Skip to main content

The Neural Basis of Free Will: Criterial Causation by Peter Ulric Tse (MIT Press, 2013)

I love Tse’s book. It has literally set me free.

This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American



What is free will, and how can it be generated by a neural circuit? The problem may seem, at first, like a fairly trivial issue. It seems evident to each of us that we wield free will. It feels like the world is our oyster—that with our wit and our will power we can overcome adversity. But the problem is actually deep, and it even precedes science in that it’s fundamental to religion as well. If God is perfect and knows all our fates, the Jesuit quandary goes, then the future is known, at least to Him, and therefore how can there be free will since every decision for the entire expanse of eternity has been laid out, unchangeable? There can be no real decisions, if true, no matter how free they seem. No matter how random or how spur of the moment any given decision was made, God knew you were going to do that.

The scientific version of the same quandary is not too different. Laplace imagined a similar scenario based solely on a universe of particles. Think of the universe as a billiards table, and all the particles in the universe so many billiards balls. If you could know all of the vectors of all the balls on the table, and all of the physical rules by which balls interact with each other—and with the table itself—you could predict any future position for the balls on the table. All possible interactions could be simulated forward to any future point in time. But even in a billiards table (or other closed system, like, say, the universe) where you don’t know every vector exactly, this line of thought suggests that although you cannot to predict the future accurately, you nevertheless know that future is predictable. That everything in such a universe is fated to happen.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


If our universe is deterministic in this way there can be no free will because you were destined to make that same decision—every single one of your decisions—from the very moment of the big bang. It’s not that you don’t make decisions: you do. But you’ll make them the same exact way in two different universes that have identical big bangs. It means that the universe conspired from its very inception to bring you and your significant other together. It’s quite romantic, actually, so long as you’ve been fortunate enough to have a nice life. But if not, you’re truly screwed, and the universe has been literally plotting your demise for the last 14 billion years.

For some physicists, these issues are not really issues at all because the universe, they claim, is not deterministic. That is to say, quantum level effects on particles are truly random. Therefore, the same big bang, if it occurred twice in two different universes made up of exactly the same particles having identical initial vectors would result nevertheless in different futures, because quantum level random effects change things up. Other physicists, including Stephen Hawking, poo-poo the quantum gambit, instead concluding that quantum effects may randomize particle vectors a bit at a low level, but only on such a minute scale that irrelevant to human life, and that the universe is, for all intents and purposes, deterministic. That rings true with me—quantum mechanics predicts that everything is possible with varying levels of probability, but that means that the way things are would probably happen again and again, given the same big bang and the same universe of particles, in each iteration.

So the universe is either deterministic or non-deterministic, and your free will to choose the red versus the green sauce for your burrito hangs in the balance. First, why would there be a problem with free will? A blade of grass rotates to follow the sun throughout the day in a process known as heliotropism. But does it intend to follow the sun? No, it’s just a chemical reaction, its basis is known, and it’s straightforward to explain. To be clear, I reject the idea that the blade of grade is conscious at some low level. That’s silly. Ok, now, what about if you take a neuron—a brain cell—and stick it in a dish with all the nutrients it needs, then spritz neurotransmitter on it. It responds with a vigorous burst. Did it intend to burst? “Well, ah, noooo, I guess not”, you might say, with some discomfort, as you realize that your brainpan is merely a dish that provides all the nutrients your brain needs, and the neurotransmitters on each neuron are all provided from other neurons in the same dish.

Making your will even less free, research in John-Dylan Haynes’s lab at the Bernstein Center for Computational Neuroscience in Berlin, Germany has shown that an upcoming decision being made by human volunteers can be predicted from brain scanning data, at levels better than chance, up to 10 seconds before the volunteers know what they themselves are going to decide. So unconscious processes are thus actually in control your decision making long before you even know you’re making a decision. Whoa!

So that’s it, we’re done, there is no free will, right? The universe can’t sustain it, your brain has made up its mind before you have, and you have merely experienced the illusion that you’ve been in control since the moment you learned to crawl and discovered the feeling of agency.

Not so fast. Enter Dartmouth neuroscientist Peter Tse, who has found a middle ground in his new book “The Neural Correlates of Free Will: Critical Causation” (MIT Press). Tse has thought through this enormous problem and realized something important that brings free will back to the realm of the living. Remember that determinism is an unavoidable fact of the universe at the macroscopic but not the quantum level. Well what if the macroscopic universe is not deterministic because the brain is designed to amplify quantum level particle effects to the macroscopic level through the action of specialized neuronal channels that make decisions potentially truly stochastic?

There are chemical receptors on most neurons that receive neurotransmitters (globs of chemicals secreted by other neurons), that then respond by opening ion channels, causing neurons to create neutral impulses (aka: macroscopic real world events normal people call “brain activity”, or “thought”). Well, in a deterministic universe… so what? You could have predicted every idea I’ve ever had, before my birth, if had enough data about the universe. Right? Tse says no, because some chemical receptors, called NMDA receptors, are actually blocked by a single atom of magnesium, that must first be released before ions can flow to cause brain activity. Because macroscopic brain activity is therefore dependent on the position of a single atom, which is itself a quantum-level creature, it means that these neurons amplify the quantum level activity of the magnesium atom to the level of neural circuit behavior and real life. Thus our behavior is indeed subject to quantum effects and the universe cannot be deterministic.

I love Tse’s book. It has literally set me free. It explains these ideas in full glory, in exquisite detail, and much better than I can summarize here.

Added post-hoc: Author Peter Tse contracted me and thought that some potential reader questions could be clarified by noting that the brain works as a coincidence detection machine at the neuronal, dendritic and receptor level, and that quantum level randomness has its effects by randomizing spike timing, and therefore this determines what will count as a coincidence. Also, a key idea of the book is that certain brain circuits associated with planning can change synaptic weights extremely rapidly, so that neurons can respond differently to same inputs, even within a short period of time based on the rapid synaptic weight changes brought about by quantum level effects. This can change the information that neurons are responsive too, so that the neural code becomes not just a spike code, but a synaptic weight code. For example, executive circuits might set the criterion that memory circuits should recall a politician. Margaret Thatcher might come to mind, because she meets that criterion. But from the exact same initial conditions, Obama might also have come to mind. Which came to mind was a matter of chance. This is not deterministic, because the outcome could have turned out otherwise. But it is also not random, because the outcome had to be a politician.

Stephen L. Macknik is a professor of opthalmology, neurology, and physiology and pharmacology at SUNY Downstate Medical Center in Brooklyn, N.Y. Along with Susana Martinez-Conde and Sandra Blakeslee, he is author of the Prisma Prize-winning Sleights of Mind. Their forthcoming book, Champions of Illusion, will be published by Scientific American/Farrar, Straus and Giroux.

More by Stephen L. Macknik