I find it fascinating that the field of physics has yet to decide on this matter, the existence of free will, that is. Of all the concepts worthy of study and experimentation, you would think that a demonstration or falsification of free will would be at the top of anyone’s list. But no. Relatively little is being written about free will in the physics community. Since January 2010, only 20 papers having the phrase “free will” in the title have been published in the online physics journal arxiv.org. Compare this to 143 papers with “superluminal” (faster than light) in the title and 663 papers on “dark energy” in the same time frame. Clearly, free will is being neglected by the scientificats in favor of more newsworthy and otherworldly research. I don’t understand. It’s not like the issue has been settled, or that it’s beyond the realm of physics to settle it.
(Note: I do not claim that no one is studying free will. As one example, the philosophy department at Florida State University is engaged in a four-year project to “improve our understanding of free will in … science, philosophy, and theology.” This effort is being funded by a grant from the John Templeton Foundation, a matter of some controversy among scientists. The Foundation has been accused of advancing an “agenda of reconciling religion and science” and “persistently seek[ing] to muddy the waters and keep religion credible in lay eyes.” Hopefully, results from the free will project will be free of such taint.)
The Meaning of Free Will
Before I go on, I should explain what I think “free will” means, because different persons have different concepts. For me, free will is the ability to select one outcome or another, intentionally (not randomly), when presented with a given set of conditions at a particular instant in time.
To elaborate: let’s say you walk into your friendly tavern in Pittsburgh, Pennsylvania, on Sunday at 4:27 pm (after the Steelers game) and you decide to order an Iron City. Now let’s also say that I am God (why the hell not!) and I can rewind the universe, so you are presented with exactly the same conditions at precisely the same moment with exactly the same mindset as when you first walked in. If you were now to decide Fat Tire instead of Iron City, this might be indicative of free will (as well as good taste).
The question is, do we live in a universe where one is able to choose Iron City vs Fat Tire? Or does the physical state of the universe (including your state of mind) at that particular moment in time determine your choice, such that you could not have chosen otherwise?
Based on the explanatory and predictive power of physics, I have concluded that under most circumstances you could not have chosen otherwise. (Sorry, Fat Tire drinkers.) Physical states and physical processes lead to physical results. We are physical beings subject to physical laws. It is unclear to me how a person would be able to “step outside” physics and make extra-physical choices.
But that is more assertion than argument. So let me introduce the principle known as Occams’ Razor: the simplest explanation is most often the best. As such, the burden of proof should be on those who believe extra-physical processes are necessary to explain human decision-making. The cleanest and most straightforward approach is to view our brains as highly-evolved bioelectrochemical evaluation units. Computers, that is.
Figure 1 (click to enlarge) shows the simplicity of this model. The current state of the universe (at least the part of it that you can observe) is filtered by your sense organs and input to your evaluation unit. Your state of mind at that moment — all the synaptic wiring, hormones and neurotransmitters that encode your memories and emotions — is the other input to the system. Your evaluation unit processes these inputs and eventually produces an output, which I call your decision.
You are not necessarily aware of all the input. In fact, you are hardly aware of any of the input. If awareness were necessary, every choice would turn into a deliberation and your decision-making would be much slower than it is.
But maybe your free will does its work inside the thing I call your evaluation unit. So let’s take a closer look at it (Figure 2, click to enlarge). The first part of any rational decision-making algorithm would consist of some type of situational assessment. What kind of decision is this? How urgent is it? What needs will it satisfy? Based on this determination, the software in your brain goes to work. First, it adjusts the weights of the factors to be considered — for example, if you are deciding what to wear for a funeral, the factor “makes me feel sexy” would be given a low weight. Your brain then sorts the factors by weight so that only the most important factors rise to the top for further consideration.
With the factors and weights in hand, another circuit carries out the task of generating a list of relevant alternatives. Each alternative is scored based on how well it is expected to satisfy the important factors. Will this be boring or exciting? Is this illegal? Will it get my clothes dirty? The score for every factor is multiplied by its weight, and all the results are added up to produce a grand total for that alternative.
When all the alternatives have been scored, your evaluation unit picks the one with the highest score and outputs it as your decision. The result may also be reported to your “awareness center” at this point so that you may become conscious of it.
(Note: studies have demonstrated that brain activity prior to “willful actions” — such as the decision to lift one’s finger — precedes one’s conscious awareness of the decision to act by a few tenths of a second. If what we typically call free will operates outside of our conscious control, it hardly deserves the name.)
The actual sequence of steps in the decision-making process may differ from what I have described — for example, some steps may be performed at the same time as other steps — and it is likely your evaluation unit does not do addition and multiplication per se in order to arrive at the final result. But my point is that this represents a plausible brain-based algorithmic decision-making process. Where is free will in this process? Where is free will necessary in this process?
Veto Corleone! (Does My Brain Have a Godfather?)
Some (e.g. Libet, 2003) argue that free will is a kind of rubber-stamp, allowing you to consciously approve or veto the tentative decisions of your subconscious (see Figure 3, click to enlarge). But where would this veto come from? What would its rationale be? What would replace the vetoed decision? The simplest answer is that your decision to veto would have been preceded by a parallel, internal evaluation of the factors involved in the original decision, maybe with different weights, but essentially the same mental process. In other words, nothing new or extra-physical involved.
What would trigger a re-evaluation, if not free will? It’s easy to imagine a circuit
that serves the purpose. Let’s say that the scores of your top alternatives are very close to each other — so close that the decision is not clear-cut. In this case, your evaluation unit would start over, tossing additional factors into the mix, and would probably report the problem to your awareness center to bring additional processing power to bear.
Stop me when I say something that sounds like free will.
Wait! Where do the weights come from? Easy. They come from the positive and negative reinforcements from past experiences that “burn in” various mental pathways.
Hang on! When I consciously deliberate and then arrive at an answer, isn’t that free will? Not really. When you deliberate, your consciousness is patched into the decision-making process so you are aware of many of the goings-on, but the algorithm is about the same. Deliberation itself is triggered by the initial assessment stage — does the decision at hand call for a deliberative process? Would it benefit from analytical thinking? Should emotional factors be suppressed and be given lesser weight?
Hold the phone! How about whims? I have whims, like when I pick Fat Tire instead of Iron City — tell me that is a calculated decision! OK. That is a calculated decision. Whim can be thought of as a relatively large weight assigned to your “I hate to be bored” factor. Imagine that the weighting for the boredom factor grows with every mundane thing you do until averting boredom becomes a prime consideration. You don’t have to be aware of it. You follow “whims” because you hate to be bored and it’s time to be rewarded.
Come on. You can’t predict what I do. That is free will! The weather is unpredictable (more or less) but no one seriously attributes free will to weather. We recognize that the unpredictability of weather is based on the lack of knowledge of initial conditions, along with the complexity of the computations. Your behavior is unpredictable (more or less) for the same reasons. (Actually, it would shock most of us to know how predictable others find our behavior, in spite of our vaunted free will.)
Free to Be… Agents and Me
I think a large part of our difficulty dealing with this topic involves the confusion of free will with the concept of agency. I would define an agent as an embodiment with the ability to gather information about its environment, store that information, process the information and take action based on its information, in order to pursue a particular goal.
It is clear to me that finches and squirrels are agents, as are worms and mosquitoes. I might consider viruses to be agents, based on their ability to mutate. While I don’t regard your iPad as an agent, I think a robot could be an agent, if it carries a program that causes it to seek goals. But such a robot would still not have free will. I think we are more like robots (and squirrels) than we care to admit.
And there’s the rub. Most people don’t want to admit this. The illusion that we will our actions into existence is powerful — because so much of the process is unobservable to us.
We tend to think of agents as having not only freedom of action but also “personalities.” The personality of an agent derives from its unique store of experience — its experiences are unique because no two agents travel the same path. When our paths diverge, so do the experiences that shape the patterns in our brains. It is hard for us to think of agents and personalities without throwing free will into the mix, but these are separate concepts.
I Think, Therefore I Will
To preserve free will, some try to bring exotic physics or metaphysics into the picture. For example, the well-known physicist Roger Penrose and his associate Stuart Hameroff, director of the Center of Consciousness Studies at the University of Arizona and a former anesthesiologist (!), posit the existence of tiny fibers in the brain (“microtubules”) that serve as sites for quantum-level decision-making. They call the process “orchestrated objective reduction” or OrchOR. As Hameroff tries to explain:
Our “free will” actions could be the net result of deterministic processes acted on by hidden quantum logic at each OrchOR event. This can explain why we generally do things in an orderly, deterministic fashion, but occasionally our actions or thoughts are surprising, even to ourselves.
To the reader: Hameroff is saying that our two alternatives, Iron City and Fat Tire, are superimposed in your mind (something like listening to two songs at the same time) and that a quantum-level computation is harnessed by your brain to collapse the choice down to one. Which may surprise you.
Luckily, most physicists agree with The 100 Billionth Person that it neither necessary nor accurate to treat the brain as a quantum computer, and that a classical algorithmic model such as the one I outlined above is fundamentally sound. However, this is not to say that one’s decisions are immune to unpredictable quantum phenomena. Here I am thinking of radioactive decay or gamma rays, both of which are random probabilistic physical events. If you happen to be in the middle of making a decision when a gamma ray from a solar flare strikes one of your neurons, your choice may in fact differ from the one you would have made if the gamma ray had been emitted a second earlier or later. With respect to my “would you make the same choice if I could rewind the universe” scenario, this is the only exception I can imagine. And it would be rare.
The influence of random quantum events inside one’s brain should not be called free will — it may be “free” but there is no “will” involved. Again, I use the same definition of free will that most people do, that is, the ability to consciously and intentionally select (the will part) an arbitrary course of action (the free part).
Humans, and scientists in particular, need to stop looking for ways to defend our illusions. Free will is like the sun rising in the east and setting in the west — it is a true statement insofar as it describes what we see and how it feels, but it does not convey how the world actually works.
The Indispensible Illusion?
Defenders of free will inevitably bring up the topic of moral responsibility for one’s acts. Their line of argument goes something like this:
- Without free will, we would not be in control of our actions.
- If we cannot control our actions, then we are not responsible for what we do.
- If we are not responsible for what we do, then a just society cannot hold us accountable for doing wrong.
- If we are not held accountable, there will be no deterrent for bad behavior.
- If there are no deterrents, humanity will descend into a brutal state of nature.
- If humanity descends into a brutal state of nature, life will be hell.
- If life is hell, then what’s the point?
- There must be a point.
- Therefore, free will must exist.
I agree with Points 6 and 7. I disagree to various extents with the others.
Meta-Comment: This free will manifesto is already nearly 2500 words long. If I were to discuss my views on responsibility and justice at the same level of detail, it would double the length of this essay and distract from its main point. Nobody wants that. Please remember this in the following sections when I appear to make bald assertions and leap to grand conclusions. I’m doing it for you.
I venture that many more people fear the loss of a basis for moral responsibility than fear a lack of free will. For the former, the cascade from Point 2 (lack of responsibility) to Point 6 (life is hell) is steep, slippery and scary. So free will becomes the rope held onto.
When physics does eventually confirm that free will is an illusion, will we slide down that cascade to pointlessness? What’s to stop our descent? Answer: it stops at responsibility. Though we do not consciously control our behavior, we are responsible for it nonetheless. You are scratching your head, I can hear it.
When most people think about the word responsible, they have this definition in mind: “Able to make moral or rational decisions on one’s own and therefore answerable for one’s behavior.” My definition is more narrow, one that separates the social connotations from the physical fact: “Being a source or cause.” As an agent, I am responsible for actions that originate with me. No one or no thing other than me could be their source or cause.
If you think that being an originator denotes free will, it might help to return for a moment to the people-are-computers concept. An ordinary computer accepts input, processes it, and produces output we call the answer. Although the answer is produced as a result of a completely deterministic process, it did not exist before the computer did its work, so the computer is responsible for originating the answer. Human actions are answers.
Accountability is clearly a separate issue from responsibility. Accountability is determined by society — not by the agent — after responsibility (cause) is established. Society evolved sanctions for destructive acts hand-in-hand with rules for applying the sanctions; it is those rules that spell out our accountability.
Accountability as we know it often calls for different sanctions depending on the capability (not free will) of the agent. For example, we levy sanctions against children according to the developmental level of the child. Why should this be so? David Arredondo explains:
Indeed, there may be paradoxical or … negative developmental consequences of incompetent or developmentally inappropriate sanctions by a juvenile court. Simply put, there is the very real risk that the system can do more harm than good to a child who is still in the process of neurobiological, psychological, social, and moral development. Because of this, the negative consequences of careless sanctioning may be more enduring for a child (and for society) than they might be for an adult.
Note Dr. Arredondo’s mention of consequences for society. Note also that he said nothing about children having less “free will” than others. If free will is the ability to consciously select an arbitrary course of action, then children would seem to have as much of that ability as anyone. Clearly, free will is not a basis for accountability — social functioning is. Accountability will persist even if free will does not. The social norms that evolved over the eons will remain in place, serving as inputs to our evaluation circuits. We will not turn into blank-eyed robots, wandering aimlessly through the streets, when our connection to free will is severed.
But in the end, no matter what physics finds or what our intellects tell us, the free will illusion is too strong. To resolve our cognitive dissonance, we will keep acting like we have free will, whether we do or not. Remember this the next time you visit your local tavern.
After all these words, some may still feel I have glossed over the issues of responsibility, accountability and fairness. In particular, how can it be fair to punish anyone if our actions result from computations over which we have no control? You may not like my response, but, to paraphrase Tina Turner, what’s fair got to do with it? Moral responsibility evolved to serve social, not individual, needs. Standards of accountability evolved to be as fair as necessary to maximize voluntary compliance and maintain social order.
There was no point in time when humans “discovered” moral responsibility. Instead, our behavioral norms and concepts of social justice evolved as we did, to serve needs at hand. We became transactional beings, calculating payoffs and anticipating risks from our social interactions. We developed an array of strategies for dealing with other agents, including avoidance, conflict, wariness and cooperation. As Rod Serling might have said, if I had been one of his writers:
Wariness, that hard-scrabble mining town halfway down the track from Conflict to Cooperation, where the benefits and risks of engagement are weighed like diamonds, and trust is just a rough-cut stone. Wariness, our next stop — in the Twilight Zone.
Well, that was fun to write but completely off-topic. It was just a whim. So I had to do it.
Cooperation by our Stone Age ancestors may have inched us toward social responsibility, but my Uncle Ogg (bless his Cro-Magnon heart) never pondered free will. He took it for granted that agents had intentions, including agents like woolly mammoths, bison and the weather — anyone and anything that affected his prospects for survival. Including that man on the other side of the valley, eyeing the same bison, eyeing me. Does he have stones as sharp as mine? If I kill the bison, will the man kill me? If we hunt bison together, maybe we could share the meat and both walk away. Is he thinking the same thing I am? How can I trust him?
We cannot cooperate without trust. So we learned to monitor others for betrayals of trust. The social rules that evolved to punish defectors, the betrayers of our trust, constituted our earliest form of justice. Those basic rules and others that rapidly followed (notably the Golden Rule) are now ingrained in humankind. As Ken Binmore says in “Justice as a Natural Phenomenon”:
I think that the first step on the road to understanding the human thirst for justice lies in the recognition that variants of the do-as-you-would-be-done-by principle are already firmly entrenched among the instincts and customs that regulate our lives. The relevant norms do not survive because we consciously cherish them. On the contrary … most of our habituated behaviour is acquired via processes that operate below the level to which our conscious minds have easy access.
To put it plainly: we do not know love or justice or moral behavior because we engineered them
into existence. These behaviors arose because they are fundamental to productive, rewarding human interactions. We practice these behaviors and teach our children to imitate them. So social constructs like morality and fairness will not disappear when we agree that free will is an illusion — they are too important, too valuable, too imbedded in our cultural and genetic code. That is, they have high weightings in our evaluation units.
You may be interested in various sources I consulted, some of which actually support my point of view. In particular, I would like to recommend “The Mind’s I” by Hofstadter and Dennett, one of the truly influential books in my life. The other readings on this list are articles and papers available online.