Free Will Flashcards
- Definition of sense of agency, central and peripheral experiences
= experience of controlling one’s own motor acts and through them the course of external events (i.e. the feeling occurring before, during and after actual muscular movement)
Central experiences: cognitive aspects, like experience to intend to act, choose or initiate an action
Peripheral experiences: conveyed by the somatosensory receptors in the body
Involuntary movements (like reflexes) possess the peripheral, but not the central aspect, and are not accompanied by sense of agency (but sense of ownership)
Sense of agency involves both ownership and feeling of being the source of the action
- Definition of sense of ownership
= Sense of mineness, feeling that a body part or mental state is specifically linked to oneself
- Explicit measure of sense of agency
Explicitly asking, “Did you do that?”
• Consistent bias: tendency to overestimate one’s own agency
• This bias is stronger for positive outcomes
• Social judgement (was it me or someone else?)
- Intentional binding paradigm (conditions involved to measure the sense of agency implicitly)
Implicit measures
• Advantage: measure without requiring people to explicitly think about agency avoid biases and desirability effects
• Tend to be only weakly correlated with explicit measures
• Less social, more “me vs. the world” distinction
Intentional binding paradigm
• Report the perceived time of a voluntary action or a subsequent sensory event
• Voluntary, but not involuntary actions are perceived as shifted towards the outcomes, while outcomes seem shifted towards the voluntary action
• sense of agency can be quantified as perceived compression of time between action and outcome
- The comparator model
Models of motor control:
• efference copy of motor command is passed to internal predictive model
• prediction of bodily states is compared to actual sensory feedback
• difference is the prediction error
Same principle applicable to sense of agency:
• if an action is self-generated, the prediction error should be zero*
• a non-zero prediction error is evidence for a non-self-generated action
• *Given the internal predictive model is correct
People have a sense of agency over events that can be predicted given their motor commands
Evaluation of the comparator model
• Can explain the phenomena of “non-agency” quite nicely, as a mismatch signals there is something going wrong (e.g. pressing the light switch, but the light does not turn on)
• However, cannot explain the normal “buzz” of agency as nicely, as no mismatch does not generate any neural signal – so where does the feeling of agency come from?
• Imply attenuation or suppression of our action outcomes we only perceive what we cannot predict
o This is functional as it prevents perceptual overload
o However, it’s quite counterintuitive to agency idea to suppress the very goal of our actions (there would never any feeling of successfully acting)
• May apply primarily to immediate sensorimotor effects of actions and cannot be the only process involved in agency
- Prospective and retrospective agency
Prospective: choosing an action
• Must be prospective, because action selection precedes the action
• DLPFC thought to be involved in assembling a response space
• Process of selecting between alternatives in the frontal lobe contributes to sense of agency
• Possible explanations how choosing increases sense of agency:
o Increasing number of options increases likelihood that the outcome will match the desired outcome
o Action selection itself boosts the sense of control over the outcome
o Action selection produces a metacognitive signal reflecting the level of conflict related to the choice
Retrospective: Sensory feedback
• Comparator: sense of agency can be computed only after a (delayed) feedback reached the comparator
• Being primed with the output of an action increases the sense of agency
Disentangling prospective and retrospective agency in intentional binding
• Varying the probability that an action produces a tone
• binding effect is stronger, if a tone is actually presented tone retrospectively altered perception of the action (‘postdiction’)
• If there is no tone, the binding is stronger if the general probability for tones is higher prospective component
• sense of agency is linked to both prediction of action outcomes and retrospectively inferring agency once the outcomes are known
- Disorders of the sense of agency
Hyper- vs hypoagentic disorders
Depression
• Feelings of less control over outcomes (Hypoagentic bias)
• One view proposes this difference is due to hyperagentic bias in healthy individuals (patients being more realistic)
Schizophrenia
• Feelings that own thoughts and actions are caused by external actions
• One explanation: failure in prediction of own action outcomes no attenuation
• Less able to detect when visual feedback of their actions has been modified
• Lack sensory attenuation of self-generated action consequences
• Intentional binding: patients experience stronger intentional binding than healthy controls
o Contradiction? No – this effect is only present when a tone is actually presented
o In healthy controls, the effect depends on the mere probability
o schizophrenic patients base their sense of agency only on retrospective attribution, prospective component of healthy individuals is missing
- Implications of sense of agency for society
Idea of responsibility for own actions is strongly rooted in our societies, is the basis for punishment and reward
Law
• Responsibility for a criminal act should involve the experience of a sense of agency
o Actions that one consciously decided to perform with a reasonable amount of understanding of the outcomes
• In some cases, diminished responsibility is acknowledged in court
o Self-defence, persistent abuse neurobiological control or just social acceptance of extreme actions in some cases? Combination of both
Obeying orders – coercion reduces sense of agency
• Study: participants delivered shocks to one another, while estimating time interval between action and shock (paired with a tone)
• implicit measure of agency: time interval experienced longer when experimenter gave coercive instructions
• Coercion also resulted in lower ERP responses to the tone
• Consistent with observation that number of action alternatives increases sense of agency coercion constrains free choice
Health
• Strong sense of agency at work is a major determinant of health
• Loss of sense of agency through disorders impairs quality of life
