Nick Griffin

and 2 more

When placed in front of a button and instructed to press it only when they “feel the urge” to do so, humans typically make a “spontaneous” decision to act at seemingly random times, usually between 5 and 30sec from trial start. In such a situation, how does our brain make a decision that the time to act has come? To investigate this question, we used a 6-area brain-constrained neural-network model that accounts for neurophysiological and anatomical features of relevant areas of the cortex.To replicate the experimental settings under which volitional button-presses are performed, we analysed the network activity in absence of any “sensory” input: we observed spontaneous ignitions of previously-learned “perception-action” circuits (linking “visual” information to corresponding “hand/finger movement”) caused by the accumulation of noise within them, which we took as model correlates of self-generated decisions to act. The time to first spontaneous circuit ignition was then used to build a simulated waiting-time histogram.We found that, for select values of the parameters, the network autonomously “decided” to act (or to wait) in a way that accurately reflected participant behaviour under such conditions. Additionally, we found that when the model’s macroscopic behaviour matched that of participants, its microscopic activity prior to a spontaneous ignition event also reproduced the pre-movement slow ramping neural signal (readiness potential, RP) experimentally recorded using electroencephalography.Reproducing both behavioural and brain indexes of spontaneous voluntary movements, the present brain-constrained architecture offers a neuro-mechanistic explanation for the emergence of endogenous action decisions in the human brain in the form of memory traces’ spontaneous ignitions. Our results also speak to the ongoing debate over the interpretation of the RP, suggesting that, rather than reflecting motor preparation (a decision outcome), the emerging RP is pre-decisional, with overt movement being the result of a distributed perception-action circuit spontaneous ignition.