JUSTIFICATIONS, EXCUSES, AND THE BINARY NATURE OF JUSTIFICATION

An agent is justified in committing a criminal offence if there is sufficient reason for him to do so, rendering his conduct right or at least permissible under the circumstances. To use two oft-cited illustrations, stealing a ladder may be justified if it was intended to be used in rescuing people trapped in a flaming building, and assault may be justified when it is needed to ward off a more harmful attack. Excuses, by contrast, form a more fluid category. The law may excuse the agent on varying grounds, including lack of capacity,Footnote ⁴ lack of free choice or lack of fair opportunity to conform to the requirements of the law,Footnote ⁵ out-of-character behavior,Footnote ⁶ and other such conditions.Footnote ⁷ In discussing the idea of partial justification, we need not commit ourselves to any single conception of excuse. With the aims of generalization and simplicity in mind, we adopt a definition by which an agent is excused if his conduct is wrong, all things considered, but there are reasons why blame is not apt or less blame is apt than would normally be the case. For example, in situations involving coercion one may be excused for killing another in order to alleviate a threat made to one's own life, and similarly one may be excused for somewhat exceeding proportional retaliationFootnote ⁸ in the exercising of self-defense.Footnote ⁹ Commentators often conclude that justifications center around the deed while excuses focus on the doer.Footnote ¹⁰ Perhaps most usefully, justifications are regarded as rules of conduct, providing citizens with ex ante guidance for rightful behavior, whereas excuses are decision rules (or rules of adjudication), directing the court on how to treat the agent fairly ex post.Footnote ¹¹

The distinction between justification and excuse may have great practical significance when allocating legal responsibilities and rights to other participants in the act in question. Justifications typically grant a universal privilege:Footnote ¹² neither the victim nor third parties are allowed to interfere with justified conduct, and whoever aids the justified agent is justified as well.Footnote ¹³ Conversely, excuses are predicated on particular personal attributes or specific states of mind that arise from the circumstances; therefore they are personal and not universal. An excused person should not be aided, and anyone assaulted by him may be justified in defending himself.Footnote ¹⁴

Are justifications and excuses concepts of an either-or type, or do they lend themselves to fragmentation and scaling? Assigning excuses to the realm of blameworthiness, we should have no difficulty in conceding intermediary levels between full excuse and complete blame.Footnote ¹⁵ One can easily envision states of affairs in which the reasons for excusing the agent's wrongful conduct are not unmitigated or are outweighed by other considerations. Justifications, by contrast, seem to be indivisible and by nature binary: conduct is either justified and permissible or it is not.Footnote ¹⁶ Justifications are generally grounded in either maximizing consequentialism or deontological absolutism, or in their derivatives and variants. A consequential calculus would advise us whether the act is good or bad, all things considered.Footnote ¹⁷ Thus there is no meaning to partial justification under consequentialism, apart from saying that within the calculus, some good consequences of the act were weighed against the bad ones before the balance was found to be unfavorable,Footnote ¹⁸ which can be said of most bad deeds and is in the nature of the calculus.

The notion of partial justification hardly fits into deontological theory either. Under the strict form of this theory, even the notion of complete justification is inapplicable: if an act is prohibited by some deontological standard, no circumstance may justify it.Footnote ¹⁹ Nonabsolutist deontology may accept the existence of justification where the act in question is “a good thing, or the right or sensible thing, or a permissible thing to do, either in general or at least in the special circumstances of the occasion.”Footnote ²⁰ Kent Greenawalt explains this position by demonstrating that “the law permits much behaviour that is regarded as morally less preferable than its possible alternatives.”Footnote ²¹

This account of justification incorporates notions of good and right with the alternative notion of permissibility.Footnote ²² It should, however, be obvious that just as action can be either right or not, it can also be either permissible or impermissible. In fact, permissibility may be the more strictly binary concept of the two, because while action may arguably be neutral—neither good nor bad—permissibility is not subject to neutrality. There is no intermediate category between permissible and impermissible, because the rules of prohibition must classify all borderline cases one way or the other.Footnote ²³

Some commentators stress the idea that an evil remains an evil even when it is justified by other circumstances. John Gardner endorses this account of justification. In his words, a justification of wrongdoing is a “cancelling permission” that “does not cancel the reason not to commit the wrong—it only defeats that reason—but does not cancel the reason's mandatariness.”Footnote ²⁴ This account, too, does not challenge the binary character of justification. The difference between viewing a justified action as good and right and viewing it as partly evil simply depends on the point of view, that is, on looking at different segments of the action. Taking a wider angle, the entire act may be seen as good overall; looking more closely at the (lesser) evil committed, it may still be viewed as a bad act that, given other circumstances, should be avoided. What Gardner seems to suggest is that we should not neglect the latter of these viewpoints. Arguably, one should note that although the act is justified from the perspective of the perpetrator and of society as a whole, still there has been an infringement of the rights of the victim of the lesser evil.Footnote ²⁵ But this, again, may rest on the question of what factors we incorporate into our account, because under a wider account it may alternatively be argued that the victim has no right against justifiable infringements, which are part of the everyday dangers associated with social life.Footnote ²⁶

Notwithstanding the strength of the foregoing analysis, one may still wish to find some terminological use for partial justification. In this case, it will not do to charge one with incoherence or unintelligibility. Indeed, language may be stretched to fit the term, and as is attested by the abundance of endeavors to conceptualize it, it apparently holds much intuitive appeal. Law may also find a place for partial justification. After all, a strong theoretical foundation has never been the benchmark of legal doctrine, nor does it have to be. Yet in order to qualify as a sound legal concept, a notion of partial justification must offer more than merely a linguistic possibility; it must further the aims of criminal law and maintain some consistency with the attributes of complete justification as commonly understood in this context. Thus the test of any conceptualization of partial justification in criminal law is not altogether its coherence but rather its usefulness and its consistency with the characteristics of complete justification. In our view, none of the conceptions that have been offered to date for partial justification meets the mark. More specifically, we find that each of these conceptions is predicated on one of eight conceptual errors implicit in the different meanings attached to the idea of partial justification. We identify and discuss these fallacies in the next eight sections of the article.

CONCEPTION 1: PART OF THE CONDUCT WAS JUSTIFIED EQUALS PARTIAL JUSTIFICATION

The term partial justification originated laconically in J.L. Austin's seminal work on excuses,Footnote ²⁷ where it was incidental and benefited from neither illustration nor explanation. Soon after, commentators started referring to Austin's mention of the term without providing a comprehensive definition for it.Footnote ²⁸ The first to recognize the awkwardness of the term partial justification and to attempt to establish a conceptual framework for its use was Kent Greenawalt.Footnote ²⁹ Greenawalt offers two scenarios that serve to demonstrate the viability of partial justification. In the first scenario (the second scenario is discussed in the next section), an agent retaliates with reasonable and proportionate force to an assault, but after subduing the assailant, continues to pound him. Greenawalt claims that because the first phase of the conduct is completely justified, the conduct in its entirety, including the second and forbidden phase, can be described as partially justified. In Greenawalt's words:

Note, however, that by combining an earlier, justified conduct with a later, unjustified one into one action, Greenawalt simply plays with the level of individuation of the conduct in order to reach his conclusion that the action is partially justified.Footnote ³¹ Following this line of logic, one might say that the conduct of a killer who happened to give charity to a street beggar just moments before he shot and killed him (or, somewhat more remotely, before killing a person other than the beggar) is partially justified as a whole. Absent a test for demarcating conduct, it is not difficult to see how this trail of thought leads to a complete analysis of all of the defendant's virtues and vices—as in virtue- or character-based theories of accountability.

Greenawalt may respond by stressing that he does not mean for his analysis to be stretched this far. The two phases of conduct he describes—reasonable and proportionate resistance at first, followed by an unwarranted assault—should be examined together because they are closely related and constitute a single set of events, as opposed to the beggar incident or to a full-fledged virtue analysis. But even so, such an analysis does not produce a useful concept of partial justification. When discussing justification, we should generally strive for a narrow individuation of conduct, separating the different phases of the agent's conduct as much as we can, because of the guidance implicit in categorizing an action as justified (remember justifications’ role as rules of conduct).Footnote ³² Lumping together several stages of conduct and declaring them to be partially justified as a whole usually does not provide the agent with meaningful guidance as to which parts or aspects of his conduct were desirable and which were not. Generally speaking, a narrowly construed conduct analysis conveys a more useful message, focusing on each phase separately and differentiating exactly which phase of the conduct is allowed and which is not.Footnote ³³

Moreover, even if we follow Greenawalt and categorize several phases of conduct as one, it does not mean that the action as a whole is partially justified. In fact, Greenawalt performs an invalid leap from “part of the conduct is justified” to “the conduct as a whole is partially justified.” For even if we choose to take a holistic view of the two-phase retaliation, as Greenawalt suggests, we must conclude that on the whole, using excessive force in warding off an attack is wrongful and disallowed, and therefore unjustified, although perhaps at times excused.Footnote ³⁴

A similarly invalid leap from “part of the conduct is justified” to “the conduct enjoys partial justification” is made by Kenneth Simons in his treatment of victims’ consent.Footnote ³⁵ Simons contemplates a partial justification for an agent who rapes a victim who has agreed to intimate sexual contact short of intercourse: “Arguably the defendant who engages in nonconsensual intercourse should be considered partly justified in his conduct and should receive a lesser punishment.”Footnote ³⁶ Surely the law would send a more useful message if it partitioned the defendant's conduct in order to communicate that his actions up to the intercourse were justified,Footnote ³⁷ whereas the intercourse, which is the offence with which he is charged,Footnote ³⁸ is not, and if mitigation of punishment is warranted, this would not be attributed to partial justification.Footnote ³⁹ If, nevertheless, we are to view the conduct as a whole, it must be characterized as sexual contact that exceeded what the victim consented to and therefore it is wrongful and disallowed or, in other words, unjustified. Under a wide individuation of the conduct in scenarios of the two-phase type, all that remains to be said is that the conduct as a whole was indeed wrongful, although perhaps it was not as wrongful as it could have been under different circumstances, or conceivably there was some justification for it. We examine these types of rationale under conception 2 and conception 3.

The two-phase conception of partial justification is also inconsistent with the base concept of complete justification, as it does not retain the key features that distinguish justification, which are its conduct-allowing function and its universal nature. Viewing both phases as one whole, the agent's allowed act is subsumed under an overall disallowed conduct and loses its distinctiveness. Furthermore, the conceptual difficulty inherent in the leap from accepting that part of the conduct is justified to concluding that it enjoys partial justification extends to scenarios involving victim retaliation and third-party intervention. We see above that justifications are typically universal, meaning that their availability or lack thereof is derivative to victims and third parties.Footnote ⁴⁰ How does this implication hold up in the hypotheticals offered by Greenawalt and Simons? If subdued Ben or the rape victim try to ward off the attack directed at them in the second phase of the event, which is alleged to be partly justified, should the law correspondingly hold them and anyone coming to their aid as partially culpable for the retaliation?

Although it may conceivably be argued that Ben is partly to blameFootnote ⁴¹ for the attack directed at him by Ann (although it would be more useful to say that he is fully blameworthy for the first phase of the attack and not blameworthy at all for the second part), the victim in Simons's example is certainly not to be blamed for being raped and should be viewed as fully justified in retaliating as long as she satisfies the requirement of proportionality. In both illustrations, surely, any third party who intervenes to help the retaliating victim is not partly culpable for so doing merely because the victim initially attacked the person now attacking him (in Ben's case), or has consented to certain sexual contact (in the case of the rape victim). If the third party intervenes in the second phase of these scenarios, any proportional harm caused is fully justified, notwithstanding the fact that part of the victim's conduct was justified. Thus a third party should be fully justified in intervening in aid of subdued Ben or the rape victim, and, correspondingly, the agents in both cases, that is, Ann and the rapist, should not be thought of as partly justified for their conduct in any meaningful way.Footnote ⁴² There is no good reason to make the leap from “part of the conduct is justified” to “the conduct as a whole is partly justified,” because the conduct as a whole is both punishable and nonuniversal.

CONCEPTION 2: REDUCED WRONGFULNESS EQUALS PARTIAL JUSTIFICATION

The term partial justification has alternatively been used to signify an act that was performed in a manner that rendered it “less inappropriate than it would alternatively be” or “not as wrong as it might otherwise have been.”Footnote ⁴³ Greenawalt contemplates a scenario in which Ann hits Ben because Ben mistreated Ann's close friend. If Ann's reason for action was reprimanding Ben for his misbehavior toward her friend, the assault in this scenario is indeed unjustified, but to a lesser degree than it would have been had the assault been performed out of selfish motives.Footnote ⁴⁴ Greenawalt classifies this form of mitigation as partial justification because the reason for which we view Ann as less blameworthy, that is, the reason for the reduced level of wrongfulness of her action, is “the same sort of reason that might fully justify her act; it differs only in that it is less powerful.”Footnote ⁴⁵ Greenawalt is aware that in light of the binary character of justification his rationalization of partial justification creates a conceptual challenge, noting that “the term justification has an either-or quality that makes people hesitant to speak of a partial justification when no aspect of the action is fully justified.”Footnote ⁴⁶ He nevertheless uses this rationalization of the term, noting its theoretical use, for example, in explaining the doctrine of provocation.Footnote ⁴⁷

It seems that Greenawalt downplays the conceptual unsoundness of referring to conduct that is wrongful, although not as wrongful as it conceivably might have been, as being partially justified. It may be argued that except perhaps for the most heinous acts, every act is not as wrongful as it could have been. For example, the reckless causing of harm is not as wrongful as the intentional causing of the same harm. Thus, according to Greenawalt's conception of partial justification, manslaughter can be conceived as “partially justified murder” rather than a less severe conduct than murder on the scale of wrongfulness. Under this account, murder, too, is partially justified homicide provided that it was not carried out with the most objectionable of motives (whatever those may be) or that it did not constitute the most horrific manifestation of that offence.Footnote ⁴⁸ Generally speaking, for all acts, save the most horrendous ones, there exist those that are more wrongful. This does not mean that all acts but few should be labeled relationally as partially justified, and if it does, this voids the term partial justification of any helpful meaning.

Criminal conduct may be graded, and wrongfulness is scalar, but categorizing an offence as partially justified merely because its physical manifestation was not the worst possible or because the mental element could have been more severe is tantamount to equating the current practice of offence grading with an alternative practice of viewing each criminal act as a more or less justified version of the most severe form of committing it.Footnote ⁴⁹ Categorizing forbidden action as partially justified obscures the distinction between allowed and disallowed conduct and runs the risk of communicating the wrong message, for example, that the act of reckless killing is justified at some level because it was not intentional. The false leap here is from “not as bad as Φ” to “partially justified in relation to Φ.”Footnote ⁵⁰

Again,Footnote ⁵¹ the conceptual difficulty extends to scenarios involving retaliation by victims or intervention by third parties. Imagine that Ben proportionally defends himself against Ann's attack. Is he partly culpable for this act because, given Ann's motives in attacking him, her conduct was not as bad as it could have been? True, in this hypothetical situation it is assumed that Ann's reason for acting against Ben had to do with Ben's mistreatment of Ann's friend, and therefore Ben is not entirely without fault. But this is not a necessary part of Greenawalt's theory. Ann's conduct would have likewise been “not as wrong as it might have been” had she harbored good motives for action that had nothing to do with Ben, as for example, if Ben had recently inherited great wealth and Ann attacked him as an act of social protest. In this case, Ben is entirely without fault. But even if we take the illustration at face value, Ben is entirely without fault for defending himself. He is culpable for mistreating Ann's friend, but that does not make him culpable for defending himself against Ann's attack. This point is perhaps best illuminated if we imagine that Ben has already been punished for his mistreatment of Ann's friend.

Current law rightly allows Ben to defend himself against Ann. Just as clearly, the law is right not to regard any third party that intervenes to help Ben as partly blameworthy because Ben had at some point mistreated Ann's friend or indeed because Ann's conduct in attacking Ben was not as bad as it could have been.Footnote ⁵² Concurrently, if Ben is fully justified in retaliating against Ann and any third party is fully justified in helping him, Ann's act is not to be thought of as being partly justified. It may be concluded that the notion of equating partial justification with lesser degrees of wrongfulness retains neither the utility nor the implications of full justification.

CONCEPTION 3: THE CONDUCT ENJOYS SOME JUSTIFICATION EQUALS PARTIAL JUSTIFICATION

At times, a prohibited act blends both bad and good traits and therefore has some justification. May we concede the appropriateness of partial justification for this category of cases? Susan Rozelle bases the provocation defense on partial justification following this type of rationale:Footnote ⁵³

At first glance this account looks like a version of conception 1, leaping from the proposition that “part of the conduct was justified” to “the conduct enjoys partial justification.” But Rozelle's envisioned conduct is not phased into distinct stages, as in that conception. Rozelle's account should not be confused with conception 2 either, although it is closely related to it. Rozelle does not center on the bad or negative aspects that are missing from the agent's conduct, rendering it less wrongful than it could have been otherwise, but rather stresses the good or positive aspects that are present in the agent's conduct. The conduct has something to commend it and therefore it is not all bad, and consequently (this is the leap), it is partially justified. But finding some positive aspects in what is generally a bad deed is not a unique characteristic of provoked killings. Indeed, evil rarely comes in pure form, and in this sense almost all bad acts have something to commend them. As shown above, balancing the evil caused with the evil spared (or good done) in the agent's conduct is in the nature of the justification calculus.Footnote ⁵⁵

Assessing the availability of a justification generally means that some good aspects or consequences of the conduct under scrutiny are weighted against the bad. A favorable balance means that the conduct is justified, that is, it is right or at least permissible. If, however, the balance is found to be unfavorable, the conduct is not justified and should not be performed. Partial justification plays no part in this account. The idea that some justification renders conduct partially justified, much like the idea that reduced wrongfulness leads to partial justification, results in substituting the practice of offence grading according to varying levels of wrongfulness with an array of partial justifications. It is therefore subject to the criticism applicable to the conceptualization of partial justification under conception 2, as it undermines the guiding force inherent in the binary nature of justification without gaining anything that the concept of wrongfulness does not already provide, and without retaining any of the utility or the implications of complete justification.

Another depiction of partial justification that builds on the commendable aspects of the prohibited conduct is offered by Stephen Garvey.Footnote ⁵⁶ Referred to as a theory of “disproportionate response,” this account states that a provoked agent is partly justified for killing the provocateur because the act of provocation merits a punitive response.Footnote ⁵⁷ Accordingly, “the adequately provoked actor is therefore guilty of a wrong, but only because the punishment he inflicted usurped state authority and was disproportionate to the provocateur's crime.”Footnote ⁵⁸ This version of the invalid step (ranging from the statement that the conduct has some merit to accepting the conceptual viability of partial justification) is subject to the same line of criticism as the one directed at Rozelle's version of this conception. Moreover, note the perverse use of counterfactual analysis explicit in this theory as Garvey understands it: “the provocation facing the actor must be adequate inasmuch as the law's own norms, or the norms of some other system incorporated into the law or to which the law defers, would permit the use of non-lethal violence in response.”Footnote ⁵⁹ This means that the agent is partially justified because the state authority (and not necessarily the actor himself) would have been justified had it used lesser force than the agent did.Footnote ⁶⁰ Put in this way, it is clear that this line of reasoning does not lead to a useful concept of partial justification.Footnote ⁶¹

CONCEPTION 4: JUSTIFIED EMOTIONS PARTIALLY JUSTIFY THE CONDUCT BASED ON THOSE EMOTIONS

Garvey proposes a further theory of partial justification that he names “worthy motives” and explains in the context of the provocation defense:

This account may be viewed as a specific case of conception 3, in which the conduct is commended by the agent's motive. In this case, the theory is subject to the type of criticism applicable to Garvey's “disproportionate response” theory of partial justification.Footnote ⁶³ Alternatively, the theory may be coined in the terms relating to conception 2, with the worthy motive making the conduct less wrongful in the sense that the agent is less culpable for this conduct than he would have been otherwise. If so, what justifies the leap from the judgment that the agent is “less culpable” because of his motive to the conclusion that he is partially justified?Footnote ⁶⁴ Can worthy motives support a useful concept of partial justification? As under conception 2, it is doubtful that anything is to be gained by the worthy-motive theory beyond what the concept of wrongfulness already affords. The provoked agent lacks the evil motive of the hateful or otherwise badly motivated killer, but this does not partially justify him; he is simply culpable of the lesser offence of manslaughter, or subjected to a lesser sentence, or both.Footnote ⁶⁵

But is there something more to the worthy-motives theory? Recently, Christopher Bennett offered a sophisticated portrayal of partial justification based on emotions.Footnote ⁶⁶ For Bennett, too, emotions can play a partly justifying, not merely excusing, role. He states that “the reasons that count in favor of our emotion count also in favour of action done from that emotion,”Footnote ⁶⁷ and therefore that action is partly justified. Bennett illustrates his position using hypothetical situations in which the harmful act was brought about by fear (lashing out at someone mistakenly perceived by the agent as an attacker) or by excitement (hugging a nearby stranger upon receipt of good news over the phone).Footnote ⁶⁸ In both instances Bennett counts the motive guiding the actor as favorable and therefore as partly justifying the act.Footnote ⁶⁹

Bennett distinguishes nonrational emotions, such as phobias, from rational emotions, such as fear (when there is something to be frightened of) or anger (when there is something to be angry about). The former excuse, whereas the latter (partly) justify.Footnote ⁷⁰ But many of the bad things we do have rational reasons, perhaps more often than not. A robber may have a craving for wealth, which under Bennett's taxonomy can only be characterized as a rational emotion. Bennett alludes only to “considerations that would surely count, at least partially, in favour of an action,”Footnote ⁷¹ but entrepreneurship may count as such, and the entrepreneurial thief is not wholly or partly justified or even excused to any extent.

In other instances, possibly including Bennett's illustrative examples, the rational emotion can serve as a form of partial excuse but not as a justification. We do not always have the capacity to do the right thing in the eyes of the law, because of irrational as well as rational reasons. This is what is at the heart of the concept of excuse. Contending that a good motive or a rational emotion partly justifies wrongful conduct is to say in effect that if the conduct was less severe, or if the reasons for it were more pertinent, or both, the action would have been justified. But as it stands, that action is not justified, even if it could have been worse. Perhaps it is not as wrongful as other potential types of action, such as an unwarranted attack, because it was triggered by a good motive or a rational emotion. We understand the agent's reason for action and are ready to give him some leeway on its account. But this is the nature of partial excuse, not of partial justification.

Although it may be tempting to follow the apparent symmetry invoked when claiming that a justified emotion translates into a partially justified action based on that emotion, it is not even clear that emotions are at all “justified” and not simply “fitting,” or relevant, or a common reaction to the situation. Based on such an account, all symmetry is lost. And just as important, when we describe emotions as fitting or common, we immediately find that such emotions have nothing magically justifying about them. They are simply understandable human responses that may (partly) excuse the conduct they engender. If under certain circumstances it is important for the law to convey that the agent's wrongful conduct derived from a particularly understandable type of emotion, the law can fully excuse the agent. Arguably, this would achieve a greater expressive effect than “partially justifying” the conduct.

Furthermore, it is difficult to see how any reason for having an emotion can justify an action. Actions are twice removed from the reasons or stimuli that produce them (reason-emotion-action), and whereas the connection between the reason and the emotion is a strong one (although, as explained above, it is more a matter of a relevant or fitting emotion than a justified one), the link between the emotion and the action does not have to be so. There is a multitude of possible reactions to every situation and the emotions it may stir. Some are justified, others are not, and of the latter, some may be excusable. For example, when one is being provoked, anger is an understandable reactive emotion (justified in Bennett's terms), but one may act on that emotion in several ways. One may struggle with one's anger and ignore the provocateur, perhaps graciously offering him or her the other cheek; one may retaliate in anger with some provocative behavior of one's own; or one may lash out in an attack of varying degree. The last reaction is not justified by the emotion, reasonable as that emotion may be.Footnote ⁷² This should be taken as part of the reason why criminal law sanctions action and not emotions as such. Perhaps the link between the emotion and the ensuing action is not as obvious or automatic or uncontrollable as that between situations and emotions (although at least some of these can be mastered and self-regulated as well).

Revisiting Bennett's hypothetical, embracing a nearby stranger following the excitement caused by receiving a piece of good news is only one possible course of action. It may not be the worst kind (one may conceivably fire a handgun in excitement), but not acceptable nevertheless. And if the person being embraced is repulsed or simply unmoved by the agent's excited elation, and feels wholly, not only partly, harmed by the invasion, we must honor that feeling. This means that the agent has only an excuse vis-à-vis society, whereas a justification would apply generally (universally) and is expected to be recognized by the victim.Footnote ⁷³ To retain a part of the universal attribute of justification consistently when viewing the excited embrace as partially justified, the victim who reacts to the embrace by pushing off the assailant must be marked partially culpable for so doing—to correspond to the assailant's partial justification. If the notion of the victim being partially wrongful and therefore blameworthy in this illustration is normatively unsound, as we think it is, it follows that the assailant should not be deemed partially justified. It would be more practical to demand of the victim to proportionate her response by taking into account the harmfulness of the assailant's action, not his reasonable-under-the-circumstances emotions.

This type of emotion is exactly the mark of excuse and not justification, and it is this kind of practical implication that should be decisive in criminal law. Therefore the illustration shows the weakness of the idea that justified emotions partly justify actions based on these emotions within the context of criminal law.

CONCEPTION 5: COMPLETE JUSTIFICATIONS DO NOT NEGATE WRONGFULNESS, THEREFORE JUSTIFICATION IS A SCALAR NOTION, AND PARTIAL JUSTIFICATIONS EXIST

In his thought-provoking article on partial defenses, Douglas Husak contends that the key to understanding partial defenses is in finding their analogues in full defenses.Footnote ⁷⁴ We scrutinize this theory in the next section of the article, under conception 6. Here we focus on one step in Husak's argumentation, in which he sets out to establish the existence of partial justifications. Husak observes that full justifications do not completely purge the act of all harmfulness but simply diminish wrongfulness and blame to a sufficient degree to preclude liability.Footnote ⁷⁵ From this statement Husak concludes that justifications are relative by nature, and therefore partial justifications exist. Husak defends this concept of partial justification by contrasting two illustrative cases:

It is doubtful whether the victim's intent to steal a car fully justifies the act of breaking his leg, but we assume for the sake of argument that it does. We can easily envisage less arguable cases in which the agent's response is proportional to the victim's threat. The more fundamental difficulty in Husak's conclusion that Smith's conduct is at the same time both wrongful and justified is that the analysis suffers from an inconsistent individuation of the act being analyzed: it is broad when discussing wrongfulness (breaking someone’s leg is wrongful) and narrow when discussing justification (breaking someone’s leg in response to his threat is justified). Consistent individuation of the act yields different results. Under a narrow individuation, we view the act as a proper response to the victim's threat, and therefore it is both not wrongful and justified. Under a wide individuation, we focus on the agent's aggressive act independent of its reasons or background and therefore view it as wrongful and, by excluding the justificatory explanation of the act, also as unjustified.

Perhaps Husak deliberately shifts the level of individuation in order to illustrate his point, that is, that an act may be justified despite the fact that some other act may be even more justified, which proves that justification is a scalar notion. We agree that Husak is successful in proving the existence of a scale, but it is not a scale of justification but rather of wrongfulness or of its converse, righteousness. Our criticism here proceeds from our analysis of conception 2 (reduced wrongfulness equals partial justification). Just as for virtually every wrongful act, excluding the most heinous ones, there exist acts that are potentially even more wrongful, so for virtually every deserving act, excluding the most righteous, there may exist even worthier ones. This in itself does not prove that justification is scalar. If, all things considered, the act is justified overall, the fact that the agent could have chosen some even more favorable line of action does not make his act less justified. For beyond justification there lie not higher degrees of justification but rather mercy and clemency or perhaps compassion and charity. A better conceptualization of Husak's illustrative cases would thus hold that Smith is justified, whereas Black is both justified and merciful.Footnote ⁷⁷

In sum, the fact that a justified act is not the best possible course of action does not necessarily challenge the binary nature of justification; just as actions may be more or less harmful or wrongful, so they may be more or less merciful or charitable—but they may still be either justified or unjustified.

CONCEPTION 6: PARTIAL DEFENSES THAT HAVE AN ANALOGUE IN A COMPLETE JUSTIFICATION ARE PARTIAL JUSTIFICATIONS

Following his argument that justification is a scalar concept,Footnote ⁷⁸ Husak introduces his description of partial justification as a component of what he terms a “unifying hypothesis” of mitigating factors. Under this hypothesis, a circumstance qualifies as either partial justification or partial excuse because it has a corresponding analogue in a complete justification or a complete excuse.Footnote ⁷⁹ Partial justifications justify partially because they have counterparts in complete justifications, which justify completely. Partial excuses, in turn, excuse partially because they have counterparts in complete excuses.

This sophisticated argument for partial justification relies explicitly on counterfactual analysis, that is, on asking what would have been the case had the circumstances faced by the agent had been more serious or had the agent's response been less harmful (or a combination of both). In Husak's words, “a given circumstance provides a partial justification or excuse whenever it would suffice to preclude liability altogether were it present to a greater degree.”Footnote ⁸⁰ This means that had the circumstances been otherwise, a full justification would have been available, which is undoubtedly true. But Husak continues to infer from this fact that as things stand, a partial justification may be afforded to the agent. In no other area of culpability are counterfactuals used in this manner or play such a decisive role. Criminal law resorts to counterfactual analysis in determining the causal connection between the agent's conduct and the ensuing result, but even here it is relevant to the factual, not the normative aspect of causation. No good reason is offered for regarding counterfactual analysis as governing the normative question of justification. Husak is not blind to this idiosyncratic element in his theory, but as far as he is concerned, the problem lies not in the use of counterfactual analysis per se but in the use of what he regards to be the wrong basis for comparison.

“A partial justification reduces the wrongfulness of an act,”Footnote ⁸¹ writes Husak, invoking conception 2 (less wrongful equals partially justified). Husak is, however, doubtful and asks, “relative to what baseline is the wrongful act or the blame of the agent reduced?”Footnote ⁸² He answers:

Husak rejects counterfactuals that allude to the way things might have been but is happy to rely on those that refer to a standard or paradigmatic offence. But this is still to claim that an act is partially justified because it could have been worse, only this time in comparison not with what the offender might have done but with what others involved in similar conduct usually do. This variation of the counterfactual examination still falls under the “less wrongful equals partially justified” fallacy.

Furthermore, it is open to debate which the preferred baseline is for applying the counterfactual analysis: that which the agent could have done, or the paradigmatic commission of the offence. The latter seems to make more sense, as Husak rightly points out. But imagine a case in which the agent, because of certain external constraints, did not have the opportunity to act as badly as in the paradigmatic commission of the same offence. If that agent performs the act in a manner that was as wrongful as opportunity allowed him to perform it, should his action be judged relative to the paradigm or to what he could have done? Compared to which baseline is he partly justified? Husak does not discuss this possibility, but we believe that it points to the conclusion that the problem is inherent in counterfactual analysis, whatever baseline is used. The reliance on counterfactual analysis in turn points out that the aspect of the unifying hypothesis here discussed is a sophisticated variant of conception 2.

An apparent shortfall of the unifying hypothesis is the difficulty in defending criteria to decide when a given circumstance finds its analogue in a complete justification.Footnote ⁸⁴ But the unifying hypothesis faces more fundamental difficulty because it is predicated on the assumption that incomplete justifications logically generate partial justifications. This does not have to be the case, however, and often incomplete justifications produce excuses. The explicit position of the German Criminal Code, for example, differentiates necessity as justification from necessity as excuse, the former applicable to the use of proportionate means to avert danger, the latter to necessary means that exceed proportionate response.Footnote ⁸⁵ Thus disproportionate reaction to necessity is neither desirable nor allowed and is therefore incomplete as a justification but may nevertheless be fully or partially excused. Put in general terms, the full defense may be a justification, but its partial derivatives may stem from excuses. This is a bug in the unifying hypothesis, and Husak ignores it when he writes that:

The shift from justification to (partial) excuse is not mystical or mysterious but rather prosaic and common. It is the shift from the right and the permissible to the understood and the conceded. A proportionate use of a defense is justified. Slight deviations from proportionate defense may be understandable and therefore generate an excuse.Footnote ⁸⁷ Any greater deviation from proportionality is derived from that excuse. Therefore, partial justifications are not the necessary or even the proper outcome of incomplete justifications.

CONCEPTION 7: UNREASONABLE BELIEF IN THE EXISTENCE OF A JUSTIFICATION EQUALS PARTIAL JUSTIFICATION

For Re’em Segev, “justification—and accordingly wrongfulness—is a matter of degree and therefore an action that is justified in the strict sense is merely the peak of a continuum.”Footnote ⁸⁸ It should be clear by now that this account, akin to conception 2 (reduced wrongfulness equals partial justification), mistakenly equates the concepts of justification and wrongfulness. In discussing conception 2, we show that the leap from viewing wrongness as a scalar concept to holding that a graded notion of justification is a useful and consistent concept within the context of criminal law is invalid. Wrongfulness is scalar, while justification serves as a threshold beyond which the act is good or allowed. Until that point, the act may be forbidden with differing degrees of wrongfulness, but not of justification.Footnote ⁸⁹ But Segev continues to offer an account of partial justification according to which one was partially justifiedFootnote ⁹⁰ if one formed an honest though unreasonable belief concerning the availability of justifying circumstances, that is, if one believed that one acted with justification but did not collect enough information or did not put enough effort into learning otherwise. Thus one may be fully justified if one made all the reasonable inquiries and reached a rational conclusion that one's harmful act was proportional with the threat it was designed to counter, but only partially justified if one reached the same conclusion following a flawed or partial inquiry into the facts of the situation.Footnote ⁹¹

Is this truly an account of partial justification? It seems to us that it serves not to explain the degree of justification but rather to elucidate the separate question of determining the threshold of justification. Segev reminds us that the purely objective and the purely subjective standards for determining the validity of a justification are not exhaustive, and that justifications may also be determined “rationally,” that is, based not on objective reality or strictly on the agent's subjective understanding of it but on the circumstances the agent could have reasonably known. This introduces a standard of negligence or perhaps of willful blindness for assessing the validity of justification, and Segev gives a sophisticated account of this standard in his article, but his argument does not support the conclusion that justification is a matter of degree, only that different standards may apply in determining justification.

A lenient examination of the method used by the agent to collect the information on which she bases her evaluation of the situation she faces constitutes a move toward the subjective standard for determining justification. Alternatively, we can choose an objective standard and still concede the agent's unreasonable belief and fully or partly excuse an agent who acted under such a belief.Footnote ⁹² Segev's account takes an intermediate, third course, according to which the threshold of justification lies in making a rational assessment of the validity of the justificatory circumstances. In sum, the objective, rational, and subjective standards of justification do not form a continuum of justification. Rather, the difference between them is categorical: they form distinct standards by which to assess the threshold of justification. Under the rational standard, an actor who acts out of unreasonable belief in the availability of justification is not justified, although her act may be less wrongful than it would have been had she acted out of base motives.

Alternatively, Segev's unreasonable-belief account of partial justification may be interpreted not as forming an intermediate standard for assessing justification but as a strictly subjective standard, as it is focused on the agent's state of mind. Under this interpretation, action based on an honest but unreasonable belief in the availability of a justification is fully justified, but the agent is separately sanctioned for omitting to follow the full course of the rational decision-making process in the given situation. This interpretation of Segev's view is especially plausible if we consider that Segev seems not to be interested in the factual existence of the justifying circumstances, which suggests that according to his theory, an agent who was not mistaken in his choice to act would nevertheless not be fully justified if he did not follow the rational process to its end, and therefore he would be sanctioned purely for failing to follow that process.Footnote ⁹³ Under this mode of interpretation too, justification is not shown to be a scalar concept.

CONCEPTION 8: OBJECTIVE REQUIREMENTS IN PARTIAL DEFENSES BRAND THEM PARTIAL JUSTIFICATIONS

Under the common law, as well as in many present-day jurisdictions, a successful claim of provocation must rest on two cumulative elements: the homicide is committed in the heat of passion; the heat of passion was the result of adequate provocation. This dual requirement, and especially the objective condition of adequate provocation, leads some commentators to view provocation either as affording a partial justificationFootnote ⁹⁴ or, more commonly, as amalgamating partial justification and partial excuse. These commentators contend that whereas heat of passion is a subjective condition and may serve to partially excuse the agent who has experienced it, the requirement for an adequate provocation dictates an objective examination of the agent's conduct through the prism of the reasonable person, suggestive of a partial justification of this conduct.Footnote ⁹⁵

But none of the commentators adequately explains precisely how the victim's adequately provocative conduct justifies the agent's lethal response. In “A Plea for Excuses,” Austin asks what we mean when we plead provocation: Do we mean that “he having done me such injury, I was entitled to retaliate (justification)”? But under the law, the provoked is not entitled to retaliate and would be subject to criminal liability and grave punishment if he did. This “plea for justification” illustrates the damage that the use of the term partial justification does to the conduct-guiding function of the criminal law.Footnote ⁹⁶ Andrew Ashworth maintains that “an individual is to some extent morally justified in making a punitive return against someone who intentionally causes him serious offence.”Footnote ⁹⁷ He notes that “ordinary language reflects this approach, with characteristic phrases such as ‘he brought it on himself,’ ‘she asked for it’ and ‘it served him right.’”Footnote ⁹⁸ Thus Ashworth anchors the objective test on the victim's culpability.

Similarly to other elucidations of partial justification that we encounter in the previous sections, this retaliatory or punitive version of provocation (partly justifying the act as punishment for the victim's harmful and often unlawful conduct) is also based on an odd counterfactual analysis. In this case, even if the retaliation is proportionate and appropriate, which, by the very definition of the provocation doctrine, it is not (proportionate and appropriate retaliation falls under self-defense, not provocation), its partial justification rests on the logic that were the agent a punishing authority implementing a sentence that was legitimately meted to the victim, she would have been justified in so doing. But one cannot accept the justifiability, even if partial, of such vigilante justice.

Another explanation offered for the victim's adequately provocative conduct justifying the agent's lethal response is that by acting provocatively, the provocateur has reduced his right not to be harmed, meaning that he has partially forfeited his right to life.Footnote ⁹⁹ This is a questionable theory as far as respect for the value of human life is concerned, and its dubious character is amplified in partial defenses. Is the life of the victim partly forfeited because he derogated the agent or had adulterous relations with his spouse (both of which may serve as basis for a successful claim of provocation)?Footnote ¹⁰⁰

The particular shortfall of conception 8 is the leap from viewing justification as an objective defense to concluding that any objective element of a defense is necessarily justifying. The objective standard in provocation, the requirement that the provocation be adequate, need not be understood as meeting the demands of a justification; it may alternatively serve to curtail the partial excuse afforded by the provocation doctrine, making it unavailable to agents who are prone to lose their self-control. The objective aspect of provocation thus sets a standard of self-restraint: only diminished control by those who have enough regard for the law to brace themselves is generally accepted, and not the rage of those who do not possess enough respect for other people's rights of self-expression and act murderously in face of behavior or speech that is not to their liking. This analysis places the objective standard alongside the subjective one within the realm of excuse.

This particular view of provocation is endorsed by numerous commentators.Footnote ¹⁰¹ Joshua Dressler maintains that “under excuse theory we do not (fully) blame a person who (partially) loses his self-control if, but only if, he is not to blame for his anger and for his homicidal actions which result from it.” Dressler exemplifies this by showing that voluntary intoxication is not an excuse even if it affects the agent's power of self-control.Footnote ¹⁰² Referring to excusing conditions in general, Jeremy Horder advocates integrating objective factors into the analysis of excuses, stating that “objective standards, in one form or another, are best understood as part of the necessary conditions, rather than as part of the sufficient conditions, for excuse.”Footnote ¹⁰³ Even Ashworth, the originator of the idea of provocation as partial justification, finally agrees that:

Thus an objective element of a partial defense serves as a normative qualifier of that defense understood as a partial excuse;Footnote ¹⁰⁵ it does not transform the defense into a partial justification.Footnote ¹⁰⁶

Can this logic also go the other way? One may argue that just as we can integrate the objective requirement of provocation into an account of that defense as a partial excuse, it is also possible to integrate the subjective requirement (heat of passion) into the analysis of justifying factors, thus viewing provocation as a partial justification, notwithstanding its subjective aspect. This may be achieved by maintaining that a justified emotion, such as heat of passion, partially justifies the act of killing. Note, however, that this argument falls under the fallacy of conception 4, predicating the notion of partial justification on the “justified” emotion that has triggered the action.

We see above that there is no convincing explanation for how acts such as uttering insults or perpetrating adultery can partly justify killing; it is more convincingly argued that the adequacy requirement in provocation may be subsumed into an analysis of excuse, as a qualification for the availability of excuse. Thus the partial defense of provocation is better understood not as a partial justification but as a partial excuse. But to complete the analysis, we must note further that as it stands, provocation may not afford a partial defense of any kind at all, but rather a complete defense applicable to the offence of murder. After all, a successful plea of provocation completely exculpates the agent from liability for that offence.Footnote ¹⁰⁷ Moreover, provocation may not constitute a defense at all, partial or otherwise. In its common-law origins, provocation was not treated as a defense but as a rebuttal of the presumption of malice aforethought, which was the required mens rea for murder. If the presumption was lifted, the level of mens rea proven could amount only to that of manslaughter. Understood in this way, provocation simply demarcates these two levels of homicide on a scale of wrongfulness.Footnote ¹⁰⁸