Is free will a scientific problem?

Posted on July 18, 2013

An American neuroscientist claims to have solved the problem of free will. Peter Tse, who works at Dartmouth College in New Hampshire, says that the key to free will lies in how neurons can rewire each other. But I would argue the problem of free will is a conceptual problem, not a scientific one.

In an article he recently published in New Scientist, and at much greater length in his book The Neural Basis of Free Will, Tse sets out his theory according to which neurons rewire each other. They can form temporary circuits, and alter the criteria to which they respond in the future.

Freedom in the circuits

What does this have to do with free will? Well, it suggests that how we respond to information is dynamic: the same input may cause us to respond in different ways at different times, depending on how our circuits have come to rewire themselves in the interim.

Our responses depend on us.

Further, the process is neither deterministic nor random. It is not deterministic because – Tse speculates – neural rewiring and firing may be partially random, as some physicists have suggested. But even if rewiring and firing is partially random, how we respond to information is not.

Circuit rewiring sets criteria for what inputs they process and for what outputs they produce. Within the constraints set by these criteria, the result may be random – in Tse’s example, when I am asked to think of a politician, it may be any politician that comes to mind – but the constraints ensure that it will be a politician who comes to mind.

Whether Tse is right that the brain processes information in part by setting and resetting criteria for possible inputs into temporary and longer-lasting neural circuits I don’t know.

The neuroscience is intriguing, if far from settled, and deserves further investigation. And Peter Tse is well qualified to conduct it. But I don’t expect to see the problem of free will illuminated by that research.

Free information processors

When we pose the problem of free will, we suppose that there is some scientific story about how the brain processes information and that the story will detail how our responses are not random.

Assuming that some such story can be told, we ask whether it is enough for us to count as free. Tse doesn’t begin to address this question.

There is no doubt that human beings are sophisticated information processors (indeed, so are many other animals). Why isn’t that enough for free will?

The traditional problem, as Tse mentions, concerns determinism. Some philosophers think that if determinism is true, we can’t be free (roughly because if determinism is true, then how I will act is settled by the laws of nature and events before I was even born).

Tse thinks that determinism is false, and he might be right. But indeterminism doesn’t help all by itself.

It is not enough that how I act is not already settled; indeterminism adds to my freedom only if I control whether I perform one act or another, when it is undetermined which I will perform.

A recurring problem

Tse may think that his story gives us the ability to exercise this kind of control, but it doesn’t. On his account, the rewiring of criteria constrains the choices we will make, but which one of the options available to us we ultimately select seems to be genuinely random.

It might be determined, say, that I will choose either to give to charity or to spend my money selfishly on myself (and that I will not burn it or attempt to eat it), but it is genuinely random whether I give it to charity or spend it selfishly.

It doesn’t help to point out that the setting of criteria for choice might itself have been undetermined. The same problem arises for every choice and every mental event, including the event that sets the criteria.

It will be true that the set of possible alternatives is constrained, and that within those constraints, how we choose is genuinely undetermined.

The limits of science

Now, this story might be enough for free will in any case. But it is enough for free will only if determinism is not a threat to free will.

If it is enough that we are sophisticated information processors who respond intelligently to the situations we confront, then there is no particular reason to be threatened by any element of Tse’s story.

After all, nothing in the story suggests we are coerced or compelled to act: rather, we act as we see fit on the basis of our own values.

That might be enough for free will, as many philosophers believe. But the irony here is clear: Tse’s account gives us free will only if the problem he aims to solve does not exist.

If it is enough to be an intelligent information processor to be free, then the details don’t matter much: so long as our brain implements information processing somehow or other, it doesn’t matter how it does it.

The problem of free will, as I argued at the outset, is a conceptual problem, not a scientific problem. We already know, by observing our own behaviour, that we are sophisticated information processors. It is philosophical argument (and not neuroscience) that will answer the question whether that’s enough for free will.

Neuroscience can tell us about when and how brains don’t work as they should, or about local failures of free will. It can’t answer the philosophical question itself.

Source: The Conversation, story by Neil Levy