Monday, 2 July 2012

Patrick Haggard: Volition and Agency: What is it, and What is it For?


      Abstract: Voluntary actions are often defined as actions that are internally-generated, rather than directly triggered by an external stimulus.  The capacity for voluntary action gives control of human behaviour a 'freedom from immediacy' (Mike Shadlen's term), that other animals may lack.  But voluntary actions are also characterised by a special relation with conscious thought.  On a classical, rationalist model, we consciously deliberate, form conscious intentions, and these drive our actions.  The crucial link in this chain is the transition from mind to body, when an intentions-in-action is transformed to a motor command, and a bodily movement.  I will discuss a number of studies of this process.  First, I will consider whether the experience of being about to act is a direct readout of ongoing neural preparation, or a retrospective narrative to explain our actions post hoc.  I will then ask the same questions about sense of agency - i.e., the feeling that our actions cause events in the outside world.  Both prospective and retrospective components are shown to exist.  The retrospective component presumably aims at providing a coherent and description of our own behaviour and self-consciousness.  The prospective component is harder to account for, and I will consider two possible functions.  First, being aware of what we are about to do just before we do it might contribute to the control or veto of action. Second, it might improve complex instrumental learning.  Both accounts suffer from the normal difficulties of ascribing causal roles to consciousness, and the existence of prospective aspects of intention and agency cannot save concepts of 'conscious free will'.  Finally, I will consider the implications of recent work on action awareness for moral and legal responsibility.

      Haggard P (2008).  Human volition: towards a neuroscience of will.  Nature Reviews Neuroscience, 9, 934-946. http://www.ncbi.nlm.nih.gov/pubmed/19020512
    Chambon V, Wenke D, Fleming D, Prinz W & Haggard P. (in press).  An online neural substrate for sense of agency.  Cerebral Cortex, in presshttp://www.ncbi.nlm.nih.gov/pubmed/22510529

Comments invited

24 comments:

  1. Doctor Haggard’s research seems to illustrate the fact that conscious agency, whether prospective or retrospective, still happens because of dynamical decisions made by groups of neurons. “Free will” and conscious action thus seems to be effects, and not causes. This is very problematic for anyone who has done Action Theory. Indeed, it points to the idea that mental events cannot cause anything: conscious volition and conscious agency are instead both caused by action selection processes. Even if the feeling itself emerged at exactly the same time as the decision was made to act in a given way—as an identity theorist would have it—it still remains that the felt experience itself would have no incidence on my actual behavior. My felt agency would merely be epiphenomena of “intentional fluency.” Given that the feeling of agency seems to be correlated to decisions made by ensembles of neurons which coordinate motor behavior, I am left with the feeling that “free will” (or mental causation) is an elaborate illusion.
    A fascinating talk. Many thanks.

    ReplyDelete
    Replies
    1. Thanks. Action selection is interesting philsophically because it is counterfactual. I did A but could have done B. Experience seems to be made at the point where the brain chooses between these alternatives, and I think that's a clue that experience may be part of a mechanism for improving the brain's choices.

      Delete
  2. Haggard mentioned the example of epilepsy patients feeling an urge for an action when certain cortical regions are stimulated during surgery. The patient reports her conscious experience, and she cannot see or sense what areas are being stimulated. If her report did not reflect a real urge (because she is actually a zombie, incapable of feeling this urge), then she would not respond in a reliable way to this stimulation. Imagine we had a standard map of where this urge could be elicited: could it be a new test for consciousness?

    ReplyDelete
    Replies
    1. A zombie could be program to do any reliable or unreliable answer to mimic humans, but this would not anything about its feeling.

      Delete
    2. But the zombie would still need some kind of input to know what behavior to mimic. If it cannot see the stimulation, and it cannot "feel" it (in a "touch" sensory receptor sense, since there are no sensory receptors in the cortex) then it could not be programmed to give the proper output behavior (reporting the conscious experience).

      Delete
  3. I want to understand if we can say precisely, what brain areas are responsible for agency. May be Dr Hallgard already discussed about this fact and I don't remember it.

    Actually, I try to understand that the agency are not an impairment in autistic persons because, some brain areas responsible with this functioning are touched.
    Thank you to help me to understand.
    (Camelia Dascalu, Paris 3 University)

    ReplyDelete
    Replies
    1. To answer your question, Patrick Haggard demonstrated that feeling of agency could be reduced by perturbing the activity of angular gyrus, which is part of the parietal cortex.
      I am not sure to understand your question or reflexion about autism.

      Delete
    2. Furthermore: we know that damage to the SMA also leads to perturbed sensations of agency, like the anarchic hand syndrome, where you move although you didnt want to. Also, regarding the sense of agenc the literature suggests involvement of the sma, premotor cortex, inferior parietal, super temporal sulcus, dorso-lateral prefromtal cortex, etc... Not to forget the cerebellum and basal ganglia. So, we can conclude that the brain is active during the sense of agency. A nice summary you can find by Nicole David et al, 2008 in Consciousness & cognition i think.
      Also, remember that the inferior parietal cortex might be rather signaling the "not - me"-ness of an action. There is also a study showing that the higher the temporal discrepancy is, the higher the activity of the inferior parietal cortex is.

      Delete
  4. sorry, I must clarify some statement: that the agency are an impairment because some brain areas are touched.
    thank you

    ReplyDelete
  5. When I was first reflecting on Dr. Haggard’s tall, I thought it was odd that when primed, the subject’s felt they were more in control of their decision, even though it seems they were influenced by the primed stimuli (less control). However, one question that was brought up at the end of the talk, was about that maybe because we have more information, we actually unconsciously take that information into account and that is why we feel more in control. Which view would you agree with? Are we actually taking more information into account that we are not aware of and then feel more in control, or is it more in the sense that there is less conflict in our decision and we feel more confident, therefore more in control? (perhaps a way to relate this to Dr. Morsella’s talk and conflict resolution in consciousness)

    Would there be any way to prime a robot in the same way that we can prime a human? Would the robot take this information into account? Could they tell us that they had not seen/been aware of the primed stimuli or would robots be aware of everything we showed them?

    Izabo Deschênes

    ReplyDelete
  6. The questioner pointed out that agency may increase as evidence increases: that's a clever point, and suggests that agency should = certainty. I'll need to think about how to test that experimentally. Experts often make decisions and select actions based on very little evidence, but they can feel a strong sense of control over the situation: one example would be quality control decisions on an assembly line. The thinnest suspicion of a tiny defect in the product does not stop you feeling in full control when you stop the line.

    ReplyDelete
  7. Looking at Patrick Haggard's talk in the context of pragmatic representations (Paul Cisek) brings forward the question of the relationship between sense of agency and reward/punishment. Does the outcome of ones actions influences the sense of agency? From an adaptive view it seems plausible that reward/punishment would enhance the sense of agency (maybe reward and punishment would influence sense of agency in different ways). In this note, does the motivational baseline of the organism bias the sense of agency? e.g. depressed individuals tend to feel less control over their environment.

    ReplyDelete
    Replies
    1. Ana, I think there is a difference between agency and sense of control of the environment. I can feel like I totally own certain behaviors and yet feel like I these behaviors don't allow my to control my environment. There may be a link between the two, but I still think they're dissociable concepts.

      Delete
  8. I don't know what it's like for others, but for me, self-inflicted pain is easier to endure than pain inflicted upon me by others. When I pinch myself, and I start off softly and squeeze progressively harder, the fact that I can hold on for longer than when my friend does the same exact thing to me - is this due to the sense of control I believe I have?

    ReplyDelete
    Replies
    1. What you are describing seems to be an "attenuation of self-produced tactile stimulation due to the sensory predictions made by an internal forward model of the motor system" http://journals.lww.com/neuroreport/Abstract/2000/08030/Why_can_t_you_tickle_yourself_.2.aspx
      I think that in the diagram that Haggard showed about sense of agency, sensory-motor contingency in action was *part* of the prospective/retrospective error computation that would result in the sense of agency. But I guess this is not the defining characteristic of sense of agency. One can imagine "sense of agency" without sensorial feedback, e.g. brain-computer-interfaces.

      Delete
  9. POSTED on FB

    ROXANE CAMPEAU
    How the "prospectivism" of Haggard does invalidate the "retrospectivism" of Dennett? To call the retrospectivism "confabulationism" is clearly a statement of invalidation, but I dont understand why can't we consider both like two sides of the same coin...?

    MARTHA SHIELL
    My impression of Dennetts retrospectivism was that it applied to our awareness of sensory experiences, while Haggard is talking about awareness of decision making. Are they the same thing?

    ROXANE CAMPEAU
    Retrospectivism and prospectivism = access to C?

    ANDY NDK
    Well Haggard makes the point that these judgments of agency or control are done on a prospective basis, so pre- feedback, which is a bit against the typical model that the match between the predictive component and the feedback arising from that action. However, i dont know much how Dennet framed it exactly.
    2 juillet, 11:27

    ReplyDelete
  10. POSTED ON FACEBOOK BY DIEGO MENDOZA-HALLIDAY :

    "ON VOLITION. Here's an explanation to why we seem to own our decisions more when we're subliminally primed towards one: Maybe when there are factors that bias our decision (be subliminal or consciously accessed), the decision becomes easier and thus we think we own it because of the feeling that there must be reasons why we're drawn to one choice more than the other. In contrast, if the decision is totally random because there is no factor that biases our decision, the choice becomes based on totally random processes in the brain, which feel like we don't own them. This resolves the apparent irony."

    ReplyDelete
    Replies
    1. LAURENCE DUMONT :
      "The impression of certainety surely contributes à great deal to agency in that sense"

      PAULINE CLAUDE :
      "I don't think the decision we make when there is no factor that biases them are based on random processes in the brain. Let's take aplysia to explain my point. Neuronal processes rely on mechanical principles that are the same in every single living species with neurons. As shown ans explain by Prof. Sossin in aplysia (and then in all other species with neurons), to be fired, neurons need a starting point, a stimuli that will start a series of biological reactions. The stimuli that will trigger choices for appearent random decisions might be unconscious, but it has to have one, otherwise neuronal firing will never occur."

      Delete
  11. This comment has been removed by the author.

    ReplyDelete
  12. Very interesting! Dr. Haggard mentioned at the end of his talk that schizophrenic patients were not ‘fooled’ by the subliminal priming in the ‘free choice’ part of the experiment (as opposed to normal subjects). In other words, their sense of agency was similar when their choice was incompatible with the prime and when it was compatible. I am not in psychology and I am not familiar with the literature on schizophrenia, but is there an obvious explanation as to why this would be the case? Or, if not, could results like this potentially help us understanding some of the mechanisms responsible for schizophrenia?

    ReplyDelete
  13. ON VOLITION. Here's an explanation to why we seem to own our decisions more when we're subliminally primed towards one: Maybe when there are factors that bias our decision (be subliminal or consciously accessed), the decision becomes easier and thus we think we own it because of the feeling that there must be reasons why we're drawn to one choice more than the other. In contrast, if the decision is totally random because there is no factor that biases our decision, the choice becomes based on totally random processes in the brain, which feel like we don't own them. This resolves the apparent irony.

    ReplyDelete
  14. I am wondering how agency of automatic behaviour such as habits is built... Would it be only a retrospective judgement, a generalization of intention across time? I guess it would differ a lot if it is a desirable habit (brushing your teeth after a meal) versus an undesirable one (biting your nails)

    ReplyDelete
  15. I am trying to disentangle volition and consciousness during sleep, and I am wondering if any parallels can be drawn between retrospective dream recall and the proposed role of the retrospective element of volition.

    ReplyDelete
  16. COPIED FROM FACEBOOK :

    Pier-Éric Chamberland:

    Isn't ironic that subliminal priming increases the experience of agency, since traditionnaly, priming has been used to influence subjects? I do understant that it is the subjective experience itself, as if the person already knew what to do


    ...Or do I confuse the feeling of agency with free will, the doing vs the willing?

    In Professor Haggard's experiment, concurently to intentional fluency, did the primed subject were less exhausted? Did the subjects do other tasks after the experiment where they had impeded performance?

    Or, did the not primed subjects felt the experiment was more effortful, compared to primed subjects?

    ReplyDelete