Moral Rationalism and Moral Motivation

Abstract

Several prominent philosophers believe that moral facts are facts about what reasons we have, and that this entails that moral judgments are necessarily and inherently motivating. According to this argument, if morality cannot move us, then it is hard to understand how it could be sensibly regarded as action-guiding or normative. That is, they endorse a traditional argument for motivational judgment internalism based on moral rationalism. This paper criticizes this argument, and argues instead that there is no necessary or conceptual connection between moral facts and motivation. First, I formulate MJI as the thesis that moral judgments are necessarily and inherently motivating, and introduced several refinements designed to accommodate some plausible exceptions to the initial formulation. I then introduce MR as the thesis that moral facts are identical with or analyzable in terms of facts about requirements of reason. MR is ambiguous between an interpretation that analyzes moral requirements in terms of motivating reasons, and is also subject to various possible refinements, and an interpretation that proceeds in terms of justifying reasons. Finally, I argue that neither interpretation entails any interesting or plausible formulation of MJI. If the argument of this paper are sound, then there is no important connection between moral rationalism and motivational internalism.

1 Introduction

Several prominent philosophers have argued that the normativity of morality is closely related to its power to motivate us—to move us to action. According to this argument, if morality cannot move us, then it is hard to understand how it could be sensibly regarded as action-guiding or normative. For these philosophers, this connection between motivation and normativity takes the form of an argument for motivational judgment internalism (or MJI, the view that moral judgments are necessarily inherently motivating) that is premised on moral rationalism (or MR, the view that moral facts are or entail reasons for action). This argument is particularly appealing to those philosophers who accept a version of MJI that makes moral motivation conditional on practical rationality. According to this idea, morality yields reasons for action in practically rational agents, and these reasons must be capable of motivating those agents, which means that moral judgments are inherently and necessarily motivating.

The purpose of this paper is to show that this argument is unsound. There is no way of understanding MR such that it is capable of playing an important role in this argument for MJI. In what follows, I will present and explain motivational judgment internalism, moral rationalism, and the argument for MJI based on MR. I will then distinguish between several possible ways of understanding MR, and argue that there is no interpretation that yields a sound argument for internalism.

2 Motivational Internalism and Externalism

Many philosophers accept the thesis that the normativity of morality consists in its being inherently or intrinsically motivating. Parfit calls this the motivational conception of normativity.¹ Motivational judgment internalism (MJI) is a family of views according to which there is an “internal,” i.e., necessary, conceptual, or intrinsic, connection between one’s moral judgments and one’s motivational states. Though not all internalists see normativity as being intimately connected to motivation, the versions of internalism that are the topic of this paper are instances of this motivational conception.² MJI is to be understood as the claim that moral normativity is to be explained by or conceived as a relationship between our moral judgments and our motivational states. On this view, moral judgments are necessarily inherently motivating, in that it is not generally possible to form a moral judgment without being motivated to act in accordance with it.³

As a first pass, let us formulate this thesis in the following way, where ϕ is an act token and S is its agent:

Motivational Judgment Internalism (MJI): Necessarily, if S judges that she morally ought to ϕ, then S is motivated to perform ϕ.⁴

MJI says that necessarily, if S makes a moral judgment that S ought to ϕ (and, perhaps some other conditions are satisfied), then S will thereby be motivated to perform ϕ. This conditional is intended to be conceptually true—it is part of the concept of a moral judgment that sincerely assenting to it is necessarily and inherently motivating—and this conceptual necessity is designed to capture the idea that one’s moral judgments are intrinsically or inherently linked to one’s motivational states. MJI is ambiguous between strong and weak interpretations. According to strong internalism,

Strong Motivational Judgment Internalism: Necessarily, if S judges that she morally ought to ϕ, then S is overridingly motivated to perform ϕ.

Many philosophers recognize that SMJI is too strong. Shafer-Landau argues that not all sincere moral judgments produce action, and that this shows that the motivation generated by moral judgments is defeasible and can be overridden.⁵ Perhaps S recognizes that she is obligated to ϕ, but also recognizes that ϕ-ing would cost a lot of money or would hurt really bad, and so, although she is conflicted, she ultimately doesn’t do it. This suggests that a weaker version of MJI is more plausible. According to it,

Weak Motivational Judgment Internalism: Necessarily, if S judges that she morally ought to ϕ, then S is motivated to some degree to perform ϕ.

Some internalists argue that even WMJI is too strong. They argue that there are several possible conditions that might result in an agent making a moral judgment without forming any motivation to act in accordance with it: lack of virtue, weakness of will, sociopathy, etc. Michael Smith, for example, claims that although a good, strong-willed person will necessarily form some motivation to act in accordance with her moral judgments, a person who exhibited some form of practical irrationality might not. These considerations lead Smith to advocate a still weaker version of internalism, which he names the practicality requirement.

Smith’s Practicality Requirement: If S judges that she morally ought to ϕ, then either S is motivated to some degree to perform ϕ, or S is practically irrational.

According to Smith, moral judgments are necessarily motivating in the absence of “the distorting influences of weakness of the will and other similar forms of practical unreason on their motivations.”⁶ SPR accordingly restricts the motivational influence of moral judgments to agents who are not exhibiting such forms of practical irrationality.

Finally, consider Dreier’s moderate formulation of internalism. Dreier acknowledges the commonplace that SMJI and WMJI are subject to counterexample based on various kinds of “amoralist,” where an amoralist is someone who makes moral judgments without forming the corresponding motivational states. One example Dreier considers comes from Michael Stocker, and involves a politician who was formerly committed to easing the suffering of the less well-off all over the world, but who has become so disillusioned that he is no longer interested in helping anyone other than his close family and friends. Another example involves a group of Sadists who recognize various objects their society appraises as “morally good,” but who are repulsed by them and do not form even the slightest motivation to seek or promote such things.⁷ Dreier points out that each case involves an amoralist who is abnormal in a key way, against a background or more-or-less normal moral agents who characteristically make, and are motivated by, moral judgments. “I think it is crucial to the examples that they do present their exceptions by reference to a normal case, that it is necessary that such [normal] cases be present even to get the exceptions off the ground. So internalism can be saved by qualifying it so that it posits a necessary connection between the believed good and motivation only in the normal case.”⁸ Dreier defends Moderate Motivational Judgment Internalism:

Moderate Motivational Judgment Internalism: Necessarily, if S judges that she morally ought to ϕ, then S is normally motivated to perform ϕ.

MMJI, like Smith’s Practicality Requirement, is not affected by counterexamples involving amoralists of various kinds and is capable of accommodating the occasional agent who is afflicted by weakness of will or sociopathy.⁹

The alternative to MJI is motivational judgment externalism, which is the view that there is no necessary connection between a person’s moral judgments and her motivational states. This view is the denial of MJI:

Motivational Judgment Externalism: It is not the case that there is a necessary connection between S’s judgment that she morally ought to ϕ and S being motivated to perform ϕ.

According to MJE, moral judgments are not inherently or necessarily motivating, even if moral agents are usually disposed to be motivated to comply with their moral judgments. If it is generally or usually true that moral agents are motivated to comply with their moral judgments, this is not because of any necessary or inherent connection between their moral judgments and their motivational states. On externalism, something might count as a moral judgment even if the person whose judgment it is was not motivated to any degree to comply with it.

3 Moral Rationalism and the Argument for MJI

According to moral rationalism, there is an important connection between moral facts and moral reasons—MR is the view that moral facts are, or constitute, or otherwise entail, reasons to act. Smith, for example, formulates MR in the following way:

Moral Rationalism: If S morally ought to ϕ, then there is reason for S to perform ϕ.¹⁰

Although Smith formulates the principle as a conditional, he is explicit that he intends moral rationalism to be an identity claim. “In other words,” he writes, “moral facts are facts about our reasons for action; they are themselves simply requirements of rationality or reason.”¹¹

Prominent internalists have often defended MJI by appeal to MR. For example, Smith writes:

It is a platitude that an agent has a reason to act in a certain way just in case she would be motivated to act in that way if she were rational. And it is a consequence of this platitude that an agent who judges herself to have a reason to act in a certain way—who judges that she would be so motivated if she were rational—is practically irrational if she is not motivated to act accordingly. For if she is not motivated to act accordingly then she fails to be rational by her own lights. But if this is right then it is clear that the third form of internalism [moral rationalism] entails the second [motivational judgment internalism in the form of SPR].¹²

Informally, the idea is that moral facts and/or judgments are properly reason-giving and action-guiding only if they are able to motivate moral agents to act in accordance with them. Smith claims that it is a platitude that S has a reason to perform ϕ just in case she would be motivated to perform ϕ if she were rational. If so, then when S judges that she has a reason to perform ϕ, she has judged that she would be motivated to perform ϕ if she were rational. If she judges that she would be motivated to perform ϕ unless she were irrational without also being so motivated, then, by her own lights, she is irrational.

Later, Smith argues that normativity is connected to motivation on the basis of “the truism that we expect agents to do what they are morally required to do.” The idea is that we have this expectation not only in the (trivial) sense that we think agents should do what they morally ought to do, but also in the sense that their doing so is predictable—we think that they will in fact act as they morally ought.¹³

Christine Korsgaard advances a similar argument. According to Korsgaard, “if [practical reason claims] are really to present us with reasons for action, must be capable of motivating rational persons.”¹⁴ In defense of this view, she borrows Williams’ point that an external, justifying reason cannot be used to explain why S has performed ϕ—only a motivating reason can do that—and Nagel’s point that a justifying reason cannot prompt S to perform ϕ unless that justifying reason is, or is accompanied by, a motivating one. On this view, the purpose of moral discourse is to move us to action, and so if moral facts generate reasons, these reasons must be of a sort that is capable of motivating us.¹⁵ This suggests that if MR is true, and moral facts provide moral agents with reasons to act, then it must be possible for these moral agents to be motivated by the reasons those moral facts provide. If so, then MR entails MJI.¹⁶

Mark van Roojen makes a related point:

Rationalism suggests a very tight necessary connection between sincere moral judgment in rational people and motivation. It explains this connection in much the same way that one might explain why rational persons will do what they believe they have most reason to do. For, according to a rationalist, to a first approximation, the belief that ϕ-ing is right for one is equivalent to the belief that one has overriding reason to ϕ. And it looks like it is a requirement of rationality that one be motivated to do what one believes one has overriding reason to do. The problem for rationalism is thus explaining the weaker version of the theory suggested by moderate internalism. … there are reasons to think that even rational people can be unmotivated by what they regard as true moral judgments. This is why it is necessary to include Dreier’s “normally” in the correct statement of internalism even if our quantifiers already range over only rational people.¹⁷

van Roojen argues that MR entails that moral judgments are analytically equivalent to judgments about what reasons there are and what it would be rational to do, because MR is the thesis that moral truths are truths about reasons and rationality. And it is manifestly irrational for S to judge that it would be rational for S herself to perform ϕ, or that ϕ is what S has most reason to do, without S coming to form a motivation to perform ϕ.

This argument claims that morality is normative and action-guiding, in the sense that it provides reasons for action of a sort that might move agents to act, and that this rational normativity supports the motivational thesis contained in MJI. According to MR, moral facts are or constitute reasons, and so recognition of a moral fact is or constitutes the recognition of a reason. But no rational agent can recognize a reason without being moved to comply with it—if she is not motivated to comply with the reason, then either she has not recognized it, or she is irrational. The various exceptions to this general rule of thumb can all be ascribed to practical irrationality on the part of the agent, which means that, so long as the agent is behaving rationally, one expects that she will act in accordance with her moral obligations. If so, then moral judgments are inherently and necessarily motivating in rational agents, and MJI is true.¹⁸

4 Why Moral Rationalism Does Not Entail MJI

This traditional argument for MJI based on MR is unsound. There is no coherent way of understanding moral rationalism that entails motivational judgment internalism, even in its moderate, qualified formulations. In this section, I will distinguish between several interpretations of MR, and show that none of them yields a cogent argument for MJI. These interpretations of MR stem from a corresponding set of interpretations of what reasons are, or what having a reason amounts to. Let us, then, draw a familiar distinction between motivating reasons, which are motivationally-relevant considerations that can be used to explain behavior, and justifying reasons, which are considerations relevant to justification and have no in-principle connection to motivation. The idea is that someone, S, has a motivating reason, M to perform an action, a just in case M could be cited in an explanation of S’s intentionally performing a; whereas M is a justifying reason for S to perform a just in case M could be cited in an explanation of why S would have been justified in performing a, whether or not M played any role in motivating S to perform a. Moral rationalism accordingly admits of interpretation as a claim about motivating reasons, and as a claim about justifying reasons. That is, between:

Motivating Reason Moral Rationalism: If S morally ought to ϕ, then S has a motivating reason to perform ϕ.

and

Justifying Reason Moral Rationalism: If S morally ought to ϕ, then there is a justifying reason for S to perform ϕ.

Very simply, motivating reason moral rationalism (MRMR) is the claim that there is an important connection between ϕ’s being morally required and its agent’s having a motivating reason in favor of performing it, and justifying reason moral rationalism (JRMR) is the claim that there is a connection ϕ’s being morally required and its agent’s being justified in performing it. JRMR has no direct implications about motivating reasons. The key premise of the argument for MJI therefore also admits of (at least) two corresponding interpretations. Let us consider each interpretation in turn.

4.1 Motivating Reason Moral Rationalism

It is unlikely that motivating reason moral rationalism expresses the principle that Smith and others have in mind—Smith’s view is that moral facts are facts about justifying reasons, rather than motivating reasons.¹⁹ One might nevertheless suspect that an argument from MRMR to MJI would be intuitively plausible, because MRMR and MJI are each claims about our motivating reasons. Despite this initial plausibility, however, MRMR does not entail MJI. MRMR and MJI yield conflicting results in situations in which ϕ is morally required for S to do without S realizing it. Belief is logically independent of truth, in the moral realm no less than any other, and so the moral judgments of moral agents will sometimes fail to reflect the facts. Even the most wise and practically rational agent will occasionally fail to realize which action is required, perhaps on the basis of misleading evidence.²⁰ Perhaps her evidence is so misleading that, although ϕ is required, she comes to believe that it is wrong. If so, then MRMR implies that she will nevertheless be motivated by ϕ’s being required to perform it, despite the fact that she believes that ϕ is wrong.²¹

MJI has the contrary implication: it implies that if S believes that ϕ is wrong, she will be motivated to refrain from performing ϕ. In cases in which S’s moral judgment is incorrect, the implications of MRMR and MJI diverge. MRMR implies that S’s motivational states with regard to ϕ will vary according to ϕ’s actual moral status, but MJI implies that S’s motivational states will vary according to her judgments concerning ϕ’s moral status, and it is possible for S’s judgments about ϕ’s status to fail to reflect ϕ’s actual status. MRMR therefore implies that, in such cases, S will be motivated to perform ϕ, but MJI implies that she will be motivated not to perform it. Restrictions to moral agents who are practically rational, as Smith has proposed, or to those who are otherwise normal, as in Dreier’s proposal, are of no help, for an agent who is practically rational or otherwise normal will occasionally assent to false moral judgments. In such cases, MRMR entails that she will be motivated by the facts, whereas SPR and MMJI entail that she will be motivated by her judgments.

MRMR therefore does not entail MJI, and does not provide evidence that MJI is true. It does not imply that moral judgments are inherently or necessarily motivating. It is the view that moral facts are related to motivation, and is silent as to whether judgments are. MRMR implies that S’s motivational states will track ϕ’s actual moral status, but there is no necessary or internal relationship between ϕ’s being right and S’s judgments about it. Indeed, it implies that agents will be motivated against acting in accordance with their judgments in an important range of cases, in which the agent’s judgments are contrary to the facts. The problem is not that MRMR is incompatible with MJI, or that the two principles generate incompatible results. The problem, rather, is that they generate results that are independent of one another.

Furthermore, in addition to showing that MRMR does not entail MJI, these data show that MRMR is not true. The influence of moral facts over the motivational states of moral agents is highly indirect and subject to influence by other factors in a way that is inconsistent with MRMR. In particular, MRMR implies that the rightness of ϕ exerts such an influence on the will that S will be motivated to perform it even in cases in which she has no doubt that ϕ would be egregiously, disastrously wrong. This is implausible. In cases in which a person’s judgments are at variance with the facts, one does not expect her motivational states to reflect the facts.

We might attempt to revise MRMR in order to accommodate this, as follows:

Motivating Reason Moral Rationalism′: If S morally ought to ϕ, and S knows that she morally ought to ϕ, then S has a motivating reason to perform ϕ.

The objection raised against MRMR does not work against MRMR$^{\prime }$, because MRMR$^{\prime }$ has no implication for cases in which S’s judgments are mistaken. Knowledge entails truth, of course, so S knows that ϕ is required only if it is required. But MRMR$^{\prime }$ does not entail MJI, because, unlike MRMR$^{\prime }$, MJI is not restricted to cases in which S’s judgment counts as knowledge. Suppose that S believes that she ought to ϕ, and that her warrant for this belief is insufficient for knowledge. If so, MJI implies that S will be motivated to perform ϕ, and MRMR$^{\prime }$ has no implication for this case. It is compatible with MRMR$^{\prime }$ that S lack motivation to perform ϕ in such a case, which means that MRMR$^{\prime }$ does not entail MJI.

In light of this, consider a second revision of MRMR:

Motivating Reason Moral Rationalism′′: If S morally ought to ϕ, and S judges that she morally ought to ϕ, then S has a motivating reason to perform ϕ.

This revision brings the principle in line with MJI in cases in which S believes but does not know that ϕ is right, but it still does not entail MJI. As was noted above, MJI entails that S will be motivated to ϕ in cases in which S’s moral judgment is incorrect, whereas MRMR$^{\prime \prime }$ does not have this implication. For, unlike MJI, MRMR$^{\prime \prime }$ contains a restriction to cases in which S’s judgment that she ought to ϕ is correct. If so, then MRMR$^{\prime \prime }$ is silent in the case in which S’s judgment is incorrect, and is compatible with S lacking motivation to ϕ in such a case, while MJI is not. MRMR$^{\prime \prime }$ therefore does not entail MJI. And again, restrictions to agents who are free from various forms of practical irrationality and otherwise normal are unhelpful, for normal, practically rational agents nevertheless remain susceptible to errors in moral judgment.²²

A further revision that eliminated the reference to ϕ’s being right would give up the ghost of moral rationalism. Moral rationalism is a thesis about moral facts. The idea is supposed to be that moral facts are facts about what one has reason to do. But this version of moral rationalism would not assert any connection between moral facts and reasons for action; instead, it would assert a connection between an agent’s moral judgments and her motivational states. It would, in short, be equivalent to motivational judgment internalism. It would therefore fail to provide any independent evidence for MJI. Such an argument would be trivial or question-begging.

Still, one might suspect that there is something to the argument, put most clearly by Smith, that if moral requirements are rational requirements, then an agent who believes herself to be morally required to ϕ without being motivated is irrational by her own lights. If moral requirements just are rational requirements, then to believe oneself to be morally required to ϕ is to believe oneself to be rationally required to ϕ, and it would be paradigmatically irrational to fail to comply with what one takes to be one’s rational requirements.

Although this argument might appear to be plausible, it is not sound. For the agent in question might not endorse moral rationalism, which means that she might believe that she is morally required to ϕ without thereby conceiving of this moral requirement as a requirement of reason. Belief is opaque to this kind of conceptual analysis. Just as a person could have various beliefs about water without realizing that water is H₂O, or could have various beliefs about knowledge while believing (contra Gettier) that knowledge is justified true belief, a person could have various moral judgments without believing that moral requirements are requirements of rationality, even if that is ultimately what they are. For property identity does not entail conceptual identity. If so, then a person could accept that ϕ is right without accepting the rationalist analysis of “ϕ is right,” even if that analysis is in fact true. If so, then S might believe that ϕ is morally right without believing that performing ϕ is a requirement of rationality, which means that S might believe that ϕ is morally right without seeing herself as rationally required to do it.

If she does not believe that moral requirements are rational requirements, or conceive of them as such, then she might believe that ϕ is right without (thereby) believing that it would be irrational for her to refrain from performing ϕ. If so, then it is not the case that she is irrational by her own lights. Perhaps refraining from ϕ would in fact be irrational, but she is not herself committed to that view. There is nothing in her set of beliefs that is necessarily connected to a motivational state on her part, or to the view that she, herself, is irrational if she lacks such a motivational state. This means that MR does not rule out the possibility that S believes that ϕ is morally right without thereby being motivated to perform it. And since there is nothing in the case as described that entails that S is abnormal or otherwise irrational, it is possible for MR to be true while MMJI is false.

This suggests that MRMR entails MJI only on the further assumptions that (i) moral requirements are conceived of by their agents as rational requirements, (ii) agents always correctly identify their rational requirements, and (iii) (practically rational) agents are always motivated to comply with what they see as their rational requirements. But these assumptions—(i) and (ii) in particular—go well beyond the content of any formulation of moral rationalism, and, taken together, are equivalent to MJI. Any argument for MJI that relies upon these further assumptions is question-begging. Moral rationalism therefore does not provide any independent evidence for MJI.

However, one might think that further refinements to moral rationalism introduced by van Roojen are useful. He distinguishes between objective and subjective senses of rationality: what is rational in an objective sense is not relativized to the actual features of the agent’s circumstances and makes use of various idealizing assumptions about how a perfectly rational agent with full information and perfect reasoning skills would behave; various ways of being subjectively rational are less idealized, in that they are relativized to the agent’s available evidence, limited time for deliberation, etc.²³ “While this complicates the kind of judgment internalism rationalism will underwrite,” he writes, “the complication will again be welcome. It will enable us to explain [certain kinds of] amoralists… Since these sorts of amoralists were the primary motivation for accepting the normalized internalist claim embodied by moderate internalism, the complication will cause our rationalist explanation of judgment internalism to match what needs to be explained.”²⁴

van Roojen’s refinements to moral rationalism are designed to accommodate the fact that actual agents can be limited in various ways: not only can they misidentify their moral permissions and obligations, but they can also misidentify the nature of moral requirement. That is, an otherwise rational moral agent might form moral judgments without realizing that such judgments can be conceptually analyzed as judgments about what is rationally required. If this is correct, then MR is compatible with the possibility of the occasional rational person, S, believing that ϕ is right without being motivated to perform it, when S is not in a position to recognize that the property being right is identical with the property being rationally required.²⁵

Subjective Reason Moral Rationalism: If S judges that she morally ought to ϕ, and S believes that moral requirements are requirements of practical reason, then S normally has a motivating reason to perform ϕ.

van Roojen then points out that, in a community in which no one was moved by their “moral” judgments, these judgments would not be judgments with genuine moral content. That is, if someone believes that an action is right, this belief depends on her being appropriately related to a background moral community in which the rightness of the action normally motivates rational agents.²⁶ This suggests that according to SRMR, the fact that moral requirements are identical to or analyzed as requirements of reason normally generates motivation in rational agents to comply with their own moral judgments. If this is right, then it seems plausible that SRMR entails Dreier’s MMJI, and possibly Smith’s Practicality Requirement.

But despite this initial plausibility, SRMR does not show that moral rationalism entails motivational internalism. First, van Roojen’s discussion yields the conclusion that rational moral agents can fail to be motivated by their own moral judgments if they fail to recognize that moral facts can be conceptually analyzed as facts about practical reason. If so, then whether SRMR entails MMJI depends on whether moral agents normally accept moral rationalism. However, I strongly suspect that most people have not thought about whether moral rationalism is true, and so people do not normally believe that moral judgments can be conceptually analyzed as judgments about rational requirements. And even if this is not actually the case, such a community certainly seems possible. Since SRMR predicts that rational agents will be motivated to comply with their moral judgments only when they recognize their moral judgments as judgments about rational requirements, this means that SRMR does not support MMJI.

Second, SRMR does not support Smith’s Practicality Requirement. SPR says that agents are motivated to comply with their moral judgments unless they are practically irrational. SRMR, on the contrary, shows how an agent could fail to me motivated by her judgments despite being fully practically rational. She could simply fail to recognize that the content of her moral judgments is identical to or analyzable in terms of reasons for action. SRMR implies that such agents might rationally fail to be motivated by their own moral judgments.

Finally, the argument in favor of MJI based on the necessity of a background community whose members are generally motivated by their own moral judgments is a separate, direct argument for MJI.²⁷ The purpose of this paper is to rebut the argument that moral rationalism entails MJI. The Background Community argument is therefore beyond the scope of this paper.

4.2 Justifying Reason Moral Rationalism

Justifying reason moral rationalism, unlike MRMR, is a thesis about the relationship between moral facts and the existence of justifying reasons. According to JRMR, if ϕ is morally right, then there is some (potentially defeasible) justification for S to perform it. But S might not know that ϕ is right, or might not care about or be sensitive to this fact, and so might not form any motivation to perform it.

Interpreted in terms of justifying reasons, moral rationalism spells out a consequence of the fact that ϕ is right, but this consequence is restricted to the realm of justification and has no direct implications for the logically or conceptually necessary connection to what agents will actually do, or what their motivational states will be. Although JRMR says that moral facts are or entail reasons to act (or refrain from acting), it does so in a way that has no inherent or necessary relationship with our motivational states. As long as there is a conceptual distinction between S judging that ϕ is required and ϕ being required, and as long as it is possible for S to judge that ϕ is required when it is not (or vice versa), then it is also possible for ϕ to be required when S has no motivation to perform it. For these reasons, I suggest that JRMR is the most plausible interpretation of moral rationalism.

JRMR does not entail MJI, because justifying reasons are distinct from and in-principle independent of motivating reasons. S might be justified in performing ϕ without being motivated, and might be motivated to perform ϕ without being justified. On JRMR, the rightness of ϕ entails that there is a justifying reason in favor of its agent performing it; this has no direct, conceptual, or otherwise necessary bearing on her motivational state. This suggests that Smith’s argument from moral rationalism to motivational internalism is unsound if moral rationalism is conceived of as a thesis about justifying reasons. This argument, which was rehearsed in §3, is based on the idea that moral facts and judgments are facts about reasons, and these reasons are properly action-guiding only if they are able to motivate moral agents to act in accordance with them. But according to JRMR, moral facts are facts about justifying reasons rather than motivating ones, and justifying reasons lack any conceptual, intrinsic, or inherent connection to our motivational states. For a consideration might in fact justify an agent to perform ϕ without her being aware of it, in which case she will lack motivation in spite of the fact that ϕ is justified. As previously mentioned, many moral agents are not moral rationalists of any kind, and are therefore not committed to the view that moral judgments are equivalent to judgments about justifying reasons, even if that view is in fact true.

Similarly, a person might see that a concern is potentially justifying—that a reasonable person with a reasonable set of priorities would regard it as justifying—but without sharing that set of priorities and without being motivated by whatever justificatory force the consideration might have. That is, an agent might be aware of the fact that ϕ is justified, or justifiable, without being motivated to perform it, if she does not happen to be interested in whether her actions are justified or justifiable, or if she believes that there is another action, ψ, such that ψ is equally justified and is preferable for other, non-justifying reasons.

For example, a certain kind of baseball coach could be aware of the current research showing that an offensive strategy that emphasizes getting on base, drawing walks, minimizing strike-outs, and hitting for power leads to a higher rate of runs per game and increases the probability of winning, but, without contradiction or incoherence, remain unmotivated by this awareness to pursue this strategy. This kind of coach sees “small ball” strategies that emphasize base-hits, stolen bases, sacrifice bunts, and speed on the base-paths, and which de-emphasizes walks and home-runs as being the right way to play the game, and would rather lose playing the game the right way than win playing the wrong way. It’s not that she doesn’t understand the arguments against small-ball strategies, or why those considerations are alleged to confer justification. She simply doesn’t want to win if it means winning the wrong way. Although she recognizes that increasing the likelihood of winning is a consideration that could potentially confer justification, she does not see this consideration as being intrinsically important. What is important to her is seeing to it that the game is played the way, in her view, it is meant to be played.

5 Conclusion

I have argued that no interesting version of moral rationalism entails any interesting version of motivational judgment internalism. Moral rationalism does not entail that there is a necessary or conceptual connection between moral facts and motivation, nor any such necessary or conceptual connection between moral judgments and motivation. This argument proceeded in several steps. First, I formulated MJI as the thesis that moral judgments are necessarily and inherently motivating, and introduced several refinements designed to accommodate some plausible exceptions to the initial formulation. I then introduced MR as the thesis that moral facts are identical with or analyzable in terms of facts about our reasons for action—that moral requirements are requirements of reason. I pointed out that this thesis is ambiguous between an interpretation that analyzes moral requirements in terms of motivating reasons, which was also subject to various possible refinements, and an interpretation that proceeds in terms of justifying reasons. I then argued that neither interpretation entails any interesting or plausible formulation of MJI. If the argument of this paper are sound, then there may be a true version of moral rationalism, but not one that supports motivational internalism.²⁸