The chapters are apparently independent, I skipped ahead as I was curious what he was going to say about free will.

He gives the argument that in a deterministic universe (with some randomness), there is no coherent theory in which we have a sort of "naive" free will (my term), where you are stepping outside the chain of causation to decide to do something that could have happened otherwise. Even though it sure feels like that's exactly what we do, hence the complete unwillingness for many people to even entertain this line of reasoning. Of course, we have no choice to entertain it or not! it's difficult and often paradoxical to try to think about all the implications.

But Blackburn wants to salvage some notion of free will, his argument is similar although I think a little different from Daniel Dennett (I mentioned Dennett's debate earlier with Sam Harris, where Harris is taking a hardcore determinist position.) I'm not sure Blackburn is entirely successful.

You can have a variety of free will within the chain of causality. I read all of Daniel Dennet's arguments in varieties of free will worth wanting, and may have to go back to it and re-read it. I'm not sure if he touched on this yet.

First, the chain of causality isn't a chain, but a web so complex that we still don't fully understand it. From what I can tell, we can build worldviews that understand this web to greater or lesser degrees. What such worldviews allow is a prediction of the future web of causality. We can see a small distance into the future, and we do it all the time, via predictions of all sorts. Based on what we predict, we can modify our behavior. There is something to this idea that I can't pin down, but whenever I try to conceptualize it, it's as if the web of causality has "eddies" in it. I'm a hardcore determinist like Harris even still.

The thought experiment I ponder is that of an advanced supercomputer that can predict the future with great accuracy. It can give a prediction, but that prediction will change the future. If this computer can also hold a model of what it wants the future to look like, it can give a series of predictions that lead to the fulfillment of the model. In this sense, it can mold the causal web even while being subject to it.

We do this same thing, to a very small and localized degree. I love how difficult this is to ponder, I come back to it constantly.

Interesting thought experiment, I'm not sure if you can say the computer will change the future, or rather, just like anything else in this debate, it depends what you mean by that.

Maybe the sheer computational difficulty makes it impossible even in theory, given the number of possible connections in one human brain alone, let alone the rest of the universe. Obviously we can make predictions on imperfect information, but that's a different sort of "prediction."

I was going to say that this is kind of like time travel, in that thinking about seems to lead to paradoxes. Or maybe we just don't have a good enough way of thinking about it.

I was going to replace the word "change". It doesn't work. It's curious that we can deliberate possible futures, and select between them. Causality becomes not only something we're subject to, but also a tool we can use. I sense that there's a great idea itching to break out here, but it's fiendishly evasive.



I don't think that we do, and that's why it's so fascinating.