Ethics and free will

It is often argued that the notion of ethics fails without "free will". This view is nonsense. It is true, however, that our notion of ethics has to change substantially without free will, but change it must. If we abandon the notion of "free will" then we must change our ethical thinking to be about causal efficacy rather than an incoherent and irrational notion of deserts.

"Free will" is a fundamentally vacuous, unfalsifiable notion. In all of my philosophical investigations, I've never seen a definition of free will that would permit me to determine if I myself — much less anyone else — actually had free will. Just like "God", free will seems to mean something, but when we investigate the notion carefully, it evaporates into vacuous platitudes.

Any specific behavior is the result of some causal process, physical or non-physical, or it is not the result of some causal process. If it is the result of a causal process, our traditional notion of free will is facially contradicted. But what does it mean that a behavior is not the result of a causal process? The opposite of a causal process is a random event, an event that cannot be correlated to anything previous in time. If we cannot hold someone accountable for a caused behavior, by what virtue do we hold someone accountable for a random behavior? The view that the antinomy between causality and randomness makes free will fundamentally ineffable or mysterious just digs the philosophical hole deeper: By what virtue do we hold someone accountable based on a phenomenon we not only do not understand but cannot understand?

The notion of consciousness as a non-physical phenomenon fails to rescue free will per se. Non-physical consciousness would just establish non-physical causality. It is the causality per se that contradicts free will, not the physical causality.

(There's an additional confusion about causality that's peripherally relevant: determinism and predictability: A deterministic system is often held to be ipso facto a mechanistic system, i.e. a system whose behavior is completely stereotyped, rigid and predictable. This notion is nonsense. First, there's good evidence from Quantum Mechanics that all systems are in some sense mixed, the result of an interaction between random quantum events and deterministic physical laws. Secondly, chaos theory has conclusively shown that even systems with very simple, rigidly deterministic behavior are radically, perhaps even fundamentally unpredictable. The notion that determinism leads to "sphexish" behavior is definitely contradicted by both facts and logic.)

So we have to adjust our notion of ethics: We must hold people accountable, at least in some sense, even though their behavior is causal or a mixture of causality and randomness. How can we do so?

We can see an amusing example of the issue in the scene from Fawlty Towers (I'm firmly convinced that comedians are, in general, better philosophers than most philosophers) where Basil beats his car with a stick because it won't start. Basil's action is obviously absurd and funny. But it's funny not because the car doesn't have free will and is therefore not responsible for its failure, it's funny because hitting a car with a stick is a spectacularly ineffective way of making it start.

When my car fails, I do indeed hold the car itself "accountable". If it doesn't perform as I require, then I fix the car. I don't merely conclude that since it doesn't have free will, there's nothing I can do about it. Human beings are, of course, vastly more complicated than cars, but the principle applies.

The way to think about ethics in the absence of free will is just in this causal sense. Use scientific reasoning to figure out the causal mechanisms that underlie human behavior, and then use that understanding to intervene and bring about the results we desire. In just the same sense we use science to determine the laws of physics, and then engineers manipulate the laws of physics to build desirable technology.

Taking this causal view, it becomes incoherent to imprison or punish murderers because in some mystical, ineffable and unfalsifiable sense they "deserve" punishment. We should imprison or punish murderers because that's an effective way of persuading people not to engage in the undesirable activity of killing their fellow human beings. And if imprisonment and punishment were not effective, or some other mutually exclusive activity were more effective, then it would positively irrational to continue imprison and punish murderers. Not because we condone murder; precisely the opposite. If we really do condemn murder, then the rational choice is do that which is most effective at reducing it.