Aaron Schurger earned his neuroscience Ph.D. from Princeton, working in Jonathan Cohen's Neuroscience of Cognitive Control Laboratory (NCC Lab). He is currently in France at the Cognitive Neuroimaging Unit of the Institut National de la Santé et de la Recherche Médicale (INSERM). Cohen and Joshua Greene (a former NCC Lab researcher and now professor of psychology at Harvard University) published an important paper challenging conscious will, with the evidence based primarily on Benjamin Libet's famous experiments. Now Schurger and his French colleagues, Jacobo D. Sitt and Stanislas Dehaene, have produced significant evidence that the readiness potential (RP) seen before voluntary self-initiated movements is not necessarily causal and determinative of the action. In their 2012 article in the Proceedings of the National Academy of Sciences, they say:
The premovement buildup of neuronal activity apparent in the RP and the assumption of causality invested in it have become a cornerstone in the study of volition. Notably, Benjamin Libet (9, 15) tried to measure the temporal relationship between the onset of the RP and the feeling of an “urge” to move. The results of Libet et al.’s experiments (9, 15) suggested that the objective neural events in the brain that cause movement precede the urge to move by 300 ms or more. A recent experiment, using Libet’s paradigm, confirms the same preurge buildup at the single-neuron level (16). Such demonstrations have had an unrivaled influence on the prevailing view that movement is initiated preconsciously and the feeling of intending to move is grafted on after the fact. In fact a gradual increase in neural activity preceding spontaneous movements appears to be a very general phenomenon, common to both vertebrates (8, 16, 17) and invertebrates (18) alike. Why do both humans and crayfish (18) exhibit the same 1- to 2-s buildup of neural activity in advance of self-initiated movements? Kornhuber and Deecke’s (12) interpretation of the RP as a sign of planning and preparation for movement fails to explain what specific neural operations underlie the spontaneous self-initiation of movement and why these operations are reflected in the specific exponential shape of the RP.Schurger et al. find that the shape of the readiness potential can be explained if the brain uses a common machinery for decision making, specifically a threshold applied to the output of a stochastic neural accumulator. they say:
Decision-making tasks are typically modeled in terms of the accumulation of evidence. What is unique to the spontaneous-movement task is that subjects are specifically instructed to not base their movement decisions on any specific evidence, sensory or otherwise. One simple solution, given these instructions, is to apply the same accumulator-plus-threshold decision mechanism, but fed solely with internal physiological noise.Alfred Mele is a powerful critic of interpretations of the Libet experiments which claim our decisions are made in our brain long before our conscious awareness of the decisions. Mele criticizes the interpretation of the Libet results on two grounds. First, the mere appearance of the RP a half-second or more before the action in no way makes the RP the cause of the action. It may simply mark the beginning of forming an intention to act. In the two-stage model, the rise of the RP might simply reflect the considering of possible options. Secondly, Libet himself argued that there is enough time after the W moment (a window of opportunity) to veto the action, but Mele's second criticism points out that such examples of "free won't" would not be captured in the classic Libet experiments, because the recording device is triggered by the action (typically flicking the wrist) itself. Thus, although all Libet experiments ended with the wrist flicking, we are not justified in assuming that the rise of the RP (well before the moment of conscious will) is a cause of the wrist flicking. Libet knew that there were very likely other times when the RP rose, but which did not lead to a flick of the wrist, so his experiment could not detect them. Schurger et al. note this backward selection bias, that only epochs ending with an actual movement are subject to analysis (ibid, p.6). Their Libetus interruptus is exploring those time intervals when the RP might rise, their accumulator model might get to, or even surpass?, the threshold, and yet there might be no wrist flick. They offer a new model for what the RP represents, beyond the vague phrases of the past four decades of research, that it reflects "planning and preparation for movement." Their model for the RP is divided into two nonlinear components: an early precommitment phase (or stage in our terminology) dominated by stochastic fluctuations (with an evolving spatial distribution) and a late postcommitment motor execution phase (the last 150 ms). Schurger et al. challenge the notion that the early buildup of activity biases supposedly "voluntary" decisions (as argued by Soon et al., among others) They say that their model is consistent with such predecision biases, but suggests that they may reflect stochastic fluctuations rather than an intentional (preconscious) decision process:
It is widely assumed that the neural decision to move coincides with the onset of the RP (which, given its slow nonlinear character, is difficult to pinpoint) (11). Our model challenges that assumption by suggesting that the “neural decision to move now” might come very late in the time course of the RP. Prior research shows an involvement of motor areas, including primary motor cortex, in motor imagery, in the absence of overt movement (40). Thus, movement-specific activity in motor cortex, even primary motor cortex, although it might vary with the probability that a movement will occur, does not necessarily signal the final commitment to produce a movement now...Thus, according to our model, uncued movements in a task like Libet’s tend to be preceded by a gradual increase in neural activity [measured at the scalp (8, 9) or the single-neuron level (16)] whose causal role is incidental—not directed (consciously or otherwise) at producing a movement."Finally," say Schurger et al., although their "model is silent with respect to the urge to move and its temporal relation to motor decisions, it helps dissolve another puzzling question that seemed to arise from Libet’s paradigm. Libet himself found that subjects were able to estimate the time of a tactile sensory decision in relation to a quickly rotating clock dial with only about 50 ms of error on average (9). Why then should there be such a long and variable gap between the time of a motor decision and the subjective estimate of the time of the motor decision, whereas no such gap exists for sensory decisions? In fact, this question arises only when we assume that the motor decision coincides in time with the onset of the RP. We have argued that this need not be the case and that the neural decision to move may come much closer in time to the movement itself (e.g., −150 ms). We propose that the neural decision to move coincides in time with average subjective estimates of the time of awareness of intention to move (9, 11) and that the brain produces a reasonably accurate estimate of the time of its movement-causing decision events." (ibid, p.7) Thinkers (e.g., Daniel Wegner, Patrick Haggard) who claim that the Libet experiments prove that our conscious will and subsequent actions are caused by prior neural activity - this is the popular view that "my neurons made me do it" - are simply wrong. Note that the abrupt and rapid decisions to flex a finger measured by Libet bear little resemblance to the kinds of two-stage deliberate decisions for which we can first freely generate alternative possibilities for action, then evaluate and select (in an adequately causal way) which is the best of these possibilities in the light of our reasons, motives, and desires - first "free," then "will."