On Act-Types and Act-Tokens: or, the Distinction between Practical Reason and Practical Judgment in Kant

Introduction

It is a truism that ethical theories aim to tell us how to act, that is, to tell us what actions are right or wrong. It is also a truism that practical reason is the capacity to determine what we should do. And yet both these formulations are ambiguous concerning the degree of generality or specificity with which they fulfill these functions. Do they only tell us about the kinds of actions that we should perform, or do they also issue in concrete actions? If they are only capable of the former, then one might worry about their ability to be action-guiding in the fullest sense. Indeed, such a concern has motivated critiques of principle-based ethics like Kant’s, such as those posed by particularists (Dancy 2004). Since principles are inherently general, the argument goes, they do not spell out exactly what we are to do in every situation.

Defenders of Kantian ethics as an ethics of principles (most prominently, Onora O’Neill and Barbara Herman) have appealed to judgment: rather than assuming principles themselves can provide us with complete answers if they are to be efficacious, or that the Categorical Imperative provides us with an algorithmic decision procedure, we should instead recognize the ineliminable function of judgment in moral agency. And yet despite this emphasis on the essential role of judgment, these commentators espouse what is arguably a quite limited conception of practical judgment—one whose primary tasks are the derivation of duties and the formation of maxims. On these accounts, practical judgment is not concerned with picking out the concrete actions we should perform. Its primary object is thus the act-type, not the act-token.

While there is a vast literature on Kant’s conception of practical reason, there is comparatively little on his conception of practical judgment. In addition, commentators have yet to take up the question of what the distinction between practical reason and practical judgment amounts to, for Kant. My aim in this paper is to provide an answer by bringing the type/token distinction to bear on Kant’s theory of moral agency. I argue that practical reason is concerned with determining the moral status of act-types, while practical judgment is concerned with picking out some act-token that instantiates it.

In this, I wish to highlight the role of the power of judgment [Urteilskraft] as a faculty that assists reason in its efforts to be practical. In its determining use, it subsumes particulars under universals, that is, applies general rules to concrete cases. But I will also suggest that the activity of reflection, for which it is also responsible, has a role to play in practical judgment. What I hope all of this will reveal is that an answer to the question, ‘Is pure reason capable of being practical?’ must acknowledge the essential role of judgment in the determination of the human will. Indeed, as I will somewhat provocatively claim, for creatures like us, reason cannot be practical without the help of judgment.

My discussion proceeds as follows. In section 1, I motivate the problem of the distinction between practical reason and practical judgment in Kant. I show how commentators not only tend to conflate practical reason and practical judgment, but have neglected to raise the question of what the distinction amounts to. In section 2, I step back in order to discuss Kant’s conception of the faculties of judgment and reason in general, as well as their practical uses. While practical reason is the capacity to derive an action from an abstract moral principle, practical judgment is the capacity to determine whether an action falls under such a principle. What this points to is a certain ambiguity in the notions of an ‘act’ and ‘action’, which I discuss in section 3. I claim that we can clear this up by appealing to the distinction between an act-type and an act-token: practical reason is concerned with act-types, while practical judgment is concerned with act-tokens. I show that commentators focus almost exclusively on the moment of moral agency in which we determine the moral status of an act-type, neglecting the subsequent moment in which we pick out an act-token that would instantiate that type. Yet it is this moment, I go on to argue, which most deserves to be called ‘practical judgment’. In section 4, I consider the interplay of practical reason and practical judgment from two different angles: first, in terms of the practical syllogism, and second, in terms of reflecting judgment. The former provides us with a ‘top-down’ model in which actions are justified from principles. On the latter, we get a ‘bottom-up’ approach—one that I argue accords better with the phenomenology of moral agency. Rather than deducing an action (act-type) from the Categorical Imperative, we reflect on an action (act-token) via its maxim. I conclude, in section 5, by making explicit the essential role that judgment plays in the overall process of practical reasoning. An upshot of this account, I contend, is that it provides us with a compelling response to the charge of rigorism: that certain types of actions are intrinsically right or wrong does not entail that there is no flexibility for practical judgment in determining a token act that instantiates it. On the contrary, it is precisely because there is more than one act-token (in most cases) that judgment is needed.

1. Practical Judgment: An ‘Aspect’ of Practical Reason?

In both the secondary literature on Kant and contemporary normative ethics more generally, the notion of practical reason has received far more attention than that of practical judgment. In addition, almost any time the notion of judgment is invoked in the context of discussions of practical reasoning, no sharp distinction between the respective activities of judgment and reason is drawn. Because the issue of the distinction and relation between practical reason and practical judgment tends not to be raised, it allows for a conflation of these two notions. While likely unintentional, it is nonetheless unfortunate, for it suggests that there is nothing too specific that needs to be said about the role of judgment in moral agency. This is all the worse for our understanding of Kant’s ethics, as it suggests that his sophisticated theory of judgment has nothing to say here.

The term ‘practical reason’ tends to be used as an umbrella term for anything pertaining to moral agency, deliberation, exercises of the will, and instrumental (means-ends) reasoning. Hill’s definition is representative: ‘Practical reasoning, in general, is reasoning about what I ought to do’ (1989: 363). Such a broad conception of practical reason raises the question of what room there is for practical judgment, and, by extension, what the difference between the two might be. And yet we find almost identical language tends to be used to define both practical judgment (the activity ‘by which we determine what we are supposed to do’, Longuenesse 2005: 236) and ethical theory itself (a good ethical theory should show us ‘how to determine which acts a person ought or might do’, O’Neill 2014: 39).¹ One might think, with O’Neill, that practical judgment is simply an ‘aspect’ of practical reason—and that there is no sharp distinction to be drawn between the two (2018: 84; cf. 109, 121).²

None of this should be surprising, since Kant is guilty of using terms inconsistently in his Critical faculty psychology (Wuerth 2014). In addition, Kant says little about practical judgment compared with practical reason.³ Despite the centrality of the notion of judgment for his Critical philosophy, he gives far less attention to it in his practical philosophy than in either his theoretical philosophy or his aesthetics. In the Critique of Practical Reason, he devotes only a few pages to the topic—in a section entitled ‘Typic of the Pure Practical Power Judgment’ (‘Typic’ hereafter; KpV 5:67–71).⁴ As we will see in the next section, the power of judgment is the faculty of thinking the particular under the universal. To be sure, there is a sense in which all our higher cognitive faculties ultimately consist in judging. As early as the pre-critical period, Kant refers to both the understanding [Verstand] and reason [Vernunft] as a faculty or capacity to judge [Vermögen zu urteilen] (FS 2:59).⁵ Moreover, in both the Groundwork and the second Critique, Kant characterizes reason as something that judges (G 4:404; KpV 5:16, 75, 78, 93, 159). Yet in the Typic, Kant explicitly invokes the role of the power of judgment [Urteilskraft] as something distinct from practical reason and concerned with applying the laws of the latter (KpV 5: 68–69; cf. 5:160).⁶ In other words, it seems as if Kant conceives of both practical reason and the power of judgment as things that judge, but also conceives of both in a narrower and more specific sense. In what follows, I propose that we conceive of ‘practical reason’ and ‘practical judgment’ similarly to the way the term ‘understanding’ is defined in the first Critique—both broadly, as referring to all three of our higher cognitive faculties in general, and narrowly, as one of those three faculties. Similarly, I will suggest, we can distinguish practical reason in a narrow sense from practical judgment—seeing both these activities as jointly constituting practical reason in the broad sense.

2. Reason and Judgment in Kant’s Faculty Psychology

To get clearer on what practical judgment is and how it might be related to practical reason, we can briefly step back and consider Kant’s conception of the faculties of reason and judgment more generally in order to better understand their operation in the practical domain.

The faculty of reason [Vernunft] receives a variety of definitions throughout the critical philosophy—most notably, as ‘the faculty of principles’ (KrV A299/B356).⁷ It is also ‘the faculty for the determination of the particular through the general (for the derivation from principles)’, and, similarly, ‘the faculty of deriving the particular from the universal and thus of representing it according to principles’ (EEKU 20:201; Anth 7:199). Reason allows us to say something new about a particular case based on a general rule, or to arrive at something unknown on the basis of what is known, and is defined as ‘the faculty of inferring’ (LL 24:693). Accordingly, Kant associates it closely with the syllogism:

[J]udging mediately (through the subsumption of a condition of a possible judgment under the condition of something given). The given judgment is the universal rule (major premise). The subsumption of the condition of another possible judgment under the condition of the rule is the minor premise. The actual judgment that expresses the assertion of the rule in the subsumed case is the conclusion. The rule says something universal under a certain condition. Now in a case that comes before us the condition of the rule obtains. Thus what is valid universally under that condition is also to be regarded as valid in the case before us (which carries this condition with it). (KrV A330/B387; cf. FS 2:59)

Kant’s most extensive discussion of the syllogism comes in the first Critique’s Transcendental Dialectic, particularly, in a section concerning what is referred to as the ‘logical’ use of reason, in contrast with its ‘real’ use (KrV A299/B355). Here Kant considers the exercise of reason in abstraction from the content of this or that cognition, focusing on its inherently inferential function. What is most notable about Kant’s remarks on syllogism in this section is that each of the three parts of a syllogism are associated with one of the three higher cognitive faculties, with each faculty being chiefly responsible for one of the parts. First, the understanding provides the major premise, which Kant also calls the ‘rule’ (e.g., ‘All human beings are mortal’). Second, the power of judgment provides the minor premise, in which a particular cognition is subsumed under the condition of the rule (e.g., ‘Socrates is a human being’). Finally, reason draws a conclusion about the particular on the basis of the general rule (e.g., ‘Socrates is mortal’) (KrV A304/B360; cf. Refl 16:99).

In the first Critique, Kant defines the power of judgment [Urteilskraft] as ‘the faculty of subsuming under rules, i.e., of determining whether something stands under a given rule or not’ (KrV A133/B172).⁸ Since rules are inherently general, which is to say, they can be applied to more than one case, there must be a separate faculty responsible for recognizing when a rule applies in a given case. A gap always remains between the generality of a rule and the particularity of a case, which can never be closed by rules themselves. One might grasp a rule, but fail to apply it correctly. Kant uses the example of doctors and lawyers who possess relevant theoretical knowledge (say, of anatomy or a legal code) but are unable to apply it to actual cases. A doctor may possess the concept ‘typhoid’ but be incapable of diagnosing a patient who has it; a lawyer may grasp the difference between ‘homicide’ and ‘manslaughter’ but be unable to discern which one applies to the defendant before him. Moreover, the task of distinguishing whether something stands under a certain rule cannot be governed by a rule, on pain of regress: this rule would require another rule ad infinitum. Kant claims that the power of judgment stops this regress, calling it ‘a special talent that cannot be taught but only practiced’; a ‘sharpened’ power of judgment is a skill that can only be acquired through experience (ibid.).

Kant provides a similar definition of the power of judgment in the third Critique: ‘the faculty for thinking of the particular as contained under the universal’ (KU 5:179; cf. EEKU 20:211). However, Kant introduces a distinction here between two uses of this faculty, which he refers to as ‘determining’ and ‘reflecting’. The distinction hinges on whether a universal is given. When it is, the task of the power of judgment is to subsume a particular under it. This is determining judgment. We can take this to involve something like predication, that is, attributing a property to a thing. For example, I might possess the concepts ‘red’ and ‘coffee mug’, and thus say of some object in front of me that it is a red coffee mug. However, if no universal is given, then we must search for one under which the particular could be placed or thought. To continue the example: the first time I saw a coffee mug, I lacked the concept necessary to see it as a coffee mug. It was only after reflecting on the particular object as such (and, presumably, other coffee mugs) that I arrived at the empirical concept ‘coffee mug’. In seeking out a universal for the particular, the power of judgment is reflecting.

We can now consider the activity of both these faculties in the practical domain. We must note that the term ‘practical’ means something specific for Kant. It is the domain where the higher cognitive faculty of reason legislates the moral law for the will (KpV 5:197–98; EEKU 20:245–46). Kant defines the will as the ‘faculty either of producing objects corresponding to representations or of determining itself to effect such objects…that is, of determining its causality’ (KpV 5:15; cf. KpV 5:9n). To have a will is to be able to bring about an object that one desires by representing it to oneself prior to its existing (MM 6:213).

If practical reason receives anything like a definition, it is with Kant’s famous equation of it with the will. What is instructive, for our purposes, is his rationale for this identity claim: namely, because ‘reason is required for the derivation of actions from laws’ (G 4:412). While the will is a capacity to act according to principles, reason is the faculty of principles. Practical principles, for Kant, are—as he says in the opening lines of the second Critique—’propositions that contain a general determination of the will’ (KpV 5:19). Chief among these are ‘maxims’ (subjective principles of volition). While there is extensive debate among commentators about the nature of maxims, which I cannot address here, the only thing that is relevant to note, for the purpose of my argument, is that maxims, as kinds of practical principles, are general (i.e., applicable to more than one case)—no matter how specific their formulation.⁹

When we turn to practical judgment, we should note that Kant defines the power of judgment ‘in general’ as a faculty for bringing particulars under universals. This being the case, we should want to know whether there is anything distinctive about practical judgment, or whether it is simply the exercise of a more general capacity for theoretical judgment. We will be in a better position to address this matter in the next section. For now, we can consider the specific activity performed by the power of judgment, which Kant treats (albeit briefly) in the Typic section. There, Kant describes ‘practical judgment’ [praktische Urteilskraft] as a matter of determining ‘whether an action possible for us in sensibility is or is not a case that stands under the rule’, and thus the act ‘by which what is said in the rule universally (in abstracto) is applied to an action in concreto’ (KpV 5:67). In addition, he describes practical judgment as the ‘subsumption of an action possible to me in the sensible world under a pure practical law’ (KpV 5:68). What Kant means by ‘possible action’ is not immediately clear, though most commentators have interpreted this in general terms—i.e., as descriptions of possible types of actions, expressed in maxims and other practical principles. This is the subject of the next section.

3. Types and Tokens in Kant’s Theory of Action

We can now consider the ambiguity I have been referring to, one that is inherent to the notion of an action. The words ‘act’ and ‘action’ can be taken in at least two senses. The first sense refers to act-types—that is, categories or kinds of actions. For example, ‘flying to Paris’, ‘running a mile’, and ‘telling a lie’ are all act-types; they are general, and thus applicable to more than one thing. The second sense refers to act-tokens, which are specific or concrete actions. Kahn defines an act-token as ‘a particular, concrete action performed by a particular, concrete individual in particular, concrete circumstances’ (2023: 575). For example, ‘taking an Air France flight from New York to Paris, which departed at 7:05 PM last Tuesday’, ‘the mile I ran this morning on the treadmill’, and ‘telling my grandmother a lie last Christmas when she asked if I liked the sweater she bought me’ are all act-tokens.

One might be initially skeptical as to whether bringing the type/token distinction to bear on Kant’s theory of moral agency is fruitful.¹⁰ Since O’Neill invokes the type/token distinction in her work, I will make use of it in what follows. For O’Neill, practical judgment is future-oriented and concerned with enacting principles. This being the case, she claims that ‘we do not “have” a particular act—an act-token—until the deed is done…and then the practical problem is over’ (2018: 91fn9). In this, she rejects as incoherent the idea of a ‘possible’ act-token. Likewise, she describes practical reason as what agents use in order to ‘shape’ and ‘guide’ future action:

Since practical reason has to bear on action yet to be done, it cannot bear on act tokens: there are no relevant, individuable act tokens at the time that practical reasoning takes place. So practical reasoning has to bear on act types… [i.e.,] to provide reasons for thinking that certain types of action…are required or forbidden, recommended or inadvisable. (2004: 94)

We can draw two conclusions from O’Neill’s remarks. First, O’Neill does not seem compelled to draw a clear distinction between practical reason and practical judgment. Rather, these terms appear to be used interchangeably in her work. At most, as we saw above, O’Neill will sometimes suggest that practical judgment is simply an ‘aspect’ of practical reason—but it is not obvious what this means. Second, O’Neill does not see either practical reason or practical judgment as concerned with act-tokens. Recall also her statement that ethical theory is concerned with determining ‘which acts’ we ought to do. We can now reread this with the type/token distinction in mind. O’Neill seems to see both practical reason and practical judgment as focused on act-types and thus as operating at the level of maxims.

If O’Neill is right, then this means that the Typic is only concerned with the question of how we determine the moral possibility of act-types.¹¹ Nuzzo expresses a similar view, writing that, in the Typic: ‘practical judgment remains at a level of generality that is insufficient to decide the problem of the law’s application to particular empirical actions’ (2014: 254). Yet if these commentators are correct, then it is hard to see how the Typic is about practical judgment at all. That is, it would entail that Kant leaves unanswered the question of how we determine which specific actions to perform—that is to say, how we pick out act-tokens that instantiate the act-types whose moral possibility we have established.¹² Yet when Kant defines practical judgment in the Typic, he speaks both of ‘cases’ and actions ‘in concreto’. This, I think, suggests that he has act-tokens in mind. Insofar as act-types are general kinds of representations, they are not capable of serving as a particular in a judgment.

What I want to suggest is that it is, in fact, the move from the act-type to the act-token that is the primary focus of the Typic—and that this is what characterizes practical judgment. Relatedly, the prior move—namely, from the Categorical Imperative to a maxim—might be helpfully called ‘practical reason’ strictly speaking (or in a narrow sense, as distinguished from practical judgment). This will allow us to begin to appreciate the essential role that judgment plays in the exercise of practical reason. As already noted, Kant famously identifies the will with practical reason ‘[because] reason is required for the derivation of actions from laws’ (G 4:412). We can now reread this passage with the ambiguity concerning the notion of an ‘action’ in mind. What I now want to suggest is that we can understand the distinction between practical reason and practical judgment in terms of the distinction between an act-type and an act-token. In short, practical reason tells us that a certain type, or kind, of action is good, moving from the moral law to a maxim (from the concept of duty ‘as such’ to a principle of duty), while practical judgment tells us which act-token would instantiate this type, moving from a maxim to a concrete action.

Kant points to such a division of labor between reason and judgment in the practical domain when he says that:

[T]he law can prescribe only the maxim of actions, not actions themselves; this is a sign that it leaves a playroom (latitudo) for free choice in following (complying with) the law, that is, that the law cannot specify precisely in what way one is to act and how much one is to do. (MM 6:390)

What is particularly striking is that Kant then connects this latitude with the need for judgment: ‘Ethics…unavoidably leads to questions that call upon judgment to decide how a maxim is to be applied in particular cases’ (MM 6:411). In other words, while being the source of a priori laws, reason alone cannot tell us which act-tokens to perform. Rather, this is the task of practical judgment—without which, we would be stuck at the level of general representations (that is, act-types).

What emerges, then, are two distinct moments of moral agency. Most commentators have focused on the first, in which we derive specific principles of duty from the Categorical Imperative via the Formula of the Universal Law. The outcome of the universalizability test concerns the moral possibility of certain kinds of actions, or act-types.¹³ To the extent that the notion of judgment is discussed in such contexts, the suggestion is that this is what practical judgment is. The further implication seems to be that once we have determined the permissibility of some act-type, or formulated and adopted a maxim, there is nothing left to be done (or, if there is, there is nothing philosophically interesting to say about it). But without an account of the second moment, a gap remains.¹⁴ For the move from the supreme principle of morality to a specific principle of duty, or from the FUL to a maxim, is ultimately just a move from a general principle to a more specific but nonetheless still general principle. But we should want to know what we do once we have a maxim in hand.

To get clearer on this, recall that maxims are general determinations of the will—no matter how specific they are. They say: ‘When in C-type circumstances, do A-type actions’. Insofar as they contain act descriptions, then, these are of act-types—not act-tokens. Allison puts the point nicely: they are ‘rules dictating action types rather than particular actions’, and so ‘general with respect to the number of possible…(actions) falling under them’ (1990: 90). ‘Accordingly’, he continues,

there are always…a number of distinct ways in which an agent can act upon a maxim…. This indeterminacy leaves scope for practical judgment, both in deciding whether acting on the maxim is appropriate in given circumstances and in determining how best to carry out the general policy in a particular situation. (Ibid.)¹⁵

Acting on a maxim, then, is not a matter of simply being committed to a given maxim. I may have resolutely adopted a maxim, but not know what specific course of action to take. In this, I would be like the doctors and lawyers that Kant speaks of in the first Critique, who grasp the rules in their generality but cannot apply them to concrete cases. Such individuals, Kant says, lack judgment. In other words, there is a practical analogue to the theoretical case—both of which stem from an inability to move from the general to the particular. As moral agents, we must pick out an act-token that instantiates the act-type. And it is this that I take practical judgment to consist in.

Having distinguished two moments in moral agency, and highlighted the latter (practical judgment) as the move from an act-token to an act-type, we can turn to the issue discussed earlier concerning whether the activity of judgment here is distinctively practical. In light of the similarities just noted, one might wonder whether what is ultimately going on in so-called practical judgment is that practical reason, after settling on a maxim, draws on the resources of theoretical judgment to pick out a concrete act-token. Call this the ‘deflationary view’. On this view, practical judgment is not fundamentally different from theoretical judgment—or is even just an exercise of theoretical judgment as applied to practical matters. This being the case, the first and second moments (which I have called ‘practical reason’ and ‘practical judgment’, respectively) would be represented as follows: the first moment, where one derives and adopts a maxim, is the genuinely practical (or moral) moment; the second moment, which occurs once we have the relevant moral principle in hand, is a matter of theoretical judgment stepping in to pick out a concrete action (act-token) that falls under it. This view would emphasize the common function of judgment to subsume particulars under universals, seeing the only major difference between theoretical and practical as pertaining to the nature of the particulars and universals in question (intuitions and concepts, in the former case; actions and principles, in the latter).

If the deflationary view is correct, then my overall claim in this paper—that practical reason depends on practical judgment—might seem underwhelming. Indeed, this would not appear to be a philosophically interesting conception of practical judgment at all, and it might explain why commentators such as Herman and O’Neill have not had much to say about it. To make matters worse, Kant’s overall lack of attention to practical judgment (evidenced by the notoriously short Typic section) is prima facie evidence for such a view.

However, there are some strong reasons in favor of a more robust view of practical judgment—one that is distinctively practical and irreducible to theoretical judgment. This view would resist assimilating the activity of practical judgment to the theoretical domain and, likewise, reject seeing the first moment (practical reason in the narrow sense) as the only distinctively moral moment.

An implication of the deflationary view is that the exercise of practical judgment is ultimately just instrumental reason, that is, identifying the means to a given end. Yet Kant, by the third Critique, distinguishes the ‘technically’ practical from the ‘morally’ practical—and importantly says that only the latter truly deserves to be called ‘practical’ (KpV 5:171–73; cf. EEKU 20:195–201). Of the former, he says that it is reducible to theoretical considerations. It includes what Kant calls ‘rules of skill’ and ‘counsels of prudence’—imperatives whose ‘ought’-ness is grounded in the fact that ‘[w]hoever wills the end also wills…[the] means’ (G 4:416–417). The technically practical merely says what actions are required in order to bring about a certain effect (KpV 5:26n). When I set certain ends for myself, the specific actions that I ought to do to achieve these ends are entirely derivable from theoretical philosophy (e.g., empirical knowledge, or facts about the world). What separates the technically and morally practical is the determining ground of the will; in the former, it is a concept of nature, rather than a concept of freedom. This division maps onto the better-known distinction between hypothetical and categorical imperatives. While certain judgements might seem to be practical by being about action, what matters is whether their principle is a law of nature or a law of freedom. The proper way to individuate judgements, according to Kant, is by their principles. Thus, Kant’s distinction between the technically and morally practical entails that (i) the deflationary view is true for the technically practical, which, in turn, entails that (ii) the robust view must be true for the morally practical.

This does not yet tell us what it means for practical judgment to be distinctively practical. To be sure, this is a much larger question that I cannot take up fully in this paper. Nonetheless, a few initial observations are in order. First, insofar as judgment is a skill that requires practice (as Kant famously describes it in the first Critique), possession of this skill in the theoretical domain (say, by being a good diagnostician) is not transferable to the practical domain; it does not make one good at picking out act-tokens. (We could extend this point to the aesthetic sphere, as well.) A ‘sharpened’ and mature power of judgment requires experience, exposure to a myriad of cases, and so on. The kind of experience, along with the relevance of the cases, matters a great deal with respect to honing the skill in question. Cultivating one’s capacity for practical judgment takes place independently of any expertise one may have in the theoretical domain. Yet if the deflationary reading were correct, it would suggest that a doctor’s ability to expertly judge whether a patient has pneumonia has some bearing on the ability to expertly judge whether some speech act is a lie (or whether a painting is beautiful). There is, of course, no reason to suppose that this is the case.

Second, while the general skill (and structure) of judgment—that of connecting particulars and universals—is the same in all forms of judgment, the particulars and universals are importantly different. Theoretical judgment operates on pre-existing objects (particulars that are, in Kant’s terms, ‘given’), whereas practical judgment must conceive of an action that is yet to be performed (it must, in some sense, ‘give’ the particular). Rather than tracking the world, as is the case with theoretical judgment, practical judgment is about shaping the world. That is, we are not simply responding to a state of affairs and trying to classify it, but conceiving of many possibilities and determining which one to realize. Thus, the act of subsumption in the practical case is, by extension, distinctively practical. And, as we will now see, it is also something that requires an act of reflection.

4. Bottom-Up versus Top-Down: Reflection and the Practical Syllogism

I have distinguished practical reason (understood in the narrow sense) from practical judgment by appealing to the distinction between act-types and act-tokens. Practical reason concerns itself with act-types, while practical judgment concerns itself with act-tokens. I have also suggested that we can think of these two moments of moral agency as jointly constituting practical reason overall (i.e., in the broad sense). Central to both these processes is the notion of a maxim. Practical reason arrives at maxims, while practical judgment seeks to apply maxims to concrete situations. We can now consider what it means to act on a maxim under two different models, both of which illustrate the distinct roles of reason and judgment, respectively. The first is in terms of the practical syllogism, while the second is in terms of reflective judgment.

While syllogism enjoys a special closeness to reason, Kant also conceives of a syllogism itself as a kind of judgment—namely, a mediate, rather than an immediate, one (FS 2:59). We can represent practical judgment syllogistically insofar as it is a kind of determining judgment, that is, one in which a universal is given. In any syllogism the major premise states a general rule (namely, that a certain predicate applies to anything that meets a certain condition), while the minor premise states that this condition in fact applies to a certain something. As we noted when considering Kant’s remarks on the syllogism in general, each of the three parts is associated with one of the three higher cognitive faculties. It is the power of judgment that is responsible for the minor premise.

Kant characterizes the practical syllogism as:

proceeding from the universal in the major premise (the moral principle), through undertaking in a minor premise a subsumption of possible actions (as good or evil) under the former, to the conclusion, namely, the subjective determination of the will (an interest in the practically possible good and in the maxim based on it). (KpV 5:90; cf. MM 6:313)

We can unpack this further in light of what we have said so far. First, the major premise is a practical principle of some kind. Recall, again, Kant’s definition of practical principles as ‘propositions that contain a general determination of will’ (KpV 5:19). Kant subsequently notes that there are two main kinds of practical principles: maxims and laws, distinguished in terms of their subjectivity and objectivity, respectively. Practical principles are ‘subjective, or maxims, when the condition is regarded by the subject as binding only for his will’ (ibid.). A ‘maxim’ is thus a subjective principle of action; it is the rule according to which I act. Practical principles ‘objective, or practical laws, when the condition is cognized as objective, that is, as holding for the will of every rational being’ (ibid.). As an objective principle of action, laws specify the rule according to which I ought to act—in virtue of being the rule according to which all rational beings ought to act. Practical rules, Kant says, are ‘always a product of reason’ (KpV 5:20). A rule of reason is called an ‘imperative’ insofar as it contains an ‘ought’; it is binding on wills that are not subjectively necessitated to be determined by this principle alone (cf. G 4:421n).¹⁶

The minor premise is the subsumption of a possible action (i.e., an act-token) under the concept of good or evil. This is an act of the power of judgment. To have a concept of an object of practical reason is to represent some object as the kind of thing one could bring about through the will (KpV 5:57). Kant contends that there are only two concepts of pure practical reason: ‘The only objects of a practical reason are therefore those of the good and the evil’ (KpV 5:58). These concepts function as the condition of the rule stated in the major premise (a rule defining what is good ‘in general’). In the minor premise, we say of some object that we could bring about through our will—and thus some action that we could perform—that it is morally good or not.

The conclusion states whether I should bring about the action. Having stated a general moral principle (in the major premise) and judged that a specific action is good or evil (in the minor premise), we can then infer whether the action is morally required. This concrete imperative should not be confused with the imperative of the major premise. Insofar as a practical principle is a general determination of the will, it cannot tell us what to do in a specific situation. As the conclusion of a practical syllogism, it would be quite unhelpful. Rather, we are specifying what we ought to do, as Kant says, ‘in the case at hand’ (MM 6:313). I take the goodness of the action, judged in the minor premise, as a reason for judging that the action should be performed. Hence, the conclusion of a practical syllogism must be something that could serve as the determining ground of my will. To summarize: in a practical syllogism, the major premise is a rule concerning act-types, while the minor premise is the subsumption of an act-token under the condition of this rule. The conclusion states whether I ought to perform this specific act-token.

We can now begin to see the essential contribution that the power of judgment makes to a syllogism—and, more generally, appreciate the way in which it assists reason in its efforts to be practical. One could not deduce anything from a general rule alone, nor from a single premise. Using the above example: one could not simply state that all human beings are mortal and then immediately conclude that Socrates is mortal. This is because one has not subsumed ‘Socrates’ under the concept ‘human being’ (the condition of the rule)—the task performed by the power of judgment (cf. LL 9:120). Only following this subsumption can reason make use of the general rule in order to make the inference that ‘Socrates’ also goes with ‘mortal’. Similarly, one could not merely posit a categorical imperative (or some abstract moral principle) and then, merely on the basis of this, conclude what one ought to do in one’s present situation. Simply possessing the adequate rule or maxim does not allow one to deduce what one ought to do in any given case. Here we need only to remember again Kant’s remark in the first Critique about those who grasp a rule but are unable to apply it. Merely supplying the major premise is insufficient; an act of judgment is needed.

While considering practical judgment in terms of the syllogism helps to illuminate the distinctive role of the power of judgment, it might also give the impression that the process is somewhat mechanical, or quasi-deductive in nature. One might even wonder how this way of thinking about practical reasoning in general leaves any room for the genuine exercise of judgment. However, the fact that we can represent the process of practical reasoning syllogistically does not mean that our moral deliberations take place by merely plugging things in and deducing what we should do. Indeed, if this were the whole story, then this would seem to rule out the possibility of moral dilemmas and interpersonal moral disagreement. This worry dissolves when we observe that it conflates the major and the minor premises—what one’s duty is, on the one hand, with how one is to fulfill one’s duty, on the other. This is precisely why it was important to introduce the distinction between act-types and act-tokens: it is easy to think of cases where two people are equally committed to acting on a moral rule yet disagree about which specific action to take in service of that rule. This is because the act of judgment, by which we move from the rule in the major premise to the conclusion about what specifically we ought to do, is a ‘special talent’ that requires practice—a skill that is ’sharpened’ by experience.

Highlighting this distinction also provides a compelling answer to the charge of rigorism—that is, the idea that Kant’s emphasis on universal moral principles allows for no flexibility whatsoever in our decision-making. Availing ourselves of the distinction between act-types and act-tokens, we can say that: (i) certain types of ‘actions required and prohibited by moral rules are right and wrong intrinsically’ (Schapiro 2006: 35; my emphasis), but that this does entail that (ii) there is always exactly one token act that I must perform. Some act-types are obligatory or prohibited; but once we have determined this, there is still work left to do. We must then exercise judgment to pick out an act-token that instantiates the act-type. So, while there may not be flexibility at the level of the rule or maxim (which concerns act-types), there will certainly be flexibility at the level of the concrete action taken in service of the rule.¹⁷ That lying is always wrong (and that there is no leeway whatsoever here, for Kant) does not mean that the specific action I must take in order not to lie will be apparent or uncontroversial in every case.

An account of practical judgment in terms of the practical syllogism, then, must be supplemented by a deeper look into the reflective dimension of the power of judgment. In contrast with what we might call this ‘top-down’ account (by which we descend from the universal to the particular), we can also consider a ‘bottom-up’ account—one that involves reflection (ascending from the particular towards the universal).¹⁸

Most accounts of practical judgment that invoke reflection (Kantian or otherwise) see it as a perception of the particularities of a situation. For example, Herman provides a Kantian account of ‘rules of moral salience’, which help us pick out features of a situation that require moral attention (1993: 78–98). McDowell provides a more Aristotelian conception of deliberation as ‘a capacity to read the details of situations in the light of a way of valuing actions’ (1996: 23). For both, reflection is a matter of reflecting on the specific circumstances one finds oneself in with an eye toward their morally relevant features. To be sure, understanding one’s context is undoubtedly an important aspect of moral agency that bears on practical judgment; moral reasoning that failed to make reference to this would remain at the level of act-types. However, this does not explain the precise sense in which practical judgment is reflective, for Kant. While we can incorporate aspects of the background conditions of an action into our description of an act-token that we consider performing, we cannot subsume it under a moral rule.

The notion of reflection appears in a variety of contexts across Kant’s critical philosophy.¹⁹ Kant defines ‘reflection’ as follows: ‘To reflect…is to compare and to hold together given representations either with others or with one’s faculty of cognition’ (EEKU 20:211). It is worth noting that Kant provides this definition of reflection immediately after he distinguishes between the determining and reflecting uses of the power of judgment. In other words, reflection is an act of the power of judgment. What’s more, as early as the first Critique’s Amphiboly section, Kant states that all judgments require reflection (KrV A261/B317).²⁰ The structure of the activity of reflection is that of holding up and comparing various representations. While aesthetic judgment is paradigmatic of reflection insofar as the activity of ‘mere’ reflection occurs in the absence of a universal, reflection also occurs when a universal is given. In the former, we hold up and compare representations provided by the free imagination to the understanding, which does not provide a concept but only a general demand for unity and lawfulness. In the latter case, we hold up and compare representations of particulars and universals. That is, reflection is required to determine whether the particular ought to be subsumed under it. Consequently (and contra Herman and McDowell), the relevant particular in a practical judgment is the possible action itself. As should be clear by now, possible actions in this context should be understood as an act-token. It is this upon which we reflect, and which we seek to subsume under a universal in a practical judgment.²¹

Highlighting the role of reflection in practical judgment lends itself to a bottom-up, rather than a top-down, account of moral deliberation. Such an account, I think, accords with the phenomenology of ordinary moral agency, and functions as a supplement to an account of the structure of moral reasoning that is primarily concerned with the justification of moral principles and the derivation of duties. Consider the familiar story about Kantian ethics—that we go about determining whether an action is morally obligatory, permissible, or prohibited by testing the maxim of the action to see whether it is universalizable or not. But this can take place in one of two ways. On the top-down account, we begin by testing the maxim to see if it passes the universality test. If it does, we then proceed to determining which specific action to perform. On this way of conceiving of practical judgment (as embodied in the practical syllogism), we start with the act-type and the question of its moral possibility—after which we pick out some act-token that instantiates it.

However, more often than not, we instead begin by conceiving of a relatively specific action that we wish to perform, or propose to undertake. That is, we start with an act-token in mind—and reflect on it in order to find some maxim under which it could fall (or which could capture it), which is to say, what act-type it instantiates. In this, we ascend from the particular to the general—rather than descend from the general to the particular. Once we have ascended to a maxim, we then reflect further: on whether this maxim could be universalized. That is, we reflect on the maxim in light of the Categorical Imperative in order to determine whether the maxim is valid. In this way, the Categorical Imperative functions as a second-order principle, the role of which is to assess maxims, or act-types.²² If this is correct, then it means that the universalizability test is not simply a logical test, but a reflective test as well.

We in fact find Kant using such language in the second Critique. The supplying of a ‘type’ of the moral law is described as the ‘comparison of the maxim of [one’s] actions with a universal law of nature’ (KpV 5:69; emphasis mine). Moreover, Kant says that in practical judgment, ‘reason…always holds the maxim of the will in an action up to the pure will, that is, to itself in as much as it regards itself as a priori practical’ (KpV 5:32, cf. 5:354; emphasis mine). We can apply this language to the moment of practical judgment: we hold up a representation of an act-token to a representation of an act-type to whether it belongs under it. We hold up a possible action to a maxim to see whether the former could be said to instantiate the latter.

5. Why Practical Reason Needs Practical Judgment

Recall that the conclusion of a practical syllogism is something that can then serve as the determining ground of the will. ‘Every determination of choice [Willkür]’, Kant says, ‘proceeds from the representation of a possible action to the deed [Tat]’ (MM 6:399). We can contrast the ‘general’ determination of the will that characterizes Kant’s definition of a maxim (and which is present in the major premise of the practical syllogism) with a specific determination of the will. We now know that it is the power of judgment which deals with possible actions. And here Kant tells us that this act of representing a possible action—that is, an act-token—is necessary for fully determining the will. That is, it is only through an act of the power of judgment that we get from a general moral rule to a subjective interest in it. It is through practical judgment, as Kant says in the Groundwork, that we provide the moral law with ‘access’ to our will—bringing ‘an idea of reason’, he says, ‘closer to intuition…and thereby to feeling’ (4:436).²³

This is not to say that there is not a kind of determination of the will that occurs insofar as I make it my maxim to do such-and-such type of action. But it cannot explain how it is that we come to carry out concrete actions. The lack of attention to this aspect of moral agency gives the unfortunate impression, I think, that all the morally salient work happens at the level of the formulation and adoption of maxims—and that what happens after is both philosophically uninteresting and morally irrelevant. But we can observe the difference between my vague commitment to not doing A-type actions (say, lying) and my concrete determination not to do some particular act-token (e.g., telling my mother that I like the sweater, even though I do not). As I said above, acting on a maxim is not simply a matter of being committed to the maxim. Yet now we can also say that I might not even be able to fully commit to a maxim if I do not know what specific actions (act-tokens) it would require me to perform. Accordingly, the will can only be determined in the fullest sense once we have a fairly specific sense of what it is that we ought to do. I might say that I will always tell the truth, but then shy away from doing so once I find out what it looks like in a particular instance. As we have seen, judgment in general is a skill that is honed, in part, by exposure to a multitude of cases. This is no less true of it in its practical activity. If all of this is correct, then it is not just the maxim (and thus the act-type) which is the object of moral assessment—but also the act-token. And, moreover, the ability to select a particular act-token that instantiates a moral rule is not a morally neutral moment.

In conclusion, consider the following passage from the Groundwork, which I think embodies the view I have been articulating:

All moral philosophy rests entirely on its pure part and…gives [us] laws a priori; which of course still require a power of judgment sharpened by experience, partly to distinguish in what cases they are applicable, partly to obtain for them access to the will of the human being for performance, since he, as himself affected by so many inclinations, is indeed capable of the idea of a practical pure reason, but not so easily able to make it effective in concreto in the conduct of his life. (4:389)

While reason is the source of a priori laws, which borrow nothing from experience, Kant ascribes to the power of judgment the twofold task of applying these laws to concrete cases and assisting reason in exerting influence on the will. Kant’s pure ethics is concerned with determining and justifying the Categorical Imperative as the supreme principle of morality. The Categorical Imperative gets described by Kant as the ‘standard’, the ‘canon’, and the ‘foundation’ of our judgments; it is, he says, the ‘guideline and supreme norm by which to judge [actions] correctly’ (G 4:390, 403–4, 424). Nothing I have said here threatens this. And yet once the content and validity of the moral law have been secured, we must then be able to apply it concretely and allow it to motivate the will. While reason legislates for the will, it cannot apply its law on its own. Just as there can be no rules for the application of rules ad infinitum, there can be no principles for the application of principles. For this the power of judgment is needed. Judgment is necessary because the human will ‘does not straightaway do an action just because it is good’ (G 4:414). And so, for creatures like us, judgment is necessary for pure reason to be practical.

Acknowledgements

Versions of this article were presented at Johannes Gutenberg-Universität Mainz, Universität Greifswald, and the University of South Carolina. I am grateful to these audiences for their questions and discussions. I am especially thankful to Jens Timmermann, Thomas Pendlebury, and Keren Gorodeisky for reading drafts of this paper and providing valuable feedback. Thanks also to the anonymous referees from this journal for their helpful comments and suggestions.

Competing Interests

The author has no competing interests to declare.