Many philosophers have thought Lewis Carroll’s wonderful fable “What the Tortoise Said to Achilles” teaches us an important lesson.¹ What that lesson is, is less clear. Here I will argue that the fable has important implications for the relationship between constructivism and particularism in an adequate theory of practical rationality. By “a theory of practical rationality” I mean an account that contributes to our self-understanding as creatures that are capable of responding to normative demands in what we do. Such a theory could and should contribute, not just to our understanding of ourselves, but to our understanding of the normative demands of rationality itself.

Constructivists have (rightly, in my view) drawn on insights like those the fable seems intended to convey to maintain that we ought not to see ourselves as standing primarily in an epistemic relation to normative facts. But this, I argue, is only part of what the fable suggests. If we understand ourselves as normatively sensitive, practically rational agents by focusing on general or formal principles – as neo-Kantian constructivist accounts do – we will not have digested its lesson. It’s not that general principles have no place in our lives as rational agents. It is that that place reflects only one of the elements of normative sensitivity that a theory of practical rationality needs to encompass. Constructivism points us to something different, and something more, in understanding our natures as reasoners.

It may be surprising that my target is what we are like as practical reasoners, while Carroll’s example is of a bit of theoretical reasoning. While the absence of a sharp demarcation between theoretical and practical reasoning is beyond what I can argue for in this essay, I do think the lesson of the fable carries straight across: it is about the nature of the norms governing our reasoning, full stop.² Perhaps this is because theoretical reasoning really is a species of practical reasoning. (It is, after all, reasoning about what to do, albeit doing something about our beliefs.) I am inclined to this view, but even if it is mistaken, theoretical reasoning certainly models important features of practical reasoning. In any event, for present purposes I will move back and forth between the theoretical and practical domains of rationality as though there were no significant boundary between them.

1 Lessons from the Tortoise: received view

Let’s begin with Carroll’s story itself. The Tortoise invites Achilles to consider a simple bit of deductive reasoning. From two premises, A and B, which Achilles accepts, the Tortoise wonders how a skeptical reasoner might reason to conclusion Z, which (Achilles believes) follows logically from A and B. I’m going to model the reasoning in question a bit differently from Carroll, to make its structure as modus ponens obvious:

$$

The Tortoise innocently asks whether, and Achilles foolishly agrees that, one might be under no obligation to accept Z unless one grants the truth of a further premise (call it C), which licenses Z as a conclusion from the two premises in I:

$$

So we now have a new argument:

$$

But of course poor Achilles is now compelled to grant that one need not accept II as valid unless one accepts the truth of a further premise D, which licenses Z as a conclusion from the three premises in II, and the race is on. Obviously, it is one Achilles cannot win.

Now, the Received View on what we should learn from this fable (one I do not quarrel with) is that we cannot treat inference rules (such as modus ponens) merely as premises in arguments.³ Why? Well, we must do something with the premises (namely, infer), and inference rules are norms for that thing we must do. If we try to capture the work those norms do as premises, we are left without the inference rules we need to license our inferring. No matter how many premises we accept, we cannot reach a conclusion so long as we insist on treating inference rules as premises.

We can put this point another way. Seeing that modus ponens is valid is different from reasoning with modus ponens. They are distinct cognitive operations. If we reason from (P → Q) and P to Q under the normative license of modus ponens, then because we recognize that we reason validly when we do so, we can write (and know to be true):

$$

But we must not make the mistake of thinking that we get the rule we have just written down first, before we recognize that we can reason in the way it represents. One might think: “Well, it’s simple enough, once one knows that MP is a valid inference rule, to apply it to particular cases.” While that is true, if we think that we may or must apply it correctly because we know it is a valid inference rule, we walk right into the trap the Tortoise laid for Achilles. If we are licensed in inferring in virtue of our knowledge of inference rules, we are stuck with only premises and no way to do anything with them. It must be because we may validly reason in the way that MP represents that MP is valid, not vice versa.

2 Lessons from the Tortoise: constructivism

Kantian constructivism is motivated by this insight. By “constructivism” here I mean views modeled on the kind of ontological outlook John Rawls defended for principles of justice. The contrast as Rawls has it is (roughly) over our understanding of the ontology of normative facts: views that deny that such facts exist “prior to and independent of” our engagement with them are constructivist.⁴ Views which hold that such facts are there prior to and independent of our engagement with them, in some sense awaiting our recognition, are (we might say) recognitionalist.

Constructivists diagnose Achilles’ mistake as one of taking the relevant norm (modus ponens, in this case) as an object of knowledge. That’s what his writing down of the norms as premises reflects – they are objects of knowledge that he now recognizes. If Achilles thinks that the primary relation he has with modus ponens is as a Recognizer of its validity, he is handcuffed when he comes to the movement from premises to conclusion. These are, again, distinct cognitive operations. Achilles will be left without normative guidance in doing what he must do with those premises, namely engaging in an inference necessary to reach a conclusion. Because it is an activity, something he can engage in in better ways and worse – not merely a causal process – that normative direction is necessary. The crucial point is that (i) guiding our movement from the premises to the conclusion, as a matter of (valid) inference, on the one hand, and (ii) knowing that the proposition stating that that movement is valid is true, on the other, are distinct, and we must not confuse them. Recognizing the validity of modus ponens is different from reasoning with it.

Christine Korsgaard especially has developed this point in ways that don’t trade directly on Carroll’s argument.⁵ She has, for example, used an analogy with maps to argue that the recognitionalist’s way of thinking about practical norms reflects “a deep confusion between knowledge and action”:

Now, her application of this lesson is to moral reasoning, but we can see straightforward application of her thought to Carroll’s case. Consider her point about morality:

We might straightforwardly translate that to a claim about what Achilles needs:

Grasping a bit of knowledge is one thing, and doing something with it is another.⁸ The distinction is endemic to normativity. It arises any place we engage in activity that is normatively guided, or which is subject to normative assessment.⁹ In some of her most recent work, Korsgaard observes that “what Philosophy wants is always some piece of knowledge,” while pointing out that norms of practical rationality cannot do the work we need them to do if we take them to be bits of knowledge, or premises in arguments.¹⁰

3 Lessons from the Tortoise: multiple tasks for norms

All that is a critical part of the case for constructivism, as opposed to the view on which our primary relationship with norms is as Recognizers of their truth or authority. But the example shows us something more, which is that there are two different ways we need norms for reasoning (practical or otherwise). To put it a better way: there are (at least) two different tasks we need norms for reasoning to perform. These correspond to two distinct standpoints or perspectives we can take in reasoning. Which of these we occupy at a given moment makes all the difference as to what we need from norms at that moment. One is the critical standpoint, from which we can see that modus ponens (say) is truth-preserving and thus a worthy standard for correct inference. This standpoint is one that can be occupied by an observer or other judge third-personally assessing what we do. (That includes us ourselves acting as observers, critics, or judges of what we do.) The other is a deliberative standpoint, from which we first-personally must determine what to do.¹¹

How is this lesson implicit in Carroll’s example? Modus ponens is a general principle that we can (rightly) Recognize, from the critical perspective, to be valid and to fix a standard for correctness in deductive reasoning. But that much is compatible with seeing that we are entitled – in the critical standpoint – to write down MP as a characterization of what we (rightly) do from the standpoint of deliberation. As we have seen, however, that does not give Achilles what he needs. What he needs to do is to conclude Z, and the crucial thing is that he see that concluding Z is the thing to do, not that he see anything about MP. The former is something he must undertake from the deliberative standpoint, while the latter is something either he or anyone else can undertake from the critical standpoint. MP, as a rule we can characterize from the critical perspective, does him no more good in the deliberative standpoint than does adducing further premises reflecting its implementation. It does not advance Achilles’ project one bit.

4 Kant and the Tortoise’s lesson

What serves our rational purposes from the critical perspective will not do so from the deliberative perspective. That is an important lesson. But there is more to understand about why not, and we need to understand that something more if we want to take our deliberative need seriously. This is not an entirely novel lesson, but it is one that has not been sufficiently taken to heart by contemporary constructivists.

Kant in fact was quite concerned to account for judgment in what I am calling the deliberative standpoint, and his exposition of what we now recognize as a general problem is a model of lucidity (even if the context in which it occurs is not). His first engagement with the issue occurs in his first Critique attempt to motivate and secure for “transcendental logic” the capacity to arrive at synthetic a priori judgments.¹² In characterizing “transcendental logic” – central to the project of the first Critique – Kant considers the capacity for “general logic,” construed as a body of rules. Judgment, he says, is a necessary tool for such rules; it consists in the “faculty of subsuming under rules.”¹³ He observes that judgment so construed cannot itself be rule-governed, since the application of any proffered rule could itself be undertaken only under the guidance of another rule, which itself could be undertaken only under the guidance of some further rule, etc. In this way he anticipates just the problem the Tortoise poses for Achilles. The resulting regress leads Kant to claim that judgment “is a peculiar talent which can be practiced only, and cannot be taught.”¹⁴

In the third Critique, Kant actually taxonomizes the varieties of this “talent,” and here he himself is pellucid that there is more than one kind of cognitive operation in normative response. The work of judgment in bringing particulars under universals he calls “reflective”; finding a particular that renders a universal determinate he calls “determinative” judgment.¹⁵ This distinction may not get much traction in practice. We may think the problem Achilles faces is knowing under which universal rule to bring his grasp of the particular premises he has accepted; that would make his problem one of “reflective” judgment. But we might equally well suppose his problem is determining whether the universal rule MP can be made determinate in a way which would yield his conclusion. Either way, however, Kant grasps the distinct natures of these operations.

Now, Kant thinks this problem can be solved for “transcendental logic,” which involves “the correcting and securing of judgment” in the “pure understanding.”¹⁶ The solution in the first Critique rests with the “transcendental schemata,” which are mediating representations that are “homogeneous” with both the rule (the “category”) and the particular (the “appearance”).¹⁷ I cannot claim to understand Kant’s proposal as to how the “schematism” actually solves the problem; indeed these elements of his Transcendental Analytic involve some of the most difficult of problems in Kantian interpretation.¹⁸ To be specific, I do not know how Kant’s proposal here would address the conundrum faced by Achilles.

At any rate Kant recognizes that norms for judgment aren’t just a concern for theoretical reasoning, as is his concern in the first Critique. However, his only explicit attention to the deliberative problem in his practical philosophy lies in the “Typic of Pure Practical Judgment” in the second Critique. But the view he offers there is that the norm for practical judgment just is the categorical imperative as a test for maxims.¹⁹ As we have seen, this cannot be the entire story, if it is right at all. “Pure” moral philosophy consists in laws (in particular the categorical imperative) which may direct a will grounded not in any elements of our natures, but “in concepts of pure reason.”²⁰ They must hold “not only for human beings but for all rational beings as such.”²¹ Kant’s own work on general logic should have induced him to recognize that this is going to leave a task for judgment, the nature of which he has left unexplored. Indeed, Kant admits, such laws “require a judgment sharpened by experience, partly to distinguish in what cases they are applicable and partly to provide them with access to the will of the human being and efficacy for his fulfillment of them.”²² Even by Kant’s lights, his second Critique proposal simply sidesteps the crucial difficulty, because the problem for applying principles of pure practical reason to “anthropology” is isomorphic to the problem he has diagnosed for general logic. His account of practical rationality simply does not address what we confront in deliberation.²³

5 Lessons from the Tortoise: moving beyond principles

Kant’s constructivist heirs follow Kant’s own theorizing about practical reasoning in trying to straddle the distinction between (what I am calling) the critical and deliberative standpoints. They have undertaken what I am calling a “straddle” here in various ways. Korsgaard, for example, maintains the Kantian aspiration of showing that reasons must be “universal” – that is, exceptionless, and not merely general – along the lines of the familiar Kantian proposal that principles of practical reason are universal laws. But she recognizes a deliberative problem with this proposal: “every occasion” (she observes) “may be different in relevant ways from the ones we have previously encountered”:²⁴

To think of a principle as provisionally universal is, as Korsgaard has it, to “think it applies to every case of a certain sort, unless there is some good reason why not.”²⁶ She maintains this is only “marginally” different from treating principles as universal tout court, but “essentially” different from treating them as merely general, in a way which is manifest when we encounter exceptions:

Korsgaard is squarely facing a problem any account of deliberative rationality must confront, one which arises from the fact that life is a matter of constant change, sometimes in directions that are not only unforeseen but practically unimaginable. That ensures that we will face deliberative challenges. Korsgaard’s acknowledgement that new deliberative conditions may require us to revise principles is a tacit recognition of the fact that, from the deliberative perspective, the principles codifying our previous maxims are by their very nature inadequate: we must keep returning to them to refine and revise.

Nor is this an isolated or occasional issue: it is endemic to practical reason from the deliberative standpoint generally. Even when we do not need to revise our principles, we must judge that no such revision is necessary. Deciding which cases are which is, of course, a deliberative problem, and that is an issue on which principles give no guidance on at all. Principles tell us nothing about where or when they are to be applied; that is precisely why the recognition paradigm for practical norms cannot be primary. In a word, that is why Korsgaard thinks we must be constructivists.

Given that the roots of Korsgaard’s constructivism are thus found in the deliberative standpoint, this lacuna in her neo-Kantian account is difficult to understand. From the deliberative perspective the “provisionally universal” principles which are ostensibly the centerpiece of Korsgaard’s account are manifestly unhelpful. The deliberative condition requires that we somehow make judgments about particular circumstances and what we have reason to do when we are in them. If our principles are provisional, and we must determine whether or not to revise them based on what reason demands in a particular situation, then those principles cannot be our guide as to what reason demands in that situation. The judgments must in a sense be logically prior to the principles that report them, in just the way that Achilles’ inferences are logically prior to the principle MP that he can write down as a principle of inference. The way in which we are sensitive to normative demands in deliberation cannot be thought to be satisfied by principles, even if their universality is provisional.

Nor is this lacuna filled if we revert to a view on which it is maxims, as subjects of scrutiny for suitability as universal laws, that are thought to mediate between concrete and particular conditions of action and general (or universal) principles. As critics of the Kantian project since Hegel have argued, maxims do not of course come without the exercise of judgment, either. We must come to an understanding of what they are, and that project too is subject to normative demand and assessment. We still have distinct kinds of cognitive operations without distinct rational norms for deliberative guidance.

This is a point that, to her credit, O’Neill has faced squarely. She emphasizes, following Kant, that there are distinct cognitive operations in judgment, and that a significant part of the difficulty in deliberation is posed by the relation between the abstract and the concrete. Not only all language but all thinking depends on abstraction from concrete particulars, as she observes.²⁸ The movement from particular to abstract and back creates exactly the problem we have observed in trying to deploy critical standards to meet deliberative needs; as O’Neill remarks, “Principles are always abstract, but they are not the whole of practical reasoning. They must always be applied in ways that take account of actual context; and they never determine their own applications.”²⁹

This is indeed a problem, and O’Neill is astute to point it out. But her solution to the problem is in effect a counsel of despair. Principles are reflected in maxims, and she believes that a plurality of maxims allows the indeterminacies of the various principles at work to be mutually limiting, narrowing the field of possible determinate applications.³⁰ But how to balance and “trade off” the competing demands of various principles? Here rationality simply stops. O’Neill dismisses “fantasies” that there is some “metric” for doing so,³¹ and quite rightly, because the need to apply such a metric would land us right back in the problem Achilles faced, and which Kant has so clearly accounted for. What then? Well, we throw in the rational towel: “in the end there is always a point at which mere picking has to take place.”³² Now, either the way we balance and resolve the competing normative demands of principles is itself subject to normative scrutiny – either there are better and worse ways of doing so – or it is not. If it is not, since these kinds of conflicts are endemic to human practical rationality and agency, in the end the entire enterprise would rest on a non-rational basis. The notion that we are actually constructing norms for the exercise of practical rationality would involve foundations of sand. If on the other hand there are better and worse ways of balancing competing principled demands, then obviously our constructivism cannot merely be of principles of practical reason. In fact, it is doubtful if principles really can “come first,” as O’Neill insists, though saying just this hardly excludes any useful role for principles in practical rationality.³³ If we believe our lives as practically rational agents are lived in response to norms that are themselves constructions of rational agency, we have to think of those norms in ways that transcend principles.

6 Lessons from the Tortoise: from constructivism to particularism

Let us return to O’Neill’s observation that moving from the abstract and general to the concrete and particular (and back) is of the very essence of practical rationality. If, drawing on a particular experience, I say, “Well, I learned my lesson!”, I am thinking that I have discovered (say) not to X when conditions A, B, and C hold. But that thought has already abstracted away from the innumerable other conditions that held, along with A, B, and C, to warrant the conclusion that X was a bad idea. That abstraction itself requires a work of judgment. But so does going in the opposite direction. For if I want to apply the lesson in conditions where A, B, and C once again hold, I have to be sure not only that they in fact hold, but also that there is no condition D that changes what I ought to do – perhaps making X just the ticket. So I must perform the counterpart act of judgment to the abstraction, in determining that the conditions in the abstraction are all and only the relevant conditions for rendering the general principle concrete. Crucially, both these moves are acts of cognition which I can perform better or worse. The norms which establish what counts as better and what as worse do not come for free with any general principle, nor can any general principle instruct me in how to undertake them.

This is recognizably the same problem Korsgaard faces with her “provisionally universal” principles, and it is also recognizably the root of the problem Kant worries about but does not solve. It is, I believe, a crucial point that follows on the activity of practical reasoning to which constructivism rightly points us. Precisely because the move from abstract and general to concrete and particular requires a cognitive move that cannot be specified by a general principle, the norms to which we respond in deliberation cannot take an abstract and general form. Somehow they must be concrete and particular.

This, I think, is something the Tortoise knew from the start, though it was not obvious that this is so. Achilles, deliberating, is uncertain what is the thing to conclude. Modus ponens cannot by itself tell him what to do. It is a general principle, licensing inferences to conclusions from premises that satisfy certain formal requirements. Those requirements are abstract but must be applied to concrete and particular premises. Are these formal requirements instantiated in the premises Achilles accepts? Now it is obvious to us that they are, but we should not be deceived by the simplicity of the case into thinking that somehow we have a situation in which the rule governing which premises fit the form of modus ponens is somehow self-applying. Trivial as the matter is in Achilles’ case, seeing that his premises fit the formal requirements of modus ponens requires the exercise of judgment just as any subsumption of concrete to abstract does. The fact that this is a qualification on even the work of modus ponens ought to alert us to the presence of a philosophical concern of the first order.

Modus ponens is a simple and intuitive inference rule, but we know we cannot take even it for granted as a deliberative inference rule.³⁴ With sufficient imagination we can use it and other principles of deductive logic to draw inferences from premises that satisfy the formal requirements, yet reach conclusions we know are problematic. Vann McGee gave us a nice example of this sort of problem, premised on the 1980 US Presidential race, in which Republican Ronald Reagan led Democrat Jimmy Carter by a comfortable margin, with Republican John Anderson a distant third. McGee observed that the conditions support accepting the following premises:

1. If a Republican wins the election, then if it’s not Reagan who wins it will be Anderson.

2. A Republican will win the election.

But these premises do not support the conclusion that

1. If it’s not Reagan who wins, it will be Anderson.³⁵

The fact that such counterexamples are quite exceptional, so that modus ponens itself (almost) invariably yields valid inferences, tempts us to suppose that there is nothing to do in governing our inferences other than to possess the knowledge – affirmed in the critical perspective – that modus ponens is an inference rule which is generally truth-preserving. But that is surely a mistake. An appreciation of what these concrete and particular conditions require by way of governing rules is inescapably necessary.

If the problem is severe for modus ponens, it is critical in ordinary practical reasoning, where modus ponens rarely comes into play. If anything, what we use is much closer to “default reasoning” – an analog of the “defeasible general rules” Korsgaard proposes for the Kantian project. Default reasoning differs from classical logic in being non-monotonic. Classical logic is monotonic in this sense: if a set of premises S supports a conclusion C via valid inference, then any superset of S will also support C via valid inference. This is not true of default reasoning. Here is a simple example of the contrast: consider argument III:³⁶

We could formulate III’s first premise as a conditional:

1. (1*) For all x, if x is a bear, then x is dangerous

and with modus ponens have, in classical logic, a valid deductive argument, warranting the truth of (C). But that would be unwise. If the way we ordinarily reason about Teddy and other bears were monotonic (as in classical logic), then adding additional premises – supplementing our knowledge of facts about Teddy – would do nothing to undermine the conclusion that Teddy is dangerous. But suppose we add a new premise:

1. (3) Teddy is old, tame, sick, toothless, declawed, and listless.

Now the truth of (C) patently does not follow. We could, of course, insist that if (3) is true, (1*) cannot be true; that’s a way to preserve the validity of modus ponens. We could also try to refine (1*) in such a way as to leave room for bears like Teddy, in which case we wouldn’t have an argument at all without additional premises to warrant those refinements. But of course it is true that (1) bears are dangerous, so it is not quite clear what we need to do to reconcile how we handle cases like Teddy’s with classical logic.

We need not try to settle the matter here; how to reconcile default reasoning and deductive logic is beyond our purview. Instead, I take two things to follow. First, default reasoning makes transparent the normatively sensitive work done in deliberation – in fact, just the work we are tempted to ignore in Achilles’ case. If we are, from the deliberative perspective, to apply a given rule to a particular case, we must know, inter alia, that there are no special conditions – extra premises – that would undermine the applicability of the rule in question. We must judge that the conditions licensing the application of the rule are exhausted; call this a judgment of closure.³⁷ This is true in both theoretical and more broadly practical contexts. It is easy to miss this from the critical perspective, because in easy cases (say, Achilles’) it seems that no such judgment is required. The simplicity and artificiality of Achilles’ case invites us to ignore the fact that closure is taken for granted.

The second is that most practical reasoning is messy in the way that the Teddy case is, rather than a matter of simply applying rules of classical logic within a framework of closure.³⁸ This is in effect what motivates Korsgaard’s insistence on “provisional” universality; it is a problem endemic to the deliberative standpoint. Korsgaard’s point is that, however many models we consider of the implementation of a rule, we cannot be certain that we have closure; life and constantly changing conditions may confront us with what we must take as a new premise, undercutting the truth of the conclusion we had thought secured.³⁹ We can and in many conditions we might rightly reformulate our rule so as to confine it to cases in which these new conditions do not apply, but the point is that judgment as to whether to regard new cases as confirming the authority of the rule, or as demonstrating that new information can undercut the conclusion we believed the rule warranted, always is necessary.⁴⁰ Until we have a judgment of closure, we cannot reliably apply general principles. But no story about general principles can tell us when we are entitled to such a judgment.

We might advert to Korsgaard’s map analogy here. As Korsgaard argued, as objects of knowledge maps do us no good unless we do something with them, and positing another map for normative guidance in what we do won’t do the trick. We might also add: nor will abstract and general rules. If we want to use the map to figure out where to go, we must bring to it a grasp of what in the concrete and particular conditions we are in matters for our task. And that in turn requires judgment that no abstract and general rules can guide. Constructivists believe that norms for successful practical reasoning are products of what we do in reasoning. So constructivists above all must attend to norms for exercising such judgment.

7 The way forward

Where does that leave us? I think there are important conclusions to be drawn for advancing constructivist understandings of ourselves as practical reasoners.

The first is a general observation about the way that constructivist theory can contribute to thinking about successful practical rationality. Philosophers have a natural grasp of the critical perspective. The abstract tools philosophers deploy are especially useful at the level of, well, the abstract, which is precisely what is needed for critical reflection on our reasoning, practical and otherwise. It is not as useful for what we need as deliberators. When the only tool you have is a hammer, everything looks like a nail, and precisely our capacities to be reflective and informative about the abstract have obscured the need to understand how to cope with the particularity and concreteness of deliberative rationality. What we should acknowledge is that producing more general and abstract theories about successful practical rationality will fail to grasp the living heart of deliberative rationality,

That is a task for which constructivism is very well suited.⁴¹ The acknowledgement that deliberation imposes essential and distinct requirements on our understanding of ourselves as reasoners and of the norms to which we respond in deliberation ought to shape our thinking about practical rationality, and that is something that emphasizing the constructive, rather than recognitional, nature of practical rationality can and should take in its stride.

That in turn means placing the work of judgment at the core of constructivist theory, rather than having judgment be the acknowledged but subordinated stepchild of theories of general principles. The theoretical legacy that does so is Aristotle’s, rather than Kant’s. Aristotle makes practical wisdom central to his account of virtue, and engagement with particulars central to his account of practical wisdom. Practical wisdom, with the other virtues, directs us to apprehend or perceive the normative demand for action in particular cases – under concrete and particular conditions.⁴²On this outlook, the task of practical reason (in conjunction with virtue of character) is to direct us to apprehend the appropriateness of particular actions, and the canon of success in doing so is contributing – both in changing the world and in changing ourselves – to the project of living a good human life. I believe a natural model for understanding how this occurs comes by thinking about fitting responses to conditions of action. Constructivism is ideally suited to incorporate the kinds of cognition that are involved in judgments of fittingness, though the task for making sense of norms of fittingness is challenging on any approach.⁴³

Still, that’s just a characterization of what we do in judgment, rather than a theory of how to exercise judgment. But we should by now have been warned off expecting much more. Moreover, thinking about deliberative rationality along the lines of that model can and should help reshape our conception of the norms of reasoning in the critical perspective in ways that respect principles but do not suppose they are the key to understanding how rational norms are constructed. Spelling that all out is, obviously, a project for another day. What I hope to suggest here is that we would do well to attend to these features of an Aristotelian approach in constructivist thinking about practical rationality, where they are less well appreciated than they are in normative ethics. The motivations for being constructivists should move us to be particularists as well.⁴⁴