Physics, Reductionism, a...

Physics, Reductionism, and Emergence

(1:20:02) Anomalies. I think people haven't found those anomalies yet, but at least it's subject to an empirical test. You know, at any time in the development of physics, we think that there are certain things we can safely predict and certain things we can't. You know, once Newton had figured out the laws of gravity, we could predict what was going to happen in the solar system to a pretty good approximation. These days, we have quantum field theory. We have what we call the Standard Model of particle physics, which is the theory of electrons and quarks, electromagnetism, and the nuclear forces, and all that stuff. And we also have Einstein's general theory of relativity, which is our theory of gravity that works not only in the solar system but also in the expansion of the universe, for example. Together, these are called the Core Theory: the quantum-field-theoretic version of the Standard Model of particle physics plus general relativity.

It doesn't work when gravity becomes really strong. It doesn't work for black holes or for the Big Bang. But in our everyday lives, it works, and as far as we can tell—in fact, we have very good evidence—it's true. The Core Theory governs absolutely everything in your everyday experience. So anything that you personally have ever seen with your eyes, touched with your fingers, etc., is completely compatible with the predictions of the Core Theory. And we have very good arguments, again, that say that if there were anything outside the Core Theory that could possibly have an influence on you as a biological organism obeying the laws of physics, we would have noticed it by now. We know how to make new particles, new forces, and new phenomena that could couple to us, interact with us strongly enough to matter for our everyday lives. We've looked for them, and they're not there.

So, as far as the fundamental laws of physics governing our everyday lives are concerned, we're done. We have those figured out. The new laws of physics, which will absolutely be necessary, are necessary for the cosmic view: for when gravity becomes very important and for when energies become very high—not for our everyday lives.

Reduction is a good thing. At an operational level, science works in a reductionist way for good reasons. Meaning that we have a complex problem, and what we're going to do is break it down into pieces. We're going to try to figure out the individual pieces, which is easier, and then hopefully we're going to be able to put together those pieces in a coherent whole. That's how science works, ever since at least Descartes, and it's obviously worked very well. But that's a methodological kind of reductionism. It's not a metaphysical version; it's not a strong reductionism of any kind.

Strong emergence is defined as the appearance of truly novel behavior once systems become complex enough or interact in a complex enough fashion. It's hard for me to explain strong and weak emergence because, to me, emergence is weak emergence. That's what I mean by emergence, even though I love emergence and I use it all the time. I love the weak form of it. The weak form of it is basically saying there's another way of talking about things at the higher level, but there's not something there at the higher level that is completely different and incompatible from what was going on at the lower level.

Roughly speaking, the idea of weak emergence is: in principle, I could take what happens at the lower level, put it on a computer, give it the laws of physics, simulate what's going to happen, and go back up to the higher level. That's the same thing that would happen at the higher level itself.

Strong emergence says I can't even use a perfectly powerful computer to simulate what would happen, because the higher level is just intrinsically different. There's something that emerges that can't be reduced down to stuff going on at the lower level. In my theory, I look at human beings through dynamical systems theory, which is a theory about how various complex systems in the world operate. I think of the self here as an information-responsive, complex dynamical system. Dynamical systems are now well known in nature, and they are systems in which higher levels of complexity make the system capable of exerting certain kinds of constraints on their parts, including the atoms. So, the complexity here actually can have downward causation.

I definitely believe in downward causation in this sense. Is there downward causation? I would say yes. Your thought process can change your brain, even though that requires a brain. I think there's a famous person who treats obsessive-compulsive disorder, and he's done experiments on this. He says he'll train people to think about their obsession in a certain way as a way of treating themselves. He goes back and records their brain; their brain has changed. So, yes, I'd say there's downward causation. Now, how that works, I have no clue, but I believe it does.

Downward causation is one of these concepts that gets involved a lot when you talk about strong emergence: the idea that there are things going on that need a higher-level description to be captured, and they affect what's happening even at the lower level.

Personally, I don't think there is any such thing as downward causation. Examples that are typically given are: if you want to understand how certain molecules of water get arranged into a snowflake, you have to understand the higher-level, six-dimensional hexagonal symmetry of the snowflake. I think that's simply false. If all I knew were the fundamental laws of physics, that water molecule would obey the fundamental laws of physics. It might be convenient—it might be extraordinarily explanatory—for me to say it's part of a snowflake, but if I were Laplace's demon and just knew what every molecule was doing, I would predict that snowflake just as well without any new stuff. So, I don't think that, in the real world as we currently understand it, there's any reason to believe in strong emergence or downward causation.

My answer is, I have no idea. I have no idea how it works. It's a very unusual thing in the world. The brain is the most complex entity that maybe ever existed. Consciousness is somehow dependent but different, and they could discover, well, here's how the brain gets into the state where you can be conscious. I don't think that's known yet, and there's no contradiction there. But it's got to be an emergent property.

I think that it's really important to appreciate what it would mean to have emergence that was more than just weak emergence. Because if I have the laws of physics, as I think we understand them in the Core Theory, and I believe that you are made of a collection of atoms and forces, then that theory makes predictions for what you're going to do. And the question is: do you think those predictions are correct or not? Those are the only two options, right? I think that those are the only two options here. If you think those predictions are correct, then there's no such thing as strong emergence; there's no such thing as libertarian free will. You just admitted that you obey the laws of physics. If you think those predictions are incorrect, then you're saying that the laws of physics that we use to describe particles and forces are wrong somehow. So tell me how they're wrong. That sounds really, really hard to do.

One of the features that we've learned about physics is that you can't just mess with the fundamental dynamical laws. Once you break them a little bit, they tend to go completely haywire. So, it's much harder than a lot of non-physicists appreciate to just say, well, there's something at the higher level that creeps in that is not there at the lower level of particle physics. That's a really big burden to meet.

When we talk about free will, we're really talking about conscious intention. Because if I say to you, you know, your unconscious has free will—you know, your brain is deciding; you're the last to know about it—nobody likes that concept of free will. They're looking for that conscious feeling of, "I've actually am the master of my own decisions."

I think the problem with libertarian free will is that it involves a misunderstanding of causation. Take a coin flip. It's our very paradigm of a random event, but of course it's caused. It's caused by factors that include the location of every atom and electron in the visible universe. Many events in a deterministic world are chaotic. They're not controllable, and they aren't controlled, and it's wrong to think you can identify the cause.

So, I think what the libertarians are confusing is thinking that their wills can't be free if their wills are part of the causal fabric of the universe. But no, that just doesn't follow. In a deterministic universe, some things are determined to be chaotic, some things are determined to be self-controlled, and some are determined to be remote-controlled. But if you are determined to be self-controlled, then you're self-controlled, and that's what you would want. Your fervent wish should be that your brain should be caused by the information streaming into it to make the decisions that, given your needs, desires, wants, and quirks, most want to serve your purposes. What could be freer than that?

Compatibilists redefine free will as self-control. For instance, philosopher Harry Frankfurt does this. He comes up with this notion of first-order desires and second-order desires, where my first-order desire is to eat this cookie right now, but I have a second-order desire to, say, lose weight. If my second-order desire wins out over the first one, then that's me exercising my free will.

Of course, that just pushes the question back: what determined my second-order desire? Right? That doesn't get you out of the problem; that just leads you to an infinite regress, where you're positing a thousandth-order desire by now that is itself explicable in terms of causes that you didn't have control over.

If determinism is true, our actions are inevitable as a result of factors beyond our control. And how could I deserve blame or punishment for them? The compatibilist thinks that compatibilism consists of the fact that we're immensely complex in our mentation, so that we have to take in a number of different inputs before we give an output, which is our decision or our behavior. And somehow, according to Dan, this constitutes free will. He's never specified what level of complexity it has to be. Chess-playing computers can do pretty much what humans do. They take an enormous number of inputs, and they do these complicated calculations about, well, what would happen if you did that—the same thing that humans do. And in fact, these computers are pretty much doing what humans do, and they do it better than humans because they can beat most humans now. But nobody says that these computers have free will. So, I'm curious as to why Dan thinks that a complex nexus between inputs and outputs in a brain that's evolved to be complex constitutes some kind of agency.

There is a socially constructed notion of free will, which is just as real as, say, dollars. Notice that we had to start off with gold and silver as our coinage. It had to be something that people felt had intrinsic value. It doesn't have any particular intrinsic value, and we've learned now that you don't have to have gold, you don't have to have silver, and you don't even have to have dollar bills. Way over 99% of all the money in the world is simply virtual money. It's on your credit card, and it's information. It doesn't have any particular physical presence; it can be in any form you like.

Does that mean that people should abandon money or that money isn't real? No. Money is real. It's a social construction, and the rules regarding it are very, very potent and very important. Very few people would want to do without it. The same thing is true of free will. I like to think of it as membership in the moral agents' club. The moral agents' club is a social artifact, something we've created. We've created it for very good reasons, even if we haven't understood what those reasons were. It's what permits us to live in security.

So, what we do is we set up moral presumptions and rules and make it very clear to our fellow human beings: you want to be a member of the moral agents' club? Here are the rules. And if you don't obey the rules, you're going to be punished. We will not consider the fact that you were determined to do that as an excusing condition, because it isn't. Excusing conditions are about specific defects of causation, not causation in general.

(1:34:34) Let me start by saying compatibilists have done a lot of good work, I think, on identifying different capacities of agents that give them different degrees of autonomy or control. I think these are very important distinctions, but I think they're important in a different place than the compatibilists. I don't think these distinctions are enough to ground belief in free will and belief in basic desert moral responsibility.

Okay, a common theme among free-will skeptics is to say that nobody ever deserves any punishment. And I say, well, you know what about red cards in soccer, or penalty boxes in hockey, or flagrant fouls in basketball? Are we going to get rid of all of that because, after all, nobody deserves anything, and nobody could help—no, no, no.