Interview by Richard Marshall.
'Kant starts the Analytic of the Sublime (in the 1790 Critique of Judgment) with the remark that “we call sublime that which is absolutely great”. '
'This brings Kant to the idea that true sublimity cannot lie in nature. For if we are right to call what is sublime absolutely great, and nothing in nature can actually be absolutely great, then the absolute greatness we are enjoying must lie in us: it is an appreciation of our rational nature. '
'In the German rationalist tradition, beauty is not neatly distinguished from sublimity. For them it was inconceivable that raw, massive, unorganised, destructive nature could have something appealing about it. Part of what is at work here is the influence of broadly Stoic ideas about the logos-infused cosmos: anything disordered, or seemingly disordered, will just be repugnant. As a result, sublimity figures in this tradition as an necessary aspect of beauty, where beauty is conceived (very roughly) as a kind of sensible presentation of rational organisation.'
'Mendelssohn argues that the feeling of the sublime cannot be astonishment, which he understands as a kind of thrilling shock. True sublimity, Mendelssohn argues, is felt as a kind of admiration: there is a positive attraction to something. This profoundly influences Kant’s ultimate view that our enjoyment of natural sublimity is rooted in feelings of admiration, which figures as a kind of analogue of respect (Achtung) for our rational nature and the possibility of its perfection (virtue) which can only be conceived through the moral law.'
'Kant understands that, we are attracted to a power or possibility of reason — most fundamentally, that it can conceive of what can never be presented to the senses. Ultimately, some of these conceptions can be a kind of knowing. This is the moral background to Kant’s theory of the sublime. Kant is a moral cognitivist: the good is the object of practical reason — we know it as the object of the will, as the good to be made actual in what we do. '
'Kantian moral psychology borrows quite a lot from the Stoics, including the basic conception of the good as the object of practical reason. In the second Critique (5:57ff.), Kant points out that German has ready resources to distinguish good from well-being (Güte from Wohl), and bad from ill-being (Böse from Übel). This is exactly the distinction that the Stoics drew, when they argued that the only truly good thing is virtue; anything else might have a kind of positive (or negative) value for planning things out so that things go well (or ill) for one. The Stoics, like Kant, think that human beings typically suffer from chronic misvaluing, where we confuse what is preferable in this latter way for what is genuinely good, and what is dispreferable in this latter way for what is genuinely bad. The only genuinely good thing is virtue, say the Stoics.'
'Mendelssohn celebrates a certain automatism in the execution of skill (not entirely unlike recent discussions of “flow”), which he took to be instrumental in meeting an objection that might be lodged against the perfectionist, agent-based ethics of his tradition: namely, that a virtuous person might seem to act for the sake of realising his own perfection in everything that he does, thereby taking a morally inappropriate interest in himself. Kant rejects Mendelssohn’s appeal to automatism, on grounds that it would make virtue mindless and unreflective.'
'From a Kantian point of view, the knowledge in question most fundamentally concerns the value of persons. Murdoch, like me, takes this knowledge to admit of degree: “The more the separateness and differentness of other people is realised, and the fact seen that another man has needs and wishes as demanding as one’s own, the harder it becomes to treat a person as a thing” (Sovereignty of Good, 64). Kant’s ontology most fundamentally divides into persons and things. His ethics is very much about how easy it is to, as it were, confuse the one for the other.'
Melissa Merritt works primarily on Kant and contemporary ethics. She's examined how the Kantian idea of virtue is rooted in an account of the essentially reflective nature of the rational mind. Her next project aims to get to grips with the Stoic influences on Kant, particularly in ethics and moral psychology -- and from this historical foundation to reassess the possibilities of moral rationalism in contemporary ethics. Here she discusses Kant's sublime, precursors, eighteenth century Anglophone work on the sublime and Moses Mendelssohn and German aesthetic rationalism, the sublime and his moral psychology and ethics, Stoic influences, the importance of reflection and the reflective life for Kant, why the idea that “every judgment requires a reflection” belongs to an area of logic that Kant calls “applied”, whether virtue is a skill for Kant, why virtue has to place love at its centre, the link between Iris Murdoch and Kant, the importance of Kant’s idea about the Enlightenment Ideal for the unity of apperception, and Kant’s views about practical reason.
3:AM:What made you become a philosopher?
Melissa Merritt:I would probably trace it to a philosophy class I took with Irad Kimhi as an undergraduate at Yale, in 1991. It wasn’t what I would call a typical philosophy class, in part because Irad is an especially creative philosopher, and in part because we did not read any self-styled works of philosophy. The course was called “Art and Its Subject”, and we mostly read art criticism at the juncture of high modernism and post-modernism: Clement Greenberg, Michael Fried, Rosalind Krauss. I was majoring in art at the time, and thought of myself as a painter.
It was particularly Michael Fried’s work, especially “Art and Objecthood”, that was crucial. I didn’t understand the essay for most of the term, but eventually I did. My final essay for the course was about Ad Reinhardt’s black paintings, and my experience of those paintings (there was a retrospective at MoMA running), and how they in many interesting ways both endorsed, yet in other ways effaced, what Fried valued in aesthetic experience. I remember feeling intellectually and emotionally exhausted after writing that essay in some way that registered as ground-shifting. It was a new kind of tired. I guess a satisfying kind, since I went on to major in philosophy — though it was a couple decades before I started to think of myself as a philosopher, and then mostly just as a convenient label. As in: philosopher, rather than chemist or lawyer. I feel a bit self-conscious calling myself a philosopher, really.
3:AM: You’re a Kant expert and one area of his work you’ve written about is his take on the sublime. His starting point is the idea of greatness – how does he understand greatness when linked to the sublime and how does this enable him to deny that a stormy sea or the Alps are not sublime and that sublimity lies in the mind rather than nature outside of the mind?
MM:Right: Kantstarts the Analytic of the Sublime (in the 1790 Critique of Judgment) with the remark that “we call sublime that which is absolutely great”. There are several things going on there. First, this appears to be some kind of report on common usage, and Kant wants to bring out its philosophical commitments. Second is the idea of absolute — non-comparative — greatness. And third is the issue you raise initially: what sort of greatness is this? Kant glosses the notion of greatness with the Latin magnitudo, which in its most basic literal meaning is physical greatness of size or quantity. But its main figurative meaning is a greatness of significance, of character, conduct, and the like. This richness of meaning suits Kant’s purposes: sublimity is not just a topic for aesthetics, by his lights — in some sense, its natural home is in ethics.
The terms for “sublime” in Greek, Latin, German (and English, via Latin) all derive from verbs meaning to raise or lift up: ὑψόω, sublīmo, and erheben. Thus what is sublime is something elevated. In ethics, the sublime figures through the ideal of virtue — a state of cultivated, perfectly good character. This is “elevated” because a person has had to raise herself up from some given, default state to get there. This is a more general point, not restricted to Kant’s ethics per se. The Roman Stoic, Seneca, is a someone who thinks of virtue as sublime in this fundamental way.
That said, in the Analytic of the Sublime, Kant is engaging directly with the aesthetic tradition on the sublime, much of which came to Germany through the British writers. And that tradition focuses on the sublimity of nature, and is not especially, or overtly, concerned with sublimity in the moral register. So it is fitting that Kant begins with magnitudo in its most literal sense, a greatness of size. But since, for Kant, nature is not merely a domain of inert extended things, but indeed a system of forces, Kant finds its appropriate to distinguish between the “mathematical” and the “dynamical” sublime when writing about our experience of natural sublimity. What is sublime in the first mode is what figures as absolutely great in size, and what is sublime in the second figures as absolutely great in power. These are ways in which physical nature can strike us.
Of course, nothing in nature can in fact be absolutely great. Nature is the phenomenal order, the order of appearances. It consists entirely of relations. Any phenomenal entity can only have a size greater or lesser than something else, and even the largest thing we can imagine — Kant mentions the Milky Way — has a magnitude that can be assessed as some finite iteration of units of measure. Same goes for considerations of power. Of course, any given experience of natural sublimity might work in both modes at once: the surf at Nazaré is both massive and crushing. And both of these experiences challenge, generally, our grip on ourselves as agents able to move about in normal, self-perspicuous ways. The mathematical sublime is disorienting, and this figures as an assault on embodied agency just as the threatening guise of the dynamical sublime does, though perhaps somewhat less obviously.
So while we enjoy natural sublimity — we are attracted, and linger over it — it is not straight, unadulterated liking. It necessarily has some element of painful aversion. The liking has to be the governing aspect of the experience; otherwise we would just run away. Kant memorably likens this mixed state to “a vibration… a rapidly alternating repulsion from and attraction to one and the same object” (Critique of Judgment 5:258). The unpleasant thing is recognising my physical smallness or weakness or powerlessness. The pleasurable aspect comes with the recognition that I am something more than this physical being, something over and above it. There’s something in me that cannot be destroyed by this sort of assault. This is where the idea of sublimity as an exaltation comes in, since Kant thinks that these experiences put us in touch with our rational nature, which is not phenomenal in essence. We are persons in a kingdom of ends (or “space of reasons”, to use Sellars’s language), and not — or not merely — things in the domain of nature.
This brings Kant to the idea that true sublimity cannot lie in nature. For if we are right to call what is sublime absolutely great, and nothing in nature can actually be absolutely great, then the absolute greatness we are enjoying must lie in us: it is an appreciation of our rational nature. There are a set of questions that arise here about whether this is egoistic or problematically self-regarding — the complaint goes back to J.G. Herder, who incidentally had been one of Kant’s students. I think this worry is fairly easily dealt with. Really, what we are revelling in is the capacity for morality, for autonomous self-determination. This is not some given, introspectible fact: rather, it is (in Kant’s view) the fundamental practical problem for the human being, the rational animal. (Kant’s account gives rise to more serious worries, stemming from his logocentrism about value — only rational nature has absolute or unconditional value — but I don’t think there are any easy answers to them.)
3:AM:He came quite late to this discussion, so how original is he when he denies that the sublime is not purely aesthetic but involves issues regarding the nature of human reason and its cultivation in light of our finitude and embodiment?
MM:I don’t think he is really that original in those claims, and others, in his theory of the sublime. Let me elaborate, and clarify my point. The sheer amount written on the sublime in the ninety or so years before the Critique is mind-boggling. It is as if every educated person with some literary pretentions felt the need to publish a little essay on it. There were a lot of ideas about the sublime “in the air”, and thus it is unsurprising that almost any aspect of Kant’s theory of the sublime, at least if considered in isolation, will seem to have some antecedent in an earlier view. And in some sense, it does. There is much in Kant’s theory of the sublime that comes from elsewhere. But these bits and pieces are transformed in his hands, because he was such a consummately systematic thinker, and because the conception of that system itself was, I think, deeply original.
That said, I was particularly struck when I came across a 1762 letter that the classicist Elizabeth Carter wrote to Elizabeth Montagu, wherein she describes a walk she took that gave her experience of “the true sublime”. Here’s a bit of it:
'I rambled till I got to the top of a hill, from whence I surveyed a vast extent of variegated country all round me, and the immense ocean beneath. I enjoyed this magnificent spectacle in all the freedom of absolute solitude. Not a house, or a human creature was within my view, nor a sound to be heard but the voice of the elements, the whistling winds, and rolling tide. I found myself deeply awed, and struck by this situation. The first impression it gave me was a sense of my own littleness, and I seemed shrinking to nothing in the midst of the stupendous objects by which I was surrounded. But I soon grew more important by the recollection that nothing which my eyes could survey, was of equal dignity with the human mind, at once the theatre and spectator of the wonders of Omnipotence. How vast are the capacities of the soul, and how little and contemptible its aims and pursuits?' (Letters from Mrs. Elizabeth Carter to Mrs. Montagu, London: F. C. and J. Rivington, 1817).
What she describes pretty nearly illustrates the core ideas of Kant’s theory, which didn’t appear until almost thirty years later. The main reason for this, I surmise, is that Carter was a translator of Epictetus: and she, like Kant, appears to have been influenced by Stoic moral psychology. I don’t know enough about Carter to substantiate that properly; but I do hold the view that Kant’s aesthetic theory of the sublime is ultimately an extension of his moral psychology, and that his moral psychology is deeply influenced by Stoicism.
3:AM: How important was both eighteenth century Anglophone work on the sublime and Moses Mendelssohn and German aesthetic rationalism for his position?
MM:So I think I’ve already given a sense of the significance of eighteenth-century Anglophone work on the sublime. But let me elaborate a bit on some of the details. One important figure is Edmund Burke, since he was the first to offer a principled distinction between beauty and sublimity. Kant does this as well, and while his account of this distinction is greatly different from Burke’s, Burke nevertheless does provide a kind of model for going about it. Many Anglophone writers also suggested, in a range of different ways, that what we really enjoy about stormy seas and jagged cliffs and so on is the way in which they arouse and exercise our minds. Why this exercise is enjoyable, and what exactly this exercise involves, and what exactly is being exercised, was not always terribly clear. But it is the antecedent of Kant’s claim that it is not the large or mighty natural entity that is sublime, strictly speaking, but rather some power of our own minds.
Yet the rationalist aesthetics developed in Germany was at least as important for the development of Kant’s view. In fact, you have to turn there to find anyone talking about sublimity as absolute greatness — this idea is nearly absent in the Anglophone tradition, though a glimmer of the idea crops up in Thomas Reid’s discussion. But in the German rationalist tradition, beauty is not neatly distinguished from sublimity. For them it was inconceivable that raw, massive, unorganised, destructive nature could have something appealing about it. Part of what is at work here is the influence of broadly Stoic ideas about the logos-infused cosmos: anything disordered, or seemingly disordered, will just be repugnant. As a result, sublimity figures in this tradition as an necessary aspect of beauty, where beauty is conceived (very roughly) as a kind of sensible presentation of rational organisation. Baumgarten and Meier, the canonical figures in this tradition, take sublimity (or “aesthetic magnitude”) to be an aspect of beauty, and beauty to be a perfection of sensible cognition; so really, on their view, there is no beauty without sublimity. And sublimity is only possible if this cognition represents “great, important, noble objects” in a manner appropriate to such dignified content. Baumgarten also distinguished between the relative greatness of anything we call sublime in nature — forests are paradigmatic for him — from the idea of absolute greatness that, Baumgarten hints, lies at the basis of our enjoyment of such natural sublimity. Baumgarten points to, and endorses, Seneca’s view that “Only virtue is sublime and exalted [sublimis et excelsa]”(De Ira I.21.4). Baumgarten’s account of how all this hangs together is somewhat muddled (at least to me), but the sorts of distinctions he draws certainly informed Kant’s later account.
Moses Mendelssohn, among the German rationalists, is particularly important for Kant. Mendelssohn was an extremely original, yet also extremely modest, thinker. He takes himself to be simply carrying on the German rationalist tradition, but really he is transforming it in highly interesting and creative ways. Kant also deeply respected Mendelssohn. Unsurprisingly, Mendelssohn sees nature as a logos-infused perfection in the broadly Stoic vein. His paradigmatic example of natural sublimity is the starry night sky, which he sees not as disordered spray of stars, but as the paradigmatic manifestation of divine perfection. Kant will draw on this in his famous “starry heavens” passage at the end of the Critique of Practical Reason — you know, the one that goes “Two things fill the mind with ever new and increasing admiration and reverence, the more often and more steadily one reflects on them: the starry heavens above me and the moral law within me” (5:161). I’ll come back to that example in a minute.
First let me note some of the particular lessons that Kant draws from Mendelssohn about the sublime. The main one concerns the emotional register of the sublime: that is, how it is felt. There was a French and British tradition of thinking that the proper feeling of the sublime must be some kind of astonishment, terror, or pity — Dubos, Dennis, and Burke provide examples. My favourite example of this is from John Dennis’s 1693 account of his hike in the Alps: “The sense of all this produc’d different motions in me, viz., a delightful Horrour, a terrible Joy, and at the same time, that I was infinitely pleas’d, I trembled” (Miscellanies in Verse and Prose, 134). An implication of this sort of view is that we enjoy the sublime because it provides a kind of thrill or relief from boredom. If we accept that, then we cannot make sense of our enduring interest in the sublime. Mendelssohn began with the principle that we have to be positively oriented towards something in our enjoyment of the sublime: the lingering, the being held, is a kind of attraction to something substantive — something with a positive, important content. For Mendelssohn, the stimulus to such an experience cannot be raw, disordered nature — jagged peaks, stormy seas and so on. For, on his analysis, we are attracted to rational perfection: its possibility, or even its apparent manifestation. That must be part of the account if we are to understand the experience as elevating our own powers of mind in a manner that stands to perfect them. This is crucial for Kant, who later emphasises that our appreciation of the sublime gathers strength the more that it is sustained. It does not simply overwhelm and then exhaust us, which we enjoy because it is thrilling.
This is why Mendelssohn argues that the feeling of the sublime cannot be astonishment, which he understands as a kind of thrilling shock. True sublimity, Mendelssohn argues, is felt as a kind of admiration: there is a positive attraction to something. This profoundly influences Kant’s ultimate view that our enjoyment of natural sublimity is rooted in feelings of admiration, which figures as a kind of analogue of respect (Achtung) for our rational nature and the possibility of its perfection (virtue) which can only be conceived through the moral law. A really nice example of Mendelssohn’s influence on Kant can be found in the famous “starry heavens” passage, which invokes the sustained attraction to the sublime, and that this is felt as admiration (“ever new and increasing admiration and reverence”), as well as the stimulus in the contemplation of rational articulation (“an unbounded magnitude with worlds upon worlds, and systems of systems”). He’s closer to Mendelssohn in that famous passage than he is in the Analytic of the Sublime two years later, where he consistently speaks of the natural stimulus to our enjoyment of the sublime as raw and disordered.
3:AM:You link his view of the sublime to his moral psychology and his ethics don’t you? Can you sketch for us how these connections are understood by you?
MM:Kant’s account of human moral psychology is rooted in the ancient definition of the human being as rational animal. As Kant sees it, we are animal rationabile by given dispensation, and we are called to make ourselves animal rationale (see his Anthropology 7:321). Of course this is actually quite difficult, and is for all we know never fully realised: the standard of perfection here is virtue, which is an ideal.
To be a rational animal is to face a practical problem of how to live and what to make of oneself. Kant particularly stresses that this problem is never “solved”: it is not as if we can get to a point where those questions never press — or no longer need to press — on us. It is also as if we can get to a point where it is no longer difficult to live well. Kant stresses this. The enjoyment of the sublime makes this difficulty thematic.
Inasmuch as we linger over the sublime in nature, the state of mind must be governed by the element of attraction. As Kant understands that, we are attracted to a power or possibility of reason — most fundamentally, that it can conceive of what can never be presented to the senses. Ultimately, some of these conceptions can be a kind of knowing. This is the moral background to Kant’s theory of the sublime. Kant is a moral cognitivist: the good is the object of practical reason — we know it as the object of the will, as the good to be made actual in what we do. When we live a life that is governed or shaped by such knowledge, we live well and are good. Of course, we can never know if we actually are good, because this is not a given, theoretical fact of some kind. Yet by Kant’s lights the standard of the good, the moral law, is graspable from where we stand, from the deeply imperfect or “fallen” human condition. So we have this as a point of orientation. At the same time, according to Kant, our rational embodiment renders us irremediably liable to view as “good” whatever promises pleasure or personal happiness. We are liable to take this as objectively good, when it is something more like subjectively preferable (but not genuinely good). So this attraction and repulsion is a basic feature of moral life, no matter what. Something in us is attracted to the ideal of virtue that is thought through the moral law, and wants to realise it in what we do and how we live; and something in us is always liable to find meeting its demands rather repelling or even abhorrent. The experience of natural sublimity puts us in the frame of mind to feel a kind of reverence for our rational nature while at the same time fully registering the reality of our human imperfection.
Most commentators think that only part of Kant’s account in the Analytic of the Sublime has this kind of moral significance. The standard reading is to treat the distinction between “mathematical” and “dynamical” sublimity to be quite sharp and deep — taking only the latter to be concerned with reason in its practical power. But I think that if you look closely at the Analytic of the Sublime, you will see that he takes both the mathematical and the dynamical sublime to be underpinned by these sorts of broadly practical considerations.
3:AM:You see Stoic influences in Kant’s approach to both his moral psychology and therefore his account of the sublime. Seneca is a key figure here isn’t he? Can you explain the link?
MM:Well, what I have just said points to an important difference between the two. In fact this is arguably the one criticism Kant makes of Stoic moral psychology. The Stoic ideal of the sage is supposed to have transcended human nature, and become truly godlike: the sage is no longer subject to the sorts of chronic misvaluing that plague ordinary human beings. The liability to misvalue has been extirpated. Kant thinks this is morally dangerous, and emphasises the impossibility of transcending human nature. Our own rational nature presents a problem for us, and one that we can never resolve once and for all. Living well is difficult. Failure to appreciate this opens the door to grotesque forms of moral conceit.
But Kantian moral psychology borrows quite a lot from the Stoics, including the basic conception of the good as the object of practical reason. In the second Critique (5:57ff.), Kant points out that German has ready resources to distinguish good from well-being (Güte from Wohl), and bad from ill-being (Böse from Übel). This is exactly the distinction that the Stoics drew, when they argued that the only truly good thing is virtue; anything else might have a kind of positive (or negative) value for planning things out so that things go well (or ill) for one. The Stoics, like Kant, think that human beings typically suffer from chronic misvaluing, where we confuse what is preferable in this latter way for what is genuinely good, and what is dispreferable in this latter way for what is genuinely bad. The only genuinely good thing is virtue, say the Stoics.
Seneca provides Kant with an example of handling the relation between natural and moral sublimity — indeed, we find Seneca in his letters and essays returning several times to the sublimity of the starry night sky, which was not only one of the perennial “stock” examples of natural sublimity (it crops up again and again in writings on the sublime), but also was particularly important for Kant. For Seneca, the night sky provides a consolation during his exile: he revels in what for him registers as the manifestation of perfect, divine rationality — so markedly different from the calculating, self-serving type of rationality always on display in the cutthroat political circles of Rome. Now, as I have noted, Kant is closest to this way of thinking in the famous “starry heavens” passage. Yet, more generally, for Kant the presentation of immense natural power challenges our default sense of what is worth going for, what is valuable. We enjoy this power “not insofar as it arouses fear, but rather because it calls forth our power (which is not part of nature) to regard these things about which we are concerned (worldly goods, health and life) as trivial [klein : literally “small”], and hence to regard nature’s might (to which we are, to be sure, subjected in regard to these things) as not the sort of dominion over ourselves and our authority to which we would have to bow if it came down to our highest principles and their affirmation or abandonment” (Critique of Judgment 5:262). He’s pointing here to the chronic misvaluing that Stoics took to be the human default, as well as the possibility of correct valuing that is expressed in some kind of respect for our rational nature.
My next project is to try to get to grips, in a more thoroughgoing and systematic way, with the Stoic influences on Kant’s ethics and moral psychology — so I hope to be able to answer your question better in a couple of years.
3:AM: I'll get back to you then! In the meantime - another area you’ve looked at is Kant’s arguments for the importance of reflection and the reflective life. One thing that immediately strikes anyone as implausible is the idea that anyone would argue for a position that says we should reflect on every thought we have. You’d agree with this and say that any claim like that has gone seriously awry – Kant’s smart and shouldn’t be read as making that sort of mistake. So what does Kant take reflection to be, why has his conception been so poorly understood and how does a distinction between constitutive and normative requirements help us to see how you interpret Kant’s reflective ideal?
MM:The first principle of all Kantian philosophy, I would say, can be found near the start of the Transcendental Deduction of the categories in the Critique of Pure Reason: “The I think must be able to accompany all of my representations.” It is a claim about the essentially reflective nature of the rational mind. It pointedly does not say that the I think must actually accompany all of my representations; it says rather that it must be able to. Even if you don’t stop and think about why you take things to be a certain way in a given instance, or whether you ought to take things as they immediately strike you as being (and so on) — what makes you a rational being, what makes your mind a rational mind, is that you have some kind of standing power to do this. Your existence is point-of-viewish. You must have some tacit grip on yourself as the source of a point of view in order to be genuinely thinking at all. All of this, I take it, is packed into that first principle, the apperception principle.
But if we have this power, what is it good for? The answer, very roughly and pretty obviously, is that there can be more for us than the default way things strike us. So there certainly is something like a reflection as a capacity to step back from the default way things strike us, so that we can consider whether this view is well-founded, sound — whether we should endorse it or reject it. This is true and important. It is also where we find a normative notion of reflection in Kant — i.e. the idea of a reflection that is required of you, something you ought to do. It is in this space that we find Kant saying things like: “every judgment requires a reflection” (e.g. in logic lectures, and the Critique’s Amphiboly chapter). This is a crucial claim, but it is hard to take it seriously without ending up with some abhorrently precious caricature of a reflective human being. One of my overarching goals in Kant on Reflection and Virtue is to show how we can take it seriously without that outcome.
3:AM:You argue that Kant is making a normative claim when he claims that all judgments require reflection but at the same time this shouldn’t be seen as a deliberately undertaken activity of some kind. Is that right? Can you sketch for us how this works?
MM:Yes, that’s basically right: it needn’t be taken as a deliberately undertaken activity.
The claim that “every judgment requires a reflection” belongs to an area of logic that Kant calls “applied”. Kant distinguishes “pure general” logic, which deals with the constitutive requirements on coherent and consistent thought, from “applied general” logic, which deals with norms for thinking given the human tendency not to make good use of our cognitive powers our susceptibility to prejudice; or liability to distraction; or to direct our attention inappropriately, and so on. Applied logic deals with our relation to epistemic standards given the whole mess of what we are. It is in this general space that Kant, I argue, develops the normative requirement to reflect in terms of the “three maxims of healthy human understanding”, which he elaborates in his later work (the Critique of Judgmentand the Anthropologyare important sources, but they appear throughout the record of his logic elsewhere, and partially and implicitly in some of his popular essays, like “What is Enlightenment?”). The three maxims provide the resources to understand the normative requirement to reflect as a matter of taking the appropriate interest in one’s own cognitive agency.
My idea is that taking this interest can govern cognitive activity in a global way, rather than being a kind of special action one performs on every (or some) occasion of judgment. So, to recap, the constitutive notion of reflection is the typically tacit grasp one has on oneself as the source of a point of view. The normative requirement to reflect can be understood as a matter of caring about that “source”: how is one minded? This allows us to see how reflection can be an expression of an abiding interest in one’s mindedness, without its being something that one is required to do on the occasion of each and every judgment.
3:AM:So what’s the best way to understand how the constitutive and the normative requirements to reflect are related to each other?
MM:The constitutive requirement to reflect is bound up with the essentially reflective nature of the rational mind: the default way things strike us is not the end of the story; there is always a possibility of other considerations entering in, and questioning whether things really are as they immediately strike us as being. Having said that much, I have already pointed to the normative requirement to reflect: we have to take some kind of practical interest in our own mindedness. In effect, the constitutive notion of reflection is key principle in the account of what we are as rational beings, while the normative notion of reflection is essential to the story about what we ought to make of ourselves as rational beings.
3:AM: How does this link with Kant’s views on reason and epistemic normativity? Does your reading end up with Kant being some sort of virtue epistemologist?
MM:In fact it was quite normal in Kant’s day to speak of epistemic character. I was trained in a way that did not especially emphasise the details of historical contextualisation. Almost all the philosophy staff at the University of Pittsburgh had serious interests in historical figures, which was a wonderful thing. I was trained in a department that did not marginalise the history of philosophy at all. Methodologically, the approach to the history of philosophy was — broadly speaking — somewhat in the vein of someone like P.F. Strawson. That is, there was serious, rigorous engagement with the arguments of the text, but not so much thought about who Kant (say) was reading, or corresponding with, and so on. As a result, I was unfortunately somewhat late in coming to take an interest in these sorts of facts; but it was as if the scales started to fall from my eyes once I did. There were several passages that I had long admired in Kant passages, broadly speaking, about epistemic normativity and the spontaneity of judgment — that I very belatedly discovered had clear sources in the logic textbook from which Kant lectured over several decades, G.F. Meier’s Auszug aus der Vernunftlehre. And finally paying proper attention to things like Kant’s friendship with Mendelssohn has had an impact on how I think about Kant. These are just some examples that have been important for me.
Do I read Kant as a virtue epistemologist? I don’t know. I suppose it depends on what you mean by “virtue epistemologist”. There is a very real sense in which it is anachronistic to call Kant an “epistemologist” at all: Kant endorses the Stoic division of philosophy into physics, ethics, and logic (see the Preface to the Groundwork for the Metaphysics of Morals) — and epistemological concerns crop up throughout all three divisions. So, it is important to read Kant in historical context. Doing so, moreover, shows that his concerns about epistemic normativity did not arise out of thin air — he wasn’t that creative — even if, as I think, he had very important moves to make, most of which stemmed from his particular way of thinking about the essentially reflective nature of the rational mind, as well as his particular understanding of the discursive and “synthetic” nature of human thought about objects.
3:AM: Is virtue a skill for Kant and does this link with Kant’s attitude towards practical reason and knowledge?
MM:In the Doctrine of Virtue in the 1797 Metaphysics of Morals, Kant takes up the question of whether virtue could be a certain sort of skill or Fertigkeit. This is an idea that Mendelssohn celebrated in his Philosophical Writings. Mendelssohn celebrates a certain automatism in the execution of skill (not entirely unlike recent discussions of “flow”), which he took to be instrumental in meeting an objection that might be lodged against the perfectionist, agent-based ethics of his tradition: namely, that a virtuous person might seem to act for the sake of realising his own perfection in everything that he does, thereby taking a morally inappropriate interest in himself. Kant rejects Mendelssohn’s appeal to automatism, on grounds that it would make virtue mindless and unreflective.
Anyone, I think, would have to recognise that there is necessarily something habitual about virtue. But habitual does not have to mean mindless and unreflective. The question is how this habit is understood. By Kant’s lights, the key idea is certain habits of thought and ultimately direction of attention that are essential to good character. By my lights, working through Kant’s qualified endorsement of the skill model of virtue — for he does not reject the model wholesale, but only certain ways of understanding it — allows us to work out a Kantian account of the embeddedness of virtue, and the idea that virtue engages (or indeed as Kant says “possesses”) the whole person.
This may not be an idea that Kant himself fully self-consciously embraced, but there are clear traces of his endorsement of the skill model of virtue in his elaboration of virtue in the Metaphysics of Morals. Following this thread allows us to articulate what is, I think, both a genuinely Kantian position in ethics and an independently attractive one.
3:AM:You’ve argued that it’s wrong to think that Kantian ethics disregards individuals on the grounds that we’re to value the universal quality of personhood rather than individuals. But you disagree and think that as in Iris Murdoch so too for Kant, virtue has to place love at its centre. Can you sketch for us your thinking here and why it’s not a position usually attributed to Kant. After all, it doesn’t sounds very Kantian!
MM:The key idea here goes back to the essential difficulty of moral life, which Murdoch brings out so well. I think there is a tendency among Kantians to think that, by Kant’s lights, it is easy to work out what morality requires of us, and only difficult to act on this knowledge because of those pesky sensuous inclinations and so on. I don’t think this is right. The whole damn thing is difficult. It is difficult in part because the knowledge in question is practical — genuinely “getting it”, knowing the good, is ipso facto being moved to make this good actual. It is difficult because the knowledge in question admits of degree — we can “get it” more and less, and part of failing to get it well is failing to be appropriately motivated. From a Kantian point of view, the knowledge in question most fundamentally concerns the value of persons. Murdoch, like me, takes this knowledge to admit of degree: “The more the separateness and differentness of other people is realised, and the fact seen that another man has needs and wishes as demanding as one’s own, the harder it becomes to treat a person as a thing” (Sovereignty of Good, 64). Kant’s ontology most fundamentally divides into persons and things. His ethics is very much about how easy it is to, as it were, confuse the one for the other.
Why has this sort of point been missed? I think there is some underlying resistance to the idea that knowledge could admit of degree — I don’t know why. Also, for generations, Kant’s ethics was thought to boil down to the “Categorical Imperative test” for maxims. The story was roughly that, in the course of deliberation, a reflective person was supposed to identify the maxim on which she proposed to act, and was meant to run the universalisation test on this maxim: can I will this maxim as a universal law? Maybe it was hard to work out what exactly a maxim is, or what indeed one’s own particular maxim for the proposed course of action is, but once that is in hand, perhaps it was thought that the assessment of the universalizability should be relatively straightforward — I don’t know: I find the whole picture rather absurd. But this sort of picture, I think, is what has led people to assume that, for Kant, morality is epistemically easy but executively difficult: i.e. not so hard to work out what are duties are, but quite hard to act and live accordingly. But if Murdoch is right — that default human psychology makes it actually quite hard to fully recognise the reality of other persons — and if the fundamental substantive principle of Kant’s ethics is to treat others as ends and never merely as means, then, once again, the whole business of living well should be quite difficult. Murdoch did another wonderful thing — she challenged what can be a vacuous celebration of “agency” in some Kantian circles. I don’t actually think Murdoch is anti-Kant at all; she is very careful to distinguish the bad latter-day Kantians from what is, by her lights, at the very least ambiguous in Kant’s own texts. But part of her point in challenging this celebration of “the free agent” was to say that our agency is perhaps fundamentally expressed in how we pay attention to one another.
OK, so I haven’t yet gotten to the issue of love. Now, it is widely thought that what we owe one another — from a Kantian point of view, and in terms of regard — is respect. Not love. We respect the bare reason in one another’s breasts (as Murdoch put it, roughly). Respect, it is thought, suits the requirement of universality: it is conceivable to respect everyone, but certainly not to love everyone. Love seems like a non-starter as a fundamental ethical concept, at least in an ethical system that celebrates impartiality and acting on universalisable principles. But then I read some very interesting papers by Kieran Setiya on love — agapic love — that helped me better understand how Murdoch takes love to be fundamental to ethics, and made me think about how Kant might, too. Basically the idea of agapic love is a kind of welcoming attitude, a positive gladness that another person is there. Respect gets you toleration, and not-offending. But it doesn’t get you a positive celebration of one another’s existence. Kant, I argue, thinks that genuine respect is only possible inasmuch as we are capable of love, in this sense. This is controversial, and I think it may require some kind of metaethical realism in Kantian ethics. But this is my view. And it is not crazy: Kant, after all, broadly deems our ethical duties to others duties of love. In some ways, the lingering issue is whether this deserves to be called “love” — but I think it clearly does, in the agapic tradition.
3:AM:How does Kant’s idea about the Enlightenment Ideal help us understand Kant’s argument for the principle of the synthetic unity of apperception as well as how he views issues that arise when dealing with practical judgment and agency?
MM:Well, this is a difficult question, and it depends upon what is meant by “the enlightenment ideal”. I have an earlier paper where I argue that Kant, perhaps developing a suggestion from Hume, conceives of the activity of thinking as a kind of synthesis (rather than a kind of inner perception or direct consciousness of mental contents). One of the things I was trying to do in that paper (“Kant’s Argument for the Apperception Principle”) was to work out how the really abstract arguments of the Transcendental Deduction relate to more ground-level considerations about epistemic normativity. But there I was mostly focused on the details of the Deduction, and I am not satisfied now with how I tried to follow the plumb line back down to Earth, as it were.
It also struck me that there was a need to understand where the principle of the synthetic unity of apperception comes from — is it self-evident, or is it established through some kind of argument? It seemed to me there was a lot of confusion about this. Roughly I came to the view that the apperception principle — i.e. “The I think must be able to accompany all of my representations” — is admitted as given or self-evident on a certain conception of reason. But the apperception principle is not yet the principle of the synthetic unity of apperception. How does that work? The first-person voice there is reason, the agent and subject of the critical self-examination. Now, the crucial question on which the Critique depends (as Kant put it in his letter to Christian Garve) is “how is synthetic a priori judging possible?”. This question comes from Hume, who challenged our epistemic entitlement to make claims that hold of necessity (i.e. are a priori) and yet pertain to matters of fact (i.e. are synthetic). The possibility of combining these two was ruled out by Hume’s way of thinking about the possibilities of judgment, where we are either working out necessary entailment relations among our concepts, or we are making contingent claims about what is the case. One or the other. But Kant thinks that there are ample examples of synthetic a priori judging — examples that seem to be viable, non-fraudulent, from mathematics and “pure natural science” — and he thinks that metaphysics, if it could exist, would have to consist of synthetic a priori judgments. So he asks how such judging would be possible.
In the Prolegomena he claims he wants to show how it is possible by “taking as given nothing except reason itself”. So his starting idea is the idea of an a priori judging reason. What would have to be true of such a power? Well, it would have to have a sensible faculty constituted a priori — a claim which he takes himself to substantiate in the Transcendental Aesthetic. Anyway, it follows at this point of the Deduction that this first-person voice of reason recognises that sensible representations are among its representations, and thus must be subject to the necessary conditions of being thought, which stem from the apperception principle. This is how Kant argues that there is an a priori synthesis by which experience is possible.
As I recall, in that paper I tried to link this to the enlightenment ideal of reflection by examining some of the ambiguities in Hume about reflection and the self-determination of the mind suggesting that Hume had something like a synthesis model of thinking, and that such a model is required to be appropriately critical about one’s view of the world. I argued that you needed something like Kant’s broader systematic commitments to follow through on these ideas, and so you needed the Deduction and the Critique and all that. I am not sure where I stand on that paper now, though.
3:AM: As a take home then, how would you summarise Kant’s views about practical reason and why is your account an improvement on those that have gone before?
MM:An important idea for me is that thinking is something that we do, and so it must in some sense by guided by practical reason. The details of that are complex, and the interpretation is undoubtedly contentious and sometimes outright controversial. But this is the governing thought that I have. I don’t treat our agency as knowers and our agency as doers as entirely separate issues.
I am also, by instinct, a dot-connector. That’s not to say that I don’t care about how the details work, but it is always at least as important for me to understand how a whole picture hangs together. This puts me somewhat at odds with normal practices in academic philosophy, which call for being really precise about the terrain of a given discussion, and knowing how to rule things out-of-bounds as “not my topic here”. I am not very good at that, but it is certainly a viable way to work. I just work differently. It is important to understand how things hang together. After all, if philosophy is rational reflection par excellence, then what we say in one domain should bear on and possibly constrain what we can say in another.
As a general point, my approach to Kant very much considers the whole person. Again, I am not at all uninterested in the “core” arguments of Kant’s critical philosophy: I have written entire papers about a few lines of the Transcendental Deduction. I just think that a successful interpretation of those arguments should carefully consider their implications at ground-level. What does it mean concretely, for human life? Without keeping this sort of point in mind, I fear that the alternative is to speak in terms of what power or faculty of the mind “does” what in the production of knowledge and the like. I don’t find that sort of thing satisfying. To me, it risks fabricating some kind of fantastic Kant-world — the philosophical significance of which, and relevance to the actual world, can be hard to make out.
3:AM:And finally for the readers here at 3:AM, are there five books other than your own that we should be reading to go further into your philosophical world?
MM:I’ll take as given the works of Kant, so I don’t cut into my allotment of five!
Iris Murdoch, The Sovereignty of Good(Routledge). This is one of my favourite books of all time.
It can be difficult to read, but it continually rewards close study, rereading, and returning to year in and year out. I taught it in one of my upper-level undergraduate courses, was somewhat (and happily) surprised by how well the students took to it. Murdoch didn’t write philosophy in a typical academic style; this may have appealed to my students but it is also undoubtedly a source of frustration for contemporary philosophers who do.
Christine Korsgaard, Sources of Normativity(Cambridge University Press). Korsgaardis one of those writers who never leaves you in doubt about why philosophy matters, and should matter, to anyone. Like any great philosophical work, this one is flawed. Perhaps it is because I have found it so compelling — and still do, on many points — that it was only with much time, and something like a sense of loss, that I realised that I deeply disagreed with its core constructivist claim about value. In some ways she is the contemporary manifestation of the latter-day Kantians that were Murdoch’s target in Sovereignty, so this entry on my list is a kind of counterpoint to the first.
Cicero, Cicero on the Emotions: Tusculan Disputations 3 and 4, translated and with commentary by Margaret Graver (University of Chicago Press). Cicero wrote this work, which largely defends the Stoic view of the emotions, during a period of intense and extended grief following the death of his daughter, Tullia, in childbirth. Though he was not a Stoic himself, he was sympathetic to Stoic ethics and moral psychology — and an extremely important source for our knowledge of Stoicism (given that most of the original Stoic texts have been lost). Graver’s translation and commentary are excellent.
Lucius Annaeus Seneca, Letters on Ethics, translated by Margaret Graver and A.A. Long (University of Chicago Press). Seneca writes to a young man, Lucilius, advising him to take up, and stay on, the Stoic path. Sometimes Seneca grates on my nerves, but there are gems throughout.
Michael Frede, A Free Will: Origins of the Notion in Ancient Thought(University of California Press). Frede argues that the notion of will emerges first in Stoicism, and the notion of free will in later Stoicism (especially Epictetus). I always learn so much reading any of Frede’s work, and this book is especially exciting.
ABOUT THE INTERVIEWER
Richard Marshall is biding his time.