KATERINA DELIGIORGI | The Scope of Autonomy: Kant and the Morality of Freedom | Oxford University Press 2012
By Katerina Deligiorgi
I am immensely grateful to both my critics for the care and seriousness with which they have treated the arguments in my book The Scope of Autonomy. Their criticisms helped me think both about the detail of the argument and about the structure and implications of the position I defend.
I start with Jeppe von Platz’s comments because they present me with the opportunity to explain the motivation for the overall project. Setting out briefly the aims of the book will help explain my strategy and go some way towards answering von Platz’s broader concerns. I then seek to deal with the specific questions he poses.
The second part of my response is devoted to Adrian Piper’s comments. Besides offering a critical appraisal of a key thesis in the book, namely that distributive universality can adequately capture the demand for moral objectivity, Piper extends the argument of the book and offers an original contribution on the nature and scope of autonomy. Her discussion presents me with the opportunity to clarify certain features of my own position.
Jeppe von Platz states that my aim in the book is “to defend Kant’s ethics”. This is not quite right. My aim is to defend a conception of autonomy. I turn to Kant because I think that he alone offers us a morally defensible conception. The search for a morally defensible conception, in turn, is motivated by the conviction that many current theories of autonomy are morally under-resourced, empty or indeed pernicious. I begin then in partial agreement with the many critics of autonomist ethics, who are puzzled about the hold that autonomy continues to have over our moral thinking. I seek to dispel the puzzlement by pursuing a double strategy. One strand of the argument aims to show that Kant provides us with a morally powerful definition of autonomy, the second strand aims to vindicate the basic appeal of simple assertions of our right to do as we see fit, by showing that there is a discursive path leading from such simple assertions to a morally demanding position. The two strands are held together by what I call the nomological conception of autonomy, which seeks to recover the importance of the notion of ‘law’ or ‘nomos’ within autonomy.
A first and central task is to show the connection between autonomy and moral objectivity. I prioritise the formula of autonomy, because it yields precisely this connection. I employ the formula definitionally, because, as Piper remarks, Kant here puts the moral requirement in the indicative mood. This is important for my purposes because it helps with giving an account of autonomous agency as moral rational agency and a clear view of what is required to be such a moral rational agent (i.e. imperatives simply command, they do not describe). The connection between autonomy and moral objectivity is given by the notion of universal law. I interpret this in terms of universalisability. I then argue that universalisability expresses a formal conception of objectivity. This portion of the argument also seeks to show that thinking a priori about what is right is not thinking about matters of fact, actual or possible, but about matters of ought. When von Platz presses the issue of the fact of reason, he is right in suggesting that I offer a negative account only, that is, I argue that it is not an intuitionistic moral fact. As a positive claim, however, that there is a fact of reason is even less contentful, since I take it as the bare assertion that there are irreducible normative facts. Another way of putting this is that in my view the fact of reason aims to block, not to offer a positive answer to, searches for the sources of normativity. So what follows or could follow from it for rational beings and their access and subjection to such facts is a different story altogether.
Action and motivation, where I turn next, raise three distinct questions: how we put autonomy to practice, what it takes to act autonomously, and why should we do so. My answer to the first question is really rather simple, we must undertake some internal house-keeping. But this is not the usual view of things. Motivational issues have tended to dominate the reception of Kant’s ethics, so they need to be dealt with in some detail, else they distort the moral conception of autonomy I seek to defend. I discuss eight possible interpretations of motivational internalism, three of which are versions of the pro tanto version that von Platz favours. I argue that these pro tanto versions either cede ground to externalism or create trouble for themselves on the way they seek to align morality and rationality. Since von Platz does not go into the detail of the disagreement here, I shall not elaborate further. The critical examination of motivational internalism is intended to take the heat out of that discussion and focus on what I think really matters, which is agent involvement and agent control. These two notions give the answer to the second question, i.e. autonomy requires a conception of ourselves as setting ends. Practically, this means that we are involved in the choice of ends and that we control this choice. Filling in the commitments that go with these requirements takes up the rest of the book.
Regarding the third question—why act autonomously—I seek to show that attempts to establish an ‘internal relation’ between agency and autonomy fail. So the question cannot be answered in this way. I doubt the importance of the question as raised purely from outside morality. I further think it destructive to seek to answer by adducing non-moral grounds. However, I do take the question seriously as a question that arises from within the moral domain. One way not to go about it is by ceding ground to internalism about reasons, a view inspired by Bernard Williams’s internal reasons argument, which, if accepted, damages the prospects of objectivism and a genuine appreciation of the normative role of pure reason.
This deals with the negative argument. The positive contribution I have to make to this discussion is showing that there is a discursive path leading from simple assertions of our right to do as we see fit to a morally demanding position. It is only a discursive path, nothing like a metaphysical deduction or anything of the sort. Briefly it goes like this: the attraction of the simple assertion rests on a promise of agent involvement and control; the moral conception of autonomy I attribute to Kant and defend in the book makes this promise good. Still, the book is called ‘the scope of autonomy’ also for this reason, that even a defensible moral conception of autonomy, indeed one that I see, following Kant, as key to objectivist ethics, is not the be all end all of ethics.
Given the foregoing sketch of the aims and argument of the book, here are my answers to von Platz’s questions in the order in which he puts them:
1. I do not discuss the formula of humanity because I do not seek to show that Kant’s ethics is intersubjective. Such a project would overlap considerably with mine, but it is not mine. My aim is more limited, it is to show that Kant’s conception of autonomy takes others into account in its very formulation.
2. The question of the relation of the different formulae and the supreme principle of Kant’s ethics is an important topic, but again it is not mine. What I do say in the book is that for the account of autonomy given here it is important that morality is unifiable, in the sense that what is morally truly demanded forms a consistent whole; this is what I call the ‘optimistic’ view. Unifiability bears directly on Kant’s unicity claim about the moral law and, flowing from this, that the formulae that express the moral law are to be viewed as one, though of course, as Paul Guyer, Allen Wood and others have shown this is easier said than done.
3. The feeling of respect is not a distinct topic in my account because it fits equally well standard internalist accounts and the account of motivation I attribute to Kant, which is merely not anti-Humean. The key point is that any internal house-keeping we are asked to do is for the sake of what morality objectively demands of us and that presupposes a non-Humean view of agency because what is at stake is setting ends according to what reason demands.
4. I discuss Kant’s position on feelings and affects in relation to Schiller in Chapter 5 of the book, with the aim precisely to show the extent to which moral autonomy can accommodate a complex moral psychology. Because the chapter is comparative and argument focused, I do not engage in exegetical issues. I hope, however, that combined with the previous chapter on motivation, it shows that the autonomous agent is not a rational automaton.
5. Adding a notion of truth to Kant’s moral epistemology can indeed appear superfluous. It gains its position in the argument first in supporting the claim about the kind of cognitivism and moral realism I attribute to Kant. Cognitivism requires that judgements be truth-evaluable. Kantian cognitivism is a kind of cognitivist realism. It is a realism in the sense that the answers we give to what is right or wrong do not depend on our attitudes or interests but what is right or wrong is not reducible to anything non-moral. Unlike other varieties of non-naturalist realism, however, there is no object of moral cognition; we have rather an objective means of determining what is right. Objectivity and truth are tied to a method of justification, namely the universalizability test. So there is no gap between epistemic access and moral truth. The second reason why truth matters is not directly about moral epistemology but rather about semantics; what I try to show is that the formal objectivity of universalisability gives meaning to the concept of ‘morally right’.
6. The essence of this question, to the extent that it is directed to the main aims of the book, is about the relation of autonomy to freedom. This issue makes contact with the metaphysics of freedom presupposed for autonomy, which I discuss in the final chapter, and with what I call the metaphysics of agency, which refers to basic features of the rational moral agency presented in the book. The features that describe the agency are agent involvement and agent control, though of course what is thereby being described is an ideal, and as such, a norm, not a fact of agency.
7 & 8. The cumulative argument I attribute to Kant aims to show that we are best placed to assert our freedom when we act morally. One line of argument, which I pursue elsewhere, explains how the Kantian position is internally coherent. Briefly, the Wille and Willkür distinction I use has antecedents in earlier distinctions Kant uses in his lectures on ethics and metaphysics. In any case, I take the distinction to be essential to Kant’s metaphysics of freedom, which concerns powers agents have and may exercise more or less, well or badly, and facts that tell us who these agents are, by virtue of what basic characteristics they are to be seen as free. The wicked are not off the hook because once we differentiate between different senses of freedom, transcendental, empirical, and moral, we have a better handle on this thorny issue, since we can affirm one without affirming all. In the book, this discussion is subordinated to the aim of showing that Kant’s metaphysics of freedom is not subject to the criticisms that usually attend autonomist ethics.
What kind of objectivity may we expect of our moral judgements? This question provides the focus for Piper’s essay. As she rightly points out, establishing a link between autonomy and moral objectivity is a key concern of the book. Her treatment of the objectivity question helped me see the different ways in which it can be posed and to what extent in each case universalisability can serve as the answer. So what follows is partly a response and partly a way of making my ideas on the matter clearer.
1. First, the requirement for objectivity expresses the epistemic concern that our moral beliefs are worthy of being considered knowledge, as for example when we say “I know it’s wrong to…” or “I know I should…” in moral contexts. In at least some of these cases, such assertions must be correct and we must be able to spot the correct ones (that is, it’s not enough to pick what is morally right blindly so to speak, we have to have something to say about why it is morally right). The test that addresses this epistemic need is the universalisability test. Speaking of a test encourages the sort of realist conception of ‘morally right’ I seek to avoid. The universalisability test is not like a blood test where you aim to find, for example, whether there is a certain antibody present in the blood. This makes for the peculiarity of Kant’s moral epistemology: the method of enquiry about what is morally right also gives
the justification of the belief that such and such is morally right.
The identity of method of enquiry and method of justification throws up an important disanalogy with the theoretical case: the doxastic states one finds oneself in in relation to moral rightness, include being mistaken, uncertain, confused, undecided, in need of advice and so on, but not entirely ignorant about the fact that something is morally right. To continue with the blood test example, of course if one was entirely ignorant of the existence of antibodies one would not have a test for them, the difference is that one can be ignorant about the fact that something is an antibody. The ‘what’ we are looking for in the moral case is guidance about the content of ‘morally right’. This is what Piper questions in her piece, with the Galileo and Al Gore examples, showing that objective validity and moral rightness can pull apart. I doubt however that this conclusion is inescapable. We can reach it, if, as Piper claims, we allow for ‘independent grounds’ for moral judgement. On the account given just now, such independent grounds can only count provisionally, and when there are clashes about what looks morally right or between what appears morally right and objectively valid, moral reflection moves up a level. That the set of true moral judgements forms a consistent whole is a ground assumption of the optimistic view that morality is unifiable, which also means that what is objective and what is right must coincide.
How does this apply to the Galileo and Al Gore examples? The Al Gore example simply illustrates the fallibility of any local co-legislation procedure. But this is fine, because no such procedure is final or supposed to be such. The requirement of ‘actual familiarity and connection with others’ is intended to pick up a strand, which I believe exists in Kant’s thought, especially in his popular writings and lectures that encourages us to conceive of the exercise of our rational abilities, whether theoretical or practical, alongside real others. The importance of this is to establish a connection with actual practices of moral reflection and deliberation, but it is not to endow any such exercise with finality.
The Galileo example is trickier. Galileo and his critics have a dispute about science, a dispute about scientific procedure, a legal ecclesiastical dispute and alongside there is a moral case to be answered by each of the participants in these disputes. These are all conceptually distinct. What strikes me about the example is not that it is a counterexample to objectivity but rather that it takes a lot of thought to locate the moral problem correctly (without gerrymandering the case to suit any pre-given purposes: we cannot start by saying we know for sure what is morally right and then use whatever test the philosopher proposes to see whether we can catch her out or whether she confirms us in our certitude). Let’s say the moral problem is how we should treat someone whose views dismay and appall us. If we condemn him not to teach his theories, resolving thus the legal ecclesiastical dispute in one way, do we also thereby offend morality? An affirmative answer would make this case similar to the Al Gore one. A negative answer would show that enquiry and justification come into play relatively late with respect to how we live our moral lives, since what we experience and are troubled about is fine-grained but the principles by which we try to order our lives are not.
2. The second aspect of objectivity has to do with the authority of morality in our lives. This authority is not derived. It would be questionable, however, if it did not issue from objective norms. There are those who do not see it that way, who are unperturbed by diminutions of objectivity. I see objectivity and authority as connected. Kant’s insistence that the moral law is valid for all rational beings, which might seem improbably rationalistic, is about this connection of authority and objectivity, now seen from the perspective of the scope and reach of moral demands.
3. Third and related to the previous point is the difference with rules of etiquette. Rules of etiquette can hold without exception, but you only have to obey them, if you want to be part of the club. Moral authority is about categoricity and categoricity, in this context, has a logical shape of objectivity understood as something that holds for all rational beings (where ‘for all’ is intended to block any attempt to localise the hold of morality to this or that club and therefore make our participation facultative and hypothetical).
4. Fourth, there is a semantic aspect to the discussion of objectivity: what do we mean when we say that such and such is morally right? This is a question about content. It is to answer this question that I opted for a distributive conception of universality. As Piper argues, partly the underlying motivation is to dispense with frivolous examples. Partly the aim is to capture the individual address that I saw as part of the content of what is morally right (this relates to Piper’s free rider discussion). And finally, properly the aim is to distance universalisability from notions found naturally in its proximity, such as exceptionlessness, impartiality, neutrality and so on. Even so, we are still in the domain of what Piper identifies as ‘formulation’, which she helpfully distinguishes from ‘application’ arguing that using a distributive model for the latter is unwieldy and that there are important resources in the collective model.
5. The fifth aspect to objectivity relates to the ‘application’ issue: how is universalisability to shape our moral thinking? Piper argues that more work is needed here to bring into alignment individual deliberation with co-legislation. We can have individual deliberation and collective validation if we follow a procedure that parallels the intersubjective one of formulating scientific hypotheses which, if successful,
apply distributively to any phenomenon picked out by their terms or concepts. This is tricky, however, since success is not judgeable by criteria that are not already employed in the formulation (this is the peculiarity I identified in the first point above). Piper concludes that any combination of distributive and collective would falter unless it were backed up by universality as a logical requirement on valid principles. I agree with this. I also agree that the objective force of moral principles is not a matter of either formulation or application. This is the reason I sought to connect universality with moral imagination, in the discussion of perspectival ascent, because I do not see how else the tension is to be maintained between formulation and application, on the one hand, and the pure order of reason, on the other, whilst also being productive.
I want to conclude by underlining two contrasting features of the conception of autonomy I present in the book, one is its modesty, the other its ambition. Modesty has to do with scope again. Accepting universality as a logical requirement on valid principles will not get us out of the problems Piper identified with the Galileo example, which as I sought to show helps draw our attention to a different issue, the fault-line between the fine-grainedness of the moral situation and the formality of a principle. Because of this I suspect that our philosophical moral architecture is always bound to look incomplete. Philosophical reflections feed off our ordinary pre-philosophical and indeed pre-reflective moral life. Taking our moral reflections up a level gives us a more secure understanding of what it is we are aiming for and what Kant calls ‘the supreme norm’ of morals. But this supreme norm is not intended to codify moral situations, rather it is intended to guide principled moral thought.
Ambition has to do with freedom. Piper—and this ties nicely with von Platz’s point about the wicked—concludes that “surely any viable account of autonomy must include the freedom to do wrong”. The discursive path I sought to establish between simple assertions of our right to do as we see fit and a morally demanding position have to do with taking seriously the possibility that we may be deceived when we assert our autonomy in certain ways and that the moral nomological interpretation gives us genuine involvement and control.
© 2014, Katerina Deligiorgi.
Katerina Deligiorgi is Reader in Philosophy at University of Sussex, UK. She obtained her Ph.D. in Philosophy from the University of Essex, UK. Her specialisation is in moral philosophy, meta-ethics, aesthetics and, historically, Hegel and Kant. She is the current editor of the Hegel Bulletin. Recent articles have appeared in International Yearbook of German Idealism, Journal of Aesthetics and Art Criticism, and Inquiry. Apart from various book contributions, she also published the monograph Kant and the Culture of Enlightenment (SUNY, 2005) and, as editor, Hegel: New Directions (Acumen, 2006).