Limits of Free Won’t

Another strange side light on free will. Some of the most-discussed findings in the field are Libet’s celebrated research which found that Readiness Potentials (RPs) in the brain showed when a decision to move had been made, significantly before the subject was aware of having decided. Libet himself thought this was problematic for free will, but that we could still have ‘Free Won’t’ – we could still change our minds after the RP had appeared and veto the planned movement.

A new paper (discussed here by Jerry Coyne) follows up on this, and seems to show that while we do have something like this veto facility, there is a time limit on that too, and beyond a certain point the planned move will be made regardless.

The actual experiment was in three phases. Subjects were given a light and a pedal and set up with equipment to detect RPs in their brain. They were told to press the pedal at a time of their choosing when the light was green, but not when it had turned red. The first run merely trained the equipment to detect RPs, with the light turning red randomly. In the second phase, the light turned red when an RP was detected, so that the subjects were in effect being asked to veto their own decision to press. In the third phase, they were told that their decisions were being predicted and they were asked to try to be unpredictable.

Detection of RPs actually took longer in some instances than others. It turned out that where the RP was picked up early enough, subjects could exercise the veto; but once the move was 200ms or less away, it was impossible to stop.

What does this prove, beyond the bare facts of the results? Perhaps not much. The conditions of the experiment are very strange and do not resemble everyday decision-making very much at all. It was always an odd feature of Libet’s research that subjects were asked to get ready to move but choose the time capriciously according to whim; not a mental exercise that comes up very often in real life. In the new research, subjects further have to stop when the light is red; they don’t, you notice, choose to veto their move, but merely respond to a pre-set signal. Whether this deserves to be called free won’t is debatable; it isn’t a free decision making process. How could it be, anyway; how could it be that deciding to do something takes significantly longer than deciding not to do the same thing? Is it that decisions to move are preceded by an RP, but other second-order decisions about those decisions are not? We seem to be heading into a maze of complications if we go that way and substantially reducing the significance of Libet’s results.

Of course, if we don’t think that Libet’s results dethrone free will in the first place, we need not be very worried. My own view is that we need to distinguish between making a conscious decision and becoming aware of having made the decision. Some would argue that that second-order awareness is essential to the nature of conscious thought, but I don’t think so. For me Libet’s original research showed only that deciding and knowing you’ve decided are distinct, and the latter naturally follows after the former. So assuming that, like me, you think it’s fine to regard the results of certain physical processes as ‘free’ in a useful sense, free will remains untouched. If you were always a sceptic then of course Libet never worried you anyway, and nor will the new research.

10 thoughts on “Limits of Free Won’t

  1. Is it that decisions to move are preceded by an RP, but other second-order decisions about those decisions are not?

    This is what I never quite got about Libet’s “free won’t”: if the RP is the herald of some unconscious decision to move, then how does the decision to preempt it come about? Just out of the wide blue yonder?

    That’s of course just the main issue discussions of free will always run into: if the fact that our behavior being caused by certain factors renders it unfree, we can’t get around this by introducing some sort of monitoring authority that either intercedes or doesn’t, because then, what causes this authority to act, or fail to?

  2. The definition of “free will” is always an issue in these discussions. If you’re still holding out for the theological / metaphysical / libertarian version that transcends causality, then I think the Libet experiments are a problem. Decisions are a detectable physical process that happens in the brain.

    But if your version of free will is the capacity to make decisions taking into account personal interests and the consequences for others, then Libet’s and other similar experiments aren’t really relevant, and social responsibility remains a coherent and useful concept.

  3. There are two problems here. The first is the idea that free will should be a function of just the conscious part of the entity rather than the entire entity making the decision. The conscious, unconscious and body are so intricately intertwined that it is hard to separate them. That free will should be confined to the conscious seems to be a holdover from the folklore and dualist theories of consciousness where the conscious is “in charge” of everything you do and is the “real you”.

    The second issue is that in Libet’s experiment and others like it, we don’t know if the conscious part of the brain handed off the random timing to an unconscious part. It seems plausible that it did as it doesn’t seem the conscious is well equipped for this task. When you think about what you are actually doing when waiting a random amount of time, with respect to your consciousness of the act, you wait until it “feels like” the moment has arrived to act. We don’t spend our conscious time counting out from a selected random number. This seems to fit nicely with the RP observations. Until this is shown to be incorrect, it remains a question of how this Libet’s results related to free will. Of course, if you believe the first paragraph, it doesn’t really matter anyway.

  4. Jochen (#1), what Stephen said. The RP isn’t the herald of an unconscious decision to move, but something more like – this is a hypothesis from the literature – motor preparations. Libet’s instructions specifically tell subjects to wait for a random urge to move. The urge could cause, or be largely constituted by, motor preparations. It would be a good design strategy for evolution to make the go / no-go switch wait until near the end of those preparations, to give the organism more flexibility. But nerve signals have finite speed, so there has to be a time when it’s too late to call off.

    This hypothesis predicts all of Libet’s results.

  5. Without prefacing speculations from the past, lets stay present to see times arrow as ‘direction’–for the will to be free…
    …Does our direction (evolution) attract dualism’s observation in the transformation of energy here on Earth…

    Reduction by its direction is always only a physical measurement; so how can it be dualistic…What would be the use of being conscious in the past (or conscious in the future)…Neurons of imagination and mindfulness are useful but more akin to sensing, feeling, thinking and mindfulness than consciousness or observation…

  6. They were told to press the pedal at a time of their choosing when the light was green

    The thing here is that it’s actually asking the subject to choose a random time, not a time of their choosing per se. The testers haven’t actually grasped the psychology involved and are testing something different as a result.

    Why not choose to hit the pedal instantly? No, the request makes it sound as if the subject should choose some time after ‘instant’, but before the green light goes out. What time? Well, that’s been left up in the air, and as much it’s requesting a random time.

    The lag of the subjects awareness makes a lot of sense here, because to choose an apparently random time would require choosing a point that is outside of conscious perception. The subject has to draw from a part of the brain they don’t have access to, in order to give a random result (otherwise its not random). It’s not surprising the part they are triggering is outside of their conscious experience for some seconds, when the tester is asking for a random responce time.

  7. If we think of the entire nervous system as a network of intercommunicating computational parts, acknowledging they differ from our digital computers, then decisions can get made in one part of the system at a certain time and later be received as input by another part at a later time. That second part can possibly countermand the decision by failing to pass it on to a third part or modifying it in some way. That one part performs the computation we call “consciousness” is neither here nor there. Does such a system have free will? This is more a question of one’s definition of free will than the structure of the computation. A question for philosophers, not biologists.

  8. Then consciousness seems only to appear when we are present to it (dualism?)…
    …and free will is (perhaps we do not have it, but want it) in the realm of presence, conscious and observation…

    Independence Day is tomorrow, am I to wait till then for science to test if there is a will to be free…

    Our presence our consciousness should not subject itself to the past by science in this field…
    …but instead we should stay on philosophy’s mission for truth and wisdom in searching for one’s life in the here now…

  9. Does anybody seriously still think that this experiment says anything about free will ? Says a bit about how much consciousness plays in the role of thought certainly, but what else ? Not that much.

    JBD

  10. I agree with John Davey. The fact that decisions are made before we are conscious of them is interesting to have been shown experimentally but not at all surprising. Since all brain processing takes time, decision must precede awareness of the decision by some amount of time. Of course, if one believes exercise of free will is, by definition, a conscious process, this shows that to be likely false.

Leave a Reply

Your email address will not be published. Required fields are marked *