If freedom is just another word for nothing left to lose, then “free will” is just another phrase for ability to choose.
Bad, wasn’t it? But if free will is an illusion, as many scientists and philosophers have argued, then you shouldn’t blame me. On the other hand, I do blame myself. Because like most bloggers, and possibly even the several dozen humans who don’t blog, I think I decided for myself what to write. Besides, as many investigators of this issue have pointed out, it’s not so obvious that free will is illusory now that quantum mechanics has inserted some randomness into nature.
Sadly, though, that reasoning doesn’t get you very far. There’s randomness in the quantum world, all right, just like the unpredictable sequence of winning numbers on a roulette wheel. But in the long run all the numbers turn up about equally often. Free will isn’t worth much if you can’t use it to beat a casino. And as MIT physicist Scott Aaronson points out, quantum math is similar: It gives the odds about what various possible things will happen, and those odds are always predicted precisely. The probability distribution of results is always just what the quantum math says it will be. Aaronson doesn’t see any free will there.
Still, the free will question has elicited some sophisticated musing from quantum physicists who like to contemplate the interface of mentality and physical reality. It seems reasonable enough to reexamine such an old question in the light of the latest understanding of the universe. It may be that modern physics can offer a perspective giving hope for those who like to make up their own mind.
As opposed to neuroscience, which explores how decisions are motivated and made, physics seeks explanations for how decisions are caused. If the laws of physics apply to people (and surely they do), choices are determined by those laws, operating in whatever particular set of circumstances people find themselves in. Presumably there’s no freedom of choice from the physics perspective, simply the inescapable tyranny of mathematical equations.
Such arguments against free will have been around for a long time. They were forcefully articulated by Henry Thomas Buckle, a 19th century British historian, who lambasted the whole free will idea. Just believing in free will doesn’t make it so, he said. There’s no good way to tell whether you have free will or you just think you do. Examples you might cite that seem to illustrate free will can always be interpreted otherwise. No principle prevents a person’s actions from being perfectly predictable.
“When we perform an action, we perform it in consequence of some motive or motives that … are the results of some antecedents,” Buckle wrote in History of Civilization in England. “If we were acquainted with the whole of the antecedents, and with all the laws of their movements, we could with unerring certainty predict the whole of the immediate results.”
Buckle acknowledged that in practice, you could never know every single antecedent leading up to a person’s behavior. “But it is certain that the nearer we approach to a complete knowledge of the antecedent, the more likely we shall be able to predict the consequent.”
In other words, as Buckle construed it, free will requires the ability to make a choice that could not have been predicted. Quantum physics reduces the possibility of such predictions to probabilities, but nevertheless constrains those choices to obeying equations.
But suppose some actions can’t be predicted. Then maybe the free-choice door cracks open a bit. And examining predictability might get you farther in the free will debate than asking whether free will itself exists. After all, says Aaronson, proving the existence of free will is hard. If you try, you’ll just get enmeshed in the philosophical arguments that have been going on for millennia. Instead, he proposes a more limited — but scientifically addressable — question.
“I advocate replacing the question of whether humans have free will, by the question of how accurately their choices can be predicted, in principle, by external agents compatible with the laws of physics,” Aaronson wrote in a paper posted online last year.
He does not claim (or even attempt) to answer that question. He merely explores how to frame it in a way that it could be scientifically tested, someday. But he does not look for unpredictability in the mathematical laws governing how the future evolves from the past. In his view, free will would arise not from the equations, but from the conditions the equations are applied to.
Equations for predicting the future always need a starting point — the initial conditions, or boundary conditions, as input into the math. Maybe the initial conditions of the universe included elements of unpredictability that prevent equations from ever forecasting every single aspect of the future, Aaronson speculates. If so, there would be some “freedom” in the universe — “a certain strong kind of physical unpredictability: a lack of determination, even probabilistic determination, by knowable external factors,” Aaronson declares.
He calls this unpredictability “Knightian freedom” (for Frank Knight, an economist who wrote about uncertainty that can’t be quantified with probabilities). Knightian freedom by itself doesn’t mean that people have free will. But Aaronson believes you could not have free will without it.
At first glance, modern physics offers no obvious source of Knightian freedom. But perhaps, says Aaronson, a source of such freedom might have been hiding in “the microscopic, quantum-mechanical details of the universe’s initial conditions.” At the very least, whether or not such a source of freedom could be found there is a legitimate scientific question.
“It seems to me that ‘freedom’ — in the sense of Knightian unpredictability by any external physical observer — is perfectly within the scope of science,” Aaronson asserts.
Knightian freedom might be possible thanks to a quirk of quantum physics known as the no-cloning theorem. Basically that means that a quantum object — say, a particle of light — cannot be perfectly copied. Such a particle exists in a “quantum state” with its own sort of freedom — the axis around which it spins can point in any of a number of directions at once, for example. A measurement of its spin will force a choice of one direction or another, destroying the multiple possibilities. Any attempt to copy the original state has the same effect as a measurement, so that original state cannot be cloned. (By the way, the “measurement” does not have to be performed by a person. Interaction with matter of any sort accomplishes the same thing.)
Such quantum states are sometimes called qubits, an analogy to the 1s and 0s, or bits, of computer information. Floating throughout the universe, Aaronson suggests, are qubits that have never been measured — they retain their original freedom. Some of those qubits might even possess Knightian uncertainty. Aaronson designates them as freebits.
“At least some of the qubits found in nature are regarded as freebits, and the presence of these freebits makes predicting certain future events — possibly including some human decisions — physically impossible, even probabilistically and even with arbitrarily advanced future technology,” Aaronson proposes.
Pause to reflect. Perhaps from the universe’s past there have emerged some microscopic bits of Knightian freedom — possessing unpredictability. Is there any way that unpredictability could be exploited by decision-making human brains? Aaronson says yes. A freebit impinging on a brain cell could alter neural signaling by a tiny amount. Subsequent chaotic activity in the brain could amplify the original unpredictability to a macroscopic scale, rendering some human decisions unpredictable.
Bear with him. As Aaronson elaborates:
“Before their amplification, these freebits would need to live in quantum-mechanical degrees of freedom, since otherwise a cloning machine could (in principle) non-invasively copy them. Furthermore, our ignorance about the freebits would ultimately need to be traceable back to ignorance about the microstate of the early universe.”
Or in other words (still his):
“A ‘quantum pixie-dust’ left over from the Big Bang … gets into our brains and gives us the capacity for free will.”
OK. Aaronson admits it sounds crazy. But he insists that, as far as he actually goes (which is merely raising the possibility), his idea for quantum freebits is not disqualified by any known physics.
“I don’t see that any of it is ruled out by current scientific understanding — though conceivably it could be ruled out in the future,” he writes. “There are possible universes consistent with the rules of quantum mechanics where the requisite states exist, and other such universes where they don’t exist, and deciding which kind of universe we inhabit seems to require scientific knowledge that we don’t have.”
One downside is that his freebit idea prohibits any simple theory that fully describes the universe’s initial conditions. “But it seems to me that nothing in established physics should have led us to expect that such a description would exist anyway,” he points out.
Another warning: this loophole would not permit an unlimited number of free decisions.
“The nature of freebits is that they get permanently ‘used up’ whenever they are amplified to macroscopic scale,” Aaronson writes. “Only a finite number of ‘free decisions’ can possibly be made before the observable universe runs out of freebits!”
So if you’re trying to decide whether to believe any of this, you should make your mind up soon.
Follow me on Twitter: @tom_siegfried