A Critique of Ronald Dworkin

by Benjamin Studebaker

Lately I have often praised the work of Ronald Dworkin, writer of Justice for Hedgehogs, a book I have recently been reading. Indeed, Dworkin’s views on scepticism, interpretation, and the independence of value from metaphysics are all very persuasive, and I have adopted partially or completely several of his positions on those topics, as regular readers may have observed in recent posts. However, at around the halfway point in Hedgehogs, I have come upon a position of Dworkin’s I cannot accept which I believe undermines much of the rest of his philosophy.

The argument turns on Dworkin’s embrace of responsibility as a core principle in his ethical and moral theories. Regular readers may remember that I have rejected personal responsibility as incompatible with both my view of distributive justice and my determinist views. Dworkin acknowledges the danger that a rejection of responsibility on determinist grounds poses for his view:

These questions about responsibility hang like swords over Chapter 9.

As a result, Dworkin devotes his tenth chapter to an attempt to rescue responsibility from the jaws of determinism. He does this by adopting a compatibilist view, claiming that responsibility has nothing to do with what causes our actions and is instead derived from what he terms our “capacity”. In this way, he separates out two interpretive understandings of whether or not responsibility exists:

  1. The Causal View–we cannot have moral responsibility unless we are the uncaused cause of our own actions.
  2. The Capacity View–so long as we have the capacity to form true and pertinent beliefs about the world and match our decisions to our personality, we are morally responsible.

It is my view both that the capacity view is incomprehensible and invalid and that the attacks Dworkin levels against the causal view are ineffective.

Dworkin defend the capacity view on the grounds that it justifies the way we presently deal with say, insane people, or children. An insane person is someone who does not have sufficient capacity to form true and pertinent beliefs about the world or to match his decisions to his personality. By this, Dworkin means that an insane person either operates under delusions about the world (say he believes himself to be a god, for instance) or is compelled by his illness to act against his beliefs (say he believes in pacifism but is his illness causes him to be extremely violent, for instance). For Dworkin, an insane person falls below the capacity threshold, and so is not responsible for what he does.

The trouble is that capacity is not a switch. You do not either have these capacities or not have them, people exist on a continuum, ranging from the completely insane or severely mentally retarded to well-adjusted geniuses with high levels of intellectual clarity. We say that insane people are not responsible while sane people are, but we do not account for the distinctions in capacity among sane people. Some sane people have greater capacity than others, yet we do not say that some sane people are more responsible than others. One’s prison term is not going to be longer or shorter purely on the basis of one’s reasoning capacities, and if capacity did play a significant role in determining the length of prison sentences, our inclination would be to deem it an utterly ridiculous practise.

Furthermore, how do we judge sanity to begin with? We judge it based on people’s actions. A person’s capacity is measured by what they do. A person who acts in ways that we deem good and reasonable is thought sane, and a person who acts in ways we deem bad and unreasonable is thought insane. So by definition, the greater the capacity, the better the behaviour. This does not jive with the way we behave in practise. If it did, the worse the crimes a person committed, the lower capacity we would have to assume this person to have, and the smaller the punishment we would have to give them. It would entirely invert the present system.

Dworkin claims that his capacity principle provides grounds for blaming Joseph Stalin for his actions rather than attributing them to his nature and upbringing, but it does precisely the opposite. The paranoid behaviour of Stalin is evidence of his lack of capacity to perceive with clarity his level of risk, it is evidence of an inability to accurately and pertinently form beliefs about the world.  But because Stalin’s paranoia is not so severe as to classify him under a mental illness, we are supposed to give him full blame for this paranoia as if he were a reasonable person? A reasonable person would have been incapable of acting with the paranoia Stalin acted with, because a reasonable person’s capacities for understanding the world would have been, by definition, higher. The capacity view attempts to attribute credit and blame on the basis of natural characteristics. It is equivalent to praising Shaquille O’Neal purely for being very tall. The capacities we have for decision-making are not self-determined, they are the products of our nature and nurture and it is unreasonable to blame people for characteristics that they have no influence over. While we most certainly benefit from people with high capacities, because they do good things, those good things are not owned by them in any distinctive way; they are the natural consequences of a nature which happens to be benign.

As for Dworkin’s attacks on the causal view? Dworkin makes two claims about the causal view:

  1. It is impossible to genuinely believe it.
  2. We cannot derive an ought from an is.
  3. It doesn’t agree with Dworkin’s view on personal responsibility, and therefore it is wrong.

The last argument is simply too circular. The causal view proposes that Dworkin’s view of responsibility is wrong, because it attributes blame and credit to people who are not the uncaused cause of their actions. To say that the causal view is wrong because Dworkin’s responsibility view is right is merely an evasion. We should focus instead on the first two.

Dworkin claims that it is impossible to really believe that we are not responsible for what we do because we are not the uncaused cause of our actions on the grounds that when we are put in a situation in which we must decide something, we cannot help but think that some decisions would be better than others. This is a misinterpretation of what the absence of responsibility means. Say for instance that I am trying to decide whether or not to shoot someone. It is consistent to say that it is better if I do not shoot someone while simultaneously saying that I am not responsible for what I do, because I am not the uncaused cause of my actions and will “decide” to do whatever I do on the basis of natural and environmental factors that are beyond my control. Being aware of the fact that I am not responsible does not fundamentally change my nature or free me from being determined. Dworkin observes that I will inevitably make a decision one way or the other and that I will be unable to help but feel ownership for the decision, but that is not because responsibility is true, it is because non-responsibility does not change my nature.

Imagine that a robot named GLaDOS has been built to perform tests on subjects. In order to get the robot to perform as desired, it is programmed to take pleasure in testing and to experience pain whenever it intervenes in the test in a way that would void its results. The result is that the robot becomes sadistic, it runs test after test to maximise its pleasure while refusing to help or to be kind to the test subjects in any way. If I then inform the robot that it is only acting in this way because I programmed it to do so, unless the robot has other programming that makes it enjoy rebelling against what it believes to be its programming more than it enjoys following the programming (which would amount to reverse-programming, a very silly idea), this information will be unable to change the fundamental forces that determine the robot’s behaviour. It enjoys testing and it doesn’t care about test subjects, and unless I intervene in the programming itself, I change nothing.

Informing people that they are determined is a kind of environmental influence, but it is not going to override fundamental aspects of our nature. We will still follow our programming. You can tell a heroin addict everything there is to know about why he is compelled to shoot up, but at the end of the day he remains a heroin addict. In and of itself, the information is unlikely to have any significant effect. In this way, not changing our behaviour in response to knowledge of determined status is not evidence that it is impossible to believe we are determined and not responsible, it is, in contrast, evidence that we really are quite similar to computers; we cannot change our essential nature. It is for this reason that we often view as futile monastic orders that bar their members from say, sexual activity. The incentives that encourage us to have sex are very strong and genetic in origin, and it is unreasonable to expect the socialisation that “sex is bad” to always be stronger than the genetic impulse.

Dworkin’s appeal to Hume’s Guillotine, arguing that determinism in itself can have no direct influence on what is moral, is a clever move. However, determinism does not in and of itself change what is moral. If we agree that we cannot be responsible when we are not the uncaused cause of our actions and we come to believe that we are never the uncaused cause of our actions, our moral principles have not themselves changed. Just as Dworkin proposes that we say “people who lack certain capacities are not responsible”, we say “people who are not the uncaused cause of their actions are not responsible”. The size of the group is a function of brute facts, but the moral principle itself is entirely divorced from this; it retains its independent status. If we said “deer are not responsible for what they do” and we took “deer” to mean only one very specific subspecies of deer or to mean all the varieties of deer in the world, our moral principle would remain the same. It is merely our definition, our understanding of the brute fact, which would change. Moral principles still apply to real people and real things. While those real people and real things cannot determine moral principles, they can influence their scope for application. We are not saying that responsibility is never sensible; we are saying that responsibility is only sensible when we are the uncaused cause of what we do, and since the scope for the truth of that is zero, the scope for responsibility is correspondingly zero.

It is unfortunate that Dworkin buys into the compatibilist evasion and consequently the responsibility view; in many other respects, his work is sterling.