Saturday, May 30, 2009

Dissection of the will

An experiment separates the threads of consciousness and action

One of the deepest sources of supernaturalism is the mystery of consciousness- the sense that our soul/spirit is so unmaterial, shiny, and pure that it could not possibly be sullied with evolutionary origins, messy neurological mechanisms, and eventual death. There is an undeniable and to some an unbridgeable tension between this remarkable experience and its rather obvious dependence on our brains.

One expression of this tension is the doctrine of "free will", which has two senses. One theological dilemma is, if god is so good and powerful, why are we not His obedient automata? Because He gave us free will, with which to be bad and yet also with which to worship Him, if we are smart, because he likes to be freely loved, not slavishly obeyed. Related to this theodicy is the second sense of free will- that our inmost soul yet still partakes somehow in the divine to the extent that it is immaterial, has a smidgen of moral knowledge (often ignored), and is eternal, even if, looking backward, it has its origin in a mysterious and divine implantation process concomitant with sperm and egg intromission.

At any rate, we feel free in our actions, while at the very same time physics tells us that no action happens without causes- causes that would obviate the felt authorship of our actions. The answer in scientific terms is to posit an "illusion of free will", where our actions, when not obvious responses to conscious exigencies, are caused by unconscious promptings which simultaneously arise in consciousness as 1: the idea which "pops" into one's head to move one's arm, and 2: the action itself, activated in parallel. In this scheme, our actions all have causes, just not always causes we are aware of. And the sense of authorship is a fiction conjured simultaneously as the actual causes activate planning and action.

The motor areas of the brain are known to form a crescent across the top midline from ear to ear, where past experiments have generated a "homunculus", or map of body parts moved by stimulating brain locations in this region. A completely separate sensory homunculus maps bodily sensations to brain locations, and is not involved in this work.

A group in France experimented on seven patients getting brain surgery for other reasons, stimulating the surfaces of their brains to figure out how actions arise and how they are perceived. Stimulation in the areas shaded purple below, which are in the motor area, caused body parts to move, but without any cognition that the person had caused the action, that they had wanted to cause the action, or indeed that the action had taken place at all.

Conversely, stimulation in the areas shaded yellow and orange below (the volitional field) caused patients to experience urges to begin an action, and at higher levels of stimulation, even the sensation they had carried out an action, though none had actually taken place. One patient, stimulated with 5 mA at site a in the figure, said "I felt a desire to lick my lips", while at 8 mA at the same location was quoted "I moved my mouth and talked, what did I say?" (Doubtless these quotes are translations from the French!) No action took place, however, and electrical EMG recording of the corresponding muscles gave no trace of any activity, confirming that the feeling the patients had was not connected to their bodies.

Locations of experimental stimulation, in the motor area (purple) and volitional area (orange, yellow), also called Broca areas 39 and 40. The SMA lies slightly forward of the motor area.

Of these two fields, the motor field is easier to understand. The output end of the process could be followed down the spinal chord and even to the muscles with electrical stimulation, creating actions authored by the experimenter rather than by the patient, who would have no sense of volition. The paper mentions that lesions in this area of the brain are sometimes associated with "patients who obstinately claim that they can move their paralyzed limbs." This paralysis is not physical but due to a disconnect between volition, which is intact, and motor execution, which is not. Additionally, the paralysis is not apparent to the patient, indicating that, as the paper hypothesizes, perception of action is more dependent on the internal sense of volition, which assumes as a matter of course that an action has taken place as planned, than on feedback from the body, surprisingly enough.

Tracking down the sense of will and authorship is more interesting. A consistent model is that a normal action engages a more extensive network of brain activity, of which only tiny portions were activated in these experiments. One terminus of a consciously or unconsciously generated action is the motor areas that actually control their fulfillment (purple), but another terminus elsewhere (yellow, orange) creates the consciousness first of wanting to do the action, and if intense enough, of having performed it. This second area may later integrate sensory feedback to enhance the sense of authorship and allow learning, but based on these experiments, this feedback is not critical to a sense of authorship, matching the conclusion of Wegner. The volitional area maps closely to the mid-brain regions that were pointed out in a recent blog to be the best guess for those responsible for consciousness.

There is another area adjacent to the motor field, called the supplemental motor area (SMA), where others have done experiments of this type. Here, electrical stimulation generates both what is sensed as an urge to perform an action contrary to one's will, and also, at higher currents, the action itself. The key distinction is that here, the urge is recognized to be contrary to the patient's will (as happens in various tic and involuntary movement syndromes), while stimulating the volitional area creates such will directly, without generating the action. The SMA may be a preparatory area for motor action, placed after the unconscious originating signal has separated between paths to the motor area and to the volition/will/feedback area, but still having partial input to non-volitional consciousness.

The bottom line is that the sense of authorship is conjured in an area of the brain that doesn't actually cause the action, but is a post-processing epiphenomenon. This sense is essential to learning, to a coherent sense of self, and to moral responsibility, but it is not causally essential. Given that the causes of action are often "reflex", or "spontaneous", or otherwise unconscious, it would require impossible circular logic to have our sense of authorship be mechanistically equivalent to actual authorship.

Incidental links
  • Philosophy bites podcast does neuroscience- the senses, blindsight, mirror neurons, alienation, anarchic hand, bodily coherence, etc.
  • New Yorker profiles neuroscientist Vilayanur Ramachandran, who deals with paralysis and other neurological oddities.
  • Science mag review of above paper.
  • Feminist atheist activist ... Taslima Nasreen
  • Supernaturalism explained, and podcast.


  1. Hello,

    Interesting post. I think your conclusion about epiphenomenal authorship is only partially correct.

    What is clear, I think, is that there are many sorts of actions in which the sense of authorship is consequent, and not causally efficacious. But it would be a mistake to thereby assume that the sense of authorship is not causally efficacious at all.

    The sense of authorship must lead to behavior; otherwise, we would have no sense of it. As you say, the sense of authorship is crucial for the formation of our sense of self. It is necessary for the story of our lives. And those stories have consequences.

    It is highly implausible that some actions, such as the writing of a personal letter, are not in some way caused by one's sense of being the author of that action.

    I thus view the sense of authorship, not as an epiphenomenon or fiction, but as a regulative principle necessary for certain types of behavior.

  2. Hi, Jason-

    One has to ask how, in close detail, behaviors like decisions and writing letters, take place. They arise from unconscious processes, like desires, motivations, even prior decisions stored in menory. Suppose I used a piece of paper to do a calculation to help me consciously decide a course of action. One could still hold a consistent model of all the actions and motivations arising from unconscious sources, with consciousness watching and monitoring all the while. In a way, this is consistent with your claim, per Dennett, that consciousness is a much smaller phenomenon than we usually give it credit for being.

    The feeling of knowing what is going on is not quite the same as direct responsibility, even though we have the quite justified faith that our unconscious processes are properly trained and responsive to outside logic and to internal motivations. That is, until they are not, as in the case of addictions, or strokes, where we may tell that we want to do X, but can't or won't. There are enough neurological syndromes and dreaming/drug states that separate the sensation of authorship from the actual doing to show that they can go their independent ways.

    I am not saying that this is all nailed down, but that this consciousness-as-caboose model is one that reasonably accounts for causality and lack of free will in a physics/neuro sense while still allowing for psychological free will, as well as being consistent with key cognitive science experiments. Daniel Wegner's book is pretty good front.

  3. "your claim, per Dennett, that consciousness is a much smaller phenomenon than we usually give it credit for being."

    I don't recall ever making that claim. I'm not sure Dennett would say that, either, though it depends on what you mean by "smaller."

    Also, I have trouble accepting the claim that decisions, desires, and motivations are unconscious processes. We talk about things like decisions, desires, and motivations because we lack a robust model of our brains. It would be a category error to use those terms to refer to neurological states or processes, unless you clarified exactly what you were talking about in neurological terms--and thereby change their definitions.

    Also, I'm not sure what sort of "faith" you are talking about. I tend to steer clear of that word, unless it is very clearly defined.

    Of course, your interest in this post is specifically with free will. I haven't addressed that topic, and I wasn't intending to. However, I will say a word about it, since it is motivating your discussion.

    Let's say Dennett is more or less right, and when we talk about consciousness and intentionality, we are adopting a stance--a framework for making predictions about complex behavior which depends upon a tangible ignorance of the causal processes producing that behavior. The point is, our notions of "self" and "desire" and so on, these are all ways of talking which help us ignore everything we don't know about what really causes our behavior.

    This creates a certain amount of flexibility, in terms of what is or isn't in the provice of one's control. What one could or couldn't have done differently.

    The notion of free will is the notion that one could have acted differently if one had wanted to. Yet, and I think Dennett makes this point well--if we examine this notion closely, we realize it doesn't pose any problem to a completely deterministic notion of behavior. For the point is, if you had wanted to act differently than you had acted, then the causal nexus would have been different, and so you would have acted differently.

    If we try to match up our notion of free will to our notions of causality, there is no real problem at all. The only problem occurs when we think free will must mean the ability to act without cause at all. But surely nobody actually means that when they claim to have free will. For acting without cause is acting without desire or intention, and that would not be will.

    So it seems that the philosophical "problem" of free will is not a problem at all. The only question is, what is or isn't in our control in any given instance? What should we hold people accountable for?

    These can be difficult questions, but they don't seem to have any profound philosophical or scientific significance.

  4. Hi, Jason-

    Insofar as Dennett claims essentially to dissolve the concept of consciousness, by your telling, I thought that claiming it to be "smaller" followed that line of thought.

    I knew the word faith would be provocative, but it is a matter of trusting our own minds to be responsive to the world, physically, socially, morally. As dieters and other addicts know, that faith is often misplaced, so it is a constant struggle, especially in childhood, but throughout life, to align our thoughts and behaviors with our longer-term models of what is good for us. I know this is a bit esoteric, so no need to get into it.

    Thanks for your presentation of free will- it is very persuasive. The only comment I would make is that the key concept of "wanting to act" in some way is not commonly understood to involve causes at all. We may "want" spontaneously as though we were radically free to choose, with no reasons necessary. That relevant causes are nevertheless plentiful and mostly unconscious is not always appreciated.

    Best wishes!

  5. Okay. If "smaller" means soluble, then I agree; according to Dennett (and myself), it is smaller than many have made it out to be. However, I think Dennett and I would agree that dissolving the concept is only worthwhile if we change the focus and function of our discourse. The notion of "consciousness" has currency in our language, regardless of whether or not we depend upon it to explain humanity.

    As for our "faith" and "trust" in our minds . . . I don't follow your reasoning. I'm not sure exactly what you are trying to get at, or why you have chosen the word "faith" (apart from wanting to be provocative.)

    Re free will, I suppose the idea you are getting at is this: For those who believe in free will, the important thing is that our wants and desires are our own--that they are not determined by something outside of us. So, while our actions might be caused by our wanting, our wanting should not be caused at all. At least, not caused by anything outside of us.

    But this notion falls apart just as easily. For, if our wanting is not determined by our circumstance, by the world in which our wanting makes sense and has meaning--well, then what is the value of such wanting?

    If "free will" means the freedom to act on desires which have no connection at all with the world, then who would want free will?

    We do get the impression that our wants arise spontaneously, that they are ours in a mysterious way, a way which cannot be reduced to biology or physics. But, as I think we both agree, this impression is the result of the way in which the notion of wanting is constructed. It is constructed out of narratives.

    The point which I would like to stress, and which I think you may not be taking into full consideration, is that such narratives are not constructed in order to produce accurate representations of anything. They are constructed in order to establish new rules of behavior; higher level organizational principles which make things like writing pesonal letters, defending one's self in court, and other social events, possible. We can take these notions as representations of facts; but the notions are not born as representations, and we should not expect to find any clearly defined objects to which they refer. So, notions like "want" and "desire" have social significance, not biological significance--unless and until we redefine them in biological terms, and thus change the way they are used.

  6. Hi, Jason-

    We talk about things like decisions, desires, and motivations because we lack a robust model of our brains.

    Well, we know enough to say that the vast majority of what goes on is unconscious. Something like playing tennis employs vast amounts of body sense and trained reflexes, and only tiny amounts of conscious executive focus. Ditto with any other activity, I'd say in general.

    As for our "faith" and "trust" in our minds . . . I don't follow your reasoning. I'm not sure exactly what you are trying to get at, or why you have chosen the word "faith" (apart from wanting to be provocative.)

    I do like to be provocative. But this connects with the point above- that so much of what we do is mindless that we have to have faith in our own unconscious to get us to the train on time. One most striking observation to me is our internal prioritization/scheduling apparatus. On a regular day, we go smoothly through a routine of tasks, then at work adopt some hierarchy of -things that must be done- often aided by physical lists and calendars. But those are optional. Mostly, things just pop into our heads as they bubble up from the unconscious and turn out to be important things to do at that time. That is the kind of thing I mean when speaking of faith. Often it is misplaced, but when it is radically lost in various addictions, obsessions, & depressions, a person's life is hell.

    If "free will" means the freedom to act on desires which have no connection at all with the world, then who would want free will?

    Everyone seems to. The common intuition is that we can want things as the consequence of nothing at all- which is the "free" in free will. As you say, this is easily shot down. I was just articulating the position.

    So, notions like "want" and "desire" have social significance, not biological significance--unless and until we redefine them in biological terms, and thus change the way they are used.

    I don't strongly disagree, but would offer that you may be getting a bit excessively lexical about this. "Wanting" does indeed correspond to biological, i.e. brain, states. They are not terribly well defined or constructed, but brain studies are really putting some definable states behind these words. What is going through our heads has physical instantiation without exception, so your shyness about linking the two seems misplaced. Indeed, the fixation on behavior seems almost to be a project to substitute a new teleology for one lost with supernatural theology.

  7. Part of the problem I'm having with your approach is that you adopt the terms "unconscious" and "conscious" without clarifying what they mean.

    You say, "the vast majority of what goes on is unconscious." But, what does that mean? Does it mean that much of what goes on in our brains is not reported, or reportable? Or does it mean something about "felt experience?" And how could you investigate felt experience apart from what is reported or reportable?

    You want to separate the way we talk about consciousness from what we can say about reports. This makes your discussion of conscious/unconscious processes unclear, and apparently without significance.

    So, again, when you talk of "faith in the unconscious," I have no idea what you are talking about. The terms "faith" and "unconscious" here lack definition.

    About free will, I'm not so sure people who claim to want free will are claiming to want to have desires which have no basis in reality. Most likely, they don't know exactly what they are claiming to want. They just want to be able to maintain some notion of agency. And I suppose we agree that this notion does not conflict with a causal account of behavior. Though I'm not sure if you still think we need to think of one's sense of agency as an epiphenomenon, or fiction.

    You say "wanting" corresponds to biological states. But, as with the case of hunger, when we normally talk about wanting, we are not referring to anything. We are performing social functions. While we can correlate the language of "wanting" with biological states, it makes little sense to say that "wanting" refers to those states; or that when we refer to wanting, we are referring to those states.

    The language of "wanting" is a language based on social behavior, and which makes sense in terms of social behavior. (And this does not change when you are talking to yourself outside of anybody else's range of hearing.) You cannot understand that language by ignoring what the body does in social situations.

    And, of course "what is going through our heads" has physical instantiation. There is no sense in postulating some non-physical events here. In so far as anything happens, it happens physically. So I think you have wrongly accused me of shyness. I am not trying to separate the mental from the physical.