Feelings of control: Contingency determines experience of action

Abstract

IntroductionThe evolution of human intelligence has allowed not just deeper understanding of the world, but a greater capacity to act on it. Such operant actions imply the ability to know that one is performing an action, and to represent its consequences (Dickinson & Balleine, 2000). Research on the epistemic content and conscious experience of action has identified two distinct processes underlying this ability.According to ideomotor theories (e.g. James, 1890), actions are internally represented by reference to their external consequences. On this view, making an operant action involves a prediction of the action goal, an idea supported by recent models of 3 computational motor control (Blakemore, Wolpert, & Frith, 2002;Berti et al, 2006). Alternatively, the conscious experience of operant action may be inferred from sensory evidence (Wegner, 2002). In particular, spatial and temporal correlations between thoughts, physical movement, and external events may lead to us to infer that we have caused an external event.Voluntary actions have strong effects on the subjective passage of time (Haggard, Clark, & Kalogeras, 2002). Temporal effects provide a common measure allowing the predictive and inferential contributions to experience of action to be compared directly.When a voluntary action, but not an involuntary movement, is followed by an external event, people perceive the action as shifted in time towards its effect, and the effect as shifted earlier in time towards the action that caused it (Haggard, Clark, & Kalogeras, 2002). This 'intentional binding' involves a predictive element, because omitting the effect does not prevent the shift in perceived time of action, as long as the probability of an effect given an action is sufficiently high (Moore & Haggard, 2008). It also involves an inferential "postdictive" element, because the tone's occurrence shifts the perceived time of action, even when tone probability is low (Moore & Haggard, 2008).This suggests that the human mind builds internal models of action-effect relations, which determine the experience of action. Here, we investigate whether rules thought to govern causal learning in animals might also underlie the experience of agency in humans. Contingency is an index of the causal relation between events, and predicts patterns of operant learning in animals (Hammond, 1980). Contingency is defined here 4 as the probability of an effect (E) given an action (A), minus the probability of E in the absence of A. This index is known as ∆p (Allan, 1980):Where ∆p > 0 the effect is more probable in the presence than in the absence of action.Conversely, where ∆p < 0 the effect is less probable in the presence than in the absence ProcedureParticipants chose on each trial whether or not to press a key with their right index finger.The element of choice was included because computing ∆p requires trials with and without actions. They viewed a rotating clock hand (period 2560 ms, see Libet, Gleason, Wright, & Pearl, 1983), and used this to judge the time of their ac...

Temporal binding and internal clocks: No evidence for general pacemaker slowing.

“…However, considering Wenke and Haggard's (2009) design reveals that we have to be careful in interpreting their results. It is possible that the higher discrimination thresholds they found in causal trials simply reflect a transient, effector-specific modulation of the pacemaker induced by motor action.…”

“…Given that previous research has shown that temporal binding is rooted in causality (Buehner, 2015(Buehner, , 2012Moore, Lagnado, Deal, & Haggard, 2009), occurs across various stimulus modalities and tasks (Haering & Kiesel, 2014, 2015, and does not merely reflect processes related to motor preparation and execution, we focused our efforts on a simple causal versus non-causal interval distinction: In the experiments reported here, causal intervals were delineated by a participant's key press, which always caused a sensory outcome after a delay, while noncausal intervals were delineated by two sequential sensory outcomes. This manner of operationalizing causal and noncausal intervals presented the most efficient method of manipulating causality.…”

“…Wenke and Haggard (2009) combined the temporal binding procedure with a temporal discrimination paradigm: In a causal condition, participants pressed a key to generate a tone after 600, 800 or 1000 ms; in a corresponding non-causal condition, their finger was passively pulled down by a motor, which was followed by a tone -the computer scheduled both events, making it obvious that there was no causal connection between the passive movement and the tone. In addition, participants experienced two electric shocks administered via electrodes on their fingers.…”

Intentional binding and the sense of agency: A review

“…By computing contingency information, organisms are able to purposefully control their environment. Moore, Lagnado, Deal and Haggard (2009) investigated the relevance of contingency learning for intentional binding, by measuring its impact on the action component of intentional binding using the Libet clock method. They used the same probability manipulation as Moore and Haggard (2008) described above, with 75% and 50% outcome probability conditions.…”

On the functional anatomy of the urge-for-action

“…Interestingly, both of these disorders have been linked to damage of the anterior insular cortex of the right hemisphere (Baier & Karnath, 2008;Karnath, Baier, & Nagele, 2005). However, it is important to note that this sense of agency may in fact have a significant postdictive or reconstructive component (Moore, Lagnado, Deal, & Haggard, 2009;Wegner, 2002) and thus is not necessarily an unambiguous index of intentionality. …”