Friday, October 08, 2004

The Phenomenology of Free Will

I'm anything but an expert on the philosophy of free will. I'm familiar with the major positions and arguments, but not with most of the issues. I've never been incredibly interested in the area, because it has seemed to me like most of the positions rest on faulty theories of individuality or self (more on that later, maybe). However, after reading this paper (which I learned about from this post on Experimental Philosophy), I thought I'd comment on the phenomenology of free will, or will in general. In particular, since some theories of free will, particularly libertarian theories, are dependent on the introspective feelings of free choice and willing, some recent research in psychology might cause problems for these theories. What if it turns out that our conscious feelings of will are often mistaken? What would the effect on such theories of free will be?

How might our feelings of will be mistaken? Here is a scenario from Dan Wegner:

What if our minds keep showing us the same set of appearances, leading to an impression of conscious will again and again, but never revealing to us how our actions are actually caused? One way this could happen is if both the thought about action and the action itself are caused by unperceived forces of mind: you think of doing X and then do X – not because conscious thinking causes doing, but because other mental processes (that are not consciously perceived) cause both the thinking and the doing. Based on your conscious perceptions of your thoughts and actions, it would be impossible to tell in any given case whethe your thought was causing your action, or something else was causing both of them. Could it be that the deep intuition we all have about the power of our conscious will is the result of this ‘sleight of mind’? Perhaps we experience conscious will when we infer that our thought causes our action, although we can’t really know that this is the causal path1.
Wegner argues that the experience of will is something that goes along with action, rather than being the cause of it. As evidence, he has conducted several studies in which people observe an action with an external cause, but are tricked into believing that they themselves caused it. They are able to do this because they can simulate the conditions that cause us to believe that we have consciously caused an action. For a person to believe that he or she is the cause of an action, his or her thoughts must occur prior to the action, be consistent with the action, and be exclusive (i.e., there are no immediately available plausible alternative causal explanations). Here are some examples research by Wegner and others:

  • In one experiment2, participants sat in front of a computer with various objects on the screen. The participants wore headphones, and kept their hand on a board that controlled the mouse. Across from them sat confederates, who were also able to control the mouse (see photo below). Participants would observe an action, with the cursor moving across the screen to one of the objects, either before or after they heard a word in the earphones. If they heard the word one or five seconds prior to the movement of the cursor, most participants felt that they had caused the action, when in fact the confederate had caused the action . If they heard the word thirty seconds prior to the action, or one second after it, they did not feel like they had caused the action. This indicates the necessity of priority of a thought for the experience of willful causation. In addition, if the word participants heard in the headphones was inconsistent with the object on which the cursor stopped, they did not feel that they had caused the action. This indicates the importance of consistency for the feeling of causation.


  • From Wegner & Wheatley (1999). Click to enlarge. Posted by Hello


  • In another set of experiments3, participants were placed in front of a mirror. A experimenter stood behind them, out of view. The experimenter's hands were placed in front of the participant, where his or her hands would be (see photo below). As in the experiment above, participants observed an action, in this case the movement of one of the confederate's hands in the mirror, either before or after hearing a consistent or inconsistent set of instructions. If the instructions were heard prior to the movement, and were consistent with it, participants experienced a feeling of consciously causing the movement, even though they were aware that the hand they were observing were the experimenter's. Furthermore, in this condition (prior and consistent instructions), when participants subsequently observed a rubber band slapping the experimenter's hand, their physiological response was similar to the response caused by a slap on their own hand.


  • From Wegner et al. (2004). Click to enlarge. Posted by Hello


  • In third set of experiments by Wegner4, participants placed their hands on the arms of an experimenter, and were told to feel the experimenter's muscle movements. After feeling the muscle movements, the participants were told to type out answers to questions for the experimenter. Participants usually answered correctly, and attributed the answer to the experimenter, even when the experimenter had not heard the question.

  • Nielson published a paper in 19635 in which he described an experiment similar to Wegner's. In this experiment, participants placed one of their hands into a box that they believed would allow them to view it. Instead, they actually saw the hand of an experimenter in a mirror. Participants then used the hand in the box to trace a line on a piece of paper. The experimenter would deviate from the from the line, and the participant would mistakenly correct for the deviation. Not knowing that the hand they watched was the experimenter's, participants had no alternative explanation for the movements they observed, and therefore felt as though they were causing them.


  • In the first two and fourth set of experiments, participants mistakenly felt like they were the cause of an event of which they were not. In the first Wegner experiment, and the Nielson experiment, participants even believed that they had caused the events that had actually been caused by an experimenter. In the second set of experiments by Wegner, the ones with the arm movements, participants felt like they had caused the arm movements even though they knew that the arms were not their own, and their physiological responses to the rubber band strikes indicate that their body reacted as though the arms were their own when they felt like they had caused the movements. In the one set of experiments in which participants did not mistakenly feel like they had caused an action, they mistakenly attributed the cause of the action to the experimenter when they themselves were in fact the cause. The lesson to be learned from these studies is that our experience of agency is a poor indicator of whether we are in fact causing an event. What does this mean for theories of free will that rely on the experience of willing? I'm not sure. It probably indicates that our introspective feeling of willing cannot be the only justification for holding a particular theory of free will. Reliance on the phenomenology of free will alone is problematic since our experience of willing is unreliable.

    1 Wegner, D. M. (2003). The mind's best trick: How we experience conscious will. Trends in Cognitive Science, 7, 65-69.
    2 Wegner, D.M. and Wheatley, T.P. (1999) Apparent mental causation: sources of the experience of will. American Psycholist, 54, 480–492
    3 Wegner, D. M., Sparrow, B., & Winerman, L. (2004). Vicarious agency: Experiencing control over the movements of others. Journal of Personality and Social Psychology, 86, 838-848.
    4 The Illusion of Conscious Will by Dan Wegner.
    5 Nielson, T. I. (1963). Volition: A new experimental approach. Scandinavian Journal of Psychology, 4, 215–230.

    4 comments:

    Anonymous said...

    "Reliance on the phenomenology of free will alone is problematic since our experience of willing is unreliable."

    Chris, I'm not sure this really follows, any more than reliance on vision is 'problematic' since we are susceptible to optical illusions; or that our experience of causality is 'problematic' because someone might occasionally getting the feeling that it's almost as if turning on the light switch was causing the telephone to ring (or whatever). The fact that there are situations in which we are fooled doesn't mean that we are fooled all the time, nor does it mean that the particular sort of information on which appeals to the phenomenology of free will are based is really affected by the attribution problems Wegner and others look at. It could mean that, of course; but a more elaborate argument would be needed. In particular, I'm not convinced that people, in appealing to the phenomenology of free will, are appealing to it to decide whether we are causes of events in a given particular case.

    I think a more serious charge against reliance on phenomenology is that we really can't do it in sophisticated cases unless we're willing to do sophisticated analysis at least occasionally. And I think this is a flaw that characterizes far more than proponents of free will. I think, for instance, Wegner occasionally faces problems with an excessive reliance on phenomenology. The Illusion of Conscious Will is a great book, but it's always seemed to me to suffer from a lack of analysis of what's really involved in the experience of conscious will - things are all sort of piled together in a vague way as having to do with 'the experience of conscious will' when one can actually argue (sometimes on the basis of the experience itself) that they are actually quite distinct. At the very least, there's a lot of danger of equivocation about what is meant by 'conscious will' and it's often the case that neither proponents of free will nor people who think it illusion take the proper safeguards against it. 

    Posted by Brandon

    Anonymous said...

    Brandon, I think you're right, if we just view Wegner's examples as similar to errors of judgement or "deceptive" sensations. However, if we view it as telling us something about how, or at least when, we perceive ourselves as being the cause of events, then it might be more problematic for theories of free will that rely on the phenomenology of will. What I take it as showing is that the experience of will is a post-hoc experience based not on perception of my own will, but on perception of certain environmental facts, and comparison of those to my own thoughts. If that is the case, then the phenomenology of free will is useless as a ground for theories about free will, because it tells us nothing about whether we've actually willed anything, just that the conditions that we associate with willing are present.

    If that is the case, then you are right, to assess whether we are willing at all, in any given situation, we're going to have to do a more sophisticated analysis of the situation, but we will have to do this instead of relying on the experience of willing.

    I think Wegner has done a pretty good job of sticking to the data (his, and data from the last 30 or so years) when coming up with what he sees as the components of the phenomenology of free will. He probably is missing some aspects, but further data will likely bear them out. The reason I find his work interesting is that, while it may not give us a complete picture of what the experience of will is like (it's not meant to), it does a very good job of showing in what sort of situations it occurs (which is what it's meant to do), and it's that part of his work that I think has implications for free will theories. 

    Posted by Chris

    Anonymous said...

    "He probably is missing some aspects, but further data will likely bear them out. The reason I find his work interesting is that, while it may not give us a complete picture of what the experience of will is like (it's not meant to), it does a very good job of showing in what sort of situations it occurs (which is what it's meant to do), and it's that part of his work that I think has implications for free will theories."

    I suspect you're right that further research will give results similar to Wegner's, from what little I know about it. One of things that concerns me about direct application to issues of will, is that it isn't always clear to me how the experiments actually relate to the issue; for instance, I don't think that on most accounts all consciousness of causing is consciousness of willing; and I don't think that on most accounts all cases of consciousness of willing are cases of consciousness of choosing, which is what would be relevant to the free will discussion; and things like that. I worry that often these distinctions aren't made (or, rather, aren't even considered as possible distinctions) when people try to apply the experimental results to the philosophical issues. Indeed, I know that they often aren't. But I do like Wegner's work a lot; one of the great things about his book is that it gives a good argument for why someone who thinks conscious will an illusion would nonetheless find it worth scientific study, and, of course, the topics discussed are absolutely fascinating. I hope more innovative work is done along these lines; although I hope, too, that people become more careful in interpreting it. 

    Posted by Brandon Watson

    Anonymous said...

    I definitely agree about the careful application of experimental results, especially in psychology, to philosophical issues. The line between data and their philospohical implications is rarely straightforward. That's why in each of the posts where I've talked about research and philosophical issues, I've been pretty noncommital. I'm not sure what the implications are.

    You're right that Wegner's work focuses on willing, and not choice. His experiments are designed to show that we feel like the causes of events when we really aren't. I think this could have implications for theories of free will, but you're right, not for theories that rely more on the experience of choosing. I suspect, however, that you could produce experimental results that were sort of the real world analogues of the Frankfurt examples, in that people feel like they chose an outcome when they really didn't. The conditions under which the illusion of choice might occur are probably pretty similar to the conditions underwhich the illusion of will occur, though there are probably some important differences. Wegner actually talks about the illusion of choice in the old research on gambling. I forget who did the original studies in the mid-70s. I'd have to look it up.
     

    Posted by Chris