Buddhist philosophy and the historical problem of free will have each been of major philosophical interest for centuries, but until recently they have been studied separately and by scholars of different traditions. In recent decades, however, there has been growing interest among philosophers in the topic of Buddhism and free will—that is, what the various Buddhist traditions have said, can say, and should say, in response to the traditional problem of free will. ¹ For good or bad, much of the focus has been on the Buddhist “no‐self” doctrine. In Buddhism (or, more accurately, the various Buddhist traditions ² ), the term anattā (Pali) or anātman (Sanskrit) refers to the doctrine of “no‐self,” which maintains that there is no unchanging, permanent self, soul, or essence in living beings. It is one of the seven beneficial perceptions in Buddhism, and along with Dukkha (suffering) and Anicca (impermanence), it is one of the three Right Understandings about the three marks of existence. The Buddhist conception of anattā or anātman is one of the fundamental differences between Buddhism and Hinduism, with the latter asserting that Atman (self, soul) exists.

Given its centrality to Buddhism, many commentators have questioned whether the no‐self doctrine leaves any room for the notion of free will. As Christian Coseru writes, “Buddhism is unique among the world's great philosophical traditions in articulating a conception of action that, it seems, dispenses altogether with the notion of agent‐causation” (2019, xi). But if the agent/self is an illusion, indeed the central illusion responsible for all our suffering according to Buddhism, how could the agent/self have free will? Some commentators have argued that the no‐self doctrine amounts to a rejection of free will since it denies that we are autonomous moral agents (see, e.g., Strawson 1986, 2017; Goodman 2002; Blackmore 2013). Others have argued that the no‐self doctrine is consistent with the control in action required for free will (Siderits 1987, 2017; Griffiths 1982; Repetti 2017, 2019; Harvey 2017; Adam 2017; Meyers 2017). Still others have argued that since Buddhism has remained mostly silent about the problem of free will for over two millennia, we should adopt Buddhist quietism about free will. Christopher Gowans (2017), for instance, argues that the main reason Buddhist philosophical analysis has not addressed the problem of free will is that it is limited to soteriological parameters—that is, whatever promotes enlightenment. And Jay Garfield (2017) and Owen Flanagan (2017) argue that absent the Christian theodicy that generated the contemporary conception of free will, the problem does not and cannot arise in Buddhism. Unfortunately, there is no universal agreement on these issues.

In his recent book, Buddhism, Meditation, and Free Will: A Theory of Mental Freedom (2019), Rick Repetti argues for a conception of “agentless agency” and defends a view he calls “Buddhist soft compatibilism.” With regard to the no‐self doctrine, he writes: “My intuition is that regardless of how much the no‐self doctrine is repeatedly asserted to be the central doctrine—if not the sine qua non—of Buddhism, it is an open question to what extent elements of Buddhism imply that there is agency or even an agent” (2019, 147–48). He argues that “[p]resent arguments against the agent‐self and autonomy are inconclusive” (2019, 6), and that Buddhism itself can provide a coherent understanding of “agentless agency.” In fact, he argues, Buddhism presupposes the existence of agency when it claims that the master practitioner of meditation can have phenomenal meta‐level control over his or her mental states, thoughts, and intentions. This leads him to conclude: “If this sort of executive control obtains at the center of conscious agency among advanced meditation practitioners, then—at the heart of Buddhism—there are grounds for thinking there is mental autonomy, whatever the correct interpretation of the agent‐self turns out to be. Even if there is no agent‐self in some coherent sense, there still appears to be agency” (2019, 9). After arguing that Buddhism does not exclude the possibility of agency, rather it offers a means of mastering it through meditative practices, Repetti then proceeds to argue that there are sufficient grounds to defend a coherent Buddhist theory of free will.

When Repetti talks of a “Buddhist theory of free will” he means, “a theory Buddhists may adopt” (2019, 145). As he puts it: “I'm more interested in what Buddhists can say about free will than what they have said” (2019, 145–46). Repetti also maintains that “Buddhism and Western discussions of free will may fruitfully unite” (2019, xiii), and in developing his Buddhist theory of free will, he freely draws on both traditions. In the end, he defends a view he calls Buddhist soft compatibilism. It maintains, first, that, despite the Buddhist denial of the self, we have the ability to increase free will through Buddhist meditation practices (2019, 7). Second, Buddhist soft compatibilism maintains, “that Buddhist external and internal history is best understood as open to compatibilism between most if not all conceptions of free will, causation, and the self” (2019, 11). Using the categories of the contemporary free will debate, Repetti explains that “soft compatibilism” is the opposite of “hard incompatibilism” (see, e.g., Perboom 2001, 2014; Caruso 2012, 2020). As he explains:

Soft compatibilism is the opposite of hard incompatibilism. Hard incompatibilism unites two incompatibilisms, hard determinism and hard indeterminism, where free will is thought to be incompatible, respectively, with determinism, and with indeterminism. Hard incompatibilists think free will is incompatible with determinism and indeterminism, they assume these exhaust the possibilities, and they conclude there is no free will. Soft compatibilists unite two compatibilisms, soft determinism and soft indeterminism, where free will is thought to be compatible, respectively, with determinism and indeterminism, they reject the (false) dichotomy that assumes these exhaust the possibilities, and so they are open to alternative conceptions of causation, such as wiggly, Humean, and other forms of causation. (2019, 146)

Repetti's soft compatibilism therefore maintains that, “Free will is compatible with causation, however construed” (2019, 152). It further maintains that, “Buddhist Soft Compatibilism rebuts the most powerful Western arguments for free will skepticism” (2019, 148) since, according to Repetti: “If determinism is true, choices are not random, but reliably related to [reasons for action], so they can be up to us, and we can have source autonomy, and if indeterminism is true, we have alternatives, can do otherwise, and thus can have leeway autonomy” (2019, 152).

At the core of Repetti's positive account of free will is his conception of “mental freedom” or “freedom of mind,” which is informed by Harry Frankfurt's (1971) hierarchical account of freedom of the will and Buddhist meditative practices. According to Frankfurt's famous account, an action is free when it is consistent with an agent's higher‐order, reflective desires—that is, when the “lower‐order” desire motivating the action is endorsed by a “higher‐order” preference to want to have that desire. Repetti contends that the kind of “mental freedom” cultivated by Buddhist meditative practices, and the kind that grounds Buddhist free will, develops similar meta‐level abilities.

By approving dharmic mental contents and disapproving adharmic ones, she cultivates a dharmic hierarchical will in Frankfurt's (1971) metavolitional sense, and more broadly a dharmic hierarchical mind. For at the meta‐level she approves or disapproves not only her first‐order volitions, which is what Frankfurt's model of free will requires, but also the rest of her mental contents—thoughts, objects of attention, emotions, sensations, perceptions, imaginings, and various other mental states. (2019, 154)

Repetti's position, then, is that the concept of free will or volitional autonomy is a species of a larger genus of mental autonomy or “freedom of the mind,” which Buddhism aims at cultivating and which involves a variety of meta‐level abilities, for example, “to dispassionately examine, thus consciously approve or disapprove of, and thus control, non‐meta‐level thoughts, volitions, emotions, actions, and related mental states” (2019, 11). It is the cultivating of these meta‐level abilities that constitutes mental freedom and, presumably, free will.

While I think Repetti's book is a major contribution to the growing literature on Buddhism and free will and a must‐read for anyone interested in the topic, I must also take issue with my good friend's account of Buddhist soft compatibilism. In the following section, I will outline some general concerns I have with Repetti's Buddhist theory of free will and his overall focus and strategy. I will then turn, in the final two sections, to a more wide‐ranging discussion of Buddhism and free will—one that foregrounds Buddhist ethics and takes seriously what the various Buddhist traditions have said about desert, punishment, and the reactive attitudes of resentment, indignation, and moral anger. The thrust of my comments will be aimed at showing that, not only is Buddhism best conceived as endorsing a kind of free will skepticism, Buddhist ethics can provide a helpful guide to living without basic desert moral responsibility and free will. In particular, I will focus on three distinctly different Buddhist stances on punishment and argue that the best way to reconcile them is to adopt something like my nonretributive alternative, the public health‐quarantine model (Caruso 2016, 2017, 2020).

BUDDHIST SOFT COMPATIBILISM: SOME GENERAL CONCERNS

My first and most general concern with Repetti's account has to do with definitions. ³ Repetti talks about free will in different senses, sometimes equating it with mind‐control, other times reasons‐responsiveness, autonomy, or a kind of meta‐level control over one's non–meta‐level thoughts, volitions, emotions, actions, and related mental states. As Repetti knows, much of the debate over free will concerns how to define free will and whether autonomy, mind control, or even meta‐level control is enough to ground the control in action required for basic desert moral responsibility. I think it's a shortcoming of Repetti's account that he does not seriously discuss desert or moral responsibility. There are only three mentions of desert in the index, mainly in connection with other people's views, and Repetti does not explain how his account of mental freedom lines up with the issue of basic desert moral responsibility. This is unfortunate since the sense of free will that has been of central philosophical and practical importance in the historical debate is the sort required for basic desert moral responsibility—or so I have argued elsewhere (Caruso and Morris 2017, 2020). As Derk Pereboom defines this kind of moral responsibility:

For an agent to be morally responsible for an action in this sense is for it to be hers in such a way that she would deserve to be blamed if she understood that it was morally wrong, and she would deserve to be praised if she understood that it was morally exemplary. The desert at issue here is basic in the sense that the agent would deserve to be blamed or praised just because she has performed the action, given an understanding of its moral status, and not, for example, merely by virtue of consequentialist or contractualist considerations. (2014, 2)

It's unclear whether Repetti accepts this definition or not, but it's important whether he does. First, if free will is defined in terms of the control in action required for basic desert moral responsibility, then a comprehensive “Buddhist theory of free will” will also need to consider what, if anything, Buddhist ethics can tell us about desert‐based judgments, attitudes, or treatments relevant to free will—such as resentment, indignation, moral anger, backward‐looking blame, and retributive punishment. This is something I will explore in detail below. Second, Repetti would be required to explain why mind control or meta‐level control is enough to ground basic desert moral responsibility—something he does not seriously address. If, on the other hand, Repetti rejects this definition, then he would need to explain why a conception of “mental freedom” divorced from any and all issues related to basic desert and the justification of basically deserved praise, blame, and punishment should be considered an account of “free will” at all and not something else.

Second, requiring only that a Buddhist theory of free will be one a Buddhist may adopt is a rather low standard. On that standard, I see no reason why a Buddhist could not adopt a hard‐incompatibilist or skeptical theory of free will. In fact, Charles Goodman (2002, 2009, 2017) has argued that Buddhism is hard incompatibilist, in that, it considers free will impossible whether determinism or indeterminism is true—though he, himself, thinks Buddhist causation is determinist. And other commentators, some of them also practitioners, have argued that Buddhism supports free will skepticism (see, e.g., Strawson 1986, 2017; Blackmore 2013; Harris 2013; Wright 2017). Free will skepticism maintains that what we do and the way we are is ultimately the result of factors beyond our control—whether that be determinism, chance, or luck—and that because of this, agents are never morally responsible in the basic desert sense. I will argue below that not only is this skeptical perspective one a Buddhist may adopt, it is one they should adopt if they wish to take Buddhist ethics seriously.

Third, while Repetti does consider the various (Western) philosophical arguments in support of hard‐incompatibilism and free will skepticism, the replies he offers to them are far from conclusive. The case I favor for free will skepticism features distinct arguments that target three rival views, event‐causal libertarianism, agent‐causal libertarianism, and compatibilism, and then claims that the skeptical position is the only defensible position that remains standing (see, e.g., Pereboom 2001, 2014; Caruso 2012, 2020). I maintain that the sort of free will required for basic desert moral responsibility is incompatible with causal determination by factors beyond the agent's control and also with the kind of indeterminacy in action required by the most plausible versions of libertarianism. Against the view that free will is compatible with the causal determination of our actions by natural factors beyond our control, I argue that there is no relevant difference between this prospect and our actions being causally determined by manipulators (see Pereboom 2001, 2014; Mele 2008; Todd 2011, 2013). Against event causal libertarianism, I advance the “luck” or “disappearing agent” objection, according to which agents are left unable to settle whether a decision/action occurs and hence cannot have the control in action required for moral responsibility (see Pereboom 2001, 2014, 2017b; Caruso 2012, 2020; Waller 1990, 2011; Levy 2008, 2011; Mele 1999, 2017). The same problem, I contend, arises for noncausal libertarian accounts since these too fail to provide agents with the control in action needed for basic desert (see Pereboom 2014). While agent‐causal libertarianism could, in theory, supply this sort of control, I argue that it cannot be reconciled with our best physical theories and faces additional problems accounting for mental causation (Caruso 2012). Since this exhausts the options for views on which we have the sort of free will at issue, I conclude that free will skepticism is the only remaining position.

In addition to these hard incompatibilist arguments for free will skepticism, I have also recently defended Neil Levy's (2011) luck pincer (Caruso 2019). The luck pincer maintains that regardless of the causal structure of universe, free will and basic desert moral responsibility are incompatible with the pervasiveness of luck. This argument is intended not only as an objection to event‐causal libertarianism, as the luck objection is, but extends to compatibilism as well. At the heart of the argument is the following dilemma: either actions are subject to present luck (luck around the time of the action), or they are subject to constitutive luck (luck that causes relevant properties of agents, such as their desires, beliefs, and circumstances), or both. Either way, luck undermines moral responsibility since it undermines responsibility‐level control.

For Repetti's soft compatibilism to be a viable option, it would need to overcome most, if not all, of these incompatibilist arguments since it maintains that free will is compatible with both determinism and indeterminism—as well as “alternative conceptions of causation, such as wiggly, Humean, and other forms of causation” (2019, 146). While Repetti does his best to respond to the kinds of arguments just outlined, I find his defense of soft compatibilism inconclusive at best. His criticisms of the manipulation argument, for example, involve some misunderstandings and fail to consider important replies as well as recent expansions of the argument by Derk Pereboom, Patrick Todd, and others. ⁴ There is also only a passing treatment of the luck pincer, but no serious attempt to address it. Since I consider the luck pincer one of the stronger arguments in favor of free will skepticism, Repetti would need to address it in much more detail if he wishes to overcome the skeptical arguments against free will. I'm also not persuaded by his replies to the disappearing agent objection or the various objections to agent‐causal libertarianism. ⁵ But rather than litigate all these arguments here, I will settle for making two more general points.

First, soft compatibilism is an extremely demanding view, one that needs to be defended on many fronts, and it will stand or fall on its ability to defend all extant accounts of free will—compatibilism, event causal libertarianism, and agent‐causal libertarianism—since Repetti wants to leave all these options open for a Buddhist to embrace. If only one of the arguments in support of hard incompatibilism succeeds—say the disappearing agent objection against event causal libertarianism or the manipulation argument against compatibilism or so on—then soft compatibilism would need to be rejected. Personally, I find the arguments for hard incompatibilism persuasive, but I will not defend them here since that would take me too far afield and require more space than is available. I leave it to the reader to judge for themselves the success or failure of these various arguments.

Second, it's unclear to me why Repetti even takes on the dual task of defending libertarian accounts of free will and trying to reconcile them with Buddhist metaphysics. I know of only one philosopher who thinks Buddhists actually embrace a libertarian conception of free will (Griffiths 1982). Almost all others agree that the no‐self doctrine excludes the possibility of agent‐causal libertarianism and there is little reason to think Buddhists embrace event causal libertarianism. Perhaps Repetti would have been better served settling on a less ambitious approach. Nicholas Gier and Paul Kjellberg, for example, write: “While the issue of free‐will does not arise in Buddhism, it is indisputable that it embraces a universal determinism: every effect, without exception, has a cause. The idea that the will is uncaused or is self‐caused violates the Buddhist principle of interdependent coorigination (pratītyasamutpāda): nothing in the universe can originate itself as substances allegedly do or the will is said to do” (2004). Pratītyasamutpāda, commonly translated as “dependent origination” or “dependent arising,” is a key principle in Buddhist teachings, which states that all dharmas (“phenomena”) arise in dependence upon other dharmas: “if this exists, that exists; if this ceases to exist, that also ceases to exist.” The pratītyasamutpāda doctrine is a fundamental tenet of Buddhism and it may be considered as “the common denominator of all the Buddhist traditions throughout the world, whether Theravada, Mahayana, or Vajrayana” (Boisvert 1995, 6–7). According to pratītyasamutpāda teachings, there is nothing independent, except the state of nirvana (see Harvey 1990, 2015). All physical and mental states depend on and arise from other preexisting states, and in turn from them arise other dependent states when they cease. It's hard to see how a libertarian conception of free will can be reconciled with this doctrine of dependent origination, since it embraces a universal determinism of cause and effect for all phenomena, except the state of nirvana.

But rather than focus on these metaphysical doctrines, I suggest that if we truly seek to understand what Buddhism can teach us about the contemporary problem of free will, the best place to look is at its ethical teachings. Repetti, along with many others who write about Buddhism and free will, tends to focus, almost exclusively, on the metaphysical aspects of the problem. For Repetti, this includes reconciling volition agency with the no‐self doctrine, dependent origination, and the like. Very little attention is paid to Buddhist ethics and what it teaches about punishment and the reactive attitudes of resentment, indignation, moral anger, and blame. I will try to correct for this lacuna in the final two sections.

BUDDHISM AND DESERT

As we've seen, free will skepticism maintains that what we do and the way we are is ultimately the result of factors beyond our control and that because of this, agents are never morally responsible for their actions in the basic desert sense. It is important to note, though, that doubting or denying basic desert moral responsibility, as skeptics do, does not mean that other conceptions of responsibility cannot be reconciled with determinism, indeterminism, chance, or luck. In fact, many free will and moral responsibility skeptics have developed and promoted other non–desert‐based conceptions of responsibility—for example, Waller's “take charge responsibility” (2011, 2014) and Pereboom's forward‐looking account of moral responsibility (2014). In this section, I will argue that Buddhist ethics is most consistent with the skeptical perspective, and that the kinds of responsibility consistent with free will skepticism are sufficient to preserve the ethical teachings Buddhists care about most. I will also argue that Buddhist ethics provides sound practical advice, and can teach us a thing or two about living without resentment, indignation, moral anger, backward‐looking blame, and retributive punishment.

To begin, we can ask: Is the assumption that we are morally responsible in the basic desert sense required for the sorts of personal relationships we value? The considerations raised by P.F. Strawson in his essay “Freedom and Resentment” (1962) suggest a positive answer. In his view, our justification for claims of blameworthiness and praiseworthiness is grounded in the system of human reactive attitudes, such as moral resentment, indignation, guilt, and gratitude. Strawson contends that because our moral responsibility practice is grounded in this way, the truth or falsity of causal determinism is not relevant to whether we justifiably hold each other and ourselves morally responsible. Moreover, if causal determinism were true and did threaten these attitudes, as the free will skeptic is apt to maintain, we would face instead the prospect of the cold and calculating objectivity of attitude, a stance that relinquishes the reactive attitudes. In Strawson's view, adopting this stance would rule out the possibility of the meaningful sorts of personal relationships we value.

Strawson may be right to contend that adopting the objective attitude would seriously hinder our personal relationships. However, a case can be made that it would be wrong to claim that this stance would be appropriate if determinism did pose a genuine threat to the reactive attitudes (Pereboom 2001, 2014). While kinds of moral anger such as resentment and indignation might be undercut if free will skepticism were true, these attitudes may be suboptimal relative to alternative attitudes available to us, such as moral concern, disappointment, sorrow, and moral resolve. Optimistic free will skeptics maintain that the attitudes that we would want to retain either are not undermined by a skeptical conviction because they do not have presuppositions that conflict with this view, or else they have alternatives that are not under threat (see, e.g., Pereboom 2001, 2014; Milam 2017; Pereboom and Caruso 2018). And, what remains does not amount to Strawson's objectivity of attitude and is sufficient to sustain the personal relationships we value.

Buddhist ethics, I contend, promotes a similar view. The Buddha, for instance, identified anger (what P.F. Strawson calls resentment) as one of the three unwholesome roots of action for all humans. The other two are greed and delusion. Their opposites, the wholesome roots, which are also present in all people, are generosity, kindness, and clarity or wisdom. All of our actions spring from one of these six sources. The family of angry emotions includes everything from minor irritation to unbridled rage. Resentment, hatred, irritation, and mild annoyance are all forms of anger. While Strawsonians believe it would be impossible or undesirable to live without moral anger and resentment, Buddhist ethics teaches us that if we take things one event at a time, we can understand anger and apply its antidotes—patience, compassion, and forgiveness. There's a useful story from the Buddha's life about nonreactivity to anger. It goes like this (from SN 7:2, translation Bhikkhu Bodhi 2000):

On one occasion the Blessed One was dwelling at Rājagaha in the Bamboo Grove, the Squirrel Sanctuary: The Brahmin Akkosaka Bhāradvāja, Bhāradvāja the Abusive, heard: “It is said that another Brahmin of the Bhāradvāja clan has gone forth from the household life into homelessness under the ascetic Gotama.” Angry and displeased, he approached the Blessed One and abused and reviled him with rude, harsh words.

When he had finished speaking, the Blessed One said to him: “What do you think, Brahmin? Do your friends and colleagues, kinsmen and relatives, as well as guests come to visit you?” – “They do, Master Gotama”

– “Do you then offer them some food or a meal or a snack?” – “I do, Master Gotama.”

–“But if they do not accept it from you, then to whom does the food belong?” – “If they do not accept it from me, then the food still belongs to us.”

So too, Brahmin, I do not abuse anyone, do not scold anyone, do not rail against anyone. I refuse to accept from you the abuse and scolding and tirade you let loose at me. It still belongs to you, brahmin! It still belongs to you, Brahmin!

Brahmin, one who abuses his own abuser, who scolds the one who scolds him, who rails against the one who rails at him – he is said to partake of the meal, to enter upon an exchange. But we do not partake of your meal; I do not enter upon an exchange. It still belongs to you, brahmin! It still belongs to you, brahmin!

The Buddha's point is that one who repays an angry man with anger thereby makes things worse for himself. But by not repaying an angry man with anger, one wins a battle hard to win—for both themselves and the other—since anger often stands in the way of reconciliation and moral formation. In fact, as is often the way in these stories, in the end Bhāradvāja is transformed by his encounter with the Buddha, becomes a monk, and, under the Buddha's guidance, eventually achieves complete awakening.

Shantideva, the sixth‐century Buddhist commentator, gives another example in A Guide to the Bodhisattva's Way of Life: “Suppose a person hits you with a stick. It does not make sense to be angry at the stick for hurting you, since the blows were inflicted by a person. Neither, he continues, does anger toward the person make sense, since the person is compelled by anger (or greed or delusion). Ignorance becomes the villain, overwhelming reason and creating suffering” (Boorstein 2014). We blame and criticize others because we don't like suffering. But according to Buddhist ethics, if we don't like suffering, we should not harm others and create more disharmony since this interferes with our own happiness. The following two verses from the Dhammapada Sutta, a collection of saying of the Buddha, capture this point nicely. In Verse 222, the Buddha explains: “Those who hold back rising anger like a rolling chariot are real charioteers. Others merely hold the reins” (translation Easwaran 2007). And in Verse 223, we are told: “Conquer anger through gentleness, unkindness through kindness, greed through generosity, and falsehood by truth” (translation Easwaran 2007). And in the Brahmajala Sutta, the Buddha says:

Monks, if anyone should speak in disparagement of me, of the Dhamma or of the Sangha [the community of which they were members], you should not be angry, resentful or upset on that account. If you were to be angry or displeased at such disparagement, that would only be a hindrance to you. For if others disparage me, the Dhamma or the Sangha…then you must explain what is incorrect as being incorrect, saying: “[For this or that reason] that is incorrect, that is false, that is not our way, that is not found among us.” (translation Maurice Walshe)

To be clear, the above passage should not be mistaken as encouragement of indifference to blame and praise. Rather, what the Buddha advocates is a calm, clear, equanimous, matter‐of‐fact recognition of the nature of others’ comments and actions, followed by clarifications of the erroneous and the affirmations of the right. It is only with right mindfulness, according to Buddhism, that we can discern what needs to be addressed and respond appropriately with patient compassion. ⁶

Furthermore, none of this means that one cannot point out others’ mistakes. It's just that, according to Buddhist ethics, when one does so, it should be done with lovingkindness and compassion, and only when there is forward‐looking benefit in doing so. The Buddha makes this clear in the Ańguttara Nikāya when, in speaking to the monks, says:

It was said: “One should not utter covert speech, and one should not utter overt sharp speech.” And with reference to what was this said?

Here, monks, when one knows covert speech to be untrue, incorrect, and unbeneficial, one should not utter it. When one knows covert speech to be true, correct, and unbeneficial, one should try not to utter it. But when one knows covert speech to be true, correct, and beneficial, one may utter it, knowing the time to do so.

So it was with reference to this that it was said: “One should not utter covert speech, and one should not utter overt sharp speech.” (MN 139, MLDB 1083–84; translation Bodhi 2016)

This restriction of covert speech and overt sharp speech (which includes various kinds of moral criticism) to only those cases where it is warranted and there would be forward‐looking benefit, closely resembles Pereboom's (2014) forward‐looking account of moral responsibility.

As I indicated above, our moral practices feature a number of senses of moral responsibility, some of which do not invoke basic desert. Pereboom (2014), for instance, has proposed that when we encounter immoral action, we might ask the agent to consider what his actions indicate about his intentions and character, to demand apology, or to request reform, thereby having him consider reasons to behave differently in the future. Engaging in such interactions counts as reasonable in view of the right of those wronged or threatened by wrongdoing to protect themselves from bad behavior and its consequences. Our practice also features an interest in the wrongdoer's moral formation, and the address described naturally functions as a step in this process. Moreover, our practice also has a stake in our reconciliation with the wrongdoer, and calling him to account plausibly serves as a stage in securing this aim. Such interactions, because they address the agent's capacity to consider and respond to reasons, manifest respect for her as a rational being. On Pereboom's forward‐looking account, then, moral responsibility and moral exchange are grounded, not in basic desert, but in three non–desert‐invoking desiderata: future protection, future reconciliation, and future moral formation. Not assuming basic desert, such an account is consistent with free will skepticism. It is also consistent, I contend, with Buddhist ethics since it satisfies the Buddha's restriction of moral criticism to only those cases where there is future benefit. Hence, both Buddhism and free will skepticism prohibit purely backward‐looking blame, anger, and retribution.

BUDDHISM AND PUNISHMENT

I would now like to turn to the question of state‐sanctioned punishment and whether it can be reconciled with Buddhist ethics. I will argue, once again, that Buddhist ethics favors practices and policies consistent with free will skepticism. In particular, I will argue that Buddhists not only may but should adopt something like my public health‐quarantine model for addressing criminal behavior (Caruso 2016, 2017, 2020; Pereboom and Caruso 2018). I will frame my discussion around the following dilemma:

The Buddhist Punishment Problem (BPP): On the one hand, Buddhist ethics tells us that the intentional infliction of harm on another is an infringement of the principle of non‐violence (ahimsā) and those sovereigns who engage in punishment will accrue negative consequences in this and future lives. On the other hand, we are also told that Buddhist sovereigns have the dual duties of protecting their people and punishing evildoers. But if Buddhist kings and sovereigns must occasionally engage in punishment for the purpose of statecraft, how can they remain ethical?

Following Michael Zimmerman (2006), I will identify three different Buddhist stances on punishment and explore how each attempts to resolve this problem. As we'll see, there are tensions between the various stances. But I will argue that the public health‐quarantine model provides the best possible way of reconciling them. My exergies of the three stances will follow closely the work of Zimmerman (2006) and will be centered on texts from the earlier period of Indian Buddhism.

I will begin, as Zimmerman does, by first considering the history of ancient Indian statecraft and Brahmanic kingship. As Zimmerman explains: “As in medieval Europe, so too in ancient India there existed a rich and imaginative set of customs concerning the measures to be applied when it came to punishing criminals and violators of traditional codes of behavior. The old textbooks on jurisprudence, the dharmasūtras and dharmaśāstras, the composition of which began in the last centuries before the Common Era and clearly bear the imprints of a brahmanically dominated society, prescribe a wide variety of such punishments” (2006, 214). The relevant parts of these books prescribe detailed punishments for all different kinds of transgressions. We are also told that no one other than the king himself was in charge of dispensing justice and deciding on the punishment (see Zimmerman 2006, 214–15). Ancient Indian texts, both Brahmanic and non‐Brahmanic in nature, are in unison in charging the king with these two main obligations.

Given that the guidelines for kingly governance are clearly laid down, one can only wonder how a sovereign who considered himself a Buddhist could adopt these traditional and general rules of statecraft? This is the Buddhist punishment problem. As Zimmerman asks: “Would [a Buddhist sovereign] not have to throw overboard the first of the five precepts to be followed by all lay Buddhists—namely the abstention from intentionally killing or injuring sentient beings, one of the main tenets in Buddhist self‐perception and with which Buddhism is widely identified?” (2006, 216). As I've already foreshadowed, there is no simple standardized answer to this question: “Indian Buddhist thinkers have been aware of the difficulties posed for their tradition and have struggled to ease the tension between an eventual need for the carrying out of punishment and their understanding of non‐violence” (Zimmerman 2006, 216–17).

Zimmerman (2006) identifies three distinctly different stances on punishment found in texts of the earlier period of Indian Buddhism. He labels them the idealist, ethical fundamentalist, and compassionate stances. The idealistic view of how a Buddhist king should reign was that of the cakravartin, the wheel‐turning king, who, as he is described in the Pāli Cakkavatti‐sīhanāda Suttanta (DN III 58–77) and other texts (see, e.g., Mahāsudassana Sutta, DN II 169–98), has conquered the four quarters of the earth and established stability, rules over them without the need for punishment or other violence, and encourages his subjects to live according to the five precepts. While the story provides us with a utopian outlook of the ideal Buddhist ruler, it offers very few concrete guidelines on what to do if crimes do take place and if stability in the country is not maintained (Zimmerman 2006, 217). For this reason, Zimmerman writes, “[t]he early ideal of the Buddhist universal emperor as he is presented in the narrative thus avoids a realistic discussion of the possible need for the application of punishment, let alone its ethical and karmic implications” (2006, 217).

The ethical fundamentalist approach is more radical since the Buddhist ideal of nonviolence is here uncompromised. According to this stance, to become a king/ruler means to break the precept of ahimsā, however “good” the motivation for the decision to do so might be. While this second approach is ready to confront a less ideal society than the proponents of the cakravartin utopia would like us to hope for, it is ethically unflinching:

[Here] punishment is uncompromisingly judged as a violation of Buddhist ethics equally as unwholesome as stealing, lying, and so on. This position offers no room for a reconciliation of the issue and rigorously rejects any kind of retrenchment at the expense of the Buddhist standard of ethics, which, in this strand of thought, is believed to be universally valid and thus does not support the idea that a member of the kșatriya class would have to fulfill his particular duty (svadharma). (Zimmerman 2006, 218)

According to this Buddhist stance, there is no viable way of combining religious practice and statecraft, and, ultimately, there would be no incentive for becoming involved in ruling. Rulers who found it necessary to punish wrongdoers would therefore not be exempt from the negative karmic consequences. And the ethical prohibition on punishment, as well as the negative karmic consequences for those who engage in it, would extend to both retributive and consequentialist forms of punishment.

There are plenty of representatives of this rigid approach throughout both the more conservative schools of early Buddhist and Mahāyāna writings (see also Collins 1998, ch.6). To mention just a few, there is, for example, the jātaka of the prince Temīya, who knows and remembers by his own experience that the throne of a king can only lead to hell. He decides to act as if he were lame, deaf, and dumb, with the sole purpose of escaping the royal duty awaiting him, even at the expense of being put to death (Mūgapakkha Jātaka 538 [6], 1–30). Another example of this uncompromising stance, is Candrakīrti, the Madhyamaka philosopher from the first half of the seventh century. In his commentary on Aryadeva's Catuhśataka, its fourth chapter being a critical analysis of the king's role in the light of a universal Buddhist set of ethics, Candrakīrti reflects on the king's fulfilling his specific royal duties. As Zimmerman summarizes it, according to Candrakīrti's commentary:

[T]he king cannot but produce negative results for his soteriological situation. The king's axiomatic guideline is the view that the fulfillment of his proper duty as a ruler—namely protecting his subjects by punishing evildoers—would come along with spiritually wholesome after‐effects for himself. This, however, cannot work, say Candrakīrti, since the king punishes without empathy: and the application of such violence does counteract the dharma (in its universally valid Buddhist meaning), just as butchers and fishermen are unaware that they produce unwholesome effects by killing animals in the belief that they have to follow their designated lineages assigned by birth. The outcome for the ruler thus cannot be positive: “A ruler without empathy has no merit at all since [his] violence is enormous” (CTt 82al). (Zimmerman 2006, 220)

These are just two examples of the ethical fundamentalist Buddhist stance. Candrakīrti's commentary, for example, is clearly aimed at showing that the Brahmanic conception of kingship is utterly unacceptable from a Buddhist standpoint. It should be noted, though, that several times in his commentary Candrakīrti emphasizes that royal violence is problematic because the king acts without empathy (sanskrit: dayā). This leads Zimmerman to question whether violent punishment with a compassionate motivation could be an appropriate alternative (2006, 222).

This brings us to the third Buddhist stance on punishment. Here, Zimmerman notes that both Mahāyāna and Pāli sources stress the beneficial role of the king for his subjects. One of the best known sets of guidelines for such a ruler in the Pāli sources is the list of the 10 so‐called “royal virtues” (rājadhamma), which usually comprise alms‐giving (dāna), morality (sīla), liberality (pariccāga), honesty (ajjava), mildness (maddava), self‐restriction (tapas), nonanger (akkodha), nonviolence (avihimsā), patience (khanti), and nonoffensiveness (avirodhana) (see PTSD S.V. rājadhamma; Zimmerman 2006, 224). The virtue of nonviolence in this list would appear to preclude such violent acts as warfare and punishment, yet many Pāli texts leave no doubt that punishment of evildoers is indeed part of the king's business (see Zimmerman 2006, 225). It would seem, then, that we are once again confronted with the Buddhist punishment problem.

Focusing on Indian Mahāyāna sources, Zimmerman argues that the rules for punishing in some of those texts contain an important additional element that cannot be found in the traditional brahmanic law books: compassion. He writes:

The inclusion of this element, the central notion of Mahāyāna ethics, as one of the guiding principles for the king, modified the ideas about the implementation of punishment in at least two decisive ways. One is the idea that punishment, more than satisfying feelings of retaliation, has to serve the improvement and rehabilitation of the evildoer in this life. The second is a tendency toward the application of milder forms of punishment and, in the best case, the absolute exclusion of certain forms of punishment that in their results are irreversible. (2006, 227–28)

He points to the royal policy chapter of the Ratnāvalī, attributed to the second‐century philosopher Nāgārjuna (RĀ 4.100; see 1995), as a representative of this kind of argument. In this work, Nāgārjuna advises a king on how to rule his territory based on Buddhist principles.

With regard to prisoners, he admonishes the king to treat them with compassion (especially those who have committed the most horrible deeds like murder) and to take good care of their physical needs with barbers, baths, drinks, food, medicine, and clothing. He advises the ruler to look at evildoers just as he would look at his children, whom he would punish with compassion to make them improve their behavior and not out of hatred or desire for wealth. Nāgārjuna further elaborates that the king should not kill or torment a criminal but, instead, banish a murderer from his territory. (Zimmerman 2006, 228)

Another example of the compassionate stance can be found in the Bodhisattvabhūmi, a part of the vast Yogācārabhūmi, and early Yogācara work (see Zimmerman 2006, 229–30). One final example can be found in the Bodhisattva‐gocaropāya‐visaya‐vikurvana‐nirdeśa‐sūtra. There we are told that: (1) if a matter can be solved without the application of “harsh forms of punishment,” the king should simply declare the crime of the lawbreaker (this could involve a simple public proclamation); (2) the king should never kill wrongdoers (hence, a prohibition on the death penalty); and (3) when punishing, the king should cultivate a mental state of “friendliness and compassion.” The text also explicitly states that a king loyal to the dharma should try his best to rehabilitate offenders and treat them like his children. In fact, the king is even compared to a physician who without anger applies himself to the treatment of the patient (see Zimmerman 2000, 196).

While this third stance has a number of advantages over traditional retributive punishment, and perhaps goes some distance in resolving the Buddhist punishment problem, it's still hard to see how it can be reconciled with the ethical fundamentalism of the second stance. While the compassionate stance avoids the death penalty and other irreversible forms of punishment, as well as prioritizes the rehabilitation, improvement, and well‐being of wrongdoers, it still condones punishment in milder forms. I propose instead that the most consistent position for a Buddhist to adopt is something like my nonretributive, nonpunitive alternative: the public health‐quarantine model.

Very briefly, the model takes as its starting point Derk Pereboom's (2001, 2014) famous account. In its simplest form, it can be stated as follows: (1) Free will skepticism maintains that criminals are not morally responsible for their actions in the basic desert sense; (2) plainly, many carriers of dangerous diseases are not responsible in this or in any other sense for having contracted these diseases; (3) yet, we generally agree that it is sometimes permissible to quarantine them, and the justification for doing so is the right to self‐protection and the prevention of harm to others; (4) for similar reasons, even if a dangerous criminal is not morally responsible for his crimes in the basic desert sense (perhaps because no one is ever in this way morally responsible) it could be as legitimate to preventatively detain him as to quarantine the nonresponsible carrier of a serious communicable disease.

The first thing to note about the theory is that although one might justify quarantine (in the case of disease) and incapacitation (in the case of dangerous criminals) on purely utilitarian or consequentialist grounds, Pereboom and I want to resist this strategy. Instead, our view maintains that incapacitation of the seriously dangerous is justified on the ground of the right to self‐defense and defense of others. That we have this right has broad appeal, much broader than utilitarianism or consequentialism has. In addition, this makes the view more resilient to a number of objections and provides a more resilient proposal for justifying criminal sanctions than other nonretributive options (see Caruso 2020). Second, the quarantine model places several constraints on the treatment of criminals. First, as less dangerous diseases justify only preventative measures less restrictive than quarantine, so less dangerous criminal tendencies justify only more moderate restraints. Second, the model demands a degree of concern for the rehabilitation and well‐being of the criminal that would alter much of current practice. Just as fairness recommends that we seek to cure the diseased we quarantine, so fairness would counsel that we attempt to rehabilitate the criminals we detain. Rehabilitation and reintegration would therefore replace punishment as the focus of the criminal justice system. This, it should be noted, is a point strongly endorsed by the Buddhist compassionate stance. Lastly, if a criminal cannot be rehabilitated and our safety requires his indefinite confinement, this account provides no justification for making his life more miserable than would be required to guard against the danger he poses.

In addition to these restrictions on harsh and unnecessary treatment, the public health‐quarantine model also advocates for a broader approach to criminal behavior that moves beyond the narrow focus on sanctions. It places the quarantine analogy within the broader justificatory framework of public health ethics. Public health ethics not only justifies quarantining carriers of infectious diseases on the grounds that it is necessary to protect public health, it also requires that we take active steps to prevent such outbreaks from occurring in the first place. Quarantine is only needed when the public health system fails in its primary function. Since no system is perfect, quarantine will likely be needed for the foreseeable future, but it should not be the primary means of dealing with public health. The analogous claim holds for incapacitation. Taking a public health approach to criminal behavior would allow us to justify the incapacitation of dangerous criminals when needed, but it would also make prevention a primary function of the criminal justice system. The public health framework I adopt also sees social justice as a foundational cornerstone to public health and safety (Caruso 2020). In public health ethics, a failure on the part of public health institutions to ensure the social conditions necessary to achieve a sufficient level of health is considered a grave injustice. An important task of public health ethics, then, is to identify which inequalities in health are the most egregious and thus which should be given the highest priority in public health policy and practice. The public health approach to criminal behavior likewise maintains that a core moral function of the criminal justice system is to identify and remedy social and economic inequalities responsible for crime. Just as public health is negatively affected by poverty, racism, and systematic inequality, so too is public safety. So instead of myopically focusing on punishment, the public health‐quarantine model shifts the focus to identifying and addressing the systemic causes of crime, such as poverty, low social economic status, systematic disadvantage, mental illness, homelessness, educational inequity, exposure to abuse and violence, poor environmental health, addiction, and the like.

The public health‐quarantine model offers, I contend, an ethically defensible and practically workable alternative for dealing with dangerous criminals, one that is more humane and effective than retributivism. It also provides a possible resolution to the Buddhist punishment problem and helps reduce the tension between the three stances just discussed. The public health‐quarantine model captures the essential components of the compassionate stance by prioritizing rehabilitation and reintegration, prohibiting the death penalty and other harsh forms of punishment, and requiring that the mental states or intentions of the punishing authority not be retaliation or retribution but instead be guided by compassion. But unlike the compassionate stance, the public health‐quarantine model can also be reconciled with the complete abandonment of punitive practices and policies, making it more compatible with the second (i.e., ethical fundamentalist) stance. Since legal punishment requires the intentional imposition of a penalty for conduct that is represented as a violation of a law of the state (Boonin 2008; Zimmerman 2011), and since the public health‐quarantine model does not involve punishment in this way, Pereboom and I consider it a nonpunitive alternative to treatment of criminals (see Caruso 2020). When we quarantine an individual with a communicable disease in order to protect people, we are not intentionally imposing a penalty for illegal conduct. The same is true when we incapacitate the criminally dangerous in order to protect people. The right of self‐defense and protection of harm to others justifies the limiting or restricting of liberty, but it does not constitute punishment as standardly understood. This allows us to see how the second and third stances could be united and how the duty to protect public safety could be made more consistent with Buddhist ethics. Even the idealist stance can be made some sense of on the public health‐quarantine mode, since it is theoretically possible (though not likely) that by adopting the right set of preventive practices, policies, and attitudes, we could altogether eliminate the need for punishment and/or incapacitation. And even if this is only an ideal, it is something to strive for.

CONCLUSION

In conclusion, the skeptical perspective is not only one a Buddhist may adopt, it is one they should adopt if they wish to take Buddhist ethics seriously. While Repetti's account of mental freedom provides important and interesting insights into Buddhist meditative practices and how they can enhance degrees of mental autonomy, we should reject the idea that such “mental freedom” amounts to a “Buddhist theory of free will.” For one, if free will is defined in terms of the control in action required for basic desert moral responsibility, then a comprehensive Buddhist theory of free will also need to consider what, if anything, Buddhist ethics can tell us about desert‐based judgments, attitudes, and treatments relevant to free will—such as resentment, indignation, moral anger, backward‐looking blame, and retributive punishment. This is exactly what I attempted to do in the “Buddhism and Desert” and “Buddhism and Punishment” sections, where I turned to a more wide‐ranging discussion of Buddhist ethics and what it has to say about desert, punishment, and the reactive attitudes. I argued that, not only is Buddhism best conceived as endorsing a kind of free will skepticism, Buddhist ethics can provide a helpful guide to living without basic desert moral responsibility and free will. In the “Buddhism and Desert” section, I argued that Buddhists, like free will skeptics, reject backward‐looking blame and anger. I also discussed some practical advice on how to eradicate these harmful reactive attitudes. I then concluded, in the “Buddhism and Punishment” section, by examining state‐sanctioned punishment and whether it can be reconciled with Buddhist ethics. I discussed three different Buddhist stances on punishment and argued that the best way to reconcile them is to adopt the public health‐quarantine model.