Many a Scylla and Charybdis threatens the navigations of the dutiful Groundwork reader. By focusing on a clarification of some of the very different meanings of “self-determination” in Kant’s work, the following apologetic interpretation seeks to steer a middle path between two extreme but common ways of reacting to the Groundwork’s account of moral self-determination as autonomy. In this case, the Scylla objection claims, in view of the “auto” component of Kantian “autonomy,” that to speak of the moral law as rooted in self-legislation is to be too ambitious, and overly subjective, and to do an injustice to the essentially receptive character of our reason. Here the contention is that Kant misunderstands how reason is a capacity that basically appreciates reasons to act given to the subject by what is outside of it. The contrasting Charybdis concern stems from a worry about what can appear to be an overly close connection drawn between morality and freedom as autonomy. Here the critic’s contention is that the “nomos” component of self-determination in the Groundwork is too restrictive, and in a sense overly objective, insofar as it makes our action appear so thoroughly law-oriented that it seems to leave only the options of being forced either by our reason to follow the moral law or by the “natural necessity” of our sensibility to go against it, and thus – in contrast to Kant’s own later works – it does injustice to our faculty of free choice, or at least our ability to act in ways more complex than these two narrow options.
Unlike “autonomy,” the components of “self-determination,” as well as those of its German correlate Selbstbestimmung, are everyday terms in their native languages, and ones that have many similar meanings and ambiguities. The verb bestimmen (“determine”) is used repeatedly in numerous contexts by Kant, and yet, like casual English speakers, he generally does not bother to make explicit the quite different senses that the term can have.
One basic ambiguity concerns two distinct philosophical senses of “determination,” namely, an epistemological (E) and a causal (C) sense. We can say, in a first, or E sense, that we determine something when – even without having any relevant effect on it – we simply learn something informative about it, for example, when we cognitively determine the fact that a surface appears warm. We can also say, in a second, or C sense, that we determine it when we simply bring about that something beyond our immediate situation is the case, for example, when we causally determine that a surface is warm – even when, in the relevant sense, we may not at all know what we are doing. It can, of course, also happen that cognitive and causal kinds of determination combine in one complex event; we can come to learn that something is warm in the very act of making it warm. (In English, these meanings are combined in a further sense when we use knowledge in a decisive way to try to bring something about, as when we say, for example, that, “no matter what,” we are “determined to” heat a surface.)
In addition to these basic E and C senses of “determine,” there are, especially for the noun form of the term, what I will call its basic F and N senses, namely a formal or definitional sense,1 as well as a normative sense, one that, for Kant, ultimately is to be understood as having a complex moral and teleological meaning. For example, in the course of determining the composition of a metal, in the E sense of merely finding out some things about it, we may eventually arrive at its determination in the more exact F sense of a formula defining its basic nature.2 (Here the English term has roots in the French verb determiner and the process of fixing a thing’s boundaries and gaining a relevantly complete notion of it.) In Kant’s tradition, the nature of something can, furthermore, be something more than a mere physically defined arrangement, in a broadly mechanical sense, for this nature can need to be understood in terms of an ideal practical form such as, above all, the notion of a moral telos or destiny.
It is this biblical and broadly Lutheran sense that is most relevant when, after J. J. Spalding’s very popular 1748 volume Die Bestimmung des Menschen,3 Kant and numerous other German philosophers, including especially Fichte, focus on Bestimmung in the N sense of our essential “vocation,” or “calling,” at a species as well as individual level. The term “determination” does not have this normative meaning in English, and thus its relation to the other terms can often get lost in translation, but this sense must always be kept in mind when reading Kant and his use of various forms of the term bestimmen, for it is this kind of determination that always is of greatest significance to him.4
Kant’s very early works, such as his 1755 Universal Natural History, go along with the dominant broadly Leibnizian view of his era, which stresses that human beings have a significant normative determination but maintains a compatibilist doctrine of freedom, one that denies absolute free choice. This view distinguishes, as basically a matter of mere degree, our rational essence as human beings with this kind of (merely relative) freedom from the broadly mechanistic nature of lower kinds of beings, while still allowing that, according to a more inclusive meaning of the term “nature,” human beings are thoroughly determined as parts of nature in the E, C, F, and N senses. Although Kant holds to this view throughout his earliest works, he then, after the fundamental revolution in his thinking upon reading Rousseau and achieving philosophical maturity at the age of forty in the early 1760s, adopts a very different conception of the relation of nature and human freedom.5 From that time on, Kant believes that our own nature is unique in having a non-compatibilist Bestimmung in its pure moral vocation, a vocation that cannot be understood as being fulfilled, as Leibnizians and other compatibilists claim, simply by attaining higher degrees of clear representation and consequent power.
Although the notion of “determination” will be my main focus, it is also necessary to add a few preliminary observations about the “self” component in the complex term “self-determination.” In a Kantian context, it is of course crucial to keep in mind that his use of the word “self” is not limited to ordinary empirical particulars. When he speaks of “simple acts of reason,” that is, our fundamental logical capacities, as being found “in my own self,”6 he clearly has in mind, in part, a general and pure faculty that cannot be explained as the product of empirical actions or capacities. It is then, I believe, an additional – and of course still much disputed – feature of Kant’s ultimate moral metaphysics that it favors affirming that the self (of each of us) has not only a range of pure general capacities (for pure intuiting, pure understanding, pure theoretical and practical reasoning, and even for generating feelings that in part have a pure origin) but also a kind of pure and particular independent form of existence, that is, an immortality conceived of as in itself lacking any sensory qualities, spatial or temporal.7
In addition to these basic empirical and pure senses of “self,” which I take to include substantial as well as functional characterizations, there is a complex reflexive meaning to the term “self” that has a fundamental significance in the context of self-determination.8 To begin with, this reflexive meaning needs to be understood as having at least a threefold structure with implications at both empirical and pure levels of determination (and concerning all E, C, F, and N senses). For Kant, to say that we are self-determined reflexively is to say, at the least, that, at both levels, the self is determined (1) in (or, one could also say, of) itself and (2) by itself as well as (3) for itself.
At the first level, this means that human beings, individually and as a group, are commonly understood to be acting with empirical effects that are in part within them, and that are caused by empirical sources in them, and that exist for the sake of empirical ideals concerning them. Thus we can speak, as Lincoln did at Gettysburg, of a government “of the people, by the people, and for the people.” But Kant would go on to insist that we speak, in addition, in terms of three parallel forms of pure reflexivity, and thus affirm, at a second level, pure effects, pure causings, and pure ideals – all to be understood as part of our own self-determined existence and not merely a possibility for divine beings.
The mere general or structural feature of reflexivity thus does not by itself capture what Kant takes to be most important about us. That is, the three kinds of Gettysburg empirical reflexivity just listed are by themselves merely empirical, and they could exhaust the capacities of the kind of agents that Kant memorably stigmatizes in terms of the image of a mere “turnspit” (Bratenwender) (KpV 5:97). In saying this, he realizes, of course, that even at the empirical level human beings are not literally mechanical turnspits, for, as rational animals, their reflexive acts have a conscious intentionality aimed at complex ideals. But if Kant had lived long enough to hear Lincoln’s threefold reflexive remark about government, and understood all that it was directly saying as a merely empirical statement, presumably he still would have maintained what he says in his 1783 review of Pastor J. H. Schulz’s “well-intentioned” quasi-Leibnizian tract on penal reform, namely, that by itself it still misses our essential (for our Bestimmung) and absolutely pure (in E, C, F, and N senses) freedom to act and to think (RezSchulz 8:13),9 which is denied in all compatibilist systems, no matter how sophisticated their picture of us as conscious, rational, and power-enhancing agents.
Once the full context and multiple meanings of Kant’s Groundwork Section II discussion of autonomy as reflexive self-determination by reason has been spelled out, and once the pure normative sense of autonomy is understood as its essential meaning there – in contrast to merely political and psychological senses of “thinking for oneself,” or being “self-governing” according to just any rational principles that contrast with merely reacting to “threats and rewards”10 – it becomes possible to deflect common objections to Kantian autonomy as overly subjective. Explaining this sense can also help clarify aspects of the Groundwork’s difficult transition from Section II to Section III, and this can set the stage for responding to objections that Kant has an overly objective or law-obsessed understanding of action in general.
The title of the first subsection of Section III is “The concept of freedom is the key to the explanation of the autonomy of the will (Wille)” (GMS 4:440). This title might suggest to some readers that we already have a distinct concept of freedom at hand, and now we can directly apply this concept to explain a mysterious feature called “autonomy of the will.” This kind of approach is problematic, however, because the previous Section culminates in an argument that already elucidates a principle of autonomy, whereas it is the nature and existence of freedom, especially in its fundamental philosophical sense, that is, a transcendental causal one, that has not yet been addressed in a direct way. In other words, at the outset of Section III, there is an at least partially well understood notion of autonomy that Kant is taking as given at this point – one involving self-determination basically in the E, F, and N senses – and it is now his goal to introduce a direct discussion of freedom that may begin to shed light on further features of autonomy – features that abruptly shift the discussion of determination largely from its previously mentioned senses to its C sense.11
Prior to this new causal discussion, autonomy is treated in strict normative terms, as in the title of Section II’s subsection: “The autonomy of the will as the supreme principle of morality” (GMS 4:440). That title expresses an initial and relatively non-mysterious idea of what Kant means by autonomy, namely, a way of characterizing the normative principle of morality as necessarily supreme. Here the most basic feature of the norm that Kant is concerned with is that it not only meet the condition of definitely concerning a principle that is “supreme” within morality, but also that it not endanger the claim of morality’s principle to be practically supreme overall. Kant is looking for a principle that is not threatened by even the possibility of being normatively derivative, and hence is necessarily supreme in the sense of being wholly unconditioned in its value, even if it may in other respects depend on general non-moral features. For this principle to be able to concern, as Kant has already argued that it must, an imperative that is categorical, it has to be such that it does not get its distinctive normative status from outside, from “something else” (GMS 4:433).12 The supreme principle is therefore a kind of essentially reflexive principle in a new and axiological sense. At this point the idea of a will with autonomy is basically the idea of the faculty of the will as something that does not take the value of its supreme norm from outside, that is, merely through faculties external to Wille. In this way the principle can be said to have a value that holds true of the will not merely in some kind of psychological sense but in a reflexively normative sense, that is, in terms of its own basic resources, and thus purely by or through it, as opposed to on account of some other source of standards (such as mere sensation).
Given this context, it is understandable that the end of Section II treats what is outside primarily in negative terms, as when it argues that traditional factors external to the will, such as empirical conditions – whether turned psychologically inward or not (i.e., involving feelings for others and not just oneself) – or dogmatic theological or teleological considerations, whatever significance they may have otherwise, have a kind of externality and contingency13 that conflicts with the pure standard of necessary value that other parts of the Groundwork already connect in categorical moral terms with the notion of the will. Hence, when we then turn to what appears to be the only option left, namely, to what is inside the will, it turns out that it cannot, after all, be internality in any ordinary sense that carries the weight of Kant’s argument. This is because, if we were to try to focus on features that seem in an ordinary sense internal to the will but contingent, we would immediately have to concede that, as conditional, these features are still inappropriate for determining, in an E, F, or N sense, what Kant requires of a “supreme principle” of morality.14
Therefore, instead of loading Kant’s idea of autonomy with the weight of some kind of extra and mysterious boot-strapping “willful” process – which readers are understandably still tempted to do15 – it is essential to see that the underlying claim of Section II depends not on an appeal to internality or even reflexivity in a traditional, general sense but simply on the specific need for finding a basic faculty that is normatively relevant because of an at least possibly appropriate connection to an unconditionally necessary principle of value. On this interpretation, Kant’s basic thought is that we have no adequate access to such necessity from some faculty altogether outside reason (hence also his constant attacks on mystical intuition), whereas reason, the faculty that concerns the unconditioned in general, also belongs, in particular, to Wille, that is, the pure practical side of the self.16
For Kant, Wille essentially has such a special feature simply because it is defined as a faculty of practical reason, and by this he means pure practical reason, in contrast to mere instrumental rationality, let alone mere arbitrium brutum (GMS 4:412).17 Although this feature, the appreciation of absolute necessity, is in one sense internal because, on Kant’s view, it is intrinsically needed for us to be what we most fundamentally are, and thus it reflects what one always is in one’s “ownmost” self,18 this is not a matter of internality in any kind of ordinary psychological, subjectivistic, or humanistic sense. Hence, insofar as it rests on a previously affirmed respect (in principle, in the second formula of the categorical imperative) for the absolute value of rational agency in this pure sense,19 the normative self-determination of Kantian morality, as explained in the Groundwork’s discussion of the supreme practical principle of autonomy, can be read as the very opposite of what it has appeared to be to many unsympathetic readers – and even to many others who have been trying to be sympathetic. Because Kant argues for the principle from the basis of a respect for absolute necessity, the burden is on others to show that his notion of self-determination has the ultimately subjectivist and limited character that is attached to it in most contemporary uses of the notion of autonomy,20 and even in many otherwise perceptive discussions of Kant himself.21
There are, of course, passages that can understandably lead readers astray and make it appear as if Kant himself goes on to undercut the fundamentally objective position just discussed. The most frequently cited text of this sort is a passage from Groundwork II that expresses a principle of autonomy as normatively reflexive pure self-determination, which I will call NRSPD: “Hence the will (Wille) is not merely subject to the law [as it would still seem to be on moral theories rooted in contingent factors such as fear or good feeling] but subject to it in such a way that it must be viewed as also giving the law to itself (als selbstgesetzgebend) and just because of this as first subject to the law” (GMS 4:431).22 Taken out of context, NRSPD might appear to be stressing, after all, an act of arbitrary imposition. The context of NRSPD, however, as indicated by the word “hence,” shows that it is meant to follow from preceding considerations, and thus, methodologically considered, it does not invoke mere imposition (or, to be precise, what the rest of the sentence calls the will’s “regarding itself as the author”) – in the loose popular sense of autonomy – as an Archimedean point. The immediately preceding sentence, and the logical precondition for NRSPD, is that “all maxims are repudiated” that are inconsistent with “the will’s own universal-law-giving” (GMS 4:431).23 In addition to the special significance of the qualification “universal” (discussed further subsequently) in the essentially unified term “universal-law-giving,” there are two other basic points here that must be reiterated whenever trying to understand sentences like this in Kant.
The first point is that the term Kant uses throughout for “will” here is Wille and not Willkür (choice),24 which means that it does not at all have the common casual and contingent English meaning of a derivative capacity or arbitrary act – or, for that matter, of anything characterized independently of the rigorous conditions of what Kant calls pure “practical reason.” The second point is that by such reason, in this context, Kant precisely also does not mean any kind of casual and contingent reasoning about merely accidental ends25 – in contrast to almost all English uses of this phrase. What he means is not just any form of practical rationality but instead the strictly universal “legislation” of pure practical reason, which intends a law that applies by unconditional necessity and not as a matter of mere general empirical fact, as in Lincoln’s political phrase. What pure reason alone allows for is a determination of not just any kind of maxims but ones appropriate for what Kant calls “lawfulness” – that is to say, law as such, which, in the pure context of morality, signifies its having the “form” of absolute necessity – unlike the accidentally posited laws that characterize our merely empirical existence and “counsels of prudence” (GMS 4:416).
Unfortunately, Kant tends to signal this condition of strictness by simply calling the relevant kind of law “universal,” and this has led to considerable confusion about what most concerns him. Kant’s frequent use of the term “universal” is understandable in a sense, as a reminder that laws that are merely posited do not in fact tend to be universally valid, nor are they generally even meant to apply universally (and, even if they have a general intention, as with the principles of a rational egoist or an advocate of mere prudence, this is not an unrestricted universality, but conditional, Kant would say, on limited interests not shared by all). But this is just an accidental truth, although it can function as a convenient touchstone, for if something can be shown to be in no way universal, then it cannot be necessary. Kant’s fundamental concern, however, as he makes explicit at least on some occasions, is with not just any kind of universality but rather a condition of “strict universality”26 tantamount to necessity. Moreover, in this case, it is a practical necessity that is absolute, involving a law that holds even for divine nature, and hence it goes beyond even the transcendental schematized Kantian necessities of the Analytic of the first Critique, which apply merely at the sensory and ultimately contingent levels of our existence.
Only once all these qualifications are appreciated can one properly begin to understand what Kant intends by repeatedly speaking here of the “universal law” as a matter of Wille’s “own giving.” This reflexive claim is made in both sentences of the short paragraph that contains NRSPD, as well as in the concluding sentence of the long immediately preceding paragraph. The reason why Kant insists on calling the law a matter of Wille’s “own giving” is basically that he is trying to find a way to express, as he says in this sentence, that a proper normative principle of Wille must not be rooted in something that would not allow it to serve “as supreme condition of its [i.e., the will’s] harmony with universal practical reason” (GMS 4:431). In other words, the “own giving” by Wille here is not a free-floating feature but is one directly tied to Kant’s attempt to characterize its principle in such a way that it makes possible a “harmony” with practical reason insofar as such reason is strictly universal, that is, “fit to be a law” (GMS 4:431).
Kant’s concern here with “harmony” is tied to his thought that the principle of morality, in accord with the general organizational principle of reason, must have a consistent threefold specification in “form,” “matter,” and “complete determination” (GMS 4:436). This harmony has a transcendental faculty assignment aspect27 as well as, derivatively, a concrete intersubjective aspect. First, practical reason as Wille, unlike the other basic faculties of mere sensibility and mere understanding, just is the only faculty that is, as Kant goes on to say, harmonious in the sense of “well suited” (4:432; see the contrast of reason and understanding at 4:452) for such universal norms simply because reason is defined as the faculty alone appropriate for expressing and systematizing unconditional necessity. In this regard, it alone is not possibly dependent in its authority on contingent factors, what Kant here calls the “interests” of the other faculties.28
This is why, secondly, he goes on to note that its norms can always be intended to apply harmoniously in a “complete determination” or structural specification of an entire ethical commonwealth (Reich der Zwecke).29 As he stresses in the universal-law-giving passage right before the NRSPD passage, its norms equally concern “every rational being” (GMS 4:431) as an agent and thus, as a Kantian Lincoln might say, they can be understood as having validity in a pure sense, and are necessarily not only “of” and “by” but also “for” each rational being as Wille. It is this interpersonal but a priori sense of normativity, and not any empirical process, that is crucial to Kant’s understanding of moral authority. Because it is the precondition driving Kant’s overall argument toward NRSPD and is sufficient for his distinctive purposes, the idea of a strict moral necessity and independence of Wille as a faculty, as expressed in NRSPD itself, should not be read as characterizing some kind of extra process of literal “giving,” in either a humanist or supernaturalist sense, for this would replace the supreme principle of moral law with what would have to appear to be a mere quasi-necessity of arbitrary acts of authorship that could claim no more than ultimately subjective validity. In other words, Kant’s autonomy formula builds on, rather than undercuts, the thought that the moral law, and a person’s being an “end itself,” is something that has a value “in itself,” with an unrestricted validity for all agents as such.30 Kant’s third basic formula for morality can be understood as simply meant to express the point that this value must not only concern (i.e., be “of” and “for”) beings with reason but also cannot be explained independently of being rooted in the faculty of will, which alone can be at once pure – unlike mere feeling – and practical – unlike mere understanding.
The full final clause of the sentence immediately prior to the paragraph of NRSPD is: “from this there follows now the third practical principle of the will, as supreme condition of its harmony with universal practical reason, the idea of the will of every rational being as a universal-lawgiving-will” (GMS 4:431).31 Note that in this sentence, which is the crucial step supplementing the first two basic formulations of the Categorical Imperative, Kant is taking NRSPD itself as something that “follows.” I propose that this means that, for Kant, to fill out normative reflexive self-determination in transcendentally reflective, intersubjectively “universal,” and “complete determination” terms,32 is simply to reiterate, in the new language of the third formulation, our need to resist any reliance on any contingent use of faculties that would undermine a kind of already assumed necessary practical “harmony.” Rather than imposing on Kant an odd and invalid extra meaning to the notion of self-determination, one can read him as basically just repeating a point that is made throughout his work and that is systematically elaborated, in an explicitly negative manner, in the concluding subsections of Groundwork II, “Heteronomy of the Will as the Source of all Spurious Principles of Morality,” namely, that if one were to try normatively to account for the necessary authority of morality in terms of exercises of faculties that are manifestly contingent, such as our sensitivity to either external or internal empirical pressures, or even theological concepts characterized in a merely arbitrary fashion (concerning a desire to please the whims of a tyrannical superpower), then this would be tantamount to sacrificing the normative necessity of the moral law and its chance for harmony with universal reason.33
Note that although it is true that there is a contingent causal relation between our awareness of such mere pressures and the existence of particular stimuli for them, it is not the relational causal contingency of the pressures that is the key to Kant’s objection to them; what matters is the immediately evident contingency of their value relevance.34 There is, for example, no reason to think that the prestige often associated with social rank is necessarily a moral good. But if contingent sources of normativity do not as such harmonize with the strict modal and universal nature of the moral demands of practical reason, some kind of fitting and necessary location for the possibility of this harmony needs to be sought. From a Kantian perspective, there is one and only one obvious alternative here, namely, to look toward practical reason itself. Reason in general is characterized throughout Kant’s philosophy as precisely that faculty which determines (in E, F, C, and N senses)35 all strictly necessary truths, and hence it only makes sense to say that the practical necessities of morality must be sought within the faculty of practical reason, what Kant also calls Wille.
Given that these necessities are unavoidably valid, it may be disconcerting at first that Kant uses an active voice here and speaks of “the will” as “giving” the law36 rather than simply seeing, understanding, or appreciating it. But there are understandable reasons for his use of the active voice here. The most obvious one is that he wants to mark a strong contrast with what he takes to be the manifestly passive and inadequate putative sources of unconditional value that others tend to rely on: mere sensation, tradition, threats and such. Moreover, even when, in a moral context, Kant does use, and even emphasizes, a term that is translated as “impose” (auferlegen), he also uses it in part in a passive voice, as something imposed “upon the will.” That is, he states that for actions (e.g., not lying to someone about a truth that they have a right to know) out of “immediate respect,” “nothing but reason is required to impose them upon the will,” since “these actions need no recommendation from any subjective proclivity … to coax [erschmeicheln, that is, lure by mere flattery] them” (GMS 4:435).37 Here again it is clear that the cash value of the term “impose” is simply to sharply oppose the idea of accepting only manifestly contingent sources of value. As Wille, we “give ourselves” the law most basically insofar as we cannot, as beings of reason, let a “supreme principle,” no matter how flattering, be contingently imposed upon us as normatively decisive. We understand that mere efficient causal determination, as a contingent fact about events, cannot be the same thing as the normative determination of a necessary standard of value – and this is true even if the causation is a matter of our own active imagination.
Throughout his philosophy, Kant makes use of a basic distinction between Tun and Lassen (GMS 4:396), that is, between being active in a paradigmatic initiating sense, in contrast to allowing something to happen. But even “allowing” is understood in this context as also a kind of action, and it is clearly Kant’s general view that, in the context of our relation to the status of norms, for us even to merely allow any of these to hold sway is to engage in a kind of act and to determine oneself in a “self-incurred” way.38 Hence, intentions in which one chooses to ignore the claim of morality and to accept as basic what Kant calls the merely heteronomous standards of sensibility and self-love must also involve a kind of act on our part, even if not in an explicit phenomenological sense.39
A common objection at this point is to say that even if the value of a law is not something to be merely “taken” in the sense of a natural process that is undergone totally passively, this does not mean that we should say it is self-given either, for one might want to characterize it as simply recognized as authoritatively present.40 Against this gambit, a Kantian might at first argue that we should speak of the faculty of reason in active terms simply because of considerations that go back to a long-standing Scholastic and rationalist tradition of understanding intentionality in general as active because at least it is implicitly propositional (and thus involving synthesis, in contrast to mere sensation and primitive feeling), although by no means in a necessarily arbitrary way. Here, however, one must distinguish between reason’s general normative (N and F) determination of the standing of a practical law, and the cognitive and appreciative acts in which a particular reasoning subject determines itself, through reason in a concrete E and C sense, to be committed to a maxim in a way that takes an actual stance on the law. Even though the latter kind of determination, on each occasion, is understandably always a matter of activity rather than mere passivity, this may leave it unclear why the general formal and normative determination of the law’s status as supreme should be said to be self-given. Nonetheless, there remain the reasons already given for speaking of even the mere formal determination of the law’s standing as something that is self-determined, in a non-subjectivist, pure, and distinctively internal sense, rather than other-determined. Kant’s view is that, even before trying to ground the synthetic claim that the moral law is in fact binding on us, the philosophical analysis of what the acceptance of such law would entail41 does point directly to a non-subjectivist understanding of NRSPD. The key point here, once again, is simply that it must be within practical reason itself, and neither of the two other faculties distinct from it, namely, mere feeling and mere theoretical understanding, that such a strict standard for practical life would have to reside.
It is, to be sure, a bit of provocative language to speak of this necessary harmony between pure reason, as a basic faculty, and pure morality, as a practical standard with content, in terms of reason’s “authoring” and “legislating” morality’s pure law (cf. GMS 4:448), for this might suggest to some readers the existence of something like an independent being, such as a person or a government, engaged with a totally independent other item, that is, an entity that need not be. Reason, however, is not a separate individual but is just Kant’s term for a pure general faculty, and as such it has a necessary relation to content that has lawful form, and hence it can be said to be “legislating.” Kant sees that in the case of practical reason, the “legislator” and its work are not only in close harmony but are in a necessary and reciprocal relation, for without any work, without some content in necessary laws, the faculty of reason would be an Unding, far emptier than any mere thought of a thing in itself. All the same, pure reason itself is responsible simply for the lawful aspect of value, the universal conditions required to respect the absolute value of being an agent with reason. Kant recognizes that the manifold empirical contents of the particular and conditioned values and reasons that arise in everyday life, prior to regulation, are not themselves rooted in pure reason, let alone the bare notion of its universal legislation. His speaking of reason’s authorship can thus be understood as a technical move, limited to a very specific meaning concerning faculties, and as having only partial, metaphorical overlaps with familiar notions of empirical authorship and legislation. The point of the Groundwork, as a “groundwork,” is basically just to express general formulae for a necessary practical principle, and to ask the question of which faculty can be consistently understood as correlated with such a principle, and so in this context it is not mysterious to propose that reason is crucial, for it is the only faculty that can be consistently regarded as the relevant authority.42 In a textual sense, this means that the formula of autonomy need not be regarded as itself independent or methodologically autonomous, insofar as it depends thoroughly on a prior acceptance of the first two formulae. Moreover, the independence, or strict lawgiving capacity, of reason by itself is only a crucial necessary, but not a sufficient, condition for moral guidance in the complexities of human life.
The fundamental normative meaning of “heteronomy,” as the opposite of the purely normative conception of autonomy, must be kept in mind when turning back to the text at the very beginning of Section III, which is the prime exhibit of Kant’s mixing, without a detailed warning, two quite different notions of determination. In Section II, Kant makes frequent normative references to Wille but does not begin to provide a formal ontological exposition. The first sentence of Section III abruptly starts such an exposition by saying, “Will (Wille) is a kind of causality of living beings insofar as they are rational” (GMS 4:446). The mystery of why this quite different sense of determination is being brought in is clarified by the remainder of the sentence, which introduces the topic of freedom, the concept that the heading of the first subsection indicates will be a “key” to the “explanation”43 of “the autonomy of will.” The notion of causality has to be brought in because, given Kant’s general categorial theory, freedom is basically a kind of causality, a causality at first described here as one that “can be efficient independently of alien causes determining it” (GMS 4:446).
Kant realizes full well that this is not the only way a philosopher might try to characterize freedom. As noted earlier, Schulz’s characterization, as well as Kant’s own earliest work, adheres to a compatibilist doctrine of human freedom that allows for the presence of “alien causes” (at least of certain types). Here in Section III, Kant goes on immediately to indicate that his first remark about freedom does not amount to a proper definition but simply provides what he calls a partial and “negative” “explication.” He moves toward a positive characterization by making a connection between freedom and lawfulness. Freedom is a kind of causality, but, contrary to what others have held, and what is assumed in Kant’s own earliest works (an assumption that is later criticized, in the dialectic of the first Critique, e.g., A447/B475), as the dogmatic presumption that freedom would have to be “lawless”), it should not be thought of as possibly a matter of “lawlessness.” Causality as such, he now contends (not uncontroversially, to say the least), requires lawfulness,44 that is, a kind of necessity, for, given a cause, “something else, namely an effect, must be posited” (GMS 4:446). The alternative he calls an “absurdity” (Unding), although presumably not because it is meant to be impossible on trivial analytic grounds.
Up to this point, the subsection has been discussing determination only in an “efficient” sense. It continues along this line in the next sentence but takes a surprising turn by introducing a new phrase and saying “natural necessity was a kind of heteronomy of efficient causes” (GMS 4:447).45 The use of “heteronomy” in this way is surprising because one might have assumed, from the extensive discussion at the end of Section II, that it is simply a normative notion. “Heteronomy” is treated earlier not in causal terms but as a matter of the approval of a kind of ultimate “principle” (GMS 4:443) of value, one that comes from “something else,” that is, from a contingent value rather than one that is necessarily authoritative because it is essential to (respect for) one’s own self as an agent with reason.
Although heteronomy is introduced in this way as a normative concept, a linking complication here is the fact that any principles and choices that are heteronomous in the normative sense also have implications at a causal level. To approve a heteronomous standard as supreme is to be ready, above all, to move one’s will, as an efficient cause, to generate intentions and external events with an aim to satisfying this standard and attaining what Kant calls merely subjective ends. Furthermore, the typical way that Kant appears to be assuming that people adopt heteronomous standards is by a process of incorrectly allowing factors that they merely passively experience through efficient causation (such as appealing sensory temptations) to count by themselves as providing sufficient grounds for moral decisions. Nonetheless, in this paragraph Kant surprisingly does not use the term “heteronomy” with reference to a situation of normative decision and choice about what counts, for he refers simply to instances of “natural necessity.”
Perhaps because of unusual texts like this, it is sometimes thought that, at least in this period of his work, Kant does not have a robust view of agency, one according to which actions not in line with morality are free choices rather than merely reactions in line with natural necessity.46 The problem here may rest in part on the mistake of thinking that when Kant speaks of a “heteronomy of efficient causes” as a kind of “natural necessity,” one can infer that he is committed to thinking that heteronomy in genuine human action must be a matter of mere necessity. I do not subscribe to this interpretation. This is not only because of features of the Groundwork itself, but also because of the often neglected fact that if one looks at Kant’s other main works immediately prior to the Groundwork, one finds that they are distinguished by a new and noteworthy explicit concern with the issue of free choice in general, and, in particular, a concern with not allowing immoral actions to be regarded as matters of natural necessity, such as ignorance, innate pressure, or mere external force. In taking time out then, in the most intense period of his career, to publish a critical review of the relatively unknown figure, J. H. Schulz, Kant’s obvious preoccupation is with insisting on a general rejection of compatibilism, and in particular of the idea that “vice” and “moral good or evil” can be explained as a mere matter of “degree” (RezSchulz 8:12), of nature keeping us relatively ignorant and weak, so that there would be “no free will” and “all remorse is idle and absurd” (RezSchulz 8:11). Even though Kant’s arguments in the review of Schulz are too brief to be persuasive and are formulated in some ways that contrast with his other writings,47 the review’s clear insistence at this time on absolute and general human causal freedom, absolute moral value, and an absolute notion of reason is consistent with what is already indicated in the first Critique (1781), and it defines a position that has to be taken as at least implicitly present in each of the succeeding writings of the mid-1780s.
If the Critical Kant had the belief that our practical errors were simply a matter of failing naturally to try to do the good, solely because of the natural necessity of a lack of sufficient knowledge or power, then there would be no reason for him to be upset, as he manifestly is, by a theory that appears to have no room for “remorse” with respect to the relevant “frame of mind” (RezSchulz 8:11). Precisely because Schulz’s book is primarily about punishment, that is, the negative side of human action, Kant wants to draw attention to the fact that a compatibilist account here conflicts with what he takes to be the obvious proper belief that vice is evil in a sense that calls for “just” (RezSchulz 8:12) retribution, rather than being regarded as simply bad, that is, illegal or weak in its perceptual underpinning. The review mocks Schulz’s view for turning “all [NB] human conduct into a mere puppet show” (RezSchultz 8:13), a mockery that would be out of place if Kant’s own view were that when we act immorally we are simply by nature failing to follow the moral law of reason, rather than freely rejecting it by adopting a maxim contrary to its supremacy.
In the Groundwork itself, “transgression of a duty” is similarly described not as a mere failure to do the right thing, or the result of an inevitable force that makes us do less than the best for ourselves. Kant’s account is that, in not being moral, we “really will … that the opposite of our maxim should instead remain a universal one, [and] we take the liberty (Freiheit) of making an exception to it for ourselves” (GMS 4:424). This “real willing,” “taking freedom,” and “making” an exception is introduced precisely as relevant for moral blame rather than excuse or mere regret, because insisting on “blame” makes sense for Kant only on the assumption that agents are exerting freedom of choice even when doing evil rather than good. All this is only to be expected, given Kant’s crucial reference to blame already within the 1781 edition of the first Critique’s third Antinomy discussion of absolute free choice (KrV A555/B583).
Furthermore, Kant’s 1784 essay on enlightenment has a similar underlying preoccupation with free evil, even though it is expressed in diplomatic terms that have led many readers to suppose that Kant’s topic here is merely political and concerned with choice in a relative and empirical sense. The essay’s initial, fundamental, and most striking claim is that our lack of enlightenment is “self-incurred” (WA 8:35), hence a matter of our own activity and presumably something to be blamed in a non-Schulzian way. This claim is meant to directly counter the excuse that mere natural stupidity, internal “laziness,” or “cowardice” (WA 8:36), which would be the last word in other theories, are the ultimate causes of our problem. In saying, twice and with emphasis in the first paragraph, that our general attitude is “self-incurred,” Kant is stressing that it us up to each of us to determine the kind of concerns we in fact give absolute and not merely relative priority, namely, either the understandable but merely local and contingent “private” demands to obey only local figures such as “the officer,” “the tax official,” and “the [state-appointed] clergyman,” or, instead, the cosmopolitan call to follow, as supreme, the norm of the free “public use of one’s reason in all matters” (WA 8:36).
This public use is normatively determined by principles of justice valid for the world as a whole, that is, the proper kind of law that Kant says a people as such “could impose upon itself” (WA 8:39). The “could” is crucial, for it implies that Kant is not asking about the mere empirical question of what a group of citizens might happen to do. He is asking about a form of “law” that they could properly formulate and accept just as “a people,”48 that is, a community of human beings regulating their actions by a common faculty of reason for discerning rules that can be universally and necessary valid and that respect persons as rational beings.49 This is also why Kant’s injunction to dare to “think for oneself” (“Sapere aude!” WA 8:39) is not redundant – on the ground that, as some have objected, on his own theory human beings are in fact always in some way (empirically) thinking for themselves50 – nor is it absurdly anarchic, as some have objected, as if Kant is proposing that persons should never listen to or learn from others. Rather, the essay’s point is an appeal to free agents to listen, above all, to the voice of pure reason in them, which, as noted before, is characterized already in the first Critique as that which reflects “one’s own self.” Admittedly, Kant does not explicitly say here that any individual giving priority to maxims that transgress the rules of reason is to be blamed for an evil use of free will,51 but the stressed and unrestricted scope of his use of the term “self-incurred,” especially at this anti-Schulzian time in his career, leaves this conclusion as an obvious inference. Moreover, this line of interpretation is substantively supported by the fact that the very same concern is evident in Kant’s other work immediately prior to the Groundwork, namely, the caustic reviews of his former student Herder, in which Kant repeatedly goes out of his way to mock a theory of human nature that relies merely on a hierarchy of natural forces rather than any reference to absolute individual freedom of choice.52
This is not to deny that it is only later, in his book on religion and numerous related essays (such as his 1786 essay, “What is Orientation in Thinking?”), that Kant goes into full detail about a theory of will and free choice by characterizing as “radical evil” humanity’s self-determined rejection of morality. An appreciation of his earlier commitment to the underlying idea here can be – and has been – complicated by the fact that, already in the beginning of Section III of the Groundwork, Kant supplements the negative component of his account of freedom, the independence from “alien causes,” with a positive component, defined as respect for the moral law, and he goes so far as to say “a free will (freier Wille) and will (Wille) under moral laws, are one and the same” (GMS 4:447, similar formulations occur in the second Critique). Readers can fall into a trap here by skipping over the key term “under” – which can apply to a potentially free and evil as well as a good will – and then thinking that Kant is identifying freedom of choice only with action that is positively moral. He is not doing that, for he is not using the term for choice (Willkür) but is basically just reiterating that his concept of Wille is a normative concept, and it has within it positive standards of pure practical reason, rather than being a mere neutral or indifferent “rational” source of effects. A merely prudent maxim, such as that of the shopkeeper at the beginning of the Groundwork, is clearly rational in an ordinary sense, but Kant’s main point in introducing such a maxim is precisely to claim that, when it is given priority over the moral law by an agent, we should not say this leads merely naturally to a life with a lower degree of goodness, power, and intelligence. Instead, we should presume that this agent, like all rational agents, has Wille and therefore should know better, and so is to be condemned for a blamable and therefore freely chosen maxim that is not a matter of natural necessity’s “heteronomy of efficient causes.” The merely prudent shopkeeper’s decision is a “self-incurred” determination in an absolute sense, one that the rest of the Groundwork aims at length to show is to be regarded as a self-determination that is an uncaused causing, and not like the effect of a mere turnspit.
In sum, I have been arguing that proper reflexive self-determination by the basic faculty of practical reason, which is neither a matter of overly subjective arbitrariness nor of overly objective mere subjection to law, is the essence of the Groundwork’s doctrine of autonomy, and – contrary to the implications of numerous common interpretations – is simply meant to follow immediately from reflection on what is needed to maintain the earlier formulations of the principle of morality. This is, of course, not to say that Kant is demonstrably correct in holding to this principle, or that the principle also demands the kind of absolute metaphysical freedom in the making of good or evil choices that he insists upon, but that is another topic, and, in Fontane’s phrase, ein weites Feld.53
1 See, e.g., GMS 4:461: “autonomy – as the formal condition under which it alone can be determined.”
2 See, e.g., GMS 4:436: “a complete determination (Bestimmung) of all maxims by that formula.”
4 See, e.g., GMS 4:396, “the true vocation (Bestimmung) of human beings must be to produce a will that is good.”
6 KrV Axiv: “ich demütig gestehe … ich es lediglich mit der Vernunft selbst und ihrem reinen Denken zu tun habe … weil ich sie [deren ausführlichen Kenntnis] in mir selbst antreffe … alle ihre einfache Handlungen.”
7 See especially Kant’s criticism of the notion of resurrection in Religion within the Boundaries of Mere Reason (RGV 6:128–9 n.).
8 See Prauss 1989:253–63 and O’Neill 2013:282–7. I take the general picture of autonomy that I am developing here to be consistent as well with most of the other essays in this helpful recent collection.
9 Here Kant at first calls this a freedom to “always act as if one were free [and such that] this idea also actually produces the deed,” and then he adds that “the understanding is able to determine (bestimmen) one’s judgment in accordance with objective grounds that are always valid,” and hence we must “always admit freedom to think, without which there is no reason” (RezSchulz 8:13). These ways of characterizing the absolute freedom to act and to think are not clearly in line with the best formulations of Kant’s position, but they vividly disclose, in an initial way, the topic that he is most concerned with writing about right at this time. See the end of his essay “An Answer to the Question: What is Enlightenment?” (1784) and his reviews of Herder (1785) as well as the Groundwork (1785). His “Idea for a Universal History with a Cosmopolitan Intent” (1784) also has a basic, although indirectly expressed, concern with absolute freedom. See Ameriks 2012: chapters 9 and 10.
11 Hence the title of the Groundwork itself, and of Section III, which introduces the notion of a “Critique [i.e., explanation] of Pure Practical Reason.”
12 See also GMS 4:458 and 4:427, where Wille is described as “the capacity to determine itself to action in conformity with certain laws … the objective ground of its self-determination is the end.” The term “end” makes clear that the point of speaking of the will’s (“objective”) self-determination is to stress a matter of normative determination.
13 See GMS 4:425, which says the ground of value cannot be in any “special natural predisposition of humanity.”
14 See GMS 4:432–3: “[When] one thought of [oneself] as subject to [unterworfen, i.e., merely passively subject in contrast to “legislating”] a law … it had to carry with it some interest or constraint … necessitated (genötigt) by something else [because not arising from the will’s own law] in conformity with a law … a certain interest, be it one’s own interest or another… But then the imperative also had to be conditional.” Later Kant also speaks of “interests” generated by reason itself, in which case they have an intrinsically necessary status.
15 A similar common and understandable, but also self-defeating, approach is often taken to the metaphysics of Kant’s idealism, as if somehow a special process of human “making” could provide a consistent Kantian explanation of the necessary conditions of our grasp of spatio-temporality itself.
16 Kant therefore stresses later in Section III that reasons still need to be given for the synthetic claim that we do have will in a strong sense, or at least in some persuasive sense must regard ourselves in this way.
18 See, e. g., GMS 4:455, “das moralische Sollen ist also unser eigenes notwendiges Wollen als Glied einer intelligiblen Welt,” and 4:457 and 4:458, “das eigentliche Selbst.”
19 I take this absolute value of being an end in itself to reside for Kant neither in actually acting with a perfectly good will, nor in simply setting whatever ends, but in having the capacity always to set ends that meet the conditions of pure morality.
20 On the need to sharply distinguish Kantian autonomy from less demanding uses of the term, which concern contingent political or psychological matters, see again the helpful essays in Sensen 2012. This point about the absolutely necessary character of Kantian autonomy is compatible, I believe, with the argument there by Paul Guyer (2012:71–86), that Kant also develops an empirical account of how humanity gets better over time at committing itself to autonomous principles.
23 “der eigenen allgemeinen Gesetzgebung des Willens.” My translation substitutes for the Cambridge, “the will’s own giving of universal law,” because the latter translation (see also note 31) might suggest a contingent relation between the terms, as many Anglophone interpretations tend to assume. For criticism of this tendency, see Ameriks 2012: chapter 6.
24 This term has a common connotation of arbitrariness in German, e.g., at GMS 4:428.
25 These are ends that one could be “subject to,” so as to meet the first, but only the first, part of the key phrase, just cited, characterizing autonomy at GMS 4:431. I bracket here the vexed external issue of whether happiness or universal well-being in general, rather than either accidental particular ends or the Kantian notion of pure duty, may by itself be an absolutely necessary value.
26 Cf. GMS 4:430–1: “because of its universality it applies to all rational beings as such.” This phrase surely must be understood as expressing a necessary essence, and not a universality reflecting mere contingent applicability. See also: GMS 4:426, “it is a necessary law for all rational beings…”
27 This is part of Kant’s general project of demarcating the transcendental “location” of the diversity of our faculties, in opposition to empiricist and rationalist “single root” tendencies that eliminate any non-derivative conception of will.
28 4:432: “the principle of … universally legislating … is founded on no interest, and thus can alone, among all possible imperatives, be unconditional.”
29 The presumption of this harmony is overly swift. As later work in logic has revealed, even seemingly necessary formal principles of theoretical reason can lead to paradoxes and a multiplicity of incompatible options, and so one should keep in mind that even Kantian practical norms based on pure reason may be vulnerable to similar problems.
30 This worry is raised by Larmore (2011:8–9, and 19), who raises the common, and self-defeating, worry that Kant is literally turning reason into an “agent.” I take my reading of the Groundwork, as basically just trying to give moral principles their proper faculty location, to entail all that Larmore wants from his own (allegedly more realistic) normative theory, especially insofar as Larmore goes on to state that what is valuable is not to be thought of existing in a totally isolated way but as in correspondence with our reason. Larmore himself says, “reasons have a relational character,” that is, involve relations to “possibilities of thought and action” that need to be “discovered” (Larmore (2011:20) – presumably by agents with the faculty of reason. Anti-Kantians tend to believe this kind of response is ruled out by Kant’s characterization of heteronomy as a matter of allowing the “object” to determine the (moral) law (GMS 4:441), but this is to overlook that what Kant is rejecting is simply the thought that a normatively contingent or indeterminate “object” could be law-determining; in other contexts he is willing to speak of the law itself as the proper “object” of practical reason.
31 This again is my modification of the Cambridge translation, which reads, “will giving universal law,” and thus does not as exactly reflect the German “allgemein gesetzgebenden Willens,” a phrase that is found on the next page and elsewhere without a break between the terms characterizing will: “allgemeingesetzgebenden Willens.” The combining of the terms without a break best expresses the crucial point that Kant is making an essential and not an accidental characterization of what he calls Wille.
32 This three-step structure dominates the Groundwork from the beginning, although sometimes in a partially inverted order. The three principles of Section I are introduced heuristically in the order of, first, “subjective” (that is, existing in the subject) content, that is, the good will and its necessary value (the notion of necessary value is also placed first in the Preface, GMS 4:389), then “objective” form, that is, having a right (universalizable) maxim, and, third, “determination” through “pure respect for practical law,” which “outweighs” all mere inclination (GMS 4:400) and is expressed later in terms of the formula of autonomy. In the initial presentation of the three basic formulae of the categorical imperative in Section II, and then also in the summary at 4:431, the order becomes (1) the “objective” form of universality, (2) the “subjective” content of the necessary value of being an agent with reason, and (3) the unity of these in the notion of a “legislating” rather than simply passive Wille – a third point that “follows” on reflection because the preceding two points about the universality and necessity of the supreme principle of morality cannot be understood in terms of a merely contingently determined will. See also Allison 2011:124, and note 42.
33 Here Kant has a special problem insofar as he must concede that, of the four basic options, the perfectionist theory of value need not be vulnerable to the objection of relying on contingent values at its base. This may be part of the reason why Kant is especially interested in the feature of the universalizability of maxims, which he thinks gives his theory a special advantage, given what he takes to be the inescapable indeterminacy of the notion of “perfection” alone.
34 Hence I assume there is concern about a judgment (ultimately involving freedom) of value, and not a mere causal relation, at work in passages such as this (GMS 4:460): “it is not because the law interests us that it has validity for us (for that is heteronomy and dependence [normative!] of practical reason [this is a point about reason, not mere psychology]…).”
35 This statement about “reason in general” is compatible with allowing that reason “in particular,” that is, as it is actually taken up on a particular occasion by a person reasoning in action, is part of what allows that person to be causally effective. The causality of practical reason has been emphasized in recent work by Stephen Engstrom.
36 One should also keep in mind that what look like German uses of the term “give,” that is, geben, are often translated more properly in non-activist terms. Es gibt does not mean “it gives,” but simply “there is,” just as in English, when we say “it rains,” we really are not speaking of a separate “it” but just mean that “rain is falling.” I suspect that Kant is most attracted to the word “give” here simply because he wants to use a verb that contrasts with “take,” which in this context signifies merely taking over from an external source in a normatively lazy way. Another complication is that here “give” and “take” have connotations that contrast with how they are generally used in relation to the English philosophical notion of the “myth of the (merely passively) given.”
37 I have inserted the phrase “these actions … proclivity” from an earlier part of the paragraph, for grammatical and explanatory reasons. Without the insertion, the translation of Kant’s phrase reads, “to impose them upon the will, not to coax…,” and here one sees perhaps even more directly how Kant’s main aim is simply to make a contrast with contingent sources such as mere “coaxing.”
41 Note Kant’s cautious language in this section: “if there is a categorical imperative” (GMS 4:432).
42 A fundamental and very different objection to Kant is to insist that it is not clear that reason itself can have practical content, for it can seem that it is at most a faculty for testing consistency or, as in its theoretical use, for illegitimately positing unconditioned conditions for items given to it by other faculties. This kind of objection is not clearly relevant at this point in the text, however, for Kant introduces the notion of reason’s practical self-determination only after he has already characterized morality in terms of what he calls its “formal” and “material,” or “objective” and “subjective” aspects, the two aspects that need to be understood as being in “harmony” through a relation to a common third factor, our faculty of reason’s “complete determination” (GMS 4:431). By these two aspects he means, first, the formal “objectivity in the rule” that is given with the first formulation of the moral law and the determination of the categorical imperative in terms of a condition of (necessary) universalizability (GMS 4:431); and, second, the “matter” or content that is there, as he says, “subjectively,” meaning (using the pre-modern sense of the term) that is present with subjects that can and (normatively) must have the “end” of rational agency, which is the need to observe the second formulation of the categorical imperative and the demand to respect “rational being … as an end in itself” (GMS 4:431). The content of the theory is thus provided by the basic conditions for preserving and enhancing rational being in this sense.
43 Erklärung, a term that on this topic Kant tends to use in the sense of providing a detailed causal explanation. The subtle structure of section III is that it introduces the notion of freedom as if it might be used to provide such an explanation, but the section concludes by stressing that we can only defend but not employ, in any particular explanation, the Idea of freedom as an efficient determining cause (“wirkende … bestimmende Ursache,” GMS 4:462). Freedom can, however, help to “explain,” in a formal, or purely conceptual sense, what is central to the notion of moral autonomy. Similarly, what we find out, after the worry is raised at the end of section II that the moral law, and the freedom that is its condition, might itself be a mere “phantom of the brain” (Hirngespenst), is that what is really a Hirngespenst is rather the thought that we need to and can obtain an Erklärung of how freedom “works” (GMS 4:462).
45 A similar phrase, “heteronomy of nature,” is used again at GMS 4:452 (cf. 4:458), but it is important to see that at this point Kant is speaking explicitly of regarding our actions as being “appropriate to” (gemäß) a specific value “principle” (either “pure will” or “happiness”) of action (Handlung). Here Kant is pointing to N and not mere C determination, and this can be lost in a translation that simply says “conform entirely with the natural law of desires…” See earlier note 12, regarding GMS 4:427.
46 This criticism is often identified with Henry Sedgwick, but others, from Kant’s time to our own, have shared this worry.
47 Here I have in mind especially Kant’s linking of arguments for freedom from conditions of mere thinking, to those for action, and his speaking of “reason” in general terms without distinguishing the demands of pure reason, and its “Idea,” from ordinary rational considerations.
48 I say “just as a people” because of the noteworthy fact that here Kant asserts that even the “legislative authority” of a monarch derives from the “collective will” of the people (WA 8:40; see also TP 8:304; ZeF 8:381; RL 6:313, 329, 342). This broadly democratic sentiment contrasts only superficially with the pragmatic advice concerning obeying the king that Kant gives at this time, which has tended to mislead critics into supposing that his position is inherently conservative. Here my reading of Kant is somewhat to the left of the helpful analysis by Katrin Flikschuh (2012). Because I take absolute moral autonomy, involving the universal necessary values of public reason and not mere “modest” self-governance (i.e., mere independence of pressures from other people), to be the concern already of the Enlightenment essay, my reading also differs from Larmore 2011: 8.
49 An especially good indication that Kant has this kind of pure normative notion in mind can be found in sentences immediately prior to this remark, in which Kant discusses the possibility (quite relevant to his own situation with regard to the state church in Prussia) of a presumably mature and uncoerced “ecclesiastical synod” getting together and agreeing that its doctrines would henceforth never be allowed to be subject to revision by critical reasoners. Although one might think that such a group is a paradigm of trying to bind and “give oneself a law” in an empirical sense, Kant declares the idea of such a contract philosophically “absolutely null and void, even if it is ratified by the supreme power,” for it is directly contrary to “public” practical reason’s necessary respect for our “original vocation” (des Berufs jeden Menschen) (WA 8:39).
50 See Bittner 1996:345–58. Bittner’s further worry, about how Kant could consistently think that people were “mature” before they fell into this state of “immaturity,” can be answered by looking at the details of his reaction to Rousseau, where Kant gives a complex account of how what he mostly has in mind is the special tendency of modern agents to get so caught up in the “luxury” of intellectual and scientific pursuits that they have begun to undermine the healthy common sense that they were born with. At a deeper level, of course, the problem is the old theological issue of how human beings can start by being created with a “good seed” and yet, in any context, tend toward radical evil, until their self-incurred inversion of priorities is reversed.
51 In passages that directly invoke the key terms of the Schulz essay, the end of the Enlightenment essay speaks of our having, above all in our nature, “the calling to think freely” (WA 8:41) and to engage “freedom in action” (WA 8:42). In their context, these phrases have a significant empirical and political sense, but Kant also expresses, in a final phrase, his absolute moral claim, the need to treat human beings above all in accord with their “dignity” (WA 8:42).
53 For the purpose of textual orientation and a very brief review of the main issues under discussion, I provide a bare bones outline of what I take to be the main arguments relevant here, which concern Kant’s Groundwork of the Metaphysics of Morals, 4:431–48, that is, the end of its Section II and the beginning of its Section III. The argument sketches are listed under the headings: IIb, IIIa, IIIb, and IIIc. The last step of IIIc, namely 4*, is very important in its own right, but it has been much discussed elsewhere and is bracketed on this occasion. The main concern of my discussion is indicated under heading IIIb, which simply notes the ambiguity of “self-determination” and the underappreciated difficulty of directly connecting Kant’s reasoning toward the end of Section II with his remarks right at the beginning of Section III:
1 Suppose there is a “supreme principle of morality.”
2 This principle requires a normative rule that is absolutely necessary.
3 An absolutely necessary rule cannot be normatively determined (i.e., “N-determined”) by what we access through faculties other than reason.
4 So, the supreme principle of morality must be N-determined through reason.
5 The will, as Wille, is the faculty of practical reason.
6 As such, it N-determines principles that are normatively necessary.
7 In N-determining such principles it cannot be N-determined by what is outside it, and so its N-determination must be internal, by its “own universal-law-giving.”
8 Hence the “supreme principle of morality” is an internal N-determination of “the will,” a matter of its “autonomy” in the sense of normative self-determination.
1 Wille is a faculty of being an efficient determining cause (i.e., “C-determining”).
2 Causality is always C-determination in accord with some law.
3 Free Wille is a freely acting efficient cause.
4 Free Wille is therefore efficient C-determination in accord with some law.
5 As a free causality in a negative sense, Wille cannot be efficiently C-determined from outside by “natural necessity,” that is, the “alien causes” covered by the laws of nature.
6 As free causality in a positive sense, Wille needs some “non-alien” causal law, so (given 5) the law in accord with which it is C-determined must be internal.
7 Free Wille as a cause is therefore internally C-determining in a non-alien causal sense, that is, it has autonomy in the sense of causal self-determination.
Even if arguments IIb and IIIa are accepted, it is not clear yet how they, and their various senses of “self-determination,” relate to each other and to the overall strategy at the end of the Groundwork. A natural interpretive hypothesis is that Kant has an encompassing argument of the following form:
1 There is a supreme principle of morality valid for us only if we have Wille in the sense of being autonomous as fully self-determined.
2 To be fully self-determined is to have Wille that is at once normatively self-determined and causally self-determined in an absolute sense, that is, with uncaused efficient causing.
3 We can now understand what it is to be normatively self-determined, as well as what it is to be causally self-determining in an absolute sense, but we cannot understand how to reflectively affirm autonomous morality as valid for us until it is shown that we should in fact regard ourselves as in fact having Wille that is self-determining not only normatively but also in the absolute causal sense.
4* It can be shown that we should in fact regard ourselves as causally self-determining in an absolute sense (argued for only in the remainder of Section III – and in a controversial way that Kant appears to have retreated from soon after), and so, as fully self-determined, we should affirm that we are autonomous.