Freedom flighters

May 25, 2007

Today at lunch a colleague cited “a recent study” purporting to have demonstrated that fruit flies exhibit free will. A quick search indicated that the study has been making headlines at Reuters, MSNBC, Slashdot, and a great many other news sources.

I don’t think I am committed to the view that fruit flies do not have free will, but since the alleged findings of this study immediately struck me as unlikely, and even comical, I must in fact think that they don’t. It’s worth taking a few minutes to look into the matter more closely.

First of all, if you read the scientific paper that has spawned the news articles you’ll find that the phrase “free will” appears nowhere, and, it seems to me, with good reason. The experiment went something like this: they placed fruit flies in a completely white environment in which there could be no visual input that would provoke a fly to zig in one direction rather than zag in another. Then they mapped the actual flight paths of the flies, tried to model those paths using deterministic or random models, and found that they could not. They concluded that the flies exhibited “spontaneity”, which they attributed to “brain circuits which operate as a nonlinear system with unstable dynamics far from equilibrium”. I am not an expert in this area, but it sound like they are claiming that the behaviour is chaotic (in the technical sense). It seems worthwhile to point out that if the fly brain is a nonlinear system with unstable dynamics, then it is deterministic, though unpredictable owing to fine sensitivity to initial conditions and perturbations.

Given those results, it is surprising, in an interesting way, that at least one of the authors (Björn Brembs) would start talking to the press about “free will”. His reasons for speaking in this way are unclear, but a plausible reconstruction might be something like this: the fly’s behaviour matches neither a random model nor a (predictable) deterministic one; it therefore looks spontaneous, or free; therefore flies have free will. That Brembs thinks this conclusion compatible with his more sober claim that the behaviour is caused by a nonlinear system in the fly’s brain is quite revealing. Since he does not hesitate to simultaneously claim that the behaviour is both free and determined, on condition that it is unpredictable, he evidently identifies freedom and unpredictability. They are for him the same thing.

This raises some interesting questions. A preliminary, and perhaps slightly cheeky, question is whether more advanced techniques in the analysis of nonlinear systems could render the behaviour non-free; in that case a fly’s possession of free will would be contingent on a technical shortcoming, which would be odd, to say the least.

But it is more interesting to observe how the conception of free will implied here differs from the common view. The common view, if I’m not mistaken, and roughly speaking, is that a being has free will if it is able to choose intentionally and without necessity between alternatives. This power of choice and freedom is, in the West at least, deeply bound up with our ideas about personhood, for freedom is usually taken to be one of the distinguishing marks of a person. This power, moreover, is usually said to have its seat in the soul. I’m not sure I know what is really meant by that, and I am reluctant to endorse a dualistic view of the human person, and I honestly don’t have a clear understanding of what freedom is or how it is possible, but I can say that whatever I think free will is, the “free will” discovered in this study isn’t it. In my mind, freedom and determinism, whether unpredictable or not, are incompatible. On some days I don’t see how one can be committed to a materialistic view of the world and still admit the possibility of genuine freedom; on such days I evidently regard free will as a spiritual power of some kind. A materialist, it seems to me, would have to engage in equivocations on “free”, which seems to be exactly the strategy adopted here.

In fact, given the authors’ commitment to materialism it is worth noting that even if fruit flies exhibited genuine free will, understood as some kind of spiritual power, they would be unable to see it. Instead, they would attribute the free behaviour to “a nonlinear system with unstable dynamics”. This phrase functions for them not as an explanation, but as a way of saying “We don’t know”.

I said above that the usual understanding of freedom is involved in some way with intention. This seems to me a crucial point, for it highlights the close relationship between freedom and interiority. How is it possible that we attribute free will to one another? I think it is because each of us has first-hand experience of his own freedom, and extends that freedom to others on the basis of their being so much like himself. If that is true, it is not clear how we could ever confidently claim that a creature like a fruit fly has free will, for we have no access to their subjective experience.

I’ll make one further remark about the concept of freedom underlying this experiment and the interpretation of it. It is telling that when they designed an experiment to test for “free will” they thought to present the fly with no visual stimuli, reasoning that anything else would introduce bias in the fly’s flight path. In other words, they conceive of freedom as unconditioned choice. This is a pervasive, but peculiarly modern, view of freedom. I don’t know when it originated; I associate it with Kierkegaard’s (unfortunate) insistence that faith, in order really to be faith, must be without reason. If you have reasons for belief, you don’t have true faith; if you have reasons for your choices, you don’t have true freedom. On this view, a rational person cannot be free because he always has reasons for doing what he does. On this view, freedom and goodness are at odds with one another, for our attraction to the latter takes away the former. This is very muddled thinking.

The older view is that the truly free man is the one who consistently chooses the good. Each of us is confronted every day with a variety of goods. The goods themselves exist in a hierarchy: the good of existence is higher than the good of a satisfying meal, for instance. We are, by nature, attracted by the good. We are biased. Yet it remains within our power to reject the good, or rather to choose a lesser good in place of a greater, thus distorting in ourselves the true reality of things. In this view, the good man is predictable: he always chooses the good. And he does so freely, which is to say, out of love. This pre-modern conception of freedom strikes us as somewhat strange, and I’m aware that I don’t fully grasp its implications, yet I believe it has much to recommend it.

Do flies have freedom in this sense? This experiment can’t help us decide.

But I suspect not.

2 Responses to “Freedom flighters”

  1. Mike Says:

    Hi Craig! I found this fruit fly study a week or so ago and was similarly a little perplexed at it. Björn Brembs posted a comment to Slashdot explaining himself a bit:

    Of course, our original study makes no mention of free will, it is not a scientific concept. However, spontaneity even in flies makes us ponder what, if anything, this might entail for our subjective experience of free will in a macrocosm we believe to be largely deterministic. Therefore we addressed the issue with an ironic question in our press release: “Do fruit flies have free will?”
    Of course, the media will drop the question mark, because questions don’t sell. Some journalists even told me their editors told them to emphasize the free will thing precisely for this reason. That’s fine with me. The debate got re-ignited and that’s a good thing, I believe.

    I would dismiss the talk of “free will” to be self-promotion to build up prominence of the paper, and nothing else. The study itself suggested that fruit flies’ behaviour followed a Lévy distribution, which is interesting, but orthogonal to any discussion of free will.

    As to free will itself, I suspect any definition of “free will” will necessarily be unfalsifiable, and thus unsuitable for scientific study. I think the best we could hope for would be to define “free will” as “making a decision in a way that can’t be modelled in our current scientific understanding”, but of course as soon as we were to study and understand it, it would cease to be “free will”.

    P.S. excuse any formatting errors, as I don’t appear to have a preview button 🙂

  2. cburrell Says:

    Thanks for your clarifications, Mike. So I was correct in thinking that they are using “free will” as a kind of stop gap to cover what they don’t understand. Or, rather, they are using “nonlinear system with unstable dynamics” in that way, and “free will” comes in, as you say, for its sensationalism.

    That the journalists would run with the theme, and do so in a sloppy manner, is deplorable but not unexpected. We’re all aware that journalistic standards, perhaps especially in science journalism, are not very high. That Brembs himself would stoke the flames by speaking in this way I consider most unfortunate. It is bound to be misunderstood.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: