1. Frankfurt and Watson: The Discrepancy Worry and the Challenge of Passivity

In his treatment of autonomy, Harry Frankfurt aims to critique a too simplistic Humean belief-desire conception of agency. His strategy is to show how the simple Humean picture of our psychology cannot account for the alienation experienced by unwilling addicts, nor other people who act on desires that they do not reflectively endorse as adequate reasons for action. In an evocative example, Frankfurt focuses on a desire to injure, which one might act upon, but where this would not amount to an action one freely willed:Footnote ¹

“It means that he does not want to injure him, even though the desire to injure is something that he experiences.” (Frankfurt, Reference Frankfurt1998b, 68)

Frankfurt is focusing on the experience one might have of aspects of one’s own psychology as something alien to one’s self-reflective stance on one’s agency. Frankfurt identifies alien psychological influences, in this case desires or passions,Footnote ² by considering whether they accord with our self-conscious evaluations about which motives we wish to be influenced by. Alien psychological influences are those which I upon reflection would not wish to be moved by; when there is a discrepancy between what I would want to influence my actions and those passions that are poised to exert such influence.

While disapproval might be a necessary condition for externality, it surely cannot count as a sufficient one. As Frankfurt (Reference Frankfurt1998a, 65–66) points out, our attitudes toward our passions cannot themselves figure as the definition of alienness, as these evaluative attitudes may themselves be alien influences in our psychology. Only when my disapproval is grounded in non-alien aspects of my psychology, does such disapproval serve as a criterion of the alienness of a passion.Footnote ³ This instigates an immediate danger of a vicious regress. We are trying to explain the alienness of a passion by reference to its discrepancy with other evaluative psychological states, which then again themselves need to be of a non-alien kind. But what establishes these further psychological states as non-alien?

Frankfurt’s own account takes the evaluative psychological states to be a form of higher-order desires. Yet, the regress-concern only increases when it is our further desires about what desires we desire to have, which serves as the evaluative psychological state. Partly in an attempt to avoid such a regress, Watson (Reference Watson1975, 209–210) has suggested that the troubling discrepancy should not be conceived as one between different orders of desires. Instead, this type of psychological unfreedom occurs when we act upon desires that differ from our value judgments.Footnote ⁴ Watson employs the example of someone whose choices are still governed by his sexual desires even though he judges such desires to be impure or corrupt. In other words, he does not value sexual activity but is nevertheless moved by his sexual desires. The core of Watson’s account is to employ the distinction between desiring something and judging it to be good. He then argues that only those actions which are motivated by our judgments of what is good are reflectively free and count as expressions of our Will.

Ultimately, we will argue that Watson’s account pushes us in the right direction by changing the motivating evaluative psychological state from an emotive state to a cognitive judgment of goodness. However, his account falls short when he allows mere value judgments to form the non-alienated evaluative basis.Footnote ⁵ Watson (Reference Watson1975, 211) is concerned about the source of our wants, whether they originate in our desires or our values. Yet, this fails to address another worry, the alienation one feels when the very process of self-conscious reflection is experienced as ineffective in occasioning what one does. Our idea of reflective freedom is an idea involving a causal role for one’s Will, and neither Watsons’s nor Frankfurt’s early accounts adequately address this.

Frankfurt’s and Watsons’s presentations of reflective unfreedom highlights what we can call the discrepancy worry. However, they overlook a further worry about unfreedom that seems to arise from the perspective of self-conscious reflection. As Kant points out (1997 [1770s], 28: 269), it is difficult to experience one’s activities as freely willed, while one conceives of one’s actions as fully occasioned by elements of one’s psychology that one simply empirically discovers as present. Reflectively free action presupposes an active casual role attributed to the reflective practices of the subject (Kant, Reference Kant, Gregor and Korsgaard1998 [1786], 4:446). This is undermined by the picture where I simply discover the desires that drive me as brute unalterable facts. Yet, this same worry about passivity can equally arise with respect to my evaluative judgments. Maybe Watson is right that we to some degree identify more fully with our value judgments than with our mere desires, and in that sense those actions that are motivated by our values count as more fully ours.Footnote ⁶ However, it still seems that one can be as passive in these cases as when one simply exhibits Humean instrumental rationality in the pursuit of passively discovered desires. Upon self-reflection, I may simply discover that I judge certain pursuits good and others not. In Watson’s account, psychological freedom then consists solely in the instrumentally rational pursuit of living up to my passively discovered value judgments. While I might not feel as alienated from my value judgments as I do from my mere desires, it is difficult to see why my reflective freedom has increased if I, as a reflective being, is equally passive in both cases.

Watson might follow Kant (Reference Kant and Smith2007, [1787], A68/B98) in thinking that judging is inherently an active endeavor and, therefore, think this worry cannot get off the ground. If that is so, then his theory is close to ours, but in this case his elucidation fails to draw the full consequences of his position. What these are will be elaborated later. Alternatively, Watson may follow more ordinary modern philosophical terminology in which “judgement” and “belief” is used loosely and somewhat interchangeably. In this way of speaking, beliefs and judgments are something that we can discover within ourselves. Self-reflection may lead us to realize that we already value a certain set of actions or outcomes; in Watson’s terminology, we discover that we already conceive of them as good. Our current values are themselves potential objects of empirical psychological discovery, just like our passions are. If that is the case, then it is difficult to see why, from the point of self-reflection, one does not feel as bound by one’s discovered value judgments as one can be bound by one’s discovered passions. To return to Watson’s own example, one may feel as oppressed by the motivation stemming from one’s own value judgment that sexual activity is impure and corrupt, as one feels oppressed by the sexual desires that one cannot resist acting upon.Footnote ⁷

Reflections as those previously given serve to highlight the second worry that guides our approach to reflective freedom—the causal efficacy worry. When explaining reflective freedom, our self-reflective capacities can neither be relegated to mere determiners of Humean instrumental effectivity, nor to mere observers. Rather, reflection itself must, in Kantian terminology, be the determining ground of our choice to perform an action.

It is important to distinguish two ways in which our reflective capacities may be causally idle in determining our actions. They might be fully causally idle, in so far as it makes no causal difference to what we do that we self-consciously observe a discrepancy between what our passions leads us toward and what our second-order desires or value judgments propose. Frankfurt and Watson need not accept this fully epiphenomenal form of casual idleness. They might claim that our mere passive discovery of a discrepancy between what we Will and what our passions lead us toward causally serves to make us more resilient toward the influence of those rogue passions.Footnote ⁸ However, nothing in their theories states how or why freedom requires that our reflective capacities can influence the very determination of our second-order desires or our value judgments. Their theories allow for full reflective passivity in the determination of what it is we reflectively think we should aim for, either in the sense of second-order desires sense or value judgments. These governing aims are simply passively discovered empirical facts about our own psychology. Freedom then consists in using reflection to ensure that we most effectively act in accordance with the governing aims, rather than simply in the pursuit of rogue passions.

Viewed in this light, it becomes clear how narrow the philosophical distance is between such theories and the full Humean reduction of rationality to mere instrumental effectiveness (Hume, Reference Hume and Nidditch1978, [1739], 415). On this expanded form of Humean reasoning, reflection might not solely have the task of making us instrumentally effective in uncritically satisfying all our passions. But it remains restricted to the task of making us instrumentally effective in uncritically satisfying a subset of our empirically discovered psychological states, be they second-order desires or existing value judgments. The feeling of alienation experienced when reflection is ascribed merely this passive role shows that Kant (Reference Kant and Gregor2015 [1788], 5: 16) is right in claiming that the role of the Will in action cannot simply be of the subservient kind imagined by Hume. Rational self-reflection must be more than the instrumentally efficient slave to the guiding hand of pregiven aspects of our psychology; aspects which we simply empirical discover, be they passions as in Frankfurt and Hume, or existing value judgments as in the case of Watson.

2. Velleman and Sartre: The Efficacy Worry and the Challenge of Arbitrariness

The causal-efficacy worry has led us to conclude that we only experience ourselves as reflectively free if the Will possesses an efficacious motive force that casually influences what we do. Moreover, this influence must reach beyond the mere role of increasing instrumental efficacy. The will must be one of the drivers behind what actions we reflectively decide to do. This means that the role of reflection cannot simply be to provide a cognitive evaluation of various activities as, say, good, interesting, depraved, or pleasant. Reflection itself must also serve to motivate us toward pursuing some specific actions because they are so evaluated.

In his extensive discussions with Frankfurt, Velleman (Reference Velleman2015c) also alights on the lack of an efficacious motivational force deriving from the Will itself. A force which determines what is to be pursued and does not merely lend further strength to motives one passively discovers as already privileged in one’s psychology. The Will is identical to the self-conscious perspective of practical reflection. It is the perspective from which those experiences of psychological unfreedom that we wish to diagnose arise. Hence, if that very process is experienced as subservient to anything external to it, or as ineffective in determining our actions, then from our deliberative perspective we experience our Will as unfree.

Frankfurt himself has also attempted to address the potential passivity of one’s Will in his further developments of his early theory. In these developments, Frankfurt (Reference Frankfurt1988, 54) thinks that:

The notion of “identification” is both meant to highlight the active aspect of the role of reflection in free agency and to figure as that which establishes the non-alien character of that which is identified with. Velleman (Reference Velleman2015c, 118) interprets Frankfurt as using “identification” to pick out an adjudicating role where the Will is active, as it is our faculty for self-reflection that performs this form of adjudication.Footnote ⁹ However, what Velleman correctly points out is that the norms determining this form of adjudication cannot simply be that one should follow the pursuit of some desires nor the enactment of some values. For, as it is the very normative authority of these values and desires that we are trying to reflectively evaluate, we cannot rely on their aims as those setting the norms of adjudication. As a consequence, Velleman (Reference Velleman2015c,120) follows Kant in thinking that: “The deliberative processes constitutive of agency requires a distinctive motive of their own.”Footnote ¹⁰ Velleman’s solution comes in two steps. The first step highlights the analytical entailment that the Will, that is practical reflection, is governed by practical reasons. From the reflective self-conscious perspective, to ask the question what one ought to do is simply asking what reasons one has for and against pursuing various courses of action. As Velleman (Reference Velleman2015c, 120) writes: “What animates practical thought is a concern for acting in accordance with reasons.” It is inherent to the process of deliberation that it is a process of discerning and being motivated by reasons. Hence, by relying on reasons in determining which desires or values we should allow to guide our lives, the Will is adjudicating on its own premises without relying on the authority of anything alien to it.

Thus far Velleman’s theory is also in accordance with the Kantian account of freedom as autonomy.Footnote ¹¹ According to Kant, to act freely must be to act in accordance with reasons. If acting based upon self-reflection is the epitome of reflective freedom, then nothing external to that process of self-reflection can be granted normative authority over the outcome. Any such external authority would from the perspective of the Will, that is from the perspective of self-conscious practical deliberation, simply be empirically discovered when we self-reflect. Consequently, our hostage to its verdicts would amount to subservience to an alien influence. Where Velleman differs from Kant is in his conceptions of the nature of those practical reasons. Where Velleman ascribes a mere formal motive to the Will, Kant requires the motive to be substantial. Thus, while Kant and Velleman superficially agree in rejecting Humeanism by ascribing a motive to the Will, they differ in their interpretation of what this amounts to.

For Kant, the constitutive goal of acting in accordance with reasons is identical to a noninstrumental reflective pursuit of the good. To be reflectively free requires that one choose the acts one reflectively knows to be good and, importantly, that one chooses so because one’s motivation is to do what is good.Footnote ¹² For Kant, to act freely one must thus take an action’s goodness as a noninstrumental reason for performing it. Velleman disagrees in this Kantian, Platonic, and Aristotelian conception of reasons for action. We can, according to Velleman (Reference Velleman2015a, 98), freely do what is bad precisely because we can take an action’s badness to be a reason for doing it. He thinks an acceptable account of human agency must allow that we can be reflectively free even while committing pointless, outrageous, evil, or heinous acts. The rest of the paper will be concerned with why we take Velleman, and Sartre who agrees with him, to be wrong in this.

3. Substantive and Executive Conceptions of the Motive of the Will

To explain Velleman’s conception of the motive of the Will, we can draw an analogy to the distinction between substantive and executive virtues. Substantive virtues, such as justice and wisdom, can only be exercised in the pursuit of the good. In contrast, executive virtues, such as bravery, loyalty, or prudence, can be exercised in the pursuit of any aim, good or bad. Since Velleman thinks that the pursuit of any goal is compatible with acting in accordance with one’s reasons, he has to abandon the idea that the motive of the Will itself can be the pursuit of a substantive goal. This forces him to attribute a mere executive motive to the Will, a motive which one can pursue while aiming for any type of goal.

For Velleman, the distinctive motive of the deliberative process is self-knowledge about what one is doing. Velleman sees this as a more specific instance of a general intellectual motive for understanding the world.

Velleman here echoes the Kantian idea that we must conceive of our faculty of Reason as being governed by self-standing regulative ideas, which provide its motive force. Unlike Kant, Velleman does not distinguish a distinctive motive for practical Reason. In his goal of rejecting a substantial motive for practical reasons, Velleman (Reference Velleman2007, 44 & 90) in effect takes what Kant would describe as the motive of theoretical Reason to guide all deliberation, both practical and theoretical.

It may seem odd to ascribe a motive to theoretical reason. Isn’t motivation the province of practical reasons? However, to Kant, any movement of the mind requires motives that drive it. When these are external to the deliberative faculty, they are mere heteronomous influences that force change upon us, but where we cannot self-consciously acknowledge the authority of their influence. However, when the motives are internal to the self-conscious reflection we engage in, we are only bound by a law that we, from the reflective perspective, ourselves acknowledge the authority of.Footnote ¹³ Consider the following: We may theoretically discern that a set of our beliefs are inconsistent, such as the belief that P, the belief that P entails Q, and the belief that not-Q. However, in this instance, theoretical reason does not merely register this fact and moves on. The very process of theoretical reflection also drives us toward rejecting at least one of the beliefs. In this case, we do not feel alienated by our sudden inclination to change our beliefs. Rather, the pursuit of consistency is driven by a motive internal to the very practice of theoretical deliberation, the pursuit of truth. Thus, we experience ourselves as free, despite experiencing an overwhelming force to change our beliefs in light of the discovery of their joint inconsistency. Thus, like practical reflection, theoretical reflection has an internal motive that drives us to activity. The difference between theoretical and practical reason thus is not that one possesses an inherent motive and the other does not. The difference is that one provides the motivation for the formation of judgment, the other for the engagement in action.Footnote ¹⁴

The pressing question is whether theoretical reflection and practical reflection share a single motive, respectively for judgment and action, or whether each type of reflection possesses a distinct motive of its own. In the following, we argue that because Velleman solely ascribes what becomes a mere formal theoretical motive to the Will, his conception of human free Will becomes vulnerable to the challenge that, from the perspective of self-reflection, it appears arbitrary which actions we reflectively decide to pursue.

Velleman thinks that our motive of self-understanding is connected to our experience of reflective freedom through an inherent connection between my ability to predict what I will do, or what I intend to do in Velleman’s (Reference Velleman2007, 98) terminology, and then those reasons for action that such a prediction creates. From my deliberative perspective, I want to acquire self-understanding. But if I only do what I have predicted myself to be doing, then upon deciding to do something, I will also have reasons to do what I decide to do, as this will allow me to know what I am and will be doing (Velleman, Reference Velleman2007, 87 & 174). Deciding to do something is inherently self-conscious. Thus, by only doing what I have decided to do, I will also be self-conscious of what I am doing. Hence, by the very act of deciding to do A, I also have a reason to do so, as restricting my actions to what I have decided will aid me in my pursuit of self-understanding. For Velleman (Reference Velleman2007, 209), all that is required for something to be a reason for an action is thus that one has decided to perform that action.

While less articulated, Frankfurt’s account has echoes of the same structure as Velleman’s.Footnote ¹⁵ By claiming that non-alienated second-order desires are those with which we have identified or those which we wholeheartedly endorse, Frankfurt insinuates a form of stability in the trajectory of one’s actions. Identification and wholeheartedness are terms that suggest a form of persistency and consistency. If someone is wildly inconsistent or fluctuating in their aims, we would find it hard to accredit them with a wholehearted pursuit of their goals. Where their accounts differ is that for Velleman this persistency and consistency is causally brought about by our practical deliberation when such deliberation governs our actions. In Frankfurt’s accounts, we are given no clear explanation of what is different between a person who passively and accidently finds themselves pursuing a remarkably persistent and consistent course of action and someone who does so based on their active identification with these principles. Is the identification merely a form of conscious avowal with no influence on action, and if so, why does it matter to reflective freedom. Or does it change the way in which our agency functions so as to shore up this persistence and consistency in our courses of action?

Velleman’s (Reference Velleman2007, 58 & 173) account underscores the causally efficacious role that our practical deliberation serves in governing our actions. Hence, his theory does not fall prey to the charge of mere window-dressing a fundamentally Humean account where we, as deliberative beings, are mere hostages to empirically discovered motive states. However, his theory remains problematic due to his rejection of the Kantian idea that practical reasons are inherently reasons for thinking an action good. On Velleman’s account, what separates reasons for action from mere contemplated proposals is that the former in some way present us with ways of fulfilling our aims. What separates this from the Humean account is that our reasons for action are actively formed when we decide to do something, rather than passively discovered. In deciding to do A, I also acquire a motivating reason to do so, as my reflective motive of acquiring self-understanding will be aided by restricting myself to doing that which I have decided to do. The problem with Velleman’s account, however, is that in his story, the active formation of such a decision cannot help but seem arbitrary from our reflective deliberative perspective. Some of our reasons are surely due to our decisions; had I simply made a different decision, then my landscape of reasons would be equally altered. This is what is at stake when the act of promising generates new reasons to do what one has promised. What is problematic with Velleman’s account is that for him this seems to be the complete picture of our deliberative reasons. There may be limits to what one can decide to do, in so far as one cannot decide to go against desires that are so strong that one cannot overcome them (Velleman, Reference Velleman2007, 180). Just like one cannot genuinely make a promise one knows one cannot keep. However, as long as the added motive force of actively deciding can sway one’s course, then according to Velleman, one can decide to do anything one can make intelligible to oneself.Footnote ¹⁶ There are no substantive reasons for making certain decisions rather than others, nor are there any unconditional reasons that are independent of one’s existing decisions and merely passively discovered desires. Neither are there any types of substantive reasons that are constitutive of the very practice of agency as many Kantians would claim (Engstrom, Reference Engstrom2009; Korsgaard, Reference Korsgaard and O’Neill1996), beyond the very meek executive requirement that one forms some decisions, whatever they be. On Velleman’s picture, we may, qua reflective deliberators, be required to decide to pursue certain intelligible courses in life, but which course we should commit ourselves to is beyond the scope of rational reflection. In this way, Velleman’s picture shares the radical decisionism involved in Sartre’s conception of free choice as the fully unconditioned. Like Velleman, Sartre thinks that prior to our choice of forming certain intentions and decisions, what we should do is unconditioned.Footnote ¹⁷

In Being and nothingness, Sartre describes the human being as “a being who can realize a nihilating rupture with the world and with himself” (Sartre, Reference Sartre2000, 439). Sartre goes on to claim that “the permanent possibility of this rupture is the same as freedom” (ibid.). Freedom as the permanent possibility of rupture is conceived here as a transgressive force that disrupts an ineradicable tendency to self-deception or, in Sartre’s famous expression, “bad faith” (Sartre, Reference Sartre2000, 47–70). For Sartre, freedom is constantly at stake in human reality, precisely because the human being perpetually tries to refuse to recognize its freedom and act in “refusal of freedom,” which he describes in the following way:

More generally, “refusal of freedom” designates any attempt to invoke necessities either from the outside, in the form of religious or secular authorities or values that are supposed to justify our choices and ends, or from the inside in the form of appeals to motives, beliefs, feelings or desires that are supposed to act as determinate grounds for purposes and actions. Sartre thus concludes that “human reality cannot receive its ends […] either from outside or from so-called inner ‘nature’. It chooses them and by this very choice confers upon them a transcendent existence as the external limit of its projects” (2000, 443). Good faith or authenticity amounts to acknowledging that freedom is thus the possibility of a rupture that undermines appeals to any necessity as the determining ground of choices. Freedom is in the final analysis expressed in the rupture, which avoids invocations of any necessities to ground human action.Footnote ¹⁸

Like Sartre, Velleman (Reference Velleman2015a) is forced to acknowledge that there can never be any requirement qua freedom that we should pursue any particular action over any other. That something would be good to do does not automatically grant it the status as a reason for one’s action. If one has committed oneself to the pursuit of evil, then that something is good to do would be a reason to refrain from doing it. For both Velleman and Sartre, there is from the perspective of practical self-reflection nothing beyond the very fact of one’s decision or commitment that accounts for somethings being a reason for one. Sure, unlike Satre, Velleman argues that one cannot decide to do anything one imagines, as certain of our passively discovered desires are too strong to overcome. In this sense, Velleman acknowledges further constraints on choice than Sartre. But notice that these are in fact mere empirical constraints akin to the fact that one cannot simply decide to fly without the ability to do so. Likewise, I cannot according to Velleman decide to do something that goes against an overwhelming desire. But within the realm of adjudication between possible actions, the realm which practical self-reflection governs, one cannot find any substantial grounds as to why one should form one decision rather than another. We can acknowledge the causal efficacy of our Willing in determining the shape our lives, in so far as it us who set our commitments and make our decisions, which in turn determines what reasons we have. Yet, from the reflective perspective of practical deliberation, this picture undermines the very reflective importance of those reasons, as they cannot help but seem arbitrary.

4. The Need for a Necessitating Determining Ground for a Decision

The problem for Velleman and Sartre is that mere awareness that we adjudicate between which desires we allow to influence our choice is insufficient for a reflective experience of freedom. Such an experience moreover requires that we can make it intelligible to ourselves why we choose as we do, while the choice simultaneously remains a product of the Will’s influence. What these theories disregard is the demand of our practical self-consciousness for what Kant (Reference Kant and Gregor2015 [1788], 5:105) calls “a determining ground (Bestimmungsgrund) with respect to its causality.” Kant’s (Reference Kant and Gregor2015 [1788], 5:132) claim is that it is inherent to our practical self-consciousness that we conceive of ourselves as intelligible causes, not merely as unfettered causes. To experience ourselves as reflectively free, we must conceive of ourselves as causes which can account for why they do as they do, without this explanation forcing us to abandon our conception of the Will as the sufficient cause of the action. The accounts of Velleman and Sartre each in their own way render such explanations impossible.

In merely providing the Will with an executive rather than substantive motive, we are left with two inadequate accounts of why we make the choices we do. In Sartre’s account, we are simply left with no intelligible explanation at all, as any reference to a determining ground of our choice will according to Sartre be a case of bad faith where we present our choice as necessitated. Hence, to Sartre free choice must from the perspective of Will appear as arbitrary, but that conception of choice makes us lose sight of our reflective freedom itself. We here ask the reader to consider whether whatever importance they attributed to free Will is not reflectively undermined by a conception where arbitrariness of choice is the inevitable outcome of freely willing our actions.

Unlike Sartre, Kant does not see reflective freedom as obstructed by reasons necessitating our choices, it rather requires it. As he writes: “The necessitation by motives is not opposed to freedom, but the necessitation by stimuli is wholly repugnant to it.” (Kant Metaphysics L: 28:256). The important point for Kant is that this necessitation cannot come from forces external to the Will, which is why reflective freedom requires the ascription of a necessitating motive internal to the Will. According to Velleman, the Will does have an internal motive which can drive our choices, namely, self-understanding. However, as this is a mere executive motive, we can only in a deficient sense make intelligible the choice of A over B in terms of the contribution of the Will. We might be able to causally explain why we managed to perform A over B in terms of the causal influence of the Will. But we cannot reflectively make sense of why we choose A over B in terms of the influence of the Will, as the Will is only governed by the motive that we should choose, not with what we choose.

On Velleman’s (Reference Velleman2007, 176) account, the only intelligible explanation of why we choose A over B is to refer back to the initial strengths of our desires for A and B, combined with certain executive limits on choice which allow for prediction of behavior, such as consistency and simplicity, and so forth. Hence, when faced with our experience of adjudicating between A and B, Velleman returns to the Humean picture, where such choices are made intelligible in terms of the relative strength of the desires; desires we simply empirically discover within ourselves. But in that sense, we have lost the very idea of reflective freedom, which precisely centered on the idea of non-alienness, which Kant expresses as the idea that the Will “must necessarily be able at the same time to agree (einstimmen) to that to which it is to subject itself” (Kant, Reference Kant and Gregor2015 [1788], 5:132). According to Velleman, the only conditions internal to the Will which restrict such agreement is if a choice is deemed to lack the mere executive virtues that allow for self-understanding. However, such mere executive virtues might be possessed by several competing actions. Notice that the choices in question are not mere quaint choices, such as choosing between two well-respected careers. A life in the pursuit of evil and a life in the pursuit of good are, to Velleman (Reference Velleman2015a, 98), equally viable choices with respect to the motive of the will, as long as both of them are consistently pursued. The theory that such fundamental choices can only be explained by recourse to accounts such as, I simply found myself more attracted to evil, just reinstates the worry that from a self-conscious perspective we experience ourselves as hostages to the forces of our empirical psychology.

Kant (Reference Kant and Gregor2015 [1788], 5: 33) explains autonomous actions as those that are performed as the conclusion of a process of practical deliberation that is self-sustaining. It is self-sustaining in the sense that one’s choice of an action is fully motivated by one’s self-consciousness that this choice satisfies the internal motive of the deliberative practice one is engaged in. Our earlier reflections show that this motive of the Will cannot simply be a form of executive motive. Rather, the Will must have a substantive motive of its own, such that our choices can be necessitated, and thus intelligibly explained as nonarbitrary, by reference to reasons whose governing force is internal to the Will (Kant, Reference Kant, Ameriks and Naragon1997b [1790–91], 28: 587–588).

In the case of theoretical reasoning, the substantive requirement on judgment is truth, in contrast to mere executive requirements such as consistency. The term for self-conscious satisfaction of this norm of theoretical deliberation is knowledge of the truth of one’s judgments.Footnote ¹⁹ If one judges based on knowledge of the truth of one’s judgment, then one’s judgment is necessitated simply by demands internal to our theoretical deliberative practice. The equivalent substantive requirement on action is goodness, where mere executive norms, such as efficiency and consistency, fall short of a full account of the motive of the Will. One’s action is self-consciously necessitated from one’s reflective perspective when one’s choice of an action is fully motivated by one’s knowledge that it is good (Kant, Reference Kant and Gregor2015 [1788], 5: 59–60; Korsgaard, Reference Korsgaard2008, 214). In such cases, the Will is the sufficient cause of my action, as it is the substantive motive inherent to my Will that reflectively necessitates my action, thereby making its choice intelligible to myself as caused solely by my Will.Footnote ²⁰ As Kant (Reference Kant, Ameriks and Naragon1997a [1770s], 28: 258) writes: “Were we also at the same time subjectively necessitated, then we would still be just as free, because this subjective necessitation arises from the objective one. We are subjectively necessitated by a condition, because the action is objectively Good.” It is crucial for Kant that the choice is fully motivated by its satisfaction of the internal motive of the Will. Our cognition of the action as good must be a necessitating and thus sufficient cause of the choice of the action, rather than being seen as a mere necessary requirement and thus an insufficient cause of the choice of action. Kant (Reference Kant, Gregor and Korsgaard1998, 4:397) is adamant that we can act heteronomously even when we do what is good because we appreciate it as good. This occurs whenever the internal motive of the will, that is the action’s goodness, is not our incentive for doing what we do. In Kant’s example, a shopkeeper may be honest because he knows honesty is the good thing to be, but crucially he adopts honesty as his practice because acting like good people do is advantageous for his long-term business. Such a shopkeeper still acts heteronomously as his deliberation is, if not fully then at least partly, under the sway of a merely empirically discovered desire for success; a desire whose authority over his choices he has not managed to justify from within the self-conscious perspective of his practical deliberation.Footnote ²¹

In contrast, if I decide to φ solely because I have, through self-conscious deliberation, come to know that φ’ing is the good thing to do, then no authority outside my practical deliberation is sustaining my choice. My self-conscious knowledge of the goodness of φ’ing is sufficient to support the action and is simultaneously the necessitating cause of the action; I do it because I know it to be good (Engstrom, Reference Engstrom2009, 69). Crucially, at this stage, the term “good” is not loaded with moral content. It is simply a placeholder for an action satisfying the substantive internal motive of practical reasoning (Kant, Reference Kant and Gregor2015 [1788], 5: 63). As Korsgaard (Reference Korsgaard2008, 214) explains it, when acting in this sense “you are motivated by the awareness that you have appropriate grounds for action.” However, if we follow the Kantian reasoning to its end, it is clear that this internal motive of the Will must in some sense be an objective motive shared by all free rational practical deliberators. If our choices are to be free from influences alien to our self-conscious deliberation, then the internal motive of the Will must also be independent of my merely empirical psychology which figures as alien from the self-conscious deliberative perspective.Footnote ²² This means it is a motive one must have simply qua being a rational practical deliberator. Kant expresses this as follows:

To which degree this unempirical motive of the Will accords with the specific Kantian conception of the good as encapsulated in the categorical imperative is beyond the scope of this paper.Footnote ²⁴ But what does follow from the issues discussed is that all reflectively free deliberators must share a common substantial motive internal to the Will, just like all theoretical deliberators share the substantial internal motive of judging truly. Thus, when we state that the motive of the Will is the good, we minimally commit ourselves to the Will possessing a substantive objective motive shared by all rational deliberators.

Crucially, this explanation of reflective freedom does not entail that one can never explain one’s action as reflectively free while justifying one’s choice by referencing the strength of an existing desire. Part of taking our own worth seriously is to acknowledge that we also deserve happiness (Kant, Reference Kant and Gregor2015, 5:110). Hence, the satisfaction of our own desires also counts as part of the good. In certain cases, doing the good will then be to satisfy one’s desires, and in that case, other things equal, one might as well satisfy the strongest of these. However, in this case, one’s Will is still the intelligible cause of one’s action, as one has reflectively determined the pursuit of one’s happiness to be what is good to do. In Kantian terminology, there are two different ways in which our desires can be represented to ourselves as motives within our practical deliberation. Either as mere incentives (Triebfeder) when they appear as empirical givens or as interests when they appear as reasons whose relevance is considered and determined from within the perspective of self-conscious practical deliberation. When they are further understood as good to pursue, they are moral interests. According to Kant, “the moral interest is a pure sense-free interest of practical reason alone” (Kant, Reference Kant and Gregor2015, 5:79). When understood as a moral interest, one’s action is not experienced as an empirically determined by incentives simply discovered in one’s psychology, it is rather understood as sufficiently and intelligibly caused by the Will, and thus reflectively free.

5. Knowledge of the Good and the Intransparency of Reflective Freedom

As mentioned, this paper will not defend the particular Kantian conception of the good. We merely aim to establish that we can only experience ourselves as reflectively free if we attribute a substantive good as the objective motive of the Will. This good must be substantive in the sense that it takes into account the content of the actions one performs, and not merely their formal properties such as coherence, simplicity, and consistency. However, we do wish to briefly discuss how the idea of freedom as acting out of knowledge of the good is reconcilable with our epistemic fallibility and accords with how we explain previous failures in practical deliberation.

In light of our earlier treatment of Watson, one might argue that reflective freedom simply requires that one acts in accordance with one’s fallible judgments as to what is good, not in accordance with one’s knowledge. Admittedly, it will at any given moment be impossible for the deliberating subject to distinguish what is truly good from what one judges to be good. Given the internal aim of judgment is truth, one can only judge what one takes to be true (Shah & Velleman, Reference Shah and Velleman2005). However, when adopting a diachronic perspective, we can see that upon realizing that a given value judgment of ours fell short of knowledge, we simultaneously experience those actions it occasioned as reflectively unfree. The moment one realizes that a given value judgment falls short of knowledge, one is simultaneously compelled to explain the making of that judgment in terms of the effects of influences external to one’s deliberative perspective.Footnote ²⁵ Any judgment that falls short of knowledge will, from the deliberative perspective, be based on a deficiency of reasons. Hence, influences external to one’s reasons must at least partially account for one’s making of the judgment. This can either be the force of psychologically internal influences, such as desires and habits, or of psychologically external influences, such as indoctrination, social norms, upbringing, and so on.

However, in realizing that I only judged φ’ing good because of the influence of say, my sheltered upbringing, I cannot help but feel reflectively unfree to the degree that I can no longer treat what occasioned my judgment as a genuine reason. Hence, I must acknowledge that something external to my Will was at least a partial nondeliberative cause of my ensuing action. Hence, while we may not, in the moment, be able to distinguish the necessitating force of alien influences from the necessitating force of our practical reasons, subsequent evaluations of our freedom disclose that genuine reflective freedom requires that our actions are caused by our knowledge of the good, not simply by our current opinions as to what is good. Our fallibility in discerning the good does not mean that reflective freedom can make do with less than knowledge as its determining ground. It simply means that we do not at every moment possess transparent insight into whether an action of ours is fully an exercise of our capacity for freedom. It is always possible that we come to discover that what we took to be an autonomous action was to some degree a case of succumbing to alien influences.