Tuesday, March 27, 2012

Unpredictable God, Unpredictable Me

I was discussing free-will and determinism with some friends. I have a lot to learn about what others have written on the subject. It has been suggested I look into Compatiblism, in particular, but I put down some thoughts while they are fresh.

We were viewing free-will and determinism in the context of our observable environment, and some interesting ideas were brought out. I'm inclined to believe that if everything we are is determined by completely predictable natural laws, then all we have is the illusion of free-will. A friend argued that if our actions are ultimately governed by random events (such as quantum events) then our choices are even less free, in that they are not really our choices at all. We just flipped a coin and did what it said. I still struggle to see true freedom in either case. If our choices can be predicted in every detail in any practical sense, then we are not choosing, but only following a predetermined course. This still seems an inevitable conclusion to me from total predictability. At the same time, I'm unwilling to claim that God exists outside of natural laws (even though I readily admit I don't come close to understanding all of those laws), so I don't like my previous proposal that free-will is some super-natural control over the interface between random events (quantum events?) and apparently deterministic, macroscale events.

Continued discussion brought out thoughts from another friend that have given me hope of finding a solution that satisfies me. I'll quote, without permission, some parts of the exchange:

CB: "That is (to put this in more prescriptive terms), if non-determinism can beget determinism [e.g. random behavior of individual photons resulting in predictable behavior of many photons], can determinism beget non-determinism?"

JLC: “We call that chaos theory, and the brain certainly qualifies. But it appears that determinism can't beget non-determinism, it can only beget EFFECTIVE non-determinism. Something that is effectively non-deterministic is deterministic, but its behavior is so complex, that the best way to figure out what it will do is often just to let it do it and see. So this is a type of non-determinism in the sense of 'predictability' but it is still deterministic in the sense of whether it would do the exact same thing if placed in the exact same situation.”

CB: “. . . would you agree that taking a descriptive rather than prescriptive position on (non-)determinism allows for non-determinism to emerge from determinism? And if so, what are the objections to a descriptive approach?”

JLC: “Yes, I do agree that you can create SEEMING non-determinism from determinism. Pseudo random number generators in most programming languages do exactly that.

“My only objection is that such seeming non-determinism doesn't provide the type of freedom that Jonathan and others like him seem to want. If I seed the number generator with the same seed, I get the same result every time I run it. It SEEMS non-deterministic, until I start playing with re-running the program and playing with the seed.

“This doesn't bother me for my definition of freedom, but it would bother a Libertarianist. Because ultimately there IS something fundamentally different between true non-determinism, and apparent non-determinism. Just because it is CURRENTLY undetectable, does not mean that it would remain undetectable. If I can create a simulation of your brain, that always does the same thing on the same input, given the same seed, then suddenly the determinism becomes apparent, and the apparent non-determinism vanishes, even for a radical empiricist.

“I guess you can summarize that last criticism by saying that apparent non-determinism can vanish when we learn more, even for a radical empiricist.”

JGC: “Does it change something if you can prove that, while possible in theory, it will never be possible in practice to learn enough to perfectly simulate a sufficiently complex system? In that case could apparent non-determinism become effective non-determinism? Would that allow us to live in a world that is fundamentally deterministic but effectively non-deterministic from the point of view of conscious will?”

JLC: “Perhaps, but I don't see how you prove that about the human brain. It appears to me that you will be able to effectively simulate the brain in the next 20 years or so.”

So I've given up, for the time, my view of free-will that requires supernatural influence. I don't like that view. I am left to ask if free will might be found in effective non-determinism? Can the human brain, or the human being, be simulated to the point that a person's every choice can be predetermined? I'm inclined to think not. Some years ago I gave up the idea that I am in control of every choice I feel I should be able to control. Most, almost all, of my motivations and subsequent actions are determined by emotions and habits that, in a given moment, are out of my control. What I like to think remains in my control is the ability to shape my habits and emotions over time. I can make small choices that result in a happier, more productive me a year, or 10 years from now. Experience forced me to give up the belief that I am capable of making any choice I want at any given moment in time. I'm highly predictable, but I cling to the notion that I am not totally predictable.

So what happens in 20 years when my brain can be mapped so that any input given it will show with 100% accuracy what outputs will result? Either I give up my illusion of will, or I conclude that my brain cannot be perfectly simulated. I may live in a deterministic reality, but my brain is effectively non-deterministic—the only way to know absolutely what it will do is to start it going and watch. How can this happen if I concede that my brain can be duplicated? If I allow that chemistry and engineering can advance to the point that my brain could be simulated down to the atom, and that every neural impulse that determines what I do can be copied perfectly to respond to every input in exactly the way I would? How can I maintain a belief in effective non-determinism? I want a cosmology that strictly obeys natural laws, but I want to be me. I don't want to be predetermined. I concede that I'm highly predictable, but I want to be at least effectively unpredictable—at least a little bit.

The start of my hope is hidden in the very claim that we will be able to effectively simulate the brain. Simulating the brain itself is not really sufficient. My brain—a physically finite object that takes in measurable inputs and produces measurable outputs—isn't really the system we're interested in. We want to know how I will interact with the world. My brain is a open system that can assimilate a huge variety of inputs. To effectively simulate my brain and my future choices, you must be able to effectively simulate all of the future inputs to my brain. This very quickly becomes a computationally intractable system as you try to simulate more and more inputs, possibly requiring more computing power than could be harnessed by turning all matter in our universe to the function of computing the possible inputs and outcomes. Maybe the best way to see what my brain will do is to make my brain and let it run?

I can see one strong objection to my proposal. There are bound to be many irrelevant inputs to predicting my behavior. In fact, almost all inputs are irrelevant—most are too far away or otherwise undetectable to my brain, and many that my brain does detect it ignores. This leads me to conclude that almost every action I will take will be predictable by other humans in the not too distant future. What chance remains that it won't be every action? Is there some real hope that I am at least a tiny bit unpredictable? I invoke the web of human relations to maintain my unpredictability. Not only must my brain be simulated, but every other brain that is going to give me inputs, and every brain that is going to give them inputs that will influence the inputs they give me. The only practical approach might be to make our brains and see what they do. Or maybe we can posit a day where we can simulate using more power than is in our known universe? Maybe God can perfectly predict all my actions, and my free-will is limited to my finite perspective, and is predetermined in God's eyes?

I venture out into the realm of the unmeasurable. I believe God and spirit are physical. They are made of matter and energy like everything else that exists. For reasons given elsewhere, I believe that we are in a stage of development where we are unable to detect this type of matter or energy except by its consequences on some of our complex, subjective, measurement devices—our minds, hearts, and actions. Here is a set of inputs my brain simulators can only simulate by random guesses. They can trigger all of the religious inputs of my simulated brain and record every action that might result, but they have no way to predict what inputs God and the angels will send me. This still only pushes the predictability back a level. When we become gods and angels we will be able to measure and predict these inputs. 

I'm not convinced it is true in any practical sense that even God and the angels can predict all of the inputs. To predict the inputs into my human brain, you must predict the inputs into the brains of God and the angels that will influence me. To predict those inputs, you must predict an ever expanding web of relationships in an ever expanding web of universes. Does God know what I'm going to do? Pretty much. Does God know everything He will do in the future? I like to think He still has choices. If He might do something unpredictable, until He does it, He doesn't know with absolute certainty what I will do. Do I know what God will do? Sometimes. He's pretty predictable. But no more than I am.


  1. Johnathan, I really appreciate this post. Wonderful thoughts. Personally, I have a suspicion that genuine free will is an illusion. But I think that it is pragmatic to pretend that genuine free will is a reality.

  2. I think I'm admitting defeat for my gut feeling that I have some will not dictated by natural laws. In practice, I assume I have some free will and don't worry about it. It's really only an intellectual dilemma for me. Has someone written down some of your reasons for believing in free will even if it is an illusion? They might give me an alternative solution to my dilemma.

  3. Jonathan -

    I also quite enjoyed your post. All I have time to add at the moment is that with all the reading and thinking of done about this, I honestly can't see the reason why we would actually doubt our actual free will. I just don't see compelling reasons why we should believe that our actions are ultimately predictable by natural laws (rather than occuring in accordance with natural laws, as expressions of free will), nor that they are governed by random events (a la quantum events).

    I also think compatiblism is just a fancy way of saying determinism.

  4. I think you may be right, Leonard. Free will may be a fundamental natural law of existence, and thus unprovable and underivable. I would be good with that. The reason I don't just accept this outright is that neuroscience, psychology, and biology are still making progress by assuming that everything is deterministic. I don't want to pigeon hole free will into the parts we can't explain, because that would make it smaller and smaller every time we learn more about the workings of the mind. It feels like the God of the gaps argument in evolution.

  5. An approximate comment from Auden: "Give me a behaviorist and an electroshock machine and I'll have him reciting the Athanasian Creed in a week." I think this question of determinism is one where our human limitations become clearest. If I am forced to think that life is deterministic, then I will be forced to think that, and that's the end of the story. But the FEELING is that I CAN CHOOSE and that I have a moral obligation to choose, that I have no ethical right to accept any conclusion that absolves me for my choices. To me, THAT is the end of the story.

  6. Thanks for your comment, Dad. I personally take it as a given that I have some free will. That makes my post more an argument about what consequences the assumption of free will leads to if I mix it with some kind of physical determinism. And I suppose it is a rationalization of my belief in free will.

    Bertrand Russell talks about physical causal laws and psychological causal laws in a series of lectures I've been reading. I'm sure he wouldn't agree, but maybe free will is a psychologically causal law that is neither deterministic nor random.


  7. Very interesting to hear everyone's beliefs on this favorite topic of mine. We've started a survey topic on this issue at Canonizer.com, to track what and why people believe on this issue, concisely and quantitatively. So far, the compatibalists have taken an early consensus lead. It'd be great to get some of the current beliefs mentioned here included in this survey topic, (including the I don't yet know, camp, if so, and why...) so we can track them going forward as we all progress:


    Brent Allsop

    1. Probably no one else will see your comment, Brent. I, since, have come to feel like Leonard in that Compatibilism is just a semantic game to retain the words free will (which they recognize as existing perceptually and in everyday experience) while fundamentally rejecting it in favor of a deterministic bias. I'll take a look at the canonizer topic.