It is always hilarious when people rail against free will, who are especially flummoxed that Common Man believes in free will, and say “If only people realized their choices weren’t free, they would make better choices.”
The fallacy is usually buried in a long string of propositions, the length of which causes one to forget where one began, which is the premise “actions aren’t free.” Yet if that’s true, then nothing matters, and no choices can ever be made, right or wrong. Indeed, there is no right or wrong: there is nothing. Even your desire that this should not be so is nothing.
Yet these fans of science still believe that better choices would be made if folks knew they couldn’t make choices, and that all was absent of free will.
This silly argument is not limited to our progressive pals. It’s seen on the right (the Wrong Right), too. Z-man, God bless him, is a frequent proponent. He says “The concept of free will has been essential to Western thought since the Greeks and it is an essential element of Christianity.” True enough, but, Zed, it is essential everywhere if it is true. Which it is.
Not so, he claims. Free will is a “myth” (he uses the word as a synonym for false, as unfortunately many do) because the choices people make are “so easily predicted by behavioral genetics.” As evidence for this, he points us to somebody called Jayman (this makes me B-man, I suppose). “No, You Don’t Have Free Will, and This is Why” insists J. He opens with this bit of hilarity:
Slate recently featured an article written by Roy F. Baumeister, Do You Really Have Free Will? In it, he claims that human do indeed have free will, something that regular readers will know that I have emphatically argued against.
Why, dude. If people don’t have free will, then there is no reason to argue, emphatically or like a lady, that they don’t. People can’t make better choices if they can’t make choices.
Anyway, Jayman quotes statistics showing “all human behavioral traits are heritable”. By heritable, he does not mean necessarily passed on, but merely sometimes in certain measure passed on, where most measurements of behaviors are all forced quantifications of the unquantifiable. Jayman appears to take weak correlation as complete causation.
It is true and was always obvious until yesterday different races have different distributions of proclivities and behavior, and that some of these differences are biological, i.e. innate. Thus that race of folks who have different spleens (“the Bajau takes free diving to the extreme, staying underwater for as long as 13 minutes at depths of around 200 feet”) will react differently on average than, say, Arabs to being tossed into the drink. But this does not eliminate free will. The words used when finding oneself dunked are still freely chosen from a conditional subset of words. So free will is conditional, but so what? Most things are conditional, including cause, probability, and the rightness or wrongness of many acts. That right and wrong are sometimes conditional does not mean there is no right and wrong. There is an infinite gap between conditional and determined.
Now most of our bodily activities are given over to automation, including those activities, like walking, where robust and active free will was initially necessary to learn the activities. Where next do you place your foot? At first we think hard about it, but eventually not at all. Eventually there is no free will in each step. But you may, at any moment, decide to take a skip instead of a step. The potential for free will is always there.
If you decided to model, using the latest deep-learning neural net massively parallel AI, to predict with near certainty that, on a walk, after I take a step with my left foot I then will take a step with my right, and vice versa, you have not proved the absence of free will. Nor would you have disproved free will if you hooked an fMRI to me while walking and asked “When did you choose to use your right foot?”
The fMRI would (more or less) show that part of the talking automation taking place in the brain, where to choice might seem to come after the “decision”. But this is because this is not a real instance of free will. You have to expend real mental effort to overcome the automation. You’re not really making a choice of step, even though the experimenter put the act in those terms. You might even try and stutter steps, which is a free act of will, but then you have given over to the automation the orders “Stutter steps”, and again you’re not quite exactly making precise choices of each stuttered step. But you did use free will to start the process. The fMRI would not capture any of this.
Same kind of thing happens when you learn a video game. At first you carefully and freely plan which button to press, but after a while automation takes over. A good thing, too, for it frees the mind to think of other things. Like strategy.
This puts into proper context articles like “Decoding the contents and strength of imagery before volitional engagement” by Roger Koenig-Robert and Joel Pearson in Nature: Scientific Reports. They hooked a small group of folks to a video game asking them, in a horribly convoluted process (see Fig. 1), to press a button indicating a choice of what kind of interleaved stripes would appear on a screen. The people practiced, like in a video game, until they got good at it.
A model based on fMRI images was able to predict “choices” made at that game to a good but imperfect degree. The fMRI did not measure, and couldn’t, the process whereby the participants told the automation to do its thing. Free will was not disproven.
You may predict with even greater certainty what your wife will say when you forget to do the task assigned to you for the umpteenth time. But this does not mean your wife does not have free will. No, in order to prove the lack of free will, you need to demonstrate without error the causes of every action. And that will never be possible.