Meir Dan-Cohen interviewed by Richard Marshall.


Meir Dan-Cohen is a hard-core Harvard-tough philosopher of law. He has written the books Rights, Persons and Organisations: A legal Theory for Bureaucratic Society and Harmful Thoughts: Essays on Law, Self, and Morality. He is inspired by Kant’s Kingdom of Ends, and thinks that the ideas that we create create us. He thinks legal positivism is a bad thing and dignity better than autonomy. His armchair is definitely not burning.

3:AM: You clerked for the Supreme Court in Israel before you became a philosopher? Is that right? What was it that drew you to philosophy and what are the rewards of philosophical studies and practice for you?

Meir Dan-Cohen: Yes, I did clerk for a year as part of my legal training. In hindsight, this looks like a step in a futile attempt to escape fate. I must have been doomed (or destined) early on to do philosophy by my mother, who, quite atypically for parents, used to chastise me as a child: “Don’t read so much; think for yourself!” This indeed captures the convenience of philosophy as well as one of its main rewards: the feeling that you’re making up stuff. 

3:AM: In 2002 you published Harmful Thoughts, a collection of essays that discussed aspects of a moral system that emphasises reasoning, associated with Kant and labeled deontological. It’s an approach to moral theory usually contrasted with utilitarianism, the theory that emphasises the consequences of actions rather than the reasons motivating actions. There you begin your thinking about this approach by identifying two strands that contemporary liberal moral thought draws from Kant. One is the liberal idea of individuals having freewill and autonomy and being able to freely choose their actions. The other is the idea of ‘communication or dialogue as a medium for forging principles of justice or fair terms of social cooperation … an insistence inspired … by Kant’s uplifting fantasy of the Kingdom of Ends.’ You argue that the conception of a self that is asked to play a role in these liberal theories is problematic, don’t you? Can you say something about why the non Dan-Cohen Kantian liberal self is problematic?

MD: The point is not so much that these Kantian themes in contemporary liberal theory are grounded in an inadequate conception of self, as that they are for the most part grounded in none. Kant himself does of course link his moral theory to a conception of self, namely what he calls the noumenal self, defined by a rational free will, whose freedom is a matter of being exempted, on metaphysical grounds, from deterministic causal laws of nature. Morality has a grip on us qua noumenal selves and by virtue of this radical, metaphysically grounded freedom. Relatedly, the Kingdom of Ends is a forum in which noumenal selves are envisaged as engaging in moral self-legislation by spelling out the practical implications of their shared humanity. So the noumenal self with its metaphysical underpinnings plays an absolutely crucial role in Kant’s own moral theory. But no one seems to subscribe to Kant’s metaphysics anymore, and consequently the noumenal self has pretty much dropped out of the vocabulary of contemporary normative Kantians.

A good example of the resulting theoretical gap is provided by the most influential contemporary liberal treatise, John RawlsA Theory of Justice. As you know, Rawls explicitly models his procedure for deriving principles of justice on Kant’s approach. Specifically, the original position, Rawls’ central constructive device, is supposed to resemble Kant’s Kingdom of Ends by serving as the forum in which abstractly conceived individuals generate through joint deliberation shared moral principles. On a closer look, however, Rawls’ approach turns out to be fundamentally different from Kant’s.

Since the Kingdom of Ends is inhabited by noumenal selves, it is designed to capture an essential truth about human beings and morality’s hold on them. By contrast, Rawls’ original position, cutoff from Kant’s metaphysical moorings, is conceived as merely a hypothetical meeting of imaginary representatives. Ingenious though it is as a constructive device, the original position turns out to be just an invitation to conduct a thought experiment. But this invitation can be easily refused. Without the metaphysical support of a noumenal self, or some suitable alternative, it is not at all clear why actual, fully informed individuals would care about the pronouncements of hypothetical others blinded by a veil of ignorance.

3:AM: So if those are the kinds of challenges the conception of self sets, you propose a theory that requires we see normative engagements as not just presupposing a self but as shaping it. Is that right? Can you explain your idea?

MD: Much as I would like to get the credit, I’ve been actually trying to draw some implications of an old and multifaceted theme, human self-creation, which goes back at least as far as the fifteenth century Renaissance philosopher Pico della Mirandola. What according to Pico distinguishes humanity from the rest of creation is that human beings have no essence, and so must create their own. Various more recent schools of thought, such as existentialism, postmodernism, and communitarianism, as well as the talk by sociologists of the social construction of the self, can be all seen as variations on this basic theme.

Now when we think of ourselves as self-creating, we don’t think of ourselves as biological organisms; organisms don’t create themselves, nor do we create ours. The idiom of self-creation belongs to the domain of meaning: only if we think of the self as constituted by meanings does it make sense to think of it as self-creating. This links self-creation to another cluster of partially overlapping approaches which conceive of the self in narrative terms, and so as an unfolding story, or dramaturgical terms, as enacting scripted roles. The core idea common to all these otherwise diverse bodies of literature can be rendered in the form of a slogan: that the meanings we create, create us.

Substituting this broad idea for the noumenal self provides a platform on which I reexamine certain aspects of the Kantian strand in contemporary liberal discourse. The point you mention is probably the main implication. If humanity is self-creating, and the self-creation is a matter of spinning out meanings, our normative engagements must be seen as principal constructive engines. That’s how law and morality don’t only guide what we do but also shape who we are.


3:AM: So this socially constructed self brings about other commitments, such as the contingency of meaning (contrasted with metaphysically underwritten freedom of choice), and a shift from personal autonomy to human dignity (replacing volition with identification). Perhaps you could say something about both of these. So to take the first one first, what does it mean to say we replace “a metaphysically underwritten freedom of choice” with “a recognition of the contingency of meaning”? Can you give an example?

MD: I’d like to comment first on your use of the “social construction” label. I do indeed invoke this idea, but I view it as part of the broader one I’ve just mentioned, that of human self-creation. “Society” is an intermediate category between a universal perspective encompassing humanity as a whole, and an individual perspective. Self-creation takes place at all three levels: the universal, the social, and the individual.

As to your question. It addresses primarily the notion of responsibility and relatedly that of choice. The phrase “metaphysically underwritten freedom of choice” alludes to Kant’s picture, which creates a metaphysical link between the two: responsibility is a matter of the noumenal self’s will being radically free, that is free of the natural causal order. If we don’t buy into this metaphysical picture, we face the challenge of reconciling our responsibility with the possibility that causal determinism prevails. Different “compatibilist” views which attempt such reconciliation have been proposed. According to the one I favor, human beings constitute themselves as subjects or agents primarily by avowals and ascriptions of responsibility. To say “I did A,” where A is some object or event, is to take responsibility for A, and so to constitute oneself as (among many other things) the author of A. This in turn amounts to treating certain bodily movements as meaningful and so as actions.

For bodily movements to be meaningful in this way, however, alternative interpretations must be conceivable. This is the sense in which meanings are always contingent. What is not, however, required for our responsibility is that the bodily motions themselves could have been different in a sense that implies exemption from the causal order. The same is true about choice. The crucial thing is that my will be aligned with the option I realize, not whether or not I might have realized other options instead. In order for my choice to be meaningful, alternative possibilities must be imaginable, but nothing in principle depends on these possibilities being actually available to me, since my identity is in any event progressively shaped by the path taken, not the ones foregone.

3:AM: What about the other shift you advocate, from a focus on autonomy to dignity?

MD: Yes, this too is a result of giving up Kant’s metaphysics. In general liberalism holds individual choices to be sacrosanct, and in the Kantian strand of liberalism, this attitude is linked to the value of autonomy. But when contemporary Kantians drop the notion of a noumenal self with its metaphysical underpinnings, an aberration results. On Kant’s view, the choices that exhibit our autonomy are made in our capacity as noumenal selves, and are conceived as idealized choices which take proper account of all moral considerations. Kant’s theory is therefore far from providing a blanket validation of all individual choices as such.

However, once we abandon the strict separation between the idealized autonomy of a noumenal self and the psychology of the individual chooser, we can no longer assume that choice will be consistent with the demands of morality in general, and those of dignity in particular. When such a gap between choice and dignity opens up, dignity must dominate. Individual choices are themselves valuable only insofar as they are consonant with the moral worth of the individuals themselves.

3:AM: This really is, as you say, an ‘insider’s critique of liberal theory’. The political state is usually presented therein as an evil to be resisted. You find liberal theories of resistance to the state complacent, don’t you? Can you say what the problem is and how your critique avoids this?

MD: The liberal tradition includes some enthusiasts for the state and some opponents of it. So the tradition as a whole can be seen as exhibiting an ambivalence toward the state. I myself espouse the ambivalence, rather than one or the other of the contrasting polarities. The charge of complacence you mention addresses a fairly common style of dealing with what is one of the most salient, necessary, and yet abhorrent aspects of the state, namely its coerciveness. Not surprisingly, a central question discussed by liberal thinkers (as well as many others) is when is coercion justified. What is often missing is an awareness that coercion is at best a lesser evil, and so even when justified, an evil nonetheless. Absent such awareness, we tend to cultivate toward the state attitudes that are blind to the fact that the state often tramples our autonomy, and that its hands are soaked in blood.

3:AM: There are philosophers such as Josh Knobe and Brian Leiter who contend that the notions of agency that are required by Kantians to justify their approaches are not supported by empirical evidence. How do you respond to this kind of criticism which basically seems to be saying that Kantians are relying on a fantasy of agency? 

MD: I prefer not to engage in indirect debate with the people you mention, but simply respond to the point you make on their behalf. Understood in a certain way, to speak of “a fantasy of agency” is not to voice an objection. Other ways of making the same point would be to describe agency as a product of the imagination or as a projection of the mind. These are different ways of claiming that the constituents and the incidents of agency are not the recordings of an independent observable reality, in the way in which our senses can be said to record such a reality. The challenge is not to find empirical support for the idea that people are responsible, since no amount of observation would reveal that and no tools for that purpose exist. Empirical evidence, or its absence, has accordingly nothing to do with it. The challenge instead is to replace Kant’s metaphysics with a more satisfactory story, and then explore the implications of that story for the precise shape of agency and its contours.

3:AM: And again linking with Brian Leiter but this time more specifically in the domain of jurisprudence, Leiter famously argues for naturalising jurisprudence. Presumably you resist this. What do you say to the arguments that say that naturalised law is a better description of what legal systems do than Kantian descriptions?

MD: Once again, let me comment on the general project you refer to rather than on any particular version. You won’t be surprised to learn that the project of naturalizing any normative domain strikes me as misguided. We know from the start that any account along these lines can at best tell a parallel story that fits the normative facts without however retaining their normative import. The “ought” in such accounts must lose its force. But if what we’re interested in is precisely the normative import, such accounts will be unsatisfactory. This is not a theoretical point, but a meta-theoretical one. No theory can set its own conditions of adequacy. What interests us, and so what type of theory will give satisfaction in any given domain, is not itself the product of such a theory but a precondition for engaging in the theoretical enterprise.

More specifically, the ambition of naturalizing law, as I understand it, is to fit law within the causal order. The interest in such an approach is to explain things in terms of common sense causation rather than in terms of some esoteric or ad hoc conception. And although common sense causation is not an altogether clear notion, three tenets define it if anything does: that causation can’t be instantaneous, retroactive, or distant. But when a judge in, say, Sacramento renders a decision annulling a contract made between two Berkeleyans three months ago, she violates all three tenets in one fell swoop.

3:AM: So perhaps I can ask how your theory fairs with something like the Occupy Wall Street movements of protest? Is it possible for your critiqued Kantian liberalism to show how these protesters are acting as good liberal citizens? What do you see as the salient issues to be considered? 

MD: An interesting question, though temperamentally I’m averse to being relevant. I’m quite old fashioned in believing in the value of reflection that keeps a certain distance from daily headlines. Even so, let me indicate one connection between things I’ve written about and the present upheavals. The Occupy movement targets the corporate world. In doing so, it’s important not to destroy what is, for all I can tell, an enormously useful social instrument that can serve, when properly handled, as an immense source of prosperity.

But it’s equally important, and vital to the Occupy agenda, to insist that corporations are no more than that, i.e., social instruments. Specifically, they are not human beings, endowed with dignity and capable of autonomy. So it is a grave mistake, and indeed a category mistake, to extend to corporations Kantian rights designed to protect these values. The law commits this mistake on a massive scale. Indeed, in a recent decision (Citizens United v. Federal Election Commission) the Supreme Court has carried this perversion to a new level by extending to corporations rights to participate in the political process.


3:AM: There’s a lot of discussion about global politics and of ‘importing’ ideas of rule of law and democratic liberal political systems into places with little or no experience of this kind. Does liberal theory have a universal mandate as these operations presuppose? 

MD: One of the aspects of Kantian moral theory that I find particularly appealing is its universal scope. (It’s not the only variant of liberalism with such scope; in an important, though somewhat different sense, utilitarianism too is a universal doctrine). But liberalism originates and has its main habitat in the West. Any doctrine that makes normative claims that reach beyond the cultural, historical and political boundaries within which it has developed is a potential peril. The danger is exacerbated in this case because of the West’s abhorrent record of colonialism and imperialism. Missionaries of human dignity can in principle be every bit as naïve or arrogant or befuddled and at any rate as destructive as missionaries of Christ have often been. Indeed there is a significant overlap between the two.

So a Kantian liberalism must tread in this area with great care. The fundamental value that defines on my view Kantian liberalism, human dignity, applies to every human being on the globe. That much in non-negotiable. But dignity is an expressive value: it concerns the meanings of actions and practices, insofar as the actions and the practices must convey respect for each individual. And unlike the value itself, which is universal, the meaning of actions and practices, and so what counts as respectful, is cultural and local. So the universality of the Kantian position at the same time requires sensitivity to cultural variation. 

3:AM Does liberal theory as you critique it handle interstate disputes as complex as Israel and Palestine, for example?

MD: It would be surprising indeed if the theory had no implications in such matters, though for the reasons I’ve already mentioned, it hasn’t been my goal to spell out these implications. But I welcome the opportunity to draw one rather obvious connection between a dignity-based liberalism I favor and the issue you raise. As you may know, human dignity is now enshrined in Israel’s Basic Law (the equivalent of a constitution). This constitutional provision has had laudable effects on Israeli society through numerous Supreme Court decisions and by other means. But the effects are for the most part limited to Israel proper.

Palestinians’ dignity is regularly being infringed in innumerable ways. The degradations and oppressions visited on them, large and small, are explained by the security exigencies that arise in the occupied territories. These exigencies are indeed real enough, and so often appear to justify the measures used. But the argument really ought to go in the opposite direction. In light of Israel’s own professed values and moral aspirations, the fact that the present state of affairs is taken to require the trampling of human dignity and so leads to the tarnishing of Israel’s moral self-image provides all by itself a compelling reason for ending the occupation. 

3:AM: I’m always attracted to arguments that reveal inescapable dilemmas. Yet often people appear to want reasons and arguments and theories that can be the last word on the subject. Your liberalism doesn’t however presuppose that having the right moral theory is enough to solve moral dilemmas, does it? In fact, you think that there will always be dilemmas. One place where you show why is in your discussion of what you call ‘acoustic separation in criminal law.’ Can you say what this is and why it leads to dilemma and why you think having really hard dilemmas will survive any good moral theory?

MD: I’m glad to find in you a soulmate in this respect, Richard. Theories are as likely to reveal unfamiliar problems as they are to help cope with familiar ones. In any case theories don’t solve practical problems. Every theory involves a simplification, and the insight each offers is in part the product of its restricted tunnel vision. It is eventually up to the readers to compose out of many such partial visions what would hopefully be a richer and more accurate picture of the overall lay of the land. And it’s up to people who possess practical judgment, which theories themselves don’t supply, to cope with problems and dilemmas that theories may help identify.

Turning now to my acoustic separation paper, the main observation is that the law speaks to different audiences in different voices, which need not fully coincide. For example, the law addresses the general public, guiding our behavior in part through the threat of sanction, while also addressing officials, specifically judges, instructing them to impose sanctions when we misbehave. But those two, the threat and the execution, may diverge: the law’s bark can be worse than its bite. That is so as long as the condition I call “acoustic separation” obtains. This simply means in this case that for a variety of reasons people may be unaware of the courts’ greater leniency compared to the threats addressed to the public. However, when a dog’s bark is worse than its bite, the dog is bluffing. So here’s a case in which we can mitigate law’s harshness by compromising its candor. I don’t propose how to resolve this dilemma, but insist that it’s perennial and make its details visible. The resulting picture of law is not meant to be a pretty one. This goes back to the point we discussed earlier in connection with our proper attitude to the state in light of its pervasive coerciveness. 

3:AM Finally, are there legal philosophers that the savvy non-specialists here at 3:AM should be reading? Other than your own books, are there five books in this realm that you’d recommend to enlighten us?

MD: I’m afraid that I’m going to give you a longer answer than this simple request seems to invite. The reason is that Anglo-American legal philosophy has been stuck for awhile in a rut of its own making, and though some excellent people are writing in the area, few of the results would be of general interest. The name of this rut is legal positivism, a fairly disastrous theory of law that came to dominate the field. In its original form, the basic idea is, roughly, that law is the expression of a sovereign’s will, and so in principle independent of morality or other substantive, rational considerations. Now positivism combines an insight with a mistake. The insight consists in alerting us to a significant pathology to which law is susceptible. The mistake is to treat pathology as physiology. It’s as though having discovered polio one were to consider paralysis as the essential nature of the human organism, or view diabetes as the key to metabolism. The problem with positivism is actually even worse than these analogies suggest, since adopting this attitude to polio and diabetes doesn’t by itself spread the disease. But in the case of law, the theory is to some degree self-supporting. If you teach people in power that whatever they fancy, just because they fancy it has potentially the meaning of law, they are likely to give freer rein to their fancy in exercising power than they would otherwise. 

This theory has two unfortunate effects on the practice of legal philosophy. First, having found itself in this rut, legal philosophy has invested great ingenuity into contriving ways of digging itself out. But all the intricate footwork is bound to be lost on the general reader who isn’t charmed by positivism to begin with. Second, law conceived along positivist lines presents a rather narrow and dull philosophical subject matter. The pronouncements of politicians or judges are not as such of great philosophical moment, and so a theory of law that holds these pronouncements to represent what law is, provides a rather narrow and impoverished philosophical agenda.

The conclusion is that your readers might do better to look beyond what counts as the current canon of legal philosophy, and consider books that aren’t a standard part of it. Here are a few illustrations. No one needs me to recommend reading John Rawls, but I’d like to start with him, since re-reading him as a preeminent legal philosopher, not just in his The Law of Peoples but also, and even more importantly, in A Theory of Justice, may prove illuminating both about Rawls and about the field. Similarly Jurgen Habermas’, Between Facts and Norms; not an easy read, but worthwhile, by an author who treats law as part of a broader vision of society, politics, and morality. The third suggestion is Ronald Dworkin’s new book, Justice for Hedgehogs. As I’m sure you know, Dworkin is part of the canon, perhaps its most prominent active representative, but was a critic of positivism from the start, and writes with a breadth and flair that give him a broader appeal. The two other suggestions belong to a somewhat different genre. Since the authors aren’t card-carrying members of the legal philosophy club, their philosophical approach is enriched by an interdisciplinary perspective; and the books themselves don’t attempt a general theory of law but explore one or another aspect of law with attention to the kinds of substantive issues it faces: Elizabeth Povinelli’s The Cunning of Recognition, and Shai Lavi’s, The Modern Art of Dying.

Richard Marshall is still biding his time.