Abstract
The aim of this paper is to defend the interpretation that Kant is committed to psychological determinism and reconstruct his deterministic account of human action. I begin by showing that some arguments against Kant’s commitment to psychological determinism are not sound. I then argue that Kant’s psychological determinism is grounded in his argument for the universal causal principle of the Second Analogy. After that, I examine Kant’s conception of the empirical character, arguing that the empirical character, construed as a complex of distinctive dispositions to adopt certain maxims and act on them, constitutes the supreme ground of a deterministic explanation of human action. Finally, I suggest that Kant is also committed to the predictability thesis of human action, albeit in a limited sense.
1 Introduction
Immanuel Kant famously claims that human action is in the world of sense necessarily determined, and when he attempts to reconcile the conflict between free will and determinism in the resolution to the Third Antinomy of the Critique of Pure Reason (henceforth the first Critique), he thinks that the genuine problem merely consists in whether human action can also be seen as transcendentally free (KrV, A536/B564).1 Kant thus develops his theory of freedom both in theoretical and in practical philosophy. While much scholarly attention has been devoted to Kant’s conception of freedom and to its compatibility with determinism, less attention has been devoted to understanding exactly what determinism means for Kant. In particular, given that human action is not directly caused by physical events but rather by mental events, it is often insufficiently clear whether Kant is committed to some kind of psychological determinism, what the exact nature of this determinism is, and how it relates to the predictability of human action. To bring greater clarity to these issues, I propose to investigate Kant’s deterministic account of human action.
In the literature, most commentators mention Kant’s endorsement of psychological determinism without going into detail about this doctrine (Allison, 1990, p. 34; Bojanowski, 2006, p. 171; Indregard, 2018, p. 675; Onof, 2018, p. 1109; Scholten, 2022, p. 89; Kohl, 2023, p. 57; Onof, 2024, p. 96). Other commentators argue that Kant’s theoretical philosophy, especially his theory of causality, does not allow him to accept psychological determinism. For example, according to Michael Friedman (1992, p. 182), the transcendental causal principle of the Second Analogy of the first Critique can be applied only to spatial objects, which implies that he denies mental causation and causal determinism in the psychological field. In addition, Robert Hanna (2006, p. 440) contends that Kant does not take the psychological laws governing mental phenomena as deterministic because he notoriously rejects the scientific status of empirical psychology. What is more, Kenneth Westphal (2004, p. 230) maintains that the joint conclusions of the Analogies and the Paralogisms of the first Critique implicitly entail Kant’s denial of psychological determinism.
The aim of this paper is to defend the interpretation that Kant is committed to psychological determinism and reconstruct his deterministic account of human action, arguing that Kant is justified in claiming that human action is psychologically determined. In so doing, I shed light on the essential role of Kant’s conception of the empirical character in his psychological determinism. The paper proceeds as follows. I begin by considering some arguments against the view that Kant accepts psychological determinism (Section 2). I then attempt to show how causal determinism relates to the principle of the Second Analogy and what is characteristic of psychological determinism (Section 3). After that, I turn to examining Kant’s conception of the empirical character, focusing mainly on its relation to psychological determinism (Section 4). Furthermore, I clarify Kant’s view of prediction of human action (Section 5). Finally, I conclude by summarizing my reconstruction of Kant’s account of psychological determinism and pointing out some open questions (Section 6).
2 Some Arguments against Kant’s Commitment to Psychological Determinism
There is no doubt that in the resolution to the Third Antinomy of the first Critique Kant endorses a form of determinism and devotes himself to making freedom compatible with it (KrV, A541/B569). It is also indisputable that in many passages of his writings Kant makes some statements about psychological determinism, albeit at times implicitly. For instance, he claims that the “human soul” is “subject to natural necessity” and “is not free” (KrV, Bxxvii); human actions “as belonging to the world of sense must … be understood as determined by other appearances, namely desires and inclinations” (GMS, 4: 453).2 However, some commentators argue that despite his own statements Kant in fact does not accept psychological determinism. Their arguments for this interpretation can be summarized as follows.
In the Second Analogy, Kant does not justify the universal causal principle (Lovejoy, 1906, p. 402; Strawson, 1966, p. 138),3 nor does he prove the universal causal determinism elsewhere (Westphal, 2012, p. 260), so he does not argue for psychological determinism.
Regarding the problem of free will and determinism, Kant holds a libertarian view (Kane, 2005, p. 33; McKenna & Pereboom, 2016, p. 243), thus he rejects determinism in any form, including psychological determinism.
The universal causal principle of Kant’s Second Analogy cannot be applied to mental phenomena, so for Kant there exist no causal relations and causal laws in the psychological field (Meerbote, 1984, p. 156; Hudson, 1994, p. 69; Nayak & Sotnak, 1995, p. 149; Friedman, 1992, pp. 182–183; Ertl, 1998, p. 164; Westphal, 2004, pp. 230–238; Hanna, 2006, p. 440; Pollok, 2013, p. 737). Consequently, Kant denies causal determinism in the psychological field.
In his Metaphysical Foundations of Natural Science (henceforth Metaphysical Foundations), Kant claims that empirical psychology is not a proper natural science; hence he must also think that mental phenomena, which belong to the subject matter of empirical psychology, are not subject to any causal determination (Meerbote, 1984, p. 155; Nayak & Sotnak, 1995, p. 147; Westphal, 2004, p. 235; Bojanowski, 2006, pp. 172–176; Friedman, 2013, p. 6).
Since the first two arguments are not directly related to Kant’s own understanding of psychological determinism per se, I will not discuss them in this paper.4 The last two arguments are in fact closely interrelated, but I shall treat them respectively due to the difference of their argumentative force.
Admittedly, the third argument seems sound at first glance. For in the second edition of the first Critique – especially in the Refutation of Idealism and the General Note on the System of Principles – Kant does point out that the legitimate theoretical application of the categories requires outer intuition (KrV, B291), from which it appears to follow that mental phenomena as objects of inner sense cannot be determined by categories such as substance and causality. In other words, all nonphysical objects must be excluded from Kant’s epistemology. This is the key point when Westphal (2004, p. 232) takes a realist position to interpret Kant and claims that the joint conclusion of the Analogies and the Paralogisms disproves psychological determinism.5 And this is also the gist when some commentators attribute to Kant the Davidsonian Anomalous Monism, arguing that in Kant there are neither psychophysical nor merely psychological laws, since a mental event as token is identical with a physical event as token that is governed by physical causal laws (Hudson, 1994, pp. 67–70; Meerbote, 1984, p. 156).6 Both interpretations obviously belong to a radically physicalist view, according to which the lawfulness and knowability of the mental are completely ruled out.
However, the emphasis on outer intuition as a cognitive condition is only part of the story in Kant’s theoretical philosophy. In fact, Kant never denies the existence of the phenomenal self and the possibility of inner experience. As he repeatedly puts it on various occasions, we can cognize ourselves as we appear to us (KrV, B68, B152, B156, A379, A382, A690/B718). This implies that, alongside our bodies and behaviors, our mental phenomena can also be objects of empirical self-knowledge. But how is it possible to apply the categories of substance and causality as well as the universal causal principle to the phenomenal self and its mental events?7
One possible answer to the question is that the transcendental idea of the soul serves as a guiding thread under which all mental phenomena can be ordered in accordance with the categories and principles of pure understanding.8 More specifically, on the basis of the systematicity contributed by the idea of the soul, we can cognize our own mental events and their relations to actions as being governed by psychological causal laws, and this happens in various a posteriori ways, including observation, comparison, abstraction, and so on.9 As Kant puts it, the idea of the soul as a principle of the systematic unity plays a regulative role “in explaining the appearances of the soul, namely by considering all determinations as in one subject, all powers, as far as possible, as derived from one unique fundamental power, all change as belonging to the states of one and the same persisting being” (KrV, A682/B710). Although Kant here mentions the change of mental states, he does not spell out how the psychological causal laws governing this mental change can be systematically discovered under the guiding idea of the soul. I now want to briefly illustrate this procedure by analyzing an example.
A person ate an apple for the first time and it tasted delicious. Then she had several similar experiences, and this makes her recognize a causal relationship of different representations: Once she sees an apple, the representation of the apple brings about a representation of deliciousness; this representation of deliciousness in turn gives rise to a desire to eat it; and this desire finally leads to an action of picking up the apple. Here it is easy to see that the person builds an existential, content-based, and causal relationship among the representations. But the application of the category of causality and the universal causal principle to the representations is possible only if one presupposes that the different representations in the causal judgments are attributed to one and the same subject and organized in the systematic unity of this subject. Yet the underlying subject is merely a transcendental idea under which the empirical self can be constituted.10
The analysis of the example shows that because mental events take place in time and must presuppose something preceding, they all stand in a causal chain governed by empirically discovered psychological laws. As Thomas Sturm (2001, p. 170) rightly points out, Kant “accepts causal laws governing the alterations of perceptions and of other psychological phenomena as well. The law of association of ideas by spatiotemporal contiguity is just such a law.” With these considerations in hand, I now turn to the problem of the scientific status of empirical psychology.
In the Preface to the Metaphysical Foundations, Kant famously claims that “the empirical doctrine of the soul,” that is, empirical psychology, does not have “the rank of a properly so-called natural science,” since “mathematics is not applicable to the phenomena of inner sense and their laws” (MAN, 4: 471). Kant then goes on to point out that this in turn is because mental phenomena are constructed only in one-dimensional time. He thus regards empirical psychology as “an historical doctrine of nature,” “a natural doctrine of inner sense,” or “a natural description of the soul” (MAN, 4: 471). These remarks seem to support the fourth argument mentioned above. But I do not think that Kant’s attitude toward empirical psychology here threatens the reading of Kant as a psychological determinist at the phenomenal level.
First, the Anticipations of Perception of the first Critique clearly shows that perceptions as intensive magnitudes can be mathematized (KrV, A166/B209). For instance, inner perceptions as objects of psychology can of course be analyzed by applying arithmetic methods. Therefore, what Kant is denying in the aforementioned passage of the Metaphysical Foundations is only the applicability of geometry to the phenomena of inner sense.11 Gary Hatfield (1992, p. 221) also points out that “the reason Kant ruled out the possibility of a scientific psychology was not a claim that mathematics could not be applied to inner sense at all, but that it could not be applied a priori.”
Second, Kant’s denial of the scientific status of empirical psychology does not entail that he denies the existence of psychological causal laws.12 In the pertinent passage of the Preface to the Metaphysical Foundations, Kant only speaks of “the law of continuity in the flux of inner changes” (MAN, 4: 471), but as many commentators note (Hatfield, 1992, p. 219; Sturm, 2001, p. 170), the law of association of representations, which Kant mentions in the first Critique (KrV, A112, B142), is supposed to be the most representative one of psychological causal laws. In the Anthropology from a Pragmatical Point of View (henceforth Anthropology), Kant even calls empirical psychology a “sum-total of inner perceptions under natural laws” (Anth., 7: 141).13 I shall revisit this issue in the next section.
Finally, given Kant’s considered view that mental phenomena are law-governed and can be systematically investigated, we are entitled to say that Kant does not completely dismiss empirical psychology as a pseudoscience; instead, empirical psychology is for Kant a systematic, self-contained study of the phenomenal self and thus qualifies as an “improper natural science.”14 Consequently, if psychological determinism presupposes empirical psychology as a science, then it is justified for us to think that Kant’s attitude toward empirical psychology in the Metaphysical Foundations does not disprove his commitment to psychological determinism. The fourth argument mentioned above is thus not plausible.
To sum up, I have examined some existing readings of whether Kant is committed to psychological determinism, arguing that the argument from the principle of the Second Analogy and the argument from the nonscientific status of empirical psychology do not suffice to support the view that Kant denies psychological determinism. In the next two sections, I shall attempt to show that the universal causal principle of the Second Analogy constitutes the ultimate ground of Kantian psychological determinism, and such a determinism can be fleshed out with Kant’s conception of the empirical character.
3 The Second Analogy, Causal Determinism, and Psychological Determinism
According to Kant, everything that happens in space and time is “determined without exception in accordance with laws of nature” (GMS, 4: 455). In the transcripts of his metaphysics lectures, Kant also says: “All things in nature, be they inner or outer events, have their determining cause, they all happen according to natural laws and are also determined according to them” (V-Met-L2/Pölitz, 28: 582). These two remarks show that Kant is committed to global or universal causal determinism, according to which both physical and mental events in the sensible world are causally determined, and this causal determination is necessarily governed by laws of nature.15 But how exactly does Kant account for the relationship between laws of nature and causal determinism? In what respect is psychological determinism a form of causal determinism? This section is devoted to addressing these two questions.
3.1 The Universal Causal Principle as the Ground of Causal Determinism
As generally recognized, causal determinism is the thesis that everything that happens in space and time is necessitated by some preceding events in accordance with laws of nature. In this characterization three points are noteworthy. First, causal determinism concerns all spatiotemporal events, which means that both token-events and type-events are causally determined, and this causal determination is thus absolutely universal. Second, causal determinism entails that the connection between an event as cause and another event as effect is necessary; that is to say, given the existence of the former, it is impossible that the latter does not result from it. Third, causal determinism requires that the causal nexus of events is governed by laws of nature, which implies that this nexus is constant or regular. But in fact, the first two points can be seen as united in the third one because the concept of universality and that of necessity are integral parts of the concept of law in the strict sense. Accordingly, an understanding of causal determinism turns essentially on the notion of laws of nature.
When in the Third Antinomy Kant sets out to reconcile the conflict between free will and causal determinism, he refers back to the most fundamental causal law of nature, that is, the universal causal principle of the Second Analogy: “The correctness of the principle of the thoroughgoing connection of all occurrences in the world of sense according to invariable natural laws is already confirmed as a principle of the transcendental analytic and will not suffer violation” (KrV, A536/B564). This suggests that Kant views the problem of causal determinism as already settled in the Second Analogy, and he thus needs not do anything with respect to the certainty of causal determinism.16 However, it is not clear to what extent Kant links causal determinism to the universal causal principle. According to Westphal (2012, p. 261), Kant in effect does not justify causal determinism in the Second Analogy. Lucy Allais (2023, p. 30; 2018, p. 717) recently maintains that Kantian causal determinism “does not rule out the possibility of there being more than one way the world could unfold in space and time.” Onof (2024, p. 143) also contends that the human agent has the ability to do otherwise in time, that is, the temporal leeway. In what follows, I shall argue that Kant’s account of the universal causal principle in the Second Analogy already implies a justification for causal determinism, and such a determinism does not allow an open future.17
According to a convincing interpretation of Kant’s theory of causality, particular empirical causal laws of nature are grounded in or determined by the universal causal principle of the Second Analogy (Buchdahl, 1965, p. 198; Friedman, 1992, pp. 171–175).18 For only by injecting the transcendental necessity of the universal causal principle into the inductively observed regularities or empirical rules can we cognize or discover them as genuinely empirical causal laws. Thus, every event is necessitated by some previous event in accordance with both particular empirical causal laws of nature and the universal causal principle. For example, the falling of an apple is governed directly by the law of gravitation, the necessity of which in turn has its basis in the universal causal principle. Consequently, the universal causal principle lies at the most fundamental level of all kinds of causal laws and is central to causal determinism.
Kant’s argument in the Second Analogy is, roughly speaking, that knowledge of determinate objective sequences of events is possible only if the sequences of events are subsumed under the schema of the category of causality and thus are governed by the universal causal principle of understanding “everything that happens has a cause” or, as Friedman (1992, p. 186) describes it in a clearer way, “all events of the same kind as A are followed by or result in events of the same kind as B.”19 Kant’s account of this principle sheds light on the essential features of causal determinism: universality, necessity, and lawlikeness.20 He writes:
Thus if I perceive that something happens, then the first thing contained in this representation is that something precedes, for it is just in relation to this that the appearance acquires its temporal relation, that, namely, of existing after a preceding time in which it did not. But it can only acquire its determinate temporal position in this relation through something being presupposed in the preceding state on which it always follows, i.e., follows in accordance with a rule: from which it results, first, that I cannot reverse the series and place that which happens prior to that which it follows; and, second, that if the state that precedes is posited, then this determinate occurrence inevitably and necessarily follows. (KrV, A198/B243–244, emphasis added)
Kant here makes it clear that the occurrence of an event B is temporally determined in relation to an antecedent event A, and this successive determination is irreversible. He also emphasizes that the event A is “inevitably and necessarily” followed by or results in the event B. And he concludes that both the irreversibility and the necessity of the sequence of the events A and B are based on a rule, that is, the universal causal principle.21 I take it that Kant’s argument for the universal causal principle establishes a form of causal determinism, according to which the occurrence of every event is fixed in terms of its temporal relation and causal necessity. For the universal causal principle entails that if some antecedent event is given, then another event inevitably follows in time and is invariably determined in terms of its content. Once again, the crucial point here is that the only possible way in which an event happens is necessarily determined by a particular empirical causal law, which in turn is grounded in the universal causal principle. As a result, Kantian causal determinism, construed on the basis of the Second Analogy, allows only one possible future.
3.2 The Characteristic of Psychological Determinism
Human beings are a part of nature. Given that human actions as appearances are caused by mental events that happen in time, it is indisputable that human actions are subject to the conditions of time and thus governed by the universal causal principle of the Second Analogy. As Kant succinctly puts it, “the preceding time necessarily determines the following time,” and “the appearances of the past time determine every existence in the following time” (KrV, A199/B244). Now, because the past mental events are not up to us, neither are the various causal laws of nature up to us; all human actions are also not up to us. That is to say, human actions are causally determined.
However, such a “short argument” for the deterministic feature of human action is not satisfying, although it completely squares with Kant’s theory of causality. For, while physical events, including neurophysiological events, fit well into the causally deterministic picture depicted above – call it physical causal determinism – the case of mental events is quite another thing and requires further clarification. What characteristics then does psychological causal determinism possess?22 To put it another way, in what sense can human action be regarded as psychologically determined?
Admittedly, when Kant in many places speaks of causal determinism concerning human action, he seems to intentionally ignore the peculiarity of this determinism. That is, he does not pay much attention to the psychological dimension of human action. He simply takes it for granted that causal determinism is true not only in the physical but also in the psychological realm, which appears to show that Kant commits himself to a form of naturalism with respect to human mind. For example, in the Prolegomena to Any Future Metaphysics That Will Be Able to Come Forward as Science, Kant remarks that a universal natural science “would have to bring nature in general – whether pertaining to an object of the outer senses or of the inner sense (the object of physics as well as psychology) – under universal laws” (Prol., 4: 295). In the Critique of Practical Reason, he points out that “the sensible world, as a sensible nature (in what concerns rational beings)” is subject to “the mechanism” (KpV, 5: 43). He then reiterates that “in the explanation of events in the world and so too of the actions of rational beings, I grant the mechanism of natural necessity the justice of going back from the conditioned to the condition ad infinitum” (KpV, 5: 49; see also 5: 97).23 That Kant treats human action exactly the same way as treating physical events appears to suggest that for him psychological causal determinism is as strict as physical causal determinism. Nevertheless, as a matter of fact, Kant’s empirical account of human action already discloses the main characteristic of psychological determinism.24
The most striking feature that distinguishes psychological determinism from physical determinism and even the causal determinism concerning the behaviors of animals is that human beings act or can always act with deliberations, whereas the motions of material objects and animal behaviors are nonintentional and unconscious. More specifically, when choosing something, making decisions, or performing actions, human agents actively let some beliefs function in their mind, that is, they conduct a deliberating procedure. Indeed, human actions involve a kind of reflexive awareness.
Kant recognizes this characteristic quite well. For example, he defines human will as “the capacity to act according to the representation of laws” and contrasts it with nonhuman being that merely “works according to laws” (GMS, 4: 412; see also KpV, 5: 32). In the second Critique, Kant explicitly remarks that mental events have “psychological instead of mechanical causality, that is, produce actions by means of representations and not by bodily movements” (KpV, 5: 96). In the Religion within the Boundaries of Mere Reason (henceforth Religion), he also claims that human will “cannot be determined to action through any incentive except so far as the human being has incorporated it into his maxim” (RGV, 6: 23). Kant’s terminology here indicates that laws must be represented, and incentives must be incorporated by the agent so as to be able to determine the will to act. That is to say, only by becoming conscious of some determining grounds such as normative constraints or sensuous inclinations can the human agent bring about actions in order to achieve aims.25 For Kant, the causal relation between mental events and human actions is different from the one between physical events and bodily motions. Accordingly, the way in which human action is causally determined is not mechanistic in the narrow sense, and neither is it reducible to a physical or neurophysiological procedure. Indeed, Kant’s causal determinism about human action is not a physicalist one, but rather properly called psychological determinism.
It might be objected, however, that this interpretation does not accord with Kant’s conception of freedom, especially when the passage cited above from the Religion is associated with the so-called Incorporation Thesis (RGV, 6: 23), according to which a desire can determine the will only if the agent incorporates it into his or her maxim. For example, Richard McCarty maintains that the commentators who explicitly or implicitly support the Incorporation Thesis (e.g., R. Jay Wallace, Onora O’Neill, Christine Korsgaard, Henry Allison) deny that Kant embraces psychological determinism (McCarty, 2009, pp. 62–67). For to say that an agent deliberates and incorporates an incentive into his or her maxim is the same as saying that the agent freely chooses to do an action. Nevertheless, in what sense Kant is committed to the Incorporation Thesis remains in the literature a matter of controversy.26 Here I just want to point out that every occurrence of desire or inclination is an event that takes place in time, and the endorsing of a desire or the adoption of a maxim in practical reasoning is also a temporal process. This clearly shows that the explanation of human action in terms of maxim-adoption definitely has a phenomenal dimension. In other words, even if the Incorporation Thesis were really at the heart of Kant’s theory of human action, it would not rule out Kant’s acceptance of a psychologically deterministic account of human action. How Kant makes this phenomenal deterministic aspect compatible with the noumenal free aspect of human action is another question.
Up to this point, I have analyzed the essential features of causal determinism, arguing that the universal causal principle of the Second Analogy constitutes the ultimate ground of explaining the causal determination of every event. Moreover, I have suggested that physical determinism and psychological determinism are two subsets of causal determinism, and the fundamental difference between them lies in whether the determining grounds of an event are consciously represented by the acting substance, which I call the reflexive awareness. Therefore, to understand Kant’s psychological determinism, we need to spell out the psychological characteristic of this determinism in more detail. As we shall see in the next section, Kant’s conception of the empirical character offers a way of accounting for the psychological and causally deterministic aspect of human action.
4 The Role of the Empirical Character in Kant’s Deterministic Account of Human Action
In the contemporary debate about the problem of free will, determinism is generally defined in terms of past events and laws of nature. For example, Peter van Inwagen (1983, p. 65) offers the following definition: “Determinism is, intuitively, the thesis that, given the past and the laws of nature, there is only one possible future.” And he puts forward the famous Consequence Argument: “If determinism is true, then our acts are the consequences of the laws of nature and events in the remote past. But it is not up to us what went on before we were born, and neither is it up to us what the laws of nature are. Therefore, the consequences of these things (including our present acts) are not up to us” (van Inwagen, 1983, p. 16). I think these characterizations of determinism perfectly apply to material objects and physical events. Yet, when it comes to mental events and human actions, one could raise many questions for van Inwagen’s account of determinism.27 A serious problem is: Where does the agent or the agency go? It seems to me that van Inwagen’s thesis of determinism overlooks the bearer of mental events, the executor of natural laws, or the author of human actions. That is, he does not take into account the constitution of the agent itself, which is supposed to be one factor that plays a vital role in explaining human action. To use contemporary terms, the aforementioned description of determinism faces “the problem of the disappearing agent” or “the homeless agent problem” (Fischer et al., 2007, p. 102; Onof, 2024, p. 3).28 Then, how to address that shortcoming in a deterministic account of human action? I submit that Kant’s conception of the empirical character can make a contribution to this issue,29 since he attributes to the empirical character of the human agent a decisive role for bringing about actions. The aim of this section is to show that this role can be specified by interpreting the empirical character as a complex of distinctive dispositions to adopt certain maxims and act on them. In what follows, I shall first clarify what exactly Kant means by “the empirical character.” I then attempt to spell out the crucial role of the empirical character for explaining human action. Finally, I will deal with the problem of whether the empirical character is changeable.
4.1 The Concept of the Empirical Character
In the resolution to the Third Antinomy, Kant famously draws the distinction between the empirical character and the intelligible character of the human agent in order to reconcile the conflict between free will and determinism. His basic idea is that, insofar as human beings can be considered both as appearances and as things in themselves, they possess both an empirical and an intelligible character. In terms of the empirical character human action is causally determined, whereas in terms of the intelligible character one and the same action can be regarded as transcendentally free (KrV, A536/B564). For now, I focus only on the notion of the empirical character. I shall revisit the relationship between the two characters in Subsection 4.3.
When Kant initially introduces the concept of character in the resolution to the Third Antinomy, he uses it in the generic sense to refer to a distinctive feature of an effective cause.30 He writes: “[E]very effective cause must have a character, i.e., a law of its causality, without which it would not be a cause at all” (KrV, A539/B567). Here Kant describes character as a law that constitutes the essence of a cause, regardless of whether the cause is a physical or a mental event or something else. Only later does he apply it to the human agent that is endowed with reason and will. In light of the transcendental distinction between the phenomenal and the noumenal aspect of human beings, the specific application naturally yields the two significant concepts in Kant’s theory of human action: “the empirical character” and “the intelligible character.”
What, then, is the empirical character of the human agent? There are two readings in the literature. The first follows directly from Kant’s own equation of character with “a law” (KrV, A539/B567) of the causality of the cause. In this case, the empirical character of the human agent is nothing other than a set of empirical causal laws governing the events that happen in inner sense. As some commentators note, it is “a singular summation of a person’s various maxims” (McCarty, 2009, p. 148) or “the law of the agent’s psychological causality” (Onof, 2024, p. 143; see also Kuehn, 2001, p. 147; Frierson, 2005, p. 24). Obviously, this “causal laws” reading is in accord with Kant’s literal expression31 and captures the nomological feature of causality, namely lawlikeness.32 Yet it is unclear why Kant needs to invoke such a new term, given that he has ubiquitously employed the somewhat synonymous concepts like “law,” “rule,” “maxim.”
The second reading has been recently proposed by Jonas Jervell Indregard. He objects to the “causal laws” reading and argues that the empirical character is a “set of causal powers of mind” (Indregard, 2018, p. 666), because he thinks this allows us to account for the possibility of moral change. On this “causal powers” reading, human beings as phenomenal substances possess abilities or powers that give rise to effects in the actual world, and these powers as a person’s empirical character are gradually formed, reformed, and cultivated, which account, according to Indregard, can avoid a fatalistic understanding of the empirical character. However, it seems to me that the emphasis on the changeability of the empirical character hardly matches with Kant’s equation of character with a law, which suggests that whatever the empirical character may be, it must be something stable and lawlike.
Admittedly, Kant himself does not say much about the exact meaning of the empirical character in the resolution to the Third Antinomy or elsewhere in his writings. It is here just a matter of Kantian exegesis open to us. And in the philosophical study of character in general, it is indeed difficult to define what precisely a person’s character is. The two aforementioned readings do have their interpretative merits.33 Nevertheless, as an alternative and especially for the sake of a deterministic explanation of human action, I propose that the empirical character of human agent is best understood as a complex of distinctive dispositions to adopt certain maxims and act on them.34 Let me explain what I mean by this. A person, construed as an agent at the phenomenal level, has an empirical character if and only if he or she possesses an integrated summation of tendencies to respond in given situations and act in accordance with the maxims endorsed by himself or herself. Maxims here are the particular empirical laws that are subjectively valid in the psychological domain. Because every individual’s dispositions or tendencies are fundamentally different, these can be regarded as distinctive ways of seeing the world and thus reflect a person’s moral or nonmoral values. In this respect, we can say that it is the empirical character that constitutes an individual’s core identity or the empirical real self. To care about a person consists essentially in caring about his or her empirical character. To use some vague terms in our ordinary life, a person’s empirical character is roughly a set of attitudes toward things and habits of doing actions.
Kant’s concept of the empirical character, construed dispositionally, is a causal one, as the two existing readings likewise claim. For Kant attributes it to rational agency that is able to make a difference in the world or bring about actual behaviors through its reason and will. When Kant identifies character in general with “a law” of the causality of the cause, I take him to be referring to the manner in which a cause produces its effect. The term “law” (Gesetz) is not used here in the conventional sense of a normative or juridical principle, but rather in its more literal sense, derived from the participle II of the verb setzen (gesetzt), namely, something set or fixed. Thus, a character as a “law” of the causality of the cause is not something universal and necessary, as the concept of law in the conventional sense designates, but rather something that differentiates a cause from other causes. It has distinctiveness and individuality. Indeed, a person’s empirical character is a complex of individual features in terms of which he or she as a rational agency can be distinguished from others. In Allison’s words, the empirical character is nothing other than the “modus operandi” of the causal agent (Allison, 2020, p. 269). And in Watkins’s words, it is a set of “natures” that an individual human being possesses (Watkins, 2005, p. 302).
Clearly, this dispositional account of the empirical character differs from the two extant readings. On my reading, the empirical character is precisely not a set of causal laws that are intersubjectively valid but rather a complex of distinctive dispositions to adopt certain maxims and act on them. Neither is it identical with a set of causal powers; instead, it is the mode of the causal agent according to which causal powers exercise or give rise to actions. In the next subsection, I shall draw on my “causal dispositions” reading of the empirical character and attempt to spell out the role that it plays for a deterministic explanation of human action.
4.2 The Empirical Character as the “Supreme Ground of Explanation” of Human Action
How does the empirical character relate to a psychologically determined action? Kant provides some remarks in the resolution to the Third Antinomy. He maintains that “for a subject of the world of sense we would have first an empirical character, through which its actions, as appearances, would stand through and through in connection with other appearances in accordance with constant natural laws, from which, as their conditions, they could be derived; and thus, in combination with these other appearances, they would constitute members of a single series of the natural order” (KrV, A539/B567). This passage suggests that human actions, insofar as they are regarded as spatiotemporal appearances, must be seen as consequences of three factors, that is, past events as “other appearances,” psychological laws as “constant natural laws,” and a complex of distinctive dispositions as “an empirical character.” The cooperation of these three conditions renders the resultant human action necessary in “a single series of the natural order.” In other words, once a human action takes place, one has to appeal to these three factors in order to explain why this action happens. As noted at the beginning of this section, in the popular understanding of determinism, for example in van Inwagen’s version, only the first two factors are taken into account; the last one is missing.
Kant further points out that the empirical character is quite decisive to the deterministic feature of human action. For example, he claims that “all the actions of the human being in appearance are determined in accord with the order of nature by his empirical character and the other cooperating causes” (KrV, A549/B577). He even asserts that “in regard to this empirical character there is no freedom” (KrV, A550/B578). And he reiterates: “Even before it happens, every one of these actions is determined beforehand in the empirical character of the human being” (KrV, A553/B581). Obviously, apart from recognizing the contributions of the antecedent mental events that serve as “the other cooperating causes” and the psychological laws that constitute “the order of [the mental] nature,” Kant particularly emphasizes the role of the empirical character in order to completely account for the determinacy of human action. Given the fact that both past events and psychological laws count as “external” factors that influence an action, Kant even comes to argue that the empirical character, which can be regarded as an “internal” factor, is “the supreme ground of explanation” (KrV, A546/B574) of human action.35 To explain an action is to understand why it happens, and this in turn requires a discovery of various determining grounds of the agent’s will. But why is the empirical character the highest one?
Let us first briefly consider the roles of the past events and the psychological laws for explaining human action. The physical world can affect us and arouse mental events such as beliefs, inclinations, and desires, which form our mind and influence our actions. As a natural being nobody can escape from his or her personal history. In our daily life we usually say that knowing a person consists significantly in knowing his or her past. What is more, human actions are governed by laws of nature. Our bodily activities comply with the physical and the physiological laws, and our mental phenomena necessarily stand under the psychological laws, that is, the subjective rules or maxims that hold for human agents in a comparatively universal way. To explain why an action occurs, one has to appeal to the underlying laws. As Kant puts it, “where determination by laws of nature ceases, there all explanation ceases as well” (GMS, 4: 459).36
However, roughly speaking, these two “external” factors merely concern the input information that an agent receives from the external world and his or her personal experiences. They might impact on every agent in the same way, but different agents could have different reactions to them, and it is the way of treating the input information that is decisive for human action. In other words, what brings about a human action is in fact the agent itself or, more precisely, the agent’s modus that processes the input information. It is exactly in this sense that we say that an agent itself is the cause of his or her action.
Now we turn to the examination of the role of the empirical character. Recall what we noted in the previous subsection. On my reading, a person’s empirical character is neither a set of causal laws nor a set of causal powers but rather a complex of distinctive dispositions to adopt certain maxims and act on them. But how does such an empirical character work together with the past events and the psychological laws so as to bring about actions? The answer lies in the essential nature of the empirical character. That a rational agent has distinctive dispositions to adopt certain maxims and act on them means that this agent is disposed in some fixed manners and possesses stable tendencies to respond in given circumstances and follow some maxims rather than others. In other words, every individual has his or her own consistent ways of seeing the world, and this renders him or her distinct from other individuals.
More specifically, when a person is affected by past mental events and is at the same time conscious of some subjective psychological laws or maxims, whether he or she performs an action hinges essentially on what dispositions or tendencies he or she possesses, since this person must first approve the mental events and then incorporate them into a maxim that he or she endorses, which is, admittedly, a highly complicated psychological procedure.37 In the case that past mental events and psychological laws are similar, different individuals will act differently because they have different empirical characters. Consequently, a person’s empirical character that consists of distinctive dispositions to adopt certain maxims and act on them is the key element for a rational agent to do actions in given situations, and this is the reason why Kant calls the empirical character “the supreme ground of explanation” of human action.
4.3 The Consistency and Changeability of the Empirical Character
If my “causal dispositions” reading is correct, it follows from this that a person’s empirical character is not an object we directly perceive or scientifically observe, neither is it something that continuously changes in time or can be modified at will. Instead, it is such a complex of dispositions that is relatively consistent and stable in the life of the human agent. Nevertheless, according to some commentators, the empirical character is of itself changeable, mutable, and unstable (Frierson, 2005, p. 26; Indregard, 2018, p. 663; Kohl, 2023, p. 67). As an interpretation of Kant’s conception of the empirical character, I do not think this view is tenable. In a central passage from the resolution to the Third Antinomy, Kant unequivocally states that the empirical character is consistent. He writes:
Now let us stop at this point and assume it is at least possible that reason actually does have causality in regard to appearances: then even though it is reason, it must nevertheless exhibit an empirical character, because every cause presupposes a rule according to which certain appearances follow as effects, and every rule requires a uniformity in its effects, grounding the concept of a cause (as a faculty), which, insofar as it must come to light from mere appearances, we could call the empirical character, which is constant, while its effects appear in alterable shapes, according to the differences among the conditions that accompany and in part limit it. (KrV, A548f/B576f, emphasis added)
Here Kant first links reason to causality in the empirical sense, and then claims that this reason exhibits an empirical character, since, as noted above, without a character a cause “would not be a cause at all” (KrV, A539/B567). In so doing, Kant highlights the lawfulness and uniformity of causal relation, and this finally leads him to assert that the empirical character, which is central to reason as an empirical cause, is constant or consistent. Kant also emphasizes this constancy or consistency by contrasting it with the alterability of the various effects that the empirical character generates under different conditions.
In fact, the consistent feature of the empirical character is already implicitly articulated when in the resolution to the Third Antinomy Kant first introduces the term “character” and equates it with a law of the causality of the cause (KrV, A539/B567, A40/B568). For the term “character” here is invoked in a causal sense. No matter how the “law” here is interpreted, either as a law in the usual sense or as a fixed way in which a substance exercises its powers on another substance, it refers to something stable because causation is not something singular or one-time but rather entails repeatability and regularity. What is more, Kant’s designation of the empirical character as “the mode of sense” (Sinnesart) (KrV, A551/B579) also implies that the empirical character is consistent, since a mode is not a changeable object of our senses but rather something abstract and structural behind the concrete sensible things. In addition, Kant regards an agent’s empirical character as “the empirical cause of all his actions” (KrV, A552/B580, emphasis added), which also indicates the consistency of the empirical character. For a variable character cannot be seen as responsible for all actions of an agent.38
Nevertheless, when it comes to the relationship between the empirical character and the intelligible character, Kant holds that the empirical character is in some sense changeable. His basic idea is that since the empirical character is ontologically grounded in the intelligible character, the change of the latter necessarily results in the change of the former. More specifically, according to Kant, the empirical character is nothing other than the manifestation or “sensible sign” (KrV, A546/B574) of the intelligible character; conversely, the intelligible character is “the transcendental cause” (KrV, A546/B574) of the empirical character and thus “determine[s]” (KrV, A551/B579) the latter. On the basis of this ontological grounding relationship between the empirical character and the intelligible character, Kant goes on to claim that “another intelligible character would have given another empirical one” (KrV, A556/B584). This suggests that a person’s empirical character would be different if and only if his or her intelligible character were changed, because the former depends essentially on the latter (KpV, 5: 98). It should be noted here that for Kant it is not permissible to infer the intelligible character from the empirical character, since the grounding relationship in question is one-way in light of Kant’s transcendental idealism.39 However, I think that this view on the special changeability of the empirical character does not undermine Kant’s insistence upon the consistency of the empirical character, because the former is based on a transcendental distinction and pertains to the noumenal domain, whereas the latter is an empirical claim. Accordingly, insofar as the human agent and actions are considered at the phenomenal level, Kant is entitled to hold that the empirical character is consistent.
Let us finally evaluate Kant’s overall account of the changeable or unchangeable nature of the empirical character.40 Kant’s view on the consistency of the empirical character might strike one as odd, since from the perspective of ordinary people a person’s character seems to be subject to changes from childhood to adulthood. My response is that Kant’s view in fact does not contradict our common understanding on this point. Here I want to make two remarks. First, Kant does not contend that the empirical character is absolutely unchangeable, which would avowedly run counter to the empirical nature of the character; rather, on my reading, he just claims that the empirical character is relatively consistent, which does not rule out the possibility that a person gradually reforms or cultivates his or her own empirical character, although this usually happens in an unconscious way. The crucial point here is that for Kant this empirically manifested changeability of the empirical character is due to or ontologically grounded in the change of the underlying intelligible character, which is unknowable to us because it happens in the noumenal world. As Kant himself makes clear in the analysis of the famous example of a malicious lie in the first Critique:
But why the intelligible character gives us exactly these appearances and this empirical character under the circumstances before us, to answer this surpasses every faculty of our reason, indeed it surpasses the authority of our reason even to ask it; it is as if one were to ask why the transcendental object of our outer sensible intuition gives precisely only the intuition of space and not some other one. (KrV, A557/B585)
Second, our common understanding that a person’s empirical character changes over time needs to be clarified. In my view, what we observe as changing is, precisely speaking, not the empirical character per se, but rather its empirical effects, that is, the particular actions that happen in space and time. Only by inductive methods such as observation, comparison, and abstraction can we infer a person’s empirical character from his or her various actions and the events impacting on them. For, as noted above, the empirical character is an integrated and unified whole or complex of dispositions, which cannot be experienced directly. In Kant’s words, it is “the mode of sense” (KrV, A551/B579) or “the way of thinking” (Anth., 7: 291). As a matter of fact, we do recognize the relatively consistent nature of the empirical character in our daily life. For instance, the proverbs say: “Old habits die hard,” “The child is father of the man.” In sum, as far as the notion of the consistency of the empirical character is limited in the relative or comparative sense, Kant’s view on this point is not as odd as one might think.
Having examined the different aspects of the empirical character, we are now in a position to describe the psychologically deterministic process of human action in a more comprehensive way. Every human action is jointly determined by the past mental events, the psychological laws, and the empirical character. But only by virtue of the empirical character can the past mental events and the psychological laws be unified and determine the will of the agent to give rise to actions. For the empirical character is a complex of distinctive dispositions to adopt certain psychological laws in the given situation of the past mental events. As Onof (2024, p. 95) rightly points out, “the empirical character is that through which actions are related to other appearances according to natural laws.” Now that for every action at a certain time the past mental events are on longer up to the agent, neither are the psychological laws under his or her control,41 and most importantly, his or her empirical character is relatively consistent and stable; the agent is at the time not free. In short, human action is psychologically determined.
5 The Problem of Predictability of Human Action
Determinism is usually regarded as closely related to the notion of prediction. The most representative of this view is presumably the eighteenth-century French theorist Pierre Simon de Laplace. He writes:
We ought to regard the present state of the universe as the effect of its antecedent state and as the cause of the state that is to follow. An intelligence knowing all the forces acting in nature at a given instant, as well as the momentary positions of all things in the universe, would be able to comprehend in one single formula the motions of the largest bodies as well as the lightest atoms in the world, provided that its intellect were sufficiently powerful to subject all data to analysis; to it nothing would be uncertain, the future as well as the past would be present to its eyes.42
Here Laplace claims that all states of things stand in a causal chain and are determined by some “antecedent state” (some past event) as well as “the forces acting in nature” (laws of nature). He then seems to equate this universal causal determinism with the thesis that a superpowerful intelligence or demon would know with certainty the states of all things in the universe at any time, if it were able to conduct data analysis based on its full knowledge of all the natural forces and the instantaneous positions of all things. This suggests that Laplace appears to simply identify the determinist thesis with the thesis of prediction.43 However, according to Earman (1986, pp. 7–8), while determinism might imply the thesis of predictability, they should be clearly distinguished, since the former is a metaphysical thesis, whereas the latter is an epistemological one. Whether we can foresee an event does not influence how the event itself is determined, namely the truth of determinism. I think this idea can apply to all theories about determinism, including Kant’s psychological determinism.
Because determinism and predictability are closely related issues, though not fully identified, I believe that a comprehensive account of Kant’s psychological determinism also requires a treatment of his notion of predictability.44 Then, my concern in this section will be with the question: To what extent is human action predictable for Kant?45
Like Laplace, Kant links prediction directly to determinism and is firmly committed to the predictability thesis concerning human action.46 In Kant’s writings, there are two relevant passages that are often cited in the discussion of determinism, one from the resolution to the Third Antinomy, the other from the second Critique:
Because this empirical character itself must be drawn from appearances as effect, and from the rule which experience provides, all the actions of the human being in appearance are determined in accord with the order of nature by his empirical character and the other cooperating causes; and if we could investigate all the appearances of his power of choice down to their basis, then there would be no human action that we could not predict with certainty, and recognize as necessary given its preceding conditions. Thus in regard to this empirical character there is no freedom, and according to this character we can consider the human being solely by observing, and, as happens in anthropology, by trying to investigate the moving causes of his actions physiologically. (KrV, A549–550/B577–578)
One can therefore grant that if it were possible for us to have such deep insight into a human being’s cast of mind, as shown by inner as well as outer actions, that we would know every incentive to action, even the smallest, as well as all the external occasions affecting them, we could calculate a human being’s conduct for the future with as much certainty as a lunar or solar eclipse. (KpV, 5: 99)
In the first passage, it is striking that Kant speaks of the determinacy and the prediction of human action in one context and specifically emphasizes the role of the empirical character for understanding the deterministic feature of human action. This suggests that the empirical character is likewise crucial to the predictability of human action. Although we have pointed out that determinism is to be distinguished from the thesis of predictability, it should be recognized that, as Earman (1986, p. 8) notes, the former does have some “epistemological implications.” In other words, if we believe the truth of determinism and are able to cognize all the conditions under which a future action happens, then we can foresee this action.47 In light of our analysis in the previous section, the prediction of a human action requires that we know the past mental events, the psychological laws, and the empirical character of the agent. And because of the centrality of the empirical character to the deterministic explanation of human action, which we discussed in Subsection 4.2, I argue that, for Kant, the predictability of human action consists essentially in cognizing the empirical character of human agents.
By cognizing the empirical character of an agent, I mean that we acquire knowledge of the agent’s distinctive dispositions to adopt certain maxims and act on them. In other words, we cognize how the agent reacts to the past events that affect him or her and adopts maxims under some given circumstances. To put it more simply, given that the past events and the psychological laws as external factors are generally recognized, the question of predicting a person’s future behaviors hinges significantly on how much we know about his or her empirical character. This is also the most crucial point for distinguishing between the prediction based on physical determinism and the one based on psychological determinism. For a future physical event, say a lunar eclipse, to be predicted, we just need to know the current position of the moon in relation to the sun and the earth – maybe plus some information about the history of the solar system – and the natural laws governing the motions of planets. That is to say, the past physical events and the laws of nature are sufficient for the prediction of a future physical event, and there is no need to know something like “the empirical character” of the object in question. But if we want to foresee how a person makes decisions and conducts actions, it is obviously not enough to merely appeal to what has occurred in the past and what psychological laws hold. Knowledge of the person’s empirical character is rather what we really need here, and it is only decisive.
But how can we cognize the empirical character of the human agent?48 Because the empirical character is ontologically grounded in the intelligible character, we would completely cognize it if we cognized the intelligible character, which is yet for Kant impossible since the intelligible character is an unknowable unconditioned in the noumenal world. Hence, the only way we cognize the empirical character is to conduct empirical study of both the physical and the psychological phenomena of the human agent. As Kant remarks, “[w]e notice it [the empirical character] through powers and faculties which it expresses in its effects” (KrV, A546/B574). I take it that by powers and faculties here Kant might mean the strength or weakness of human will, the firmness of doing something, or the endurance of the mind. In the first key passage cited above, Kant also points out that the empirical character “must be drawn from appearances as effect, and from the rule which experience provides” (KrV, A549/B577). Again, this suggests that the empirical character, which I have interpreted as a complex of distinctive dispositions, cannot be directly perceived but rather indirectly cognized through “appearances as effect”49 and the regularities we experience in an agent’s various actions. More specifically, in his famous example of a malicious lie in the resolution to the Third Antinomy, Kant states that “one goes into the sources of the person’s empirical character, seeking them in a bad upbringing, bad company, and also finding them in the wickedness of a natural temper insensitive to shame, partly in carelessness and thoughtlessness; in so doing one does not leave out of account the occasioning causes” (KrV, A554/B582). This shows that one can infer and cognize the empirical character from the empirical conditions that influence an agent’s life, that is, by observing, comparing, reflecting, and abstracting the empirical data. At this point, Kant does not say much but rather simply ascribes the investigation of the empirical character to anthropology, as he points out in the last sentence of the first passage cited above.50
Finally, it is worth noting that in both the aforementioned passages Kant uses the German subjunctive II to make counterfactual statements about the prediction of human action, which implies that the epistemic requirements for predictability Kant introduces are practically unrealizable. For instance, in effect, we are not able to investigate “all the appearances of [the] power of choice down to their basis,”51 and we are not actually capable of having a “deep insight into a human being’s cast of mind,” neither can we really “know every incentive to action,” so it is for us virtually impossible to predict human action “with certainty” or “calculate a human being’s conduct for the future with as much certainty as a lunar or solar eclipse.” The underlying reason for our inability to do accurate prediction is that human beings are finite rational beings whose cognitive capacities are limited so that their knowledge about appearances – including physical and psychological phenomena – and natural laws are always conditioned and incomplete.52 Our explanations of human actions and even physical motions are by no means exhaustive.53 Nowadays, with the development of technology, we can predict some physical events almost with a hundred percent certainty or at least high probability – for example a lunar or solar eclipse – but this kind of prediction or explanation remains incomplete. The uncertainty of weather forecasts is also due to the incompleteness of our meteorological knowledge. Therefore, we have to admit that the unknown domain of the material world is still immense, let alone the immaterial phenomena of our mind. In the latter case, the prediction is much more difficult because mental events, psychological laws, and the empirical character are not direct objects of our outer intuitions and thus lack absolute objectivity. At this point, we can say that the so-called Laplacean superintelligent is just an ideal.
6 Conclusion
In this paper, I have argued that some existing arguments against the interpretation that Kant embraces psychological determinism are not sound. In my view, Kant’s commitment to psychological determinism can be traced back to his argument for the universal causal principle in the Second Analogy, and this principle in fact serves as the ontological ground of Kant’s causal determinism, which in turn can be divided into psychological determinism and physical determinism. I have shown that the essential difference between these two forms of determinism lies in whether deliberation or reflexive awareness is involved in the occurrence of an event. Human action is performed by representing psychological laws, whereas physical events take place merely in accordance with laws of nature. At the heart of the paper, I have developed a “causal dispositions” reading of the empirical character of the human agent, according to which the empirical character is a complex of distinctive dispositions to adopt certain maxims and act on them. In so doing, I have maintained that human action is causally determined by past mental events, psychological laws, and the empirical character of the agent, and because the last factor plays a central role for an agent’s performing an action, it constitutes the supreme ground of explanation of human action. In addition, since determinism and predictability are usually seen as closely related issues, I have addressed the problem of predictability of human action within Kant’s philosophical framework, whereby I have suggested that Kant is also committed to the predictability thesis concerning human action, although he recognizes that due to the epistemic limitation of human cognitive faculties an accurate prediction is in principle unrealizable.
In the end, I want to point out that this paper is restricted to an investigation of the phenomenal dimension of human action, which implies that it does not rule out the assumption that Kant can still argue for human freedom at the noumenal level.54 As to how a deterministic account of human action can be compatible with a libertarian account, whether psychological determinism threatens the ground of moral responsibility, and whether Kant is really a compatibilist, answers to these questions are beyond the scope of this paper. I believe that Kant’s commitment to psychological determinism is not only harmless to his entire philosophical system but also significant for his effort to save human freedom.
Acknowledgments
I thank the two anonymous reviewers for the Journal of Contemporary Chinese Philosophy for their searching and insightful comments on the drafts. I also thank the participants in the Practical and Transcendental Freedom in Kant conference at Tsinghua University and the Kantian Ways of Thinking conference at University of Santiago de Compostela for their feedback. This research was funded by the General Projects of the National Social Science Fund of China, Grant Number: 22BZX098.
Note on Citations
In this paper, citations to Kant’s works give an abbreviation, volume number, and page references, except the Critique of Pure Reason, which is cited according to the pagination of the standard A/B edition: I. Kant (1900–), Kant’s gesammelte Schriften, De Gruyter. Translations are based on P. Guyer & A. Wood (Eds.) (1992–), Cambridge Edition of the Works of Immanuel Kant, Cambridge University Press. I use the following abbreviations:
Anth. Anthropologie in pragmatischer Hinsicht
RGV Die Religion innerhalb der Grenzen der bloßen Vernunft
GSK Gedanken von der wahren Schätzung der lebendigen Kräfte
GMS Grundlegung zur Metaphysik der Sitten
KrV Kritik der reinen Vernunft
KpV Kritik der praktischen Vernunft
KU Kritik der Urteilskraft
MS Metaphysik der Sitten
V-Met-L2/Pölitz Metaphysik L2
MAN Metaphysische Anfangsgründe der Naturwissenschaften
Prol. Prolegomena zu einer jeden künftigen Metaphysik
Refl. Reflexion
V-Met/Dohna Vorlesungen Wintersemester 1792/1793 Metaphysik Dohna
V-Anth./Mensch Vorlesungen Wintersemester 1781/1782 Menschenkunde
References
Allais, L. (2018). The Compatibility of Kantian Determinism with an Open Future. In V. L. Waibel, M. Ruffing, & D. Wagner (Eds.), Natur und Freiheit: Akten des XII. Internationalen Kant-Kongresses (pp. 713–728). De Gruyter.
Allais, L. (2023). Kantian Determinism and Contemporary Determinism. In D. Heide & E. Tiffany (Eds.), The Idea of Freedom: New Essays on the Kantian Theory of Freedom (pp. 24–38). Oxford University Press.
Allison, H. E. (1990). Kant’s Theory of Freedom. Cambridge University Press.
Allison, H. E. (2020). Kant’s Conception of Freedom: A Developmental and Critical Analysis. Cambridge University Press.
Beck, L. W. (1978). Essays on Kant and Hume. Yale University Press.
Bojanowski, J. (2006). Kants Theorie der Freiheit: Rekonstruktion und Rehabilitierung. De Gruyter.
Buchdahl, G. (1965). Causality, Causal Laws and Scientific Theory in the Philosophy of Kant. British Journal for the Philosophy of Science, 16(63), 187–208.
Cohen, A. (2009). Kant and the Human Sciences. Palgrave Macmillan.
Earman, J. (1986). A Primer on Determinism. D. Reidel.
Ertl, W. (1998). Kants Auflösung der „dritten Antinomie“: Zur Bedeutung des Schöpfungskonzepts für die Freiheitslehre. Karl Alber.
Fischer, J. M., Kane, R., Pereboom, D., & Vargas, M. (2007). Four Views on Free Will. Blackwell.
Friedman, M. (1992). Causal Laws and the Foundation of Natural Science. In P. Guyer (Ed.), The Cambridge Companion to Kant (pp. 161–199). Cambridge University Press.
Friedman, M. (2013). Kant’s Construction of Nature. Cambridge University Press.
Frierson, P. R. (2005). Kant’s Empirical Account of Human Action. Philosophers’ Imprint, 5(7), 1–34.
Frierson, P. R. (2006). Character and Evil in Kant’s Moral Anthropology. Journal of the History of Philosophy, 44(4), 623–634.
Frierson, P. R. (2014). Kant’s Empirical Psychology. Cambridge University Press.
Guyer, P., & Wood, A. (Eds.). (1992–). Cambridge Edition of the Works of Immanuel Kant. Cambridge University Press.
Hanna, R. (2006). Kant, Science, and Human Nature. Oxford University Press.
Hatfield, G. (1992). Empirical, Rational, and Transcendental Psychology: Psychology as Science and as Philosophy. In P. Guyer (Ed.), The Cambridge Companion to Kant (pp. 200–227). Cambridge University Press.
Hudson, H. (1994). Kant’s Compatibilism. Cornell University Press.
Hutton, J. (2021). Kant, Causation and Laws of Nature. Studies in History and Philosophy of Science, 86, 93–102.
Indregard, J. J. (2018). A Gradual Reformation: Empirical Character and Causal Powers in Kant. Canadian Journal of Philosophy, 48(5), 662–683.
Kane, R. (2005). A Contemporary Introduction to Free Will. Oxford University Press.
Kannisto, T. (2017). Kant on the Necessity of Causal Relations. Kant-Studien, 108(4), 495–516.
Kant, I. (1900–). Kant’s gesammelte Schriften. De Gruyter.
Keil, G. (2001). Wo hat Kant das Prinzip vom nomologischen Charakter der Kausalität begründet? In R. P. Horstmann, V. Gerhardt, & R. Schumacher (Eds.), Kant und die Berliner Aufklärung. Akten des IX. Internationalen Kant Kongresses (pp. 562–571). De Gruyter.
Keil, G. (2017). Willensfreiheit. De Gruyter.
Kohl, M. (2023). Kant on Freedom and Rational Agency. Oxford University Press.
Kraus, K. T. (2018). The Soul as the “Guiding Idea” of Psychology: Kant on Scientific Psychology, Systematicity, and the Idea of the Soul. Studies in History and Philosophy of Science Part, 71, 77–88.
Kraus, K. T. (2020). Kant on Self-Knowledge and Self-Formation: The Nature of Inner Experience. Cambridge University Press.
Kuehn, M. (2001). Kant: A Biography. Cambridge University Press.
Lovejoy, A. (1906). On Kant’s Reply to Hume. Archiv für Geschichte der Philosophie, 19, 380–407.
McCarty, R. (2009). Kant’s Theory of Action. Oxford University Press.
McKenna, M., & Pereboom, D. (2016). Free Will: A Contemporary Introduction. Routledge.
Meerbote, R. (1984). Kant on the Nondeterminate Character of Human Actions. In W. L. Harper & R. Meerbote (Eds.), Kant on Causality, Freedom, and Objectivity (pp. 138–163). University of Minnesota.
Nayak, A. C., & Sotnak, E. (1995). Kant on the Impossibility of the “Soft Sciences.” Philosophy and Phenomenological Research, 55(1), 133–151.
O’Neill, O. (1989). Constructions of Reason: Explorations of Kant’s Practical Philosophy. Cambridge University Press.
Onof, C. (2018). Kant’s Resolution of the Third Antinomy and Contemporary Determinism. In V. L. Waibel, M. Ruffing, & D. Wagner (Eds.), Natur und Freiheit. Akten des XII. Internationalen Kant-Kongresses (pp. 1107–1116). De Gruyter.
Onof, C. (2024). The Problem of Free Will and Naturalism: Paradoxes and Kantian Solutions. Bloomsbury Academic.
Pollok, K. (2013). Naturalism and Kant’s Resolution of the Third Antinomy. In M. Ruffing, S. Bacin, A. Ferrarin, & C. La Rocca (Eds.), Kant und die Philosophie in Weltbürgerlicher Absicht: Akten des XI. Kant-Kongresses (pp. 731–742). De Gruyter.
Scholten, M. (2022). Kant Is A Soft Determinist. European Journal of Philosophy, 30(1), 79–95.
Strawson, P. F. (1966). The Bounds of Sense. Methuen.
Strohmeyer, I. (2013). Kausalität und Freiheit: Eine Untersuchung des quantenmechanischen Indeterminismus im Lichte der kantischen Freiheitsantinomie. Kant-Studien, 104(1), 63–99.
Sturm, T. (2001). Kant on Empirical Psychology: How Not to Investigate the Human Mind. In E. Watkins (Ed.), Kant and the Sciences (pp. 163–184). Oxford University Press.
Timmermann, J. (2003). Sittengesetz und Freiheit. De Gruyter.
Timmermann, J. (2022). Kant’s Will at the Crossroads: An Essay on the Failings of Practical Rationality. Oxford University Press.
Van Inwagen, P. (1983). An Essay on Free Will. Clarendon Press.
Watkins, E. (2005). Kant and the Metaphysics of Causality. Cambridge University Press.
Watkins, E. (2019). Kant on Laws. Cambridge University Press.
Westphal, K. R. (2004). Kant’s Transcendental Proof of Realism. Cambridge University Press.
Westphal, K. R. (2012). Die positive Verteidigung Kants der Urteils- und Handlungsfreiheit, und zwar ohne transzendentalen Idealismus. In M. Brandhorst, A. Hahmann, & B. Ludwig (Eds.), Sind wir Bürger zweier Welten? Freiheit und moralische Verantwortung im transzendentalen Idealismus (pp. 279–310). Felix Meiner.
On how Kant’s works are referenced throughout, see the Note on Citations at the end of this article.
In the transcripts of his metaphysics lectures, Kant is reported to have said: “That a human being has freedom cannot be proven psychologically, but rather morally. … If I wanted to prove freedom psychologically, then I would have to consider a human being according to his nature, i.e., as natural being, and as such he is not free” (V-Met-L2/Pölitz, 28: 773). This statement makes it clear that Kant is committed to psychological determinism, although he himself does not use this term. For further textual evidence, see KrV, A550/B578, A554/B582; Prol., 4: 295; KpV, 5: 94, 99; V-Met/Dohna, 28: 682.
Those who reject Kant’s argument of the Second Analogy do not explicitly claim that Kant denies psychological determinism. But this can be seen as one of the upshots of their interpretation because, as we shall see, if the universal causal principle were unfounded, psychological determinism would be impossible.
I do not think that Kant’s argument for the causal principle in the Second Analogy is a failure, and neither do I agree with the libertarian reading of Kant’s position offered by some contemporary free will theorists. Yet these issues are beyond the scope of this paper. For convincing interpretations of Kant’s Second Analogy, see Friedman (1992, pp. 161–199) and Watkins (2005, pp. 203–216); for a detailed discussion of Kant’s rejection of the libertarian free will, see Scholten (2022, pp. 87–89).
For a criticism of Westphal’s interpretation, see Frierson (2014, pp. 11, 22–26).
The Davidsonian reading of Kant has been criticized by many scholars, e.g., Kohl (2023, p. 56).
I shall not address the issue of the applicability of the category of substance to the phenomenal self because psychological determinism primarily concerns the category of causality and the causal principle. For Kant’s assertion about the persistence of the phenomenal self, see KrV, B415.
For the purposes of this paper, I rest content with briefly pointing out this interpretive approach. For a detailed investigation about the role of the idea of the soul in empirical psychology, see Kraus (2018, p. 82; 2020, p. 238).
Note that, roughly speaking, in Kant’s architectonic system of the transcendental ideas, the systematic unity of the soul corresponds to the categorical judgment or the category of substance, whereas the systematic unity of the world corresponds to the hypothetical judgment or the category of causality (KrV, A334/B391). Here I link the category of causality to the idea of the soul because we are now concerned with the causal relation between the mental events that happen in one and the same subject.
For a practical account of the applicability of the causal principle of the Second Analogy to the representations of inner sense, see Onof (2024, p. 96).
Nowadays, both arithmetic and geometry are massively applied to psychological research, and in the contemporary classification of disciplines, psychology usually counts as a natural science. Nevertheless, one should not be anachronistic because psychology in Kant’s time is significantly different from psychology in the twenty-first century.
Bojanowski (2006, p. 171) also holds a similar view.
For further textual evidence, see Prol., 4: 295; KpV, 5: 33, 43.
I adopt this term from Kraus (2018, p. 85). This natural science is described as “improper” because it does not have the universality in the strict sense as the proper sciences like physics do.
For a lucid account of the relationship between necessity and law, see Watkins (2019, p. 14).
Admittedly, as Earman (1986, p. 5) points out, both the concept of causality and the concept of determinism are of themselves vague and obscure. Hence, the appeal to the former to clarify the latter remains question-begging. But I shall confine myself to Kant’s own framework without addressing the fundamental questions.
In recent literature, Matthé Scholten (2022, pp. 85–87) provides an account of Kant’s commitment to causal determinism, focusing on whether the argument of the Second Analogy is valid for scientific objects or objects of everyday experience.
However, when it comes to the problem of how Kant justifies the necessity of particular empirical causal laws, Buchdahl and Friedman hold different views.
For Kant’s own formulations, see KrV, A189/B232. Beck uses two shorthands to refer to the two versions of the causal principle: the weak causal principle “every-event some-cause” and the strong causal principle “same-cause same-effect” (Beck, 1978, p. 120). Whether these two formulations are substantively different and where Kant justifies them are still under dispute. For example, Keil (2001, p. 566) and Hutton (2021, p. 93) hold that they are in essence one and the same principle, and Kant has argued for it in the Second Analogy. By contrast, Watkins (2005, p. 216) and Kannisto (2017, p. 503) contend that they are in fact two distinct causal principles, the weak one of which is successfully proved in the Second Analogy, whereas the strong one requires arguments from other texts. In my view, no matter which interpretation one adopts, the grounding relationship between universal causal principle and causal determinism remains intact, and we are entitled to claim that all events are governed at the most fundamental level by the two principles, which are in turn instantiated in particular empirical causal laws or natural uniformities.
One might think that quantum mechanics, especially the principle of indeterminacy, threatens Kant’s argument in the Second Analogy and the resultant thesis of causal determinism. But, according to Beck (1978, p. 157), the principle of indeterminacy still presupposes the causal principle of the Second Analogy. For detailed discussion of this issue, see Strohmeyer (2013, pp. 72–78).
In the Second Analogy, Kant repeatedly emphasizes that the universal causal principle entails the necessitation of every event by some antecedent event. See KrV, B234, A193/B238, A199/B244, A201/B246.
Throughout this paper, I use the phrase “psychological determinism” to refer to psychological causal determinism concerning human action. It is different from other forms of determinism, e.g., divine determinism, historical determinism, fatalism, etc.
In one place Kant distinguishes between the mechanical and the psychological causality (KpV, 5: 96).
One might discover many characteristics of psychological determinism from Kant’s texts. For instance, mental events are just one-dimensional; inner sense plays an important role for explaining human action; human action has a teleological dimension, and so on. A detailed study of these characteristics is beyond the scope of this paper.
In this paper, I understand by human action a bodily movement that is intentionally and deliberately performed by a human agent. I am not concerned with the synthesis or combination of understanding, which Kant also calls an action in the Transcendental Analytic of the first Critique (KrV, B130). Neither do I consider unconscious actions, actions from instinct, and bodily forced actions.
For a lucid and illuminating discussion of the Incorporation Thesis, see Timmermann (2022, pp. 94–99).
For example, what exactly counts as events in the remote past? Why do the events before we were born matter for our present acts? What kind of laws of nature does it concern? Should we not distinguish between physical events and mental events as well as between physical laws and psychological laws? What does the phrase “not up to us” mean?
Note that contemporary philosophers coin these terms to challenge event-causal libertarianism. Here I simply use them in the literal sense.
Note that I am not trying to place Kant in the contemporary free will debate; rather, I confine myself to Kantian exegesis.
In Kant’s writings, the term “character” in a broad sense means a property or feature that differentiates a thing from other things (GSK, 1: 35; KrV, A2, A225/B273; KU, 5: 369). It can apply to material things, human beings, events, concepts, and so on. For instance, Kant speaks of “the persistence belonging necessarily to the character of a substance” (KrV, B418). But in Anthropology, Kant particularly uses this term to refer to the sensible, bodily feature or the rational, moral feature of human beings and calls it “mark of differentiation” (Unterscheidungszeichen) (Anth., 7: 285). In his moral philosophy, Kant is concerned specifically with the morally good or evil character of rational beings. For a detailed study of human beings’ moral character, see Frierson (2006, pp. 623–634).
In a passage from his Reflexion, Kant also associates character with a law: “Here one must first wait for a character, and then one has a law in order to explain appearances” (Refl., 18: 257).
Kant claims that “the concept of causality brings with it that of laws” (GMS, 4: 446).
According to Onof, the empirical character can be seen both as a law and as a power. He writes: “But my empirical character is, like all natural causality, a power that is an unchanging ground of temporal determinations, here, of mental states, e.g. of intentions” (Onof, 2024, p. 149). An anonymous reviewer also suggests that the two readings are just two sides of the same coin.
Allison (1990, p. 33) mentions this interpretation in passing, but he does not develop it.
By the term “external,” I do not mean something independent from us in the absolute or transcendental sense; rather, it simply means something the origin of which does not exclusively lie in us. According to Kant’s transcendental idealism, something external is still mind-dependent in the empirical sense. Thanks to an anonymous reviewer for pressing me to clarify this point.
Watkins emphasizes the necessity of laws by saying that “whatever grounds and causal laws have held in the past will not change in the future” (Watkins, 2005, p. 290).
For a detailed discussion of the empirical character in relation to cognition, feeling, desire, and action, see Frierson (2005, pp. 23–29).
In the Anthropology, Kant also underscores the consistency of a person’s character in general. He writes: “But simply to have a character signifies that property of the will by which the subject binds himself to definite practical principles that he has prescribed to himself irrevocably by his own reason. Although these principles may sometimes indeed be false and incorrect, nevertheless the formal element of the will in general, to act according to firm principles (not to fly off hither and yon, like a swarm of gnats), has something precious and admirable in it; for it is also something rare” (Anth., 7: 292). For other textual evidence, see KpV, 5: 97, 100; V-Anth./Mensch, 25: 1175.
As to how the intelligible character can be altered, Kant offers his detailed account in the Religion. An investigation of this issue goes beyond the scope of this paper. For a discussion of the changeability of the empirical character in relation to the intelligible character, see Ertl (1998, pp. 194–197).
Thanks to an anonymous reviewer for pressing me to address this issue.
Kant clearly articulates the determinacy of these two factors in the second Critique (KpV, 5: 94–95).
I cite this passage from Earman (1986, p. 7) and draw on his interpretation of Laplace’s determinism.
How to understand Laplace’s determinism is still under dispute in the literature. Here I rest content with mentioning his basic idea. For discussions on this issue, see Westphal (2012, pp. 268–270) and Keil (2017, pp. 18–21).
Determinism seems to be associated with both the explanation and the prediction of human action. Roughly speaking, the former is retrospective, whereas the latter is prospective. An investigation of Kant’s psychological determinism should pay due attention to these two aspects of human action. I addressed the problem of explanation in the previous section of the paper; now I am concerned with the problem of predictability.
Although the question of how Kant’s commitment to the predictability of human action is compatible with his conception of freedom is also important, I shall not address it because this paper is restricted to a study of the phenomenal dimension of human action. For a detailed discussion of defusing the threat of predictability to free will, see Kohl (2023, pp. 58–64).
Some commentators argue that Kant in fact does not embrace the predictability thesis with respect to mental events and human action. For example, according to Hudson, human actions are predictable only if this is based on physical events, and any prediction at the psychological level is impossible (Hudson, 1994, p. 69). See also Ertl (1998, p. 164), Westphal (2004, p. 241), Kohl (2023, pp. 55–56), and Onof (2024, p. 150).
Note that I am not claiming that all commentators who favor Kant’s psychological determinism would endorse Kant’s predictability thesis, nor vice versa. For example, Onof maintains that Kant implicitly makes a “distinction between determination (implied by determinism) and predetermination of action (which entails predictability in principle)” (Onof, 2024, p. 150). He then goes on to argue that Kant just endorses the former and rejects the latter.
Within Kant’s philosophical framework, it is indisputable that the phenomenal self, including its empirical character, is cognizable in principle because it belongs to the domain of possible experience, not the domain of things in themselves. See KrV, B156, B157, B277, A379, B430.
The context of the cited passage makes it clear that the term “effect” refers back to the term “appearances,” not the term “this empirical character,” although the singular form of the term “effect” might support the latter option. That is to say, what Kant here means is simply that we cognize a cause from its effect. For a helpful discussion of this issue, see Onof (2024, pp. 97–98).
For a detailed study of human action in Kant’s anthropology, see Cohen (2009, pp. 61–108).
It is controversial what exactly Kant means by the phrase “down to their basis.” Timmermann (2003, p. 132) interprets the “basis” as the intelligible character, whereas Bojanowski (2006, p. 176) understands it as the special empirical causal laws. In my view, the phrase simply means investigating all the appearances in question exhaustively, thoroughly, or completely.
In many passages of his writings (GMS 4: 407–408; MS 6: 441), Kant points out that the maxims upon which we act and our own dispositions are hidden even from ourselves, although they pertain to the knowable phenomenal domain. For example, he writes: “The real morality of actions (their merit and guilt), even that of our own conduct, therefore remains entirely hidden from us. Our imputations can be referred only to the empirical character. How much of it is to be ascribed to mere nature and innocent defects of temperament or to its happy constitution (merito fortunae) this no one can discover, and hence no one can judge it with complete justice” (KrV, A551/B579).
For a specific emphasis on the incompleteness of our naturalistic explanations, see O’Neill (1989, p. 68).
Note that this paper focuses on the deterministic feature of human action, but is not intended to develop an exclusively naturalistic reading of Kant’s theory of human action. For detailed objections to a purely naturalistic reading of Kant, see Pollok (2013, pp. 731–736).
