1. Introductory Remarks

In the Groundwork of the Metaphysics of Morals, Kant attempts to prove the moral law in some way. But, in the Critique of Practical Reason, he admits that the moral law cannot be proved by any deduction; and he argues instead that the moral law is given as a fact of reason. For example, he writes:

In addition, Kant’s principle of morality is widely taken to be a ‘formal’ principle, for it must apply to every rational being, and so it is bound to be insensitive to the concreteness of human existence in our world. To put the point a little differently, the moral law is supposed to be universal and necessary, and so desires, which are relative and contingent, cannot ground the moral law. Then two questions naturally arise. First, how should we understand Kant’s claim that the moral law is given as a fact of reason? Second, insofar as Kant’s principle of morality is formal, how can it be reconciled with our presumably substantial moral end?

The purpose of this paper is to address the above two pressing questions for Kantians. First, contra the standard view, I will argue that there is a non-dogmatic way of defending Kant’s claim that the moral law is given as a fact of reason. In particular, I will provide a coherentist defence of this claim. And then I will argue that Kant’s principle of morality can be construed as a formal principle in the sense that it provides correctness conditions for moral laws, rather than providing specific moral laws, so that it can be theoretically harmonized with our substantial moral end.

2. The Objective Reality of the Moral Law

In the Groundwork of the Metaphysics of Morals, Kant attempts to derive categorical imperatives by analysis of our concept of moral duty. On his view, it is part of our concept of moral duty that it applies to all rational agents regardless of their desires. But this kind of conceptual analysis does not, by itself, establish a synthetic proposition that there are indeed categorical imperatives. It is one thing to say that Pegasus purports to be a winged horse, but it is another to say that there is such a thing as Pegasus. In a similar vein, it is one thing to say that moral duties purport to be categorical imperatives, but it is another to say that there are indeed categorical imperatives. In this regard, it is noteworthy that we cannot simply rule out the possibility that morality is “a chimerical idea” (see GMS 4:445), and so our moral common sense is not self-vindicating. Therefore, the objective reality of categorical imperatives cannot be established by mere analysis of our concept of moral duty. In addition, on Kant’s view, a categorical imperative cannot be shown by an a posteriori way either; for our empirical intuitions are always subject to the conditions of sensibility, but the normative necessity of a categorical imperative cannot be an object of the senses.

Along these lines, in the Critique of Practical Reason, Kant admits that his early search for a deduction of the moral law was unsuccessful, and that no deduction of the moral law is possible. ^{Footnote 2} For instance, he writes:

Of course, this does not mean that Kant succumbs to moral scepticism. In the Critique of Practical Reason, he argues instead that no deduction of the moral law is necessary, and that the moral law is given as a fact of reason (Faktum der Vernuft). In other words, instead of attempting to prove the objective reality of the moral law, Kant appeals to a fact of reason; on his view, it is a fact of reason that we are conscious of the moral law as the supremely authoritative and regulative law for us. He writes:

The question then is whether the moral law can really be justified as a fact of reason. Regarding this question, it is worth considering whether Kant is a moral realist or a moral constructivist. Unfortunately, however, it is not easy to settle this question because there are not only many different forms of moral realism, but also many different forms of moral constructivism. Nonetheless, the relevant distinction between moral realism and moral constructivism I am concerned about in this paper is roughly the following. According to a strong form of moral realism, there are values that exist prior to and independent of the moral law, and so we can start with an independent conception of value and then defend the moral law by appealing to such independent values. By contrast, moral constructivism denies such independent values that precede and ground the moral law. On this constructivist view, to say that something is (morally) good is to say that it accords with the moral law; accordingly, if there were no moral law, there would be no good things either. Therefore, we must start instead with a conception of practical rationality and then defend the moral law without appealing to independent values. Notably, Barbara Herman, Christine Korsgaard, John Rawls, and Andrews Reath read Kant as such a moral constructivist. ^{Footnote 3} But many Kant commentators such as Karl Ameriks, Paul Guyer, Rae Langton, and Allen Wood disagree. ^{Footnote 4} On their realist views, the constructivists cannot really defend Kant’s clear commitment to the unconditional value of persons just by elucidating the requirements of practical rationality. Thus, unfortunately for Kant, we may say, it is widely considered that his doctrine of the fact of reason fails to provide morality with a rational foundation. For example, Ameriks argues that a disturbing feature of dogmatism remains in Kant’s doctrine of the fact of reason. He writes:

Contra the above view, however, I will argue in the remainder of this paper that there is a non-dogmatic way of defending the moral law along the lines of Kantian constructivism, that is, in a way that does not appeal to an independent order of value that precedes and grounds the moral law. A cautionary note: the main goal of this paper is to show that there is a defensible version of Kantian constructivism that can avoid the aforementioned realist worry; therefore, it is not my intention to make my Kantian account in what follows faithful to Kant’s own view in every respect.

3. A Coherentist Defence of the Moral Law

As pointed out in the previous section, the objective reality of categorical imperatives cannot be established by mere analysis of our concept of moral duty. In what way then can Kant (or a Kantian constructivist) go beyond mere analysis? We can find an important clue from Rawls. He writes:

I agree with Rawls that Kant (or a Kantian constructivist) can offer a coherentist defence of the moral law. But, unfortunately, Rawls did not provide a sufficiently detailed account of how the moral law can be defended along coherentist lines. Thus, in this section, I will develop his coherentist interpretation a bit further to provide a non-dogmatic way of defending the moral law as a fact of reason.

Above all, we can distinguish two types of arguments: theoretical arguments and practical arguments. Theoretical arguments are concerned with what we ought to believe, whereas practical arguments are concerned with what we ought to do. And a moral norm (or the moral law) is concerned with what we ought to do. Thus, as Kant has shown, we cannot justify a moral norm by a theoretical argument. The question then is whether we can justify a moral norm by a practical argument. It is important to notice at this point that our concept of justification is a normative concept. In our epistemic discourse, we evaluate beliefs as justified or not on the basis of epistemic norms. Thus, epistemic (or theoretical) justification depends on epistemic norms, and so, if there were no epistemic norm, then there would be no epistemic justification either. In a similar vein, practical justification depends on practical norms, and so, if there were no practical norm, then there would be no practical justification either. Here we may distinguish three types of practical norms: prudential norms, institutional norms, and unconditional practical norms (i.e., moral norms). Clearly, we cannot derive an unconditional practical norm from a conditional practical norm. Accordingly, we cannot justify a moral norm by a non-moral practical argument that depends on some prudential or institutional norm. In addition, due to the problem of circularity, we cannot justify a moral norm by a practical argument that depends on a moral norm. Thus, we cannot justify a moral norm by a practical argument either. Hence, it follows that we cannot justify the moral law by a theoretical argument, nor can we justify the moral law by a practical argument. ^{Footnote 7} Nonetheless, this does not mean that the moral law is not rationally defensible. As I will argue shortly, there is a different way of defending the moral law. And this different way, as Kant argues, is to appeal to a fact of reason. Let me elaborate.

All living creatures have to deal with some problem or other in their lives. We are no exception. But, unlike mere animals, we are rational agents. On Kant’s view, it is our conception of ourselves that we are rational agents, who have not only theoretical reason but also practical reason; to put it another way, we are rational agents, who can engage in theoretical and practical reasoning in order to decide what to believe and what to do. What then is our distinctive way of dealing with our practical problems as rational agents?

An illuminating way of addressing the above question, I think, is to consider what Thomas Hobbes calls “the state of nature” in which there are no moral laws, and so people are free to do anything they please, even taking the property and lives of others. According to Hobbes, such a state of nature is bound to be the constant state of “war of every man against every man,” and so life in the state of nature would be intolerable. ^{Footnote 8} We may say that in such a state of nature we are governed by the law of the jungle: the stronger prey on the weaker. Instead, we (as rational agents) avoid the state of nature in such a way that we govern our mutual relations by moral laws. For example, if a person harms another person for no reason, he or she ought to be punished for violating a moral law. Accordingly, if we cannot interact with a certain creature in such a way that the mutual relations between the creature and us are governed by moral laws, then we cannot take the creature to be one of us. Only a rational agent who can be subject to moral laws can be one of us. Along these lines, we may say that we are different from mere animals in that our mutual relations are governed by moral laws. In other words, our distinctive rational nature is such that our mutual relations are governed by moral laws.

Let me elaborate a bit further on why all of us are bound by moral laws without exception. Consider an organization in which all members are subject to the regulations of the organization. If one is a member of such an organization, one is bound by its regulations. For example, if a man is a police officer, he is bound by the police regulations. And he can free himself from the normative force of the regulations only by quitting being a policeman. Thus, we may say that one’s activity as a member of an organization commits one to complying with the regulations that define what one should do as a member of the organization. Therefore, insofar as one participates in an organization as a member, one is subject to the regulations of the organization. In a similar vein, we may say that, insofar as one participates in the moral community as a member, one is thereby subject to moral laws. Yet, there is an important difference between being a member of an organization such as the police force and being a member of the moral community. As noted, we are different from mere animals in that we are by nature such rational agents whose mutual relations are governed by moral laws; consequently, it is not due to our desire or choice that we are members of the moral community. In this regard, it is also worth observing that we are born and raised as such social beings in a human society. Therefore, each of us cannot quit being such a rational person. Hence, even if a certain person refuses to accept moral laws, insofar as he interacts with us as a rational person, he is still subject to moral laws. Consequently, if he harms one of us for no reason, he ought to be punished for violating a moral law, regardless of the fact that he refuses to accept moral laws. This is why moral laws apply to all of us without exception.

A cautionary remark is in order here. The line of reasoning presented above should not be understood in a way that overlooks a crucial difference between Kant’s view and Hobbes’s view. On Hobbes’s view, good things (such as self-preservation or the satisfaction of desires) exist prior to and independent of moral laws; and it is rational for us to follow moral laws because it is to our own advantage to live in a society where moral laws are enforced. Accordingly, Hobbes ultimately justifies moral laws (or the social contract) on the ground that we can promote our self-preservation by following moral laws. Therefore, one important implication of his social contract theory is this: if we are no better off than we were in the state of nature, then we no longer have a reason to abide by the contract. By contrast, according to Kant, “the concept of good and evil must not be determined before the moral law … but only … after it and by means of it” (KpV 5:63). Thus, on his view, there is no moral value prior to and independent of moral laws. In this regard, it is noteworthy that terms such as ‘good’ and ‘bad’ are normative terms. And our normative evaluation of an action as either (morally) good or bad cannot be justified without recourse to some relevant norms; for the justification of a normative evaluation implies that there are some relevant normative reasons, which in turn imply some relevant norms. Thus, as Kant argues, there is no moral goodness prior to and independent of moral laws. As a consequence, without appealing to any moral norm, we cannot justify self-preservation as a good thing that we ought to pursue; instead, self-preservation would be nothing more than an animal instinct. Observe that animal instincts are by themselves neither good nor bad. Hence, contrary to Hobbes’s view, we cannot justify moral laws on the ground that we can promote our self-preservation by following moral laws. Instead, we are what we are (in part) because we govern our mutual relations by moral laws. To put the point another way, the source of categorical imperatives lies in the fact that we are by nature such rational agents whose mutual relations are governed by moral laws, rather than in the fact that we can promote self-preservation by following moral laws.

Now it is worth recalling that the objective reality of categorical imperatives cannot be established by mere analysis of our concept of moral duty; for it is one thing to say that moral duties purport to be categorical imperatives, but it is another to say that there are indeed categorical imperatives. The question then is: how can we justify the moral law as a fact of reason? For this question, it is also important to recall Rawls’s claim that Kant holds a coherentist account of how practical reason is authenticated. Let me elaborate on this claim.

For Kant, reason is the highest court of appeal concerning all questions about itself and its principles. Accordingly, reason must answer any questions about facts of reason from its own resources on due reflection. He writes:

To put the point somewhat differently, it is our conceptual framework that provides the norms, criteria, or rules for defending (or criticizing) any claim, and so it is inevitable to address any justification question on the basis of our conceptual framework. And it is the key coherentist insight that we have no way of judging whether a belief is true but on the basis of reasons within our conceptual framework. Therefore, to address the question of whether the moral law is rationally defensible, we have no other way but to rely on our conceptual framework. And, as argued before, on our conception of ourselves, we are by nature such rational persons whose mutual relations are governed by moral laws. In this regard, it is important to consider a key difference between thin conceptions of a person and thick conceptions of a person. Gottfried Wilhelm Leibniz’s idea of a person as a knower and David Hume’s idea of a person as a bundle of desires are two paradigmatic thin conceptions of a person, whereas Kant’s idea of a person as an autonomous rational agent is a thick conception of a person. Rawls explains this difference as follows:

On the aforementioned thin conceptions of a person, one’s nature as a rational person is independent of being bound by the moral law. By contrast, if my arguments so far are on the right track, we are what we are (in part) because, unlike mere animals, we are by nature such rational agents whose mutual relations are governed by moral laws. On this thick conception of a person, being subject to moral laws is part of our nature as rational persons as distinct from mere animals. It is important to observe two things regarding this thick conception of a person. First, on this Kantian constructivist view, morality is ultimately grounded in basic features of the way we conceive of ourselves as rational persons. But the aforementioned thin conceptions of a person are not strong enough to ground morality. In other words, such a thin conception of a person does not provide a sufficient basis on which to dispel the realist worry discussed in the previous section, namely that the constructivists cannot really defend the rational bindingness of moral laws just by elucidating the requirements of practical rationality. Second, the status of this thick conception of ourselves is so fundamental that it forms a basis of our conceptual framework; accordingly, this conception of ourselves is not derived, but is instead a fundamental part of our conceptual framework. Therefore, we can take this conception of ourselves to be a kind of conceptual fact on the ground that we can accept it by an analysis of our concept of ourselves. Notice that this kind of conceptual fact is not such that it can be justified on the basis of a more fundamental rational ground either by theoretical reasoning or by practical reasoning; in other words, this fundamental conceptual fact neither admits nor requires further proof. Instead, it can be defended only by way of a proper analysis of our concepts related to what kind of creatures we are. In this sense, we can take this fundamental conceptual fact as a ‘fact of reason.’

But here we should not overlook the fact that on the coherence theory of justification there are no such things as basic beliefs. As W.V.O. Quine insists, “no statement is immune to revision.” ^{Footnote 10} Our conception of ourselves is no exception. Thus, the aforementioned conceptual fact (or the fact of reason) is not exempt from a demand for justification. In other words, even this fundamental conception of ourselves can, at least in principle, be challenged. Nonetheless, as mentioned before, we have no other way but to justify any claim on the basis of our conceptual framework. Besides, as Robert Brandom argues, we should understand the structure of justification along the default-and-challenge model of justification, because the infinite regress of justification is impossible. ^{Footnote 11} Moreover, as for beliefs that have a default positive justificatory status, we can shift the burden of proof to the challengers. ^{Footnote 12} The aforementioned thick conception of a person is indeed such a case. As noted, the status of this conception of ourselves is so fundamental that it forms a basis of our present conceptual framework. Accordingly, we are entitled to hold this conception of ourselves unless we are given some positive reason to think otherwise. Thus, regarding this conception of ourselves, we can shift the burden of proof to the challenger of the status quo. In other words, until and unless this thick conception of a person is proven to be wrong, we are entitled to take ourselves as such rational agents whose mutual relations are governed by moral laws. ^{Footnote 13} In this regard, it is worth mentioning that on Kant’s view we can defend this fact of reason insofar as we can assure ourselves that there is no proof of its impossibility. He writes:

Besides, according to Ian Proops, we can read Kant’s Faktum in terms of the distinction between the ‘question of fact’ (quid facti) and the ‘question of right’ (quid juris) criminal lawyers used during the 18^th century. On this reading, the question of right in a given case rests on a prior proof of a question of fact, and Kant uses the term ‘Faktum’ in the sense that it “ought to have some probative force—in the way, perhaps, that legal evidence has defeasible, non-demonstrative force.” ^{Footnote 15} My construal of Kant’s fact of reason as a kind of conceptual fact is in line with this interpretation of Proops in that such a conceptual fact can play a justifying role, insofar as we are not given some positive reason to think otherwise. ^{Footnote 16}

Let me sum up this section. Along the coherentist lines presented above, we can defend the aforementioned thick conception of ourselves as a kind of conceptual fact. This conceptual fact then can play the role of a fact of reason in our defence of the moral law. That is, until and unless this fact of reason is proven to be wrong, we are entitled to take ourselves as such rational agents whose mutual relations are governed by moral laws.

4. Our Substantial Moral End and Correctness Conditions for the Moral Law

In this section, I will discuss why our moral end is substantial. And then I will argue that Kant’s principle of morality can be construed as a formal principle in the sense that it provides correctness conditions for moral laws, rather than providing specific moral laws, so that it can be theoretically harmonized with our substantial moral end.

Any rational thinker should follow epistemic norms such as modus ponens in order to achieve our epistemic end, and so all rational beings are normatively bound by epistemic norms. But this does not imply that our epistemic end is not substantial. Our epistemic end is (at least in part) to obtain an adequate picture of the world by having true beliefs and avoiding false beliefs, and so it is bound to be substantial. In a similar vein, we can say that any rational agent ought to follow moral norms (or moral laws) in order to achieve our moral end, and so all rational agents are normatively bound by moral norms. And this does not imply that our moral end is not substantial. What then is our moral end?

We can interpret our moral end both negatively and positively. As argued in the previous section, our distinctive rational nature is such that we avoid falling into the state of nature in a way that our mutual relations are governed by moral laws. Thus, we may say that our moral end is at least not to fall into the state of nature. This is the negative interpretation of our moral end, which provides us with duties of justice. For example, anyone should not harm another person for no reason. But we can also interpret our moral end positively. On its positive interpretation, our moral end is to maximize the common good insofar as it does not conflict with duties of justice. And this positive interpretation provides us with duties of virtue. For example, we should cultivate our own moral perfection and the happiness of others (as permissible ends). ^{Footnote 17} Along these lines, we can say that our moral end is substantial, even under its negative interpretation.

Now that our moral end is substantial, the question arises: how can Kant’s formal principle of morality be reconciled with our substantial moral end? On Kant’s view, one is bound by a material practical principle only under the condition that one has a desire-based end. But the fundamental principle of morality must apply to all of us, and so its normative force should not depend on any desire-based end. Therefore, on his view, only a formal principle can apply with normative necessity to all rational agents. Accordingly, the minimal sense in which Kant’s principle of morality is ‘formal’ is that it is not a material principle. However, if we characterize Kant’s formal principle of morality only in this negative sense, we cannot gain much positive insight as to why all rational agents are normatively bound by this principle. Besides, if we interpret Kant’s moral principle only in this negative sense, we can hardly claim that Kant offers us a substantial moral theory. Then, in what sense is Kant’s principle of morality a formal principle? In what follows, I will argue that Kant’s principle of morality is best construed as a formal principle in the sense that it provides correctness conditions for moral laws, rather than providing specific moral laws.

Let us first consider correctness conditions for a definition. We can take a definition to be correct just in case it is not only formally correct but also materially correct. In the first place, a definition is formally correct just in case it conforms to the logical laws governing definitions and only admissible expressions occur in its definiens as well. Here ‘admissible expressions’ means such expressions that can legitimately be used in defining the definiendum, and so the admissible expressions should be clearer in their meanings than the expressions that appear in the definiendum. In the second place, a definition is materially correct just in case it accords with the ordinary uses of the definiendum. One notable example of a correct definition in the above sense is Alfred Tarski’s definition of truth. Tarski presents a definition of truth that is not only formally correct but also materially correct. ^{Footnote 18} As mentioned before, a definition is formally correct just in case it conforms to the logical laws governing definitions and only admissible expressions occur in its definiens. Accordingly, a definition of ‘true’ of the form ‘x is true if and only if ____’ is bound to be formally correct provided that the right-hand side of the equivalence contains only admissible expressions. In addition, as also mentioned before, a definition is materially correct just in case it accords with the ordinary uses of the definiendum. Let ‘X’ be a name of a sentence of a language L, and let ‘p’ be a translation of that language into English. Then the following equivalence holds: (T) X is true if and only if p. And all instances of (T) are obviously true, and so any definition of ‘true’ that is incompatible with them would be false. Consequently, we can say that a definition of truth for L is materially correct only when it implies all instances of (T). To put the point another way, a definition of ‘true’ accords with the ordinary uses of ‘true’ only when it implies all instances of (T).

Similar remarks apply to Kant’s concept of the categorical imperative. On Kant’s view, every concept must have not only its logical form but also its material content. He writes:

The concept of the categorical imperative is no exception. What then are its logical form and its material content? To put it another way, what are the conditions under which a categorical imperative is not only formally correct but also materially correct? To address this question, we need to examine Kant’s concept of the categorical imperative. Kant offers three formulations of it.

These three formulations of the categorical imperative tell us what a moral law (or a categorical imperative) is supposed to be. Roughly put, a moral law must have the form of a universal law; a moral law must apply to all and only rational agents; and it is we who impose moral laws on ourselves. How then are these three formulations related to moral laws?

In the first place, we may say that a moral law is formally correct just in case it has the form of a universal law. As noted, our distinctive rational nature is such that we avoid falling into the state of nature in a way that our mutual relations are governed by moral laws. Observe that if some creature cannot be subject to moral laws, then we have no other way but to face the state of nature in our relationship with that creature. Accordingly, only a rational agent, who can be subject to moral laws, can be one of us. Thus, our moral end of not falling into the state of nature requires that all of us be subject to moral laws without exception. This is what the Formula of Universal Law tells us. In this regard, Kant says:

Along these lines, we can construe the Formula of Universal Law as providing the formal correctness condition for moral laws.

In the second place, we may say that a moral law is materially correct just in case it accords with our distinctive rational nature. The reasons are as follows. We are by nature such rational agents whose mutual relations are governed by moral laws. Accordingly, there are alleged moral laws by which we govern our mutual relations, whether or not these laws are really adequate. The question then is: what kind of law is materially correct (or adequate) as a moral law? Clearly, we cannot take an alleged moral law to be materially correct if it does not accord with our distinctive rational nature. In addition, as argued in the previous section, a moral law is not such that it can be derived from some more fundamental rational ground either by theoretical reasoning or by practical reasoning. In defending the moral law, thus, the only thing we can appeal to is our thick conception of ourselves. Therefore, we may say that a moral law is materially correct just in case it accords with our distinctive rational nature. There is another thing to notice. As pointed out before, our distinctive rational nature is such that we enjoy the benefits of communal life without falling into the state of nature in a way that our mutual relations are governed by moral laws; and to say that a moral law ought to accord with this distinctive rational nature of ours is tantamount to saying that a moral law must be conducive to attaining our end of enjoying the benefits of communal life without falling into the state of nature. Therefore, we may also say that a moral law is materially correct just in case it is conducive to attaining this moral end. And this is tantamount to saying that we are the ends for which moral laws are set up from the outset. ^{Footnote 19} This is also what the Formula of the End in Itself tells us. In this regard, Kant says:

Along these lines, we may construe the Formula of the End in Itself as providing the material correctness condition for moral laws.

How then are the two preceding formulas related to the Formula of Autonomy? As has been pointed out many times before, we are by nature such rational agents whose mutual relations are governed by moral laws; and it is we who impose moral laws on ourselves in order not to fall into the state of nature. What then are these moral laws that we should set up? As noted, correctness conditions for such a law are provided by the Formula of Universal Law and the Formula of the End in Itself. That is, we are required to set up moral laws that are not only formally correct but also materially correct. Here it is important to observe that we could not set up moral laws if we were not autonomous agents; in other words, autonomy is a condition for the possibility that we govern our mutual relations by moral laws. In this regard, it is noteworthy that for Kant our concept of autonomy is equivalent to our concept of freedom, and so analyzing one concept leads us to the other concept, and vice versa (see, e.g., GMS 4:447). In addition, in a famous footnote in the second Critique, Kant says that “freedom is indeed the ratio essendi of the moral law … were there no freedom, the moral law would not be encountered at all in ourselves.” ^{Footnote 20} Therefore, the reason that we can set up moral laws is precisely that we are autonomous agents. In this sense, we may say that the Formula of Autonomy is the ratio essendi of the moral law. In this regard, observe that according to Kant moral laws are completely determined by the Formula of Autonomy.

On the basis of the above considerations, we may argue that Kant’s principle of morality provides not only correctness conditions for moral laws but also their ratio essendi. And we can bolster this interpretation by explaining why Kant claims that we are not only lawgiving members of a kingdom of morals, but also subjects in it. He writes:

As argued before, the Formula of Universal Law and the Formula of the End in Itself provide correctness conditions for moral laws. And our distinctive rational nature is such that we avoid falling into the state of nature in a way that our mutual relations are governed by moral laws. Therefore, as such rational agents, all of us are subject to moral laws, insofar as they are not only formally correct but also materially correct. In addition, it is we who impose moral laws on ourselves. And this is possible only if we are autonomous agents. Therefore, the Formula of Autonomy provides the ratio essendi of moral laws. This implies that we are lawgiving members of a kingdom of morals.

For the above reasons, we can say that an adequate understanding of Kant’s principle of morality requires considering all three formulas together. ^{Footnote 21} In addition, we can understand Kant’s principle of morality as a formal principle in the sense that it provides correctness conditions for moral laws, rather than providing specific moral laws. Furthermore, if Kant’s principle of morality is understood as a formal principle in this sense, it can be theoretically harmonized with our substantial moral end. This is because the principle of morality includes the material correctness condition for moral laws; recall that the Formula of the End in Itself provides not only duties of justice, but also duties of virtue (under its positive interpretation).

5. Concluding Remarks

In this paper, I have argued mainly for the following two claims.

First, there is a non-dogmatic way of defending Kant’s claim that the moral law is given as a fact of reason. We have no other way but to rely on our conceptual framework in order to defend that there are moral laws. And, on our thick conception of ourselves, we are by nature such rational agents whose mutual relations are governed by moral laws. On the basis of this thick conception of ourselves, we can defend that all of us are bound by moral laws. Regarding this coherentist defence of the moral law, it is important to notice two things. In the first place, this amounts to providing a kind of ‘argument’ that defends the view that all of us are bound by moral laws, although it employs a defensive strategy. In the second place, it is not dogmatic to employ this defensive strategy because, as argued in Section 3, we are entitled to employ the strategy of shifting the burden of proof to the challengers regarding such a fundamental fact of reason.

Second, Kant’s principle of morality is best construed as a formal principle in the sense that it provides correctness conditions for moral laws, rather than providing specific moral laws. If Kant’s principle of morality is construed as a formal principle in this sense, it can be theoretically harmonized with our substantial moral end. ^{Footnote 22}