Chapter 5: The problem of free will and determinism
In this supplement we will consider two issues concerning how modern science impacts (or doesn’t impact) the traditional free will debate: one concerning whether quantum indeterminacy in physics creates room for free will and one concerning how certain experiments in brain science (the Libet experiments) purportedly undermine free will. In both cases we will see that there is no simple argument from science to either support the existence or denial of free will. Rather, the scientific results leave the problem of free will unsolved.
Quantum indeterminacy (QI) refers to a range of different phenomena in quantum physics (the physics that studies the subatomic level) where the physical facts themselves seem to be indeterminate. One of the most famous examples of QI is the Heisenberg Uncertainty Principle which says that it is not possible to know both the position and the velocity of subatomic particles (such as electrons). Importantly, this uncertainty is understood by physicists not as a lack of our ability to measure correctly, but as built into nature itself. This contrasts with the way that physicists such as Isaac Newton understood the world. According to the Newtonian understanding of the world, if we knew all of the forces on a particular object, we could predict exactly what would happen to that object—where it would go, for example. Consider a billiards ball knocking into another billiards ball on a pool table, for instance. The velocity of that second billiards ball and the exact trajectory that it would take could be calculated if we knew all of the initial conditions—for example, the force of the first ball hitting it (determined by its mass and velocity) and coefficient of friction of the felt of the pool table. In short, Newtonian physics is deterministic: if we knew all of the initial conditions of some system, we could predict everything that would happen in that system. This is exactly the idea behind the famous “Laplace’s Demon” thought experiment (which was referenced in “The Problem of Free Will and Determinism” chapter). In contrast, in quantum physics even if we knew all of the initial conditions, we would still not be able to predict the exact position and velocity of an electron because the indeterminacy is built into the structure of nature. That means that the indeterminacy is not merely epistemological (due to our lack of knowledge), but rather ontological (build into the nature of the physical system itself). And that, in turn, means that at the subatomic (quantum) level, reality is non-deterministic—that is, even given all the forces on an object, where exactly that object will end up is indeterminate.
How does QI bear on the problem of free will and determinism? Here is how some people have tried to do that. The idea is that if the universe isn’t deterministic (in the way that Newton thought that it was), then this leaves room for (libertarian) free will to exist. Recall that the libertarian requires that determinism be false in order for there to be free will. However, even if determinism is false, that is not sufficient to establish free will. Here’s the reason why. What is happening at the quantum level isn’t anything like what we are looking for in free will, rather it is something more like simple indeterminacy.
Indeterminacy is simply the denial of determinism. However, what we are looking for in free will is not simply indeterminism, but rather control over our actions. If I were to randomly (and without any control over the matter) simply start squawking like parrot in the middle of class, this wouldn’t be free will, even if that action were not determined by all of the preceding events going on in my brain. Rather, such a behavior would exemplify a terrifying lack of control over my behavior. Thus, indeterminism (which is all that quantum indeterminacy gets you) is not the same thing as free will. Free will requires something more—but what? Well, that isn’t a question that quantum physics is going to be able to answer for us, nor should that be its goal!
The Libet experiments
Whereas quantum indeterminacy is supposed to support the libertarian view of free will, the Libet experiments are supposed to undermine that view. The Libet experiments take their cue from a crucial distinction: the difference between it feeling like one has free will and actually having free will. It is possible that although at the conscious level we strongly feel like we have free will, that feeling is an illusion. This is what Daniel Libet’s experiments are often taken by determinists to show by hard determinists. Determinist’s don’t deny that their actions feel free; they just claim that that feeling is an illusion. In 1677 Baruch Spinoza said: “Men are mistaken in thinking themselves free; their opinion is made up of consciousness of their own actions, and ignorance of the causes by which they are determined.”
Here is how the Libet experiments work. There is a device called an electroencephalogram (EEG, for short) that measures electrical signals produced by the brain. As scientists have observed the brain’s electrical activity, they have noticed a consistent pattern which they call the “readiness potential” (or RP, for short). The RP is a rise in the electrical activity in the brain that always proceeds an action by an average of 550 milliseconds (ms). Libet brought participants into the lab and hooked them up to an EEG. He then had them watch a little clock whose hand moved around and around the face and then instructed them to flex their wrist and to note when they become aware of their conscious intention to flex their wrist. What Libet found was that participants became aware of their conscious intention to flex their wrist about 200 ms before they actually flexed their wrist. Thus the RP was reliably occurring about 300 ms before one reported any conscious intention to carry out an action. And that means that the RP itself cannot be identical to one’s conscious intention to carry out the action (in this case, the flexing of the wrist). But it also means that an unconscious brain event (that is, the RP) is involved in causing the flexing of the wrist (since the RP always precedes the flexing of the wrist).
That is what Libet found in his experiments. But how is it supposed to undermine the existence of free will? The assumption of the libertarian view of free will is that our actions our under our control. But since we cannot control something that we are not consciously aware of (such as unconscious brain events), it follows that if what is causing our actions is an unconscious brain event, then we are not really in control of our actions. Here’s that argument in standard form:
- An action is free only if its cause is under one’s conscious control
- The RP is not something under one’s conscious control.
- The RP is causing the flexing of the wrist.
- Therefore, the flexing of the wrist is not something under one’s control [from 2-3]
- Therefore, the flexing of the wrist is not a free action [from 1, 4]
Here is how Daniel Wegner interprets the results of the Libet experiments:
The RP could thus signal the occurrence of unconscious mental events that produce both the experience of wanting to move and the occurrence of actual movement. This possibility alerts us to the intriguing realization that conscious wanting…is a mental event that is caused by prior events. It seems that conscious wanting is not the beginning of the process of making voluntary movement but rather is one of the events in a cascade that eventually yields such movement (Wegner 2003, p. 55).
This is an interesting challenge to free will, seemingly grounded in good science. Does this show that humans don’t really have free will? The philosopher Al Mele has argued that it doesn’t. Consider first that if Libet’s experiment is supposed to show in general that humans don’t have free will, then the kinds of actions that participants undertake in his experiments would have to be representative of the kinds of actions human beings undertake in general. Are they? First of all, an action such as flexing one’s wrist is an extremely simple action—about as simple as you can get. In contrast, consider a decision one would make in solving a math problem or deciding whether or not one wants to continue in one’s current relationship with one’s partner, or planning a vacation for spring break. These kinds of actions obviously involve more complex decision-making and are much more likely to be influenced by our conscious thoughts than simple actions like flexing one’s wrist. Second, Libet’s participants were explicitly instructed to try to be as arbitrary as they could in choosing when to flex their wrists. Many of the actions that we would want to claim as our own (as our freely chosen actions) would be precisely those that are not randomly chosen. Rather, they would be one’s where our choices are informed by our reasons. So there are two good reasons for thinking that whatever it is that is occurring in the Libet experiments is not in general representative of all human actions. In other words, even if we grant that simple, arbitrary actions are not within our conscious control, it doesn’t follow that every human action is not within our conscious control.
But there are deeper challenges to the idea that the Libet experiments show that humans do not have free will. A common interpretation is that the RP is itself the decision to move one’s wrist that is being made unconsciously (since it precedes one’s conscious awareness). There’s a certain kind of sense to this since the electrical activity in the brain is hitting a peak at that point (that’s just what the RP is). However, it is important to understand that there is activity that is itself giving rise to the RP. Why not treat that prior activity as “the decision”? On the other hand, why not treat the activity after the peak as “the decision”? Further, why not think that the RP is just a precursor to the decision and the decision itself is actually being made around 200 ms before the action? The argument against free will seems to come from the idea that a conscious decision must not be caused by any other antecedents. However, although there are some who have held a view like this (for example,, Roderick Chisholm), it seems a really implausible assumption to saddle the defender of free will with. Instead, one could see the conscious decision as itself arising from earlier unconscious activity in the brain. How else could it possibly work? And just because the conscious decision is itself caused by previous brain events, that doesn’t mean that the conscious decision is causally impotent.
Libet himself envisioned this idea. The idea is that even if the RP were the unconscious decision being made, this still left room for a “conscious veto.” What he meant by this is that a decision could be overridden in the last second by a decision to not flex one’s wrist. So conscious decisions could still have some causal power even if we grant that the RP is the decision. Libet claimed to have found data to support the idea of a conscious veto by instructing participants to prepare to flex their wrists at a preset time (on the hand of a clock) but to then to veto the intention to flex their wrists. What Libet found was a similar kind of RP, but that flattened out about 200 ms before the preset time at which they were supposed to flex their wrists. Libet claimed that this flattening out of the RP was due to the conscious veto. However, one begins to wonder at this point whether the instructions to both intend to flex one’s wrist but then veto that intention is sufficiently confusing for participants to be puzzled as to what they are really doing. As Al Mele has noted, “how can a normal agent simultaneously be settled on [flexing their wrists at a preset time] and not [flexing their wrists at a preset time].”
Mele has also raised a simple but important criticism of Libet’s methodology which casts further doubt on the idea that the RP is the decision. Libet only collected data in cases in which participants actually flexed their wrist. So although it is true that every time there is a RP, the action occurs 550 ms after, we do not know whether there is an RP even if the action doesn’t occur (since when the action didn’t occur, no EEG data was collected/kept). Compare the following two statements:
Every time there is an RP, there is an action
Every time there is an action, there is an RP
For us to be able to conclude that the RP was the decision, we would have to be able to say that both were true. However, the data collected only allows us to verify the first statement, not the second statement. If there were an RP even when the action did not occur, then clearly we could not say that the RP was the decision. This is the kind of possibility that would be true if there was a conscious veto and as we have seen in the previous paragraph there is at least some evidence that this is true.
However, a deeper issue is whether it is correct to see the RP as “the decision.” One reason to think it isn’t is what I’ve suggested above: if what we have an evolving chain of causes in the brain, some conscious, some unconscious, then on what basis should we isolate some subset of those causes and call it The Decision? Rather, it seems more accurate to think of the whole chain, starting from the rise of the RP to the ultimately engagement of the motor neurons as the decision. Perhaps it is mistaken to think of decisions as kinds of things that could be isolated by milliseconds. Instead, perhaps the decision is something that unfolds over time. In that case, the conscious aspects of those brain states are also part of the decision and hence part of the cause of the action. These considerations point us in the direction of further empirical and philosophical questions about the nature of mental states and of mental causation. This is not surprising given that philosophical questions centrally involve the analysis of concepts.
- True or false: If quantum indeterminacy is true, then not everything is strictly determined.
- True or false: Quantum indeterminacy is sometimes taken to undermine the idea of libertarian free will.
- True or false: Libertarianism would be vindicated as long as determinism were shown to be false.
- True or false: Libet’s experiments are sometimes taken to show that determinism is true and that human lack libertarian free will.
- True or false: Libet found that the RP (that is, readiness potential—a brain event that reliably proceeds our actions) proceeded subject’s conscious awareness of their intention to act.
- True or false: The fact that RP is seen as causing our conscious intentions is thought to undermine the concept of libertarian free will.
- True or false: One criticism of the idea that the Libet experiments show that humans lack free will is that the kinds of actions used in the Libet experiments are simple rather than complex actions.
- True or false: Libet’s experiments were set up in order to be able to test the following claim: “Every time there is an action, there is an RP.”
- True or false: Libet’s experiment leaves room for what he calls a “conscious veto” and this leaves some room for free will.
- True or false: There is nothing problematic about identifying our decision to act with RP.
- There are other kinds of indeterminacy in quantum physics, including indeterminacies about energy/time, as well as indeterminacies regarding the components of spin angular momentum. Thanks to Anthony Kuchera for clarification on this point (personal communication). ↵
- You can find this line of argument in a number of places, including Robert Kane, A Contemporary Introduction to Free Will. Oxford University Press, 2005. ↵
- A good example of this strategy can be found in Daniel Wegner’s The Illusion of Conscious Will. MIT Press, 2003. ↵
- This passage is found in Spinoza’s Ethics, which can be found online here. ↵
- Here is how the experiment was practically carried out. An EMG (which measures electrical signals produced by muscles flexing) measured when participants flexed their wrists. When they did so, this also triggered a recording of the EEG. The muscle burst measured by the EMG created the point in time against which the previous events (verbal report of the conscious intention to flex was well as the RP) were mapped. ↵
- A great, accessible introduction to both the Libet experiments and Mele’s criticisms of it can be found in Mele’s short book, Free: Why Science Hasn’t Disproved Free Will. Oxford University Press, 2014. ↵
- The following two paragraphs are a rough summary of the criticism of Libet in Mele 2003. ↵
- Al Mele, Effective Intentions: The Power of Conscious Will. Oxford University Press, 2009, p. 52. ↵