Too free will

An interesting preprint on the whole free will debate thing: “Reframing the Free Will Debate: The Universe is Not Deterministic“. The argument is that based on what we know the universe is not deterministic (because quantum1) so the biggest point of contention in the free will debate – whether or not free will is compatible with determinism – is out of date. The authors noting that “In an indeterministic universe, freedom (in the broad sense of real possibilities or leeway) comes for free” therefore propose the more pressing question “where does the control come from? How can agency emerge in the presence of such indeterminacy?2”

This restatement seems like a different version of why is there order instead of chaos? At this point, 13.7 billion years after the creation of the universe, and 3.5 billion years into the existence of life on Earth, we can safely assume that there is order and our best answer for why there are complex organisms is evolution3. Taking the data we have, it seems to me that even if the universe were deterministic, randomness would still have to evolve. Randomness can be a good solution to some of life’s problems. Indeed, in game theory, randomness is often an optimal solution. If you are predictable then something can make use of that predictability to eat you. Better to be unpredictable then. At the same time, there can be advantages to predictability. If everyone you met behaved completely randomly, life would be impossible. There is a tension between the two.

Randomness also seems to be at the heart of whatever free will might mean. How free is free will if your actions are predetermined? That is to say, that the idea suggests a range of possibilities and that the precise choice is not knowable in advance. How strictly it is unknowable is an interesting question. By introspection, I might say that I don’t know where exactly my own decisions come from (or that I know to varying degrees). I might say something I believe to be spontaneous only to have my spontaneous utterance received with a big sigh and heavy rolling of the eyes: “I knew you were going to say that John”. Free will emerges as an idea in our interactions with others and our awareness of ourselves4 and the extent to which the actions of each of these are knowable or predictable. Our knowledge of others is strictly limited so we can do no better than probabilistic determinations (even in a deterministic universe). The apparent (or actual) freedom comes in part from the interplay between predictability and unpredictability. The former we might ascribe to the character or personality of the person whose will it is, the latter is the source of their freedom.

Deterministic or not, I feel we would still have to deal with the question as restated by the authors. A deterministic universe would still seem indeterministic to any finite agent living within it. That being said, their argument against determinism from a physics perspective seems strong5. I just think that we can pose their questions without the necessity of the lengthy argument.

How can agency emerge in the presence of such indeterminacy? This question seems to have a simple answer once we get past why there is order at all, and that is: evolution6. The first question is more difficult “where does the control come from?” Is the answer simply evolution and learning? By which I mean everything that happened to a person and everything that happened to their ancestors, which would cover all bases.

I’ll be interested to see their follow up papers on all these matters and morality because I can’t see a very important distinction between perfect determinism and randomness. Either way, agency (or apparent agency) arises from the interactions of atoms. But as far as morality goes, we don’t punish atoms, we punish big groups of atoms called people. It’s generally considered unfair to punish someone if they could not have done otherwise. It also seems unfair to punish a person simply for being unlucky, because they came out on the wrong side of the cosmic dice throw.

Anyway, fascinating stuff.

-fin-

1. Amongst other less compelling reasons. ↩︎
2. That’s two questions, but I’ll let that pass. ↩︎
3. There are, of course, bad arguments that aren’t evolution, but either way we have complex organisms. ↩︎
4. I read somewhere that there is a species of small squeaky mammal that does not itself know exactly which way it will run when chased by a predator. If it knew, the argument goes, then it might telegraph its intentions in subtle ways that a predator might pick up on. Better that the lunges and swerves come as a complete surprise to all parties. ↩︎
5. There are some interesting arguments in there that I’d not previously seen about the fundamental fuzziness of even classical mechanics. The argument is that a finite volume cannot contain an infinite amount of information. It seems to hinge on whether the location and characteristics of a thing are information. On the face of it, it seems like the argument supposes that our representation of the thing (a list of numbers detailing position, velocity etc) and the thing itself (a thing in space and time) are identical. This has always troubled me about physics: we have the thing, we have data about the thing and we have the laws that it is bound to obey. That seems like two things too many. Three if you allow that the formulae that describe those laws and act on the data are somehow separate. It seems wrong somehow that the universe would be divisible in this way. ↩︎
6. This is kind of answering a question with a question. It’s an observed fact that agency exists at the end of an evolutionary process, but what within that theory leads to agency particularly? ↩︎