Justification and Motivation

Abstract

According to the motivational thesis (MT), we are justified in performing an action if and only if we perform that action for the right reason(s). Proponents of MT disagree about how it is best interpreted—about what count as reasons of the right kind. In Fundamentals of Criminal Law, Andrew Simester criticises an interpretation offered by John Gardner. Here, I explore some of Simester's reasons for objecting to that interpretation, and I argue—partly on the basis of those same reasons—that Simester's own interpretation of MT should be revised. I conclude with a preliminary defence of an alternative interpretation, which I call the tripartite view of justification.

1 Introduction

Some things we do are justified. Others are unjustified. A theory of justification tells us which are which.¹

Here are two questions any such theory must address²:

The Grounding Question—What are the conditions under which P is justified in φing?

The Explanatory Question—Why are the conditions of justification as they are?

One popular answer to the grounding question is:

The Motivational Thesis (MT)—P is justified in φing if and only if P φs for a reason of the right kind.

What counts as the right kind of reason? Proponents of MT disagree among themselves. They disagree, that is, about how MT is best interpreted. Prominent interpretations are offered by John Gardner and Andrew Simester. In Fundamentals of Criminal Law, Simester argues that—given an answer to the explanatory question that is common ground between them—Gardner’s interpretation of MT requires revision. Here, I argue—partly on the basis of that same answer—that Simester’s interpretation of MT should also be revised.

Sections 2 and 3 clear some ground. Section 4 offers the argument just advertised. Sections 5 and 6 defend a further interpretation of MT. They make a preliminary case for what I will call:

The Tripartite View—P is justified in φing if and only if (i) P φs for r; (ii) r is an undefeated normative reason for P to φ; and (iii) P regards r as undefeated on reasonable grounds.³

2 Concepts

One clarification before we proceed. Claims about the conditions of justification are ambiguous. Read one way, they are claims about the conditions under which conduct is classified as justified by legal authorities. Read another way, they are claims about standards to which the classification is subject. Some of these standards are moral. Others are conceptual. The classification is subject to moral standards because the classificatory decisions of legal authorities have morally significant consequences. It is subject to conceptual standards because the idea of justification is not the creation of those authorities. It is an idea that figures in our evaluative thinking generally. One way to evaluate the conditions under which courts classify conduct as justified is to ask whether those conditions diverge from the ordinary idea of justification. The significance of any such divergence need not be purely conceptual. There may be moral standards that favour convergence between the legal classification and the ordinary idea. I mention one such standard in passing below. Whether convergence is morally favourable depends, of course, both on the content of morality and on the content of the ordinary idea. The task of clarifying the conditions of justification, as I approach it here, is the task of elucidating that idea.⁴

3 Reasons of the Right Kind

It will help to begin with a familiar distinction within the theory of reasons. Normative reasons for action are reasons that apply to us. They are facts that count in favour of the acts for which they are reasons. Motivating reasons are reasons for which we act. They are our ends in acting as we do, and our means to those ends. Normative reasons need not be motivating. Motivating reasons need not be normative. The two correspond when P’s reason for φing is a fact that counts in favour of P φing. They diverge when this is not the case.

The motivational thesis is a thesis about motivating reasons. The conditions of justification it contains are satisfied when those reasons are of the right kind. What, then, counts as the right kind of reason? This section explores some answers to this question. The next puts those answers to the test.

A—Weight

According to

The Modest Thesis—reasons to φ of the right kind are undefeated reasons.

We can compare this with

The Strong Thesis—reasons to φ of the right kind are sufficient reasons.

Both undefeated and sufficient reasons to φ are unexcluded reasons. They are, that is, reasons for which it is permissible to φ. The two are distinguished by their comparative weight. Undefeated reasons to φ are unexcluded reasons that are together no less weighty than the reasons not to φ.⁵ Sufficient reasons to φ are unexcluded reasons each of which is at least as weighty as the reasons that disfavour φing. It follows that, while all sufficient reasons are undefeated, not every undefeated reason is sufficient. A reason to φ is insufficient if it is itself less weighty than the reasons not to φ. It is nonetheless undefeated if it is part of a set of unexcluded reasons that is at least as weighty as the countervailing reasons.

B—Normativity

John Gardner endorses the modest thesis.⁶ Andrew Simester defends the strong thesis.⁷ Their debate concerns the weight of reasons of the right kind. A second debate concerns their normativity. It concerns the relationship that must obtain between P’s motivating reasons for φing, and the normative reasons for P to φ, if the former are to confer a justification upon P for φing.

On one view, the requisite relationship is one of correspondence. According to:

The Correspondence Thesis—P’s reason for φing is of the right kind if and only if P has a corresponding normative reason to φ of the requisite weight.

We can compare this with:

The Reasonable Grounds Thesis—P’s reason for φing is of the right kind if and only if (i) P regards that reason as a normative reason to φ of the requisite weight, and (ii) P’s grounds for so regarding it are reasonable.

Assume for a moment that reasons of the requisite weight are undefeated reasons. For P to regard some fact—f—as an undefeated normative reason to φ is for P to make three judgements. First: that f obtains. Second: that f counts in favour of P φing. Third: that f is a reason for P to φ that is neither excluded nor outweighed.⁸ The grounds on which P regards f as undefeated—as I use the term here—are the considerations on the basis of which P makes each judgement. Those grounds are reasonable—as I use the term reasonable—when P lives up to expectations in making these judgements on the basis on which she makes them.⁹

There are cases in which the reasonable grounds thesis is more permissive than the correspondence thesis. And there are cases in which the reverse is true. In cases of the former kind, P’s reason for φing is not an undefeated (or sufficient) normative reason to φ; P nonetheless regards it as such on reasonable grounds. In cases of the latter kind, P φs for a reason that proves to be undefeated (or sufficient); but P’s grounds for regarding it as such are nonetheless unreasonable.

An example of the first kind is

Operation 1—P performs an operation on her patient Q. P’s reason for operating is that Q will otherwise soon die. P takes the steps she ought to take, in her role as Q’s doctor, to determine both that (i) the operation is the medically recommended course of action, and that (ii) P has Q’s consent. P regards her reason for operating as an undefeated reason because she has adequate evidence of (i) and (ii). In truth, however, Q does not consent.

In Operation 1, P’s reason for operating is a defeated reason. Since Q does not consent to the operation, prolonging Q’s life is an excluded reason for P to operate. P’s grounds for regarding that reason as undefeated are nonetheless reasonable. P does not fall short in regarding her reason for acting as unexcluded—indeed as undefeated—on the basis on which she regards it as such. It follows that, though the correspondence thesis is not satisfied, the reasonable grounds thesis is satisfied.¹⁰

We can compare Operation 1 with

Operation 2—As in Operation 1, except P regards the fact that the operation will prolong Q’s life as a sufficient reason to operate regardless of whether Q gives consent. As a result, P fails to take the steps she ought to take to ensure that Q consents. Unknown to P, Q does consent, and the operation succeeds in prolonging Q’s life.

Prolonging Q’s life, we can assume, is a normative reason to operate that is at least as weighty as the reasons not to do so. Q’s consent makes it the case that it is also unexcluded. It follows that, in Operation 2, P operates for an undefeated (indeed sufficient) reason.¹¹ What does not follow is that P regards that reason as undefeated on reasonable grounds. As Q’s doctor, P is expected to base any judgement that she has undefeated reasons to operate on adequate evidence that Q consents to the operation. Since P regards Q’s consent as an irrelevance—and fails to do what she should to determine whether Q consents—P falls short of this expectation. It follows that, though P acts for (what is in fact) an undefeated reason, P’s grounds for regarding that reason as undefeated are not reasonable. Though the correspondence thesis is satisfied, the reasonable grounds thesis is not.

4 Culpability

These distinctions give us four interpretations of the motivational thesis. We can combine the reasonable grounds thesis with the strong or modest thesis. We can do the same for the correspondence thesis.

In Offences and Defences, Gardner claims that:

(A) P is justified in φing if and only if P φs for r and r is an undefeated normative reason for P to φ.¹²

This is to combine the modest thesis with the correspondence thesis. In Fundamentals of Criminal Law, Simester argues that (A) requires revision. We should prefer:

(B) P is justified in φing if and only if P φs for r and r is a sufficient normative reason for P to φ.¹³

Since sufficient reasons are undefeated reasons, Simester’s objection is not that (A)’s necessary condition is too stringent.¹⁴ His objection is that (A)’s sufficient condition is too lax. The problem is solved, according to Simester, by rejecting the modest thesis and combining the correspondence thesis with the strong thesis.

Is (A) too lax? That depends on our answer to the explanatory question. It depends, that is, on why the conditions of justification are as they are. A familiar move here is to appeal to claims about what justification negates. Here are two:

Permissibility—If P is justified in φing, it is not impermissible for P to φ.

Exculpation—If P is justified in φing, P is not culpable for φing.

More on Permissibility later. My focus in this section is on Exculpation. It is a thesis both Gardner and Simester accept.¹⁵ The conditions of justification are as they are, both claim, partly because justification and culpability are mutually exclusive.¹⁶ This is one reason why we should reject:

(C) P is justified in φing if P has an undefeated normative reason to φ.

If (C) is true, MT is false. What matters for justification are the normative reasons one has, not the reasons for which one acts. Exculpation, however, implies that we should accept (C) only if we should accept:

(D) P is not culpable for φing if P has an undefeated normative reason to φ.

The argument against (C) is that, since (D) is false, while Exculpation is true, it is (C)—not MT—that must be rejected.¹⁷

One question about this argument is whether we should accept Exculpation itself. The answer depends, of course, on whose acceptance is at issue. For criminal lawyers, the obvious candidate is the criminal courts. Let us assume that it is a criminal offence to φ, and that the offence was legitimately enacted. Now consider:

The Culpability Principle—If P is culpable for φing, P ought to be held criminally liable for having φed.¹⁸

A criminal court which holds P to be justified in φing thereby holds that P is not criminally liable for φing. It follows that, if the culpability principle ought to be endorsed by the criminal courts, those courts ought to deny that P is justified in φing if P is culpable for having φed.

My aim here is not to argue for Exculpation. I assume that it forms part of our ordinary idea of justification. My interest here is in the implications of this assumption for the conditions of justification. One implication is that, if (D) is false, so is (C). Whether (D) is false depends, of course, on the conditions of culpability. To pick as few fights as possible, let us confine ourselves to two:

Insufficient Concern—P is culpable for φing if P φing manifests insufficient concern for those affected.

Responsiveness—P φing manifests insufficient concern for those affected if P’s practical reasoning is insufficiently responsive to the interests of those φing affects.¹⁹

Now consider:

Defender 1—P kills V to eliminate the competition for a lucrative promotion. Unknown to P, killing V saves the lives of many innocent people.

Part of what it is to be sufficiently responsive to the value of another’s life, I take it, is to treat improving one’s job prospects as an insufficient reason to kill them. Since P treats improving her own prospects as a sufficient reason to kill V, Insufficient Concern and Responsiveness entail that P is culpable for the killing. P’s actions do, of course, also save innocent lives. Given our two conditions, however, this is not a fact that bears on whether P is culpable. P’s practical reasoning is in no way responsive to the interests of those she saves. So the fact that P’s actions turn out to protect those interests does nothing to show that her actions manifest sufficient concern. What it does show, at least when the number of lives saved is large enough, is that P has an undefeated normative reason to kill V. If she does, (D) is indeed false. (C) must therefore be rejected.²⁰

Recall next that according to:

(A) P is justified in φing if and only if P φs for r and r is an undefeated normative reason for P to φ.

We should accept (A) only if we should accept:

(E) P is not culpable for φing if P φs for r and r is an undefeated normative reason for P to φ.

As Simester shows, however, (E) is false. Consider:

Defender 2—V attacks P, intending to push P aside and kill E, who is standing behind P. P is unaware of the threat to E. To prevent herself being pushed, P responds with lethal force.²¹

P has two normative reasons to use lethal force: (1) to prevent herself being pushed; (2) to save E’s life. She also has a normative reason (3) not to use lethal force against V. Since (1) is less weighty than (3), (1) is an insufficient reason for P to use the force she uses. But since (1) and (2) together are at least as weighty as (3)—and since reasons to use defensive force are unexcluded reasons—both (1) and (2) are undefeated reasons. It follows that, when P uses lethal force against V, P does so for an undefeated (albeit insufficient) reason. (A) therefore confers a justification upon P for killing V.

Is P culpable for the killing? This might be denied. The denial is not plausible, however, if (C) should be rejected on the grounds on which we rejected it earlier. What goes for improving one’s job prospects goes for preventing oneself being pushed: to treat either as a sufficient reason to kill is to be insufficiently responsive to the value of human life. True, P’s act of killing V also saves E. We already know, however, that this is irrelevant here. Since P is unaware that E is behind her, E’s prospects play no role in P’s practical reasoning. So they do nothing to displace the conclusion that P is culpable for killing V. If P is culpable, (E) is false. And if (E) is false, Simester’s argument is successful. (A)’s sufficient condition is indeed too lax.

Simester claims that we should replace (A) with:

(B) P is justified in φing if and only if P φs for r and r is a sufficient normative reason for P to φ.

In Defender 2, P kills V for an insufficient reason. (B) therefore denies P a justification for the killing. We should nonetheless reject (B). We should do so because (B) is compatible with Exculpation only if:

(F) P is not culpable for φing if P φs for r and r is a sufficient normative reason for P to φ.

Yet (F) should also be rejected. We can see why by returning to Operation 2. There, recall, P operates on Q for a sufficient reason. She has Q’s consent, and successfully prolongs Q’s life. Our two conditions of culpability nonetheless entail that P acts culpably. P manifests sufficient concern for Q, when deciding whether to operate, only if P’s practical reasoning is sufficiently responsive to the value of Q’s autonomy.²² P’s practical reasoning is sufficiently responsive to the value of Q’s autonomy only if P treats Q’s consent as determinative of whether she may proceed, and only if P takes reasonable steps to ensure that Q consents before proceeding. In Operation 2, neither is the case. P regards Q’s consent as an irrelevance. She fails to do what she should to determine whether Q consents. It follows, given Insufficient Concern and Responsiveness, that P is culpable for operating on Q.²³ If P is culpable, (F) is false. And if (F) is false, (B) too requires revision.

5 The Tripartite View

Here is where we are. Simester argues—against Gardner—that (A) is too lax. I have argued—against Simester—that the same is true of (B). Neither succeeds in making justification and culpability mutually exclusive. So if Exculpation is correct—as Gardner and Simester claim—both (A) and (B) require revision.

For Simester, what afflicts (A) is the modest thesis. The remedy comes in the form of the strong thesis. If I am right, a different diagnosis presents itself. The problem with (A)—and, indeed with (B)—is the correspondence thesis. Why so? Because the correspondence thesis confers a justification on P for φing if P’s motivating reason turns out—for whatever reason—to be an undefeated or sufficient normative reason to φ. This is a problem because we have seen that P can be culpable for φing even if such normative reasons turn out to exist.

What, then, is the solution? My suggestion is this: both the correspondence thesis and the reasonable grounds thesis capture part of the truth. Each thesis, recall, consists of a biconditional. Each identifies necessary and sufficient conditions of acting for a reason of the right kind. The two cannot be combined—if one is right about sufficiency, the other is wrong about necessity. The necessary conditions they contain nonetheless can be. Putting them together gives us a further interpretation of MT. It gives us what I earlier called:

The Tripartite View—P is justified in φing if and only if (i) P φs for r; (ii) r is an undefeated normative reason for P to φ; and (iii) P regards r as undefeated on reasonable grounds.

The tripartite view avoids the objection I have pressed against (A) and (B). This is the upshot of condition (iii). We already know that this condition is not met in Operation 2. The same is true in Defender 2. Recall the three normative reasons at play in the example: P has reasons (1) to prevent herself being pushed, (2) to save E’s life, and (3) not to kill V. Since P is unaware of (2), she must regard (1) as undefeated by (3) on the basis that (3) is no weightier than (1). Does P live up to expectations in regarding (1) as undefeated on this basis? She does not. Those who find themselves under attack are expected to treat the value of their attacker’s lives as greater than the value of preventing minor intrusions on their autonomy.²⁴ It follows that P regards (1) as undefeated on grounds that are not reasonable. The tripartite view—like (B), but unlike (A)—denies P a justification for using lethal force.

So far so good. The tripartite view might nonetheless be thought to face counterexamples of its own. Since it incorporates the modest rather than the strong thesis, it might be thought to inherit the deficiencies of (A). Consider:

Defender 3—As in Defender 2, except P knows of but is indifferent to the threat to E.²⁵

Grant that P remains culpable for using lethal force. Conditions of justification compatible with Exculpation must deny P a justification. Since P acts for an insufficient reason, (B) does just that. The tripartite view, it may seem, does the opposite. Since P knows of the threat to E, P knows that her reason for using lethal force is undefeated. It may seem to follow that P’s grounds for regarding it as such cannot but be reasonable grounds. In truth, however, this does not follow. P’s motivating reason is (1)—the fact that she will otherwise be pushed. She is indifferent to (2)—the fact that her actions will save E’s life. It follows that P’s grounds for regarding (1) as undefeated by (3)—the fact that V’s life is threatened—must be the same as in Defender 2: given her indifference to (2), P must again judge (1) to be undefeated on the basis that (3) is no weightier than (1). We already know that this ground is not reasonable. It follows that the tripartite view denies P a justification in Defender 3.

6 Aptness

I have claimed that Gardner and Simester should endorse condition (iii). We know that both endorse condition (ii).²⁶ Combining the two might nonetheless be thought to create a puzzle. Imagine that, though condition (iii) is met, P is mistaken: P’s motivating reason—r—is not the undefeated normative reason P reasonably believes it to be. Imagine further that P is also unknowing: P has an undefeated normative reason to φ—call it q—of which P is unaware. If I am right so far, P is not culpable for φing. Nor is it impermissible for P to φ. Since q is an undefeated reason, P φing is not impermissible in the fact-relative sense.²⁷ Since P regards r as undefeated on reasonable grounds, P φing is also permissible in both the evidence-relative and belief-relative senses.²⁸ The tripartite view nonetheless denies P a justification. On that view, P is justified in φing only if the reason P reasonably regards as undefeated is the undefeated reason P reasonably regards it as being. The puzzle is why this should be the case.

What solves the puzzle, I will now argue, is a further answer to the explanatory question—one that is distinct from, though compatible with, the answers considered above. I first introduce that answer—which I call Aptness—then make a preliminary case for two propositions. First: that if Aptness explains why the conditions of justification are as they are, the tripartite view correctly captures the content of those conditions. Second: that Aptness compares favourably, in at least one respect, with other answers to the explanatory question.

A—Responsible Agency

Suppose P exercises an ability—she takes a shot, say, while playing a game of basketball. One question we can ask is whether the exercise was successful. A second is whether P’s success is to her credit. P shoots successfully if and only if the ball goes through the hoop. P’s success is to her credit if and only if it is owed to her skill as a shooter. This need not be so even if P is highly skilled—even if, that is, under appropriate background conditions, P reliably makes shots of the kind she takes. P’s success is not owed to her skill if, though she would otherwise have missed on this occasion, the ball is blown through the hoop by a freak gust of wind. P’s success, though it remains a success, is then a fluke. And fluky success, whatever else it is, is not to our credit.

It will help to introduce some terminology at this point. Let us say that P exercises an ability aptly if and only if P exercises the ability successfully, and P’s success is to her credit. P’s exercise of ability is not apt if it is unsuccessful, or if P’s success is a fluke.²⁹

Now it is a truism that both justified and unjustified acts are exercises of ability. They are exercises of our responsible agency: our ability to see reasons that bear upon our actions, and to make use of those reasons in acting as we do. Imagine my leg kicks out. You trip over it. If this is the result of a seizure, I am not culpable for tripping you up. This is not, however, because what I did was justified. That my leg kicked out was neither justified nor unjustified. It was not an exercise of responsible agency. And responsible agency is the ability that makes justification possible.

We exercise our responsible agency successfully if and only if our acts conform to undefeated reasons.³⁰ We exercise our responsible agency aptly if and only if our conformity is to our credit—if and only if the fact that our acts conform to undefeated reasons is owed to the competent exercise of that agency. According to what I will call:

Aptness—P is justified in φing if and only if P φing is an apt exercise of P’s responsible agency.

This, in a nutshell, is the further answer to the explanatory question advertised above. To endorse it is to claim that P acts without justification when P’s act—qua exercise of responsible agency—is either a failure or a fluke. The conditions of justification are as they are, Aptness implies, because justified action is neither of these things.

B—The Conditions of Justification

Grant, for a moment, that Aptness is sound. Like Exculpation, it implies that (A) and (B) require revision.³¹ Recall again Defender 2. P conforms to an undefeated reason. P acts for such a reason. From P’s perspective, however, this is a fluke. The fact which makes it the case that P’s reason for acting is undefeated (= the fact that E’s life is saved) is a fact to which P is oblivious. P’s reason for acting (= to prevent herself being pushed) is a reason P regards as undefeated only because she exaggerates the force of that reason, as compared with the force of the reason not to act as she does (= that doing so kills V). The exaggeration, I have claimed, is one P is properly expected to avoid. To conform to an undefeated reason, on this basis, is not to exercise responsible agency competently. It is to exercise that agency deficiently. Aptness therefore implies that P lacks a justification. (A), we know, reaches the opposite conclusion.

Much the same goes for Operation 2. P’s reason for acting is undefeated (indeed it is sufficient). This is so partly in virtue of Q’s consent. But P neither knows nor cares that Q consents. As far as P is concerned, this is a coincidence. It follows that, though P conforms to an undefeated (indeed sufficient) reason, her conformity is not to her credit. Aptness implies—pace both (A) and (B)—that P lacks a justification for operating on Q.

We can usefully compare Operation 2 with:

Operation 3—as in Operation 1, except Q consents.

In Operation 3, P exercises her responsible agency successfully. She conforms to an undefeated (indeed sufficient) reason to operate. P also exercises her responsible agency competently. Unlike in Defender 2, P is not unaware of the facts in virtue of which her reason for acting is undefeated. Unlike in Operation 2, she is not indifferent to them. P regards her reason for acting as undefeated because she believes that the aforementioned facts obtain: because she believes that she has Q’s consent to operate, and because she believes that the operation will prolong Q’s life. P proceeds only because she has adequate evidence for the truth of her belief: adequate evidence, that is, that her reason to operate is indeed undefeated. It follows that, in Operation 3, P’s success is owed to her competence: P conforms to the undefeated reason that is her reason for acting because she regards it as undefeated on reasonable grounds. Since this is so, Aptness implies, P is justified in operating. So, too, does the tripartite view.

It is in these last remarks that we find the solution to our puzzle. It is one thing to be successful. It is another to be competent. It is a third for one’s success to be owed to one’s competence. Aptness tells us that justification requires all three. The tripartite view gives us conditions of justification that make this so. Now suppose again that P makes a reasonable mistake in φing: P regards r as an undefeated reason to φ; she does so on reasonable grounds; but—as it turns out—P has no such reason. Suppose again that P is also unknowing: P conforms to q by φing; q is an undefeated reason for P to φ; but P is unaware of the facts which give her this undefeated reason.

Since P makes her error on reasonable grounds, she exercises her responsible agency competently. Since P conforms to an undefeated reason, the exercise is successful. P’s success is nonetheless not owed to her competence. The two are related only accidentally. P is successful because she conforms to q. But it is not because P believes that q obtains that P φs for r. P, after all, is unaware that q obtains. That it does, from P’s perspective, is a coincidence. It follows that P’s exercise of responsible agency is not apt. We know that the tripartite view denies P a justification. If Aptness is correct, it is right to do so.

C—The Priority of Justification

Is Aptness correct? One reason to think so emerges from a further truism: that it is better, all else being equal, to be justified than (merely) excused.³² Any (complete) answer to the explanatory question must account for the truth of this proposition. Since both justifications and excuses preclude culpability, Exculpation cannot. Permissibility might be thought to fare better. If justified acts are permissible, while excused acts are impermissible, we have a reason to prefer the former. This account, however, is extensionally inadequate. It fails to explain why justification remains preferable when questions of permissibility are not on the table. If I go out in the rain without my umbrella, say, I may or may not have a justification or excuse. What I do is not impermissible, however, even if I have neither.³³Permissibility cannot explain why it is nonetheless better if I have a justification for what I did.

Aptness succeeds where Permissibility fails. Or so I want to suggest. To endorse it is to claim that the conditions of justification are not sui generis. Rather, they instantiate distinctions we draw when considering any human exercise of ability. Those distinctions, importantly, are not evaluatively neutral. They determine the value we assign to the exercise at hand. We value success—all else being equal—over and above the corresponding failure.³⁴ We value success owed to competence—all else being equal—over and above coincidental success. Imagine cooking a meal, giving a lecture, or taking a three-point shot in a game of basketball. Better if the food tastes good, the lecture is educational, or the shot goes through the hoop. Better still if this is no accident but is owed to one’s ability—as a cook, a teacher, or a three-point shooter. Justification is better than excuse, Aptness tells us, because what goes for abilities like these ones goes for responsible agency too. Better to exercise that agency successfully than to fail. Better for one’s success to be to one’s credit than for it to be a fluke. To be clear: neither failure nor fortuity entails culpability. Excuses of various kinds remain available. Qua responsible agent, however, this is second best. Apt exercise is superior—all else being equal—to the alternatives. The conditions of justified action are the conditions under which exercises of responsible agency are apt. So justification is superior—all else being equal—to excuse.