The subject of free will came up earlier this week, and I notice that Sam Harris has a new book on the subject, which I have not yet read. Some of you have asked for me to elaborate on my remarks on this subject in The Big Questions. Here are a few bullet points:
- We all have the subjective experience of deliberating among various options and making a choice. I claim that the existence of this subjective experience is not something we can be deluded about, any more than you can be deluded about a sensation. If you think you’re sensing the color blue, then you are sensing the color blue (though of course you might be deluded about what’s causing that sensation). If you think you’re deliberating and choosing, then you are deliberating and choosing.
- I take “free will” to mean that these acts of deliberating and choosing form a useful basis for making predictions about how people are going to behave. And it seems to me that is pretty much undeniably the case. I am virtually sure that if I know what you’re thinking, I have a better chance of predicting your behavior than if I don’t know what you’re thinking.
- By that definition, then, free will is a no-brainer. Of course we have free will.
- The question, then, is whether I’ve got the right definition. But I claim that I must have the right definition, because no other definition makes any sense.
- For example: You might want to insist that in order to count as instances of free will, my deliberations and choices have to be actually causal, as opposed to merely having predictive power. But what, other than having predictive power, could “causality” possibly mean? The history of the Universe is what it is — a seamless whole. The “cause” of Hurricane Katrina lies in either a cycle of evaporation, or a complex interaction among trillions of molecules, or the laws of quantum mechanics, or in the way those laws evolved from the Big Bang, or in the mathematics that governs those laws. It makes no sense to single out one of those as the “cause” and deny the others. Causality is a useful shorthand for describing “usefulness as a predictor” and it has, as far as I can see, no other useful meaning.
- Or, you might want to insist that in order to count as instances of free will, my choices could have been different. But again, I don’t see what that can possibly mean. Does it mean that my choices could have been different in this Universe? Surely not, since my choices are part of this Universe; if they’d been different, this would be a different Universe. Does it mean my choices could have been different in some other Universe? What does that mean, exactly? Does it mean that there is some other Universe whose entire history is identical to ours up to the point when I decided to turn down my air conditioning but is entirely different thereafter? If so, I don’t see what this has to do with “free will” as it’s ordinarily described. We are very far from understanding what other Universes might be out there, let alone in this sort of detail, and discussions of free will do not ordinarily presume that we’ve first got to settle questions of cosmology that appear to be unimaginably difficult.
- Or, you might want to insist that in order to count as instances of free will, my choices must be causal in a way that runs contrary to physical law, But that’s certainly impossible, because anything that happens in the physical Universe is by definition consistent with physical law.
- Or, you might want to insist that in order to count as instances of of free will, my choices must be causal in ways that run conrary to our current understanding of physical law. But that makes free will into something that flickers in and out of existence depending on the current state of scientific knowledge. I don’t think that’s what anyone means by it.
- I therefore stand by the assertion in The Big Questions that, according to the only definition that makes any sense, we quite obviously have free will.
- I note that Sam Harris, in his new book, argues that we don’t have free will. As I said, I’ve not yet read the book, but I’ve read enough reviews to suggest that he and I agree entirely on substance and differ only in the language we use to describe that substance. But I stand by the above: If your definition of free will requires it to flout physical law, or to be inconsistent with the known history of the Universe, or to be “ultimately causal” in a sense that nothing is ultimately causal, then of course we don’t have free will. By any other definition I can think of, free will is as much a part of the human condition as sensation or consciousness, and thus just as obviously part of the fabric of reality.
I am interested in whether you think a computer running a chess program also has free will in the sense you use it. And if not, what aspect(s) of a human deliberating and choosing are missing from the computer chess example which grants one free will and the other not.
You can watch a Sam Harris video for free. You favor free will and he is against it. I would say that you disagree on the substance.
Steve says >> “The history of the Universe is what it is — a seamless whole. The “cause” of Hurricane Katrina lies in either a cycle of evaporation, or a complex interaction among trillions of molecules, or the laws of quantum mechanics, or in the way those laws evolved from the Big Bang, or in the mathematics that governs those laws. It makes no sense to single out one of those as the “cause” and deny the others.”<<
I agree, and I think this is why free will is an illusion; the universe is a complex nexus of physical interactions that will run an inevitable course irespective of human will – and this includes the physical agitatons that make up the human mind.
That said, I think Steve is right to identify 'choices' because they do seem to be a genuinely true aspect of human cognition in that it does appear we make them. Perehaps for clarity's sake we could say that choices appear to exist in conjunction with perception and experience, and that they are an evolutionary utility that places nature's cosmological inevitability in the background of experience. To use an analogy, it is rather like actors on a stage playing out improvisations while all the time being aware that the overall performance and plot pertains to a script that they are not free to change ultimately.
choose not to breathe. even if you could hold your breath long enough to pass out, your brain would get you breathing again.
why suicide? who would ‘choose’ death over life?
What is the difference with “if you think you are seeing God, then you ARE seeing God” ?
I would argue for a variation of #6 — “free will” means that even if you made a choice at a point in time X, you could have made some other choice *without* violating the physical laws of the universe.
(I’m not sure if you meant something close to that with #7. You wrote for #7, “my choices must be causal in a way that runs contrary to physical law”, but I’ve never heard anybody argue for that definition of free will. As you said, it makes no sense since anything that happens in our universe has to be consistent with the laws of the universe.)
What free will means to me is that if a copy of our universe existed with all of the same physical laws, and that universe had followed the same course of ours up to point X, and then had proceeded *differently* after you made a different choice from the one that you did make, without violating the laws physics, then we have free will.
That’s what I think free will means, anyway. Under that definition, I don’t think it’s clear from the current state of physics whether we could have free will.
If all events in the universe were determined by interactions of subatomic particles that followed completely predictable paths in accordance with the laws governing their motion and attraction, then we could *not* have free will under this definition, because the position and motion of all particles in the universe would determine the state of the universe at any point in the future.
But quantum physics tells us that some events at the subatomic level happen nondeterministically. Maybe that means we do have free will. On the other hand, it may turn out that the outcome of those quantum events can be completely predicted by laws that we just haven’t discovered yet.
@EricK — I suspect one response to that would be that the computer (probably) has no conscious perception of free will, and therefore there is no argument there for free will, in the sense that we can argue we have free will because we perceive that we do.
To which *I* would say, that under my definition (free will means you could have made a different choice without violating the laws of physics), that one’s perception of free will might be an illusion, since you could perceive free will even if the motion of the particles in the universe are moving along a pre-determined path.
@James — I think it depends on what you mean by “the universe is a complex nexus of physical interactions that will run an inevitable course”.
If you mean that the *complete* state of the universe at any future point in time, is completely determined by the present state of the universe, then that could be true, although I’d argue that quantum physics allows for non-deterministic processes which means that it also might not be true. Certainly we can’t just assume that it’s completely determined.
If you mean that the final *end* state of the universe is already determined (either heat death, or a big crunch), then that could be true, but doesn’t tell us anything about whether we have free will regarding the decisions we make in the meantime.
About my previous question: never mind, I think you already answered with “you might be deluded about what’s causing the sensation”. Thanks
Bennett Haselton:
What free will means to me is that if a copy of our universe existed with all of the same physical laws, and that universe had followed the same course of ours up to point X,
I’m not at all sure one can make sense of this (though I realize I used the same quite-possibly-senseless formulation in the post). What does “up to point X” mean? Does your definition work only in a Universe with a preferred global time coordinate?
James Knight:
I agree, and I think this is why free will is an illusion;
Do you also conclude that a cycle of evaporation is an “illusion”?
If so, where do you stop? Is every level of explanation of every phenomenon an “illusion”, just because there’s always another level?
I’d much prefer to say that a cycle of evaporation is exactly what it appears to be: A convenient way of organizing my thoughts about the Universe that improves my predictive powers. Just like free will. And clearly, both exist in as useful a sense as anything at all in the physical Universe can be said to exist.
@Steven wrote:
“I’m not at all sure one can make sense of this (though I realize I used the same quite-possibly-senseless formulation in the post). What does “up to point X” mean? Does your definition work only in a Universe with a preferred global time coordinate?”
Are you talking about the sense in which, under the theory of relativity, two events could be “simutaneous” in one frame of reference, while not being simultaneous in another frame? So we need to pick a single frame of reference, in order to make sense of “point in time X”?
OK, then yes, in that case, pick any frame of reference you want — say, the frame of reference of the person who makes the decision at point in time X. That seems to me to be only a minor clarification and not a fatal flaw in the argument.
Then I’m saying we have “free will” ONLY if it’s possible to have two copies of our universe, with identical physical laws, where each universe proceeds through exactly the same physical states up until point in time X (as qualified above), and then there’s a diversion where the person makes choice 1 in one universe, and choice 2 in the other universe — all without the laws of physics being violated in either case.
From my paper on robot responsibility (publ. Miami Law this past spring):
Various philosophers have staked out positions in the general area of compatibilism, accepting both the notions that the universe is microscopically deterministic and that there is something real that our intuitive notion of free will actually corresponds to.
The best such theory for application to machines comes from computer scientist Drew McDermott. It is based on a cognitive architecture that is squarely in the mainstream of artificial intelligence research, and yet it is general enough to be applicable to a broad range of architectures, including ultimately that of our own brains.
Now let us consider a robot which chooses its actions by means of a model — a computational simulation — of the world around it. The robot knows that if it drops a pound of lead to the floor, it will fall rapidly and hit with a thud. It knows because it can imagine — create in its simulation — such a lump and run the computational experiment. The robot similarly knows what will happen if it drops a china vase, or knocks over a glass of milk, or jumps off a cliff.
Note that in order for this “world model” to be useful to the robot, it must be deterministic: it must reliably predict the results of actions in the real world.
Besides the model, the robot needs a “utility function,” a way to evaluate and compare states of the world as represented in the model. For example, scenarios including a puddle of milk on the floor might be scored lower than ones without.
As a part of the model of the world, the robot has a model of itself. In order to decide what action to perform in the real world, the robot runs computational experiments in which its model of itself performs a set of candidate actions in the model world. For each action, the resulting state is calculated, and the utility function of that state evaluated. Then the robot undertakes in the real world the action that resulted in the highest simulated utility.
This is in essence a standard “rational agent” cognitive architecture, but McDermott (2001) adds to it a remarkable observation. Consider what happens when the designers of the robot, having gone to the trouble of building a comprehensive world model, proceed to reuse the model as a resource for the robot’s intuitive ontological judgements. To answer a question about a property of something in the world, the robot consults its model of the thing and reports the property as modelled.
In particular, the robot will perceive the dynamics of its world to be deterministic, since that’s the way the model world works.
But the robot’s model of itself cannot be deterministic. The rational agent algorithm requires that the self-model arbitrarily attempt a set of different actions from exactly the same initial condition. From the point of view of the part of the robot outside the model, its inner model of itself must have a “control knob” that selects an action — a knob that every other object in the deterministic world model lacks.
… in other words, our perception of free will is the result of a heuristic optimization in our cognitive architecture.
SL I think this is what people mean by free will. I’m in a room deciding on whether to eat steak or broccoli. A sophisticated machine is outside the room scanning the position and momentum of each particle on the room (I think we can ignore quantum effects for this. )
Can the machine predict with perfect accuracy what decision I’m going to make?
That’s it. No parallel universes or any other extraordinary conditions.
Free will does not exist in the sense that there is no “ghost in the machine”. Our thoughts, emotions, perceptions are the result of purely physical biochemical processes. To use the illustration Harris does in the opening chapter, if the host and I were to trade bodies, atom-for-atom, I would be him. There is no “me” apart from my physical makeup.
Steve –
Your definition of free will (the conditions of which are easier to meet than other definitions, like Harris’s) might go a long way toward reconciling the question of free will versus predestination in Christianity, not to mention the penumbra thereof which extends to the question of moral responsibility. I’m not sure whether you’ll be pleased or not.
But maybe you can clarify a bit, since I might be over-reading you: suppose that an act of will ‘W’ is a perfect predictor of an effect ‘E’. In the presence of an orthogonal necessary condition ‘N’, you would maintain that W predicting E admits the presence of free will. An example might be the flight of an arrow. I intend to shoot an arrow at a target (W), but the cooperation of the laws of physics (N) are necessary for the realization (E).
But imagine another case: I intend to stay firmly planted on the ground (W’). But here the physical law of gravity is a sufficient condition (S), although not a necessary one. I accomplish my intent, but the sufficiency of S might put W’ outside your definition of free will. Does it or doesn’t it?
Bennett Haselton:
OK, then yes, in that case, pick any frame of reference you want — say, the frame of reference of the person who makes the decision at point in time X. That seems to me to be only a minor clarification and not a fatal flaw in the argument.
I think this is a much bigger issue than you think. Frames of reference are local. Only in very special (and probably unrealistic) circumstances does there exist a single frame of reference that covers the entire Universe. But your criterion really does demand that we consider the *entire Universe up to some time*. And the problem is not that there are multiple frames that pick out different “entire Universes up to some time”; it’s that there are *no* frames that allow us to do this.
Coupon Clipper: It seems to me that your “sophisticated machine” criterion is a criterion for determinism, not for free will. These are separate issues.
Thomasblair:
Free will does not exist in the sense that there is no “ghost in the machine”. Our thoughts, emotions, perceptions are the result of purely physical biochemical processes.
But…..of *course* these things are the result of physical processes because they are part of the physical universe, so this holds by definition. Surely we don’t want to define the problem away this lightly.
My complaint about the ‘no free will’ crowd is that they then usually conclude that we should never hold people responsible for their actions. This conclusion depends on confusing the ‘no ghost in the machine’ notion of free will with the way Steve defines it (or I do http://www.thebigquestions.com/2012/07/12/chain-reaction/#comment-58520). So the conclusion is just an error. But it’s a remarkably pervasive error. Indeed I saw Harris himself make it once, arguing we should change the law on this basis.
This is the problem with using the same, emotionally freighted, term for very different ideas. You see the same thing with alleged proofs of ‘god’.
In my experience, it is the lack of free will that is actually an illusion.
People who don’t believe in free will typically believe that if it were possible to construct a machine that was capable of analyzing every atom of our brains and every moment of our lives leading up to Decision X, that machine could predict the outcome of Decision X.
The problem is that such a machine already exists: It is the person facing Decision X. And that is the only machine capable of being built to analyze the full data set, and the only machine that will ever really do so.
In other words, OF COURSE the “no free willers” are correct about what they’re saying. In the hypothetical case that a machine could be built that would live our lives for us, that machine would make all the same decisions we would BY DEFINITION.
But that machine is a fantasy, and so is the idea that free will doesn’t exist.
KenB, as soon as someone says what we should do as a result of some theory, they are not really members of the “no free will” crowd. Either we end up doing it or not, and we couldn’t help but make that choice.
That aside, the conclusion that we shouldn’t hold people responsible doesn’t really follow from the lack of free will anyway. By holding people responsible (or acting as if we do) the inevitable result is a reduction in the type of behaviour we are trying to prevent. As the faxt that people will be held responsible is an input into the deterministic workings of their mind which affects the outcome. And this is the case pretty much no matter what definition of free will you use.
I wonder about the choice of the term “free will.”
Landsburg regards free will [alternative-selection mechanism] as akin to physical concepts such as momentum: observing these phenomena can help you predict the behavior of objects. Landsburg does not argue that the person chooses his will, or that the person has the power to choose to act in a manner contrary to his will. So does it make any sense to refer to “free will” but not “free momentum”?
Historically discussions of free will emphasized freedom, and consequences for moral judgments about people. I sense Landsburg denies that people have such freedom, and his use of the term “free” obscures his point.
RPLomg, I would ask you the same question as I asked Steve: Would you say a computer chess program is demonstrating free will when it is contemplating and then choosing what move to make? And, if not, what is the difference when a human contemplates and chooses a chess move?
EricK –
Of course the computer program is not exercising free will. The computer program is engaging in a series of pre-programmed algorithms that result in a move.
When a human being plays chess, she may engage in a particular series of analyses that have proven reliable from past experience, or she may choose to innovate an entirely new form of analysis, which may or may not be algorithmic. Alternatively, she may make a decision based purely on an emotional sentiment. She might even choose to make a move that a good algorithm would tell her is a bad move, if she thinks the move is a good strategic gamble or if she thinks her opponent isn’t paying attention.
More importantly, there is no way for any other human being or machine which decision-making *process* she will choose. And if she knows she is being analyzed, psychological experiments have shown that she will make another decision entirely, simply because she is being watched.
Computers, on the other hand, do not do any of the same things that humans do, unless humans specifically program simulations of those behaviors into their algorithms.
The two processes are completely different, and I’m surprised that anyone would consider them to be the same thing.
In fact, the more I think about it, the more I realize that innovation is precisely what proves that human beings have free will. A really good analyst may be able to predict that I will invent something, but will never be able to predict in advance what I will invent. If he did, he would have invented it themselves, and *he* would be the one exercising free will.
You can choose a ready guide in some celestial voice
If you choose not to decide, you still have made a choice
You can choose from phantom fears and kindness that can kill
I will choose a path that’s clear
I will choose freewill
-Neil Peart, Rush-
RPLong, How do you know the human isn’t just following a (more complex) algorithm which also results in a move (or sometimes results in them tipping over the board in frustration, or walking away and making a cup of tea or whatever)? Just as in your previous post, the only way to 100% reliably predict what a particular computer chess program will do is to construct a 100% exact replica.
I think the main reason you consider them fundamentally different processes is that you (and everyone else) don’t know exactly how a brain works, so we can construct the exact replica in the one case but can’t in the other.
EricK –
Two points:
1 – The burden of proof is on you to provide evidence for such a hypothetical algorithm. Until then, it is an act of faith on your part, and it would be unreasonable to expect me to believe that such an algorithm exists without any reason to believe in it, other than conjecture.
2 – It would be a remarkable algorithm indeed that could result in the invention of an entirely new algorithm. How might that work?
So SL are you arguing in favor of human determinism but also in favor of free will?
Steve: I think you an Harris are _basically_ on the same page: free will (I have a process in me that makes choices, these choices are not directly forced by agencies currently outside of my own mind, and knowing something about myself and my habits is roughly predictive) is obvious, while Free Will (whatever the heck else one might mean by that) is incoherent.
But in the book, he actually isn’t too interested that particular debate (I think because he finds the latter definition as preposterously incoherent as you do).
His main purpose is to undermining the idea that knowing your thoughts means “you” are in control of your choices, because he doesn’t think that rabbit hole goes very deep: we don’t actually, in the moment, understand the mechanisms by which our own thoughts and decisions arise. We can often give first order accounts of why we chose some path over another, or recount the course of our thoughts narrating our way to the choice but that reasoning doen’t go very far, because it doesn’t really explain what made the difference in any given case, or why our chain of thoughts proceeded in the way it did. He also relies on various neurological studies into the nature of choice: ones like Libet or the experiments in which scientists, via electrical stimulation, MADE a person’s body move in a certain way, after which, rather disturbingly, the subjects recounted that they felt like they had deliberately chosen to make the movement themselves.
His arguments go sort of like this:
///
I just thought of a purple elephant. I don’t know why I did. There is probably some explanation that has to do with particular readiness potentials in my brain for certain concepts (purple/elephant), and a whole host of other factors that favor one concept popping out or another at any given time. One day, science may be able to explain all of that in detail. But that doesn’t change the fact that I, sitting here, don’t have any direct subjective access or insight into that particular process as it happens. It’s going on behind the scenes, and I don’t know how it works.
I can “choose” to instead think of a red badger now, but the genesis of THAT choice is likewise obscure: maybe there was some part of me that is contrarian, and wants to provide a weak counter-example, and so thought of one. Or another part of me that just wanted to keep typing. But I can’t really explain _why_, in this particular instance, those drives happened to win out over a decision to stop typing based on the fear my audience might get bored if I go on too long, or the concern that I have other important things to do. They just did. In this case.
////
He also spends a fair amount of time talking about how we would think about morality in light of the various causes that can be predictive of particular choices. In particular, he points to examples of violent people who had brain tumors (in precisely the areas of the brain that regulate and prevent violent behavior) and weighs them against people who are just violent, with no known abnormalities in their brains. He argues that there is no fundamental difference between the two: SOMETHING in the brains of people likely explains why they acted violently (the tumor is just a more obvious case): both are chains of causality, and both are ultimately not “up to” the person making the choices in any greater sense beyond simply that being who the person is at that given moment, making the choices they make, which is a fairly trivial matter.
RPLong, Do you believe that all life forms exhibit free will? If not, how do you envisage it evolved, and what, in particular, is the difference between the brains of earlier organisms which don’t exhibit free will and those of their descendants that do? Do you believe the actions of the former could be replicated on a correctly programmed computer but the latter couldn’t?
I liked the argument I saw in favor of free will. I Sam Harris really believes that he doesn’t have free will, what’s the point of writing a book? By his definition, no one can be “convinced” by his book by his own definition.
The obvious response is that Sam had no choice, but to write the book.
I find this argument as interesting as the existence of god. Not very. It’s all speculation and no one can agree on a definition to test. It’s like asking if the color that I see when I see blue is the same color you see.
RPLong: can YOU predict what you will invent, but before you actually do so (i.e. know what the particular insight or connections that you will one day make will be, before you have them)? If not, then how is your subjective experience of yourself any more in control of the insight than any other random person? It’s a mystery to them beforehand. But it’s also a mystery to you beforehand.
“More importantly, there is no way for any other human being or machine which decision-making *process* she will choose. ”
Harris point is that you don’t really know which you choose, or why, either. There are all sorts of various drives in us, many of which are competing. Sometimes we notice something (like an experimenter watching us) and sometimes we don’t. And if we do, sometimes that might prompt a feeling of resentment that causes us to pick a different decision-making process than the ordinary one, and sometimes it won’t. But why in that particular case did some drives/perceptions win out over others. It’s easy to see in theory how a scientist could one day model you, and give a causal explanation that would make sense. And it’s not clear what else you think there could be to it.
Even if we start postulating supernatural (or, as Steve might put it: natural parts of the universe which we don’t currently understand) all we seem to be doing is tossing in more causal factors to consider in the model.
Is there any alternative to that? There is: truly uncaused events. But if we go down that road, then we’ve already given up on saying that you made a choice for which you are responsible for, because uncaused means UNCAUSED. Including any cause that has your name on it.
So you can posit the existence of truly uncaused choices, but you can’t then be responsible for them, because nothing that we can possibly identify as “you” could be the determining factor.
EricK –
Setting aside the question of free will, it would be impossible to program a computer to replicate the actions of living things, due to technical constraints. Even an amoeba engages in more actions and more processes than a quantum computer is capable of analyzing.
No computer can or ever will be able to replicate the actions of any living thing, period. For this to occur, the computer would have to algorithmically select from the list of all possible actions an organism of that variety could possibly perform and then also analyze the current environmental conditions, which also involves algorithmically deducing them from a list of all possible, and then analyze all potential strategic considerations in accordance with the extent to which that particular organism’s cognitive enables it to strategize, and then finally the computer would have to apply an algorithm for considerations involving the individuality of the creature and the totality of its past experiences.
Simply stated, the belief that any computer will ever be able to perform these kinds of calculations in any semblance of real-time is a bit of a fantasy.
And, really, I think that’s what drives people to the mistaken conclusion that free will doesn’t exist. It is possible for you to imagine a hypothetical computer that can engage in what amounts to magic. From the notion of this imaginary computer, you then extrapolate that such a thing could replicate your own personal thoughts.
As I stated in my original comment, such a “computer” already exists: the living thing engaged in the process being analyzed. Only living things can replicate what living things do.
Computers can sometimes provide a good simulation of what, say, an ant might do, but no one will ever really know if the computer had it right, because no one can step into the brain of the ant and see for sure. (And even if they could, they’d have to step back out of the ant’s brain again to make that determination, anyway.)
@RPRLlong- Is your claim a theoretical or a practical one, is it that no Turing Machine could never efficiently simulate a living organism or that we personally will never have the technology to build such a computer.
Drew –
First you say:
can YOU predict what you will invent, but before you actually do so (i.e. know what the particular insight or connections that you will one day make will be, before you have them)? If not, then how is your subjective experience of yourself any more in control of the insight than any other random person? It’s a mystery to them beforehand. But it’s also a mystery to you beforehand.
And then you say:
Sometimes we notice something (like an experimenter watching us) and sometimes we don’t. And if we do, sometimes that might prompt a feeling of resentment that causes us to pick a different decision-making process than the ordinary one, and sometimes it won’t. But why in that particular case did some drives/perceptions win out over others. It’s easy to see in theory how a scientist could one day model you, and give a causal explanation that would make sense. And it’s not clear what else you think there could be to it.
You seem to be suggesting that the unpredictability of my decisions makes it “easy to see in theory how a scientist could one day model” me. That seems like a non sequitur to me. How is my unpredictability a mark against free will?
You also suggest the concept of an “uncaused” choice. What is that? A choice by definition is a selection from an array of possibilities. What makes it “free” is not the fact that it’s unpredictable (using computer modeling or any other method), but rather the fact that it is my brain doing the choosing, not an inevitable list of “factors.”
As I mentioned to EricK, the belief that a computer could be built that could correctly predict all the actions of an amoeba is essentially a belief in a magic, hypothetical computer. There would be no identifiable difference between myself and that computer.
The argument is essentially a tautology: Hypothetically, if I could manufacture something that could think like you, it would think like you.
Well, yes. Of course.
“It is possible for you to imagine a hypothetical computer that can engage in what amounts to magic.”
Aren’t you arguing that there is something “magic” happening in the brains of living creatures? If not, then why is the idea that organisms are more complex than any conceivable artificial organisms that are modeled after them any sort of argument for a “free will” beyond the .
I’m also not sure I buy the idea that the causal chains for most choices are really all THAT complex. I don’t need to simulate the action of every single cell in my body to simulate the amount of information flowing into my brain. And I’m speculating here, but I don’t think every single cell in my brain is doing much in any given choice. Nor do I think that quantum effects play much of a role on the larger scale (certainly they do in the case of particular cell functions, but the combined _effects_ of these functions are likely a much smaller set of possible reactions).
Finally, I don’t think there’s THAT much going on in my mind in any given circumstance: in fact, part of what I find so unsettling when examining my own thoughts is just how finite and time-based they are. Thinking takes time, there’s only so much time put into any given decision, so I only REALLY consider a very few things from a very few perspectives. I’m just not sure that I see how achieving the same basic results is beyond the realm of possibility.
I think the main problem with brains vs computers is that for historical reasons, computers are built on a very different computational structure than brains.
Roughly, computers access, say, a “memory” by looking up pointers to it from an index, finding out where that data is located, and then accessing it directly. Human neurology/memory, on the other hand, seems to work more like someone sending out a wide broadcast of pointers to what they’re looking for all throughout the brain, and then relying on those certain parts of the brain that are best attuned to that particular package of stimuli piping up with the information (often with many different parts of the brain piping up at different volumes, and the result having to be integrated/filtered).
These are, indeed, very different models for how to store and retrieve information. But I don’t think there’s some physical incapability against us building computers that follow the brain model instead of the traditional one. That is, in fact, exactly what researchers are now doing with neural nets: http://www.ibm.com/developerworks/library/l-neural/
My wife and I conspired to create a machine that thinks. I don’t claim we were the first.
Mike –
I am probably making both claims. But I wouldn’t call it a “Turing Machine.” The issue is not merely to build a machine that we humans *think* models us. The issue is whether we can build a machine that *actually does*.
Suppose you built a Robot RPLong, stood him up next to me, and turned him on. Suppose he made a choice other than I would make. I submit that skeptics of free will would say that, because Robot RPLong and I do not occupy the same space, the environmental conditions are slightly different, and therefore affected his decision versus mine.
But would that be true, or would that be an excuse? How would we know?
“You seem to be suggesting that the unpredictability of my decisions makes it “easy to see in theory how a scientist could one day model” me.”
The crucial point is: the unpredictability of your decisions TO YOU. The hypothetical scientist examining your decisions would have a lot more to work with than you sitting there in a chair examining your own thoughts would have access to. You can see, in your mind, the effects (thoughts, decisions, etc.). They would be investigating the causes (what parts of your brain were activated, and why, based on what stimuli they were getting and which each part was most attuned to because of its particular state).
“The argument is essentially a tautology: Hypothetically, if I could manufacture something that could think like you, it would think like you.”
If that’s all you’re arguing then we agree: “free will” is a fairly trivial and obvious matter regarding a causal story about the brain as a mechanism that makes choices according to causal laws, and any grander sense of “Free Will” is nonsense.
You do make choices, and you are responsible for them. Exactly as a free-standing computer is responsible for the choices it makes. Take away you, or take away the computer, and replace it with a different person or a different computer, wired differently, and they/it would make different choices, because they are different in key, unsurprising ways that predictably lead to different choices.
But Drew, we`re back to innovation. Computers don’t innovate. They don’t create new processes. Humans do. You side-stepped this point by saying that I can’t predict what I will invent before I invent it, but regardless, it doesn’t change the fact that what I create is new.
Computers don’t make choices, they perform algorithms. They have not merely “evolved” to perform algorithms – that is literally, in point of fact, all that they are capable of doing. They don’t engage in free will because they can’t venture beyond their programming.
But humans do it all the time. If you don’t believe in free will, you are essentially suggesting that every new product of human thought was inevitable.
“Suppose you built a Robot RPLong, stood him up next to me, and turned him on.”
Is the new version of ‘postulate a can opener’? Postulate a robot Natalie Portman?
:)
What Steve said.
I’ll add that I have read Harris’ book, and as far as I can tell he adds absolutely nothing new on the subject. Like all incompatibilists, he keeps repeating the same old arguments. Maybe incompatibilitsts don’t have free will.
One incompatibilist line, repeated over and over again by Harris, is “free will is an illusion.” With an illusion, the difference between subjective experience and reality can be demonstrated. With Adelson’s checkerboard illusion, you can cut out the squares and place them side by side. With the Muller-Lyer illusion you can measure the two lines with a ruler and establish they are the same length. You cannot do that with free will, so I wish incompatibilists would stop asserting a fact they cannot demonstrate.
“But Drew, we’re back to innovation. Computers don’t innovate. They don’t create new processes. Humans do. You side-stepped this point by saying that I can’t predict what I will invent before I invent it, but regardless, it doesn’t change the fact that what I create is new.”
I don’t know what you mean about how computers cannot innovate. Of course they can, in exactly the same way we can and do: by mixing up old concepts into new ones, resulting in outputs never before seen in existence, or recognizing patterns that no one ever recognized before, and so on. Computer programs can even evolve, using evolutionary algorithms that in many cases come up with solutions to problems or even physical structures that human engineers never hit upon on their own. What is innovation if not that? Sure, they were programmed to work like that. But the programmers did not knowingly build in the specific solutions to problems that they themselves could not solve into the program. The ability to come up with the solutions are neat trick of the program.
I don’t see how this is a compelling argument for a stronger definition of free will in any case, putting aside the fact that you haven’t even defined exactly what it is you’re talking about (if your definition of what “free will” is different from the ones I’ve noted above).
“Computers don’t make choices, they perform algorithms. They have not merely “evolved” to perform algorithms – that is literally, in point of fact, all that they are capable of doing. They don’t engage in free will because they can’t venture beyond their programming.”
You haven’t explained, then what this “free will” thing is that we’re engaging in: what is it? How does it operate? What’s the mechanism?
I make choices, and think thoughts. These choices and thoughts appear to be fairly predictable given certain known aspects of my biology, my memories, and so on. These processes make perfect sense when described as algorithms, which is to say that I can tell you a causal story about how they came about, to varying levels of specificity (as I noted, a hypothetical scientist examining my brain is probably capable of telling the causal story in FAR more detail than I am): getting outputs from inputs. It’s pretty complicated, sure, but I don’t see how that changes anything. Making these choices, thinking thoughts, is literally, in point of fact, all that I am capable of doing!
So what?
“But humans do it all the time. If you don’t believe in free will, you are essentially suggesting that every new product of human thought was inevitable.”
In the sense that all events in the universe are predictable effects from particular causes, sure. There may well be non-deterministic aspects to the universe of course and the exact outcomes may be highly contingent and chaotic, thus very sensitive and very hard to predict. But those phenomenons are all highly explicable with and predicted by mathematical theory.
You’re losing me, Drew.
Your argument seems to boil down to the idea that because your thoughts make sense ex post facto, they must be modellable. The kinds of models you describe, though, are imaginary. No one has been able to predict a human being’s decisions. You are essentially claiming that some day, somehow, somebody will figure out a way to do it. This is basically a magical, faith-based belief of yours.
You start from a believable premise: Because you have good reasons for doing the things you do, then it should be hypothetically possible for someone else to collect all the reasons that are important to you and make the same decision.
But then you take it into Santa Claus territory by hypothesizing the existence of a model that *somehow* knows what is important to you at any given moment and weighs every conceivable factor accordingly, *just somehow*.
This isn’t a theory, and it isn’t scientific. It’s a repeated claim that, because something seems predictable, it must be autistic. Step One: Analyze the data, Step Two: Santa Claus waves his magic hat, Step Three: Human consciousness is modelled to perfection.
I think it’s really silly to demand that I define free will. It’s a multi-faceted term that we all understand full well. It is telling, though, that you ask questions like “What is the mechanism?” when speaking of free will. Free will isn’t a mechanism. Neither is sentience. Would you suggest that sentience is an illusion? (That’s a trick question – if you say that it is, you are performing a variation of the Liar’s Paradox.)
If I had to try to define it, though, I would suggest that free will is an act of thought that is both rational and uncontingent on prior impulses.
You will probably suggest that rationality is a prior impulse, but you have no hope of proving this. If you jump out of the way when you see a sudden flash of light, you are engaging in instinct, but if I ask you why you jumped away, you would give me a perfectly rational answer. The difference is a priori versus a posteriori.
You seem to be making the claim that all human thought is a posteriori but that it provides the illusion of a prioristic reasoning. My response is that if this were true, humans would never be able to engage in a prioristic reasoning to invent a completely new concept. They would merely be instinctually driven to a random reassembly of existing patterns, and then rationalize it as a new invention a posteriori. If this sounds ridiculous to you, then there is probably a good reason why. ;)
Neil: “One incompatibilist line, repeated over and over again by Harris, is “free will is an illusion.” ”
I think the debate about whether or not it’s an illusion isn’t really a very useful one, mostly because people tend to mean or interpret very different things into that word, and so often talk at cross purposes. It becomes a very semantic debate. And I agree that this is part of Harris’ own over-dramatic fault, as “illusion” is an evocative word, but not really one that makes as much sense when talking about subjective experience.
In his case though, his argument still seems to be just that the sense that our choices are fully explicable to us is false: that our sense of deliberating is not, in fact, the real deal when it comes to explaining the choices we make: it’s a constructed narrative. The sense of illusion he argues for is similar to thinking you understand evolution, when in fact your understanding of it is a super-simple caricature: avoiding the gaping logical holes in your belief. In short, you are aware of your thoughts and of your preferences and how they interact, but you’re not really able to explain how particular thoughts arise and not others, or why some set of preferences wins out over others in any given situation. So the real reasons for making one choice or another remain essentially hidden and mysterious to us.
I think he makes a pretty decent case for this view, one at least worth grappling with. Why can we explain the overall narration of our thoughts, but not why particular thoughts arise and not others? What makes us stop considering and weighing options, and make a decision? Why do brain events that predict decisions seem to arise prior to decisions? And why can we so easy SIMULATE the experience of making a decision for someone else, just by stimulating particular parts of the nervous system (and not even necessarily parts in the brain proper)?
As an aside on the question of what causes us to stop weighing choices and make a decision, Radiolab did a segment once on a man who suffered a brain lesion on the part of his brain that deals with emotion. While some effects of this are as expected (he expressed, and reported feeling very little emotion), the other effect was fascinating: he found it extremely hard to make even the simplest of decisions. He would deliberate endlessly even on tasks such as which pen to use to sign his name, unable to stop weighing options and considering possibilities. So instead of become basically a sort of hyper-logical Vulcan, he became basically unable to live a normal life. This doesn’t prove anything, but it does suggest that emotional responses play a huge role in the process of cutting off deliberating and just committing to a particular choice of action.
RPLong: I don’t think what I’m arguing is all that strange. I’m just saying that our decisions, like everything else in the universe, seem to be explicable causal processes. If not, I’d like you to explain what aspect of them could not be so, and if so, how it works or what is going on instead.
The idea that these processes can be modeled is, certainly, speculation. But it’s a direct implication of it being a causal process that exists, along with all sorts of other causal processes in this universe. No, we can’t simulate the entire universe down to the smallest quantum interaction. But we don’t need to in order to understand basically how it works and why.
I don’t see how you conclude that human brains are any different. What is going on in them that IS functionally different? You tell me.
“I think it’s really silly to demand that I define free will. It’s a multi-faceted term that we all understand full well.”
It’s not silly: it’s impossible to have a discussion about whether or not it’s a valid concept unless it’s first defined. Full stop.
I certainly understand what I mean by it, using a certain definition that I’ve laid out. But you seem to disagree, and say you have a definition that is different. I want to know what that entails. I think you understand my definition. But I don’t think “we all” understand your definition.
“Free will isn’t a mechanism. Neither is sentience. Would you suggest that sentience is an illusion?”
So what is it then: magic? But I thought you were dead-set against appeals to magic.
“If I had to try to define it, though, I would suggest that free will is an act of thought that is both rational and uncontingent on prior impulses.”
Right, but this gets us into lots of problems right off the bat. If it’s not contingent on prior impulses, what IS is contingent on? What determines it? And if nothing, how is it a WILL at all? A will implies a force directed towards something. But you seem to be describing something that has no particular direction at all.
“My response is that if this were true, humans would never be able to engage in a prioristic reasoning to invent a completely new concept. ”
I’m sorry, but I don’t think you’ve established that humans ever do create “completely new concepts.” In fact, every example of humans coming up with concepts seems to fit pretty logically into extrapolations from existing concepts, pretty tightly confined to the world they inhabit and explore, their available senses, and our ability to, ironically, model the world around us in our heads and play with the models.
This is a pretty long-standing debate in philosophy though. I.e. Hume’s belief that, say, the concept of angels is something we imagined by combining the wings of birds with the bodies of humans in our minds.
Computers can, of course, do the same sorts of things.
Steve Landsburg:
Thanks for the reply. Some comments:
>>Do you also conclude that a cycle of evaporation is an “illusion”? If so, where do you stop? Is every level of explanation of every phenomenon an “illusion”, just because there’s always another level?<>I’d much prefer to say that a cycle of evaporation is exactly what it appears to be: A convenient way of organizing my thoughts about the Universe that improves my predictive powers. Just like free will. And clearly, both exist in as useful a sense as anything at all in the physical Universe can be said to exist.<<
There is one key distinction though – a cycle of evaporation belongs in the class of objects we 'perceive' – it's a class that involves things like tables, chairs, plants, and empirical objects of that kind. Free will belongs in the class of objects we 'conceive' – it's a class that involves things like 'happiness', 'justice', 'team spirit', etc, which are not physical objects of study, but concepts that help us interface with physical reality.
Given the foregoing assessment, I think 1) We might find it useful to keep in mind a category distinction between objects of perception and objects of conception. And 2) Our world appears to be a world in which nature's inevitable trajectory brings about minds that construct a perspectival first person reality within that trajectory that leads them to believe they are making free choices, when what they are actually doing is operating within a mental matrix that disallows the vast majority of information implicit in this inevitable trajectory that makes up nature's grand cosmic story.
Steve, this means that you do genuinely present me with a problem to solve, because you believe that "both exist in as useful a sense as anything at all in the physical Universe can be said to exist" – which is difficult to deny. So what is the criterion by which we can separate them? I can think of one way. The grounds on which I would say 'free will' is not a fact and a 'cycle of evaporation' is a fact is that the latter is not inconsistent with nature's inevitable trajectory. In other words, a cycle of evaporation and a hurricane can be part of the inevitable trajectory, whereas free will cannot, because one is never 'free' from that cosmic trajectory. So the free will either is part of the trajectory – in which case, it is not really acting in any kind of freedom separate from the trajectory; or free will is not part of the cosmic trajectory – in which case it is claimed to be no longer part of nature, where, at that point it falls down by not being definable (as per my initial request for a definition). Either way, the idea of 'free will' gets us in philosophical trouble.
I understand Landsburg to argue that “free will” is part of the inevitable trajectory. To synopsize:
Mathematics determines everything Landsburg experiences. Landsburg struggles to understand the whole, undifferentiated manifestation of mathematics, but ultimately must resort to a simplifying strategy of analyzing parts of these manifestations distinguished from the rest.
As part of this cognitive strategy, Landsburg distinguishes among various branches of sciences, and among various types of objects. In particular, he distinguishes an object he calls Landsburg. And he distinguishes among objects based on how Landsburg might influence them. He finds that Landsburg may be able to influence certain objects via expressions of ideas/information; he categorizes these objects as having “will.”
Other than this attribute of being capable of influence by ideas/information, objects with “will” behave much like pebbles in an avalanche. They act on others, and are acted upon by others, by deterministic laws. (I see no advantage in characterizing their actions as “free.”)
Again, these objects – and Landsburg himself – are merely manifestations of the whole. While there are pedagogical reasons to treat objects as distinct from the whole, these distinctions do not reflect the nature of the objects as much as they reflect the nature of the mind that is struggling to understand the objects.
Drew –
When you say, “So what is it then: magic?” do you mean to say that you *do* consider sentience a mechanism? Merriam-Webster defines it as a feeling. You know, like “happiness.” Is “happiness” a mechanism? What *isn’t* a mechanism to you? And more importantly, is anything that is not a mechanism magic?
It would be a mistake to impute motives or purposes to evolution. Evolution is an accident – there is no “purpose” behind it, no “mechanism.” It simply accidentally occurs through cellular mutation. Therefore, “sentience” is not a “mechanism” that serves an evolutionary “purpose” any more than thumbs “evolved so that we could grab onto things.” If you feel otherwise, you have the causality backwards. That our hands mutated resulted in our ability to grab things, not the other way around.
Similarly, that our brains developed a given structure enabled us to become self-aware and engage in free will. They are not “mechanisms.” Nor are they magic. They are *attributes*, feelings, facts, characteristics of the human brain.
What is free will contingent on? I thought my definition clearly indicated that I did not believe it was contingent on anything other than thought. Similarly, sentience is not “contingent on” anything. These things are attributes of the functional human brain. They are contingent on the fact that the brain we are talking about is both human and functional.
You say: “In fact, every example of humans coming up with concepts seems to fit pretty logically into extrapolations from existing concepts, pretty tightly confined to the world they inhabit and explore, their available senses, and our ability to, ironically, model the world around us in our heads and play with the models.”
Neither have you conclusively demonstrated this. You (absurdly) claim that computers can extrapolate. In fact, they cannot. They can perform mathematical evaluations if we first give them the models. They cannot create entirely new models based on past observations.
In short, they cannot perform a priori reasoning. The best they can do is *simulate* a priori “reasoning” algorithmically, i.e. by running an existing model, saving the estimators as parameters, and then re-calling the parameters in a pre-defined equation. To call this a priori reasoning would be a serious mistake.
Of course, human beings have created plenty of things that did not exist in the environment beforehand. The examples are virutally infinite, and the best ones involve abstract concepts that have no direct analogue in nature. Since we’re on an economics blog, I’ll give you a great example: Money.
I think most people don’t really have a clear definition of “free will”, but in practice use it to mean something along the lines of “not like how a computer works”. And when presented with something that makes choices they see if they have an idea how it makes choices, and if they can explain how it does, they proclaim it doesn’t have free will, but if they don’t know how it does it (and suspect nobody else does either) then they impute free will to that object. In that sense it is a bit like a “God of the gaps”.
That being said, I don’t think this is how Steve is using “Free will”.
I think some light can be thrown on the topic by this observation: If I ask “Why did I get up this morning?” there are two distinct ways to answer that question. I can say that I got up because electrical impulses were sent to the muscles in my legs. And they arose because of electrochemical events in the brain, which in turn were triggered by other parts of the brain and by sensory inputs from the environment and so on. Alternatively, I can answer it thus “I got up because I wanted to go to work”. And I wanted to go to work to earn money, and not get fired etc. Both answers are true. When we consider the former answer, free will is nowhere to be seen; but when we consider the latter, free will and its close relative “purpose” are everywhere. Our conscious experience only has access to the latter type of explanation. The story it tells us is one of purpose and choosing. It does not have access to electrical signals or the actions of neurotransmitters.
So we feel we have free will, because our brains tell us we do. And we feel that others have free will because our brains model them in that same sort of way. And it is very possible that the latter came first in evolutionary terms. We evolved the very useful ability to predict what our friends and enemies would do by setting up models based on high-level concepts like “purpose”, and eventually turned those models on ourselves.
@Drew #47
I do not confine the information processing done by my brain in making a decision to the conscious part. The deliberating, running simulations, evaluating consequences that define free will occur throughout my brain. My brain can call up information to the conscious part as needed in this process. I see no conflict between free will and Libet’s experiments or retrospective consciousness, but others have covered this ground.
In short, when I say I have free will, I mean my entire physical self, not just my conscious brain. The subjective feeling is just the evidence of free will, which we can share, as Steve said.
Steve, let me respond to your post point by point:
1. I agree with you here, with a caveat: it is true that if you think you are perceiving deliberating and choosing, then you really ARE perceiving deliberating and choosing, but that doesn’t mean that YOU are deliberating and choosing. There is still a question of whether this experience is merely one being received by you, or whether you are the one who is doing the activity of deliberating and choosing (regardless of whether this activity has anything to do with decision making).
2. I agree with you here.
3. I agree with you here.
4. Here’s where I disagree, as I discuss below.
5. I think predictive power is a poor definition of cause. Seeing turkey eating on Thanksgiving is a pretty useful predictor of there being Christmas trees up in a month. But clearly Thanksgiving turkey doesn’t cause Christmas trees. I think a better definition of cause is the standard one used in philosophy, which goes as follows: event A causes event B if A happens before B, and B would not have happened if A had not happened. Note the use of the counterfactual here; I think the use of the subjunctive is very important for the discussion of causality, free will, and the like.
6. I don’t think the subjunctive “could have been” requires the invocation of other universes, although your view that it does has a name, modal realism. The opposing view is known as modal fictionalism, and it states that modal notions like possibility and neccesssity can be discussed without there being actual alternate universes. Instead, such notions are primitive, so that the subjunctive mood is not reducible to the indicative mood, if you want to think of it in terms of grammar.
7. I think there is a significant issue here that doesn’t boil down to a trivial definition. Certainly if you define a natural occurrences as an occurrence that occurs in the natural world, then all occurrences in the natural world are natural. But I don’t think you can get rid of the supernatural that easily. The philosophy known as methodological naturalism states that every phenomenon that occurs in the observable world has a cause which originates in the observable world. The truth of that philosophy is presumably dependent on what universe you’re living in, so we can characterize free will in terms of that: the belief that human actions are such that they are inexplicable in terms of causes that originate in the physical world alone.
8. I agree with you here.
Keshav Srinivasan:
Seeing turkey eating on Thanksgiving is a pretty useful predictor of there being Christmas trees up in a month. But clearly Thanksgiving turkey doesn’t cause Christmas trees.
But Thanksgiving turkey has no predictive power for Christmas trees after conditioning on other relevant variables. I should have been more careful about the way I formulated this.
A causes event B if A happens before B, and B would not have happened if A had not happened.
I can’t make head nor tail of this. Changing one event in the Universe changes the entire Universe, so that it no longer makes sense to talk about event B.
Does the number 3 cause the number 8? Well, if the number 3 didn’t exist, then the number system would be entirely other than it is and I have no idea whether “8” would have any meaning at all. In other words, I can’t make sense of the question. Questions that start with “If A had not happened” seem as hard to parse as questions that start with “If 3 were not a number”.
I have actually read the book and it seems to me you largely agree with Dr.Harris. However, the difference is that you are failing to see that the definition that you provided is not the conception that people have when they speak of free will. The definition you provided roughly falls into the “compatibilist” definition which is one that defines free will as the capacity to carry out one’s deliberated intentions. However, this bears “no relationship to the feelings of agency and moral responsibility that have made the idea of free will an enduring problem for philosophy”(Harris 23).
The definition that does bear a relationship to these feelings is the one that defines free will as being “the conscious author of one’s actions”. As you have articulated, you clearly disagree with this definition since you don’t believe free will is “causal”. However, you fail to understand that people seem to believe this. To illustrate this, I will provide two examples.
1. There have been psychological studies in which people’s brains have been scanned while performing simple behaviors. Their decisions were accurately predicted milliseconds, even SECONDS, before they were CONSCIOUSLY AWARE of the decision. Such findings usually shock people when they hear of the results.
2. Our intuitions of morality are based on an illusory notion of free will. The example Harris provided is the case of two psychopaths who kill a random person, in their own words, “just for the fun of it”. One psychopath, however, has a tumor in his brain the size of a golf-ball, which changed him from being a normal man to a psycho-killer. The other psychopath has no disruptions in his brain and has always been a psychopath. Most people would say the killer without the tumor is evil, while the other is a victim of a tumor. Both, however, have the SAME EXACT SUBJECTIVE EXPERIENCE. The psychopath without a tumor is just as much a victim of his neural physiology has the one with the tumor.
Essentially, as the neuroscientists Dr. Harris quoted say: ” Most people’s view of the mind is implicitly dualist…That is, it requires the rejection determinism and an implicit commitment to some kind of magical mental causation…contrary to legal and philosophical orthodoxy, determinism does really threaten free will and responsibility as we intuitively understand them” (Harris 74).
The way I have framed free will may have been ruled out by your causal point but I am not yet convinced of that.
If my thoughts and actions are a result of the physical processes taking place within my body, it must be concluded that since I cannot change the composition of my body, I cannot have free will. Does free will require supernatural powers, or just the perception of deliberation?
I have condensed my thoughts down to a few bullet points.
A universe with regularities could be exploited by a hypothetical entity which had genuine free will.
Given a universe with regularities and which contains a self-replicating entity without free will, the evolutionary process will give rise to organisms which also exploit those regularities (as they will out-compete their rivals)
The longer the process goes on, the more of those regularities will get exploited, and so the more those organisms will start to resemble beings with genuine free will
An intelligent being observing those organisms would notice that they appear to have free will
An organism which evolves the ability to observe itself would notice that they themselves appear to have free will.
This explains why we appear to have free will even though we also appear to live in a deterministic universe. It is evidenced by the increase in apparent free will as you move from simple to more advanced organisms.
Since Steve and Mr. Harris seem to disagree only on fuzzy semantics I propose that the language be improved. Everyone has “will”. An action is motivated by “free” will when it is rational given the individual’s preferences (hey hey). Otherwise it’s not-so-free.
Alex B:
If my thoughts and actions are a result of the physical processes taking place within my body, it must be concluded that since I cannot change the composition of my body, I cannot have free will.
If evaporation is the result of interactions taking place at the quantum level, must it be concluded that there is no such thing as evaporation?
Vic: I think there are clear alternate readings for both your points.
1) There have been psychological studies in which people’s brains have been scanned while performing simple behaviors. Their decisions were accurately predicted milliseconds, even SECONDS, before they were CONSCIOUSLY AWARE of the decision.
This does not speak to the question of free will; it speaks only to the question of the relation between free will and consciousness.
2) One psychopath, however, has a tumor in his brain the size of a golf-ball, which changed him from being a normal man to a psycho-killer. The other psychopath has no disruptions in his brain and has always been a psychopath. Most people would say the killer without the tumor is evil, while the other is a victim of a tumor. Both, however, have the SAME EXACT SUBJECTIVE EXPERIENCE. The psychopath without a tumor is just as much a victim of his neural physiology has the one with the tumor
Sure, but it’s still perfectly reasonalbe for us to believe that deterrence will be ineffective (and hence pointless) regarding the guy with the tumor, but very effective (and hence wise policy) regarding the other guy. That, I think, is why we treat them differently. (Though one could certainly argue in the opposite direction — that because the guy with the tumor is harder to deter, he should be punished *more* severely.)
I’m generally in Steve’s camp but I think his definition omits a key idea: information. So let’s look at this:
I mean I can create a model of you wherein you respond to differing information in ways I can only predict using the notion of mental states and information content.
I fully concede this is a wee bit vague. But it’s why Steve can have free will and a falling rock cannot. And the vagueness is a result of the fact that, like the evap cycle, it deals with imprecise notions in human minds. ‘Human mind’ being of course an imprecise notion itself.
No, but we might conclude that there’s no such thing as free evaporation. We’re postulating a deterministic chain of events. We label one link on the chain “will.” But its action is just as determined as any other link.
To be sure, if we accept that we live in a deterministic universe, we confront the question of why anything matters. Landsburg observes that we continue to respond to incentives in a deterministic universe, and so our actions continue to have consequences – if not moral consequences, then practical ones.
So, why would a guy who believes in determinism write a book called “Fair Play”? Of what relevance is fairness in a deterministic world, where each person’s conduct is purely the function of physical phenomena beyond the individual’s power to contravene? The bully on the playground may act unfairly, but the bully could not do otherwise.
No, the concept of fairness does not give me a fair basis for judging the bully. But it still gives me an unfair basis for doing so. And people – even bullies – may yet respond to the disincentive of being disapproved of, even unfairly. Thus, “fairness” is a means of social control, even if an arbitrary one.
So, why would someone who believes in arbitrary tools of social control condemn religion?
Steven,
“If evaporation is the result of interactions taking place at the quantum level, must it be concluded that there is no such thing as evaporation?”
Thanks for the clarification. I see now that the definition is key to this whole thought process. Free will by your definition, is just the process of deliberation within a person’s mind regardless of outcome. It is not what I think most people think of when they conceptualize free will; that someone is free to make any choice regardless of the physical limitations of their own body that ultimately limit his or her choice to one direction. People want to believe that they have full, possiblly supernatural (though they woudln’t frame it as such) control over their actions.
Is this correct?
Alex B:
Free will by your definition, is just the process of deliberation within a person’s mind regardless of outcome.
I don’t know what you mean by “regardless of outcome”. “Free will” to me suggests that this process of deliberation is a cause of subsequent actions.
@Ken B #43: The good news is it seems highly likely in some other universe, Ms. Portman is into you. Even more likely if she doesn’t have free will.
Does the weatherman “cause” it to rain tomorrow?
SL: “‘Free will’ to me suggests that this process of
deliberation is a cause of subsequent actions”
But science sez that this process of deliberation is
deterministic, mere neuron activity, per the laws
of ochemistry, without room for choice by “I”.
The consternation arisies from the fact that people
intuitively, immutably, believe there’s an “I” which
operates independently of these impersonal laws.
“I’m not a robot, dammit!”
Yes you are –
Paul T:
But science sez that this process of deliberation is
deterministic
Actually, “science” does not say that anything at all is ultimately deterministic, but the question of determinism is, in any event, not the same as the question of free will.
mere neuron activity, per the laws
of ochemistry
An ocean wave is “mere molecule activity”. Does that mean there’s no such thing as an ocean wave?
I think that we should start on a deeper level, on the level of consciousness. What is it really? I am sure that consciousness exists – simply because I experience it and most other people do, too.
The question is, if someone modeled my brain, neuron by neuron, by an analog computer, did all the wiring correctly, set all the input weights, effects of the actual neurochemistry etc, would this computer experinece my consiousness, at least for a second? Would he/it experience being me? As far fetched as it may sound, this modelling is not forbidden by any law of nature and might be one day even technologically feasible.
We may never know if the computer really has acquired consiousness even for a second or so (and even the experience of being me – with the terrible disappointment of realising the truth in a while – I think that it would be actually very cruel to do this, even to a machine.. thinking of it, I would probably prefer for someone other than myself to be the modelled one :)), and even if he/it says so, we have no way to confirm that he/it is speaking the truth..
Anyway, if human brain is 100% replicable in such way, I would say that free will is an illusion. Decision making is a weighing process, with many many variables playing a role, but still a sort of mechanical, alebeit very complex, one, except we do not “see” many of those variables and can be hence deluded into thinking that we have “decided” on our free will.
On the other hand, maybe there is something special about hundreds billions of live neurons being connected by even several orders more of connections etc. Maybe there is some emergent quality, coming from this unimaginable complexity, that changes the game somehow. But I have no idea what that might be, short of “magic”..
Steve Landsburg >>An ocean wave is “mere molecule activity”. Does that mean there’s no such thing as an ocean wave?<<
I think this is the wrong way to claim free will exists. I explained this earlier, but I'll reapply it to your ocean wave analogy. The grounds on which I would say 'free will' is not a fact and an ocean wave is a fact is that the latter is not inconsistent with nature's inevitable trajectory. In other words, an ocean wave can be part of the inevitable trajectory, whereas free will cannot, because one is never 'free' from that cosmic trajectory.
So the free will either is part of the trajectory – in which case, it is not really acting in any kind of freedom separate from the trajectory; or free will is not part of the cosmic trajectory – in which case it is claimed to be no longer part of nature, where, at that point it falls down by not being definable (as per my initial request for a definition). Either way, the idea of 'free will' gets us in philosophical trouble.
James Knight:
The grounds on which I would say ‘free will’ is not a fact and an ocean wave is a fact is that the latter is not inconsistent with nature’s inevitable trajectory.
I have no idea what this means.
Steve, if nature has an inevitable path then an ocean wave can logically be part of that universal story without any contradiction. However, if free will is part of the universal story then we have a contradiction because it is not really acting in any kind of freedom separate from the trajectory of nature – so it is wrong to call it ‘free’, because by ‘free’ I take it you mean nature permits us an implicit ability to act outside any inevitability that nature may hold over us. If free will is not part of the cosmic universal story then it is claimed to be no longer part of nature, where, at that point it falls down by not being definable (as per my initial request for a definition).
Steve I just sent a reply, but now it has gone. Did you get it, or are there glitches in the system?
James Knight:
I. if nature has an inevitable path then an ocean wave can logically be part of that universal story without any contradiction.
I don’t know what “logically be part of that universal story” means. Does it mean:
a) there exists another Universe, not identical to ours, in which at least one past light cone is identical to at least one past light cone in our Universe?
or
b) there exists another Universe, not identical to ours, in which every past light cone is identical to some past light cone in our Universe?
or
c) there exists another Universe, not identical to ours, and some choice of global time coordinate such that the past of our Universe is identical to the past of that Universe (past being taken relative to some fixed time)?
or
d) there exists another Universe, not identical to ours, such that for any choices of global time coordinates, the past of our Universe relative to some fixed time is identical to the past of that Universe (the past being taken relative to some fixed time)?
or
e) something entirely different?
II. My followup question is:
What does any of this have to do with free will?
III. Ah—but you’ve said this: If free will is not part of the cosmic universal story then it is claimed to be no longer part of nature,
This, of course, settles the question. Free will is obviously “part of nature” (what else could it be part of), so whatever your definitions, it is part of the “cosmic universal story”.
Steve, I mean e) something entirely different. If nature has an inevitable path then an ocean wave can logically be part of that universal story without any contradictions. By logically I mean simply that an ocean wave’s existence does not violate any of our ideas about what nature is in some logical manner. The same could be said of other objects in nature – trees, buildings, water, happiness, etc.
For clarity’s sake, forget about other universes if possible – they have nothing to do with my argument. What we need to ask is; what could be said to be a fact about nature that would interfere with our ideas about what nature is in some logical manner? There are some obvious examples – if we said ‘The universe has no carbon atoms’ that would be a violation of the logic we know to be true of the universe (and empirical findings) because the universe does have carbon atoms. If we said ‘There are no words in the universe’ that would also be a violation of logic because we are using words to assert it.
We’ve said that nature has an inevitable story beyond the activity of human beings. Now let’s say we contend that free will exists and is part of nature. I am saying that is a violation of logic just like ‘There are no words in the universe’ is a violation of logic -because if nature has an inevitable story beyond the activity of human beings then the ‘free’ part that we ascribe to humans is in logical violation from the ‘inevitable’ part that is built into nature. Nature cannot be inevitably on a trajectory while at the same time contain humans that are within that trajectory (and part of it) yet somehow free. All of our thoughts and actions are just nature playing out the 2nd law of thermodynamics.
What nature could do, however, is bring about minds that believe they have free will because they only sparsely sample a tiny fraction of nature’s story. That’s what I think has happened.
James Knight:
Somehow we are entirely failing to communicate.
I have no idea whether the existence of carbon atoms is a logical necessity or an empirical fact. I’m not even sure what you mean by a logical necessity — a logical necessity is something that follows from a set of axioms, but I don’t know what axioms you’ve got in mind.
When you say that “nature has an invevitable story”, I *think* that all you’re saying is that the Universe is what it is, and is not anything that it is not. It’s hard to disagree with that, but it’s also hard to be sure that’s all your saying.
When you say that “the `free` part … is in logical violation from the inevitable part”, I lose you completely. I don’t know what logical argument you have in mind, but if it seems to prove that free choice is impossible, then it’s surely incorrect, since we do make free choices. But I can’t pinpoint your error, because I don’t know what your argument is.
As for “All our thoughts and actions are just nature playing out the 2nd law of thermodynamics … “, are you saying that the second law of thermodynamics, by itself, without any other laws of physics, suffices to determine our thoughts and actions? I find this a truly incredible claim and would love to know where you got this idea. I’d also love to know what you think it has to do with free will.
Steve, I think communication is a problem, because what I’m trying to convey to you is not being understood by you (although that’s probably to do with my not explaining it clearly enough). Like you, I’m a huge supporter of Aumann’s Agreement Theorem, and believe that this issue should be able to be resolved with clarification of meaning. I also would love to believe in free will, so I have no biases against it. I just think it is not true – I think it is a concept we have made up in conjunction with our gradual ability to evolve moral concepts.
By “nature has an inevitable story”, I mean more than the universe is what she is; I mean if hypothetically we could step outside of her and watch the story unfold we would see her take an absolutely deterministic path. Everything within the universe, every thought, every dream, every stone rolling down the hill, every burst of wind, every formation of a planet, every nucleosythesis, every galaxy, and moreover, the entire cosmic expansion is all nature running its course, and that course is entirely deterministic (deterministic in the absolute sense). If you knew every fact about the universe it would be a story from start to finish that couldn’t have been any other way because its laws facilitate this particular path it has taken. I don’t mean with regard to any fundamental laws of logic, I mean that to state the universe we happen to be in could (given all its laws) take another path other than the one it is on is to talk nonsense.
Let’s see if you agree with that before we go on, because that is the first important step about which we both need to be on the same wavelength.
James Knight:
Everything within the universe, every thought, every dream, every stone rolling down the hill, every burst of wind, every formation of a planet, every nucleosythesis, every galaxy, and moreover, the entire cosmic expansion is all nature running its course, and that course is entirely deterministic
(emphasis added).
1) I’m not sure exactly what “entirely deterministic” means. It presumably means that information about some parts of the history of the Universe suffices to determine the remainder of the history of the Universe. I’m unclear on exactly which parts you have in mind, hence my earlier questions about light cones, etc.
2) I’m not sure whether you’re referring to a *single* history of the Universe or an ensemble of histories. If the former, it seems very unlikely that determinism is true (in fact, I’d say that the Aspect experiment pretty much rules it out). If the latter, then some version of determinism *might* be true, but I don’t see any compelling reason to expect it.
3) None of this seems relevant in any event. As Robert Nozick likes to point out, even if determinism is true, thermostats can still control the temperature — and our volitions can still control our actions.
Steve,
I basically agree. I think the easiest way to define free will is the alignment between my definition of my self, my desires and my actions. When I look in the fridge and grab a piece of cake, the same action can be viewed multiple ways. If I identify with my thin self, I can say I was unable to resist the urge to eat cake. If I identify with my hungry self I can say I chose to eat cake. Since both definitions are valid, free will exists as a possibility.
If we know someone’s desires and context, we can anticipate their actions. We can even anticipate involuntary actions.
In general I think the squishy part of free will is actually the fuzzy definition of self. I think some people just assume the self is a permanent fixed entity. In reality, how I define my self can differ dramatically over time and as revealed above, it can be viewed multiple ways for the same event.
Harris defines away some of these self definitions.
Steve,
I don’t think the Aspect experiment or any of Bell’s theorems have very much to do with whether determinism is true, because they only relate to a particular way that we view the universe.
Steve says >>I’m not sure exactly what “entirely deterministic” means. It presumably means that information about some parts of the history of the Universe suffices to determine the remainder of the history of the Universe.<<
Yes but more than that – it means information regarding the laws of physics suffices to determine the entirety of the universe from start to finish. Given the laws the universe has – it seems its entire story from start to finish followed a path that, with complete knowledge of it, would be entirely predictable and mappable to a series of deterministic computations (as per Steven Wolfram's conjecture)
I think Steve Landsburg has a definition of free will much like Roger P's – "alignment between my definition of my self, my desires and my actions". I have to concede, if that is as much as you wish to say about free will then I have no objection to the notion that the universe could give rise to alignment between human beings' definition of the self, their desires and their actions.
I hope you would not object though to the notion that the universe's trajectory – from start to finish – is only going one way, and that way is an inevitable story that contains within it the human story which also had an inevitability about it due to the laws of physics.
James Knight:
it means information regarding the laws of physics suffices to determine the entirety of the universe from start to finish.
This seems to be in direct conflict with most mainstream interpretations of our current understanding of physical law. Why would you expect it to be true?
I hope you would not object though to the notion that the universe’s trajectory – from start to finish – is only going one way, and that way is an inevitable story that contains within it the human story which also had an inevitability about it due to the laws of physics.
I honestly can’t make head nor tail out of what “going only one way” means. If it means that the Universe is what it is and is not what it is not, then of course I agree; if it means anything else, I am baffled as to what that anything else could be.
The remainder of your statement seems to suggest that the laws of physics suffice to determine the whole of the universe, independent of any initial conditions. Even putting aside the fact that the laws of physics, as currently understood, are not deterministic, I don’t see any reason to expect that the initial conditions don’t matter.
(All of this assumes that your references to “the Universe” are meant to evoke a single history, not an ensemble of histories….but correct me if I’m wrong.)
PS to James Knight:
I don’t think the Aspect experiment or any of Bell’s theorems have very much to do with whether determinism is true, because they only relate to a particular way that we view the universe.
If you care to elaborate on this, I’ll be interested. I’d have said that the Aspect experiment is pretty close to definitive, but if you can make me understand what you mean by “a particular way that we view the universe”, maybe I’ll learn something.
Steve, I’ve read your fresh post on determinism, so I’ll incorporate the questions you asked there into this response. You’re quite right, determinism is a tricky term to define – but I reached a point a while back at which I came up with what I think is a satisfactory definition. I agree the universe has no globally defined state that can be called the present, because the present is relative to the observer. You also enquired as to why the Bell theorems say virtually nothing about determinism. As it happens I think the answer is the same for both. The way we describe the universe is through the construction of terms and ideas based on our physical perceptions of the world (the macroscopic world). What we don’t have is an adequate way to describe the universe in terms that would give a proper definition to determinism. We don’t even have a proper method of describing the quantum world – even then we resort to using terms implicitly related to the macroscopic world (particles, waves, position, locality, state, spin, collision, energy, etc). Imagine the problem in trying to define a universe in terms other than our spatio-temporal terms used. That’s why I think determinism is opaque. The best way I’ve found to circumvent this is to describe the universe in a completely different way. Every physical system can in principle be described in terms of computation – so what we have to do is imagine the entire universe in terms of pattern. This is no problem, because it requires a hypothetical description of the universe in terms of pattern storage (a very large pattern – but small compared with the whole system of mathematics). So what we can do is define the pattern in binary 1s and 0s. No doubt this is a far too complex task for any human, and would require execution times that beggar belief – so the easiest thing to do is imagine you are a being who can traverse the entire pattern from outside of the universe (let’s call the entire pattern P). Now determinism becomes clearer, because with the execution time (it’s not literally time, of course) to search the pattern we can observe its determinism.
But there’s another key point, which goes along with your comment about determinism being hard to define. We only get to grips with what absolute determinism is (as per my above definition of an overall determinism) if we can conceive of relative determinism too. Here’s how it can be done. At the human level indeterminism-determinism is a spectrum related to knowledge. Here’s a simple way to look at it; if I drop an apple the algorithm that calculates the motion is deterministic because we know what will happen – the apple will head towards the centre of the earth. If I attempt to track a particle in fluid subjected to Brownian motion then the algorithm that calculates the motion is indeterministic to us because the pattern it will take cannot be predicted. That is what I mean by relative determinism. If you know a pattern it can be said to be relatively deterministic, if you do not it can be said to be relatively indeterministic. This, of course, changes over time. If one day we work out the random patterns in Brownian motion (I doubt we will) then we could call it relatively deterministic because its patterns could be mapped to a deterministic algorithm.
That was relative determinism. Absolute determinism is different – it is an overall determinism of the entire universe, as described above. Nature has a script already written – and your conceptions of randomness, unpredictability, surprise, and cause and effect alterations are not in any way an alteration of that script – they are simply pages of that script being revealed to us. It’s just that now, instead of viewing those things in terms of the spatio-temporal, we are now considering the script in terms of a pattern of 1s and 0s. We think we are uncertain about determinism because we deal with nature in terms of its smaller constituent parts, not as a whole. With mathematics and computation nature tells us something very relevant here; providing that the search space is linked together in a nexus, with indefinite amounts of execution time, we can map anything into a descriptive algorithm or function, giving us a pattern. We could do this with nature if we had access to it from outside, and could observe the pattern. It is because we have incomplete knowledge from inside that we have radically unpredictable events, and it is because we have radically unpredictable events that the indeterminism and determinism spectrum comes in from the inside. If we had complete knowledge of the universe then its absolute deterministic path would be deterministic to us in the second (the relative) sense too, as well as in the absolute sense described above. If we define relative indeterminism as a physical system that cannot be described with an algorithm or function, then there are no (finite) systems that are indeterminate – because all systems can potentially be calculated from a basic equation if we have access to the search space and allow the necessary execution time.
Make sure you’re clear of one thing though – this concept of pattern storage is only one way (of many) to describe nature – but it’s the best way to describe it to show it is deterministic. We needed to find this way because nature cannot be described deterministically in those macroscopic terms. With the above terms, it can.
James Knight: Thanks for post 84, which looks like it will repay a much more careful reading than I have time for right now. I will study it closely later in the day.
PT: But science sez that this process of deliberation
is deterministic
SL: Actually, “science” does not say that anything
at all is ultimately deterministic, but the question
of determinism is, in any event, not the same as the
question of free will.
PT mere neuron activity, per the laws
of chemistry.
SL: An ocean wave is “mere molecule activity”. Does that
mean there’s no such thing as an ocean wave?
************************************
Weak analogy.
Because no one says “of course, an ocean wave is
the sum total of its molecular constituents. But
there’s more to it, something non-physical, its
self-awareness, which enables it to CHOOSE which
way to break, the spirit.”
Whereas, they do feel that way about brain activity;
“there’s MORE TO IT than mere neural chemeistry, which
science can’t explain.” But there’s nothing of substance
to support this.
Paul T:
Because no one says “of course, an ocean wave is
the sum total of its molecular constituents. But
there’s more to it, something non-physical,
As I said in the original post, if your claim is that our actions are caused by something “non-physical”, then your claim is so self-evidently wrong that I can’t believe this is what anyone’s actually claiming.
PT: Because no one says “of course, an ocean wave
is the sum total of its molecular constituents.
But there’s more to it, something non-physical,”
SL: As I said in the original post, if your claim is that
our actions are caused by something “non-physical”, then
your claim is so self-evidently wrong that I can’t believe
this is what anyone’s actually claiming.
*******************************
No one claimed that here.
But that is what people believe, the overwhelming
majority, in my experience. They cannot accept that
brain activity = mental activity, just billions of
neurons, blindly obeying the laws of biochemistry.
There must be more than that.
That’s what makes the issue so thorny.
Nobody wants to believe he’s a robot.
Paul T:
They cannot accept that
brain activity = mental activity, just billions of
neurons, blindly obeying the laws of biochemistry.
There must be more than that.
Well, of course there’s more than that! There are also trillions of connections.
More to the point, why would you choose the level of biochemistry at which to desribe these processes? If you’re going to say that we don’t have free will because “it’s all just biochmeistry”, why not say that there’s no such thing as biochemistry because it’s all just quantum mechanics? Or that there’s no such thing as quantum mechanics because it’s all just math? These things are as true, as false, and as silly as asserting that we have no free will.
My whole point is that attributing our actions to biochemistry is neither more nor less “scientific” than attributing them to free will. No matter what level you describe things at, someone can always come along and say you’re deluded because there’s a deeper level. If that persuades you to abandon free will, it should also persuade you to abandon all of science.
I disagree with point 1.
“If you think you’re sensing the color blue, then you are sensing the color blue (though of course you might be deluded about what’s causing that sensation).”
That is fine, your sensing is merely a description of your internal state. This would be the same if you were a simulation running a pre-determined course.
“If you think you’re deliberating and choosing, then you are deliberating and choosing.”
This is not true as I understand “to choose”. Choosing is not a description of your internal state. If you were a simulation running a pre-determined course, I do not think you could describe your action as “choosing”.
Thus you could be deluded about what was causing the sensation of blue, but still be perceiving blue, but if you were deluded about choosing, you would not be choosing.
New experiments cast doubt on an old one http://www.newscientist.com/article/dn22144-brain-might-not-stand-in-the-way-of-free-will.html
RPLong: “When you say, “So what is it then: magic?” do you mean to say that you *do* consider sentience a mechanism? ”
I think sentience is likely the _result_ of a particular mechanism yes, though this is not something I’m committed to on first principles. I’m just speculating because if it is not, then we really have no idea what it is, and can’t really draw any particular conclusions from it. It certainly IS a feeling that we have. And either that feeling is caused by some explicable causal mechanism or it is probably not explicable at all. The former implies some way to better understand it. The latter is basically the end of all possible, verifiable inquiry into what the heck it is.
“What is free will contingent on? I thought my definition clearly indicated that I did not believe it was contingent on anything other than thought. Similarly, sentience is not “contingent on” anything. These things are attributes of the functional human brain.”
So, what reason do we have, then, to say that they are not present in anything that can deliberate over choices, simulate possibilities, etc.? How are you going about denying them to, say, complex computers? If you can’t explain how they arise in the human brain, how can you look at a complex neural net and claim that it has no qualitative experience of deliberation? What grounds to you even have to deny consciousness to, say, an completely inanimate rock? You’re trying to appeal to a mysterious, utterly inexplicable quality. Ok, I say: so if we want to play that game, then I can play it too: computers, nay, even rocks have exactly the same experience. Prove me wrong.
“In short, they cannot perform a priori reasoning. The best they can do is *simulate* a priori “reasoning” algorithmically, i.e. by running an existing model, saving the estimators as parameters, and then re-calling the parameters in a pre-defined equation. To call this a priori reasoning would be a serious mistake.”
But all evidence points to human beings, human brains, functioning in exactly this way: the major advance of human beings over most of the other animals is our frontal cortex. What does that part of the brain seem to do? It imaginatively SIMULATES future states of the world, based on past experiences (that is, it imagines outcomes, given particular facts, and then judges whether those outcomes are what it wants or not).
“Of course, human beings have created plenty of things that did not exist in the environment beforehand. The examples are virutally infinite, and the best ones involve abstract concepts that have no direct analogue in nature. Since we’re on an economics blog, I’ll give you a great example: Money.”
But money is as an ingenious “neat trick” solution to mutual exchanges of value as is any neat evolutionary trick. Human beings did not create a world in which it turns out to BE a neat trick: they hit upon it as a solution to a very real problem with social interactions that involved exchange (i.e. that the person you are exchanging something with may not be able to give you something that the very NEXT person you trade with wants, directly: so money allows a common price value for _all_ goods in an economy, thus allowing you to trade for goods/services with anyone, assured that you’ll be able to exchange the money with someone else, and that you’ve both valued the exchange corrected). Money is a coordination solution for a particular situation that we’ve found ourselves in. Why do you think specific solutions to specific problems that human beings have DEMONSTRABLY hit upon via a well known HISTORICAL process (via lots of trial and error with various exchange systems) is somehow evidence that human choices are somehow unbounded by anything in their past experience, in a way that no evolutionary algorithms could possibly duplicate?
Neil: “In short, when I say I have free will, I mean my entire physical self, not just my conscious brain. The subjective feeling is just the evidence of free will, which we can share, as Steve said.”
I see how it is indeed, evidence of free will. I don’t see how it is evidence of “Free Will.” We make decisions. We’re absolutely incredible beings in that we can take in all sorts of inputs about the world, and then make decisions/choices about how to act and influence the world we live in, based on our interests, needs, goals, what have you. But the question of “Free Will” that’s debated by philosophers is much bigger than that. _IS_ there more to it than that? You tell me what that “more” would entail, specifically.