Dissecting Intentional Control (or Free Will Redux)
by Dwayne Holmes
This year we received some good news from the field of neuroscience.
No, it’s not that scientists discovered we have free will (or a will at all). That’s something we don’t require advanced technology to deal with. Rather, a recent study has potentially freed neuroscience from further misuse, by undercutting one of the major talking points of the hard-determinist, anti-free will (AFW) crowd. 
Note I said talking point and not evidence. That’s because the data emerging from neuroscience never suggested (much less constituted evidence for) what had been claimed. Nevertheless, AFW advocates liked to argue that Libet or Libet-style experiments show that brains make decisions before “we” do, and so disprove personal agency. 
Libet was a researcher who found that when subjects were asked to move their hand and note when they made the decision to move it, a signal could be detected in the brain (correlated to the movement) prior to when they felt they had made the decision. Many follow-up studies tweaked Libet’s setup and came to much the same conclusion, some placing the correlated signal full seconds before subjects reported feeling (i.e. became conscious that) they made a decision.
Moving from those interesting findings about brain activity (related to simple motions) to declarations about personal autonomy raises many questions. [2,3] One of the more obvious problems is the AFW claim that the detectable signal (called a readiness potential or RP) actually constitutes the moment of decision that determines what a person can or will do. A recent study addressed this by moving beyond questions of awareness (when do we feel we make a decision?), to testing how long people continue to exert intentional control over an action, throughout the entirety of an action.
Put simply, if RPs are the “final decision” for an action, then a person should not be able to alter that action following an RP. Authors of the new study likened the RP (under the AFW account) to the first domino toppled in a chain of dominos. If there are no other brain mechanisms employed during decision making past the RP then all of the dominoes must fall.
In that case, a computer capable of identifying RPs in real time should be able to predict what the subject will do (akin to knowing what will happen to the dominos, once the first is seen falling), allowing it to produce an appropriate response before subjects are even aware they made a decision.
With this possibility in mind, scientists created a game that set subjects head-to-head against a machine, using a Brain-Computer-Interface that would allow the system to learn to “read their minds.” The goal for the subjects was to beat the machine by not doing something (in this case pressing a foot pedal) that the computer predicts they must do, based on their own brain activity.
If machines beat the subjects… well… to my mind that still wouldn’t prove that we lack free will (that brains decide before “we” do).  But AFWs would probably continue citing Libet-style experiments with the same hubris (perhaps more) as before, since there would be no direct, obvious counter-evidence to their position from those kinds of studies. 
If subjects managed to beat the machine, then other neural processes located downstream of the RP must underlie our capacity for decision making. Indeed they must in some way facilitate the tracking and evaluation of the potential act (detected RP), with regard to our intentions (beat the machine), and exercise an overriding control over those potential acts. The simplest conclusion coming from such a result being that conscious decision making (since “feeling” we decide comes later than the RP) involves vetoing potential acts, or as Libet argued we might have “free won’t.” 
Thankfully (keeping it simple), the humans beat the machines. There is more to our decision-making than what is currently caught in MRIs and EEGs, a fact that should not be surprising to any neuroscientist aware of the limits of such technologies.
The authors briefly mention their paper’s impact on free will arguments:
The possibility of a veto has played an important role in the debate about free will… which will not be discussed further here. Note that the original interpretation of the veto was dualistic, whereas in our case veto is meant akin to “cancellation.”
While not completely clear, the second sentence appears to be denying libertarian, contra-causal (“dualistic”) accounts of free will, while leaving open compatibilist accounts. That is to say if there is conscious/intentional control of our actions, it is not manifested by anything except the physical matter and mechanisms found within our central nervous system. However, it is possible they are trying to leave the door open to a semi-AFW position that the brain decides (something “you” have no control over), and all you can do is cancel incoming commands. Then again, that would be dualism of another sort.
Their concluding summary is interesting from a neurobiological standpoint, while simultaneously (even if unstated) sounding the death-knell for Libet as an AFW talking point:
To summarize, our results suggest that humans can still cancel or veto a movement even after onset of the RP. This is possible until a point of no return around 200 ms before movement onset. However, even after the onset of the movement, it is possible to alter and cancel the movement as it unfolds.
That last sentence holds extra value. Even if we reach a point where a movement must begin (the message has been sent to muscle fibers), we still maintain some intentional control over how it will end. As damaging as that is for AFW claims that “we” don’t control our basic muscle movements, consider how devastating it is for larger time-scale “life” decisions, where discussions of intent and free will matter the most.
Intriguingly, the paper came from a lab whose lead investigator was presumably in the AFW camp. Not only was his prior work cited by AFWs, but he delivered a lecture at my own graduate school on how his work (and other studies) prove determinism. He started that talk by saying he wouldn’t discuss free will implications, yet ended by brow-beating philosophers (yes, he specifically said “philosophers” though we were all neuroscientists) for not getting with the program on hard-determinism.
It is hard for me to understand how any scientist thought the RPs detected by Libet offered an answer to the question of free will, much less that the answer was that it didn’t exist. And it is a shame that some scientists chose to promote such a lazy interpretation of the data, particularly to the public.
At least, it would seem, neuroscientists themselves are beginning to clean up that mess.
1) I admit, my statement is an exaggeration. It is likely people will continue to misinterpret and mischaracterize evidence from neuroscience. However, it is true that the study under discussion in this essay should reduce misuse by neuroscientists on topics related to free will (article on study: http://neurosciencenews.com/decision-making-eeg-free-will-3333/ and the paper itself: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4743787/). Of course there may be holdouts, provocateurs, pretenders, or counter-studies which call into question the the results of this paper (and therefore…).
2) There is a nice animated overview of Libet’s experiment and differing interpretations of the results at: https://www.youtube.com/watch?v=OjCt-L0Ph5o as well as Wikipedia’s overview: https://en.wikipedia.org/wiki/Neuroscience_of_free_will#The_Libet_experiment
3) I discussed problems with standard AFW interpretations of Libet-style studies, from a neuroscience perspective, in an earlier essay at Scientia Salon: https://scientiasalon.wordpress.com/2015/01/26/choosing-a-compatibilist-free-will-perspective/
4) Well now here I was just plain lying. Libet argued from the start that conscious decisions may involve canceling whatever unconscious drive the RP might represent (see #2). There have been studies since then suggesting we possess an ability to cancel actions linked to an observed RP, and even parsed out the timing (reportedly up to 150-200ms before an action begins). The current study cites those earlier papers for precedence, so it’s not like we lacked evidence against AFW interpretations.
The nice thing about this latest study is that it does a better job dissecting intentional actions from the emergence of an RP through the entire course of activity (body and brain) until the action is complete. This allowed researchers to discover more than one kind of “decision” a person can make with regard to controlling a potential action (and their windows of opportunity). It is also important that the study, as I discuss within the essay, comes from a group that previously appeared to support AFW interpretations.
5) Of course I am not happy with the “simplest conclusion”. Just because a signal related to an action occurs prior to our conscious observation of making a decision, does not mean conscious decisions are limited to vetoes of unconscious “decisions” or “initiations”.
One alternative is that there are many signals that we just do not see with the technology we have. Perhaps we cannot detect them (out of range), parse them out (too weak/mixed with others), or they never get associated to anything they might have caused, for the simple fact they were not chosen. It is possible that out of all the choices of action being considered (whether producing RPs or not), the one with the greatest signal merely indicates an increased relative “attention” toward that action or increased preparation/anticipation to do it. After all, subjects know beforehand the single action they are meant to engage in and so it will be on their mind more than anything else, prior to deciding to go ahead with it. In this scenario the brain processes a decision/choice/selection between actions (rather than simply vetoing one) well after attention, preparation, and anticipation have been paid to all possible actions.
It is also possible that RPs indicate a mounting degree of motivation/inclination toward taking a certain action, rather than being part of the core mechanisms/brain activity underlying decision making. To use a loose analogy, while the carrot may always precede the horse, it is merely a motivating factor for the horse (to choose to move), it is not the decision itself nor an integral piece of the decision making apparatus.
And there could be more… like if you challenge their idea of how consciousness/awareness operates: http://www.sciencedirect.com/science/article/pii/S1053810013001633