Some theorists have argued that our knowledge of the brain will one day advance to the point where the perfect neuroscientific prediction of all human choices is theoretically possible. Whether or not such prediction ever becomes a reality, this possibility raises an interesting philosophical question: Would such perfect neuroscientific prediction be compatible with the existence of free will? Philosophers have long debated such questions. The historical debate between compatibilists and incompatibilists, for example, has centered on whether determinism and free will can be reconciled. Determinism is the thesis that every event or action, including human action, is the inevitable result of preceding events and actions and the laws of nature. The question of perfect neuro-prediction is just a more recent expression of this much older debate. While philosophers have their arguments for the compatibility or incompatibility of free will and determinism (or perfect neuroscientific prediction), they also often claim that their intuitions are in general agreement with commonsense judgments. To know whether this is true, however, we first need to know what ordinary folk think about these matters. Fortunately, recent research in psychology and experimental philosophy has begun to shed some light on this.
On the one hand, a growing body of empirical evidence indicates that people think that their choices aren’t determined (see here, here, here, here, and here). It is not just that they don’t have the belief that their choices are determined. Rather, they positively think that their choices are not determined. And this belief is implicated in their thoughts about free will. For instance, in recent studies, when presented with a description of a determinist universe, most participants say that in that universe, people don’t have free will. This provides some reason to think that the everyday conception of free will is not compatible with determinism or perfect neuroscientific prediction.
On the other hand, researchers have also found that at least in certain scenarios people exhibit compatibilist intuitions about free will (see here and here). In a recent study by Eddy Nahmias, Jason Shepard, and Shane Reuter, for instance, researchers presented participants with scenarios involving neuroscientists than can “predict with 100% accuracy every single decision a person will make” and where “everything that any human thinks or does could be predicted ahead of time based on their earlier brain activity.” They found that people overwhelmingly attribute free will to agents in these contexts even when their behavior is predicted by neuroscience with absolute certainty. Nahmias and his colleagues interpret such ascriptions as evidence for the compatibilist view that there is no inherent conflict between a perfectly predictive neuroscience and the common notion of free will.
Is there any way to reconcile these conflicting results? Is there any way to settle whether ordinary folk have compatibilist or incompatibilist intuitions when it comes to the possibility of neuroscientific prediction? A recent set of studies published in Cognitive Science by David Rose, Wesley Buckwalter, and Shaun Nichols suggests an intriguing resolution. In a series of six experiments, the authors presented participants with a number of different neuroscientific scenarios and found evidence that participants fill in the scenarios in ways that undermine the inference that the ordinary notion of free will is compatible with the idea of perfect neural prediction. To begin, the authors write:
It’s useful to distinguish two kinds of filling in, what we’ll call ‘importing’ and ‘intruding.’ Importing occurs when participants fill in the scenario in ways that are consistent with the scenario, but the filling-in systematically goes beyond the information provided in the scenario. Of course, when participants read vignettes, importing will be a common occurrence. It becomes theoretically interesting when the imported information undermines the interpretation of the results. Intruding occurs when the filling in leads to misrepresentation of the scenario.
Rose and his colleagues hypothesized that if people fill in scenarios depicting perfect neuro-predication based on prior beliefs about indeterminist free will, then this could actually lead them to reject the notion that decisions are perfectly predictable in the way specified in the scenario. In their experiments they tested this hypothesis and found evidence for both importing and intruding. As they describe their findings:
Experiment 1 demonstrates that those who affirm free will in Nahmias et al.’s neuro-prediction case do so while imposing indeterminist details contrary to those stated in the story. This indicates that people misrepresent instances of perfect neuro-prediction and thus that intrusion has occurred. Experiment 2 replicates this intrusion effect and demonstrates that it plays a mediating role in the comprehension of neuro-prediction stories. Experiments 3 and 4 find the same kind of intrusion effect in a different narrative context utilizing simplified cases with more minimally matched pairs. Experiment 5 again finds intrusion effects and demonstrates that they persist across different ways of probing participants that make the predictive nature of brain activity highly salient. Experiment 6 demonstrates that filling in also occurs as a result of importing an indeterminist view of choice when presented with an adapted case of perfect neuro-prediction.
These findings suggest that what’s really taking place in the Nahmias et al. study is that “people are imposing an indeterminist notion of free will onto the situation, despite the fact that the situation is explicitly described in terms of perfect predictability.” In light of these findings, Rose and his colleagues conclude “we doubt that people are broadly comfortable with the idea of perfect neuro-prediction or that it is fully compatible with commonsense notions of free will.”
These results are a welcome addition to the growing empirical literature on the folk psychology of free will. Not only do they provide a plausible reconciliation of seemingly conflicting results, they also highlight an important methodological point: when testing folk psychological intuitions about free will, it will be important moving forward to rule out the possibility that participants are filling in—i.e., that their intuitions about indeterminism and free will are not intruding into their representations of the various scenarios and vignettes. When people provide answers to philosophical questions about free will, we need to keep in mind that they also have metaphysical commitments that often intrude on their reasoning.