In which I solve the problem of free will

I enjoyed listening to two episodes of Adam Conover’s Factually podcast, in which he hosted guests with competing perspectives on the existence of free will, a profound question that has vexed thinkers of all stripes for centuries. We have a strongly developed intuition that we choose what we do, and our theories of morality rely on this intuition. And yet, as we understand more about what humans are and how we work, it seems there is no space in us for a choosing agent.

The first guest, philosopher Robert Sapolsky, argues strongly that free will doesn’t exist. The second guest, Kevin Mitchell, argues strongly that it does. 

Well I’ve been meaning to do a bit more writing for this blog, so what better way to do that than by solving free will.

Here’s me: ‘free will’ in the absolute sense is, as both Conover’s guests agree, a nonsense. Human action is never freely determined, there is no master-ego within us with an open field of action, there are always profound constraints on the behaviour we produce. The real question is whether human action is fully explained by summative forces of domino-chain neural firing across synapses, which instantiate all the prior experiences and learning that shaped our brains; or whether there is some gap to fill.

Dr Kevin Mitchell argues persuasively that there is a gap. The future is made fuzzy by tiny proportions of randomness that multiply over time. The domino chains must admit uncertainty, which in turn allows for the imposition of some system of evolutionarily-developed cognitive control. (I am, of course, oversimplifying his account wretchedly, and compounding the crime by doing so from memory.)

I have my own account. I have not done the reading on this; it’s just the operating assumption I’ve landed on that satisfies me. Maybe it will satisfy you as well. It’s this:

I don’t think we experience free will at all. Examining our experience of conscious thought can never actually locate a moment of choice. We can experience need to make a decision; we can experience the decision as made. The transition is not something we can actually locate. This is because it is not something we do with conscious thought.

To restate in a doubling down sort of way: our choices are not conscious.

What we are in fact doing, and what we experience with our consciousness, is cycling through information, refreshing it, reweighting it, experiencing it in a new present context; and then, this process eventually tips over enough dominoes to take us down a path of having chosen. Our experience of conscious choice is actually the process of assembling sufficient information that a choice becomes made within us.

The process of considering – what we call thinking – is just dominoes. A decision-making process is itself triggered by prior dominoes, some conscious, some not. Everything that happens in that process can be causally traced: our eyes take in photons that our brain assembles into a person’s face, that trigger associations with that person which we associate as memories, which in turn weight other dominoes so they fall this way not that. Our cognitive systems cycle through pass after pass of neural patterns firing, which we experience as thinking, because that is what thinking is. One thought leads to the next, we manage to focus on one issue above others, we discard alternative hypotheses, we make a choice – that’s the conscious experience of the domino array. The hard thing to wrap your head around is how complex those dominoes are: the number of inputs is so vast, and the fuzzy outcomes and uncertainties that flow within them are so deep, that a control system that is a huge evolutionary advantage to generate behaviour out of chaos.

And we experience those vast domino arrays in our brains as consciousness. So it could be said our conscious experience is, yes, fundamentally epiphenomenal. But, and here’s the trick to that: so what?

For what it’s worth, I don’t think this perspective does require an epiphenomenal consciousness. I think the thing we experience as consciousness *is* the process of tuning and filtering that produces our behaviour. In the same way we experience a kick to the shin as pain, we experience neurons working shit out as consciousness. That’s phenomena, not epiphenomena!

That intuition of free will still isn’t satisfied though, right? It’s all just dominos and even if some of them are wobbly, where is the decision? Where’s the moment my inner self stretches out a finger and chooses whether to knock over domino A or domino B? Well with all due respect to this intuition, I think it is unproductive. And not even based on what we actually experience introspectively. What it is, is a nice story, a narrative explanation that sews together all kinds of elements of life into a single useful tapestry (this is not mixed metaphors, a narrative can be a tapestry and vice versa, also HELL YES a tapestry can be useful, have you ever tried to decorate a wall in a medieval castle???). What we’re really talking about here is, in the statement “I think about this problem and then I come to a decision”, who is “I”? Is I my conscious self, whatever that might be? Or is it the sum total of all the thinking processes I have, some of which I can talk about and experience and some which I can’t? If you’re too precious to admit there’s unconscious stuff as a fundamental part of everything you think and do, then I can’t help you baby… and if you do admit there’s unconscious stuff in the mix, then how can you account for that except by admitting that some of “I” is not within your awareness? In conclusion, consciousness might not be an epiphenomenon, but the story we tell ourselves of a willful, agented self sure is. That’s a whole other blog post though.

As for the question of how the ‘lack of free will’ affects morality: no of course it doesn’t follow that since all our actions come from wobbly dominos not a willful agented self then we should let the murderers go free. All behaviour is part of a great matrix of social information, and by enforcing rules we are tuning society to better suit our needs. So lock ’em up! (Actually carceral systems are bad for society including for most murderers.)

Anyway that’s what I was thinking as I went for a dog walk this morning. (Also, the problem of free will kind of goes away if you don’t think moving time exists, moving time I have my eye on you, I am not convinced)

3 thoughts on “In which I solve the problem of free will”

  1. I don’t think Sapolsky is a philosopher at all though? He’s a biologist with a massive specialisation in primate behaviour? (Incredible speaker if you check out YouTube and an entertaining writer too).

    I’m also not sure I am fully on board with this idea. We have all experienced the struggle against our first reaction, our instinctive reaction, and making ourselves Do The Thing. What are we doing when we overcome urges? What are we doing when Kahnemann’s system two wins over system one?

  2. (True about Sapolsky, although if you’re writing books about free will I think you qualify as a philosopher on some level!)

    To me, the struggle to assert our higher self (rational self, will, system two…) over our base urges (instinct, system one…) fits smoothly into this model. The whole thing is a vast domino array, and one layer of the dominos is the control system of consciousness – the processing that runs through here we experience as thought, and if that layer of dominos winds up contributing more to the output of the whole array, then that’s what exerting our self-control both feels like and is…

Leave a Reply

Your email address will not be published. Required fields are marked *