Causal Necessity in Kant’s Empirical Account of Human Action

In the first Critique, Kant says, “all the actions of a human being are determined in accord with the order of nature,” adding that “if we could investigate all the appearances . . . there would be no human action we could not predict with certainty,” and he gives a striking example to illustrate this general point. He says,

Let us take a voluntary action, for example, a malicious lie . . . . First of all, we endeavor to discover the motives to which it has been due, and then, secondly, we proceed to determine how far the action . . . can be imputed to the offender. As regards the first question, we trace the empirical character of the action to its sources, finding these in defective education, bad company, in part also in the viciousness of a natural disposition insensitive to shame . . . . We proceed in this enquiry just as we should in ascertaining for a given natural effect the series of its determining causes. But although we believe the action is thus determined, we nonetheless blame the agent.[1]

In the Grounding, he reiterates this: “everything which takes place [is] determined without exception in accordance with laws of nature.”[2] And in the second Critique, he insists that if we knew the relevant preconditions, “we could calculate a human being’s conduct for the future with as much certainty as a lunar or solar eclipse.”[3]

At the same time, Kant insists that human beings are transcendentally free, uncaused causes of their actions: “a rational being can . . . say of every unlawful action he performed that he could have omitted it.”[4] Kant’s defense of freedom depends on his transcendental idealism, according to which even though our actions are determined by natural law, they are nonetheless free. Allen Wood aptly describes this as a “compatibility of compatibilism and incompatibilism,”[5] explaining that while human actions can, on the one hand, be explained by “empirical observation and natural science, and “from this standpoint, our actions are causally determined,”[6] those actions also have a ground that is not capable of being observed, and that ground may be free.[7]

Kant’s theory of freedom has important implications for his empirical psychology. In a lecture on metaphysics, Kant says,

Freedom cannot be proven psychologically, but rather morally . . . . If I wanted to prove freedom psychologically, then I would have to consider a human being . . . as a natural being, and as such he is not free.[8]

Kant makes room for human freedom transcendentally, not psychologically, and thus his psychological account of human action is left thoroughly deterministic.[9] But while Kant’s theory of freedom has received a great deal of attention from contemporary philosophers, his empirical psychology has not been studied in much detail. There is still a need for a clear explanation of how Kant thinks human action can be explained from an empirical perspective.

In this paper, I lay out a clear description of Kant’s determinist account of human action, as found in his empirical psychology. The direct benefit of this description will be to clear up confusion about the relationship between empirical and practical accounts of human action by bringing to light the nature of Kant’s empirical description of human action. This paper will also be crucial for developing more historically accurate and philosophically sophisticated Kantian accounts of the emotions, moral education, cultural and historical influences on human behavior, and the role of psychology and anthropology in Kant’s moral theory.

The need for an explanation of Kant’s empirical psychology is particularly important today because many discussions of Kant’s psychology focus on Kant’s account of moral choice from a practical perspective, and this gives the sense that Kant’s empirical account of action depends on what Simon Blackburn calls a “Kantian Captain,” “free of his or her natural and acquired dispositions.”[10] When Andrews Reath, for example, discusses Kant’s “theory of motivation,”[11] he starts his account with a treatment of respect for the moral law, rather than a detailed discussion of how non-moral motives function. The result is that presuppositions of Kant’s moral theory unduly influence Reath’s psychology, so that he insists on finding freedom within a Kantian account of motivation. Thus he objects that “if the moral law determines choice by exerting a force that is stronger than the alternatives, moral conduct will result from the balance of whatever psychological forces are acting on the will . . . . It is not clear that this model leaves room for any real notion of will or choice.”[12] But given Kant’s transcendental idealism, such a model does leave room for real choice, not within the model itself, but from the perspective of practical reason. More importantly, Kant’s transcendental idealism, as articulated in the first Critique (especially the Second Analogy[13] and the Third Antinomy), shows that it must be possible to give a causal picture of the kind that Reath opposes.

Similarly, Marcia Baron argues against “speaking of acting from the motive of duty” on the grounds that “the term ‘motive’ suggests causation.”[14] She goes on:

This [causal picture of agency] is a familiar picture of agency from the empiricist tradition. Kant’s theory of agency is very different. Our actions are not the result of a desire or some other incentive that impels us. An incentive can move us to act only if we let it.[15]

As evidence for this alternative Kantian picture, Baron appeals to an important passage in the second Critique in which Kant argues for freedom on the grounds that one is always conscious of an obligation that ought to – and therefore can – be obeyed.[16] Baron is certainly correct, of course, that for Kant human beings are free and therefore incentives can move only if we let them. But this account of agency is an account of transcendental freedom, an account that, for Kant at least, is consistent with the familiar empiricist picture of agency. There is nothing wrong, of course, with focusing on looking at agency from the standpoint of freedom. Kant insists that this is the proper standpoint for moral philosophy. But by implying that there is a conflict between the freedom necessary for moral agency and empiricist accounts of motivation, Baron, like Reath and Blackburn, mistakenly ascribe to Kant’s empirical psychology a kind of freedom that Kant thinks is out of place there.[17]

Including freedom within third-personal descriptions of human action results in confusion about what Kant’s causal account of human action actually is, and this has left room open for critics of Kant such as Blackburn to accuse Kant of having an overly simplistic account of human psychology. Blackburn’s “Kantian Captain” is introduced as a prelude to his account of “the fundamental mistake about deliberation” that this Captain represents.[18] Blackburn points out that Kant’s Captain is just bad psychology. But Kant, like Blackburn, distinguishes between psychological claims and first personal moral claims, between “speaking from within a moral perspective” and “describing those who speak from within it.”[19] And Kant’s “Captain” has no place in descriptions of human action, but only in discussion of what it means to speak from within a moral perspective. Thus the Captain has no role in empirical psychology, so it cannot be an example of bad psychology. Unfortunately, simplified accounts of Kant’s psychology make it too easy for Blackburn and others to dismiss Kant. And Blackburn rightly claims that “Kant, or perhaps his translators, cannot escape responsibility for the confusion here.”[20] Kant bears some responsibility for not laying out his empirical psychology systematically in any published works. His commentators bear responsibility for never making use of the published and unpublished resources that Kant did leave.[21] By drawing on these resources to explain Kant’s empirical psychology, this paper will help to clear up some of the confusion that prompts overly swift dismissals of Kant’s moral theory.

There are two main aspects of Kant’s empirical account of human action. The first is rooted in Kant’s engagement with 18^th century faculty psychology. In the context of a tradition that describes the soul as involving appetitive and cognitive faculties, Kant develops an account of relationships among three main faculties of soul: desire, feeling, and cognition. This provides Kant with an opportunity to explain human action as the result of a faculty of desire and to explore the causes of various kinds of desires. Kant’s most detailed accounts of this faculty psychology are found in his lectures on empirical psychology, part of his lectures on metaphysics.[22] The second aspect of Kant’s account of human action comes from his engagement with emerging theories in biology and natural history and involves Kant’s account of natural predispositions that underlie human actions. This second aspect is necessary to understand both the nature and the limits of Kant’s causal accounts. The primary sources for this aspect of Kant’s account are taken from his anthropology, including his historical essays, lectures on anthropology, and his published Anthropology from a Pragmatic Point of View.

The structure of this paper is as follows: In section 1, I discuss Kant’s faculty psychology to lay the groundwork for the rest of my discussion. In sections 2 – 4, I explain the predispositions underlying different faculties of soul. This account of predispositions provides a Kantian account, as far as it goes, of the causal connections between cognitions and desires. Sections 5 – 6 complete Kant’s causal account of action by explaining the causal laws governing the cognitive faculties. The result is a complete causal account of human actions, from the experiences that give rise to them to the actions themselves.

1) Kant’s faculty psychology

Kant’s faculty psychology developed in response to three main trends in 18^th century philosophy: Wolff’s Leibnizian rationalism, Crusius’s Pietist response to Wolff, and British empiricism.[23] The overall structure of Kant’s empirical psychology is largely set by Wolff, who developed a faculty psychology in order to reduce diverse faculties to “representation” as the single “essence” of the soul. Kant’s course on metaphysics was based on the textbook of the Wolffian Alexander Baumgarten, who followed Wolff in the organization of empirical psychology. Kant takes over Wolff’s and Baumgarten’s distinctions between different faculties of soul but resists their attempts to reduce these faculties to a single essence.

Instead, Kant follows Crusius in resisting this reduction, and he shifts from a two-fold distinction between cognitive and appetitive faculties to a threefold distinction between the faculties of cognition, feeling, and desire.[24] Each of these three faculties includes several distinct basic powers, none of which is reducible to others. For Kant, the classification of different basic powers is important because “the concept of cause lies in the concept of power.”[25] Different powers reflect different specific laws of causation. Within the faculty of cognition, for instance, Kant includes distinct basic powers such as the five senses, imagination, understanding, and reason, each of which is governed by its own set of causal laws.[26] Because basic powers are the key to any causal explanation of phenomena, Kant claims that “all physics, of bodies as well as of minds, the latter of which is called psychology, amounts to this: deriving diverse powers, which we know only through observations, as much as possible from basic powers.”[27] In both physics and psychology, Kant’s goal is to reduce the variety of observable phenomena to as few basic powers as possible and to explain the laws according to which those powers operate. In psychology in particular, we are to seek “natural laws of the thinking self” based on “observations about the play of our thoughts.”[28] The result is a clear and comprehensive causal account of natural phenomena, whether bodies in physics or minds in psychology.

Finally, from British empiricism Kant adopted the practice of explaining each power in terms of laws describing regular connections between phenomena.[29] Unlike many of the British empiricists, Kant does not focus on laying out causal laws and applying them to understand various mental phenomena. Kant’s focus is on the framework of basic powers for which causal laws will have to be found. An even more important difference is that Kant does not think that an empirical account of these basic powers provides any basis for epistemology or ethics. But when Kant does describe the laws governing the basic powers, his laws are similar to those of the empiricists, including a “law of association” governing the imagination and various laws of logic and prejudice governing the understanding.[30]

When it comes to explaining human action, Kant focuses his account on desire: “all desires have a relation to activity and are the causality thereof.”[31] Insofar as a representation is the ground of an action that brings about some state of affairs, it is a desire: “the faculty of the soul for becoming cause of the actuality of the object through the representation of the object itself = . . . the faculty of desire.”[32] An “object” here is not necessarily a physical object but anything that can be desired, including physical objects but also states of affairs.[33] The object of desire is a possible purpose for an action, and desires, for Kant, naturally give rise to actions. As Kant puts it here, a desire is defined as a representation that leads to action, that “becomes cause of the actuality of an object.” And when desire is taken in this broad sense, there are no actions that are not preceded by and caused by desires.

Thus for explaining human action, the most important task of empirical psychology is tracing the causes of desires. Within this psychology, Kant engages in this task by connecting the faculty of desire with the other basic faculties of the soul. For Kant, this relationship is fairly straightforward:

Pleasure precedes the faculty of desire, and the cognitive faculty precedes pleasure . . . . [W]e can desire or abhor nothing which is not based on pleasure or displeasure. For that which give me no pleasure, I also do not want. Thus pleasure or displeasure precedes desire or abhorrence. But still I must first cognize what I desire, likewise what give me pleasure or displeasure; accordingly, both are based on the cognitive faculty.[34]

According to this structure, cognition of an object gives rise (at least sometimes) to a feeling of pleasure or pain, and that feeling gives rise (again, at least sometimes) to a desire or aversion for the object. We can trace the series of causes as follows:

Cognition à Feeling (pleasure or pain) à Desire à Action[35]

For example, one tastes a mango (cognition), that taste gives one pleasure (feeling), that pleasure causes one to desire the mango, and that desire leads one to eat (or continue eating) the mango.[36]

Within this picture, feelings of pleasure and displeasure function as transitions from cognitions to desires (and thereby to actions). Jeanine Grenberg has described the connection between these “practical pleasures” and desire in detail.[37] As she explains,

[P]ractical pleasure is itself necessarily related to and is indeed the very vehicle for the expression of the status of an agent’s faculty of desire. Technically . . . there is a distinction to be made between feeling and desire: feeling, an element of the agent’s sensible nature . . . “determines” (bestimmt) the faculty of desire . . .. For the purposes of describing action, there is, however, little distinction to be made between the possession of a practical pleasure and that of a desire.[38]

Like Grenberg, I will not focus on the distinction between feeling and desire in the rest of this paper. Given a practical pleasure, a desire will follow simply because of the nature of practical pleasures. The challenge for giving a causal account of human action is to explain the origin of those practical pleasures.[39]

In this context, the task of giving a causal account of human action involves two crucial components. First, Kant needs to provide causal laws governing the connections between cognitions and feelings/desires. This task is complicated because not all cognitions lead to feelings of pleasure or pain, and not all feelings lead to desire or aversion. Even within those cognitions that do affect desire, some lead to desires and others lead to aversions. Thus Kant needs some account of why the series goes through in some cases and not others, and why it leads to the conative state that it does. Second, even if Kant gives a causal account of these connections, he needs to explain the origins of cognitions. Since cognitions start the series, as it were, a causal account of cognitions themselves is required for Kant’s account of action to be thoroughly deterministic. Providing these two components of a causal account of action depends on a more detailed faculty psychology and a treatment of the role of predispositions in human nature. In the rest of this section, I lay out the requisite details from Kant’s faculty psychology. In the next several sections (2 – 4), I explain the role of predispositions as causal bases of connections between cognitions and desires. In the last two sections, I discuss the causal laws governing cognitions.

In his empirical psychology, Kant’s approach to both the origin of cognitions and their connection with desire involves further distinguishing between different faculties of soul. I have already noted the important distinction between the faculties of cognition, feeling, and desire, which forms the background for Kant’s overall account of human action. Cutting across this three fold distinction, however, is a further distinction – adopted from Baumgarten – between “higher” and “lower” faculties of cognition, feeling, and desire. The “lower” faculties are primarily receptive. The higher faculties are “self-active” or “spontaneous.”[40] (In these contexts, Kant generally[41] uses the terms “self-activity” or “spontaneity” to describe an empirical or comparative freedom of the higher faculties, a freedom that is consistent with the view that even those higher faculties are causally determined.[42] When Kant talks about this freedom in the context of his empirical psychology, it is empirical freedom to which he refers, and this leaves room for Kant to give a causal account of even the higher faculty of desire, as we will see in section four.) Thus we can broadly outline six different “faculties” in Kant’s empirical psychology: higher and lower cognition, higher and lower feeling, and higher and lower desire.

With respect to cognition, the lower faculty is referred to broadly as “sensibility” (Sinnlichkeit) and includes the senses (Sinne) and the imagination, each of which is further subdivided.[43] The senses include the five outer senses as well as inner sense, and the imagination includes memory, anticipation of future events, and the “productive” or “fictive” imagination. This higher faculty of cognition is often referred to by the general term “understanding” (Verstand) and includes three specific cognitive powers: reason, the understanding (Verstand) in the narrow sense, and the power of judgment.[44]

Just as “there is a higher and a lower cognitive faculty[,] so there is also a rational and a sensible feeling of pleasure or displeasure (and so it is also with the faculty of desire).”[45] For the present purposes, I focus on Kant’s account of the higher and lower faculties of desire, as these are most directly tied to action, but most of Kant’s account of higher and lower desire applies to feeling as well.[46] As with the cognitive faculty, the distinction between the higher and lower faculties of desire is based on the distinction between the senses and the understanding: “all desires are . . . [either] intellectual or sensitive.”[47] But in the case of desire, what is relevant is not the nature of the desire itself but the cognitive cause of the desire.[48] “The representations which produce determinations [of desire] are either sensible or intellectual.”[49] Insofar as a desire is the direct result of the senses or unmediated imagination, is it part of the “lower” faculty of desire. Insofar as it proceeds from the understanding or reason, a desire falls under the “higher” faculty of desire. The key difference here is between motivation by immediate intuitions and motivation by principles or concepts. As Kant explains, every desire[50]

has an impelling cause. The impelling causes are either sensitive or intellectual. The sensitive are stimuli <stimuli> or motive causes [Bewegungsursache], impulses. The intellectual are motives [Motive] or motive grounds [Bewegungsgrunde] . . . . If the impelling causes are representations of satisfaction and dissatisfaction which depend on the manner in which we are [sensibly] affected by objects, then they are stimuli. But if the impelling causes are representations of satisfaction or dissatisfaction which depend on the manner in which we cognize the objects through concepts, through the understanding, then they are motives.[51]

The distinction between higher and lower faculties of desire is critically important for Kant’s overall account of human action because the causal mechanisms governing desire operate quite differently depending on whether they belong to the higher or lower faculty. Though both faculties are determined by “impelling causes” or “incentives,”[52] the higher faculty is determined by “motives” which proceed from the understanding and the lower faculty is determined by “stimuli” that proceed from the senses.[53] As we will see in sections 3 and 4, the difference between higher and lower faculties of desire is reflected in different predispositions that underlie Kant’s causal accounts of each faculty.

Before moving on to the next section, it is worth drawing attention to one further distinction within Kant’s account of the higher faculty of desire. “Higher” desires are caused by one or more kinds of higher cognition, but these desires need not be purely rational. Although all “higher” desires have “grounds of determination . . . [that] lie . . . in the understanding,”[54] these desires can be “either pure or affected.”[55] Kant explains this distinction as follows:

The intellectual impelling cause is either purely intellectual without qualification <simpliciter . . .>, or in some respect <secundum quid>. When the impelling cause is represented by the pure understanding, it is purely intellectual, but if it rests on sensibility, and if merely the means for arriving at the end are presented by the understanding, then it is said to be in some respect.[56]

For a desire to be purely intellectual, it must be caused by the pure understanding, or pure reason (recall that the understanding in the broad sense includes reason). But a desire can be directly caused by higher cognition without being caused by pure reason when someone acts on the basis of a principle of the understanding that is directed towards fulfilling some sensible desire (or inclination). Such impure higher desires proceed from representing to oneself hypothetical imperatives as principles for action. These impure desires are still “higher” desires, however, because they are caused not solely by sensible desires but by principles or concepts of the understanding directed towards satisfying such desires.[57] The pure higher faculty of desire, because it involves desires that follow from purely rational considerations, issues from the representation of categorical imperatives.[58]

We can now summarize the results of this section by filling in Kant’s taxonomy of faculties of the soul:

The faculties (and powers) of the soul

	Cognition (representations)	Feeling (pleasure and displeasure)	Desire (impelling grounds, incentives)
Higher (intellectual, active, spontaneous)	Understanding (including the distinct powers of judgment, understanding, and reason)	“satisfactions or dissatisfactions which depend on the manner in which we cognize the objects through concepts”	Motives, motive grounds (including both pure and impure motives)
Lower (sensible, passive, receptive)	Sensibility (including distinct powers of the senses and imagination)	“satisfactions and dissatisfactions which depend on the manner in which we are [sensibly] affected by objects”[59]	Stimuli, motive causes, impulses.

For Kant, this taxonomy is the first step in giving causal laws for mental phenomena because each distinct mental power will be governed by its own causal laws. (Recall that “the concept of cause lies in the concept of power”[60]). Thus Kant’s empirical psychology must explain the underlying causal mechanisms for the origin of each kind of cognitive state as well as the mechanisms for connecting those cognitive states to the states of feeling and desire to which they give rise. The next section offers a crucial further component of that explanation.

2) Human predispositions and the limits of mechanist explanation

So far, we have seen that Kant develops his empirical account of action in the context of a faculty psychology. Kant traces the sources of particular desires to their connection with other mental states. But Kant also offers a more general account of the bases for these connections. Human beings desire some things rather than others, and this is not simply because we cognize some things rather than others. We often have cognitions that do not give rise to feelings, and feelings that do not give rise to desires.[61] To flesh out his naturalistic explanation of human action, Kant explains why some cognitions but not others give rise to desires.

As with his psychology in general, Kant’s approach to predispositions is empirical and taxonomic. Kant accounts for human tendencies with his fundamental notion of a “natural predisposition” (Naturanlage).[62] He does not give many causal accounts of the origins of these predispositions. They are purposive tendencies in human nature that should be classified but cannot (easily) be explained. As he explains in his “Conjectures on the Beginning of Human History,” “we must begin with something that human reason cannot derive from prior natural causes – that is, with the existence of human beings,” including all of their natural predispositions.[63]

In an important respect, the positing of predispositions reflects a backing off from the implicit commitment of his first Critique to the possibility of fully mechanistic accounts of human action. Even Kant’s faculty psychology itself reflects a limitation of mechanical explanation, in that Kant avoids the Wollfian attempt to reduce the powers of the soul to a single kind of power governed by a single kind of law. In fact, as early as 1782, Kant makes explicit a kind of scientific modesty when it comes to mechanistic explanations of phenomena.[64] In a revealing comparison of Descartes and Newton, Kant distinguishes two modes of study in the physical sciences in a way that points out the danger of allowing the demand for simplicity to govern scientific explanation.

There are . . . two physical modes of explanation: (1) mechanical philosophy, which explains all phenomena from the shape and the general motive power of bodies….

(2) The dynamical mode of explanation, when certain basic powers are assumed from which the phenomena are derived. This was first discovered by Newton and is more satisfactory and complete than the former. Thus to explain something mechanically means to explain something according to the laws of motion, dynamically, from the powers of bodies. With either explanation one never comes to an end. The correct mode of explanation is dynamical physics, which includes both in itself. That is the mode of explanation of the present time. The first is the mode of explanation of Descartes, the second that of the chemists.[65]

Descartes errs, according to Kant, because he overemphasizes the reduction of phenomenal explanation to a single power (the “general motive power”). By contrast, Newton and the chemists rightly postulate additional basic powers when these are necessary to explain diverse phenomena. What is more, Kant highlights here that both kinds of explanation take for granted certain causal powers, and these are left unexplained.

Kant’s psychology follows the example of the chemists rather than mechanist physics. His focus is on not overly reducing powers to a single basic one. As he says, “for our reason there must be several [basic powers] because we cannot reduce everything to one.”[66] But with respect to psychology, and biology more generally, Kant develops even more fundamental reasons for questioning the attempt to reduce all explanations to simple mechanistic ones. Kant’s modification of his mechanistic ambitions with respect to human psychology is incorporated into a general realization that the study of “organized beings” – including humans as well as birds[67] and even grass[68] – cannot always proceed mechanically. This limitation on mechanistic explanation is particularly important in the context of explaining the origins of organized beings or their specific natural capacities.[69]

In this respect, however, Kant is simply modifying a dominant strategy of 17^th and 18^th century mechanists for dealing with the generation of living beings. As Phillip Sloan has pointed out, “after 1660 . . ., mechanistic solutions to the problem of organic generation typically involved some version of preformationism.”[70] Preformationism has a strong form, according to which every living thing was created at the creation of the world and are simply encased, like “Russian dolls”[71] in either ovaries or spermatazoa. The version with which Kant was more sympathetic early in his career involved preformed “germs” (Keime) that unfold into living organisms.[72] Either version, however, retains mechanist accounts of change within nature while bracketing the problem of finding a mechanist origin for life itself. Life itself is simply treated as given, or “preformed.” The only other mechanist strategy for explaining life was mechanist “epigenesis” – according to which one gave mechanical explanations for the emergence of organic structures from inorganic nature. But by the end of the 17^th century, this approach simply looked implausible from a scientific standpoint, and in the years between the first and third Critiques, Kant came to a deeper realization of its limits.[73]

Thus in his third Critique, Kant argues against a purely mechanist account of life. In a now famous passage, he explains,

It is quite certain that we can never adequately come to know the organized beings and their internal possibility in accordance with merely mechanical principles of nature, let alone explain them; and indeed this is so certain that we can boldly say that it would be absurd for humans even to make such an attempt or to hope that there may yet arise a Newton who could make comprehensible even the generation of a blade of grass according to natural laws that no intention had ordered; rather, we must absolutely deny this insight to human beings.[74]

Kant’s account in the third Critique is not wholly negative, however. Instead, Kant finds a new way to preserve a commitment to mechanistic accounts of the world in the face of the scientific obstacles presented by biology. Although Kant initially was sympathetic with a sort of moderate preformationism, according to which there are basic “germs” (Keime) in nature that develop over time, in the third Critique, he changes his view in two important respects.[75] First, he shifts from language of germs (Keime) to the more epigenetic language of “predispositions” (Anlagen). Unlike germs, which simply unfold into organic structures, Anlagen are “dynamic, purposive predispositions”[76] in the context of which organisms grow and develop. This conception of Anlagen fits well with non-mechanist conceptions of epigenesis prevalent in the second half of the 18^th century, according to which organisms arise from interactions between various vital forces. In Blumenbach, for example, epigenesis involves the evolution of organisms from a basic “formative force” (Bildungstrieb). The emphasis on predispositions rather than germs marks Kant’s shift towards a vitalist epigenesis rather than the more mechanist – but less explanatory – preformationism. Even this concession to epigenesis, however, still involves postulating organic structures – the predispositions themselves – that fit into a “system of generic preformation, since the productive capacity of the progenitor is still preformed.”[77]

The second and more significant change in Kant’s biology in the third Critique is that Kant’s lingering preformationism is given a status that makes it compatible with a thoroughgoing and mechanist epigenesis. In the third Critique, Kant’s pessimism about finding a “Newton . . . of a blade of grass” is not a rejection of the claim that even the generation of a blade of grass must in fact be causally determined according to natural laws. As Kant insists, “the principle that everything that we assume to belong to nature (phaenomenon) and to be a product of it must also be able to be conceived as connected with it in accordance with mechanical laws nonetheless remains in force.”[78] And even in scientific inquiry, we have an “obligation to give a mechanical explanation of all products and events in nature, even the most purposive, as far as our capacity to do so” allows.[79] But as a “heuristic principle for researching the particular laws of nature,”[80] one can add to the principle of mechanical causation a “principle of final causes”[81] “in order to supplement the inadequacy of [mechanical explanation] in the empirical search for particular laws of nature.”[82] According this heuristic principle, “nothing in [an organized product of nature] is in vain, purposeless, or to be ascribed to a blind mechanism of nature.”[83] By relegating his biology to the status of a heuristic, a “maxim of the reflecting power of judgment”[84] that “is merely subjectively valid,”[85] Kant is free to adopt a vitalist epigenesis that explains natural organisms in terms of purposive structures. This purposive account can provide relatively satisfying biological explanations without undermining the fundamental principle that every event in the world in fact has mechanical causes.[86]

When Kant turns to human predispositions in particular, he remains faithful to his biological insight that these predispositions are best explained in terms of how they fit into an overall system of natural purposes. Throughout his Anthropology, Kant’s accounts of various natural powers are teleological rather than strictly mechanical. Thus with respect to the natural capacity for dreaming, for example, Kant says that “dreaming seems to be essential: unless dreams always kept the vital force active during sleep, it would go out,” and that “illusion” by which someone “who is naturally lazy” mistakes “objects of imagination as real ends” is a means for “Nature . . . [to] make us more active and prevent us from losing the feeling of life.”[87] These teleological explanations of various natural endowments are not strictly a part of Kant’s mechanist account of human action. Rather, they reflect a different kind of causal explanation. For the purposes of this paper it is not necessary to discuss these teleological explanations of human predispositions in detail. Instead, what is important is the way natural predispositions, once granted, function in Kant’s overall account of the causes of human actions. Although Kant may not give causal accounts of the origins of these predispositions, he effectively uses predispositions themselves to ground causal accounts of particular human actions. In the next two sections, I show how Kant uses predispositions to explain the causal connections between cognitions and desires, and thereby to explain human action.

3) Human predispositions in Kant’s theory of action: the lower faculty of desire

Once he allows for the use of predispositions in explaining the development and behavior of living things, Kant draws on predispositions to provide the background against which to give his causal account of human action. Predispositions fill in the explanation for why certain cognitions lead to desires, others to aversions, and others to no appetitive response at all. Thus to explain why the smell of a particular food gives rise to a desire for it whereas other smells do not give rise to any desire,[88] or why certain kinds of social interactions are pleasant and others are not, Kant appeals to basic predispositions in human nature. A full Kantian account of human action, then, must classify and give law-like form to the predispositions that underlie connections between various faculties of soul.

Kant describes basic predispositions in different ways for each different faculty of the soul. For our purposes, the most important distinction is between the higher and lower faculties of desire, since these predispositions are most directly involved in action. In the rest of this section, I make use of Kant’s biology and anthropology to give a fuller account of the causes of action in the case of the lower faculty. In the next section, I discuss the causes of action for the higher faculty. Because the connections between feeling and desire and between desire and action are relatively straightforward, as noted in section one, I simplify my account and focus on the connection between cognitions and feelings/desires.

For Kant, predispositions are not simply additional causes of a mechanical kind. Kant does not conceive of explanation here as simply adding more motive causes. Rather, he seeks to rigorously explain in the “dynamical mode” of Newton, classifying and characterizing the bases for basic powers of the soul. In this context, the role of predispositions will not be strictly causal. That is, one does not simply add a predisposition to a cognition in order to cause a desire, such that

Cognition + predisposition x à Desire

Rather, predispositions for Kant play something like the role that gravity plays in Newton’s account, where it is not the case that

Earth’s mass + Gravity à falling of apple

That is, gravity is not just another cause, like the mass of the earth. Instead, gravity is what explains why the mass of the earth causes the apple fall.[89] I capture this different kind of explanation with a vertical arrow (↑). Thus for Newton,

Earth’s mass à falling of apple

↑

Gravity

For Kant, this is the kind of explanation that will be needed for why certain cognitions lead to desires and others do not, and predispositions will play a crucial role in these explanations.

With respect to the lower faculty of desire, Kant explains the role of natural predispositions in connecting cognitions and desires terms of instincts, propensities, and inclinations.[90] The role of instincts in explaining human action is the most straightforward, so I will start with it. In his Anthropology, Kant explains, “The inner necessitation of the faculty of desire to take possession of [an] object before one is familiar with it is instinct.”[91] In his lectures on anthropology, Kant claims that “instincts are the first impulses according to which a human being acts.”[92] Kant is careful to distinguish instincts from acquired desires, and generally warns against “multiplying instincts among human beings” in our explanations of human behavior. Nonetheless, he gives ample examples of instincts throughout his lectures and published writings, including the sexual instinct, the parental instinct to provide for young, the “sucking instinct” of infants, instincts for various foods, the “natural instinct to test [one’s] powers,” and natural sympathy, which Kant treats as an instinct.[93]

The clearest example of the way in which instincts function in causal explanations of human behavior is from Kant’s short essay, “Conjectural Beginning of Human History,” where Kant discusses the role of instinct in determining which foods the earliest human beings would have eaten. He says,

Initially, the newcomer must have been guided solely by instinct, that voice of God that all animals obey. It permitted him to use some things as food and forbade him to use others. – It is unnecessary, however, to assume for this purpose a particular long-lost instinct. It could simply have been the sense of smell and its affinity with the organ of taste, along with the well-known sympathy between the latter and the digestive organs – in other words an ability . . . to sense in advance whether a given food is suitable for consumption or not.[94]

In one respect, Kant’s treatment of instinct here is atypical, in that he provides some explanation of the causes of the particular connections between sensing a particular food and desiring it, through the relationship between smell, taste, and digestion. With respect to other instincts, Kant more often makes appeal to “particular instincts” without further inclination. But in other respects, his account here is typical. A human being has a sensory – here olfactory – cognition of a particular food, and this cognition gives rise to a desire because of an instinct for that particular food. Similarly, Kant explains in the case of the sexual instinct that “as soon as one comes into society, one’s instinct will quickly find an object.”[95] In both cases, the appeal to instinct explains why the mere sensible cognition – of food or another person – becomes a desire of a particular kind. Thus one can expand Kant’s account of action from section one as follows, at least for some cases:

Sensory cognition à Feeling/Desire

↑

Instinct

For example,

Smell of a ripe mango à Desire for that mango

↑

Instinct for mangos

In cases where instinct explains the connection between cognition and desire, the task of explanation is sufficiently complete when one classifies the different instincts and describes the laws of their operation. In the context of an account of human instincts, one can causally explain any desire by appealing to the instinct as a natural predisposition (here the instinct for mangos) and the activating cause of that instinct (here the smell of mangos).

While Kant thinks that instincts explain some human actions, he does not explain most actions in terms of instinct. Even most desires associated with the lower faculty of desire are not explained by reference to instincts, but by reference to inclinations. Unlike instincts, which are relatively few in number, the types of inclination are too many and too varied to give even a partial list. Inclinations cover a wide range of human desires, from inclinations for smoking and drinking[96] to love as an inclination[97] to inclinations for honor, money, and power.[98] When explaining actions in terms of inclinations, Kant’s model is similar to that for instincts. Like instincts, an inclination is “a lasting ground of desire”[99] or a “subjective necessity of desiring.”[100] Thus for the case of an inclination to “strong drink,”[101] for example, we get:

Sensory cognition à Feeling/Desire

(sight or smell of strong drink) (desire to consume the drink)

↑

Inclination (for strong drink)

Unlike instincts, however, inclinations are not themselves natural predispositions, and thus Kant’s causal story cannot end with this picture. Inclinations are acquired, so for Kant’s account to be complete, he needs to explain the causal origin of the inclination itself.

Kant’s explanation of the causal origin of inclinations is fairly straightforward: we acquire inclinations by past experience, which develops a habitual desire, or more properly a “habitual ground [Grund] of desires.”[102] In some cases this relevant past cause of the inclination need only be a single instance of experiencing the relevant object of desire.[103] At other times, developing an inclination depends on “frequent repetition”[104] of experiencing the object of inclination. Thus a more complete account of inclination-based desires (taking drink as an example again) is as follows:

Sensory cognition à Feeling/Desire

(sight or smell of strong drink) (desire to consume the drink)

↑

Past experience with strong drink à Inclination (for strong drink)

And now, of course, there is another causal connection – between past experiences of an object and the inclination for that object – that needs to be explained. To explain that connection, Kant appeals to a different kind of natural predisposition, a “propensity” (Hang). As Kant explains, “Propensity . . . is the inner possibility of an inclination, i.e. the natural predisposition to the inclination.”[105] A propensity is a “subjective possibility of generating a certain desire,”[106] which “can be found even when there is not yet the actual desire.”[107] For example, Kant claims that “northern peoples have a propensity to strong drink,”[108] and in the Religion, he clarifies what this means (changing the relevant people-group!)[109]:

Propensity is actually only the predisposition to desire an enjoyment which, when the subject has experienced it, arouses inclination to it. Thus all savages have a propensity for intoxicants; for although many of them have no acquaintance at all with intoxication, and hence absolutely no desire for the things that produce it, let them try these things but once, and there is aroused in them an almost inextinguishable desire for them.[110]

Inclinations are not themselves predispositions; rather, they are the result of experiences of objects for which someone has a propensity. A “northern person” who experiences strong drink will develop a habit of desiring such drink; such a habit is an inclination.

In explaining a particular human action, then, one can appeal to instincts or inclinations to explain why a particular cognition gives rise to a desire, whereas another does not. For the lower faculty of desire, there are two different models for the causal origin of a desire:

(1)

Sensory cognition à Feeling/Desire[111]

(smell of mango) (desire to eat mango)

↑

Instinct* (for mangos)

(2)

Sensory cognition à Feeling/Desire

(sight or smell of strong drink) (desire to consume the drink)

↑

Past experience (with strong drink) à Inclination (for strong drink)

↑

Propensity* (for strong drink)

Again, the form of this causal account is that sensory cognitions provide the immediate natural cause of the relevant desires, and instinct or inclination provides the ground or explanation of why that cause functions in the way that it does.[112] Instinct and propensity are marked with *s to indicate that these are natural predispositions for which Kant does not give a mechanistic explanation (for reasons explained in section two).

4) Human predispositions in Kant’s theory of action: the higher faculty of desire

When turning to the higher faculty of desire, the underlying explanation for the connection between cognitions and desires is character. Kant uses the term character in several senses throughout his writings, and it is important to keep those distinct here. In the broadest sense, the character of a thing is the “law of its causality, without which it would not be a cause at all,” such that “every effective cause must have a character.”[113] In this sense, gravity reflects the “character” of matter, and one’s instincts are part of the “character” of one’s faculty of desire. In a quite different sense, Kant uses “character,” in the context of one’s “intelligible character,” to refer to the free ground – “which is not itself appearance” – of one’s appearances in the world.[114] Character in this sense has no role to play in any sort of empirical explanation of action. One can not, in particular, empirically explain actions that flow from the higher faculties of cognition in terms of intelligible character.

The sort of character that plays an important role in Kant’s empirical theory of action is distinct from intelligible character and more specific than the character of an efficient cause in general. Kant refers to this sense of character in his anthropology as “character simply [Character schlechthin],” and defines it as “that property of the will by which the subject has tied himself to certain practical principles.”[115] This more specific sense of character plays, for the higher faculty of desire, the role that instincts and inclinations play for the lower. In the rest of this paper, I use character in the narrower sense of Character schlechthin.

Just as instincts and inclinations ground a consistent connection between the lower faculty of desire and sensuous cognitions, so character grounds a similar connection with respect to the higher faculty of desire.

Character is a certain subjective rule of the higher faculty of desire . . .. Accordingly, character makes up what is characteristic of the highest faculty of desire. Each will . . . has its subjective laws, which constitute . . . its character.[116]

Kant makes the nature of this connection clearer elsewhere, explaining that “the essential characteristic of character . . . belongs to the firmness of the principles.”[117] A person whose actions are explained by reference to their “character” is someone whose faculty of desire is determined by principles flowing from the higher cognitive faculties. The relevant principles here need not specifically be moral; any principles can be practical in that they guide action. As Kant makes clear in his Anthropology, acting on the basis of firm principles, regardless of the content of those principles, determines whether or not someone has character:

Simply to have a character relates to that property of the will by which the subject has tied himself to certain practical principles . . .. Although these principles may sometimes indeed by false or defective, nevertheless the formal element of will as such, which is determined to act according to firm principles (not shifting hither and yon like a swarm of gnats), has something precious and admirable to it, which is also something rare.[118]

A person who acts from any stable set of principles has character. [119] Kant can thus explain the difference between sensuous people and those with character as follows: “the man of principles, from whom we know for sure what to expect, not from his instinct, for example, but from his will, has character.”[120] Those who act from instinct or inclination and those who act from character are both predictable – we “know what to expect” – but for different reasons. Instincts and inclinations ground a regular connection between lower cognitive states and desires. Character grounds a regular connection between higher cognitive states – principles – and desires.[121]

The picture here looks like this:

Cognition à Pleasure/Desire

↑

Character

Unfortunately, the picture as presented here is incomplete because the origin of character itself is unexplained. Unlike instincts, which are natural predispositions and thus do not need to be explained, “character comes not from nature, but rather must be acquired.”[122] In this respect, character is like inclination, and like inclination, character depends on both a prior propensity – a “propensity to character”[123] – and on various influences that cultivate this propensity into character itself. But whereas Kant gives a fairly straightforward account of origin of inclinations, he gives no equally simple account of the origin of character. This might lead one to think that there is some room for transcendental freedom in this account of higher desire. Kant even makes some claims that seem to suggest that character, unlike inclination, cannot be explained naturalistically. He says, for instance that “having character simply characterizes man as a rational being, one endowed with freedom” and that character “shows what man is prepared to make of himself” as opposed to “what can be made of a man.”[124] Moreover, even if there is some causal account of character, a complete causal account of human action depends on giving an account of the higher cognitions themselves. And while a naturalistic explanation of lower (sensuous) cognitions might seem straightforward, one might question whether it is possible to give a causal explanation of “principles,” and especially of moral principles. In fact, however, while rationality, the higher faculty of desire, and character are all associated with freedom, they nonetheless can all be explained mechanically.[125] In the rest of this section, I take up the issue of the causal origins of character, and in section 6, I discuss the causal origins of the higher cognitions upon which someone with character acts.

In some respects, this mechanical explanation will be similar to that of inclinations. But Kant’s account of the origin of character is much more complex that the account of inclination. For inclinations, Kant’s causal story is quite straightforward: Given the propensity to desire a particular kind of object, one need only experience the object (sometimes multiple times) to develop the inclination. For character, the account is not nearly as simple, but this should come as no surprise. Character explains the connections between highly developed cognitive states and their corresponding desires, and the capacity for character is one of the features that distinguish human beings from animals. Thus it is natural that its causal origin is considerably more complicated than that of inclinations. But this added complexity does not imply that its origin is not causal.

We have already seen that the character, like inclinations, is acquired on the basis of a natural propensity.[126] Just as someone with a propensity to strong drink may never have an inclination to such drink, someone with the propensity to character may never develop a character. But the very need to posit a propensity to character shows that Kant’s account of the origin of character fits within a broader biological account of human behavior in terms of natural predispositions. In that sense, the basis of character is no less natural than the basis of inclination.

Kant’s account of the factors that contribute to the development of character further emphasizes its empirical basis. Kant suggests that some human beings are better prepared for character than others by virtue of other natural predispositions, such as temperament.[127] Kant says, “not every temperament is inclined to adopt a character, e.g. the melancholy one adopts a character first, the sanguine one not so easily,”[128] and of one with a phlegmatic temperament, Kant claims, “without being brilliant, he will still proceed from principles [and hence from character] and not from instinct.”[129] Kant’s discussion of temperaments is “psychological” and empirical, and temperaments can even “be influenced . . . by the physical condition of a person.”[130] Insofar as temperament plays a role in the formation of character, this formation is at least influenced by natural causes.

Kant discusses further aids to the cultivation of character that are not natural endowments but are nonetheless empirical causes. Among these education is the most important.^[131] “The acquisition of good character with people happens through education.”[132] And even when Kant is most insistent that “the act of establishing character is absolute unity of the inner principle of our conduct” and thus “a kind of rebirth,” he points out that “education, examples, and instruction . . . produce this firmness in our principles.”[133] The “transformation”[134] whereby one’s character is established is something that is produced (bewirkt) by education.

Kant’s accounts of character are often accompanied by specific pedagogical recommendations. He suggests that because “imitation . . . greatly hinders character, in education one must never refer one’s children to the neighbors children, . . . but rather build their character directly, [using] principles of good and bad to inspire righteousness and nobility.”[135] Particular details about the kind of education someone receives can influence whether that education leads to true character or mere imitation. And in the Critique of Practical Reason, Kant even discusses the way one can cultivate not just character, but good character, in “a ten-year-old boy.” One can bring this boy to “admiration, and even the endeavor to resemble” a virtuous person.[136] The key to this education, for Kant, is to focus on the purity of the moral law. And Kant’s discussion of education here shows that repeated arguments against heteronomous accounts of the nature of morals have an important pedagogical purpose. As he explains, “every admixture of incentives taken from one’s own happiness is a hindrance to providing the moral law with influence on the human heart.”[137] Kant’s claim here might seem to be an indictment of appealing to empirical causes in explaining moral action. But his point is really that certain kinds of empirical causes – instructions that appeal to happiness – are ineffective.[138] The empirical influence that actually can give the moral law influence over a person’s heart is “a story”[139] within which the moral law is described in its purity. Likewise in the Metaphysics of Morals, Kant gives a “fragment of a moral catechism” as an example of how to teach virtue to children, and even offers an outline of “the experimental (technical) means for cultivating virtue.”[140] Character in general, and good character in particular, cannot be cultivated in the same simple ways that inclinations can be cultivated. Kant’s focus on the details of good character education show how challenging such education is. But this attention to education in the development of character makes sense only because character is something “produced” by the right kind of education.[141]

Education is the most important influence on character, but Kant also mentions other empirical influences. Politeness plays an important role in the cultivation of character by combating passions and promoting self-control.[142] In his writings on history and politics, Kant gives further explanations for how natural inclinations can give rise to various social institutions – including stable and just political regimes, peace, and even progress in the arts and sciences – that may have beneficial effects on character.[143] Thus “experience and history” provide reasons that “we should not despair about our species’ progress toward the better.”[144] The increasing presence of stable political structures and advancing culture, like the presence of stable norms of polite society, can help the cultivation of constancy in principled action. These external influences, like the more direct influence of education, help to “produce” character.[145]

Filling in Kant’s overall account of character, one gets:

The account in the case of higher desires is considerably more complicated than in the case of lower desires (whether those lower desires are explained by inclinations or instincts), but a more complex causal account is still a causal account. Because all actions are caused by either higher or lower desires, Kant has provided, at least in outline, a causal account of the connections between cognition and desire for all human actions. All that remains to have a complete causal account is an explanation of the origin of cognitions themselves. In the next two sections, we turn to that task.

5) Causal accounts of cognition: the lower faculties of cognition

In his empirical psychology and anthropology, Kant distinguishes between the higher and lower faculty of desire on the basis of the causal origin of desires. Insofar as desires are caused by higher cognitions – concepts or principles – those desires are ascribed to the higher faculty of desire and the connection between cognition and desire is explained by reference to character. Insofar as desires are caused by lower cognitions – sensations or imaginings – those desires are lower desires and the connection between cognition and desire is explained by reference to instincts or inclinations. In both cases, a fully causal account of human action must not only explain the connections between cognitions and desires but must also give a causal account of the origin of the cognitions themselves. In this section I argue that Kant is committed to the possibility of such a causal account of cognitions (at least in principle), and I give some details about the nature of that causal account in Kant. As in the previous section, I divide my discussion here according to the higher and lower faculties, and I discuss the lower faculty (of cognition) first.

In Kant’s lectures on empirical psychology, his accounts of the faculties of cognition are generally brief, and Kant sometimes refers explicitly to his anthropology for further information about the nature of cognitive powers. Thus in a late (probably 1790-1) lecture on empirical psychology, Kant briefly discusses the cognitive faculty, but adds that “anthropology will treat of this in more detail.”[146] Unfortunately, although Kant’s pragmatic anthropology and related lectures on anthropology give more detail about these faculties, they do so in a particular context. Kant’s pragmatic anthropology presents causal accounts not simply for the sake of information but in order to put to use those causal accounts of one’s faculties. In the introduction to his published Anthropology, Kant explains this difference between speculative theorizing and pragmatic anthropology:

If we ponder natural causes – for example, the possible natural causes behind the power of memory – we can speculate to and fro (as Descartes did) about traces, remaining in the brain, of impressions left by sensation we have experience. But since we do not know the cerebral nerves and fibers or understand how to use them for our purposes, we still have to admit that we are mere spectators at this play of our ideas and let nature have its way. So theoretical speculation on the subject is a sheer waste of time. – But when we use our observations about what has been found to hinder or stimulate memory in order to increase its scope or efficiency, and need knowledge of man for this purpose, this is part of anthropology for pragmatic purposes; and that is precisely what concerns us here.[147]

For Kant, both pragmatic anthropology and theoretical speculation lay out various natural causes of mental faculties. The difference is that pragmatic anthropology focuses on causes that it can use to improve those faculties.

In that context, Kant’s causal accounts of the nature and function of the cognitive powers is focused on two important tasks. First, Kant lays out causal laws that govern the normal and healthy functioning of the cognitive powers. Second, he describes ways that these cognitive powers can be improved, focusing in particular on various problems or defects that can arise with respect to one’s cognitive powers and how (if at all) these can be treated or prevented.

The lower faculty of cognition involves the senses and the imagination, and Kant explains various causal laws governing the behavior of both. With respect to the senses, Kant is brief. Unlike Hume, however, who simply argues that “impressions . . . of sensation . . . arise in the soul originally, from unknown causes,”[148] Kant at least offers some explanation of the origin of sensory ideas.[149] He says, for instance, that “the sense of touch lies in the fingertips and the nerve endings (papillae) and enables us to discover the form of a solid body by means of contact with its surface” and that “sight is a sense of indirect perception appearing to a certain organ (the eyes) sensitive to agitated matter, namely light, which . . . is an emanation by which the locus of an object in space is determined.”[150] Kant gives similar accounts of hearing, smell, and taste. The general accounts of purely sensory cognitions involve mechanical interactions between objects and sensory organs. The connection between these mechanical interactions and the corresponding sensory cognitions is simply posited in the way that basic powers are. Thus Kant explains that “Nature seems to have endowed man . . . with this organ [of touch].” The organs are the basic powers of sensory cognitions, and they are not further explained mechanically in Kant’s account.

In addition to the senses, the lower cognitive faculty includes the imagination.[151] Kant’s account of the imagination is similar to that of the empiricists. Imaginative cognitions begin with sensory cognitions, and they can never go beyond what has been made available by the senses. Kant explains that even the “productive [produktive] [imagination] is nevertheless not creative [schöpferisch], because it does not have the power to bring forth a sensory representation that was never given to our sensory powers.”[152] Moreover, Kant follows the empiricists in positing a “law of association of ideas”[153] as the fundamental law governing the relations between ideas of the imagination, and he gives several varieties of this principle to explain different ways in which one idea can lead to another. For example, in a passage that could almost be taken straight from Hume, Kant explains how “empirical ideas that have often followed each other produce a mental habit such that, when one is produced, this causes the other to arise as well.”[154]

With respect to both the senses and the imagination, Kant’s basic account of the causal laws governing their operation covers the operation of normal or healthy senses and imagination. But in both cases, Kant’s attention to the well-functioning senses and imagination is sparse relative to his treatment of various influences that can change – usually in ways that are unhealthy – the normal functioning of these faculties. Thus with respect to the senses, Kant devotes a section of his published Anthropology to “the inhibition, weakening, and total loss of the sense powers” through such causes as “drunkenness” and “fainting.”[155] And with respect to the imagination, Kant discusses the ways in which “intoxicating food or drink”[156] can influence the imagination, he lists some “faults of the imagination,”[157] and he describes various standard ways in which the imagination can lead one astray. For instance,

When one reads or hears of the life and deeds of a man who is great by virtue of his talent . . ., one is generally misled in ascribing considerable stature to him in imagination . . .. Not only the peasant, but even someone fairly well acquainted with the ways of the world, feels strange when the hero, who appearance had been judged by the deeds sung of him, presents himself as a little fellow, and when the sensitive and gentle Hume presents himself as a square-built fellow.[158]

This focus on disorders, and ways of correcting them, is what one would expect from Kant’s “pragmatic” anthropology. And Kant’s approach is no less causal for this practical focus. The result is an account of causal laws according to which Kant briefly explains the normal functioning faculties of sense and imagination and then discusses various ways of causally influencing these faculties for better or worse.

Given his account of the causal origins of lower cognitions, we can now complete Kant’s overall account of the lower faculty of desire.[159] Interactions between sense organs and the physical world give rise to sensory cognitions in accordance with natural laws governing the operation of sense organs. These sensory cognitions can cause desires in accordance with one’s instincts or inclinations, or they can give rise to further lower cognitions (of imagination) according to various laws of association. If they give rise to imaginative cognitions, these can in turn give rise to desires in accordance with instincts or inclinations. And these desires can cause actions, either directly or by affecting higher practical principles.

6) Causal accounts of cognition: the higher faculties of cognition

One might think that even if Kant can give a causal account of the lower faculty of cognition – the cognitive faculty that is supposed to be primarily “passive” or “receptive” – he will not be able to give a similar account for the “spontaneous” or “self-active” higher faculty.[160] However, although the accounts of the higher cognitive faculties are in some respects more complicated than those of the lower faculties, Kant does not think that higher cognitive faculties are any less natural, nor any less susceptible to explanation in terms of natural causes, than the lower faculties.

Kant identifies all the basic powers of the soul as natural predispositions,[161] and “understanding and reason,” the characteristic powers of the higher cognitive faculty, are specifically identified in this way.[162] Even “pure concepts” can be traced to “predispositions in the human understanding,”[163] and the three sorts of practical reasoning – technical, pragmatic, and moral – are labeled “predispositions.”[164] In the Metaphysics of Morals, Kant even explicitly includes “powers . . . whose . . . use is not drawn from experience but rather derived a priori from principles,” along with “memory, imagination, and the like” among the “natural predispositions” that one has a duty to perfect.[165] However reason and the understanding appear from within the context of epistemology and pure moral philosophy, when considered within the context of anthropology and empirical psychology, they are simply natural predispositions.

Like other natural predispositions, the higher cognitive powers are in a sense left unexplained.[166] Kant gives various teleological explanations for why human beings have these predispositions,[167] and he even offers some conjectures about the causal origins of human reasoning.[168] But Kant generally treats the higher cognitive faculty just like any other natural predisposition, as something which is itself left unexplained, a properly basic power. Just as in the case of other predispositions, however, Kant insist that there are characteristic laws that govern the connections between cognitions within the higher faculty of cognition. Kant says in the first Critique, for example, that “even . . . reason . . . must exhibit an empirical character” and suggests that “reason is itself determined by further influences.”[169] In his lectures on ethics, Kant is adamant about this further determination of reason:

Even one’s reason, as subjected to the laws of nature, can be considered devoid of all freedom . . . . Man is not set free from the mechanism of nature by the fact that in his action he employs an actus of reason. Every act of thought or reflection is itself an occurrence in nature . . . though this actus is an inner occurrence, since it takes place in the man himself . . . . So the fact that a man is determined to action on grounds of reason and understanding does not yet release him from all mechanism of nature.[170]

At least in general terms, Kant reiterates his claim that all mental phenomena are causally determined, even those that are rooted in the “spontaneous” higher faculties.

Just as lower cognitions are caused either directly or indirectly by experience, Kant insists that higher cognitions are also caused by experience. In the first Critique, Kant explains this point in a way that highlights the difference between the empirical cause of a higher cognition and its epistemic justification.

There is no doubt whatever that all our cognition begins with experience. . . . As far as time is concerned . . . no cognition in us precedes experience, and with experience every cognition begins.[171]

Kant makes this claim precisely in the context of distinguishing a priori cognitions, the most “pure” of the cognitions associated with the higher faculty of desire, from empirical ones. As Kant makes clear here, “although all our cognition commences with experience, yet it does not all on that account arise from experience,” and in that sense that there might well be “cognition independent of experience.”[172] The point here is that even if all cognitions are empirically caused, they need not be empirically justified.[173] In terms of empirical psychology, one can find a cause for any cognition, and that cause will always be an experience of some kind. But this does not mean that the content of the cognition is limited by the experience that causes it, nor that it is justified by that experience.

When it comes to actually stating the causal laws that govern the higher faculties of cognition, Kant follows the trend established with the lower faculties, briefly discussing the normal operation of the understanding, judgment, and reason, but focusing on various ways in which these faculties can have “defects” or “illnesses.” Like his accounts of the senses and unlike his account of the imagination, Kant’s discussion of the laws governing the normal functioning of the higher faculties is very brief in his anthropology (both published and lectures) and lectures on empirical psychology. Unlike the senses, however, where one needs to look to biological accounts of sense organs for more detail about proper function, Kant provides details about the proper functioning of the higher faculties of cognition in his Kant’s logic.

It might seem out of place to look for causal laws in Kant’s logic. In his lectures on logic, Kant explicitly distinguishes the study of the mind involved in logic from that of psychology. He explains,

Some logicians, to be sure, do presuppose psychological principles in logic. But to bring such principles into logic is just as absurd as to derive morals from life. If we were to take principles from psychology, i.e. from observations concerning our understanding, we would merely see how thinking does take place and how it is under various subjective obstacles and conditions; this would lead then to cognition of merely contingent laws. In logic, however, the question is not about contingent but about necessary rules; not how we do think, but how we ought to think.[174]

As Kant makes clear here, the laws of logic are necessary laws, not derived from observation, and hence distinct from the psychological laws that actually govern the connections between higher cognitions. At the same time, however, Kant claims in his logic that “like all our other powers, the understanding in particular is bound in its actions to rules, which we can investigate.”[175] Kant even compares this law-likeness to “everything in nature, [which] takes place according to rules,” such that “water falls according to laws of gravity and with animals locomotion also takes place according to rules.”[176] Thus on the one hand Kant claims that psychological rules governing the understanding are distinct from logical ones, but on the other he claims that his logic investigates rules that are like those of other natural laws.

This apparent conflict can be resolved by keeping in mind two important details in Kant’s distinction between logic and psychology. First, Kant’s separation of the two is unidirectional. Kant specifically warns against using psychological generalizations in logic, not about using logical laws as guides in empirical psychology. Second, Kant’s account of how the mind actually works is not simply a matter of logic. Logic describes how the understanding ought to operate, and this will describe how the understanding in fact operates only if the understanding is functioning properly. Often the understanding does function properly, and in those cases, Kant’s logic provides an account of the psychological laws governing the connections between higher cognitions. But human minds are also susceptible to “subjective obstacles and conditions,”[177] some of which are quite common. A complete causal account of higher cognition would have to explain the ways in which these subjective obstacles interfere with proper functioning of the understanding. In his logic, psychology, and anthropology, one finds descriptions of these variations from normal function. Such descriptions represent an important tangent when discussed in Kant’s lectures on logic, but in his anthropology, these variations and various ways to treat or prevent them are a proper focus of Kant’s attention. By structuring his overall account of mental operations in this way, Kant can apply the insights of his logic to flesh out his empirical description of human action without conflating logic and psychology.

It should not be too surprising that Kant analyzes the causal laws governing the higher faculty of cognition by reference to logic, the rules governing its proper function, and focuses in his anthropology on “deficiencies and diseases of the soul with respect to its cognitive powers.”[178] Even in his accounts of the senses, Kant’s focus is on “the inhibition, weakening, and loss of the sense powers,”[179] rather than the causal laws governing their normal operation. And in his lectures on logic, although Kant insists that logic must be distinct from psychology, he consistently gives at least some attention to ways in which the actual operation of the understanding differs from how it ought to conduct itself. Thus he asks, for instance, “How is it possible for a power to depart from its own laws?” and even claims that “the understanding itself cannot err,”[180] suggesting that the laws governing the actual conduct of the understanding are just those that govern the way it ought to be, at least in the absence of influences external to the causal laws governing the understanding itself.

Moreover, even the most ardent empiricist accounts of mind (at least in the 18^th century) turn to apparently logical laws when explaining the operation of healthy higher cognitive faculties.[181] When Hume sets out to determine the faculty responsible for “the transition from an impression present to the memory or sense to the idea of an object, which we call cause or effect,” he asks “Whether experience produces the idea by means of the understanding or of the imagination; whether we are determined by reason to make the transition, or by a certain association and relation of perceptions.”[182] To determine whether or not reason is the faculty responsible for this transition, he appeals to a “principle” that, in accordance with basic laws of deductive logic, would justify this transition. Hume uses the apparently logical claim that the transition is not deductively justified to justify the psychological claim that the cognitive faculty responsible for the transition is not the understanding (or reason). In this context, Hume treats the laws of logic as the causal laws governing the operation of reason. In appealing to logic as a source of causal laws of a properly functioning understanding, Kant’s account is no less causal-determinist than Hume’s.[183] And in explaining various ways in which the higher faculties can fall short of proper function, Kant’s causal account is actually more sophisticated than Hume’s.[184]

Kant’s causal account of the healthy understanding is fairly ordinary, rooted in his broadly Aristotelian account of logic. As in the case of the lower faculty, there are different laws governing different specific powers of cognition. Thus the understanding in a narrow sense is governed by laws regarding the formation of concepts and judgments and laws governing connections between various judgments.[185] With respect to connections between judgments, for example, Kant distinguishes between “immediate” and “mediate” inferences, where the inferences of the understanding, properly speaking, are all “immediate” and include such inference-types as the “universal rules of conversion,” such as the inference from “all men are mortal” to “some of those who are contained under the concept mortal are men.”[186]

With respect to the formation of concepts, Kant claims that there are three “logical actus of the understanding, through which concept are generated . . .: 1. comparison . . ., 2. reflection . . ., [and] 3. abstraction.” Kant gives the following example:

I see, e.g., a spruce, a willow, and a linden. By first comparing these objects with one another I note that they are different from one another in regard to the trunk, the branches, the leaves, etc.; but next I reflect on that which they have in common among themselves, trunk, branches, and leaves themselves, and I abstract from the quantity, the figure, etc. of these; and thus I acquire a concept of a tree. [187]

This account of concept formation is not specifically designed as a causal account, but as a normative logical one. Kant’s point here is not an empirical generalization about how people in fact arrive at concepts, but about how one should acquire concepts. But the account also functions as Kant’s explanation of how a properly functioning understanding can acquire new concepts.

While comparison, reflection, and abstraction feature prominently in Kant’s logical account of how one should acquire concepts, they are not the only way that one in fact acquires concepts. In fact, concept acquisition through instruction is an area in which Kant’s anthropology contributes significantly to understanding the normal functioning of the understanding. One can acquire concepts from other people through instruction, and this source of concepts is is prominent in Kant’s anthropology. As he explains, “Instruction can enrich natural understanding with many concepts and equip it with rules.”[188] In his lectures on pedagogy, Kant discusses the “cultivation of the mind”[189] and in particular the “cultivation of the higher faculties of cognition.”[190] For example, “the understanding may at first be cultivated . . . by quoting examples that prove the rules, or . . . by discovering rules for particular cases.”[191] Kant breaks down the factors involved in education in his Anthropology, pointing out in particular the role of language.[192] Kant gets into even more detail about the causal nature of this communication by connecting the transmission of concepts specifically with physical causes that influence one’s sense organs: “it is by this medium [air], when it is put into motion by the vocal organ, the mouth, that we can most readily and fully share in one another’s thoughts . . . . [W]ords are the means best adapted to signifying concepts.”[193]

Kant’s commitment to this causal basis for the development of the understanding goes so far, in fact, that he makes hasty and unfortunate claims about those who lack the sensory capacities for hearing: “a man who, because he was deaf from birth, must also remain dumb (without speech) can never achieve more than an analogue of reason” and “a man born deaf . . . does [nothing] more than carry on a play of . . . feelings, without really having and thinking concepts.”[194] Kant’s point here is clearly overdone, and one would hope that more experience with the deaf, or even just more thoughtfulness, would have corrected his opinion. The deaf are more than capable of thinking with concepts and have more than an analogue of reason. But Kant’s conviction that they would be impaired in the higher cognitive faculties because of a hindrance in sense organs shows his commitment to a causal explanation of concepts rooted in the senses, even while he maintains an epistemic position that some of these concepts are a priori.

In his anthropology, and even to some degree in his logic, Kant goes beyond a description of the way that the higher cognitive faculty ought to proceed and discusses some of the ways in which its operation differs from this normative standard. This practice of giving a normative account supplemented by various disorders is consistent with Kant’s practice throughout his anthropology, including his treatment of the lower cognitive faculty and even the faculties of feeling and desire.[195]

When Kant turns to non-normative accounts of cognition in his logic, he focuses on the role of “prejudices” in causing the higher faculty of cognition to operate in a way that is different from the standards of logical perfection. Prejudices, in general, are “provisional judgments . . . accepted as principles,”[196] and they affect the way in which other cognitions arise. In that sense, they play a causal role in one’s higher cognitions. For example, “the prejudice of the prestige of the age” leads us to favor the writers of antiquity more than we should, thereby “elevating the relative worth of their writings to an absolute worth.”[197] This prejudice leads people to adopt principles that they otherwise would not adopt, simply because they read of those principles in the ancients. Not only do prejudices themselves play a causal role in Kant’s account of the understanding, he explicitly seeks the “causes [Ursache] by which . . . prejudice . . . is created and sustained,”[198] and he finds that there are three “principle sources of prejudices . . .: imitation, custom, and inclination,”[199] all of which are empirical influences on higher cognitions. Thus even in his logic, Kant gives empirical-causal explanations of the factors that play a role in bringing the understanding to operate in non-normative ways.

In the Anthropology, Kant focuses less on common prejudices and more on uncommon “weaknesses and illnesses [Schwächen und Krankheiten]” affecting the cognitive powers, as well as various “talents” that are relevant to cognitive faculties. This focus is in keeping both with the popular approach of Kant’s anthropology, wherein he seeks to present “phenomena and their laws” in a way that is “entertaining and never dry,”[200] and with the connection between his anthropology and 18^th century natural history. With respect to the former, the focus on disorders of these kinds, rather than fallacies of prejudice, give Kant the opportunity to present numerous colorful examples of mental weakness and disorder, which his students would have found entertaining and could “compare with their . . . experience.”[201] With respect to the latter, the focus on weaknesses, illnesses, and talents brings Kant’s discussion of the higher cognitive faculties in line with his more general anthropological approach, where he focuses on natural endowments and quasi-biological classification of different types of people.

In keeping with his anthropology and empirical psychology more generally, Kant’s approach to these disorders is more taxonomic that strictly causal. He distinguishes mental weaknesses, where someone’s reason simply “has not enough control over itself to direct” thoughts, from mental illness, wherein someone’s “stream of thoughts follows its own (subjective) rule, which is contrary to that (objective) [rule] that conforms to the laws of experience.”[202] Mental deficiencies include being dull, simple, or silly, lacking judgment, and distraction. The differences between these deficiencies depend on which cognitive faculty is affected (reason, the understanding, or the power of judgment) and in what ways that faculty fails to function properly. In each case Kant explains ways in which these deficiencies affect one’s cognitive function, and in each case the taxonomy marks different causal stories of the flow of cognitions. For example, someone who is “dull” lacks intelligence, the “faculty for discovering the universal for the particular;”[203] thus such a person cannot effectively acquire concepts for him or herself through the process of comparison, reflection, and abstraction. Such a person can still be taught, however. So, for example, “Clavius, whose schoolmaster wanted to apprentice him to a blacksmith because he could not make verses [showing that he was dull] . . . . became a great mathematician as soon as he got hold of a book of mathematics.”[204]

With mental illnesses, Kant’s approach is similar. He distinguishes “melancholia (hypochondria)” and “mania.” The former “can cause an illness because human nature, by virtue of a peculiar quality lacking in animals, can strengthen or sustain a feeling by centering attention on certain local impressions.”[205] Thus one focuses on certain sensations and conjures up imagined illnesses, which can eventually dominate one’s life, making one “unable to come to grips with his imaginings.”[206] With respect to the latter, Kant pessimistically claims, “It is difficult to bring systematic classification to what is essential and incurable disorder.”[207] Mania covers such a wide variety of divergent and irregular mental illnesses that Kant despairs of providing an exhaustive catalog of them. Moreover, in the context of his pragmatic anthropology in particular, “it is of little use to occupy oneself with [mania] because all methods of cure must be fruitless . . .. Still, anthropology requires at least an attempt at a general outline of this most profound degradation of humanity which seems to originate from Nature.”[208] In this context, Kant distinguishes between several different kinds of “derangement,” focusing on broad classifications rather than specific varieties of each.

1. Madness [Unsinnigkeit; amentia[209]] is the inability to bring ideas into mere coherence necessary for the possibility of experience . . ..

2. Dementia [Wahnsinn; dementia] is that disturbance of mind wherein everything which the insane person relates is in accord with the possibility of experience, and indeed with the formal laws of thought; but because of falsely inventive imagination, self-concocted representations are treated as if they were perceptions . . ..

3. Insanity [Wahnwitz; insania] is disordered faculty of judgment in which the mind is deceived by analogies, which are being confused with concepts of similar things . . ..

4. Lunacy [Aberwitz; vesania] is the sickness of a disordered reason. The patient disregards all the facts of experience and aspires to principles which can be entirely exempted from the test of experience . . .. This fourth kind of madness could be called systematic. For in this last kind of mental derangement there is not merely lack of order and deviation from the rule for the use of reason, but also positive unreason; that is, another rule is present.[210]

In each case, the relevant higher power of cognition – understanding for the first two, judgment for the third, and reason for the last – fails to operate according its normal rules of operation, operating either in “essential disorder,” according to causes that cannot be easily subsumed under rules, or according to “another” rule that that which is normative for the respect faculty, as in the case of “positive unreason.”

For a complete causal account of the higher cognitive faculties, Kant should give causal accounts of the origins of these mental illnesses, or should show that they are natural predispositions in those who have them. With respect to derangement, Kant focuses on biological bases, arguing that “the germ [Keime] of derangement develops together with the germ of reproduction, and is thus hereditary.”[211] Kant’s idea here seems to be that such derangement simply sets on at a particular time due to biological factors. The “definite object” that becomes “the subject matter about which the person will rave” is based on an “accidental encounter;” this object is simply “what first comes into the mind at the (usually sudden) outbreak of a crazy disposition.”[212] With respect to other mental illnesses, Kant gives a bit more detail about contingent causal factors relevant to their genesis. He claims, for example, that “the reading of novels” can “cause many . . . mental discords” and “has the consequence that it makes distraction habitual.”[213] The failure to give sufficient details about the biological and contextual causes of various mental disorders is a limitation to Kant’s empirical account of human action. But this limitation is understandable in a philosopher whose primary focus was on normative rather than psychological issues, and the scarcity of detail here hardly compromises Kant’s overall commitment to causal necessity in human action.

Overall, then, Kant’s empirical account of the higher faculty of desire can be explained causally in a way similar to that for the lower faculty of desire. The causal account of the higher faculty of desire begins with higher cognitions. One gets these principles of the understanding either through a process of comparison, reflection, and abstraction that begins with sensory experience, or through education, including such informal “education” as conversation in which someone introduces another to a new concept or principle. Some of these principles of the understanding are practical, in that they give rise to desires in accordance with one’s character. And that character is itself the result of various empirical influences acting in the context of a natural predisposition, a propensity for character.

7) Conclusion

Whether actions proceed from the higher or the lower faculty of desire, Kant’s empirical account of these actions is determinist.[214] It is based on his classification of the human faculties into the faculties of cognition, feeling, and desire. Every action follows immediately from a desire. Desires themselves are the result of particular kinds of pleasures following from particular kinds of cognition. And desires can be caused by a variety of types of cognition, from raw sense-perceptions – as when a delicious smell draws us to eat food – to principles of reason. The causal origin of all of these cognitions, however, can be traced back to sensory experiences, either immediately (for sensory cognitions) or indirectly through laws of association (for the imagination), rules governing comparison, reflection, and abstraction (for the understanding), or principles of linguistic communication and instruction (also for the understanding). Whether a particular cognition will give rise to a desire (and thereby an action) depends on one’s “natural predispositions” and the ways that these have been cultivated to generate the inclinations and character that one has. The biological structures in human beings provide the foundation for any particular explanation of behavior in empirical psychology. And every such explanation will be thoroughly deterministic, accounting for action on the basis of desires that are caused by one’s nature, background, and circumstances.

There is more detail that could be added to this account. I have not, for instance, shown how some of the familiar terms of Kantian psychology – such as interests, maxims, choices, or the will – fit into the account I have presented here. There is also considerably more that could be said about the specific laws governing instincts, inclinations, and particular features of character. One might explore what preconditions are necessary to develop each particular inclination that one finds in human beings, what limits there are to the scope of principles that can form the basis of character, or what causes various mental illnesses. Kant’s detailed pedagogical insights could be expanded into a fuller account of the ways in which both cognitive and character development take place in human beings. Kant’s accounts of inclination, instinct, character, feeling, and desire could be compared with recent psychological and philosophical work on the emotions. Kant’s account of feeling in particular was not analyzed in detail, and there are subtleties in that account that I have not discussed.

Finally, it would be well worth applying the psychology laid out here specifically to the context of actions motivated by “respect for the moral law.” Respect functions as the intellectual (higher) feeling that causes one to act morally, and it is in turn caused by a cognition of the moral law itself. But there are notorious difficulties with making sense of this account, and the psychology that I have presented provides a framework within which to make sense of respect. In the context of discussing moral action in particular, there may seem to be specific obstacles to giving a complete causal account, and some specific discussion of Kant’s causal accounts of moral action would be valuable further fruit of this study of his psychology.

My goal in this paper, however, was simply to lay out Kant’s overall empirical psychology. I argued that this psychology is thoroughly determinist, at least as much as any other causal accounts in biology. I have shown that in his empirical psychology, Kant’s account of human beings is precisely the opposite of the popular interpretation of Kant that Blackburn captures well with his account of a “Kantian Captain” who is “immune in all important respects from the gifts or burdens of our internal animal natures, or of our temperaments as they are formed by contingent nature, socialization, and external surrounds[,] . . . . free of his or her natural and acquired dispositions.”[215] In this paper, I have given at least some sense of how Kant’s empirical psychology describes those features in our animal natures that, combined with socialization and external surrounds of particular kinds, give rise to the natural and acquired dispositions that explain human action. This empirical dimension of Kant’s account of human action provides a needed and under-appreciated complement to the accounts of human action from the standpoint of practical reasoning, accounts which properly emphasize human freedom. By adding this dimension, important interpretive problems in Kant can be discussed in more psychologically astute ways, and Kant’s philosophy as a whole cannot be as easily dismissed on the grounds of its supposed psychological naiveté.[216]

---

[1] The first passage is from A549/B577. The example is from A554-55/B 582-83, with emphasis added.

Kant reiterates this point with a very similar example in a set of lectures in 1794-5 (29:1019-20), suggesting that the point is not limited to Kant’s early Critical writings.

(Throughout this paper, reference to Kant are to the Academy Edition of Kant’s works or to the standard A and B editions of the Critique of Pure Reason. For English translations, I have made extensive use of the Cambridge Edition of the Works of Immanuel Kant, where available, and of translation of the Anthropology by Victor Lyle Dowdell (Southern Illinois University Press: 1978) and Mary Gregor (Martinus Nijhoff: 1974).)

[2] 4:455.

[3] 5:99. Despite these apparently clear statements, some have recently claimed that “there is no Kantian basis for maintaining causal determinism in the psychological realm” (Westphal 1995: 362, Gouax 1972):

Kant himself held it to be one of the cardinal achievements of the Critical philosophy, forever to foreclose on both materialist and ‘spiritualist’ explanations of the mind…. Those are the only two kinds of causal explanations countenanced in the Modern period. To foreclose on such explanations in psychology is (for . . . Kant) to foreclose on the scientific status of psychology. (Westphal 1995: 358)

There are two sorts of criticism generally raised against the notion of a Kantian account of causal laws governing mental states. The first and most radical claim, articulated here by Westphal, is that for Kant there simply are no causal laws governing human psychology. The second claim, which is often connected (e.g., by Westphal) with the first, is that psychology can never rise to the level of a science.

In general, there are also two lines of argument to establish each of these claims. The first argument is based on the requirements for human freedom. Briefly, the argument is that because of Kant’s emphasis on human freedom, he cannot allow for a science that explains human actions in terms of causal laws. Westphal thus claims that “Kant needs Critically justified theoretical reasons for denying [determinism in psychology] in order to permit himself to describe the human will as . . . liberum” (357-8). I discuss this argument with respect to Reath and Baron in the body of the paper, so I will not take it up here. The second kind of argument, primarily used to establish the impossibility of psychology as a science, is based on Kant’s disparaging of psychology in the Metaphysical Foundations of Natural Science. There Kant says,

The empirical doctrine of the soul must always remain even further removed than chemistry from the rank of what may be called a natural science proper . . .. It can . . . never become anything more than a historical (and as such, as much as possible) systematic natural doctrine of the internal sense, i.e., a natural description of the soul, but not a science of the soul, nor even a psychological experimental doctrine. This is the reason why . . . the general name of natural science . . . belongs to the doctrine of body alone. (4:471, see too 28:679)

Many (e.g. Gouax 1972) have taken this passage to imply that Kant opposes any kind of serious empirical study of the causal principles underlying mental life, and some have gone as far as to use this to argue against causal necessity in human actions.

However, Kant’s argument against psychology as a science employs very specific objections to psychology as a science, and Kant allows that psychology can be a “historical systematic natural doctrine of the inner sense” (4:471) and even a “natural science . . . improperly so called, . . . [which] would treat its object . . . according to laws of experience” (4:468). As Hatfield (1990, 1992), Sturm (2001), and others argue, Kant objects to applying to psychology a very particular conception of science, as a study “whose certainty is apodictic,” which must thus consist in “a priori principles” (4:468) and in particular in the application of “mathematics” to its subject matter (4:470). Hatfield rightly points out,

While we must give these remarks [in MFNS] their due, they should not be allowed to obscure Kant’s basic position that the phenomena of empirical psychology are strictly bound by the law of cause just as are the phenomena of physics. (Hatfield 1992: 217)

Neither Kant’s concern with freedom, nor his prohibition of a “science” of psychology show that human beings are not governed by causal laws, nor even that we cannot discern (some of) those causal laws.

[4] 5:98.

[5] Wood 1984:74. Wood’s phrase here gives a nice way to distinguish different accounts of Kant’s moral psychology. Some, such as Henry Allison, basically endorse Wood’s view (see Allison 1990:28). Others, such as Hud Hudson (and arguably Christine Korsgaard), defend Kantian compatibilism at the cost of Kantian incompatibilism. And still others, such as Reath and Baron, implicitly defend Kantian incompatibilism without sufficient attention to Kant’s compatibilism.

[6] Wood 1984:74.

[7] This defense of freedom might not be fully satisfying, but it is Kantian. Wood 1984, Allison 1989, and Frierson 2003 have offered more detailed arguments for this view. For the purposes of this paper, I simply take this general account of Kant’s theory of freedom for granted, with the little support I have already offered here. I generally agree with Wood’s account of Kant’s theory of freedom, as developed in Wood 1984. The one issue on which Wood and I disagree is that I see Kant as claiming that actions that follow from reason are just as causally determined as actions that proceed from purely sensuous influences. Occasionally (e.g. pp. 78, 83, 87) Wood seems to suggest that nature only determines action through sensuous influences. As we will see in the course of this paper, “reason” is just as natural as the senses, from within the perspective of empirical psychology.

[8] 28:773, cf. 28:682, A535/B563.

[9] Of course, this determinism is only empirical determinism, not determinism “all the way down.” Human actions are empirically determined in the sense that these actions follow from prior causes in accordance with laws of nature. But these actions are still ultimately ascribed to transcendentally free agents, where this freedom consists in those agents not being determined at a noumenal level by empirical causes. This is discussed in detail in Wood 1984 and Frierson 2003: 13-30.

It is also worth noting that at this point, there is room for Kant to tell some causal story about human action, but his transcendental idealism does not imply any particular story. Given Kant’s transcendental philosophy, freedom of the intelligible character can be preserved regardless of the picture at the empirical level. Paul Guyer has even suggested, on these grounds, that “the subjective state of one’s feelings” can, perhaps even directly, “reflect the moral choices of one’s will” (Guyer 1993: 367). In fact, Kant develops a very particular conception of human action at the empirical level, one that is influenced by and related to his contemporaries’ accounts, but not identical with them, but his overall transcendental philosophy could support a variety of different empirical psychologies. {part of footnote deleted for the sake of anonymity}

[10] Blackburn 1998: 252.

[11] Reath 1989: 287.

[12] Reath 1989: 290-91.

[13] Thus even Gouax, who argues that “the First Analogy is not applicable to objects of inner sense” (1972: 240) is forced to admit that “it is possible to apply causality [the Second Analogy] to inner alterations.” And while Gouax insists that “surely causality applied to inner alterations is a weak notion” (1972: 240), he offers no textual basis for this claim.

[14] Baron 1995, p. 189.

[15] Baron 1995, p. 189.

[16] Baron 1995, p. 189. Baron does not quote the whole passage here, and thus gives the impression that Kant’s argument for freedom is based on a kind of introspection of a felt power to choose otherwise. She cites Kant’s suggestion that “if someone claims that his lust is irrestible, ask him ‘whether he would not control his passion if, in front of his house where he has this opportunity, a gallows were erected on which he would be hanged immediately after gratifying his lust’ (PrR 30)” (Baron 1995: 189). Kant goes on to argue,

But ask him whether, if his prince demanded, on pain of the same immediate execution, that he give false testimony against an honorable man . . ., he would consider it possible to overcome his love of life . . .. He would perhaps not venture to assert whether he would do it or not, but he must admit without hesitation that it would be possible for him. He judges, therefore, that he can do something because he is aware that he ought to do it. (5:30, emphasis added)

On the basis of conjectural introspection, one “would perhaps not venture” to speculate about one’s capacities, but one can infer that one is able to do something from the moral (and thus not psychological) fact that one ought to do it.

[17] Even Jeanine Grenberg, who is much more sensitive than most Kant commentators to the details of Kant’s psychology, admits that Kant’s “language of impulsion certainly does suggest that . . . he is . . . advocating a more mechanistic theory of action” but argues that “this is not in fact the case,” that in fact human actions “do not follow the laws of nature.” (Grenberg 2001. The first quotation is from p. 151, the second from p. 175.)

Grenberg seems to locate the freedom of action in our capacity to control the representations that give rise to desires (and thereby actions). She asks “How does an appeal to representation distinguish feeling from mere physical force? That which evokes feelings is not, strictly speaking, a force completely external to an agent, but rather any state of affairs in so far as it has been taken up by an agent’s capacity to represent it to herself” (161). But of course, this contrast between internal and external causes is one with which any empiricist would agree, and one that Kant specifically argues is not sufficient for the kind of freedom that is needed for moral responsibility (see 5: 97f.). And one would not expect this merely empirical freedom to be particularly important to distinguish Kant’s accounts from his more determinist contemporaries, since Kant saw his contribution to freedom of action to lie not in a novel psychological account of empirical freedom but in a new transcendental freedom compatible with thoroughgoing natural determinism.

At times, Grenberg seems sensitive to this point. She is much clearer than other commentators about the specifically first-personal nature of freedom in Kant’s psychology, saying for example that “when [an agent] judges her feeling of pleasure to be good . . . she attributes it to her own faculty of desire, not from a third person perspective, but from a first person perspective,” and she distinguishes this “practical” task from “theoretical . . . knowledge of herself as object” (171). But this distinction does not sufficiently inform her overall treatment of Kant’s account of action, such that she still seems to think that there is a conflict between Kantian freedom and a thoroughgoing natural necessity in psychological explanations of human action.

[18] Blackburn 1998: 250. See the whole discussion on 243-261.

[19] Blackburn 1998: 106.

[20] Blackburn 1998: 256.

[21] The primary resources here are hints within Kant’s three Critiques but especially his Anthropology from a Pragmatic Point of View, as well as Reflexionen and lecture notes on anthropology and metaphysics. The lectures on metaphysics, in particular, give Kant’s most systematic treatment of empirical psychology. While there has been some question about the veracity of these lectures, they have increasingly been taken to be reliable, largely because of their overlap with each other and with Kant’s published writings. (See the introduction to the Cambridge Edition of the Lectures on Metaphysics for more details on their reliability.) I generally take these lectures to be reliable, while making some effort to back up key interpretative claims with references to published works.

[22] The recent publications of these lectures in English, as part of the Cambridge Edition of the Works of Immanuel Kant, will make the study of Kant’s empirical psychology much easier within the English-speaking world. The forthcoming publication of Kant’s works and lectures on anthropology in the same series will hopefully have a similar result.

[23] For details on the relationship of these three strands to Kant, see Beck 1969; Hatfield 1990: 21-77; Henrich 1957/58 and 1994: 20-7, 70-2; Hilgard 1980; and Schneewind 1998. For a close study of the reception of Scottish philosophy in 18^th century Germany, see Kuehn 1987. Other figures are relevant to Kant’s psychology, including Tetens, Eberhard, Mendelssohn, and Lossius, but a full discussion of the historical background of Kant’s psychology is beyond the scope of this essay. Baumgarten and Mendelssohn are particularly important in that both articulated similar three-fold divisions to Kant’s own (cf. Hilgard 1980).

[24] cf. 29:877.

[25] 28:564.

[26] Kant explains that he groups the essentially distinct powers into 3 classes “in order to treat empirical psychology all the more systematically” (28:262).

[27] 28:564. I have amended the Cambridge Edition translation, translating Geist as “mind” where they translate it as “spirit.”

[28] A347/B405.

[29] Kant would have known of Hume’s philosophy, at least through Sulzer’s translation of Hume’s Inquiry. For more on the connections between Kant and Hume, see especially Henrich 1957/58 and 1994, Kuehn 1987, and Hatfield 1990.

[30] For the law of association governing the imagination, see 7:174ff. and corresponding sections in his lectures on anthropology (e.g. 25: 883) and empirical psychology (e.g. 28:585, 674). Baumgarten too draw attention to laws of association governing the imagination (see Baumgarten’s Metaphysica, §561), but this law was particularly prominent among British empiricists (see e.g. Hume 1740: I.iii.vi.4, pp. 88-9 and I.iii.vii.6, p. 97: “we are not determined by reason, but by . . . a principle of association”). I argue in section 6 that the comparison of Kant and Hume can also help one see the ways in which some of Kant’s apparently logical claims about the understanding can be read as causal ones.

[31] 25:1514, cf. 29:1024. For Kant, all action proceeds from a prior determination of the faculty of desire. Thus within Kant’s empirical account, at least, it is not the case, as Simon Blackburn has suggested, that “motivation by means of desire was one thing, motivation by apprehension of the Moral Law a different thing” (Blackburn 1998: 214). “Desire,” as the faculty giving rise to action, is necessarily involved in any human action. (That said, Kant distinguishes between different sorts of desire, including a “desire in the narrow sense” (6:212) that is not necessarily involved in every action.)

[32] 29:1012, cf. 6: 211, 399; 7:251.

[33] One caveat must be added here. For Kant, the tendency of a representation to maintain that representation itself without bringing about a change in the world does not count as a desire. This is how Kant accounts for aesthetic pleasure, where one seeks to maintain a representation but without any desire for an object of that representation, and Kant is particularly interested in these cases of “disinterested” pleasure (cf. 28:674-5.) The fuller context of the passage quoted at this footnote reads,

We have pleasure or displeasure without desiring or abhorring, e.g. if we see a beautiful area, then it enchants us, be we will not on that account wish at once to possess it. Pleasure or displeasure is thus something entirely different from the faculty of desire. But on the other hand we can desire or abhor nothing which is not based on pleasure or displeasure. For that which give me no pleasure, I also do not want. Thus pleasure or displeasure precedes desire or abhorrence. But still I must first cognize what I desire, likewise what give me pleasure or displeasure; accordingly, both are based on the cognitive faculty. There are also many representations which are connected with neither pleasure nor displeasure, and thus the cognitive faculty is wholly distinct from the feeling of pleasure or displeasure. (29:877-8)

Kant’s aesthetics is largely foreshadowed in his empirical psychological discussions of feelings that do not give rise to any desire. This is a rich and complex part of Kant’s empirical psychology that I forego here, but for more, see Kant’s accounts of feeling in his lectures on metaphysics and anthropology, and his account of the beautiful in the Critique of Judgment and in “Observations on the Feeling of the Beautiful and the Sublime.” The nature of aesthetic pleasure has been discussed at length in secondary literature. For some examples, see Allison 2001, Ameriks 2003, and Guyer 1979, 1993.

[34] 29:877-8. This account of human action is identical for animal action. Like humans, other animals have cognitions, feelings, and desires. Kant even describes animals as having “choice [Willkühr]” (cf. 6:213, 28:588, 29:1015). The difference between humans and animals is that humans have a “higher” faculty of cognition, and hence of desire, and this gives humans a kind of empirical freedom than animals lack. This shows that at least Kant’s general picture of human action does not imply any transcendental freedom. Kant nowhere suggests that animals have transcendental freedom, so insofar as they are motivated by similar structures as human beings, there is nothing “free” about these structures in themselves.

[35] There is one complication to this picture that I will not discuss here. When discussing the “higher” faculty of desire (see below), Kant sometimes suggests that pleasure does not precede desire. Some of the strongest language here is from the second Critique, where Kant says,

What is essential to any moral worth of actions is that the moral law determine the will immediately. If the determination of the will takes place conformably with the law but only by means of a feeling, of whatever kind, that has to be presupposed in order for the law to become a sufficient determining ground of the will, so that the action is not done for the sake of the law, then the action will contain legality indeed but not morality. (5:71-2)

This passage could be read merely to refer to first personal grounds of choice rather than to empirical causes of choice, but it might also be read psychologically. The former interpretation, which fits better with the account of pleasure offered in this paper, is supported in part by Kant’s appeal later in this discussion to a “positive feeling” of “respect for the moral law” (5:73), a “moral feeling . . . produced solely by reason” that serves “as an incentive to make this [moral] law its [the will’s] maxim” (5:76). For detailed studies of these passages, see Allison 1989, McCarty 1993 and 1994, and Reath 1989. A similar tension between an intellectual feeling that causes choice and an insistence that feeling must play no role occurs throughout Kant’s lectures and other writings. Thus in the Metaphysics of Morals, Kant distinguishes between an “interest of inclination” where “a pleasure necessarily precedes a desire” and an “intellectual pleasure” that “can only follow upon an antecedent determination of the faculty of desire” (6:212, cf. 29:1024).

But in a lecture of the same period (1794-5), Kant deals with the issue of pleasure differently. He first distinguishes between “pleasure” (Lust) and “satisfaction” (Wohlgefallen), the latter of which is “more general” (29:1013) and includes both intellectual and sensible satisfaction. (In this context, “pleasure” describes the subset of satisfaction that is purely sensible.) But Kant then claims that “desiring” has its “ground” in “satisfaction with respect to the actuality of the object” (29:1013). (Even in that set of lectures, however, Kant distinguishes between the higher and lower faculties of desire by saying that the “impelling causes” of an action “lie either in the understanding as the law of action [in which case the cause is a higher desire], or in the sensibility, namely, in the feeling of pleasure and displeasure,” in which latter case the cause is a lower desire (29:1014). Here Kant seems to conflate sensibility with feeling and to associate both exclusively with the lower faculty of desire.)

In an earlier lecture (1782-3), Kant is even more explicit about the fact that pleasure must precede desire: “we can desire or abhor nothing which is not based on pleasure or displeasure. For that which gives me no pleasure, I also do not want. Thus pleasure or displeasure precedes desire or abhorrence” (29:878). Later in that lecture he reiterates that “representations cannot be the cause of an object where we have no pleasure or displeasure in it,” adding that “the faculty of desire rests on the principle: I desire nothing but what pleases” (29:894). In this context he distinguishes higher and lower faculty of desire based on whether they are caused by sensibility or the understanding (see 29:895, 28:254).

Overall, Kant seems to have two basic models for human action. On one model, Cognition à Pleasure à Desire à Action. Here the difference between respect for the moral law and action from sensuous incentives is that the relevant pleasure in the case of respect is a pure intellectual pleasure. This strand in Kant confirms those like McCarty who endorse an “affectivist” reading of Kant. On the alternative model, Cognition à Desire à Action and Pleasure, where again the relevant pleasure here is purely intellectual, but now it plays no motivational role at all and is simply an co-effect (with action) of the desire. At least in the account in the second Critique, this second picture also seems to involve a direct negative action of pure practical reason on sensuous feelings, by which it “derives self-love of its influence” (5:75) to make room for the direct action of pure cognitions on the faculty of desire.

[36] Kant suggests that smell and taste are paradigm cases here, and their connection is particularly important in this regard. See 8:111.

[37] Grenberg 2001: 160-3. For another very helpful account of Kant’s theory of pleasure in action, which makes some important distinctions that Grenberg ignores, see Morrison (forthcoming).

[38] Grenberg 2001: 163. Grenberg’s account of the connection between pleasure and desire is supported in the passages in Kant’s Metaphysics of Morals from which she draws her account. There, as Grenberg points out, Kant simply defines practical pleasure as “that pleasure which is necessarily connected with a desire (for an object . . .)” (6: 212). In his Anthropology, Kant similarly explains that

[P]leasure . . . and displeasure through the senses . . . can [be] describe[d] . . . in terms of the effect that the sensation of our state leaves on the mind. What directly . . . prompts me to leave my state (to go out of it) is disagreeable to me – it pains me. What directly prompts me to maintain this state (to remain in it) is agreeable to me – it gratifies me. (7:230-1)

Admittedly, this passage focuses on sensuous pleasures in particular, and thus need not apply to aesthetic or intellectual pleasures. For issues surrounding the nature of intellectual pleasures, see footnote 35.

As I indicated in footnote 33, I follow Grenberg in largely bracketing the issue of aesthetic pleasure. Aesthetic pleasure is not prominent in Kant’s account of action, and though it raises important issues for his overall empirical psychology, these issues go beyond the scope of both Grenberg’s paper and mine.

[39] There is more that could be done with respect to Kant’s theory of the role of pleasure in action. Two issues in particular merit further attention. First, Kant’s empirical psychology is unique in sharply distinguishing between feeling and desire, and his basis for this distinction is his account of aesthetic pleasure. I am inclined to agree with Iain Morrison, who argues that it is the nature of all pleasure to “want to maintain itself,” and that the difference between aesthetic and practical pleasure is simply that in the case of the former, “the way that the state of mind can preserve itself is . . . by simply maintaining contemplation . . ., [while] in practical matters the pleasurable state of mind maintains itself by bringing forth the contemplated object” (Morrison (forthcoming)). But this account of the relationship between aesthetic pleasure and practical pleasure in the context of Kant’s empirical psychology needs to be further developed.

Second, with respect to the higher faculties of cognition, feeling, and desire, Kant’s account seems to shift throughout his lectures and published writings. I discussed this to some extent in footnote 35, but more work needs to be done making use of Kant’s empirical psychology to develop an overall picture of Kant’s views on the nature of intellectual pleasure.

[40] 28:228, 29:880, 28:584.

[41] In some cases, Kant associates the spontaneity of the higher faculty with that transcendental freedom that is a condition of possibility of moral responsibility. Strictly speaking, describing the higher faculty of desire as free in that sense is inconsistent with empirical psychology. Insofar as one studies human action empirically, such action is, as Kant insists in his first Critique, causally necessitated in accordance with nature laws. But Kant does hold that the present of a higher faculty of desire is an indication of moral responsibility and hence transcendental freedom. (See {part of footnote deleted for the sake of anonymity}) Thus he sometimes slips into these properly transcendental discussions in lectures on empirical psychology. This effort to discuss the Critical philosophy in lectures on empirical psychology is not particularly surprising, of course. As a teacher, Kant found an opportunity within the syllabus prescribed by Baumgarten’s text for explaining some of Kant’s own more important philosophical ideas, a temptation to which he can hardly be blamed for succumbing.

[42] See Beck 1987, pp. 35-6 for a helpful account of this sort of freedom. This comparative freedom is the freedom that Kant refers to in the second Critique as the “freedom of a turnspit” (5:97). Brian Jacobs puts the point well in the context of discussion the nature of freedom of the higher faculty of desire (the “will”) in Kant’s anthropology: “the ‘arbitrium liberum’ that Kant posits against the animalistic ‘arbirtium brutum’ . . . is a practical empirical concept and one that is observable when a human being resists acting solely according to the ‘pathological’ necessity that characterizes animal will” (Jacobs 2003: 120).

[43] Cf. 7:140-1, 153ff.; 25:29f., 269f.; 28:59f., 230f., 585, 672f., 869f., 737f.; and 29: 882f.

[44] Cf. e.g. 7:196, and related sections of lectures in empirical psychology (28:73-5, 242-3, 863-5; 29: 888-90) and anthropology (25: 537, 773-4, 1032f., 1296, 1476) Kant’s placement of the power of judgment (Urteilskraft) in the higher cognitive faculty is a notable departure from Baumgarten, who places it in the lower cognitive faculty (see Metaphysica §§ 606-9). A detailed comparison of Baumgarten and Kant on the nature of judgment would reward further study but it beyond the scope of the present paper.

[45] 29:877.

[46] For complications related to bracketing the issue of freedom here, see footnotes 34 (above) and 47 (below).

[47] 29:894.

[48] Kant makes the same claim in the context of pleasure, but there Kant is careful to insist that on the one hand all pleasure is sensitive in itself (hence lower) but there is still a lower and higher faculty of pleasure. See too footnote 35 for Kant’s shifting views on the nature of higher pleasure.

[49] 28: 674-5.

[50] Kant actually says “act of the faculty of choice” here, but he has just explained that the faculty of choice is simply the faculty of desire insofar as it operates in a context where its activity can bring about its object (28:254).

[51] 28:254, cf. 29:895.

[52] 29:895.

[53] 29:885, cf. 29:1015, 27:257.

[54] 29:1014.

[55] 29:1015.

[56] 28:589, emphasis added.

[57] One way of putting this is that higher desires are those for which Allison’s incorporation thesis holds. For Kant, contra Allison, human beings can, sometimes, act purely from instinct or inclination, without incorporating such instincts or inclinations into any principle of the understanding. Kant’s language to describe such “actions” fits the lack of true agency implied by their failure to fits Allison’s account of incorporation. He refers to them as actions proceeding from “stimuli” or “impulse.” Most actions, even those that are not guided by morality, are free in the sense that they are associated with the higher faculty of desire, where one acts on principles or maxims, even if these maxims take the satisfaction of inclination as their end. But one can also “act” directly from lower desires. (This may help explain both the role of affects in Kant’s philosophy and Kant’s treatment of weakness of will.)

[58] This distinction is somewhat different than the discussion of higher and lower faculties in the Critique of Practical Reason. There Kant discusses the distinction in the context of arguing against heteronomous ethical theories, and he downplays the difference between pure and affected higher desires. In the account in the second Critique, he argues against those who describe the higher faculty of desire as one within which intellectual cognitions cause pleasure and thereby move the will. By contrast, he insists upon a higher faculty of desire as the ability for “pure reason . . . to determine the will without some feeling being presupposed” (5:24). These passages can be thought of from two perspectives. If interpreted as empirical psychology, this passage could be part of Kant’s more general shift away from the Cognition à Pleasure à Desire à Action model of motivation in the context of the higher faculty of desire. For more on Kant’s shifting views in this respect, see footnote 35. More likely, however, Kant is here arguing from within the practical perspective. From that perspective, choosing to act to pursue the pleasure acquired from intellectual activity is just as sensuous – because based on pleasure – as choosing to pursue any other pleasure.

[59] 28: 254.

[60] 28:564.

[61] Feelings that do not give rise to desire or aversion are particularly important for Kant’s aesthetics (cf. footnote 33).

[62] A natural predisposition “indicate[s] what can be made of a man” by nature, as opposed to “what man is prepared to make of himself” (7:285). The concept is closely related, for Kant, to the notion of a “germ” (Keime). (For more on Kant’s account of Anlagen, as well as the relationship between Anlagen and Keime, see Munzel 1999 and Sloan 2002.) Kant discusses Anlagen throughout his philosophy. In his lectures on metaphysics, Kant most commonly uses claims about natural predispositions to argue for the immortality of the human soul. As he puts it in a lecture from the early 1790’s,

The proof of the immortality of the soul is grounded on the principle of the analogy of nature. Nature has placed in all living organic beings no more predispositions than what they can make use of . . .. We find in us a [moral] summons to sacrifice the greatest advantages, without receiving in life the slightest advantage for it. Here is a predisposition in human nature, and this is just as purposive, according to the analogy of nature, as all predispositions of nature. We thus infer a future life where the use of these predispositions and their end first can be attained. (28:765-6, cf. 29: 916-7, 1040)

Here the notion of a predisposition is quite broad – including all end-directed “faculties” in animals, and even their “organs,” which “are not given any larger than they can make use of” (28:765). In the Religion, Kant makes use of the notion of a predisposition to good to explain the nature of radical evil. And in his Critique of Judgment, Kant discusses the natural predispositions of human beings in the context of his general argument that human beings are the ultimate end of nature (cf. 5:340-1).

[63] 8:110.

[64] This modesty is epistemic. Kant elsewhere argues that as a metaphysical matter, “it is obvious that there is only one basic power in the soul . . . But this is a wholly other question: whether we are capable of deriving all the actions of the soul, and its various powers and faculties, from one basic power. This we are in no way in the position to do” (28:262). Although Kant does not emphasize this point in other lectures, there is nothing in his later claims that precludes the possibility of all human powers in fact being reducible to a single one. But when it comes to empirical psychology, we are not justified in trying to effect this reduction. For more on the important difference between limits on human explanation and limits on metaphysical possibility, see Ameriks 2003.

[65] 29:935-6.

[66] 29:773, 822. Here again (recall footnote 64)., Kant leaves open the metaphysical possibility that “the unity of each substance requires that there be only one basic power” (29:822).

[67] 2:434

[68] 5:400

[69] It is also worth noting here that this limitation on mechanist explanation puts Kant in good company with other naturalist explanations of human beings. Hume, for example, argues,

When we see, that we have arrived at the utmost extent of human reason, we sit down contented . . .. But if this impossibility of explaining ultimate principle should be esteemed a defect in the science of man, I will venture to affirm, that ‘tis a defect common to it with all the sciences . . .. (Hume 1740, Introduction ¶9, p. xviii)

If Kant’s dependence on stipulated principles of human action without further explanation rules him out of the ranks of determinist empirical psychologists, then Hume must also be ruled out of those ranks.

[70] Sloan 2002: 232

[71] Sloan 2002: 233

[72] Sloan (2002) has detailed the role of Keime in Kant’s biology. He also argues persuasively for the introduction of Anlage as a technical term that provides Kant a “middle way” (238) between strict preformationism and vitalist epigenesis. For the purposes of this paper, the details of that account are not needed.

[73] For more detail, see Sloan 2002.

[74] 5: 400.

[75] Sloan persuasively argues that this was due to the influence of Blumenbach.

[76] Sloan 2002: 249

[77] 5:423.

[78] 5:422.

[79] 5:415, cf. 5:411

[80] 5:411

[81] 5:387

[82] 5:383.

[83] 5:376.

[84] 5:398.

[85] 5:390.

[86] Kant even seems to think that mechanist explanations are available in principle for any event in the world. Unlike Kant’s claims in the first Critique that it is impossible to know things in themselves, his claim that there will never be a Newton of a blade a grass does not seem to be a claim about what is in principle possible for human beings. Rather, he seems to think that mechanist explanations of biological phenomena are simply too complex for exhaustive comprehension for human beings.

[87] 7:175, 274.

[88] 8:111

[89] For Newton’s use of gravity in explanation, see especially his Principia (e.g. Definitions III-V).

[90] Kant also discusses passions (Leidenschaften) in connection with the faculty of desire, but these are derivative on his notion of inclination. For more on the nature of passions, see Borges 2004, Frierson 2000, and Sorenson 2002.

[91] 7:365; cf. 8:111f.; 25:796, 1109, 1111-4, 1334, 1339, 1514, 1518-9.

[92] 25: 1518, cf. 8:111f., 25:1109.

[93] For the sexual instinct, see 7:179; 8:112; 25:797, 1334, 1339. For the parental instinct to provide for young, see 7:265; 25:797, 1113, 1518. For the sucking instinct, see 25:1339, 1514, 1518. For instincts for food, see 8:111. For the “natural instinct to test [one’s] powers,” see 7:263. And for natural sympathy (Mitleid), see 25:1518.

[94] 8:111.

[95] 25:1518.

[96] 6: 29; 25:1112, 1339, 1517.

[97] 4:400, cf. too 27:676.

[98] 7:271.

[99] 25:1114, cf. too 25:1514.

[100] 25: 1519.

[101] 25:1339.

[102] 25:1114 my emphasis, cf. too 25:1514.

[103] 6:29.

[104] 25:1514.

[105] 25:1111-2. Kant adds here, “One can put instinct between propensity and inclination.” Like propensities, instincts are innate in human beings. Like inclinations, instincts provide direct explanations for various desires. (Kant is not entirely consistent about the relationship between inclinations and instincts. Generally (e.g. 25:111-2 and 7:265) he distinguishes between them as two different sorts of explanation for desire. But elsewhere he draws lines a bit differently. In a metaphysics lecture from 1782-3, Kant reportedly says, “If the stimuli <stimuli> have become habitual, then they are inclinations and their source is instinct or habit” (29:895), suggesting that the “source” of some inclinations is instinct. And elsewhere (25:1518) Kant says that the “inclination [Neigung] of parents towards their children is also instinct.” The account that I give here of the relationship between instinct, inclination, and propensity is the dominant one in Kant’s published works and lectures, but occasionally, at least according to his students, Kant relates instinct and inclination in different ways.)

[106] 7:265, cf. 25:1517.

[107] 25:1339.

[108] 25:1339.

[109] In fact, Kant periodically changes the relevant people-group in discussing this example. His overall view seems best captured by his claim that “Human beings across the whole world have a propensity to drink [alcohol]” (25:1112), as an example of which he sometimes uses “northern peoples” (25:1339), sometimes “the wildest peoples” (25:1112), sometimes “nations that have wine” (25:1518), and once even the “people in Kamtschatka, [who] have a certain cabbage, which when they eat it, works in them a kind of madness, for which they love to have it” (25:1518). The point of these examples is not to pick out any particular group, but to show that this propensity is universal. Thus the passage in the Religion, often taken as an offensive way of distinguishing “savages” from Europeans, actually reflects the fact that the propensity to drink was undeniable in the case of Europeans, but some might claim that this propensity itself is acquired, against which Kant cites the case of savage, or “raw” (rohe), people.

[110] 6:29.

[111] We can make the pictures a bit more complicated by inserting the distinction between feeling and desire. Then the account for instinct will look as follows (and the account for inclination will be altered in just the same way):

Sensory cognition à Anticipatory Pleasure à Desire

↑ ↑

Instinct* Nature of pleasure

The notion of “anticipatory pleasure” refers to the pleasure that causally effects the desire, in contrast to any expected future pleasure or past pleasures that may have cultivated one’s inclinations. For more on the nature of anticipatory pleasure, see Morrison 2004. Because the nature of pleasure will always explain the connection between anticipatory pleasure and desire, I have simplified the diagrams in the rest of my discussion by conflating pleasure and desire.

[112] What it means here to say that instinct explains the desire to suck is just that the natural scientist is leaving the desire to suck unexplained, allowing the positing of a new causal law – “Humans desire to suck.” Likewise the propensity to drink alcohol is a simply a new causal law – “northern peoples or savages will desire alcohol if they experience it.”

[113] A539/B567, cf. 25:634. For more on the importance of character for Kant’s moral philosophy, see Munzel 1999 and Kuehn 2001.

[114] A539/B567.

[115] 7:292.

[116] See Kuehn 2001:147.

[117] 25: 1175, cf. 25: 630, 651-52, 1384.

[118] 7:292, cf. 6:651-52.

[119] For more on the nature of character, and especially how one’s character can be evil and still admirable, see Frierson (forthcoming – JHP).

[120] 7:285.

[121] Kant even suggests that those who do not act consistently on the basis of principles have a kind of “bad character” (schlechte Character, cf. 25:650, 1172). This character is not a character in the strict sense, because it does not involve acting consistently on principles of the understanding, but it is a state of character in that it provides a basis for explaining why the connection between principles and desires is not constant. Thus it plays an explanatory role in an unstable causal account.

[122] 25:1172; cf. 7:294.

[123] 25:1172.

[124] 7:285.

[125] In {part of footnote deleted for the sake of anonymity}, I discuss the way in which these empirical features characterize human beings as free and morally responsible. See too Jacobs 2003.

[126] 25:651, 823, 1172, and 1176.

[127] 7:285, 290; 25:1388.

[128] 25: 1388.

[129] 7:290.

[130] 7:286.

^[131] Munzel 1999 examines the role of pedagogy in the cultivation of character in detail. See especially chapter 5.

[132] 25:1172.

[133] 7:294-5. Emphasis is my own.

[134] 7:294.

[135] 25: 635, cf. 7:325; 5:154; 25:599, 722ff., 1386. Kant even gives recommendations about different ways to teach boys and girls (25:1392), suggesting that the influences on character can get quite fine-grained.

[136] 5:156.

[137] 5:156.

[138] Kant emphasizes the empirical nature of this potential for motivation by moral reason in the second Critique. He introduces his account of educating the ten-year-old boy by claiming that “We will . . . show, by observations anyone can make, . . . this property of our mind, this receptivity to a pure moral interest and hence the moving force of the pure representation of virtue” (5:152, my emphasis). Just as one presents empirical reasons for believing in various human instincts, propensities, and inclinations, so one can show by observation that human beings have an innate possibility for moral motivation. And the receptivity that one finds in human beings for such motivation is precisely a receptivity to empirical influences, such as the telling of a vivid story of moral virtue. And this receptivity is itself the result of empirical causes, though human science may never be able to discern precisely how it arose from them.

[139] 5:155.

[140] 6:479f.

[141] Elsewhere Kant points out that even in adults, one can directly influence whether another acts on the basis of the moral law. As he explains,

This happens when the other . . . confronts the subject with . . . the moral law by which he ought to act. If this confrontation makes an impression on the agent, he determines his will by an Idea of reason, creates through his reason that conception of his duty which already lay previously within him, and is only quickened by the other, and determines himself accordingly to the moral law. (27:521)

Kant’s emphasis on the role of reason, and even an “Idea of reason,” should not distract from what is actually going on here. One person is able, through conversation, to cause another to have a new higher cognition, an Idea of reason, and this higher cognition in turn causes a new determination of the faculty of desire, according to which the person acts morally. This is, for Kant, an example of how “a person may be compelled to duty by others” (27:521).

[142] For more on the role of politeness in cultivating character, see Brender 1997, 1998, and Frierson (forthcoming – Kantian Review).

[143] 8:375. Brian Jacobs has pointed out that “as he makes clear from his ‘Idea’ essay of 1784, Kant thinks that precisely these aspects of human life [“social, cultural, political, and historico-teleological characteristics”] are as determined as natural events” (Jacobs 2003: 112).

[144] 7:329.

[145] Kant’s account of the origin of character is not limited to external influences on character. He claims, for instance, that the cultivation of one’s propensity to character comes “through understanding and reason” (25:1172). (Immediately after saying this, however, Kant reiterates that “the acquisition of . . . character . . . happens through education.” For Kant, cultivating character through reason and understanding is a pedagogical task, not a solely individual accomplishment.) In that context, Kant discusses several specific rules that one can follow in cultivating character in oneself and others:

a) Not to speak an untruth intentionally . . ..

b) not to dissemble . . ..

c) not to break one’s legitimate promise . . ..

d) not to join the company of evil-minded people . . ..

e) not to pay attention to slander . . ..” (7: 294, cf. 25:1387-88, 1392, LA 1789-90: 130)

These are all practical principles that support and constitute the development of character as such. The pursuit of these methods for developing character depends on already having at least some level of character. Unless one can act on the basis of principles, one will be unable even to follow the principles for developing character. But keeping these principles even sporadically can have some beneficial effect. The more one avoids duplicity, bad company, and slander, the easier it will be for one to stick to principles in the future. Insofar as one has some minimal level of constancy, these principles can reinforce one’s character. They are important aids to self-improvement, even if they are not sufficient.

It is important to note, however, that these are not simply maxims for self-improvement but maxims based on causal laws governing the formation of character. Some of the maxims conducive to character actually embody what character is (for more on this, see Frierson (forthcoming – JHP)). Thus refraining from untruths and dissembling is crucial to actually being consistent to who one is, to one’s principles. To have character just is, in part, to follow these rules. But the rules also depend on certain regular connections of causes and effects. Thus one should not join bad company or pay attention to slander because bad company and slander can causally interfere with one’s development of character. And even refraining from speaking untruth is an activity that can have a negative effect on one’s development of character.

The emphasis on practical rules rather than mere causal explanations is, moreover, particularly appropriate in the context of a pragmatic anthropology, which seeks not simply to “ponder natural causes” but to “use our observations” for self-improvement (7:119). (For more on the nature of “pragmatic” anthropology, see Frierson 2003: 48-56, Louden 2000: 68-70, Wood 1999: 202-7, Brandt and Stark Einleitung to Ak. 25, and Stark 2003.) For Kant, explanations of character in terms of natural causes are possible and helpful, but only if put to use. Thus it is natural for him not only to give causal explanation but to formulate these in terms of rules that one can follow in cultivating one’s capacity for character. Kant appeals to causes of character that are within a person’s control for the same reason that he focuses on causes that are within the control of educators, because these causes can be put to pragmatic purpose.

The rules for the cultivation of character suggest that there is the possibility of a causal loop in the cultivation of character. Insofar as one begins to develop a character, one can more easily act on the basis of principles. And some principles are conducive to further cultivation of character, so acting on the basis of those will contribute (causally) to a deeper cultivation of character. A causal circle is possible here where small improvements build on each other to produce a character in a full sense. (Admittedly, there is some tension here between this circle, which suggests a gradual development, and Kant’s appeal to the importance of an “explosion” (7:295) that suddenly gives rise to character, but a full exploration of problems arising within Kant’s account of the development of character is beyond the scope of this paper.)

[146] 28:585, cf. too 29:907.

[147] 7:119. For an analysis of this passage that shows the empirical nature of anthropology, see Frierson 2003: 33-4.

[148] Hume 1740: I.I.ii.1, p. 8. Hume gives a bit more explanation for refusing to consider the origin of sensory impressions, arguing that “the examination of our sensations belongs more to anatomists [here meant in the literal sense] and natural philosophers [i.e. scientists of body] than to moral [i.e. psychologists]; and therefore shall not at present be enter’d upon” (Hume 1740: I.I.ii.1, p. 8).

[149] For a detailed discussion of Kant’s account of the origin of spatial perceptions, including some reference to his empirical account and its early modern predecessors, see Hatfield 1990.

[150] 7: 155, 156.

[151] The “senses” include, for Kant, not only the 5 outer senses but also the “inner sense,” and Kant gives an account of that power in his anthropology and empirical psychology. See 7:161f..

[152] 7:167-8.

[153] 28:674, cf. 7:175-77, 182.

[154] 7:176, cf. Hume 1740: I.iii.vi.4 and I.iii.vii.6, pp. 88-9, 97. Kant goes on here to explain that this principle of association is simply a basic causal law, not something that one should seek to reduce to something else: “To try to explain this in physiological terms is futile; we are free to use some principle that will always remain a hypothesis” (7:176). This positing of a basic principle is consistent with Kant’s Newtonian conception of scientific explanation, as discussed above.

[155] 7:165.

[156] 7:169ff.

[157] 7:181.

[158] 7:173.

[159] Strictly speaking, all I have provided here is the basic framework for Kant’s account. A fully worked out causal account would need to exhaustively catalog all human instincts and propensities and explain their laws of operation.

[160] For the spontaneity of the higher faculties, see 28:228, 29:880, and 28:584.

[161] 29:915.

[162] 25:1172.

[163] A66, cf. Sloan 2002:230.

[164] Here Kant says,

Humanity is easily distinguished from all other natural beings by his technical predisposition for manipulating things (a mechanical predisposition joined with consciousness), by his pragmatic predisposition (for using other men skillfully for his purpose) and by the moral predisposition in his being (to treat himself and others according to the principle of freedom under laws). (7:322)

This list of predispositions is similar to, but importantly different from, the three forms of the “predisposition to good” in the Religion. To highlight just one important difference, the predisposition to animality in the Religion is not itself rational and is a predisposition (or set of predispositions) that does not distinguish human beings from other living beings on the earth. By contrast, the predispositions in the Anthropology all distinguish humans from animals. For a different interpretation of the relation between these lists, see Kain 2003.

[165] 6:444-5.

[166] In some lectures (e.g. 25: 780), Kant makes a distinction between two different sorts of natural predispositions, talents and capacities, and he claims that the understanding is a capacity (like the senses and imagination) while reason and judgment are talents. Kant uses this distinction to explain different ways that one can promote each faculty.

[167] See especially Kant’s “Idea for a Universal History,” Groundwork (4: 395) and Critique of Judgment.

[168] See Kant’s “Conjectures on the Beginning of Human History.”

[169] A549/B577, A803/B831.

[170] 27:502-4, c. 1793.

[171] B1, cf. 29:951-2.

[172] B1-2.

[173] Similarly in his lectures on metaphysics, Kant explains that “with human beings all representations commence with objects of experience,” such that “in order to obtain cognitions, even concepts of the understanding, our faculty of cognition must be awakened by objects of experience” (29:951). But he again quickly points out that there are nonetheless “a priori concepts . . . that cannot be derived from experience” (29:952, emphasis added).

[174] Jasche Logic, 9: 14. Hatfield (1990: 73) shows how Kant is here opposing the Wolffian reliance on empirical psychology as a source from which logical principles can be “borrowed” (cf. Wolff’s Psychologica empirica, Prolegomena, §§ 4, 5, and 9). Kant, by contrast, studies logic on its own terms, but then uses the insights of logic as a guide to his empirical psychology. Like Wolff, however, Kant maintains a connection between logic and psychology. The difference between the two is that Wolff starts with empirical psychology and uses its principles in his logic. Kant starts with logic but makes use of logical principles for developing his empirical account of cognition. In this sense, Kant’s approach to the relation between empirical psychology and logic is the opposite of Wolff’s.

[175] Jasche Logic, 9: 11.

[176] Jasche Logic, 9: 11.

[177] Jasche Logic, 9: 14.

[178] 7:202f.

[179] 7:165.

[180] For both quotes, see 24:721. Cf. too 9: 54, where Kant says, “Every error into which the human understanding can fall is only partial, however . . .. For a total error would be a complete opposition to the laws of the understanding and of reason. But how could that, as such, in any way come from the understanding and . . . be held to be a product of the understanding.”

[181] A detailed study of empiricist approaches to the “higher” faculties of cognition would be well worth studying, both for its own sake and for comparison with Kant, but such a study is beyond the scope of the present essay. One might also usefully compare Kant with Wolff here (cf. Hatfield 1990: 72-5), but Wolff is less clearly empiricist than Hume.

[182] Hume 1740: I.iii.vi.4, pp. 88-9. This passage has been discussed in detail among Hume commentators. Cf. e.g. Baier 1994, Garrett 2002, and Stroud 1981.

[183] But see footnote 9. Kant’s account is less determinist than Hume’s in that there is more to the story than determinism. Unlike Hume, Kant does not think that the ultimate ground of human actions can be identified with their empirical causes.

[184] There is a good reason for the greater sophistication in Kant’s account of the higher faculty of cognition. Hume thinks that most belief is the result of principles of the imagination, so his psychology focuses on the nature of that cognitive power. Kant’s recognition of the possibility of prejudices and disorders as problems with the understanding itself helps support his expanded account of the role of the understanding, and his sense of this expanded role makes him more interested in and attuned to problems specific to the understanding.

[185] For the rest of this section, I focus on the understanding in the narrow sense, though Kant’s accounts of judgment and reason are similar. With respect to the power of judgment, Kant claims in his logic that the operative laws are “induction and analogy,” and Kant discusses various kinds of syllogisms related to reason (9:132, 133f.). In both cases, Kant’s account is brief. (Perhaps the most extensive detail with regard to the ways in which reason can be unhealthy is in Kant’s first Critique accounts of the illusions of reason in the transcendental dialectic. In his anthropology, Kant explains that illusions can cause higher cognitive faculties to diverge from the rules that govern healthy reason: “to avoid errors . . . one must seek to disclose and to explain their source: illusion” (Jasche Logic, 9:56). And the account in the first Critique shows several examples of this divergence. For more on illusions of reason in the first Critique and Kant’s early theoretical writings, see Grier 2001.)

Again, Kant’s primary interest in his empirical accounts of the higher cognitive faculties is practical. In this context in particular, he points out that while the understanding can be “educated,” the faculty of judgment can only be “trained” (geubt) or “cultivated” (cultiviert). (For geubt, see 7:199 and 25:777. For cultiviert, see 25:1476. Cf. too 24:722; 25: 774, 1477; 29:890.)

[186]Jasche Logic, 9:114ff. This specific example is from 9:119.

[187]Jasche Logic, 9:95.

[188] 7:199, cf. 25:777, 1476.

[189] 9: xxx-xxx

[190] 9: xxx

[191] 9: xxx

[192] He says, “Language signifies the presence of thought and [is] the means par excellence of intellectual signification . . ., the most important way we have of understanding ourselves and others” (7:152).

[193] 7:155.

[194] The first quotation is from 7:155; the second is from 7: 192-3.

[195] I have discussed this with respect to the lower cognitive faculty above. With respect to the faculties of feeling and desire, Kant’s account of disorders focuses on the affects and passions. See 7: 252-275; 25:215-18, 414-26, 613-24, 1114ff., 1353-1364, 1526-1530.

[196] Jasche Logic, 9:75.

[197] Jasche Logic, 9:79.

[198] Jasche Logic, 9:79. The emphasis on “causes” is in the original. (Of course, the original is based on a set of notes, so this is not necessarily Kant’s emphasis, but the notes are Kant’s own.)

[199] Jasche Logic, 9:76.

[200] Kant’s letter to Herz, 10:146; 78.

[201] Kant’s letter to Herz, 10:146; 78.

[202] 7:202.

[203] See 7:204 for Kant’s account of dullness. This definition of intelligence is at 7:201.

[204] 7:204.

[205] 7:212.

[206] 7:212.

[207] 7:214.

[208] 7:214.

[209] For each mental illness, Kant provides both a German and a Latin term. I give both here.

[210] 7:214-16.

[211] 7:217. For more on the relationship between Keime and predispositions, see Munzel 1999 and Sloan 2002.

[212] 7: 218.

[213] 7:208, cf. 7:185.

[214] At least in the empirical sense (see footnote 9).

[215] Blackburn 1998: 248, 252.

[216] {Acknowledgments footnote deleted for anonymity.}