Many, perhaps most, readers of Kant’s moral system start with his Groundwork to the Metaphysics of Morals, a work whose familiarity to us can mask its almost impossibly abstract argument. Barbara Herman’s innovative approach in her magnificent new book, The Moral Habitat, is to begin somewhere more messy, complex, and engaged in what it is to lead a life – something many critics of Kant believe to be neglected by or even excluded from a Kantian system. She starts by examining three under-studied imperfect duties – gratitude, giving, and a duty of care – and investigating what these tell us about moral agency and the nature of the moral system within which they need to be understood. She works from there to create an account of Kantian ethics as a system she calls the moral habitat: ‘a made environment, created by and for free and equal persons living together’ (p. ix). The aim of the book is to introduce and defend the idea of a Kantian moral habitat. To my mind, there can be no doubt that it succeeds in this, and in doing so gives us a really fresh, creative, exciting, deep, and compelling look at very well-trodden ground. Starting from an unfamiliar place casts the system in new light. We learn a lot.

Imperfect duties are duties which give us a required end rather than a specified action. The imperfect duties Herman discusses feature in moral theorising less often than perfect duties, but they are familiar to life and, I think helpfully, often occur in contexts that are ordinary and without high stakes. As she argues, the key to understanding imperfect duties is the nature of the license they provide (p. 10). She resists an account of imperfect duties as limits on the apparent demandingness of morality (the view on which we have options concerning which charity to write our cheque to and then can go on with pursuing our own projects). That picture, mistakenly, sees imperfect duties as requirements on action, but discretionary ones, so misses what they are, which is an orientation to a required end. Herman suggests, instead, seeing the exercise of discretion imperfect duties offers as something necessary for realising the duty’s point or object: an integral part of realising a value in a context in which ‘it matters to what we are doing that our action reflect our judgement about what to do’ (pp. 38–39). One of the important things we get from this is that it helps us to see ‘the deliberating and morally active person at the centre of a generative moral enterprise’ (p. ix) and to move past ‘the obfuscating atomism in much modern analysis’ (p. 11). This seems to me extremely helpful.

Herman argues that to understand the license imperfect duties involve we need, further, to understand them in relation to a value that they are for, and understanding the value that they are for leads to the values which they are situated in relation to, in a system. For example, we cannot understand gratitude or gift giving without understanding what is ours to give, so without understanding ownership and rights. Reflecting on ways in which giving too much can go wrong, we realise that paying back too much can involve (amongst other things), asserting greater authority over the terms of the debt than is rightfully yours to determine. The more common starting point of thinking about imperfect duties – from the point of view of an agent considering giving in response to an ocean of need, in which giving more always seems better, but also wanting to go back to living their own life – does not seem to have similar resources to explain what is wrong with repaying too much. And thinking about the values to which imperfect duties respond and the way the discretion they involve is part of realising these, moves us towards a system, because ‘while the different ways material goods move between persons – free benefiting, exchange, borrowing – call for different kinds of moral response, the set of responses must make sense together’ (p. 21).

Herman deftly shows how starting with the active, deliberating agent, oriented to certain ends, leads to a unified and dynamical account of juridical and ethical duties as an ongoing human project. This enables her to present a deep and compelling account of the relation between the ethical and the juridical (the moral and the political) parts of Kant’s account, starting with the idea that ‘[o]ur moral lives take place in the terms of some political order or other, whose concepts enter into our understanding of what we do, on whatever principle, from whatever motive’ (p. x). The system, she argues, requires starting with Kant’s technical notion of right, because ‘accurate moral deliberation depends on prior knowledge of what we are free to do: which efficient means are available for our purposes’ (p. 87), but also because a formal idea of innate right sets conditions for realising the moral status of persons (p. 2). On this account, the two halves of the Metaphysics of Morals are the first principles of a unified system of duties for beings like us (p. 4). This is a brilliant and to my mind ground-breaking way of understanding the unity of Kant’s system. I will mention two fascinating and powerful ideas she develops from it. One, Herman uses her investigation of the duties derived from two distinct principles, one about free action and one about necessary ends (p. 86), to introduce interesting imperfect juridical duties, ending, in the final chapter, with an argument for an imperfect duty to be an agent of change. Two, perhaps the most concrete and particular aspect of a habitat is a home, and Herman uses the moral habitat system to show why being homeless is not a morally possible condition (p. 187), and how we should think about the right to a home, where a home is both a requirement of agency in terms of such aspects of life as privacy and a personal life but also a requirement of a civic life and civic agency – an address, an abode, and a place where your mail-in ballot is delivered (p. 202).

Reading this challenging book involves the wonderful experience of constantly having new thoughts, and the book also suggested to me a lot of fruitful avenues of further investigation that the moral habitat idea opens up. I will briefly discuss three: happiness, gratitude, and injustice. Herman’s account of beneficence puts the role of happiness in Kant’s system in new light. As I have noted, she resists taking beneficence as the poster-child for imperfect duties and does not start with it, but she does, later in the book, present a novel and (to my mind) compelling account of relational beneficence, according to which ‘drowning strangers are not the paradigm case from which to build an account of the helping duty, and writing checks is not the paradigm helping action’ (p. 145). Herman presents an account of Kantian relational beneficence on which it is not primarily a duty of aid directed to other persons in need, because the ‘area of concern of duty is ends, not simply needs’ (p. 133); ‘we are not self-conceived distributors of stuff, but agents of ends that are not our own’ (p. 139). The fruitfulness of her approach again reveals itself in a rich discussion of why, within a Kantian account, we have a duty to make the happiness of others our end. It has never struck me before (though I feel now it obviously should have) how surprising this actually could seem, especially the fact that it is not presented by Kant as instrumental. Kant’s moral philosophy is often thought of as being concerned only with our rational agency, with our human needs coming into the picture only to the extent that they are in the service of our rational agency. But Kant does not say we are instructed to make the happiness of others our end insofar as this supports their moral agency, but simply that we must make their happiness our end.

Herman gives, I think, a brilliant account of how treating a person as an end in themself unavoidably involves making their happiness your end, but I wondered whether the way her conclusion is presented here slightly undersells the potential in the materials she uses to reach it. She argues (if I read her correctly) that because, for Kant, happiness is an end that ‘we can no more abandon […] than we can abandon our nature’ (p. 137), it is integral to our agency. To understand another as an agent is to understand their conception of their happiness, which ‘is not a haphazard collection of interests, but a more or less ordered set of subordinate ends: our loves and projects in an updating life plan’ (p. 137). She concludes – this is where I question the emphasis – that supporting happiness is essential to the emergence of effective moral agency (p. 137) and therefore that it matters morally as ‘a condition – necessary but not sufficient – for realizing the potential for moral personality in us’ (p. 138). My concern with the idea that happiness matters morally because it is a condition of moral agency, as well as the correlative idea that pain matters morally ‘because it is a mark of impedance with rational and autonomous functioning’ (p. 131), is that it seems a kind of ‘wrong reasons’ account. This is similar to the kind of worries people have about what Kant seems to say about the grounds of moral concern with the treatment of animals, where it seems that while it may be true that cruelty to animals harms our moral agency, this leaves out the central moral ground of concern based on the feelings of the animals themselves. Of course we do not think the mere badness of pain for the agent always counts against it (some necessary treatments are painful and medicating grief can be a mistake), but the idea that happiness and pain matter only as conditions of moral agency seems to leave out something important. One might think that suffering matters not just because it impedes rational and autonomous agency but because it is bad for creatures who matter in themselves. As I read it, Herman’s account has given us the materials to show this, within a Kantian system, since she shows so beautifully that taking another person’s happiness as one’s end simply is part of what it is to take another person as an end. Supporting another’s permissible ends is not only a matter of supporting a condition of her moral agency but of supporting her, and it is her that is the end in itself. I think looked at in this way, the account has more materials for expanding to include the interests of animals – why the ‘analogue’ of our agency we find in them is significant – than if happiness only matters as a condition of moral agency.

One of the imperfect duties Herman discusses is gratitude. She understands the duties of gratitude as a way in which the beneficiary reasserts their equality and independence in response to a possible threat to this created by the benefactor carrying out the beneficiary’s purposes. My initial concern with this was that it seems somewhat anxious and overly status-concerned to see another’s benefaction as an apparent threat to the beneficiary’s agency. When I think about the content the reactive attitude of gratitude – the content of actually feeling grateful to someone – it does not seem to have anything to do with reasserting my independence in the face of their role in carrying out my projects. It seems to be about seeing them, in the light of their willing towards me, in a certain (appreciative) way. Perhaps this difference is because I approach gratitude from the perspective of the content of the reactive attitude, whereas Herman is starting from the content of gratitude as a duty. But it is fruitful to consider how the content of the reactive attitude relates to the duty. Thinking about Herman’s account of the duty has, I think, deepened my view of the content of the reactive attitude, but I think that reflecting on the content of the attitude could slightly expand her account of what is achieved by the duty.

I had previously thought of the content of the reactive attitude of gratitude as simply involving a particular affective way of appreciatively seeing the person who has benefited you. What I think I had left out (and I have got from Herman) is the idea that the content of the duty is also part of the content of the attitude. People writing on reactive attitudes like resentment have noted their communicative, ‘call-and-response’ structure – the idea that their content implicitly has an addressee and is for a response, whether or not this is expressed (see Macnamara Reference Macnamara2013: 6, 896; Darwall Reference Darwall2006: 145; McGeer Reference McGeer, Coates and Tognazzini2013; Smith Reference Smith, Coates and Tognazzini2013: 44; Walker Reference Walker2006: 136). It is easier to see this with blaming reactive attitudes, but what Herman helped me see is that a feeling of open-ended willingness oriented to being ready for the kinds of responses she describes as part of the duty of gratitude may be part of the affective content of the attitude of gratitude – as opposed to just an appreciative way of seeing the benefactor. But my concern about her emphasis on status remains. When I reflect on the content of feelings of gratitude, at least some of the time the appreciative orientation to the person I see myself as standing ready to do something for just does not seem to contain anything about the desire to reassert the equality of my agency (though it might sometimes). Rather than feeling that my equality and agency is threated, I appreciatively feel our interconnection, and even my dependence, in relation to the way they have helped further my ends. So where Herman focuses on the reassertion of independence and equal agency in the face of someone lending their agency to yours (to your projects), I think that another side of this could be to focus on our equality-as-dependents, experienced as something made non-threatening through our mutual helping. In standing ready to help you when you are in need, I need not be asserting that I am not your dependent because we are both equally agents, but that we are both equally human dependents, in need of each other, and as you were there for me I will be there for you. I do not think this is in tension with Herman’s account or her approach, since the duty to further the ends of others recognises our common unavoidable interdependence, and since what I am suggesting also involves equality. But I think that it perhaps develops a different emphasis: that our equality as agents (the side she emphasises) is inseparable from our equality as dependents, and that this need not be a threat to our standing. However, as I said, I think this is something her account can incorporate (and perhaps is already intended to incorporate).

Thinking about the content of reactive attitudes relates to an interesting discussion Herman has about whether there are motives we can be required to have. Reactive attitudes make plausible the idea of required attitudes, because, as I understand them, it is part of the content of reactive attitudes that you recognise another as a person, and, as such, as owed at least a minimum degree of good will – a recognition that seems required. But we might think that you cannot be required to have a feeling and also that affective states involve a particular focus or attending that seems optional. Even if resentment is warranted, one might think, you are not required to have the attending to the wrongdoer’s willing that resentment involves. On the other hand, the very focus and attending that these feelings involve is part of grasping the content in a way that involves caring about it. For example, grasping the content of your having done something wrong without caring about this in the way that guilt involves seems to be missing something. This raises the question of whether one can fully grasp the content without the feeling, and I think working this out might be part of understanding the non-empirical feeling of respect at the centre of Kant’s account of moral motivation.

I have questioned Herman’s emphasis on the idea that acts of benefaction threaten or appear to threaten status equality. In defence of this idea, Herman argues that it makes sense of Kant’s instructions about the care benefactors need to take when need is the result of injustice (p. 19). But that example is complicated because (I think) part of Kant’s concern is that the background injustice makes us liable to misunderstand the situation, in particular when we take it as one really involving benefaction and generosity. So status inequality is in the picture, but that is because of the failure for independence to be secured by right, rather than the ordinary context of interdependence that exists within conditions of right. However, this does leave me wondering about how we should think about gratitude outside of conditions of (minimal) justice. If gratitude resides ‘within a set of ethical duties that are to shape the actions and relations of persons already acting under the governance of duties of right’ (p. 22–23), I wonder how we should think about the workings of gratitude in conditions under which people are not so acting.

Further questions about injustice are raised by the very unity Herman so helpfully presents between Kant’s moral and political systems. She argues that ‘[t]he heart of the moral habitat idea is an account of the moral in which its work is done by two cooperating parts, interconnected yet differently structured. In one we each separately do what we ought; in the other the same we, only now acting together, are tasked to do something we cannot accomplish acting one by one’ (p. x). Herman shows that in addition to the constitutive role reason’s highest principle plays in constituting our free agency (Kant’s argument, I take it, in the Groundwork), our rational willing (practical reason) is unavoidably committed to willing as a public ‘we’. This is required by such basic conditions of action as using objects in the world in a way consistent with treating others rightly. Without our entitlements to the use of some things being determined and reconciled with each other, in using things I wrongfully constrain your use of things and make unilateral decisions I am not entitled to. But I need to use things to act. Kant holds that we cannot resolve this through our individual willing but only through an omnilateral will, which seems to mean that willing as a ‘we’ realised by a public authority is a condition of our agency to which we are rationally committed. But now it seems that on this picture, we are rationally committed to willing something that is not there, or is not properly there, when we are not living in conditions of justice, and further, on Herman’s account, we are committed to this as the background condition against which our moral and non-moral lives can make sense. This might seem to threaten the sense we can make of our moral lives under conditions of injustice. I find this a fascinating idea and think that more work is needed to determine whether it is a feature or a bug of a Kantian system (I incline to the former).Footnote ¹

As Herman notes of her imperfect duty to be an agent of change, ‘[t]his sort of talk about change and adjustment makes most sense in a system of duties not grievously unjust at the root’ (p. 183). This leads to the further project of working out what the moral habitat view tells us to do when we exist in conditions that are ‘grievously unjust at the root’. As I understand it, one way the system gives us guidance is through the thought of whether what we are doing could be consistent with the rational willing of a ‘we’, as well as consistent with getting us to a condition of right. Herman argues that while ‘the content of articulated juridical duty will vary, […] [it] will be limited by the core values of innate right, so even where there isn’t an appropriate law, a practice can be evaluated in relation to the question of whether it could be consistent with, for example, any legitimate system of property’ (p. 113). There is obviously much to be said about this. Another resource, I think, is in Kant’s notion of provisionality, and what it means for the content of rights. Herman discusses the idea of provisional human rights which she understands as signals of ‘a systematic deficit of juridical right’ (p. 116). A different notion of provisionality, as I read Kant, is a matter of the unavoidable indeterminacy of the boundaries of rights outside of a system of right. Property rights, for example, are unavoidably incomplete and indeterminate because our use of our property unavoidably affects others’ property in ways they have not consented to and affects resources not owned by anyone (air, water, a stable climate). This feature of property is ubiquitous, unavoidable, and significant, including in ways that are more obvious and pertinent now than they may have been Kant’s time (though I think indeterminacy is central to his account). We currently seem to treat property rights as premised, roughly, on the idea that everyone has an enforceable entitlement to use their property in ways that destroy resources they do not own so long as these are not owned by other individuals. This is indefensible from the point of view of right, which implies that our entire system of property and exchange is unjust. But understanding property rights, Herman shows, is the background against which we make sense of our moral lives, including what is ours to give, and what counts as generosity. Taking injustice seriously raises many questions about how we should understand our duties – questions which this brilliant book gives us a lot of resources for thinking through.