Edinburgh Research Explorer Kant on science and normativity Citation for published version: Cohen, A 2018, 'Kant on science and normativity', Studies In History and Philosophy of Science Part A, vol. 71, pp. 6-12. https://doi.org/10.1016/j.shpsa.2018.03.002 Digital Object Identifier (DOI): 10.1016/j.shpsa.2018.03.002 Link: Link to publication record in Edinburgh Research Explorer Document Version: Peer reviewed version Published In: Studies In History and Philosophy of Science Part A General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer content complies with UK legislation. If you believe that the public display of this file breaches copyright please contact openaccess@ed.ac.uk providing details, and we will remove access to the work immediately and investigate your claim. Download date: 06. Apr. 2021 https://doi.org/10.1016/j.shpsa.2018.03.002 https://doi.org/10.1016/j.shpsa.2018.03.002 https://www.research.ed.ac.uk/portal/en/publications/kant-on-science-and-normativity(f9466333-4e41-4c97-b752-fe61004ecc4e).html 1 Kant on Science and Normativity1 The aim of this paper is to explore Kant’s account of normativity through the prism of the distinction between the natural and the human sciences. Although the pragmatic orientation of the human sciences is often defined in contrast with the theoretical orientation of the natural sciences, I argue that they are in fact regulated by one and the same norm, namely reason’s demand for autonomy. To support this claim, I begin by spelling out the pragmatic nature of the human sciences. Insofar as they are directed towards human cultivation, civilisation and moralisation, they are committed to investigating human phenomena for a practical purpose, namely the realisation of human beings’ aims. What is not sufficiently acknowledged, however, is that the human sciences also pertain to the enterprise of human cognition itself: they help human beings realise their cognitive vocation by promoting the conditions of good cognition. The second section focuses on these conditions and shows in what sense they constitute normative constraints upon belief. On the reading I propose, they take the form of epistemic principles that should guide our reflective attitude upon our cognitive endeavors. I then turn to the question of whether given their theoretical orientation, the norms that govern the natural sciences and cognition in general differ from those that govern the human sciences. For one may be tempted to think that even if cognition is normatively guided, its norms are epistemic whereas in the case of the human sciences, by contrast, insofar as they are pragmatically oriented, their norms are practical. Yet the third section shows that this is not the case. On the interpretation of Kant I defend, our actions and our thoughts are subject to the same rational norm, for rationality expresses itself normatively through the demand for autonomy in thought as well as in action. However, crucially for my account, the prime locus of responsibility is not over beliefs and actions themselves but rather over the principles that should regulate them. Once we turn our attention to the role of these principles in regulating our activity, we can make sense of the Kantian picture according to which the only source of normativity is our capacity for autonomy. 1 Insofar as the following works by Kant are cited frequently, I have identified them by these abbreviations: A: Anthropology from a Pragmatic Point of View; CJ: Critique of Judgement; CPR: Critique of Pure Reason; G: Groundwork for the Metaphysics of Morals; LA: Lectures on Anthropology; LL: Lectures on Logic; MM: Metaphysics of Morals; WOT: What is Orientation in Thinking? For the sake of clarity in the references to Kant’s writings, I have chosen to use titles rather than the author/date system. I have also included a citation to the English translation in parentheses, followed by a citation to the German text of the Prussian Academy edition (volume and page reference) in brackets. 2 1. The human sciences as enterprises with a pragmatic purpose Kant begins his Anthropology with an explicit reference to its ‘pragmatic point of view’: anthropology is ‘the investigation of what the human being as a free-acting being makes of himself, or can and should make of himself’.2 The ‘makes’ points to the descriptive part of Kant’s project – what human beings actually make, or have made, of themselves. The ‘can make’ refers to the realm of possibility, i.e., the scope and limits of human beings’ influence on themselves, whilst the ‘should make’ indicates the prescriptive part of Kant’s project, which encompasses the whole realm of human action – i.e., its technical, prudential and moral dimensions. On this basis, Kant’s anthropology essentially aims at accomplishing three tasks. First, it describes human beings’ behaviour relative to their purposes. Second, it deduces from their predispositions the scope of what they can make of themselves. Third, it draws conclusions regarding what they pragmatically should do, in order to accomplish the best possible fulfilment of their purposes, whether technical, prudential or moral. For the realm of the pragmatic encompasses all the dimensions of human actions: the development of skills, the means of achieving happiness, and the helps and hindrances to morality. The sum total of pragmatic anthropology, in respect to the vocation of the human being and the characteristic of his formation, is the following. The human being is destined by his reason to live in a society with human beings, and in it to cultivate himself, to civilize himself, and to moralize himself by means of the arts and sciences.3 To accomplish this task, Kant focuses on knowledge ‘of practical relevance’, that is to say knowledge that is useful to one’s conduct in life.4 This knowledge has an extremely broad scope: it discloses ‘the sources of all the [practical] sciences, the science of morality, of skill, 2 A 231 7:119. As is now well-known, Kant calls his anthropology ‘pragmatic’ rather than ‘practical’. But in the context of the introduction to the Anthropology, the meanings of these terms coincide insofar as they both have to do with the realm of action: ‘anthropology is concerned with subjective, practical rules’ (Lectures on Ethics 42 [27:244]). Contrast this with the narrow meaning of ‘practical’ as having to do with free action (G 95 4:448). As is regularly noted by commentators, Kant sometimes calls the prudential dimension of human action ‘pragmatic’ (e.g., Louden (2000): 69-70). For instance, he writes: ‘The first imperative could also be called technical (belonging to art), the second pragmatic (belonging to welfare), the third moral (belonging to free conduct as such, that is, to morals)’ (G 69 [4:416–17]; see also MM 565–6 [6:444–6]). However, far from entailing an inconsistency, this merely implies that the word ‘pragmatic’ can be understood in two distinct senses: in a narrow sense as ‘prudential’, having to do with welfare and happiness, and in a broad sense as ‘practical’, having to do with the field of action in general. My claim is that when Kant uses the term ‘pragmatic’ to describe his Anthropology, he is using the term in the latter rather than the former sense. 3 A 420 7:324. 4 A 233 [7:122]. 3 of human intercourse, of the way to educate and govern human beings, and thus of everything that pertains to the practical’.5 Yet the fact that the human sciences are practically oriented does not entail that they do not have a theoretical dimension. On my reading, the pragmatic intent of anthropology calls for a descriptive and explanatory dimension since human beings need to understand their nature in order to be able to determine what they are capable of and how they can achieve their purposes.6 This includes the investigation of nature’s purposes for the human species as well as an account of human beings’ psychological and biological make-up.7 As summed up in the Lectures on Anthropology, ‘Anthropology is thus a pragmatic knowledge of what results from our nature’.8 The knowledge of our natural constitution is necessary for us to use nature, and in particular our nature, to realise our purposes. As a result, far from being independent of each other, or even excluding each other, as is often presupposed, in anthropology the realm of the practical necessitates that of the theoretical.9 It is on the basis of theoretical observations about the human world that anthropology can play the crucial role of providing a map for human beings to orient themselves in it and realise their aims.10 The uniqueness of its approach lies in its commitment to investigating human phenomena for a practical purpose.11 Depending on our purpose when we adopt its recommendations, anthropology can be used either towards the realisation of morality, or towards the realisation of our own happiness.12 As a doctrine of prudence, it contributes to the latter insofar as not only does it help us choose ends that are consistent with the greatest possible happiness, it also teaches us 5 Correspondence 141 [10:145]. The notion of ‘knowledge’ is of course problematic here since Kant does not mean to suggest that the knowledge at stake in anthropology is of the same kind as the knowledge in natural science. However, it goes well beyond the remit of this paper to tackle this issue. Suffice to say that for Kant, anthropological knowledge is based on empirical generalisation, induction and interpretation. For discussions of this question, see Cohen (2009), Sturm (2009) and Wilson (2006). 6 I have defended this claim in Cohen (2009): 71-84. 7 Unfortunately, I cannot get into the details of the theoretical dimension of the human sciences for Kant. For a discussion of Kant’s biological account of the human species and nature’s purposes for it, see Cohen (2006). For a discussion of his psychological account of human beings, see Frierson (2014), Introduction. 8 LA 48 25:471. 9 For a version of the reverse claim that the theoretical standpoint necessitates the practical, see O’Neill (1989): ch. 3. 10 See Cohen (2009). 11 As Louden has noted, ‘Kantian social science…is not value-free but morally guided. We seek Weltkenntnis in order to further the goal of moralisation. Knowing the world stands under the moral imperative of making the world better’ (Louden (2000): 230). 12 ‘Prudence is the capacity to choose the best means to our happiness. Happiness consists in the satisfaction of all of our inclinations’ (LA 25:413). Reason clearly indicates our moral destination, namely the realisation of the moral law: ‘reason by itself and independently of all appearances commands what ought to happen’ (G 62 [4:408]). For a defense of the claim that Kant’s pragmatic anthropology encompasses both prudential and moral dimensions, especially by contrast with Brandt (2003: 92), see Cohen (2009): 70-71. 4 how to realise these ends.13 In its moral dimension, it examines the empirical helps and hindrances to moral agency – not any empirical helps and hindrances but specifically ‘the subjective conditions in human nature’.14 By identifying and recommending the means that help the realisation of our duty and counseling against the hindrances to it, it makes us more morally efficacious.15 It is in this sense that Kant’s anthropological project is a pragmatic project directed towards human cultivation, civilisation and moralisation. The practical orientation of the human sciences is often interpreted in contrast with the theoretical orientation of the natural sciences. In the Preface of his Anthropology, Kant himself distinguishes between the investigations of ‘a mere observer’, which he calls ‘theoretical speculation’, and the knowledge of ‘how to put them to use for his purposes’ – ‘anthropology with a pragmatic purpose’.16 Thus there seems to be a prima facie contrast between theoretical and pragmatic sciences, a contrast that can further be situated within the broader contrast between the practical and the theoretical standpoints spelt out in the Groundwork.17 However, although it is not sufficiently acknowledged, pragmatic anthropology cuts across these categories and itself encompasses recommendations that pertain to the enterprise of theoretical science and cognition in general – what Kant calls ‘applied logic’. What I Kant call applied logic … is thus a representation of the understanding and the rules of its necessary use in concreto, namely under the contingent conditions of the subject, which can hinder or promote this use, and which can all be given only empirically.18 On my reading, the aim of applied logic and Kant’s anthropology of cognition more generally is to promote good cognition.19 To help us realise our cognitive vocation, they rely on the 13 See Reflection 6:45n. For a very clear account of prudence and prudential ends in Kant’s anthropology, see Kain (2003). 14 MM 372 [6:217]. ‘It would deal with the development, spreading, and strengthening of moral principles (in education in schools and in popular instruction), and with other similar teachings and precepts based on experience. It cannot be dispensed with, but it must not precede a metaphysics of morals or be mixed with it’ (MM 372 [6:217]). 15 The nature and extent of moral anthropology is the subject of numerous debates in the literature. However, it falls beyond the remit of this paper to engage with them. For helpful discussions, see Frierson (2003), Louden (2000) and Cohen (2009). 16 A 231 7:119. 17 As Kant notes, ‘The concept of a world of understanding is thus only a standpoint that reason sees itself constrained to take outside appearances in order to think of itself as practical.’ (G 104 4:458) For discussions of the two standpoints, see for instance O’Neill (1989): 51–77 and Korsgaard (1996): 171ff. 18 CPR 195 A54-5/B78-9. 19 Kant notes in the Lectures on Logic that ‘Applied logic really ought not to be called logic’ (LL 533 9:18). In this sense, applied logic rightly understood is the pragmatic dimension of Kant’s account of cognition. The first part of Kant’s Anthropology, ‘On the Cognitive Faculty’, which deals with cognition, its talents, its weaknesses and diseases, is in effect an extension of what Kant calls applied logic in the Critique of Pure Reason and the Lectures on Logic. I would like to thank a referee of this journal for helping me clarify my account on this point. 5 knowledge of how we use our cognitive faculties when we actually form beliefs. It is on this basis that they can guide us on how to make the best use of our capacities – Kant repeatedly talks of ‘the use of understanding and reason’, the ‘correct use of the understanding’ or the ‘purposive use of [the faculty of cognition]’.20 Error can be avoided if we formulate and adhere to ‘the rules of the use of the understanding under the subjective empirical conditions that psychology teaches us.’.21 Crucially, note the parallel between the role of the ‘various subjective obstacles and restrictions’ and the ‘contingent conditions of the subject’ as they pertain to good cognition in applied logic and the ‘subjective conditions in human nature’ as they pertain to good action in moral anthropology.22 Just as these subjective conditions are the key to the success or failure of the realisation of our moral aims, they are key to the success or failure of the realisation of our cognitive aims. It is thus by identifying them and spelling out how to avoid them that Kant’s anthropology of cognition can contribute to the success of our cognitive endeavours. For once we understand how error occurs, we can prevent it. While it falls beyond the remit of this paper to examine Kant’s account of error, what is crucial for my present purpose is that there are rules for the correct use of our cognitive faculties, and that being epistemically responsible consists in abiding by these rules.23 The aim of the following section is to explore them in order to determine what normative constraints apply to cognition. 20 Based on A 386-8 [7:288-91], LA 386 [25:1261], 106 [25:545], 520 [25:1481], LL 533 9:18, 577 [9:74], CJ 175 [5:295]. For an overall account of Kant’s anthropology of cognition, see Cohen (2014). 21 CPR 194 A53/B77. While Kant mentions psychology in this passage, note that what he is talking about falls under the discipline of anthropology understood as the investigation of the way human beings think and act. By contrast, pure general logic is the ‘pure doctrine of reason’, which is ‘properly scientific, although brief and dry, as the scholastically correct presentation of a doctrine of the elements of the understanding requires’ (CPR 195 A53-54/B78). For a detailed account of the distinction between pure and applied logic, see Lu-Adler (forthcoming). For a discussion of Kant’s psychology in relation to his anthropology, see Wilson (2006) and Frierson (2014). 22 LL 533 9:18, CPR 195 A54-5/B78-9, MM 372 [6:217]. 23 CPR 194 A53/B77, 195 A54-5/B78-9. Kant’s central claim to this effect is that we can avoid error by withholding unwarranted judgments until we have reached objective certainty: ‘It is certainly really prudent, therefore, to know how to withhold one’s approval in most cases, until one has enough grounds for the thing’ (LL 126 24:160). When evidence is wanting or we lack a sufficient degree of certainty, we should suspend judgment until further evidence is available. There is of course a distinction between cases in which ‘approval does not arise immediately through the nature of the human understanding and of human reason’, and cases in which it does (LL 125 24:158). In the former cases, judgment is withheld and the will is called upon to orient the inquiry. But in the latter cases, not only is the will not called upon, judgment is immediate and ‘it is always very hard, if not utterly impossible, to withhold one's approval’ (LL 124 24:158). In this sense, while Kant does argue that judgment can be suspended, this claim should not be mistaken for the claim that it is free to do so at will. As he notes, ‘In suspensio judicii there lies some freedom’ (LL 471 24:736; second emphasis mine). Contrary to what the expression ‘freedom of judgment’ may suggest, judgment is not free as such; in and of itself, it has no power of choice over its operations. Rather, Kant’s claim is that it can be withheld if the epistemic environment necessary for certainty is absent. For a defense of this claim, see Cohen (2013). 6 2. Epistemic normativity From his early Lectures on Logic to his Anthropology from a Pragmatic Point of View, Kant identifies three ‘principles of thinking’ that spell out the cognitive attitudes conducive to truth: first, to think for oneself; second, to think in the place of another; and third, to always think consistently with oneself.24 These principles of sensus communis, as Kant calls them, should guide the epistemic agent in the pursuit of knowledge: ‘the issue here is not the faculty of cognition, but the way of thinking Denkungsart needed to make a purposive use of it’. As ‘[u]niversal rules and conditions for avoiding error’, they are the principles according to which we ought to think.25 They guide the reflective attitude we as responsible epistemic agents should take to our cognitive endeavours. Thus on my reading, just as the formulas of the Categorical Imperative should guide maxim-formation, the principles of thinking are second- order principles that should guide first-order maxims of belief-formation, which I will call “epistemic maxims”.26 In this sense, epistemic responsibility is a matter of whether and how we formulate our epistemic maxims, and the source of false or unjustified beliefs turns out to be the wrong ‘way of thinking’ about these maxims. To make sense of this claim, it can be helpful to further explore the epistemic case in light of the moral case. According to Kant, wrongdoing occurs when we order our principles the wrong way round. We subordinate the moral law to self-love by valuing the incentives of our inclinations over those of morality. Thereby we make the satisfaction of our own desires the ultimate value, an unconditional principle. What is at stake is thus our hierarchy of value.27 How are our principles ordered? Do we place morality before self-love or self-love before morality? When our principles are ordered the right way around, our commitment to morality expresses itself through the decision to only act ‘in such a way that we could also will that our maxim should become a universal law’.28 By contrast, wrongdoing occurs when our 24 LL 563–4 [9:57]. See also L A 520 [25:1480], CJ 174–5 [5:294–5] and A 333 [7:228]. I cannot discuss the content of these maxims here due to lack of space, but for helpful discussions see McBay Merritt (2011): §2, Wood (2002): 103 and O’Neill (1989): chs. 1–2. 25 CJ 175 5:295. 26 For a helpful discussion of the role of the Categorical Imperative in Kant’s account of maxim formation, see O’Neill (1989): ch. 5. I will return to the parallel between moral and epistemic maxims in section 3 but in the meantime, note that the epistemic maxims I have in mind are of the sort “I will not ignore evidence in cases when it falsifies a belief I desire to be true” or “the degree of certainty of my belief ought to be proportioned to the evidence I possess”. 27 For a detailed discussion of the different ways of ordering value in a Kantian context, see Bader (2015). 28 G 94 [4:402]. 7 principles are ordered the wrong way around and we prioritise subjective values over objective ones, thereby acting on principles that cannot be universalised.29 Kant expresses the wrong ordering of our principles in terms of moral egoism. The moral egoist limits all ends to himself, sees no use in anything except that which is useful to himself, and as a eudaemonist puts the supreme determining ground of his will simply in utility and his own happiness, not in the thought of duty.30 The moral egoist makes his own happiness the ultimate value and turns it into a principle that overrides all others. What I would like to suggest is that there is an epistemic equivalent of self-love, which Kant calls ‘logical egoism’: ‘The logical egoist considers it unnecessary also to test his judgment by the understanding of others; as if he had no need at all for this touchstone (criterium veritatis externum)’.31 Just as the moral egoist values the incentives of inclinations over those of morality, thereby subordinating the moral law to his own satisfaction, the logical egoist values his own judgment over and above everyone else’s, thereby subordinating the interest of truth to his subjective point of view.32 Of course, the logical egoist doesn’t think that what he believes isn’t true. Rather, he values his subjective point of view over that of others, and this happens at the expense of truth.33 By contrast with the logical egoist, the broad-minded thinker has what Kant calls an ‘extended mode of thought’: ‘he sets himself apart from the subjective private conditions of the judgment…and reflects on his own judgment from a universal standpoint’.34 To make sense of what this universal standpoint consists in, we need to turn to Kant’s account of assent. 29 ‘Hence the difference, whether the human being is good or evil, must not lie in the difference between the incentives that he incorporates into his maxim (not in the material of the maxim) but in their subordination (in the form of the maxim): which of the two he makes the condition of the other’ (Religion within the Boundaries of Mere Reason 83 6:36). While I cannot get into the details of Kant’s account here, note that the doctrine of radical evil is meant to explain how we might freely decide to subordinate the moral law to self-love: the radically evil person ‘makes the incentives of self-love and their inclinations the condition of compliance with the moral law’ (ibid). For a helpful discussion of this claim, see Muchnik (2009). 30 A 241 7:130. 31 A 240 7:128–29. Logical egoism can take one of two forms: either indifference to others’ judgment or arrogance about one’s own judgment. ‘Logical egoism is either indifference toward the judgments of others, in that I hold the judgments of others to be unnecessary for passing judgment on my own judgment, or it is conceit and arrogance, where one allots it to himself alone to make a correct judgment about a thing for all others’ (LL 323 24:874). 32 ‘Egoism can contain three kinds of presumption: the presumption of understanding, of taste, and of practical interest; that is, it can be logical, aesthetic, or practical’ (A 240 7:128). 33 I would like to thank Yoon Choi for pushing me on this point. 34 CJ 175 5:295. Thus, a universal community of knowers is already presupposed in the agent’s testing of his maxims. Knowledge is in principle not a solitary enterprise: we ‘desire to test our judgment on the understanding of other men and to investigate. Here one makes, as it were, an experiment and checks whether what we think is universal, whether others accept it, or whether it is not in agreement with reason’ (LL 141 24:178). 8 Assent for Kant is the ‘holding to be true’ (fürwahrhalten) of a proposition – a broader term encompassing what we now call “belief”.35 It has different epistemic modes, depending on whether its grounds are subjective or objective, sufficient or insufficient. While there is no space to get into the details of Kant’s account, it is sufficient to note for present purposes that only objective grounds provide reliable information about ‘the constitution of the object’ or state of affairs in question.36 They typically consist of perceptions, memories, introspections, as well as other beliefs we hold.37 By contrast with objective grounds, which are ‘independent of the nature and interest of the subject’, subjective grounds consist of psychological processes by which a person comes to hold a belief – ‘the merely private validity of the judgment’.38 The former are objective criteria, which contain the ground for why something is really true or false. The others, however, [are] subjective criteria[,] which contain certain circumstances, by means of which one is in a position to make a supposition about the truth or the falsehood in a thing.39 The nature of the grounds of a belief determines its epistemic mode: knowledge (Wissen) is both subjectively and objectively sufficient; opinion (Meinen) is subjectively as well as objectively insufficient; and faith (Glauben) is only subjectively sufficient and objectively insufficient.40 Thus for a belief to count as knowledge, it requires sufficient subjective as well as objective grounds. Otherwise it is not knowledge but mere opinion or faith. It is thus permissible to hold opinions, but only explicitly qua opinion, ‘with the consciousness that it is’ mere opinion.41 As long as we acknowledge the sufficiency of their grounds or lack thereof, all these modes of believing are epistemically legitimate in their own right. If we fail to 35 As Stevenson has noted, ‘it has recently been common for philosophers writing in English to use the word ‘believe’ (or ‘assent’) in this wide sense, meaning any sort of holding a proposition to be true, however confident or hesitant, rational or irrational, justified or unjustified. It would thus be tempting to translate Kant’s verb fürwahrhalten as ‘believe’. In that usage, knowledge implies belief; and ‘mere’ belief, without any sufficient justification, will then be the kind of belief which does not amount to knowledge’ (Stevenson (2011): 97). See also Chignell (2007b): 34: ‘In contemporary discussions, the fundamental attitude is assumed to be belief. For Kant (as for Locke, Leibniz, and some others in the early modern tradition), the attitude is Fürwahrhalten — ‘assent’ or, literally, ‘holding-for-true’. Assent for these writers is the genus of which most other positive propositional attitudes (opining, having faith in, knowing, and the like) are species. Kant doesn’t have an exact equivalent of our contemporary concept of belief, but if he did that concept would also fit under the genus of assent.’ 36 CPR 685 A821/B849. 37 LL 574 [9:70]. 38 CPR 685 A821/B849. See also LL 9:66, CPR A820/849. To formulate this distinction slightly differently, one could say that subjective grounds show why someone holds a proposition to be true whereas objective grounds show why a belief is non-accidentally true. Since there is no space to develop Kant’s account of the grounds of cognition here, see Chignell (2007a) for useful discussions of this issue. 39 LL 67 24:88. 40 CPR 686 [A822/B850]. 41 CPR 686 [A822/B850]. 9 acknowledge the grounds of our beliefs however, we are merely persuaded, where persuasion is ‘a holding-to-be-true on insufficient grounds, of which one does not know whether they are merely subjective or also objective’.42 If we mistake the subjective ground for an objective one, error occurs.43 On this basis, we can now make sense of Kant’s claim that we should think from a universal standpoint. The epistemic principle that we should reflect on our beliefs from the standpoint of others is intended to ensure that we have adopted them independently of our own subjective private condition. The touchstone of whether taking something to be true is conviction (Überzeugung) or mere persuasion (Überredung) is therefore, externally, the possibility of communicating it and finding it to be valid for the reason of every human being to assent to it; for in that case there is at least a presumption that the ground of the agreement of all judgments, regardless of the difference among subjects, rests on a common ground, namely the object, with which they therefore all agree and through which the truth of the judgment is proved.44 Recall that as I have just shown, the objective grounds of belief ‘are independent of the nature and interest of the subject’, whereas subjective grounds are not.45 This suggests that the ‘subjective private conditions of judgment’ are not shareable and that only objective grounds can be shared.46 As Kant notes, ‘a cognition is not correct when it agrees with my private understanding but when it agrees with the universal laws of the understanding of all men’.47 Accordingly, first, we should only be epistemically certain on the basis of grounds that can be adopted by all, at least in principle, since they are the only grounds that are universally valid. Second, insofar as the only grounds that are shareable are objective grounds, the only belief that is valid from a universal standpoint is the assent to propositions whose grounds are 42 LL 576 [9:73]. 43 ‘Error is the holding-to-be-true of falsehood. … No error is unavoidable in itself, because one simply need not judge about things of which one understands nothing. … With error…we are ourselves always culpable, in that we are not cautious enough in venturing a judgment, for which we do not have enough cognition’ (LL 288 24:832). 44 CPR 685 A820–21/B848–49. 45 LL 574 [9:70]. 46 CJ 175 5:295. 47 LL 148 9:187. In this case, I have conviction (Überzeugung): ‘If it is valid for everyone merely as long as he has reason, then its ground is objectively sufficient, and in that case taking something to be true is called conviction’ (CPR 685 A820/B848). Persuasion (Überredung), on the other hand, is subjectively sufficient but not objectively sufficient: it is ‘mere semblance (Schein), since the ground of the judgment, which lies solely in the subject, is held to be objective’ (CPR 685 A820/B848). 10 objectively sufficient.48 As a result, knowing responsibly consists in ensuring the universalisability of the grounds of our assent. How can we be epistemically responsible agents in this sense? It is by reflecting on the grounds of our judgment: are they what we think they are or are we mistaking subjective grounds for objective ones? Are they the correct grounds for the kind of judgment we are making or are we mistaking opinion for knowledge? On my account, this capacity to reflect on our grounds is precisely the locus on epistemic responsibility. Although it falls beyond the remit of this paper to defend this claim, I believe that the notion of common sense (Gemeinsinn) that Kant discusses in the Critique of Judgment is key here.49 By ‘sensus communis,’ however, must be understood the idea of a communal sense, i.e., a faculty for judging that in its reflection takes account (a priori) of everyone else’s way of representing in thought, in order as it were to hold its judgment up to human reason as a whole and thereby avoid the illusion which, from subjective private conditions that could easily be held to be objective, would have a detrimental influence on the judgment. (CJ 173-4 5:293-4) Exercising common sense consists in reflecting on the grounds of our judgments. In the aesthetic case (common sense aestheticus), which is most familiar to readers of Kant, common sense consists in reflecting on the grounds of our judgment of beauty to determine whether they are feelings of aesthetic pleasure.50 In the cognitive case (common sense logicus), it consists in reflecting on the grounds of our cognitive judgment to determine whether they are objective or subjective. Thus in both cases, common sense consists in determining the nature of the ground of my judgment and whether it is appropriate to the kind of judgment I make. It is in this sense that this process is normatively guided. Given the nature of the norms that apply to our beliefs, one may be tempted to think that insofar as they are oriented towards truth, they are strictly speaking epistemic and thus 48 The case of faith (Glaube) undoubtedly complicates this picture, although there is no space to discuss it here. As Kant notes, in this case, ‘the conviction is not logical but moral certainty, and, since it depends on subjective grounds (of moral disposition) I must not even say ‘It is morally certain that there is a God,’ etc., but rather ‘I am morally certain,’ etc.’ (CPR 689 A829/B857). For an interesting account of the epistemic grounds of faith, see Chignell (2007b): 354-7. What is clear, however, is that in the case of knowledge (Wissen), shareability is limited to its objective grounds. 49 ‘One could designate taste as sensus communis aestheticus, common human understanding as sensus communis logicus.’ (CJ 175 5:294-5) 50 ‘the common sense, of whose judgment I here offer my judgment of taste as an example and on account of which I ascribe exemplary validity to it, is a merely ideal norm, under the presupposition of which one could rightfully make a judgment that agrees with it and the satisfaction in an object that is expressed in it into a rule for everyone.’ (CJ 123 5:239) For instance, when common sense reflects on the grounds of our judgment of beauty, they could turn out to be agreeable feelings instead and we may be mistaken about that. I have discussed this in Cohen (2013b). 11 radically different from what we usually think of as practical norms, as discussed in section 1 in the case of pragmatic anthropology. If so, it would turn out that although both theoretical and practical enterprises are normatively bounded, their norms differ from each other both in ground and in content.51 However, the aim of the following section is to argue that the central role of the adoption of a universal standpoint in the foundation of the epistemic principles that regulate our beliefs points to the opposite claim. Namely, our theoretical and practical pursuits are ultimately regulated by the same rational norm, reason’s demand for autonomy. 3. The norm of rationality: autonomy To make sense of the claim that our theoretical and practical enterprises are regulated by the same norm, let’s go back once again to the moral case. Famously for Kant, maxims of action are only morally permissible if they pass a universalisability test. Its function is to rule out any maxim that cannot become a universal law: ‘I ought never to act except in such a way that I could also will that my maxim should become a universal law’.52 Testing the universalisability of a maxim establishes whether it is permissible by determining whether it can become a universal law without generating contradictions. Thereby, the universalisability test stipulates what is morally wrong, obligatory, and permissible.53 In the following passage, I believe that Kant suggests that epistemic maxims should also pass a universalisability test. To make use of one’s own reason means no more than to ask oneself, whenever one is supposed to assume something, whether one could find it feasible to make the ground or the rule on which one assumes it into a universal principle for the use of reason. This test is one that everyone can apply to himself.54 If we apply this model to the epistemic realm, the formula of universal law would be formulated as follows: ‘I ought never to believe except in such a way that I could also will that my maxim should become a universal law.’ I have defended this view elsewhere and unfortunately, there is no space to do so here.55 However, what is important for my argument 51 For instance, as Louden has noted, ‘moral norms and values may well be indigenous to the practice of science itself’ (Louden (2014): 212). 52 G 57 [4:402]. 53 There is controversy surrounding the interpretation of Kant’s universalisability test. See, for instance, Wood (1999): 40–2, O’Neill (1989): 83 ff. and Sullivan (1989): 47–53. However, these debates are irrelevant to my argument, at least as it is stated here. 54 W O T 1 8 [8:146f n ]. 55 See Cohen (2014b). 12 is that on my reading, believing responsibly consists in ensuring that the epistemic maxims that guide belief-acquisition are universalisable; that is to say, epistemic responsibility consists in ensuring that our epistemic maxims pass a universalisability test. The function of this test is to identify and rule out maxims that produce unjustified or false beliefs – maxims that Kant refers to as prejudices. Whilst prejudice is commonly thought of as an unjustified belief, for Kant a prejudice is an illegitimate principle the subject has adopted as his epistemic maxim: ‘Prejudice is a maxim of judging objectively from subjective grounds’.56 On Kant’s account, there are three different ways our epistemic maxims can be prejudiced.57 First, a maxim can ground beliefs on inclinations. Yet since desires have no relationship to truth, they should not be used as objective grounds. Second, a maxim can ground beliefs on habits. Yet the fact that things have been a certain way until now does not justify the belief that they will remain the same in the future. Third, a maxim can ground beliefs on imitation. Yet parroting another person’s beliefs fails to provide any insight into their justification.58 These prejudiced epistemic maxims seem rather different, but on my reading, what they have in common is that they are incapable of being universalized. It is on this basis that they are impermissible. Whether we are talking of inclination, habit or imitation, these are entirely dependent upon the nature and interest of the subject, which limits their validity to the private sphere. As a result, any epistemic maxim that turns them into objective grounds of belief should be ruled out. To sum up, I have argued that for Kant, belief is subject to the following normative constraints. First and negatively, epistemic agents should not form beliefs based on mere subjective grounds. Second and positively, the process of belief-formation should be guided by epistemic principles that are universalisable. I would like to end this section by suggesting that these constraints in fact express one and the same demand, namely the demand to believe autonomously.59 The first requirement, that we judge freely, independently of our private condition, is in effect a freedom-from, a negative freedom: the epistemic equivalent of practical freedom in the moral case.60 It is the capacity to ground our beliefs objectively, independently 56 LL 473 [24:737]. See also LL 315–16 [24:864–5]: ‘The principal sources of prejudices are subjective causes, accordingly, which are falsely held to be objective grounds. They serve, as it were, in place of principles, because prejudices must be principles.’ 57 ‘The principal sources of prejudices are above all imitation, custom, and inclination’ (LL 316 [25:865]). See also LL 579 [9:76]. For a discussion of prejudice, see Frierson (2014): ch. 6. 58 Note that a different kind of imitation can be legitimate in an educational context. See, for instance, A 329 [7:225]. 59 I would like to thank Eric Watkins for pressing me to address this point. 60 Kant defines practical freedom as the capacity to determine the will ‘independently of alien causes’ (G 94 4:446-7). 13 of our subjective condition. The second requirement is that we set our own epistemic principles. It is a freedom-to, a positive freedom: the epistemic equivalent of moral self-determination.61 Since the only legitimate epistemic principles are those that can be universalized, directing our cognitive powers according to maxims spelt out by reason is the only way of realising our epistemic autonomy. By contrast, if we fail to direct our cognitive practices on the basis of self- legislated maxims, our mind stops being its own guide. We let it be determined heteronomously through the adoption of prejudiced maxims that use subjective grounds as though they were objective. This is true of prejudice in the epistemic case (i.e., inclination, habit and imitation) as well as self-love in the moral case (i.e., pleasure, happiness and private satisfaction). For what the analogy between moral and logical egoism suggests is that the source of false belief is the same as the source of wrongdoing, namely the adoption of maxims that are not sharable, which Kant calls the wrong ‘way of thinking’ (Denkungsart). The opposite of egoism can only be pluralism, that is, the way of thinking in which one is not concerned with oneself as the whole world, but rather regards and conducts oneself as a mere citizen of the world.62 Acting and believing as citizens of the world consists in seeing ourselves as part of a community of agents who share a world and are equally committed to reason’s demand for autonomy and thus for universalisability.63 On this basis, autonomy is the principle that grounds epistemic normativity as well as moral normativity. Contrary to what is often assumed, it is not just the remit of morality. Our capacity for self-legislation also underlies our cognitive activity: ‘the power to judge autonomously — that is, freely (according to principles of thought in general) — is called reason’.64 Just as we act autonomously if we act according to moral principles we give ourselves, we believe autonomously if we believe according to epistemic principles we give ourselves. Freedom in thinking signifies the subjection of reason to no laws except those which it gives itself; …if reason will not subject itself to the laws it gives itself, it has to bow 61 Although I am unable to defend this claim here, note that on my reading, although there is a sense in which for Kant we legislate the laws of nature and logical laws (i.e., what Kant calls pure general logic), this legislation is not akin to either epistemic self-legislation as I have defined it here, or moral self-legislation. For the self-legislation I have in mind is limited to the domain of maxims. In this sense, I would argue that the laws of the understanding, for instance, are not normative in the sense that I believe our epistemic maxims are. For a defence of a similar claim with regards to the status of the laws of logic, see Tolley (2006). 62 A 241-2 7:130. 63 As Kant notes in the Groundwork, ‘the proposition, the will is in all its actions a law to itself, indicates only the principle, to act on no other maxim than that which can also have as object itself as a universal law’ (G 94 4:447). 64 The Conflict of the Faculties 255  7 : 2 7  . 14 under the yoke of laws given by another.65 Insofar as reason’s only command is that we act and think according to principles that can be shared by everyone, it commands all, in the same way, and in all cases: ‘Thinking according to a commonly ruling maxim…is only using your own reason as the supreme touchstone of truth’.66 Although the application of reason’s authority to a particular domain, whether we are deliberating about what to believe or what to do, gives rise to moral or epistemic norms, both epistemic and moral normativity are grounded on reason’s demand for autonomy. Conclusion This paper set out to show that for Kant, the human sciences and the natural sciences are regulated by the same rational norm, namely reason’s demand for autonomy. However, the argument defended here points to a much broader claim, namely that on the Kantian picture, all human enterprises, whether theoretical or practical, should be guided by the same norms, since rationality expresses itself normatively through the demand for autonomy. There is thus a fundamental analogy between our position as agents and our position as cognizers. Our actions and our beliefs function analogically in so far as they are subject to the same rational norm. Of course a lot more needs to be said to flesh out this claim. In particular, is the analogy between belief and action sufficiently sound to support the claim that we are responsible for our beliefs just as we are responsible for our actions?67 While it falls beyond the scope of this paper to address this issue, I would like to conclude by drawing attention to the fact that on my reading of Kant, his most valuable insight is that the prime locus of responsibility is not over beliefs and actions themselves but rather over the principles that should regulate them. It is in this respect that acquiring a belief is like acting: they both ought to be guided by principles that are under our direct voluntary control. Once we turn our attention to the role of these principles in regulating our activity, whether it is theoretical or practical, we can make sense of the Kantian picture according to which the only source of normativity is our capacity for 65 W O T 16 [8:145]. Kant’s famous enlightenment motto formulates the demand for autonomy in the most striking way: ‘Enlightenment is the human being's emergence from his self-incurred minority. Minority is inability to make use of one's own understanding without direction from another. This minority is self-incurred when its cause lies not in lack of understanding but in lack of resolution and courage to use it without direction from another. Sapere aude! Have courage to make use of your own understanding! is thus the motto of enlightenment’ (‘What is Enlightement?’ 17 8:35). 66 LA 521 [25:1481]. 67 I have begun to tackle this question in Cohen (2014b). 15 autonomy: we ought to act and think ‘only in accordance with that maxim through which you can at the same time will that it become a universal law’.68 Acknowledgments An earlier draft of this paper was presented at the American University of Beirut and a session on epistemic normativity for the Society for German Idealism and Romanticism at the Pacific APA in Seattle in April 2017. I would like to thank the audiences for very helpful questions and suggestions, and in particular Andrew Chignell, Courtney Fugate, Samantha Matherne, Lissa Merritt and Konstantin Pollok for their insightful comments. Finally, I am grateful to Yoon Choi and three anonymous referees for this journal whose careful and constructive criticisms have significantly helped to improve the paper. References Bader, R. (2015): ‘Kantian axiology and the dualism of practical reason’, Oxford Handbook of Value Theory, Oxford: Oxford University Press, pp. 175-201. Brandt R. (2003): ‘The Guiding Idea of Kant’s Anthropology and the Vocation of Human Being’, in B. Jacobs and P. Kain, eds., Essays on Kant’s Anthropology, Cambridge: Cambridge University Press, pp. 85–104. Chignell, A. (2007a): ‘Kant's Concepts of Justification’, Noûs 41 (1), pp. 33–63. ——— (2007b): ‘Belief in Kant’, in: Philosophical Review, 116(3), pp. 323-60. ——— (2010): ‘The Ethics of Belief’, in: Stanford Encyclopaedia of Philosophy. Cohen, A. (2014): ‘The Anthropology of Cognition and its Pragmatic Implications’, Kant’s Lectures on Anthropology: A Critical Guide. Cohen, A. (ed.). Cambridge University Press, p. 76-93. ______ (2014b): ‘Kant on the Ethics of Belief’, Proceedings of the Aristotelian Society, vol. CXIV, Part 3, pp. 317-333. 68 G 73 [4:421]. 16 ______ (2013): ‘Kant on Doxastic Voluntarism and its Implications for Epistemic Responsibility’, Kant Yearbook. 5, 1, p. 33-50. ______ (2013b): ‘Kant on the Possibility of Ugliness’, British Journal of Aesthetics, 53, 2, p. 199-209. ______ (2009): Kant and the Human Sciences: Biology, Anthropology and History. London: Palgrave Macmillan. ______ (2006): ‘Kant on Epigenesis, Monogenesis and Human Nature: The Biological Premises of Anthropology’, Studies in History and Philosophy of Science Part C: Studies in History and Philosophy of Biological and Biomedical Sciences. 37, 4, p. 675-693. Frierson, P. (2014): Kant’s Empirical Psychology. Cambridge: Cambridge University Press. ______ (2003): Freedom and Anthropology in Kant’s Moral Philosophy. Cambridge: Cambridge University Press. Kain, P. (2003): ‘Prudential Reason in Kant's Anthropology’, In Brian Jacobs & Patrick Kain (eds.), Essays on Kant's Anthropology, Cambridge: Cambridge University Press, pp. 230-265. Kant, I. (1992): Lectures on Logic, ed. by J. Michael Young, Cambridge, Cambridge University Press. ——— (1997): Lectures on Ethics, ed. by J. B. Schneewind and Peter Heath, Cambridge, Cambridge University Press. ——— (1999): Practical Philosophy, ed. by Mary J. Gregor and Allen W. Wood, Cambridge, Cambridge University Press. ——— (2001): Religion and Rational Theology, ed. by Allen W. Wood and George di Giovanni, Cambridge, Cambridge University Press. ______ (2007a): Anthropology, History and Education, ed. by Robert B. Louden and Günter Zöller, Cambridge, Cambridge University Press. ______ (2007b): Correspondence, ed. By Arnulf Zweig, Cambridge, Cambridge University Press. Grier, M. (2001): Kant’s Doctrine of Transcendental Illusion. Cambridge: Cambridge University. Guyer, P. (2000): Kant on Freedom, Law and Happiness. Cambridge: Cambridge University Press. Kleingeld, P. (1998): ‘Kant on the Unity of Theoretical and Practical Reason,’ Review of Metaphysics 52(2), pp. 311–339. Korsgaard, C. (1996): Creating the Kingdom of Ends (Cambridge: Cambridge University Press). 17 Louden, R. B. (2014): ‘Kantian anthropology: a science like no other’, Estudos Kantianos v. 2, n.2. ______ (2000): Kant’s Impure Ethics: From Rational Beings to Human Beings. Oxford: Oxford University Press. Lu-Adler, H. (forthcoming): ‘Kant on the Normativity of Logic’, European Journal of Philosophy. McBay Merritt, M. (2009): ‘Reflection, Enlightenment, and the Significance of Spontaneity in Kant’, in: British Journal for the History of Philosophy, 17(5), pp. 981-1010. Muchnik, P. (2009): Kant's Theory of Evil: An Essay on the Dangers of Self-love and the Aprioricity of History, Lanham, MD: Lexington Books. Mudd, S. (2013): ‘Rethinking the Priority of Practical Reason in Kant,’ European Journal of Philosophy, DOI: 10.1111/ejop.12055. Neiman, S. (1994): The Unity of Reason: Rereading Kant. New York: Oxford University Press. Nuzzo, A. 2005: Kant and the Unity of Reason. West Lafayette: Purdue University Press. O’Neill, O. (1989): Constructions of Reason. Cambridge: Cambridge University Press. Stevenson, L. (2004): ‘Freedom of Judgement in Descartes, Hume, Spinoza and Kant’, in: British Journal for the History of Philosophy, 12(2), pp. 223-46. Shah, N. (2002): ‘Clearing Space for Doxastic Voluntarism’. The Monist, 85(3), pp. 436-45. Stevenson, L. (2004): ‘Freedom of Judgement in Descartes, Hume, Spinoza and Kant’. British Journal for the History of Philosophy 12(2), pp. 223-46. Sturm, T. (2009): Kant Und Die Wissenschaften Vom Menschen. Berlin: Mentis. Sullivan, R. J. (1989): Kant’s Moral Theory. Cambridge: Cambridge University Press. Tolley, C. (2006): ‘Kant on the Nature of Logical Laws’, Philosophical Topics 34 (1/2), pp. 371-407. Williams, B. (1973): ‘Deciding to Believe’, in: Problems of the Self, ed. by B. Williams, Cambridge, Cambridge University Press. Wilson, H.L. (2006): Kant's Pragmatic Anthropology: Its Origin, Meaning, and Critical Significance. New-York: State University of New York Press. Wood, A. (2002): Unsettling Obligations. Essays on Reason, Reality and the Ethics of Belief. Stanford: CSLI.