Does chaos theory square classical physics with human agency-

Metadata

Notes

Immediate thoughts

The point of this essay is to pick at what we mean when we say "cause".

In one sense, yes, since, perhaps in the deterministic world, if you changed the butterfly so that it didn't flap it's wings, perhaps the hurricane doesn't happen, or perhaps I didn't eat a brownie.

But in another sense, the answer should be no. First of all, in an intuitive sense, the answer should definitely be no. Why do we have that intuition?
Perhaps because, when we ask the question of: "how can we influence X to occur?" we don't think of stomping on a butterfly (unless one is deranged, masochistic, or both; or, of course, if the X in question is "the death of an innocent creature"). In the case of eating a brownie, well, the direct cause seems to be the decision to eat a brownie. While for stopping a hurricane, well, maybe we don't understand that well enough to assign anything real causal value (if we knew how, we'd be doing it).

Deeper thinking

So, coming back to the original question of, what is causation... When does X cause Y? One answer is via this intervention based approach to causal inference: X causes Y when doing something to X changes the probabilities of Y. At an operational level, this is the kind of causal inference a lot of scientists want: e.g. economists just want to know if an intervention is going to have some desired effect.
However, even this intervention based approach, I feel, might not be the philosophical answer we want. For instance, with all this behavioral economics, like nudge, we've seen that small interventions like placing healthy foods at eye level lead to higher rates of eating healthy food. Was this product placement the cause of me eating my veggies? At an interventional level, one would say yes. But maybe, if we'd like to hang on tight to free will, we'd like to say, no; that instead, it was still my brain state, my decision, which was the real cause.

Questions

Followups

Have the following connections/ideas been studied?

Highlights

Tiny variations in a system’s conditions, like a seemingly insignificant missing nail, can accumulate and trigger an unexpected shift – the shoe falling off, the horse being injured, the battle being lost. These sudden changes, surprising transitions within the system, are driven by underlying physical laws, yet they reveal the inherent unpredictability and complexity embedded in what might seem like straightforward events.
Note: But to what extent is this true, when one looks on average? That is, say, in the case of a battle, one may ask the question of: on average,
does a missing nail have a big influence on the outcome of the event? That is, randomizing all the other factors, how big of a change in the average outcomes does this missing nail induce?

Relation to influence from ABF?

The tension compels us closer to Russell’s radical conclusion – if nearly everything influences everything else, the word ‘cause’ begins to lose its meaning.
Note: That's why I think you need to think of all of this as average case things. To think of the causes as the variables with high influence.

Is there something theoretical to say here?

In this view, ‘causes’ are seen as ‘handles’, things in nature that provide us with a measure of control.
Note: Reminds me of the do calculus from causal inference: causation defined via interventions.

We may want to avoid being injured or getting ill, and our interest leads us to ask a specific set of questions; in turn, this leads us to the fact that much of the world becomes irrelevant.
Note: The point being here that the set of questions we ask are ones in which there's concentration of measure.
In this sense, concentration of measure is sort of like error correction.
Since another perspective on this is that, the question of "how is my health" is robust to small external changes, like where a particular butterfly was some day.

But, it actually is still the case that chaos theory implies that a butterfly flapping its wings change a thing that I see and therefore changes the decisions I make down the line. And therefore, changing my health.

What's important here is that the cause factors through my decision making. Ultimately, the true causes will be the actions that I take. In a world of free will, these are the only "true causes." In a deterministic world, there are other causes, but one might still be able to objectively draw out some implication diagram and see that there's something of "foods Daniel eats" through which all other causes factor.

Is this a new way of defining causation? Via a notion of universal propertys? That the root cause of something is the variable through which every other cause factors?

And while the butterfly could have an effect on a storm, it does not make a difference to the occurrence of storms in a way that we can predict or control.
Note: This seems key. This notion of prediction or control. This I think is related to what I was saying about average case stuff, since in the study of influence, one way of seeing it is as, here are the voters you want to flip to most shift the odds of the election.

This raises an interesting case of information vs computation: perhaps with all the information in the world, we could perfectly choose the best things to change. And perhaps, even in cases where the information we have is enough to determine the relevant "handles," it's computationally difficult to figure out what those handles are. [related:: maximizing influence in an ising model]

Maybe this is even a fundamental, physical thing. That is, it's physically impossible for an "entity" to be able to fully characterize the handles it can pull to lead to a desired (low energy) state. physics of computational complexity

Exploring the dichotomy between the perspectives of physics and human agency uncovers a paradox: our actions are simultaneously bound by the determinism of physical laws and enriched with intention, purpose and meaning that go beyond them.