fact and normative appraisals.12 But the problem takes a particularly dire form when it comes to addiction.13 If something approximating the argument laid out in the previous sections is correct, the addiction excuse is fundamentally grounded in the fact that people with addiction often lack full control over certain behaviors. And yet, behavioral control in the context of addiction is something of an elusive notion. On the most plausible view of addiction, control may be significantly reduced but is typically not eliminated—and the force of the addiction excuse depends on the correct assessment of the extent to which the actions in question were under the agent’s control. This poses a significant challenge to blamers, who must make a particularly difficult call concerning the extent to which addiction has undermined the agent’s control over the relevant actions. In practice, this is often difficult to determine given the available evidence. The most pressing ethical challenge for the person with the right standing to blame is how to navigate the epistemic precariousness of this situation.
Moreover, there are risks associated with getting the judgment wrong. Over-blaming is, of course, problematic. It is clearly unfair to blame someone more than is warranted by the extent of their actual responsibility. But under-blaming can also be problematic in its own way. On the one hand, there are instrumental reasons related to the function of blame in this context. As people with addiction struggle to gain a firmer grip on their agency, holding them accountable for their behavior can be a valuable way of supporting this effort by providing them with the right sort of feedback. In addition, there are other risks associated with under-blaming that are distinctively moral in nature. Withholding blame when blame is appropriate may convey the message that we see the other as less capable of moral agency. Thus, it amounts to denying an important form of recognition of the other’s status as a moral agent: under-blaming risks sending the message that one does not see the other as a full member of one’s moral community (Shoemaker 2022). I do not mean to suggest that a proper consideration of this issue should lead to the conclusion that the problems associated with over-blaming and under-blaming are symmetrical. It may be, for instance, that the harms that would result from over-blaming are somehow more serious. The important observation, in the present context, is that the risks associated with under-blaming are not insignificant and can be a subject of serious moral concern.14
For people with the right standing to blame—especially those in close relationships with people suffering from addiction—it is crucial to get this judgment right. And yet it is immensely difficult to do so.
5.1. How much control did the agent have?
Based on what we currently know about addiction, it is fair to say that the condition can, in some cases, significantly undermine agents’ ability to refrain from using. From the point of view of the blamer, however, what needs to be determined is the extent to which a person with addiction was in control of some relevant action at the time of acting. In a sense, this involves the ordinary difficulty of making a judgment on a matter of fact based on incomplete evidence, compounded by the fact that there are ethical consequences to getting this judgment wrong. But when it comes to addiction, there are additional complications that make this assessment more difficult for the blamer.
One is that the very concept of partial or undermined control is unfamiliar and particularly difficult to grasp. Folk-psychological lore is not well equipped to deal with the ambivalent status of addictive agency when it comes to behavioral control. Complete lack of control is much easier to grasp. We seem to have no trouble picturing that there is a purely causal explanation for the sleepwalker’s wandering or the seizure’s victim erratic movements. These are not up to the agent in any relevant sense, and they seem to have nothing to do with what she has reason to do or her preferences, and so it is doubtful, at best, that these happenings belong in the realm of action. But it is much harder to grasp that a person can do something intentionally, at least in part because she wants to, and yet that her actions are not fully under her control. Moreover, commonsensical proxies for addictive motivation risk promoting a false sense of understanding. The predicament of the person with addiction who is trying to refrain is not, despite common metaphors, like the common difficulty of abstaining from eating too many chips or too much ice cream. There is something extraordinary about the difficulty that people with addiction face. This unordinary difficulty in refraining is, to put it bluntly, the main reason for thinking of addiction as a mental disorder.15
As suggested above, the philosophical toolkit can help with the thorny issue of how to make sense of the very notion of degrees of control. Thinking of degrees of control as degrees of reasons-responsiveness offers a way to capture both sides of the coin. On the one hand, it seems true that there are always sufficient reasons (actual or counterfactual) to refrain from using that even people with severe addiction would respond to. Thus, their inclination to use is not totally unresponsive to relevant considerations—they have some control. On the other hand, the set of actual or potential sufficient reasons to refrain to which they would successfully respond is plausibly smaller than the corresponding set for a non-addicted person under otherwise similar circumstances. Thus, there are some actual or counterfactual scenarios in which they have sufficient reason to refrain and yet fail to do so. In other words, they have less control than the non-addicted person, all else being equal.
And yet this will not get us very far when it comes to making the sort of judgment that is relevant to deciding particular cases. Did the person, at the moment of action, have sufficient capacity to respond to the moral reasons for doing otherwise that were overlooked from the point of view of the blamer? The graded nature of control in the context of addiction makes it particularly difficult to answer this question with any confidence. Whatever evidence we consider for the case, it will predictably be consistent with both the presence and the absence of the relevant sort of control. This kind of underdetermination follows once we accept that the person has the ability to respond to some relevant reasons, though possibly not to all of them. Ambivalent control implies that we should expect some evidence of preserved control as well as some evidence of diminished control.
A further difficulty is that the ability to control behavior is not a fixed property of agents, but one that varies across contexts and circumstances. The evidence suggests that control in addiction is highly sensitive to relevant contextual features. People with addiction seem to find it much more difficult to abstain when they are with certain people or in certain places. For example, the tendency to experience drug craving is known to be highly context sensitive (Skinner and Aubin 2010). Thus, it is not only the intrinsic general ability to refrain that needs to be taken into account, but the specific ability that the person had in the particular circumstances under consideration. This is even more difficult to estimate.
A clinical assessment of the severity of the person’s addiction may be helpful, but it is at best an imperfect proxy for the kind of assessment that is relevant to moral responsibility. Diagnostic criteria, imperfect as they are, are developed with a specific goal in mind, namely, to identify those cases in which clinical intervention might be beneficial. Thus, if the relevant goal is to appraise degrees of responsibility, there is a real possibility that the criteria that are useful to clinicians will turn out to be an imperfect guide.
The DSM-5-TR distinguishes between mild, moderate, and severe forms of substance use disorder (American Psychiatric Association 2022, 546). In practice, the distinction is operationalized in terms of the number of diagnostic criteria met by the patient, as judged by the clinician. The “severe” level is applied to cases in which the patient meets six or more of the eleven diagnostic criteria listed in the manual. This is intended to capture an observation that amounts to clinical common sense: within the domain of cases sensibly described as ‘addiction,’ some are more severe than others. But the quantity of symptoms is at best an imperfect measure of the intuitive notion of severity. The criteria that are useful for the purposes of diagnosis are not all equally relevant to responsibility judgments. For example, whether the person is using more than intended or is failing to fulfill social roles with which she identifies, seems prima facie more relevant to responsibility than whether she is showing signs of tolerance to the drug. Furthermore, the quantity of symptoms approach to assessing severity is intended to capture a graded notion, but it does so by adding up how many of the criteria are met, and each of the criteria is decided by a categorical assessment. The very fact that the manual proposes to measure severity in this fashion is a testament to the difficulty of assessing levels of ability. The availability of a diagnosis from a competent clinician can provide guidance in making the sort of judgment that is relevant to moral responsibility, but it will often not be enough to settle the issue.
5.2 How much reasons-responsiveness did the situation call for?
Estimating with confidence how much control the agent had at the moment of action is not the only challenge for the blamer. Another particularly tricky issue arises when we consider how much control should have been sufficient to avoid wrongdoing in the situation at hand. Or, to put it differently, just how compromised control must be in order for an agent to qualify for mitigation of responsibility under the relevant circumstances.
This too is, in a way, a general difficulty we encounter when we think of responsibility itself, and of the agential abilities that make someone responsible for her actions, as matters of degree. Moral reasons are not created equal: some are more salient than others. We assume, for instance, that it requires less moral understanding to see that it is wrong to murder someone than it does to realize that a particular joke might be offensive to someone with a different cultural background, even if we consider both to be morally required. And the same goes for reactivity. We expect some actions to be so aversive to a morally competent agent that it would take less self-control to refrain from doing them. Briefly put, avoiding certain morally criticizable behaviors requires less in the way of moral competence. Just how much moral understanding and what degree of ability to govern oneself are required to avoid engaging in morally criticizable activities are not fixed parameters, and vary along different moral situations. Thus, even if we assume that an agent possesses the abilities relevant to moral responsibility to an imperfect degree, it may be the case that she is sufficiently capable to warrant the normative expectation that she does not behave in certain ways when the situation at hand is less demanding of moral competence. Fully spelling out the rationale for this would involve resolving some difficult issues in the theory of moral responsibility. But the addiction excuse seems to be more forceful in some cases than in others, depending on how high the moral stakes are. The more the consequences of a decision to use drugs involve serious moral harms, the more the intuitive appeal of the addiction excuse appears to weaken.16
Of course, judging the degree of moral competence that a particular situation calls for is a difficulty that people with the right standing to blame face in many sorts of cases that do not involve addiction. This, too, is a difficulty we are bound to face once we think of responsibility and moral competence as matters of degree. What matters in the present context is that this further difficulty adds to the challenge faced by potential blamers in addiction cases. Assuming that the person had a diminished or imperfect ability to control her behavior in a given situation, should that degree of control have been sufficient to prevent her from overlooking the relevant moral reasons for doing otherwise than she did?17
Think of Juan and Diego again. Diego rightly expected Juan to be kind and respectful to his parents who had invited him to their home, and Juan failed to respond to that expectation. When he decided to start drinking prior to their rendezvous, the prospect of letting Diego and his parents down did not exert sufficient pull on his deliberation to make him exercise his ability to refrain. However, if he had learned that there was a fire in the building he was in when he was about to pour his glass, he would probably have chosen to run from the fire instead of having a drink. And if he had believed that having that drink would, through some intricate causal chain, lead to Diego’s death, he would likely not have done it either. He had some control that he could have exercised if the situation had been dire enough. The prospect of letting Diego down and hurting his feelings, which he likely contemplated that afternoon, gave him a less salient and less compelling reason to refrain than the possibility of precipitating his death would have given him. Should that have been enough?
Note that post hoc expressions of remorse, apologies, and other attempts at relationship repair are relevant to the moral situation, albeit in a different way. Juan may use these means to demonstrate that he cares, that he recognizes the legitimacy of Diego’s expectations, and that he values their relationship. Despite the relevance of all this to their ongoing moral conversation, it speaks to a different kind of concern. Backward-looking responsibility is particularly grounded in the amount of agential control the person had at the time of action. Thus, whether Juan cares about past harms now does not substitute the need to assess whether he should have been able to respond to the relevant moral reasons at the time of action.
As before, there is no general solution to this difficulty. A person may be fully convinced that the imperfect ability to control behavior that we see in addiction can, in principle, provide a mitigating excuse for overlooking certain moral reasons in some situations. Notwithstanding this, the person with the right standing to blame will still be faced with the need to navigate the intricacies of ambivalent agential control in order to resolve what her response should be. Of course, how to respond to addiction-related moral harm is a complex question that depends on factors other than whether backward-looking blame is appropriate. For example, Diego could, perhaps should, consider which responsibility response would be more useful for them and for the relationship going forward. However, to the extent that the question is whether blame is deserved, as opposed to whether it would serve some forward-looking purpose, the search for an answer will lead one to ponder the difficulties arising from ambivalent behavioral control. This puts the potential blamer in a particularly difficult position: how much control the agent had, or how compromised her ability to respond to relevant moral reasons pertaining to the situation at hand was, is underdetermined by the available evidence. And whether the degree of control the agent had should have sufficed to respond to the relevant moral reasons calls for another challenging normative appraisal that is difficult to make with any confidence.
6. Conclusion
Can addiction be the basis for a mitigation of responsibility for addiction-related moral faults? In some cases, I have argued, it can. The reason is that addiction may partially impair the ability to control drug-related behavior. As a result, an addicted person’s responsiveness to moral reasons may be diminished when decisions to use drugs are at issue. On a plausible theory of moral responsibility, such a decrease in reasons-responsiveness affords not a full exemption or excuse, but a mitigation of responsibility for moral faults that are suitably connected with decisions to use drugs.
However, this offers only limited guidance when it comes to assessing degrees of blameworthiness in particular cases. Applying the addiction excuse involves further factual and normative appraisals that are particularly difficult to make. Moreover, getting these assessments right is often important to avoid the risks involved in both over- and under-blaming. For people with the right standing to blame, the need to navigate the intricacies of the ambivalent agential control that we see in addiction poses a significant ethical challenge.
Acknowledgments
I thank Migdalia Arcila-Valenzuela, Gerald Lang, Diego Lawler, Diana Pérez, and Tobías Schleider for their insightful comments on earlier versions of this paper. Thanks also to two anonymous reviewers for this journal. Research for this paper began while I was a Visiting Fellow at the School of Philosophy, Religion and History of Science, University of Leeds, and was completed while I was a Research Resident at the Brocher Foundation in Geneva (www.brocher.ch).