2. The Motivating Reasons Conception

The first premise of the argument—the Motivating Reasons Conception—is the theoretical (evaluative) claim that the moral status of persons—including whether or the degree to which persons are culpable—depends, ultimately, on the (final) reasons for which they act, and not on their beliefs or intentions. The appeal of this claim may be illustrated by the observation that even beliefs and intentions that may seem objectionable can turn out to be commendable. For example, the fact that an agent believes that her action would harm an innocent person, and even the fact that she intends to harm this person, do not in themselves entail that the agent is culpable (not even in one respect). Whether or not she is culpable, and, if she is, the degree to which she is culpable, ultimately depends on the final reasons that motivate her action. If, for instance, she intends to harm an innocent person only because (she believes that) this is the only way to prevent much more serious harm to another innocent person; she considers the lesser harm as a reason against the action (although one that is defeated by the stronger reason to prevent the greater harm); and, finally, she thinks that it is morally worse to allow the greater harm, she does not seem to be culpable.Footnote ⁷ Indeed, she may well be praiseworthy.Footnote ⁸

Alexander Sarch objects that a person who brings about a bad state of affairs (for example, harms another) intentionally is more culpable in one respect (albeit not necessarily overall) than a person who brings about the same state of affairs non-intentionally—although he suggests that this is the case only regarding actions that are not justified all things considered.Footnote ⁹ However, if the above conditions are met, I think that the former agent is not more culpable than the latter even in one respect. This is the case, it seems to me, both if the relevant actions are wrong and if they are not.Footnote ¹⁰ Sarch’s suggestion that intentions affect the culpability of agents when their actions are wrongful—but not when they are not—seems to me arbitrary.Footnote ¹¹ Thus, a person who harms another to prevent greater harm is not culpable, given the above conditions, even if her action is wrong (this may be the case, for instance, if there is a deontological prohibition on harming persons intentionally that is decisive in this case, or if the agent is mistaken about a relevant fact).Footnote ¹²

Another example that illustrates why culpability ultimately depends on motivating reasons rather than beliefs or intentions is the case of euthanasia. As noted above, Hurd and Moore point out that a mercy killer and a contract killer may not be equally culpable even if they have the same beliefs and intentions—the belief that a certain action will cause death and the intention to bring about this outcome—due to the different reasons for which they act. More specifically, Antill highlights the case of Dr. Jack Kevorkian, who helped terminally ill patients to end their suffering by way of assisted suicide. Despite acting with the intention to cause death, agents such as Dr. Kevorkian may not be culpable, or at least may be much less culpable than the typical intentional killer, if their ultimate reason is to prevent suffering rather than to cause harm. In this respect, the fact that Dr. Kevorkian was tried for one of the most serious offenses—murder—based on the Beliefs & Intentions Model, appears to be unjust. In contrast, Antill argues that the police officer who kneeled on George Floyd’s neck and thus caused his death may be more culpable than the typical intentional killer (and especially more than agents such as Dr. Kevorkian). This is the case, he argues, even if the police officer did not intend to kill, and even if he did not think that his action would cause death, if his decision not to consider whether this may be the case was due to the fact that he failed to ascribe proper weight to the life of Floyd. Accordingly, the implication of the Beliefs & Intentions Model—that he is liable to punishment that is much less severe than that of the typical intentional killer—seems too lenient.Footnote ¹³

The precise implications of the Motivating Reasons Conception depend on the nature of the motivating reasons that it identifies as important in terms of the moral status of agents. In this regard, there are many different versions of the Motivating Reasons Conception.Footnote ¹⁴ Indeed, prominent versions of this conception hold that the extent to which persons are praiseworthy or blameworthy depends on the relationship between their motivating reasons and the good (normative, guiding) reasons that apply to their actions.Footnote ¹⁵ These versions hold that people are culpable when, and to the degree to which, they engage with normative reasonsFootnote ¹⁶ in a way that is flawed—for instance, in a way that does not track normative reasons appropriately, or (more specifically) does not reflect sufficient concern for the interests of others.Footnote ¹⁷ These versions thus require not only an answer to the question of what are the reasons that motivate the agent but, in addition, an answer to the question of which reasons are good ones. Thus, the familiar controversies regarding the latter question (for example, between consequentialism and deontology as well as between different versions of these views) affect the details of the Motivating Reasons Conception.

The fact that there are many versions of the Motivating Reasons Conception may trigger the question of whether the Beliefs & Intentions Model may be one version of the Motivating Reasons Conception. In one respect, this is indeed the case. First, some of the mental states that the Beliefs & Intentions Model considers (such as purpose, knowledge, recklessness, or negligence) may reflect some of the reasons that motivate the agents. In addition, as noted above, certain beliefs and intentions may correlate with certain motivating reasons. However, there is an important difference between the Motivating Reasons Conception and the Beliefs & Intentions Model in this respect. Plausible versions of the Motivating Reasons Conception hold that what ultimately matters in terms of the moral status of persons is the final reasons that motivate their actions—their ends (in the strict sense of final ends) as opposed to reasons that are derived from their final reasons, such as what they consider merely as means to these ends.Footnote ¹⁸ This is because the final reasons of agents are those that explain their other reasons (at least if they are rational).Footnote ¹⁹

This does not entail that other (motivating) reasons are irrelevant to culpability, but it does entail that other reasons are relevant to culpability only if, and to the degree to which, they are related to the final reasons that ultimately matter in this respect. For example, as noted above, the fact that a person intends to harm another does not necessarily entail that she is culpable if her final reason is to prevent greater harm (and she thinks that she is doing the right thing). In contrast, the Beliefs & Intentions Model focuses on mental states that may not refer to the final reasons for which the relevant persons act (as the above examples of harming to prevent greater harm and euthanasia demonstrate). Accordingly, the following discussion is compatible with all of the plausible versions of the Motivating Reasons Conception that go beyond the common categories of mens rea to which the Beliefs & Intentions Model refers.Footnote ²⁰

Given its plausibility, it is not surprising that the Motivating Reasons Conception is widely accepted, including by criminal law theorists. Indeed, the Motivating Reasons Conception is sometimes part of additional arguments regarding criminal law (beyond the argument for the conclusion that criminal laws should consider motivating reasons to a greater degree). For example, James Edwards and Andrew Simester argue, based on this conception, that the degree to which agents are culpable does not depend on the outcomes of their actions (although the latter may affect the moral status of their actions).Footnote ²¹

Furthermore, even when the Motivating Reasons Conception is rejected, the alternative usually triggers a similar concern regarding criminal laws that reflect the Beliefs & Intentions Model. The most important alternative may be the view that the moral status of persons ultimately depends on their character. This view does not vindicate criminal laws that reflect the Beliefs & Intentions Model. On the contrary: It suggests an even more acute concern about such laws, since the divergence between beliefs or intentions and character may be even more substantial than the divergence between beliefs or intentions and motivating reasons.Footnote ²² In what follows, I focus on the argument that relies on the Motivating Reasons Conception of culpability, but significant parts of the discussion are also relevant to an alternative argument that endorses instead a character-based account of culpability.

Before considering the other premises of the argument, a final terminological clarification should be noted. The relevant literature on criminal law often uses the term ‘motives’ as a contrast to intentions. This term is ambiguous (as many have pointed out). Indeed, it is often used in senses covering many mental states not part of the Beliefs & Intentions Model. Thus, it is sometimes used in senses that refer to motivating reasons (although not necessarily the final ones). At other times, it is used in senses that focus more on emotions or on considerations that do not motivate actions in themselves.Footnote ²³ The distinction that matters for the arguments considered in this paper is the one between the (final) motivating reasons that bear on the culpability of persons (especially as agents) and the mental states that are part of the Beliefs & Intentions Model.Footnote ²⁴

3. The Culpability Principle

The Culpability Principle, recall, is the normative claim that moral culpability should be a central concern of criminal law. This proposition is attractive and widely accepted.Footnote ²⁵ However, as noted above, it is vague,Footnote ²⁶ and the way in which it is explicated may be important to the question of whether the argument for the conclusion that criminal law in the relevant legal systems should focus more on motivating reasons is sound.

Versions of the Culpability Principle may differ regarding its content, scope, and force, as well as concerning the values that underlie the principle (which entail the answers to the former questions). For example, according to a general version of the principle, criminal law should be concerned with all aspects of culpability. In contrast, more specific versions distinguish between the positive claim that there is a reason to punish those who are culpable to the degree to which they are culpable, and the negative claim that there is a reason against punishing people beyond the degree to which they are culpable. It is common, for example, to accept the negative claim but not the positive one, or to consider the negative claim as (much) more stringent.Footnote ²⁷ Even more specifically, some distinguish, within the negative claim, between the claim that there is a reason against punishing those who are culpable beyond the degree to which they are culpable, and the more specific claim that there is a reason against punishing those who are ‘innocent’, namely, not culpable at all (at least with respect to actions proscribed by criminal law). Often, the latter reason is assumed to be stronger. Indeed, some consider it as decisive in all cases, regardless of the nature and force of countervailing reasons.Footnote ²⁸ In contrast, the more general (negative) reason against punishing culpable agents beyond the degree to which they are culpable is less often assumed to be necessarily decisive, and no one seems to think that the positive reason to punish culpable agents is always decisive.Footnote ²⁹

Finally, and even more specifically, there seems to be a distinction between two ways of understanding the sense of innocence that the narrow claim employs. The focus in this regard is more often on cases in which a person did not perform an action that is attributed to her (we may refer to such cases as cases of ‘factual innocence’). Yet another sense is innocence (which may be labeled as ‘normative innocence’), which refers to cases in which a person performed the action that is attributed to her but is not culpable—for example, since (she thinks that) the action is not wrong (some cases of euthanasia may be of this type, for instance). Factual and normative innocence are different in some respects. However, both are cases of moral innocence. This is because what is important, with respect to moral innocence, is (only or also) whether a person is responsible for a wrongful action rather than if she is responsible for an action attributed to her (by someone). An accurate declaration of innocence is “I did nothing wrong” rather than “I didn’t do what you say I did.” The latter claim is significant only if, and because, it entails that the relevant person did nothing wrong. In this respect, both factual and normative are the same.Footnote ³⁰ Nevertheless, these types of innocence are often distinguished (usually implicitly). For example, it is more common to hold that the reason against punishing innocents is very strong, and even necessarily decisive, with respect to cases of factual innocence compared to cases of normative innocence.

These differences in the content, scope, and force of the relevant reasons are related to different views regarding the values that ground them. The positive claim that there is a reason to punish culpable agents is often based on the view that desert is valuable in itself.Footnote ³¹ (The same is true regarding the more general claim that criminal law should be concerned with the degree to which people are culpable, since this claim includes the positive claim.) In contrast, the negative claims, and especially the narrowest claim against punishing agents who are (completely) innocent, are more often associated with the view that there is a deontological prohibition (which is often assumed to be necessarily decisive) on punishing innocents or punishing people beyond the degree to which they are culpable.Footnote ³² (Of course, if or the degree to which agents are culpable may be important also due to other reasons, including reasons that do not consider culpability as important in itself.)

These different accounts of the Culpability Principle entail different versions of the argument from the Motivating Reasons Conception to the conclusion that laws that adopt the Beliefs & Intentions Model are unjust and should be amended. Accordingly, these differences affect the answer to the question of whether (the relevant version of) this argument is sound. For example, an argument that relies on the more general and more stringent versions of the Culpability Principle, such as the one that holds that criminal laws must always track culpability accurately, is more likely to be valid (since it is less likely that laws that reflect the Beliefs & Intentions Model consider motivating reasons to a sufficient degree given these versions). However, since these versions are less plausible (and much more controversial), the argument may not be sound if it relies on them.

It may be thought that we need not resolve these difficult questions regarding the appropriate version of the Culpability Principle in order to conclude that laws that adopt the Beliefs & Intentions Model are unjust. Consider the following argument:

This argument does not refer explicitly to the Culpability Principle.Footnote ³³ Therefore, it is important to consider if it is indeed possible to dispense with the Culpability Principle in this way.

The above argument—which compares the Motivating Reasons Conception and laws that reflect the Beliefs & Intentions Model—seems to assume that both are concerned with the same project. Indeed, otherwise, they could not be incompatible. Since the Motivating Reasons Conception is clearly, and explicitly, concerned with assessing the degree to which agents are culpable, comparing it to the laws that adopt the Beliefs & Intentions Model implies that these laws, too, are concerned with evaluating culpability.

It is a common assumption that the laws that adopt the Beliefs & Intentions Model reflect judgments about the degree to which agents are culpable. Thus, for example, Ken Simons writes that the Model Penal Code—a paradigm example of the Beliefs & Intentions Model—“views its four basic mental states or culpability terms as hierarchically ordered: all else being equal, purpose is more culpable than knowledge, which is more culpable than recklessness, which is more culpable than negligence.”Footnote ³⁴ Similarly, as Sarch points out, “the criminal law is often said to embody a culpability hierarchy. Modern criminal law recognizes four main mental states one might act with—negligence, recklessness, knowledge and purpose—each supposedly more culpable than the last.”Footnote ³⁵ Sarch himself endorses a similar assumption when he writes that when a legal system does not distinguish between cases in which an agent who commits arson knows that her action may harm a person, and cases in which she does not, “[t]here is a sense in which this jurisdiction attributes the same amount of culpability for both forms of arson.”Footnote ³⁶ Likewise, Antill writes that:

Thus, he concludes that laws that reflect the Beliefs & Intentions Model are committed to an account of culpability that is rejected by theorists who endorse instead the Motivating Reasons Conception. Indeed, he argues that this conception and the relevant laws reflect “two differing pictures of subjective culpability,” one that “attributes culpability to the proximate mental states (like an intention) behind an agent’s acts,” and another that “attributes culpability to the more distal mental states (like the reasoning behind the intention to act).”Footnote ³⁸

It is not completely clear in what sense exactly laws that reflect the Beliefs & Intentions Model are assumed to assert claims about the degree to which people are culpable. Is this, for example, a purely descriptive claim, for instance, about the intentions of the officials who made the relevant laws,Footnote ³⁹ or rather a partly normative claim about the most plausible understanding of these laws?Footnote ⁴⁰ Regardless of the answer to this question, however, the important point is that the above argument that compares the Motivating Reasons Conception and laws that reflect the Beliefs & Intentions Model is not valid even given the assumption that these laws assert claims about the degree to which people are culpable. Indeed, even assuming that the two premises of the argument are true—namely, that the Motivating Reasons Conception and laws that reflect the Beliefs & Intentions Model assert incompatible accounts of culpability and that the former is correct in this regard, it does not follow that the relevant laws are flawed and should be amended. For this conclusion to follow, the argument should include the additional premise that these laws ought to reflect (in some way) an accurate account of culpability and nothing more.

This claim appears to be false, however. Presumably, laws should reflect the balance of all applicable practical (normative) reasons that apply to them rather than the most accurate answer to a single theoretical question—including the degree to which persons are culpable. Thus, the relevant laws should be concerned with a project that is different (indeed, different in type) than that of the Motivating Reasons Conception. And, of course, if they are indeed concerned with distinct (types of) projects, they are not (and could not be) contradictory.

The above argument fails for this reason, even if (some) laws that follow the Beliefs & Intentions Model reflect a theoretical claim regarding the degree to which people are culpable and nothing more (perhaps because this was the objective of some of those who drafted or voted for the relevant laws and assuming that this objective determines the content of these laws). For if this is the case, the relevant laws are flawed, as they should not be concerned (only) with this theoretical question. To the degree to which, due to this error, these laws do not properly reflect the balance of all applicable (normative) reasons, they are flawed and should be amended. However, they need not be amended such that they assert the proposition of the Motivating Reasons Conception but rather such that they reflect the balance of all the normative reasons that apply to them.

This failure of the above argument is instructive more generally (as I explain in the conclusion of the paper). Arguments that criticize laws based only on their alleged incompatibility with a certain moral standard are deficient to the extent to which they do not consider additional applicable reasons.Footnote ⁴¹

Thus, in order to conclude based on the Motivating Reasons Conception that laws that reflect the Beliefs & Intentions Model are unjust and should be revised, the argument must include some version of the Culpability Principle. Those who endorse the above argument may implicitly assume some version of the principle. However, since there are many versions of this principle, and different versions may not be equally plausible or entail the same conclusions, such an assumption is not obvious.

Moreover, the argument is more likely to be valid if it includes the most radical version of this principle, namely, one that holds that criminal law should always impose liability in a way that accurately reflects (in some sense) the degree to which the relevant persons are culpable (and never impose liability when they are not culpable or beyond the degree to which they are culpable). Indeed, as noted above, other versions of the Culpability Principle, which do not require that criminal laws always track culpability accurately, may well be compatible with laws that reflect the Beliefs & Intentions Model, given that such laws often consider motivating reasons to some degree.

However, the most radical version of the Culpability Principle is much less plausible and much less common than other versions. Indeed, no one seems to endorse it. For example, no one seems to hold that the (positive) reason to punish culpable agents is necessarily decisive, inter alia, since this may require that we invest all resources in criminal punishment. It seems that the suggestion that the Culpability Principle is always decisive is only made regarding its most narrow version, which prohibits punishing people who are not culpable at all, and even then, the focus is usually on cases of factual as opposed to normative innocence. This version of the Culpability Principle accordingly entails only a very limited version of the argument under consideration. Specifically, this version does not apply to the much more common cases in which criminal laws impose liability in ways that do not track culpability accurately. Furthermore, even this very limited version of the Culpability Principle is not, I think, plausible. This is not the place to resolve this familiar controversy. Therefore, I only note a few brief comments that highlight the degree to which this version is radical (and uncommon).Footnote ⁴²

The view that a certain reason (such as the reason to construct and use criminal laws in a way that tracks culpability accurately) is necessarily decisive may take one of two forms. The first is that this is the only moral reason in general, or at least the only one that applies in a certain context, such as that of criminal law. The second is that while there are other (applicable) reasons, they are always defeated by the reason to track culpability accurately.

The first option is uncommon. While criminal law theorists often emphasize the significance of culpability, and accordingly the reason to construct and use criminal laws in ways that track culpability, they typically hold that this is not the only moral concern that is pertinent to criminal law.Footnote ⁴³ For example, Moore writes that “retributivists are not monomaniacal about the achieving of retributive justice” and that “there are other intrinsic goods besides giving culpable wrongdoers their due and sometimes these other goods override the achievement of retributive justice.”Footnote ⁴⁴ Similarly, Larry Alexander and Kimberly Ferzan hold that the most plausible theory of punishment must trade off retributive justice “against the values of societal welfare, distributive justice, and corrective justice.”Footnote ⁴⁵

Still, it is worth noting that the option that the reason to track culpability accurately is the only applicable reason is less implausible than it may seem. It may be argued, for example, that there is no reason to promote the well-being of a person who does not deserve more well-being. Similarly, it may be argued (and has been argued) that there is no reason to give priority to the worse off if she does not deserve to be better off. Indeed, the assumption that desert is valuable in itself (which many criminal law theorists accept) does not seem to sit well with the hypothesis that there are other (applicable) intrinsic values (although this hypothesis is also widely accepted). Consider a case in which a culpable agent deserves a certain punishment, given his level of culpability. If we assume that not only desert is valuable in itself, but also, for example, maximizing the sum of well-being or allocating it equally or in a way that gives some priority to the worse off, it seems to follow that in addition to the reason to impose the deserved punishment (given the degree to which the offender is culpable) there are also countervailing reasons, against imposing this punishment—and indeed any punishment—since this would be detrimental to the offender’s well-being or make him worse off than others (who are not culpable). Accordingly, the balance of all of these reasons may require that the offender should be punished to a lesser degree than he deserves, given his level of culpability alone. This may seem misguided. There do not appear to be such reasons against imposing deserved punishment, even if, as a result of this punishment, the offender is worse off (than others). In other words, it does not seem that the values of maximizing well-being or giving priority to the worse off provide reasons to promote the well-being of the offender, independently of what he deserves, such that these reasons should be balanced against desert.Footnote ⁴⁶

However, we need not resolve this issue here. For even if desert is the only pertinent value, a plausible account of this value (for example, one that holds that the currency of desert is well-being) entails not only a reason against undeserved punishment but also a reason to prevent other forms of undeserved harm, inter alia by using criminal laws. Thus, if criminal punishment that is more or less than what a person deserves, given the degree to which she is culpable, would deter crimes more effectively than deserved punishment, the value of desert itself entails a reason in favor of punishment that is more or less than what is deserved. Moreover, as noted above, the more common version of the claim that the Culpability Principle is always decisive does not refer to a general reason to give people what they deserve but rather to the much more limited reason against punishing those who are not culpable at all.

The latter claim has considerable initial appeal. However, it is less common than it may be thought, and I think that it is ultimately implausible.Footnote ⁴⁷ Assume, for instance, that there is a deontological reason against harming people who are not culpable, or in a way that goes beyond the degree to which they are culpable (a reason that applies, inter alia, to undeserved punishment when it is harmful). The suggestion that this reason is always decisive, namely, necessarily defeats all other moral concerns, appears to be unreasonable. For example, if all else is equal, it is presumably not always wrong to harm even a person who does not deserve to be harmed if this is the only way of preventing much more serious harm to another person—who also does not deserve to be harmed (and indeed deserves the much greater harm less). For instance, it does not seem wrong to direct a trolley that would otherwise kill an innocent person onto a side track in which it would break the finger of another innocent person (assuming that all else is equal and there is no other way to prevent the death of the first person).

The hypothesis that the reason against undeserved punishment, more specifically, and especially against punishing innocents, is always decisive may seem more attractive. However, it is doubtful if it is reasonable for similar reasons. For example, traffic offenses often impose mild punishment, such as a small fine, in ways that also apply to some drivers who are innocent (in one of the above senses—the factual or the normative), since this is the only way of saving the lives of several innocent people.Footnote ⁴⁸ While it is no doubt unjust (in one respect) to punish innocents, it is also unjust (in another respect) if innocent persons are harmed (and there is accordingly a reason to prevent this harm when it is possible to do so). Thus, if enacting and enforcing traffic offenses that impose small fines (also) on some innocent drivers is the only way to prevent serious harm to other innocent persons, it is at least doubtful if doing so is necessarily—and thus always—wrong, all things considered.

Two further points are worth noting in this regard. First, as mentioned above, it is uncommon to hold that the reason against criminal laws that do not track culpability accurately is always decisive, namely, defeats every countervailing reason and every combination of such reasons. Even the more specific claim that the reason against undeserved punishment is always decisive is uncommon. What is more common is the much more specific claim that the reason against punishing innocents—those who are not culpable at all—is always decisive. Yet it is doubtful if the distinction between punishing those who are not culpable at all and punishment that exceeds the degree to which a person is culpable is significant such that the reason against the former is always decisive while the reason against the latter is not. Indeed, the reason against punishment that exceeds the degree to which a person is culpable to a very significant degree may be stronger than a mild punishment, for example, a small fine, that is imposed on a person who is not culpable at all.

Second, the hypothesis that the reason that criminal laws accurately track culpability is always decisive is especially implausible given the assumption that culpability ultimately depends on motivating reasons. This is the case since there are likely to be additional countervailing reasons that are due to this assumption. Most obviously, actions performed by agents who are not culpable, or less culpable, given their motivating reasons, may still be wrong and undesirable (for instance, when the agents are mistaken about the relevant normative standards or facts).Footnote ⁴⁹ It is also clear that it may be more difficult and costly in various respects for law enforcement officials (such as prosecutors or judges) to ascertain whether or to what degree people are culpable given their (final) motivating reasons, compared to determining mental states such as beliefs and intentions. This is likely to be the case, since there is often more evidence regarding mental states closer to the agents’ actions, such as their immediate goals, and less evidence concerning mental states further removed from these actions, especially their final ends. This difficulty is exacerbated by the fact that agents may have several reasons that motivate their actions and that these reasons may not always be fully transparent, even to the agents themselves.

Additionally, laws that consider motivating reasons may incentivize agents to hide their true motives to avoid criminal liability. In this respect, too, it may be more difficult and costly to prevent this phenomenon, compared to laws that focus instead on beliefs and intentions.Footnote ⁵⁰ A related concern is that laws concerned with the motivating reasons of agents may be less clear than laws that refer to beliefs and intentions, since there appear to be numerous motivating reasons that may affect the degree to which agents are culpable. As a result, laws that consider such reasons are more likely to take the form of standards as opposed to rules (such as a ‘lack of ill will’ defense), or to include such standards (in addition to rules).Footnote ⁵¹ Therefore, when, and to the extent to which, there are reasons against vague laws, these reasons may be more significant with regard to laws that focus on motivating reasons.Footnote ⁵² Finally, inter alia given the previous concerns, laws that are sensitive to motivating reasons may make the relevant legal processes more complex and expensive. Overall, then, there are likely to be weighty reasons against laws that consider the motivating reasons of agents. As a result, these reasons may sometimes defeat the countervailing reasons that support such laws.

In light of the doubts regarding versions of the Culpability Principle that hold that it is always decisive, it is worth considering weaker versions of this principle, which may be more plausible and also wider in scope. One suggestion for such a weaker and wider version of the Culpability Principle was recently made by Antill as part of his argument from the Motivating Reasons Conception to the conclusion that laws that adopt the Beliefs & Intentions Model are unjust and should be revised. Although he often seems to suggest the above argument, which compares the Motivating Reasons Conception and the Beliefs & Intentions Model directly, he sometimes appears to recognize that this argument fails without a certain version of the Culpability Principle. Indeed, he emphasizes several factors that are irrelevant to the former argument. First, he acknowledges the possibility of countervailing reasons against laws that track culpability accurately. Second, he considers, in light of this possibility, the hypothesis that beliefs and intentions may be good proxies for what ultimately counts in terms of culpability, namely, motivating reasons.

Third, he rejects this hypothesis based on the claim that the discrepancy between the Motivating Reasons Conception and laws that reflect the Beliefs & Intentions Model may be substantial. For example, he argues that a common version of the Beliefs & Intentions Model is committed to the view that an agent who harms another intentionally is (always) more culpable than one who harms another unintentionally (for instance, recklessly), while the Motivating Reasons Conception entails that the former is sometimes much less culpable than the latter. This may be the case, for example, when an agent who harms another intentionally acts for reasons that are better than those of the reckless agent—for instance, when the former considers the harm to the victim as a strong reason against her action whereas the latter is indifferent concerning the harm or considers it as less important than it actually is. Similarly, he points out that Dr. Kevorkian, who caused death intentionally, may be much less culpable than the officer who killed Floyd unintentionally. Finally, Antill claims that laws that have this result—punishing much less culpable agents more harshly than those who are much more culpable—violate a “weak proportionality principle” that criminal laws should respect.Footnote ⁵³ According to this principle, “substantially less culpable agents” should not be held “substantially more criminally liable than substantially more culpable agents for the same criminal act.”Footnote ⁵⁴

This principle is thus another version of the Culpability Principle—one that does not apply only to laws that target those who are completely innocent, on the one hand, and is not always decisive, on the other hand.Footnote ⁵⁵ However, it is unclear if an argument that includes this version of the Culpability Principle, in addition to the Motivating Reasons Conception and the Beliefs & Intentions Model, is sound. More specifically, there are three related doubts regarding this version of the argument: What exactly does the Culpability Principle require if it takes the form of the above proportionality principle? Is this principle correct? And do laws that adopt the Beliefs & Intentions Model violate this principle? These doubts are of course related: Determining the content of the proportionality principle is necessary in order to consider if it is plausible and if the relevant laws violate it.

Consider first the content of the proportionality principle. As noted above, this version of the Culpability Principle does not require that criminal laws always track culpability accurately. This is an advantage since, as noted above, the Culpability Principle is not always decisive. However, it is unclear what exactly the “weak” proportionality principle requires. The above explication of this principle—based on the proposition that “substantially” less culpable agents should not be held “substantially more criminally liable than substantially more culpable agents for the same criminal act”—is vague. As a result, it is unclear if laws that adopt the Beliefs & Intentions Model violate this principle, especially since these laws do consider motivating reasons to some extent (directly and indirectly).Footnote ⁵⁶

Finally, and most importantly, the appropriate version of the Culpability Principle is that the degree to which criminal laws should track culpability depends on the conclusions entailed by the balance of all applicable (normative) reasons, including both the reason that criminal liability and punishment track culpability accurately, as well as countervailing reasons. Thus, the best articulation of the argument is the following one:

This version of the argument is valid (its premises entail its conclusion). Moreover, the first two premises of the argument (the Motivating Reasons Conception and the Balancing Principle) appear to be correct. The question is therefore whether the third premise is also correct (and accordingly if the argument as a whole is sound), namely, if laws that reflect the Beliefs & Intentions Model violate the Balancing Principle.

The answer to this question is not obvious, since the implications of the Balancing Principle—the degree to which it requires that criminal law tracks culpability—depend on complicated questions and are likely to be contingent. These include difficult normative questions regarding the content of the relevant moral standards and how they should be balanced (when they pull in different directions), and various empirical questions concerning the facts that are morally significant according to these standards—for example, the degree to which constructing or using criminal laws in various ways would deter crimes and thus prevent harm (including undeserved harm). Accordingly, the degree to which criminal law should track culpability may be different at different times and places. Finally, these implications depend on the degree to which laws that take the form of the Beliefs & Intentions Model consider motivating reasons. The next section considers this question.

4. The Beliefs & Intentions Model

The final premise of the argument is the descriptive claim that criminal law often focuses more on the beliefs and intentions of agents and less on their motivating reasons. In order to consider if the relevant legal systems—those that adopt the Beliefs & Intentions Model—violate the version of the Culpability Principle (the Balancing Principle) that requires that criminal law treat agents in a way that properly reflects the balance of all pertinent reasons, we should consider the degree to which these legal systems consider motivating reasons. The less they do, the more likely it is that they violate the Balancing Principle and vice versa. Before considering this question, however, two preliminary points should be noted.

First, the main defining feature of the legal systems that adopt the Beliefs & Intentions Model is that the central categories of ‘mens rea’ are certain beliefs or intentions (regarding facts that are legally significant). For example, these legal systems usually impose a harsher punishment when an agent intends to harm another, compared to when she believes that her action may harm another but does not intend this result (if other things are equal, including, for example, the degree of harm).Footnote ⁵⁷ This is a feature of the criminal law of many countries, including the United States,Footnote ⁵⁸ the United Kingdom,Footnote ⁵⁹ and Canada,Footnote ⁶⁰ for instance (despite some variations).

Second, the claim that criminal law often focuses more on the beliefs and intentions of agents and less on their motivating reasons refers, I assume, to the position of a legal system as a whole rather than to a certain part of the legal system, or a certain legal rule, considered in isolation. Thus, when evaluating the position of a legal system in this respect, we should consider not only the content of criminal prohibitions, but also pertinent defenses and sentencing rules, for example. Accordingly, the fact that substantive criminal law often does not consider motivating reasons—a fact emphasized by Sarch, for instance—is not important in the relevant respect, given that (as Sarch acknowledges) other parts of criminal law usually consider motivating reasons to a considerable degree.Footnote ⁶¹ To the extent that the position entailed by a certain part of the law is different from the position of the legal system as a whole, we should focus on the latter. I also assume that the position of a legal system as a whole depends not only on its explicit legal rules, but also on the way in which they are interpreted and applied.Footnote ⁶² What matters, in terms of the argument under consideration, is the degree to which a legal system considers pertinent motivating reasons, and not how this is done.

The fact that criminal laws that adopt the Beliefs & Intentions Model usually consider motivating reasons to a considerable degree is widely acknowledged.Footnote ⁶³ The slogan that ‘motives’ (which are often understood to include some motivating reasons) are irrelevant to criminal law is inaccurate.Footnote ⁶⁴ This is acknowledged even by those who argue that the relevant laws are incompatible with the Motivating Reasons Conception, although the latter sometimes downplay the extent to which criminal laws consider motivating reasons.

Laws that follow the Beliefs & Intentions Model consider motivating reasons in various ways. First, while beliefs and intentions do not always align with motivating reasons, this may be the exception. For example, if most people who cause harm to others do not do that for good (praiseworthy) reasons, beliefs and intentions may often be good proxies for motivating reasons.Footnote ⁶⁵ There are also other factors that the Beliefs & Intentions Model considers and that may be proxies for motivating reasons. For instance, Edwards and Simester suggest that the outcomes of one’s action may be indicative of how the agent deliberated.Footnote ⁶⁶

Furthermore, when beliefs, intentions, and other factors are not good proxies for motivating reasons, the relevant legal systems often take this into account in various ways. One example is the common rule that criminal prohibitions proscribe only unjustified (or ‘unreasonable’) risks. According to this rule, the answer to the question of whether actions involve unjustified risks depends also on the reasons for which they are performed. A more important reason may justify greater risk and vice versa.Footnote ⁶⁷

Another way in which criminal law considers motivating reasons is by providing explicit affirmative defenses, such as self-defense, defense of others,Footnote ⁶⁸ lesser evil,Footnote ⁶⁹ necessity,Footnote ⁷⁰ and duress,Footnote ⁷¹ which apply in some cases in which people violate legal prohibitions for reasons that are either praiseworthy or less blameworthy than usual.Footnote ⁷² Indeed, such defenses sometimes apply only if the relevant actions are performed for the right reasons.Footnote ⁷³ However, these defenses usually apply only in some cases where people act for good reasons. Typically, for example, they are limited to situations that are not covered by any other legal rule and to cases in which no pertinent public official can intervene.

Moreover, in addition to these legal rules that pertain to the question of conviction, motivating reasons are considered even more often as part of sentencing rules, such as those that refer to mitigating and aggravating factors. Indeed, even the misleading slogan that motives are irrelevant to criminal law is meant to apply only to the former type of rules.Footnote ⁷⁴ Thus, for example, hate (or bias) crimes sometimes consider certain motivating reasons as aggravating factors.Footnote ⁷⁵

Furthermore, as noted above, we should consider not only how legal rules (regarding both conviction and punishment) are formulated, but also how they are interpreted and applied. Motivating reasons are often considered by various legal officials, such as police officers, prosecutors, judges, and juries, even if the legal rules do not refer to them explicitly.Footnote ⁷⁶

Finally, when a legal system law does not consider certain motivating reasons that are relevant to culpability, this is not necessarily because these motivating reasons are (mistakenly) assumed (by those who make or apply the law) to be irrelevant to culpability. Rather, this may be due to controversies and mistakes regarding the question of which reasons are (more or less) good or bad (and accordingly which agents are more or less praiseworthy or blameworthy). More importantly, a decision not to consider certain motivating reasons may be based on the assumption that there are decisive (normative) countervailing reasons against considering the relevant motivating reasons. And this assumption may be justified in some cases.

These points may be illustrated even regarding the case of Dr. Kevorkian (which was highlighted as an illustration of the opposite proposition—that criminal law is incompatible with the Motivating Reasons Conception). Kevorkian was indeed tried and convicted of one of the most serious offenses (second-degree murder) based on the fact that he acted with an intention to kill, even though he was presumably less culpable than most (second-degree) murderers (and arguably not culpable at all) due to his reasons for action.Footnote ⁷⁷ However, it is important to notice also the following facts, due to which this example is misleading in several respects.

First, and most importantly, using criminal law in response to euthanasia, especially in a way that involves a murder charge, appears to be unusual. As Husak points out, although euthanasia may often seem to violate the prohibition on murder, and is not covered by a formal defense, criminal law is often not employed in response to euthanasia, and one explanation for this is that the agents are not deemed to be culpable or at least sufficiently culpable.Footnote ⁷⁸

The case of Kevorkian was unusual not only in that he was charged but also in other ways that may explain why he was treated differently. These include the doubts as to whether all the people whom he assisted were indeed terminally ill, suffered pain, and were firm in their desire to end their lives, as well as the fact that he helped many people to end their lives while advocating his actions in a very public manner. Moreover, even when he was finally charged, he was first acquitted several times before his conviction, and his ultimate sentence was very different than that of most murderers.Footnote ⁷⁹

In addition, to the extent to which euthanasia is not tolerated by the relevant legal systems, this is not necessarily due to the view that motivating reasons are irrelevant to culpability. One alternative explanation is the controversy, which is internal to the Motivating Reasons Conception, regarding the question of whether specific motivating reasons—such as those that may support euthanasia—are praiseworthy.Footnote ⁸⁰

Furthermore, there may be (normative) reasons against tolerating euthanasia, even assuming that the relevant agents are not culpable (or less culpable than the typical intentional killer). These may include, for example, a concern that allowing euthanasia in some cases, even ones in which it is justified, may encourage unjustified actions. Indeed, such concerns may have been especially salient in the case of Kevorkian, given the fact that he advocated his actions publicly and the doubts as to whether all the people whom he assisted suffered pain and were terminally ill. More generally, a law that imposes the same sentence—regardless of whether the agent acted for a certain reason—may be justified not based on the assumption that this reason does not affect the agent’s degree of culpability, but due to other (countervailing) reasons.

The fact that criminal law often considers motivating reasons may also be illustrated regarding another case that was highlighted as an example of the opposite proposition: the case of the police officer who caused George Floyd’s death. The officer in this case may indeed have been much more culpable than most of those who cause the death of others unintentionally and without attending to the fact that their actions may cause death, due to his motivating reasons.Footnote ⁸¹ However, he was in fact convicted of offenses (including second-degree murder) that are more serious than those that are attributed to agents who cause death negligently. And he was accordingly sentenced to a much longer prison sentence (indeed, his eventual prison sentence is likely to be much longer than that of Dr. Kevorkian). It may be objected that these facts are beside the point because they do not reflect the law accurately.Footnote ⁸² Yet, given the many ways in which the relevant laws often consider motivating reasons, it seems that the fact that they were considered in this case, too, is not an aberration.

Overall, then, the relevant criminal laws often take into account motivating reasons that are relevant to culpability to a considerable degree,Footnote ⁸³ albeit not always.Footnote ⁸⁴ And since there are sometimes countervailing reasons against doing so—and especially doing so even more than these laws already do—it is unclear if these laws are flawed in this respect. At least, the argument that is based on the Motivating Reasons Conception and the (balancing version of the) Culpability Principle does not demonstrate this. More specifically, the above laws are not necessarily committed to the assumption that agents who intend harm, for example, are always more culpable than those who do not (other things being equal). Of course, it is unlikely that the laws under consideration are perfect in the extent to which they consider motivating reasons, but for all we know they may be flawed not because they do not do that enough but because they do it too much.

5. The General Lesson

The argument that highlights the tension between the proposition that the culpability of agents ultimately depends on their motivating reasons and the fact that criminal laws often focus on their beliefs and intentions is interesting and important. However, an accurate formulation of this argument reveals several difficult normative and empirical questions concerning, for example, the force of culpability-related reasons compared to other moral concerns, and the degree to which criminal laws take account of motivating reasons, directly and indirectly, in addition to beliefs and intentions. Thus, the general conclusion that may be drawn from the above argument is important but modest: Criminal law should consider motivating reasons that affect the degree to which agents are culpable (including in the framework of its liability rules) when the (normative) reasons against doing so either do not apply or are weaker than the (normative) reasons to track culpability accurately. More definite conclusions require addressing the complex normative and empirical questions that impact the relative force of these reasons in specific contexts. Given the contingent nature of the answers to the empirical questions, these conclusions are likely to vary across different jurisdictions and legal frameworks.

The doubts regarding a more definite general conclusion highlight a more general lesson about normative arguments regarding the law. Indeed, the essence of these doubts applies to many arguments for and against various laws. Such arguments are usually very ambitious, as they require addressing two very complicated tasks. One is a consideration of all the normative reasons that are relevant to the question at hand, including general reasons that are not unique to this question.Footnote ⁸⁵ Without considering all pertinent concerns, an argument for a conclusion regarding the question of what the law should be is not valid.

The argument from the Motivating Reasons Conception to the conclusion that laws that adopt the Beliefs & Intentions Model are unjust and should be revised illustrates this point. In order to determine if the normative premises of this argument are correct, we need to resolve, inter alia, the controversies regarding questions such as which motivating reasons are praiseworthy or blameworthy (and the degree to which each reason is praiseworthy or blameworthy). In addition, we need to resolve the controversies concerning the content of consequentialist and deontological reasons that may apply—for instance, if there are deontological reasons that require (or entail) that criminal law tracks moral culpability (more or less) accurately (and, if there are, what is their exact content and force). The second task that such arguments must undertake is an evaluation of all the empirical questions that are relevant, given the applicable normative concerns. For example, in order to determine if the above argument is sound, we need to know the degree to which various alternative formulations of criminal prohibitions would deter more crimes and thus prevent more (undeserved) suffering. These tasks are not only difficult but also require different types of expertise. Yet without completing both tasks successfully, normative arguments regarding the law are not sound. Therefore, academic writing about the question of what the law should be may often be too ambitious.

To be sure, this concern is a matter of degree. Some normative and empirical assumptions are more evident than others, and certain assumptions may be fairly obvious. However, at least most of the interesting and controversial arguments rely on claims that are not of this type. More specifically, it is important to notice that the above concern is not limited to arguments that rely on a consequentialist assumption concerning the applicable normative standards. This is the case since all plausible and prevailing moral theories, including deontological ones, consider (in addition to deontological reasons) also consequentialist reasons.Footnote ⁸⁶ The only view that entails that there is no need to consider countervailing reasons is one that assumes a deontological reason that is necessarily and thus always decisive, regardless of all countervailing reasons. However, as I have pointed out above, this assumption is doubtful and uncommon.Footnote ⁸⁷ Moreover, even those who endorse it do so only regarding certain deontological reasons (for instance, only regarding narrow versions of the Culpability Principle). Accordingly, even given the uncommon assumption that certain deontological reasons are always decisive, it is unnecessary to consider countervailing reasons only in the limited number of cases in which these reasons apply.

Of course, some people—for example, legislators, prosecutors, and judges—must often make decisions about how the law should be constructed and used. In these cases, a decision that is based on limited information, and accordingly on speculations, is unavoidable. But this requirement—and the justification that it entails for such speculations—rarely applies to academic work, which is accordingly justified only to the degree that its contribution is based on a systematic discussion. This usually requires limiting the scope of the discussion to only some of the questions that are pertinent to the overall evaluation of the laws under consideration. As long as academic works nevertheless address the question of what the law should be, all things considered, they accordingly ought to be more modest and acknowledge the assumptions that they rely on but do not argue for—which, as noted above, would typically be many—and clarify that the recommendations regarding the law are accordingly conditional (namely, depend also on these unargued-for assumptions).