In contrast, the signal-in-noise view suggests AG-014699 ic50 that experience of volition occurs when an internal signal exceeds a criterion value, or crosses a threshold. Patients with GTS vary in the level of motor noise associated with tics, and also in the perceptual awareness and intentional controllability surrounding their tics. Our results show that these latter factors strongly influence the experience of volition in GTS. Therefore, patients with GTS may face a greater difficulty than controls in the crucial perceptual computation to

separate one’s own volitional actions from other movements. Could a retrospective, inferential account of intention also explain the results in GTS patients? Retrospective accounts would suggest that experiences of volition are inserted post hoc, whenever a patient moves. In GTS, this process would occur both after voluntary actions, and also after tics. This retrospective insertion might potentially explain some premonitory urges – although many urges build up over a much longer timescale than the subsecond timescales associated with retrospective insertion of

intentions.Crucially, however, a retrospective account of GTS action awareness would suggest that a patient who strongly reconstructs PD-166866 order urges should also strongly reconstruct intentions. In our dataset, high PUTS scores should then be associated with early W judgements. In fact, we found a strong effect in the opposite direction. Therefore, our results seem more consistent with the idea of perceptual learning of a premotor signal, rather than a general inferential mechanism

for retrospective insertion of intentions. A recent computational model rejected cAMP the notion of volition as a hierarchical top-down control of the motor system, and suggested instead that random fluctuations of a motor readiness signal could be sufficient to explain the initiation of voluntary actions (Schurger, Sitt, & Dehaene, 2012). Our result is consistent with the view that people also experience an intention to act when an internal signal exceeds an individual’s threshold level ( Hallett, 2007). The choice of threshold leads to a relation between the average time of conscious intention, and its trial-to-trial variability. We verified this prediction in both GTS and the control group. Setting a suitable threshold level for the neural signals that produce the thin and ambiguous experience of volition is a perceptual challenge. Setting a low threshold will regularly produce false positives. These individuals would show early detection of intention on average, but their judgements would be highly susceptible to motor noise. In contrast, an individual who chooses a high threshold would be less susceptible to noise. However, the high threshold would be crossed only late in the motor preparation sequence, leading to a delayed experience of volition. We show that this idiosyncratic variation exists in the general population, as well as in GTS.