4

In his lecture at the Royal Institution, Associate Professor of Genetics and Neuroscience at Trinity College Dublin, asserts that humans have free will; that a human can be the 'locus of causal power'.

Mitchell asserts that neither determinacy or randomness allow any space for free will, but proposes a third option, whereby, "...indeterminacy opens up a little room for organisms... to have some causal power at a macro scale" (30:00).

He describes how we develop the capacity to make what he assumes are real choices, largely via a process of perception, thought, action and learning (45:35).

He paraphrases Sam Harris, who claims that "...we don't know where our thoughts or intentions come from - they just bubble up into consciousness. We have no power over them and can't choose them. Mitchell crucially then claims that Harris is "plain wrong", and that he thinks "...we choose our goals and intentions all the time".

The reason I say 'crucially' is because one definition of free will that makes sense to me (and at least one other user of this site) is that it is the 'ability to make decisions'. This definition allows us to differentiate between two contrasting states, either of which might be the case in the world:

  1. A state where we actually originate decisions as causal agents, and
  2. A state in which what we perceive to be decisions are in fact merely events which occur as the result of determined and/or probabilistic/random sequences; events over which we exert no control.

If the 'ability to make decisions' is a valid definition, then Mitchell simply stating that we make choices routinely is assuming that which he is trying to prove. I acknowledge though that he might not share this definition, in which case, I am wondering whether his excursion into biology satisfies his claim that humans are the loci "of causal power". Perhaps I have not presented enough of his case here for a conclusion to be drawn without a viewing of his entire lecture, in which case I am hoping any interested SE users will take the time to watch it and respond. I want to steelman Mitchell as much as possible, but to do him justice would probably warrant this question becoming far too long. This is why I've focused on his claim that we make choices, for it is here that I think he has made an assumption.

In short, I don't feel as though he has explained/justified his claim - as stated earlier - that "...indeterminacy opens up a little room for organisms... to have some causal power at a macro scale"; that what he sees as 'choices' are in fact real choices and not illusory ones.

Am I misunderstanding/underestimating the strength of Mitchell's idea? Has he demonstrated that humans have free will?

1

7 Answers 7

3

It does seem like he might be assuming the very thing he's trying to prove, especially if we define free will as the "ability to make decisions."

It's that classic “ghost in the machine” problem, right? Just because our brains are buzzing with activity, and we feel like we’re making choices, doesn’t necessarily mean those choices are truly free from prior causes, whether deterministic or random.

Mitchell's argument hinges on this idea of "indeterminacy" opening up space for free will. He’s trying to carve out a middle ground between hard determinism and pure randomness, where our agency can emerge. He gestures towards biology and the complex processes of perception, thought, action, and learning – and that's where he needs to do more work.

To truly steelman his argument, we need more than just the assertion that we make choices. He needs to show how this indeterminacy at the micro level actually translates into genuine causal power at the macro level of our decisions. How exactly does this "room for organisms" to exert agency actually function? What's the mechanism?

He does touch on these complex biological processes, but it’s true that he doesn’t fully flesh out precisely how they lead to free will. He needs to bridge that gap, to connect the dots in a way that moves beyond simply stating that we choose and instead delves into the how of that choosing.

So yeah, you're right to be a little skeptical. Mitchell's lecture sparks some really interesting questions, but he doesn't quite get us to a satisfying answer about whether those choices are truly free in the sense that matters most to this debate.

1
3

"If the 'ability to make decisions' is a valid definition, then Mitchell simply stating that we make choices routinely is assuming that which he is trying to prove."

I think it's worse than that - I think his take produces an impossible infinite regress.

He paraphrases Sam Harris, who claims that "...we don't know where our thoughts or intentions come from - they just bubble up into consciousness. We have no power over them and can't choose them. Mitchell crucially then claims that Harris is "plain wrong", and that he thinks "...we choose our goals and intentions all the time".

Okay, so let's look at a moment where we choose our goals and intentions. At 13:00 today, I consciously made it a goal that I would get through my work emails, read every one by the end of the day. I consciously made that choice, right? Okay, so let's look at that - why did I consciously make that choice? Well, I had a series of thoughts, maybe something like "My emails have been unread all day, I don't want to leave it until tomorrow when they'll pile up even more, there might be something important in there, so I would be better that I deal with it today than leaving it to pile up." And then I make the decision to do it, right?

So the problem that arises is, even if it's true that I chose to prioritize the goal of reading my emails, it would seem as though I DIDN'T choose to have the thoughts that produced that choice to read my emails.

But, suppose someone says, 'nope, I chose to have those thoughts too'. Okay, so chose them based on what? Chose them because of what preceding chain of thoughts that produced that choice? And did you also choose those thoughts?

If choices are the consequence of thought sequences, then to get around what Sam Harris describes as "thoughts bubbling up", one MUST have an infinite sequence of choices. Otherwise, it must be true that the choice to read my emails was the consequence of thoughts that just bubbled up.

It seems rationally and experientially / intuitively more likely that Sam Harris has a substantial point here, and that thoughts really do bubble up, they really do sort of happen to us.

3

Mitchell presents a recent state of physicalist science about biology and the universe, all of which point to compatibilist free will. Instead of acknowledging this, he just repeatedly makes baseless assertions that incompatibilist free will is involved.

The main straw man he is burning is that compatibilist free will would only be able to display mere reactive behavior, without memory or planning ahead, which is not a claim or necessary trait of compatibilism.

You can see this most clearly maybe around minute 19 of that lecture, when he presented how a bacteriums behavior is studied in a lab understanding how a bacterium moves towards food sources. His next point is that reality is more complex more factors go into a bacteriums internal processing.

He revealingly says: "it's response depends on it's own nutrient state, it's not just an automated kind of a thing".

But that's a contradiction, if behavior depends on a nutrient state, it is very much an automated kind of thing. Automated things can be complex, incredibly complex. There is no limit to how complex "automated kind of things" can be.

The fact that bacteria have internal state determining behavior is also no more special than a good old mechanic kitchen timer having an internal state ticking away causing it to ring when the international spring reached it's end without additional input from the outside. It's just a delayed reaction.

So Mitchell presents all the physicalism evidence that supports compatibilism, but burns some strawmen to give the impression of incompatibility free will being a more satisfying model.

Other obvious mistakes he makes are claiming that compatibilist free will would require a deterministic universe (it does not), or that determinism implies predetermining (it does not), or that system states cannot "contain all the information of future states, as it is Mathematically impossible (Conway's game of life showing the opposite to be true, in a way any child can understand).

Mitchell holds his lecture without even once using the word compatibilism, which means he either didn't read up about the main contemporary view on free will (which is unlikely, but possible), or he leaves this out because it does not fit with his goal in publishing his book.

His problem seems to be that he has nothing new to offer to readers interested in compatibilism, so a book on that would not sell, but he needs the money, so he needs to make a case for incompatibilist (as those folks are desperate), but his own science points to compatibilism (at best).

3
  • Thanks for warching the video. Commented May 24 at 8:36
  • That determinism isn't true, and that evolution uses that randomness for evolutionary benefit, are both pretty significant points. Most of the anti-free-will posters here don't understand/concede either of them. Of the three classic free will positions, they refute incompatibilist determinism, and compatibilist determinism. So that a non-philosopher hasn't absorbed that there is a version of non-deterministic compatibilism should not be that surprizing. He is basically arguing to refute Sam Harris type thinking.
    – Dcleve
    Commented May 24 at 15:00
  • That this universe may not be deterministic is not new to anyone, and mostly irrelevant to compatibilism. Mitchell does not say this indeterminacy is used by evolution (unless I missed), and it would be wrong to say so. Evolution needs changes that are arbitrary and thus called random, but not indeterministic random.
    – tkruse
    Commented May 24 at 20:31
3

I generally recommend that one dig into the details of someone's thinking through their writings, not a video, as videos are much harder to backtrack and cross-check with than writing. However, I have not read Mitchell's book, only this review: https://inquisitivebiologist.com/2024/01/23/book-review-free-agents-how-evolution-gave-us-free-will/ I will try to offer an answer based on it.

As you noted upon reading the review of Mitchell's thinking, when you dig deeply into it, it is hard to distinguish from compatibilist determinism. Conifold noted that Mitchell is basically a two-stage libertarian, but even for the most noted of those, Tim Kane, I still had difficulty identifying how his views still weren't just compatibilist determinism.

From what I can tell from the review, Mitchell does a good job showing how evolutionarily we really do make choices. He apparently makes this case by a bottom up process, where he shows how neural systems can be multi-modal, and respond usefully to randomness, as well as the evolutionary utility of non-determinism in lower life forms. One could make this same argument from a top-down perspective, as System 2 of Thinking Fast and Slow is clearly tuned and selected for in order to do 2nd-guessing of choices.

Where it sounds like his argument falters is at several steps. The review dismisses the QM uncertainty transitioning up to macro scales, but does not mention the Chaos Theory principles that would lead to such leveraging. This leads me to think Mitchell left out that critical argument step.

Additionally, the review does not discuss emergence, nor the causal efficacy of emergent structures. If Mitchell did not discuss emergence, this too would represent another major argument step missing in his reasoning.

The most significant shortfall, and the reason that Mitchell seems to be hard to distinguish from a compatibilist, is because libertarian free will needs a third causal option, "agent causation", which is neither determined nor random. And if Mitchell tried to shoehorn his agency thinking into a determined/random bimodal causal logic (as Kane did as well), there really isn't any way to avoid compatibilism.

The final part of the argument for libertarian free will is to cite logical pluralism, and that NO logic is prior knowledge. Whether we have a third option or not in our world is an empirical question, and Mitchell's evolutionary evidence provides strong plausibility that agency is real and matters.

So, in my judgement, guided by the review, Mitchell understands that Libertarian Free Will needs a semi-deterministic world, with the ability of small inputs to trigger significant alternatives. AND that he needs to show both that this is the case, and that evolution makes free choosing beneficial. But that in trying to make the argument needed, he left out three critical steps, detailed above.

Note that even if Mitchell had added in each of these other three points, no he would not have "proven" free will. Whether free will is real in our world or not in an empirical question, subject not to proofs, but to Inference to Best Explanation question. With the bottom up, and top down evolutionary cases for free will, the best explanation is that it is real, and we are in a world with Agent Causation.

2
  • Thanks Dcleve. Can you explain what you mean by 'no logic is prior knowledge'? Commented May 24 at 2:58
  • 1
    @Futilitarian We cannot treat any logic as a default presumption. It must be justified by empirical support.
    – Dcleve
    Commented May 24 at 7:04
2

what we perceive to be decisions are in fact merely events which occur as the result of determined and/or probabilistic/random sequences; events over which we exert no control.

You seem to be saying that the will is only "free" if it can be shown to be brought about - by necessity - what would otherwise be - but now aren't - indeterminate proceses.

It'd be a move outside the dereminism debate, and one that I don't think is needed for moral responsibility. There are various different sorts of causation, and I don't think we need a specific one (be that action of a soul or an inviolable desire) for moral responsibility.

1
  • ofc there is more to the free will debate than that, but i hope that's of some use
    – andrós
    Commented May 22 at 14:29
2

You define free will as the „ability to make decisions“. This definition seems to me useful to discuss the qestion of free will.

  • I watched Mitchell’s video in full length. I also stumbled over his announcement

    “… indeterminacy opens up a little room for organisms … to have some causal power on a macro scale.”

    In particular his affirmative quote of Epicurus of atoms “swerving” sometimes. Fortunately Mitchell does not come back to this point. In fact, he does not explain decision making from the incompleteness of determinism. It is never the best choice to place an explanation in the gaps of a rivalry model.

  • Mitchell bases his talk on evolution, biology and a bottom-up view of information processing and decision making in living beeings. That’s rather refreshing because his viewpoint is not standard in philosophical discussions of free will. Nevertheless he goes the whole path up to the meta-level of the human organism: Thinking about one’s own thinking. Long-term traits controlling short-terms decisions.

  • To answer your question: Mitchell makes plausible how decision making is a capability of living beings. He shows how mechanisms working already in simple animals exemplify what on a higher level we call decision making.

    He shows that the capability of decision making is consistent with a causal worldview.

    I would vote even more to avoid any discrimination between “I” on one side and my genetic and neural wiring on the other side. Instead: I am a person, which includes my body and my genetic and acquired neural wiring.

-1
  1. A state where we actually originate decisions as causal agents, and
  2. A state in which what we perceive to be decisions are in fact merely events which occur as the result of determined and/or probabilistic/random sequences; events over which we exert no control.

Option 1. seems to be the case.

It is impossible to "perceive to be a decision" something that is not a decision. A random draw out of multiple options could not be perceived to be a decision as there is no agent to make that decision.

Writing your question was a lengthy series of decisions all made by you, there was no-one else. This post did not exist before you decided to write it. You are the author, you caused these words to appear here.

Your question was inspired, not caused, by this lecture. You felt the need for more knowledge about the subject and you chose this forum and these words to acquire that knowledge. None of this could happen as a causal reaction to prior events.

10
  • 1
    "None of this could happen as a causal reaction to prior events." - that seems at best entirely speculative, and at worst clearly factually incorrect.
    – TKoL
    Commented May 23 at 16:30
  • @TKoL Causal reactions cannot create anything, they don't know or feel anything, they don't need anything or strive to achieve any goals. These are simple facts. You cannot construct a working mind that consists solely of causal reactions. Commented May 23 at 20:46
  • 1
    The closest thing to a working mind that anybody has ever constructed does consist solely of casual reactions - LLMs
    – TKoL
    Commented May 23 at 21:13
  • 2
    "There is no concept of cause & effect in psychology." - just seems like you're making stuff up at this point. That's literally the opposite of the truth.
    – TKoL
    Commented May 24 at 7:48
  • 1
    "everyone knows this", he said, after making up a bunch of fictional ideas
    – TKoL
    Commented May 24 at 8:45

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .