Well, perhaps Ben Goren, who asked this question, is better described as a “fixture” than a reader. But in the latest free will post, he made a comment that deserves some discussion:
I, too, would very much appreciate an answer to this question.
It would seem obvious that there is no fundamental difference between what goes on in human brains from what goes on in any other computational device. That’s the essence of the Church-Turing Thesis, and that one’s on superbly solid ground.
So, if humans have free will, the only possible conclusions are that all other computational devices, whether biological or mechanical, also have free will — a proposition which the overwhelming majority of proponents of “free will” will outright reject. The other logically possible path to take is to reject Church-Turing and embrace supernaturalism, to go with a ghost in the machine; this is precisely what that same overwhelming majority of proponents of free will in fact do do.
If we can get a clear answer on this one from the proponents of compatabilist free will, I think it would go a long way to clarifying whether this is an argument over substance or semantics. And if it’s an argument over semantics, I think it should also make clear the perils of embracing the loaded and oxymoronic term for something that it is typically the diametric opposite of its common meaning.
Clearly, if you’re a compatibilist, then you might admit that a computer programmed to respond (deterministically) to many complicated contingencies has free will. So do robots. And perhaps a thermostat has free will. If it doesn’t, then there’s somewhere on the continuum of “response devices” where one must say, “I’ll arbitrarily decide that free will begins at this point.” And, of course, compatibilists must agree that some (but perhaps not all) animals have free will, and even some plants (after all, plants can respond to diverse environmental challenges in complicated ways).
But you needn’t do any of this if, as Ben implies, you just deep-six the whole idea of free will and use other language to express what you mean.
Now I’ll sit back . . .
sub
2
3
3.1
+
3 + 2i
If I’m a fixture, I’d hope I’m more of the type that illuminates the surroundings than one that spews forth fluids….
b&
But the more important thing to ask – are you a fixture that has free will? And if so how much 😉
Less important perhaps: Is the term “fixture” as applied (Coyne’d?) by Jerry to Ben, a product of Jerry’s pre-dispositional programming, or simple free will?
Not only am I a fixture that has free will, I abstemiously luxuriate in my free will on my working vacations at my highrise cottage in that metropolitan oceanfront villa in north-central southern Arizona border location. You’d know the one — it’s got a great view of the Kansas mountains, obscured only by the dwarf redwoods.
Cheers,
b&
Thank you Ben for getting JAC to take Church-Turing and the implications of the Turing halting problem seriously. I think a computer running self-referent algorithm has free will.
Interestingly Seth Lloyd has posted paper on arXiv on this very question earlier this month, entitled “A Turing test for free will”
http://arxiv.org/abs/1310.3225
This paper is essential reading for anyone interested in the debate on free will, eloquently argued even if the mathematical proof is a bit hard going at times.
Take his Turing test for free will yourself everyone.
Not heavily into philosophical naval-gazing, but I think I’ll give that paper a read.
Yes, but in your defense, you have no choice.
That you are a fixture is fitting.
This was answered in Dennett’s book ‘Freedom Evolves’. Free Will is a feature of any system that can make conscious choices. It doesn’t matter what the hardware is.
What is a “conscious” choice? Sounds like using one vague term to define another.
Yes I once had the pleasure of discussing this with Susan Blackmore… Her take, and I agree, is that Dennett should have titled the book “Choice Evolves” which is what he actually succeeded in arguing for, not free will.
What is “conscious choice” if not “free will”? How is this not just a re-definition? Dennett might as well have said that burning is a feature of any system that can sustain rapid combustion.
So then the thermostat might not and the computer might consciously make – or “recognize itself” making – one of the possible “determined” choices. (Such determined choices might look quite like Schrödinger’s cat.. neither there nor not there until identified.) And if so might not Dennett have described two points on the continuum where mechanistic free will is and is not? Perhaps the better the definition of consciousness, the clearer can be located the “now” arbitrary point/s of divergence.
Well put in the last paragraphs. Asked for a one sentence definition at a conference in (I think) 2012, Dennett (perhaps our most articulate compatibilist) described free will as a political – rather than metaphysical – boundary between those who have the “right sort of competencies” for choices that they’d want to recognise as “theirs” and those who don’t have those competencies.
I don’t see any reason why other animals or robots couldn’t have that, and neither why it should be called “free” will – it’s more like “appropriate will” or “responsive choices” or somesuch. Calling it “free will” seems to simply be a sop to human sensibilities around being especially complex or mysterious. It’s only role is that political one – softening the blow of what naturalism dictates – but that political role shouldn’t allow us to be fooled into thinking it’s a different sort of choice to that potentially made by other agents, like robots and so forth.
That’s useful information and I agree with you.
As Jerry’s several posts, and as many subsequent comments suggest, there is certainly some justification for arguing that the issue is rather political. However, while I tend to agree somewhat with one of your premises or arguments – that there is some common ground between the “free will” of robots and animals, and that of humans – it seems the scale and magnitudes of the feedbacks involved make the two cases not just different degrees but of entirely different kinds. For instance, my understanding is that robots of various types – from thermostats to industrial versions on production lines – are largely predicated on linear systems.
Rather different with humans, I think, where, as with other emergent processes, the feedbacks and the systems themselves are highly non-linear: the whole is truly very substantially more than the sum of the parts.
We are, I think, to a first approximation and in some cases, our own “First Causes” – so to speak. We are free, at least to some extent, to choose those courses of action which are optimal for our own survival. And in that, I think, lies some degree of “free-will”, and moral responsibility.
Reblogged this on hitchens67 Atheism WOW!! Campaign.
Do we know of anything which is free as in totally unconstrained? If free is to mean determined by factors within the system, isn’t decision making itself a process of bringing relevant additional factors into the system?
Yes, I would really like to know where “free” features in the compatibilist’s definition of “freewill”!
This question has been answered lots and lots of times. Read the many threads on this website on the topic, or do a google search on the issue.
You mean it’s been waffled about.
If by “free” you don’t mean “free from deterministic cause and effect”, then what does it mean?
What does this freedom consist of?
Then: how does the “will” achieve this freedom.
Finally, what does this “will” consist of?
If you define “freedom” that way then freedom doesn’t exist, and I presume you never use the word in day-to-day life. Let’s examine some usages:
“Free speech”. This means that the *internal* wish to speak is not overridden by external constraints. Whether the internal desires are produced by a deterministic machine (they are) is irrelevant to the usage.
“Free market”. Means that a person’s *internal* wish to trade is not overridden by *external* constraints. Whether the internal desires are produced by a deterministic machine (they are) is irrelevant to the usage.
“Freed from jail”. Means that a person’s *internal* wishes as to where they want to be are not overridden by jailers. Whether the ditto ditto ditto.
“freedom of religion”. This means ditto ditto ditto ditto ditto.
Spotting a pattern yet? You either say that “freedom” doesn’t exist and never use the word, or you use it in a compatibilist sense about a deterministic universe.
Thus: the “free” in “free will” is about the extent to which (determined) internal desires are overidden by (determined) external desires.
Yes, I’m on to you now:
The Equivocation fallacy:
http://en.wikipedia.org/wiki/Equivocation
I’ve been giving you too much credit. 🙁
Not so. The meaning of “free” as compatibilists use it is a very widespread and common usage.
We are fully aware that it doesn’t mean the same as “dualist free will”.
That’s not what “free market” means, surely. My desire to trade necessarily entails an externality — namely, another party with which to trade and which has no particular motivation to ensure I get exactly what I want.
Yes, but it means there is no external authority telling you and your willing other party whether you can trade. Again, the term is all about external constraints.
The question Ben refers to in his first sentence is this (asked by Eric):
So, let me just see if I have this straight. Coel, Sastra, Vaal, Couchloc, feel free to reply with a simple yes or no:
The CFW position is that humans have free will in the same way that airplane autopilots, and soda machines have free will. Yes or no.
Vaal has replied with “no.” Coel’s previous comments suggest “yes.” I think there’s some disagreement in the CFW camp.
In the same way, yes, but not to the same degree. Cf. a human has “intelligence” in the same way that a bacterium could be said to have it, but not at all to the same degree.
Things like intelligence and complexity are continua, and you have humans towards one end and bacteria towards the other. In between, however, there are just differences of degree.
“Free will” is not anything mystical or supernatural, it is just a label for a type of *behaviour*.
Compare with the behaviours “courting ritual” or “threat display” or “social bonding and grooming”. Clearly it only really makes sense to use such terms at one end of a continua, but these are still continua (else they couldn’t have evolved), they are not binary yes/no phenomena. In the same way you shouldn’t ask for a binary yes/no dividing line about CFW.
Really, all these questions to compatibilists are harking back to dualism, they’re asking “what plays the role of the supernatural soul in compatibilism?”, but it’s the wrong question, the answer is “nothing”.
Right, but if nothing plays the role of a causal agent outside of materialism, if all causes are material, then how is compatibilism different from materialist determinism?
It is not different at all, not even in the tiniest detail is it different from materialist determinism. Compatibilism *starts* by embracing materialist determinism in utter and entire completeness.
So what on earth are you arguing about?
Where is the “free” in FREEwill?
The “free” means that the external constraints are not preventing you from pursuing the desires that your internal machinery produce.
As in: “Did you sign this contract of your own free will or were you coerced?”.
What about internal constraints?
Surely “free” implies freedom from internal constraints (deterministic cause and effect) as well.
If not, how is this freedom?
If so, what does this freedom consist of and how is it acheived?
All this does is push the problem back to defining what counts as coercion or as an “external constraint”.
Does a man who develops a brain tumor commit murder of his own “free will”? Did the tumor coerce him? Does the tumor count as an “external” constraint?
What about genetic predispositions?
To be fair, one could be an immaterialist determinist as well. Spinoza is arguably one of these, A. Damasio not withstanding. (The bit about how “god” (=nature) has other properties beyond the mental and the physical is the contentious point.)
No.
“It would seem obvious that there is no fundamental difference between what goes on in human brains from what goes on in any other computational device.” — Ben Goren
I reject this premise of Ben’s argument. I think computationalism about human thought is false and has never been established. The fact that you can describe the brain as a computer doesn’t imply that computation is all that’s going on there. If you doubt this read Searle.
You’re referring to the Chinese Room, of course. Which is just a god-of-the-gaps argument claiming that a soul is necessary for consciousness.
Sorry.
b&
Straw man much? Searle is a naturalist I’m sorry to inform you and doesn’t believe in souls.
True, but he has never answered what it is about brains that makes them different and escape the arguments of his (fallacious, but we’ll grant it to him to allow the ad hominem argument to work) argument. He just appeals to “causal powers”. In principle he could be right; the brain could function non-computably, but he owes us an explanation this “wonder tissue” also is not only hypercomputational (or, alternatively, non-computational), but how the hypercomputational properties allow the “intrinsic intensionality” he claims to be needed. Neither of these have ever happened.
Exactly right. Quacking ducks, and all that. An argument that at least an element of consciousness is super-physical or can do things a Turing Machine can’t even in principle is an argument that’s the exact logical equivalent of a Christian soul, and is one of the two primary reasons why we know that the Christian soul is fantasy — the other, of course, being perfect lack of evidence even where evidence must be found if the phenomenon actually exists. Merely obfuscating the description doesn’t change anything, and the religious affiliation or lack thereof of the proponent is likewise irrelevant.
Besides, I’ve never heard a proposal by which a person or other entity could manage to know enough about both English and Chinese to be able to skillfully translate between the two and not understand them, or why it isn’t reasonable to simply assume that our subjective experience of understanding is the “inside-the-box” experience of having all the necessary mappings to do the job. It’s only ever made sense to me if you base your argument on dualism…which swings us right back to spirit woo.
Cheers,
b&
Dang, Prof. Coyne, that Cat/Popcorn photo made my day!
Vaal
It has certainly crossed my mind that my computer might have a subjective experience, and I murder it every time I shut it off.
Hehe… 🙂
I wonder if the blue screen of death is considered an fundamental existentialist crisis in the world of digitalism.
Agreed.
Doesn’t follow. Cf.: “Thus, if one computer is calculating the digits of pi, the only possible conclusions is that every computer is calculating the digits of pi”.
Free will is a specific type of program (or behaviour), namely one concerning: (1) the selection from an array of possible choices to achieve some aim (= will), and (2) the ability to act, to carry out that selection.
Thus an aircraft autopilot has freewill (in, perhaps, a rudimentary form compared to us), but a computer calculating the digits of pi would not.
To have “free will” the computer needs to: (1) have an aim, (2) receive information from the environment, (3) consider which of a range of options — given that environmental information — would best meet the aim, (4) implement the choice. (All of this is deterministic of course!)
One might add a fifth condition: (5) some degree of self awareness, though deciding on what minimally counts as “self awareness” is tricky. But, given that “free will” is a continuum, with some things (like us) having much more of it than other things (chess-playing computers) that is not particularly a problem.
Free will is a continuum? So there is partial free will, or free will that is not fully free? I think this just confuses things even more. The fact that hundreds of comments in, no one can understand the CFW position except its proponents, should count as evidence that it is confusing and unhelpful. Even the CFW supporters seem unable to agree with each other on what exactly they are saying.
Of course. Just like there is a continuum for any high-level behaviour of animals.
Dualistic “free will” with supernatural souls might be binary yes/no, but CFW is not. The problem is that you incompatibilists keep arguing against dualism, they never argue against CFW!
It’s actually quite easy to understand, all you have to do is try. The first step is to stop suspecting us of being closet dualists, and stop advancing anti-dualist arguments against compatibilists!
My problem is that it appears there are only two ways to make any sense out of CFW.
1. CFW is the position that the brain is entirely constrained by physics, chemistry, and biology, and the same set of inputs will produce exactly the same output, and we’re going to call this situation free will
Or
2. CFW is really a closet dualism
Some CFW supporters (Vaal), when pressed to agree with #1 disagree and say, “no, there’s real freedom in there.” This leads me to #2. Others appear to agree with #1, and in that case we agree about everything except definitions of terms.
I’ll try to get back to this in more detail later, but I think both Coel and I are taking cues from Dennett’s arguments, in terms of referring to a continuum.
The distinction I would make is that “freedom” is on a continuum, but “free will” isn’t on exactly the same continuum, because it introduces another factor “will.” And we then have to ask “what is a will and to what kind of entities does it make sense to ascribe a will?” So free willed entities will form a subset of entities that have “freedom” of sorts, along a continuum.
There’s also the “moral responsibility” facet often associated with free will, which can make for an even smaller subset.
These may seem sticky questions for compatibilism, and they are. But then, this isn’t a special problem of compatibilism.
Whether we are incompatibilist, or compatibilist, we still face having to answer questions like “does it make sense to use the term ‘choice’ or ‘goal’ to things like apes, mice, computers, autopilots, etc?
And what are the consequences of our answers to such questions.
It’s no free ride either way.
Vaal
I suppose “freedom” must mean something like rolling a dice to “choose” between various options non of which are clearly more advantageous than any other.
But where does “will” come into the picture.
This seems to be to be a dualist’s concept. Where does determinism produce “will”?
In human brains, for starters.
BillyJoe,
I can not imagine how you inferred your rolling dice analogy.
Vaal
Thanks for the flippant answer about human brains being the source of the “will”
And the “roll of the dice” analogy was only a suggestion because I can’t see what else this freedom could possibly consist of.
Seems you don’t either.
BillyJoe,
The “freedom” has been explained to you explicitly many times, and not one instance suggests anything like random roll of the dice.
We have desires, and rational minds to think about which practical actions we can take will best fulfill those desires.
I have a desire to cross the street to the other side safely. I reason that to meet this goal, I ought to stop, look both ways to ensure there is no car close enough to hit me when I walk across the road, and then cross the road at a speed I’ve calculated will get me to the other side safely.
How on EARTH you find an analogy to “rolling dice to make decisions” – as if I’m just as likely to decide to break-dance in the middle of the street – is completely beyond me.
Vaal
(And in terms of free will, to say I made the choice to cross the street of my own free will is to say it’s a decision I desired to make and I could have chosen otherwise, e.g. to not cross the street).
I’m not suggesting its a roll of the dice. I’m suggesting that is a way that something in the brain could be “free”. In this sense, “free” means free from deterministic cause and effect. That’s not your idea of “free” because you insist it’s compatible with determinism. I’m asking you to explain in what sense something can “free” but not free from deterministic cause and effect.
You mention “desires” and “rational thinking” in your response. But these are also the result of deterministic case and effect within the brain. Still no “FREEdom” or “WILLing” to be seen anywhere. Yet, somehow we have FREEWILL.
The concept, frankly, is incoherent.
Yes!
No, we’re going to call *one* *type* of behaviour produced by the above as “free will”.
In the same way, we would not call the above situation “courtship ritual” or “threat display” but we would call certain types of behaviour produced by the above those things.
BTW, while taking the UBC class I alluded to, I came across a then-recent collection from a neuropsychology journal titled _Willed Action and Its Impairments_. There was a good review article in whose name it escapes me at present. I think my papers may cite it, though I cannot check at present. In any case, there *is* work on the neuroscience of will.
This sounds as though you have defined free will as the same thing as consciousness to me. Is free will different from consciousness, and if so, how?
“(4) implement the choice. ”
All computer programs do this; your limitation to action in the physical world is arbitrary.
An aircraft autopilot in which the “output” wire to the aircraft controls had been disconnected would not do that.
Same with the pilot. Disconnect the control surfaces from the “pilot output” and the pilot is in real trouble.
How on Earth does a computer have freewill?
It follows an algorithm.
This algorithm makes it do things.
The computer does these things specifically as instructed by the algorithm.
Occasionally, in order to stop a “brute force” situation where the computer would take too long to produce the output, it is instructed to do something akin to rolling a dice to “choose” between unrefined options.
Where is the damn freewill?
It has compatibilist free will. Compatibilist free will is deterministic. Google “compatibilism” if you want an explanation.
If it’s deterministic, then it’s not free.
Freewill = FREEwill
Compatibilist freewill = notFREEwill.
😀
Can “free speech” exist in a deterministic universe?
If yes, welcome to compatibilism.
If no, I respect your viewpoint but think you have a useless definition of “free”.
Equivocation Fallacy.
1. Just throwing out the names out fallacies is not good debate.
2. It’s not an equivocation fallacy. It’s the opposite. An equivocation fallacy is the misleading use of a term with more than one definition. I do not believe that “free” has more than one definition.
3. If we ARE going to shout out names of fallacies, then you can think about my point as an accusation of Special Pleading. You are using one definition of “free” in the context of “free will”, while using another definition of “free” in every other context.
You guys keep asking “but how is [compatibilist free will] free?” and we’re responding “in the same way that anything else is free”.
Either bite the bullet and concede that – on determinism – *nothing* is free, and the word is useless (I’ve only met one person who had the intellectual honesty to do that, but I’m ever-hopeful) or find some way to defend the notion that “free will” really does need this special one-off definition of “free”.
Well, I’m of the “free will is incoherent and therefore meaningless” camp, so it might not surprise you to learn that I’m quite comfortable that nothing is free in the types of senses dualists and compatibilists try to use the term.
But I would note that the word is still quite useful in all sorts of other contexts — as is, for that matter, the term, “free will.” Even though we are clockwork marionettes dancing to the strings of physics, it still makes sense to indicate whether or not you’re signing a contract of your own free will, for example.
This, of course, bears no more relevance on the philosophical and theological questions of “free will” than does the fact that a singer can pour her heart and soul into a song without either actually having a soul in the religious sense or leaving a bloody mess on the stage.
Cheers,
b&
If anyone refers to free will, I ask, “Free from what?” But if someone says ny choices are illusory, I take issue.
When a computer performs a calculation, the answer is determined, but if it is a new calculation being performed in the most efficient possible way, the simplest way to predict the outcome is to let the computer get on with it. I maintain that the calculationis a real process, and that making up my mind is a real process, however determined the outcomes may be in either case.
As for the gradation from thermostat to (dare I use this example?) pussy cat to person, the problems here relate to perception and consciousness, not to free will as such.
Is free will deterministic?
Free from *external* constraints that prevent someone acting on their *internal* (deterministic) desires. cf. “Did you sign this contract of your own free will or were you coerced?”.
You are never fully coerced because it is determined whether or not you give into the coercion. People can refuse, you now. And the compulsion of a gun to your head, and your response to that, is exactly as determined by physics as is your supposedly “non-compulsory” decision to sign the contract. “Compulsion from others” is just part of the environment, just as are the environmental factors that lead somebody to, say, sign a lease on a Lincoln Continental.
The determinism of the situation is neither here nor there, since, whether it’s all determined or not, there will still exist scenarios where one’s choice is being made under the coercion of another, or not.
Just like, stating that the fate of animals in a zoo is “determined” has no bearing on the fact that there still exists differences between an animal caged in a zoo and “free” in the wild to describe.
To say “I signed the contract with John of my own free will (not coerced)” isn’t to say “I am free from all existing constraints.” It merely points to some *particular* constraints, e.g. John wasn’t threatening me with a gun to sign the contract, forcing me to do what I wouldn’t have done, without that type of coercion. That’s the freedom being described, and since it’s real whether it’s determined or not, objecting that “it was all determined” is beside the point.
Vaal
Full agreement.
I must also say that the line of argumentation employed here – “people can refuse, you now” – reminds me uncomfortably of libertarians arguing that it is only fair if somebody is employed under unsafe conditions for very little salary because hey, they accepted that offer when they could have refused and decided to remain unemployed and starve to death instead. And the odd thing is, those libertarians using the same line of argument promote libertarian free will to justify their conclusions…
But a gun to the head can make the difference. {internal state} + {gun to head} + {all other environment} = Action A, whereas {same internal state} + {no gun to head} + {same all other environment} = Action B.
That’s what we call coercion. What other type of “coercion” is there (or would you prefer to drop the word and the concept from the language)?
Indeed so. But what’s the relevance of this? Compatibilism entirely embraces determinism, that is its starting point.
<blockquote “Compulsion from others” is just part of the environment, …
Indeed so. And in compatibilism the “free” part is the extent to which desires resulting from internal machinery are over-ridden by external “environmental” inputs.
What about internal constraints?
Where is the freedom?
Any specific instance of “freedom” is relative to specific conditions. What internal constraints are you thinking of?
I have a desire to go outside for a walk.
If there is a type of internal constraint you are thinking of that would make it impossible for me to go outside for a walk, then what would it be? And if it’s the case it does make it impossible for me to go for a walk then *shrug* I’m not free to do so.
But, failing something stopping me, I’m free to do so. As a matter of fact…I’m going out for a walk right after I type this 🙂
Vaal
I may have misunderstood, but when you referred to freedom as being free from external restraints, I assumed you meant constraints imposed by the environment. So I asked, what about internal restraints – by which I meant the constraints imposed by the deterministic brain.
If freedom, does not mean free from deterministic cause and effect, what does it mean. What does this freedom consist of?
BillyJoe,
It’s really, really simple.
Just think of “free” in the way you or anyone else use that term normally. You already know what people mean by “free” whenever they use that word.
I’m free for lunch.
I’m free of my contract.
I’m free to go on vacation.
A western woman notes she’s “free” to drive a car, in contrast to many women in Saudi Arabia.
A newscaster explains a political prisoner has been “freed.”
You understand what all these expressions refer to right? Nothing magical or too weird for you to figure out.
Why then is the word suddenly baffling to you here when it’s being used in just the same way?
If someone asked you of the above examples “what does this freedom consist of?” what would be YOUR answer? Would you be baffled and unable to explain what is meant by those sentences?
Vaal
Well I’m going to blame the language for your misunderstanding about freewill.
Our whole language is dualistic. It’s the simplest and most compact way to communicate in everyday situations. But, it is definitely not the best way to communicate in these discussions where we are questioning the very assumption of dualism.
The problem is that it is easy to be seduced by the language. Even you, who claim not to be a dualist, have used the dualist language to defend your point of view.
I am free to go outside.
So out I go if I feel like it.
But actually no.
The brain in the body labelled “BillyJoe” feels like it because the deterministic cause and effect relations within it produces that feeling. That same deterministic brain causes the body labelled “BillyJoe” to go outside.
Difficult isn’t it, this non-dualist language.
Even with a gun to your head there are choices. You can choose to not cooperate and risk being shot, you can choose to go full “Bruce Willis” and try to get the gun off them. You could sign someone else’s name. There are always choices, there is always constraint.
This comparison ignores the element of consciousness in the free will debate.
I haven’t read through all of the comments on this post yet, but this comment cuts to the heart of a confusion that I have seen crop up repeatedly in these free will debates. Conscious choice does not depend on or validate the existence of free will (whether libertarian or compatiblist). Although consciousness is a very tricky concept, and I have no idea how to solve the “hard problem”, there is a very important distinction between consciousness and free will:
Unlike free will, the existence of consciousness stands on firmer ground, empirically, than any other phenomena.
Consciousness demonstrably exists, even if the contents of consciousness are an illusion (as in “The Matrix” or Orpheos’ cave). Free will does not. We think we feel like our actions are freely authored because we do not pay close enough attention to our experience most of the time. Take a few moments to observe the way thoughts arise in your consciousness, and you will realize that you have no more free will to determine what your next thought is than you have to determine what or who will next wander into your field of vision.
Whether we continue using the loaded and confusing term “free will” after having this little revelation is mere semantics.
First, a “nutshell’ attempt to answer:
There is nothing in principle stopping a computer/robot from having “free will” of the sort we have. In the same way it seems there is nothing in principle stopping them from having “consciousness” either (unless you are what Dennett calls a “mysterian” who thinks of consciousness as a magical thing…which I’m not).
Secondly, Ben’s assertion that the above answer is
“a proposition which the overwhelming majority of proponents of “free will” will outright reject”
Seems to at least imply some dubious assumptions. One is: which proponents of free will is he talking about? In terms of academia, a majority of philosophers are compatibilists, and it seems to me many would agree with the above conclusion.
Or, is Ben talking about religious people, or the common man on the street or whatever?
In that case, there seems to be the implication that if they wouldn’t agree that such a system had free will, then this means the free will the compatibilist is talking about isn’t the ‘real free will” this majority holds it to be.
But that wouldn’t follow necessarily. It could still be the case, and I would say it is, that we are talking about essentially the same thing with free will, but the person rejecting it’s application to machines that sufficient mirror our capabilities simply isn’t being *consistent* in his logic. He’s special pleading. In the same way that if we both held that “white people” have human rights, but someone said “but that doesn’t apply to black people.” That doesn’t mean we aren’t talking about the same thing when we talk about “human rights;” we are, it’s just that the racist isn’t being consistent in applying that concept to other entities that he should.
Same with free will. Someone saying “I think humans have free will but a similarly structured machine wouldn’t” isn’t necessarily differing on what free will is; he just isn’t being consistent in it’s application.
Just wanted to clear that up, at least from my perspective.
Vaal
Classic free will (will free of the laws of physics) doesn’t exist. I suspect most atheists agree with this.
Colloquial free will (will free of coercion or unusual constraints) does exist.
No computer, animal, etc has the first. Which has the second depends on how we define “coercion” or “unusual constraints”.
Ultimately it boils down to semantics. This is also true for consciousness, which I think is in the eye of the beholder.
Awesome cat / popcorn picture!
The problem is that the definition of freewill that neither deals with the “free” bit nor the “will” bit is a bit if a sausage.
I don’t know if you read the comments further down that addressed this but here it is.
The “will” is the process of deciding on an action by an agent.
And “free” is used in exactly the same sense we use it in every other situation ever – “not under a particular set of constraints” which is given by the context. E.g. “free speech” = “free from government censorship”, “free lunch” = “free from monetary exchange” etc etc.
That hard-determinists act like “free will” needs an entirely different definition of “free” is Special Pleading to a degree that would make William Lane Craig blush.
> And perhaps a thermostat has free will.
> If it doesn’t, then there’s somewhere on
> the continuum of “response devices”
> where one must say, “I’ll arbitrarily
> decide that free will begins at this point.”
Freewill… may very well be a pipe-dream,
but… discretion is within Human scheme.
Folk touting fate~karma~kismet – extreme.
As a compatibilist, I’ve got no problem with continua and having to draw an admittedly arbitrary line for what we call free will. Similarly, I have absolutely no reservations about a woman wanting to end her pregnancy in the early days with RU-486, but have huge opposition to her being allowed to end it 8 months later with anything other than a live birth (unless her own life or health is very seriously at stake).
The ethics of abortion have nothing to do with this discussion.
That’s true, but then I would have supposed that that was obvious. The abortion comment was made merely to illustrate a more general point: that life presents us with questions that involve continua, which at the end points can be very black or very white. Reasonable people realize that they have to draw lines and where they draw those lines will involve a certain amount of arbitrariness. The incompatibilists here keep asking for black or white on the issue of free will, and I’m seeing a lot of gray areas.
Bad analogy then.
An abortion at 8 months can be justified.
What would be unjustified is if you had any say in the matter.
Anyway, back on topic…
I don’t have a problem if NASA uses their super duper ray gun to blow up an small asteroid, or an invading death star, but I have a big problem if they use it to blow up the moon or my house.
Now, what was the topic again?
Yeah, and the effect of gamma rays on man in the moon marigolds can be pretty devastating too.
Freewill?…what?…what are we taking about?
Question 2 for INcompatibilists
If it were shown that the dualist anti-deterministic populace were more readily persuaded by the second of these two:
(1) Everything in your brain is determined by the laws of physics. You don’t have free will.
(2) Everything in your brain is determined by the laws of physics. The “will” that you have, the choices you make, are the products of low-level brain machinery of which we are not aware.
… then would your opposition to compatibilism be mollified? It’s not obvious to me that (2), explaining our perception of “free will” rather than denying it, would be a less successful approach than (1).
Personally no. I care about the truth not about convincing others. That’s more a lawyer-ly thing to do (not that I dislike lawyers :)).
What would convince me is some evidence I suppose. I don’t know what that is yet. Perhaps some more tangible evidence from neuroscience and physics.
If you’re asking for more evidence you’re thinking about compatibilism wrongly. It isn’t a postulate that there is something extra, it is a *description* of *behaviour* that we already know exists.
We don’t need any more evidence of the *behaviour* that compatibilists call “free will”, any more than we need more evidence of the existence of the behaviour that ethologists call “courtship ritual”.
Then if compatiblism is a description of a behaviour I am more convinced that the argument of free will is one of semantics and that compatiblists and incompatiblists are speaking different languages and arguing about different things.
Oh sure, I’ve long been convinced that the difference between “compatibilism” and “incompatibilism” is merely semantics.
Indeed, when incompatibilists stop railing against compatibilism and go about their day-to-day life they become compatibilists.
We will do this when compatibilists explain where in their semantical definition of freewill is “free” and “will”. Where is freedom and will in a deterministic brain?
Did you not feel the will to answer and type that comment? Did that will not exist in your brain?
Would you have typed the comment without willing to do it?
Your answer is in your very act of asking the question.
BillyJoe,
Imagine:
You are captured and held captive by “Fred.”
He has you chained to a wall all day long, a window in the room shows downtown, everyone walking around going about their business.
You plead with Fred to set you free.
Fred replies: “But all is determined. There is no such thing as making you “free” that I can do for you. After all, where is freedom in a determined existence?”
Is your response to Fred to say:
“I guess you’re right. I can’t really be free so there’s no difference between my situation and those of the people outside. May as well just let me hang here.”
If that’s not your answer, and we both know it wouldn’t be…what WOULD your answer be?
How would you explain to Fred what you are referring to by asking him to give you “freedom?”
Might you start talking about the things you want to do, but can’t because of the chains? The things you could do, but can’t because of the chains? The things people outside the window are “free” to do, that you aren’t free to do, because of your chains?
Really…the freedom we are talking about is commonly understood and it’s not that hard.
Vaal
lalifefr,
You said “feeling of freewill”. That’s sounds like the “illusion-of-freewill. I accept we have an illusion-of-freewill. But there seems to be no basis for actual freewill.
vaal,
I don’t get it.
How are you explaining freewill?
Fred behaviour is the result of his deterministic brain. That behaviour can be deterministically modified by further input into his brain. My pleading with him to release the chains, could be the input into his brain that deterministically modifies his behaviour. And my pleading with him is the output of my deterministic brain.
Where’s the freedom?
Where’s the will?
Where’s the freewill?
Seems you are not adding anything to what those who are not compatibilists are saying.
@BillyJoe,
no, I didn’t say “feeling of freewill”. I said “feel the will”. The free part I don’t like to discuss because, as I say, I don’t think that “freedom” is an empirical object. “Will”, on the other hand, is an actual phenomenon.
If you had not had the will in your brain, which is obviously instantiated by neuronal activity, you would not have written your comment. Your will to write was the cause of your writing. That there were prior causes to your will to write is also obvious. But the fact that there was a prior cause to your will doesn’t mean that your will didn’t actually exist once it had been caused.
BillyJoe,
If you tried answering the questions I asked: how you’d explain to “Fred” the difference between your being chained vs being set free of the chains, you’d be well on your way to answering your own questions.
So why don’t you try to answer?
Since you keep asking, given determinism “where’s the freedom” I keep getting you to try to understand first what “free” means in normal use. If you do this, you should realize that it is a normal, valid, useful, descriptive term that makes sense even GIVEN determinism. Because even GIVEN determinism, there is STILL a difference to describe between your situation of being chained up vs not being chained up. Right?
So…just start expressing the differences…(and I gave some suggestions) and then you’ll be talking about what I’m talking about when I use the term “free” in a deterministic world.
Further, you did not answer my other argument against your position.
I’ve pointed out that when you’ve first claimed there is no moral responsibility, you’ve left yourself with no forceful argument to tell anyone they are “responsible.” You tried to characterize trying to pin responsibility as simply “an input that can influence them” – but not to anyone with half a brain who has noticed you
are arguing incoherently, since you’ve already denied the validity of such responsibility. Hence, you can make all the noises at someone you want – but your attempt to persuade someone he has any responsibility will be so much white noise, no more cogent or coherent than all the other bad arguments in the world.
Why not leave yourself with GOOD, coherent arguments by which to persuade people?
Vaal
lalififr,
But then you don’t have freewill. You only have “will”. To claim “freewill”, you have to address the “free” part as well. In any case, the feeling of will is produced deterministically by the brain.
And it’s worse than that.
There is not even a “you”. That is an illusion also.
vaal,
I don’t know what else to say, but I’ll try.
I’m not interested in discussing what “free” means in ordinary everyday language. Ordinary everyday language is dualist to the core. I use “free” in my ordinary everyday language as well. I know what it means. It means I’m using dualist language to say something in the simplest most compact way. I just used the word “I”, but I know that there is no “I” controlling my brain, telling it what to do. That’s a dualist concept, but I have to use dualist language as shorthand way of communication if I am to get on in this world.
But I’m a materialist all the same.
I’m interested in discussing what “free” really means – divorced form its ordinary everyday use. We both agree that dualism has no legs, so how can you use dualist language to prove a point about materialism. It makes no sense.
As for responsibility as opposed to moral responsibility, I have explained it as best as I can. Perhaps if I put it in non-dualist language:
The deterministic brain in the body labelled “BillyJoe” causes that body to beat up the body of another brain for fun. The deterministic brains in other bodies with labels, perhaps including “vaal”, hold the brain in the body labelled “BillyJoe” responsible for that beating and those brains punish that brain in the body of “BillyJoe” with incarceration. This will deterministically decrease the chances that that brain will use its body to beat up some other body in the future and will also deterministically decrease the chance that other brains will use their bodies to beat up the bodies of other brains.
Simple, hey?
BillyJoe:
Well it doesn’t mean anything in your strict sense, it doesn’t exist. There is no such thing as “free” in your meaning.
Thus the “ordinary everyday use” is the *only* use of the word that makes any sense. And given that that is the only sensible use of the word, that is what the word actually *means*.
Well, then, you are a dualist (whether you realise it or not) if you are defending freewill with nothing but dualist language. If there no other meaning of “free” than there is no freewill as far as a materialist is concernec.
Nope. I am not a accommodationist. Science should not be constrained by religious sensibilities.
To believe that determinism and free will are “scientific” questions is to not really care much for science.
These are fully and unequivocally philosophical questions, not scientific questions.
A scientific question is, e.g., “under what conditions can a superconducting electronic fluid form?”, “What are the atomic and electronic events that make it superconducting?”. That sort of thing. Not “Is there free will under determinism?”.
There is no true answer to that question, because the concept of “freedom” is not a scientific object.
Nope.
Science knows about deterministic cause and effect (and probabilistic QM).
It doesn’t know about freewill.
So, no freewill in science unless and until that evidence is forthcoming.
Science only discusses the question of gauge symmetry, it doesn’t care about “whether free will exists in a deterministic world” because **there is not scientific definition of freewill**. Science requires well-crafted and clearly stated theories to frame its evidence against. There is no such concept in any scientific theory.
Your “feeling of the scientific existence of a definition of freewill” is just an illusion 🙂
You’re not even listening anymore.
The whole concept of freewill is incoherent.
There can be no definition of an incoherent concept that is not itself incoherent.
We have classical determinism and we have quantum probability. There is nothing else to satisfy the requirements of freewill.
And this is quite separate from the fact that determinism is incompatible with freewill.
Where is the dividing line between the two? Hint, people called “philosophers” and people called “scientists” are often the same people, and certain “cross-polinate” literature.
Er, that’s a total non-sequitur. I’m pretty sure you’re perfectly aware of that but just couldn’t resist making a point just in case.
The fact that some scientists are interested in the philosophy of science, and that some even try their hand at “cross-pollinating” literature is a true fact.
What is not true is 2 things:
– it is not “often the same people”. Merely sometimes, probably in the same proportion that “often” scientists explain that they don’t need philosophy.
– even if 80% of scientists also practiced philosophy, it wouldn’t change the fact that having an opinion about the meaning of freewill in a determinist universe is not an activity that has any of the recognizable features that define what science is (the protocols, intellectual frameworks, standards of truth, means of enquiry, scope of conclusions, etc. are entirely of a different class). This is the reason why I say that to claim that it’s a scientific question is to not hold science (either, in fact) in very high regard.
I don’t hold it is a scientific or a philosophical question. Instead, I hold that it is *both*, so to speak, since I regard any dividing line as arbitrary.
Consider S. McCall’s _A Model of the Universe_. McCall is a philosopher in terms of academic affiliation. However, the fundamental hypothesis of the work is one about the global topology of spacetime. Is it thereby physics he is doing? He (shows/tries to show) that this hypothesis, if true, has interesting consequences for various debates, including foundational questions in quantum mechanics. He also tries to show it bears on questions of will and action (less successfully in my view, but that’s irrelevant). So which is it? Philosophy or physics? How does one tell?
I refuse to draw a line; intellectual inquiry is of a piece. There are non-, and even anti-scientitific philosophies, but so much the worse for them, as far as I am concerned. I have reservations about the book I mentioned, but not because it is in those latter two categories.
I understand your point, and I sympathize with it.
I think that from the point of view of a general enlightened understanding of the world, that’s the right way of looking at things.
I also agree that, as with many things in the universe, where something ends and the other thing starts is often unclear at all scales. That’s an important fact about the universe (which is by the way one of the reasons why I have a problem with the mind being a finite-state machine with a tape, because I’m not sure there really are only discretely describable states of things, if at all).
On the other hand, I believe that it’s important to know categories when we see them. I don’t think understanding and being able to distinguish between the epistemological status and other properties of various kinds of praxis or theoretical frameworks prevents us from considering knowledge as a general, continuous and interconnected human activity.
Categories have subcategories, and it’s not possible to think without categories. We also need to be able to recognize where certain knowledge is operational and where it is not. I mean we can look at the earth and consider it one big whole where everything is connected, where does the ocean start and where the beach, all that, but it doesn’t change the fact that an ocean is a different thing than a beach.
Interesting question and point behind it.
I take it that the compatibilist should reply either by (1) denying the Church-Turing thesis (presumably with Chinese Room-style arguments) or by (2) admitting that a sufficiently advanced computer would have free will.
For my part, I would say that if a computer had beliefs and desires, then it could have free will.
(I also think successful arguments for dualism about the mind would probably undermine the Church-Turing thesis. I’m not sure whether dualism is true–I go back and forth–but at least I’m sure that empirical observation alone cannot refute dualism.)
I say “sufficiently advanced” because the compatibilist certainly need not say that all machines have free will. We also don’t need to draw an arbitrary line; at least, considerations about that line will have to do with vagueness considerations. My guess is that we won’t be able to draw an undeniable line, but we’ll have clear extreme points: a singularity-AI would have free will, and a thermostat wouldn’t, nor would plants. The computer would have beliefs and desires and the thermostat and plant wouldn’t.
To lay my cards on the table, I should say that I’m definitely a compatibilist about “free will” and determinism; it seems fairly obvious to me that there is at least a common sense of ‘free will’ that is compatible with determinism. For example, we know from Frankfurt-style arguments that freedom doesn’t logically require alternative possibilities.
The more interesting question–as I think Jerry recognizes–is whether determinism is compatible with moral responsibility. On that, I’m not sure.
“if a computer had beliefs and desires, then it could have free will”
And, if those “beliefs” and “desires” where the products of derministic cause and effect relations within that computer as dictated by the program algorithm, then that so called “freewill” would be bereft of “freedom” and “will”, and you would be working with a semantical defintion of freewill that doesn’t include either.
Neat!
Hi BillyJoe,
Your claims are precisely what’s at issue in the debate over the analysis of ‘free will.’ I don’t see any argument for an incompatibilist definition of ‘free will’ in your comment, but if you have one, I’d be glad to look at it.
As for your comment on moral responsibility, I’ll address that in this comment, too.
I think the best way of figuring out the connection between freedom and moral responsibility is to look at a set of claims:
Obviously, one of (1)-(4) must be rejected. For my part, I’m not sure which.
Two of the three most popular quantum mechanics interpretations are indeterministic, so that’s a reason to reject (1). Nevertheless, I think (1) and (2) can be replaced by the weaker:
Even then, (4) seems pretty obvious. Then why not reject (3)?
I don’t need a definition of freewill that is not Compatibilist because, like everyone else who is not a Compatibilist, I don’t believe freewill exists. It’s the Compatibilist who has the explaining to do.
Regarding the moral question:
You beg the question by posing #4
We are discussing whether or not we can be morally responsible if determinism is true.
You simply state without reasons that it is morally wrong to beat someone for fun.
I would simply reply it is not MORALLY wrong to beat someone for fun but that we are going to hold the beater responsible for his action. The pragmatic reason is that we want to decrease this sort of behaviour. In this sense, the behaviour is WRONG but not MORALLY WRONG.
Good to know that you won’t beat others up for fun only for fear of consequences, not because you think it’s actually wrong.
But even then, why do we hold the beater responsible for their actions then? Because we want to maximize well-being, I think is what you said elsewhere on this thread. But then why do we think it’s legitimate to maximize well-being, if it’s not in itself morally right?
I mean even if it’s evolved behavior by natural selection, then that means that our morality is evolved, and therefore caused, but why does that make it non-existent?
You seem to hold the view that provided we can find a physical cause for something, then that entails that that thing doesn’t exist. But isn’t it the other way around? The fact that we can find a physical cause for something is actually evidence that it exists, isn’t it?
Otherwise, we can follow all causes back to the beginning of the universe, to the ultimate uncaused event, if it exists. Whether there was a prior cause to the universe, or not, which one of these two options makes the universe actually exist in your book?
This is getting off-topic, but morality is simply an optimal strategy (in the sense of Game Theory) for an individual to behave in a society. And, surprise surprise, going on murderous raping rampages isn’t going to get you very far, but paying your taxes and helping your little old lady neighbor bring her groceries in is going to do you wonders. Enlightened self-interest and all….
And, yes. The evolutionary advantages to being able to unconsciously do that kind of calculus should be obvious.
Cheers,
b&
They’re obvious to me, and they’re the cause for morality as an observable human behavior to exist.
P.S.: Getting even further off-topic: for people explaining how behaving as a good moral individual gets you really far in life whereas being an amoral bastard gets you nowhere, I suggest to take a real look at how society actually works and who’s running the various shows in town… Morality is a system that can obviously be gamed easily if you don’t have any.
Morality, like evolution, operates at much larger than individual scales.
Just look at what happened to the French Nobility to see what happens in cases of excessive parasitism. That a similar infection has set in again in Western society should give everybody alive today great concern, especially the parasites.
If we’re very, very, very lucky, Snowden’s revelations combined with the just-now-surfacing pushback from the EU will allow the normal immunological response to kick in. If not, we’re in for a very rough ride, indeed, with no guarantee of survival.
Cheers,
b&
Why do we want to maximise well-being?
Because our brains have evolved deterministically and function deterministically to favour well-being as a pretty reasonable way to survive.
Nothing to do with morals.
“Good to know you don’t beat others up for fear of consequences”
I don’t beat people up.
But those who do beat people up need consequences to feed into their deterministic brains to decrease the chances that they will continue to beat people up.
“That means our morality evolved”
If there is no freewill there’s no morality, so forget about morality evolving, because you are back at your starting point.
Hi BillyJoe,
You asserted incompatibilism in your comment, so I noted that you have the burden of proof in defending incompatibilism; you made the claim.
Begging the question occurs when one’s opponent has no independent reason to accept a premise.
The independent reason to accept (4) is that it’s obvious to almost everyone.
It may not be obvious to you. But, as with perception, if you don’t seem to see something that almost everyone else sees, you should start to wonder whether there’s some sort of problem with your detection apparatus.
Also: If you don’t believe in moral reasons, do you not believe in epistemic reasons either?
You have a logical argument up until 4. Then not. “Seems” is not a logical proof.
The reason 3 is wrong is because one feels it to be. If you don’t feel it to be then it isn’t.
Sorry. Mix up. The reason it is wrong to torture is because one feels it to be; otherwise not. Is what I mean to say. Moral calculus is not part of logic.
This is to say that people suffer from ‘causal drainage’? You’re begging the question of what a person is in that case (moral responsibility would be the sort of thing had by people as independent causal factors – so then what is an independent causal factor, as causality implies dependency, unless you want to revert to a constant correlation definition?).
I think the problem in these discussions of free will boil down to a confusion between self-determination and self-efficacy. I don’t think we can claim the former; I think we have a pretty good claim for the latter, call it what you will.
FWIW, I think that sums it up pretty well, thanks.
It’s not about “free from” or “free to”, it’s about “Who is free”. But BillyJoe actually also argues that there is no self. Although I suspect his definition of self is even more of a strawman than his definition of free or of moral.
In fact, ultimately, I think it’s the definition of “existence” that should be clarified.
The self is the idea that there something in the brain that controls what the brain does. That is an illusion. The self is not real. However, the illsuion-of-self IS real.
Think checker board illusion.
Those squares really do seem to be different colours. The self really does seem to be controlling our brains. Both are illusions. Those squares are not really the same colour. There really is no self controlling the brain.
You can define concepts to conform to what you want to deny in them. But in order to do that, you have to keep at least one word sufficiently vague so as to be able to leave it open to any interpretation suits your purposes later in the conversation.
There is “something in the brain that controls what the brain does”: they’re called called neurons, glia, blood vessels, ions channels, etc..
You are operating under the delusion (not illusion) that just because a phenomenon can be explained and results from a physical substrate, that it doesn’t exist, or “is an illusion”.
The term “illusion” in the context of this conversation means absolutely, and I mean it literally, nothing. If there is an illusion, who is experiencing the illusion? Is the illusion of my self experiencing the illusion of my self? It makes no sense.
Oh, I also just realized that as far as you’re concerned, colors don’t exist.
All your arguments hinge on your definition of existence which is narrow to the point of meaninglessness. With such a definition of existence, then nothing actually exists at all. And I don’t think “science” would agree with you there.
lalifefr,
You seem to think illusions are something that don’t exist, even though I’ve explained to you a couple of times now that illusions do exist. Those squares really do look to be different colours. The self really does seem to exist. What doesn’t exist is an actual self. Those squares really are not different colours. It’s not that illusions don’t exist, it’s that the things they are illusions of don’t exist.
Colour is an illusion. That illusion exists. What doesn’t exist is the actual colour. If you think colour exists, show me where? That flower is not red, it just reflects certain wavelengths of light. The image on the retina is not red it just reflects certain wavelengths of light. There is no red in the brain, it’s just nerve impulses. There is no red. But there is the illusion-of-red. And all of us with intact eyes and brains can perceive the illusion-of-red
Just to point out that nothing in the field of morality is “pretty obvious” if you look deeply enough:
Presumably the field of torturing attracts people who enjoy the work. Such a person is employed by some Authority to achieve certain ends, such as interrogation, punishment, or terror. The torturer might reasonably believe they are doing good (especially since the Authority has lots of influence on what is considered “good”) by doing this work which they enjoy. Unavoidably, sometimes innocent people get caught up in the system and get tortured.
So, an objectively innocent person is being tortured by an authorised torturer (for “national security” or whatever), and the torturer enjoys it. Does that make this torturer more “morally wrong” than one who just does it as a 9-5 job and doesn’t enjoy it?
“The more interesting question…is whether determinism is compatible with moral responsibility”
Obviously, it is not.
But we do assign responsibility to the individual.
Broadly speaking, we do this because it then becomes part of the input into that individual’s brain that will help to prevent behaviour that adversely affects other individuals.
BillyJoe
re moral responsibility being compatible with determinism:
“Obviously, it is not.”
Ok, so here you’ve declared there is no moral responsibility.
But then you try to hoodwink people anyway:
“Broadly speaking, we do this because it then becomes part of the input into that individual’s brain that will help to prevent behaviour that adversely affects other individuals.”
But how do you expect this “input” of assigning responsibility to someone have much force if your own position undermines
the very logic of it?
In other words, if, after saying there is no moral responsibility, you then try to pin responsibility on someone, any half perceptive person will say “Uh-uh, you’ve already told me I’m not *really* responsible.”
What do you do then?
All you’ve done is left yourself with bad incoherent arguments to try to persuade dumb people who hopefully won’t notice the contradictions.
(And…are you also saying there is no “morality?” I’m curious just how deep you want to dig yourself here)
Vaal
Yes he really is responsible for his behaviour. Just not MORALLY responsible. What I mean by responsible is that he did factually commit that behaviour. That’s all. And, if that sort of behaviour is to be discouraged, then action has to be taken against persons committing that sort of behaviour. That’s what I mean by HOLDING that person responsible for their behaviour.
“Yes he really is responsible for his behaviour. Just not MORALLY responsible. What I mean by responsible is that he did factually commit that behaviour.”
Ok.
So what is your argument going to be to him for why he “shouldn’t” or “ought not” have done what he did, or shouldn’t do it again?
If all you can say is “Hey, you did this” then the answer is “yeah, thanks, I know I did that. So what?”
If you are going to start telling him what he SHOULD do, then what prescriptive force will you have behind your argument? Why should he listen to what you tell him?
You’ve already hinted at something, that some behavior “is to be discouraged.” Ok, WHY is some behavior “to be discouraged?”
I don’t see how you are going to answer without necessarily getting into value statements that are essentially moral talk anyway. And how do you say someone “ought” to act in one way over another unless you admit people CAN CHOOSE to act in one way vs another? And if you admit that it still makes sense, even given determinism, to hold that people “can do otherwise”…then what in the world is your beef against compatibilism which is saying just that about morality and our ability to choose in a determined universe?
Vaal
If he beats someone up, we hold him responsible for that beating and we incarcerate him. That should decrease the chances that he will beat someone up in the future. Do you disagree?
Why should beating up be discouraged? Because, if we don’t, the chances are increased that, sooner or later, you are going to be beaten up. And I don’t think you would like to be beaten up.
No morals to be seen yet.
The issue here is really this: by determinism and the view you are saying, we don’t actually choose to hold responsible, or incarcerate, and we certainly don’t choose to consider whether we will be beaten up or whether we would like to be beaten up or not. Thus, it is equally valid for us to hold him responsible or to NOT hold him responsible, and not because we can make any kind of logical reasonable argument for why he should be held responsible or not, but merely because that’s what we do.
Thus, it seems to me that if there is no difference between a person who shoots someone because they had a chip in their head activate to make them do it and someone who shoots them because they “feel like it”, then there’s also no difference between someone who shoots people and someone who DOESN’T shoot people. In all cases, we end up doing no more than what we are doing, and reacting the way we react to others. Thus, if someone reacts to someone shooting someone else by locking them up or by cheering them on, there’s no difference there. They’re both equally valid responses, and you can’t really consider the shooter or those who cheer on the shooting any worse than those who don’t shoot people or those who lock up shooters.
The only way to start making these distinctions is to start looking at the reasons why those actions were taken, which takes you to, at a minimum, assessing the “competence” of their internal decision-making processes. At which point, you’re at least back at the starting point of compatiblist interpretations, and moving away from Hard Determinism.
But it all makes sense if you stick with determinism all the way.
No need for freewill.
No need for morals.
Consider…
The person who beats someone up for fun:
A lifetime of inputs into his brain sets up memory stores within his brain including memories of emotions which will be triggered in certain circumstances. A new input (new circumstances) combined with his present emotional state (also determined, mostly by recent events) and combined with memories that are triggered by his present circumstances, result in the (output of) him beating someone up for fun.
The people who incarcerate him:
A life time of inputs into their brains have resulted in them living relatively in peace with others. This causes them to react against the person who beats someone up by incarcerating him. This stops him from beating someone up for fun at least while he is incarcerated, and reduces the chances of him beating someone else up once he is released (because he may find himself incarcerated again), and it also acts as a deterrent to others contemplating beating someone up.
These people are in the majority. Majority rules and this is why that person is incarcerated rather than excused because he has no freewill.
In fact, most of these people believe in freewill. As a result they are also inclined to agree that beaters should also be punished.
The people who do not believe in freewill:
They have had input into their brains that have convinced them that there is no freewill. However, because they do not wish to be beat up or see other people beat up, they agree that such people should be incarcerated. The difference is that, because they don’t believe in freewill, they also do not believe in retribution. So the incarcerated person is treated humanely, an attempt is made to provide input into his brain that will reduce the chance he will beat someone up in the future.
Moral *sentiments*. Morals are feelings.
Good show BillyJoe.
Dualism doesnt’t get you out of questions about computability. In fact, this is about the only use of Fodor’s remark that ghosts could implement automata (e.g., Turing machines) for all it matters.
I think the reason certain people balk at the idea of the non existence of free will is similar to the reason why certain Christians balk at the theory of evolution: Accepting it means that — to them — the only thing that follows from that conclusion is to jettison all morality.
Since we can’t jettison morality, therefore free will must exist/Creationism is true.
…and my response is the same as it was the last time I saw this line of reasoning:
Whatever I think “free will” is, I think it is an emergent property of sufficiently developed brain function. I also think that those who wish to jettison the concept are jettisoning this property in an exttremely selective fashion. INcompatibilists here do not jettison (or even qualify) the credit they extend to Mozart, Darwin, Feynman, or their car mechanic for other emergent properties of complex brain function. Why? If they are willing to do so, is it fair to call them nihilists?
I’m not sure I follow.
Is it Nihilistic to require evidence for an assertion?
I’m wondering exactly which mechanisms in the brain are considered free will and what is not.
How do compatibilists objectively decide what to label as free will and what to label as other brain functions?
Why should these things be separate and not comingled? The evidence for “free will”? People make decisions, rocks don’t and plants do so in only the most limited of ways. (As I said above, I have no problem with a continuum and arbitrary lines – or shades of gray.) Our host seems to think that since there is a sequence of molecular events that occurs that “causes” those decisions, the decisions are “illusions” – his favorite word to describe “free will”. I disagree and, as I said, he labels this particular product of brain function as an “illusion” and does so in a very selective fashion.
I don’t think you answered my question.
At what point between rock and plant does free will start and what unit would you use to describe it?
First, you answer the same question about “complexity”, to show us how it is done, then I’ll adopt your answer for the behaviour we call “free will”.
Complexity in what sense and compared to what?
Complexity in the sense of how complex something is, and the rock compared to the plant, and a cat.
Well, if we’re talking at what point an entity is complex enough to have consciousness, then I’d say it requires a brain of some sort, but we still don’t know quite how to define it and measure it.
That’s why I don’t think free will is a term that clarifies anything. You might as well call it mind or brain or some other broad term.
No, I’m simply asking you, at what point between a rock and a plant and a cat does an entity become “complex”? The point of the question is that the only sensible answer is that it is a continuum, with no clear line being crossed at any point. Ditto re CFW.
It completely depends on what our definition of complex is and what exact situation we’re talking about.
But regarding brain functions then I’d say that a brick is zero, a plant is zero and a cat is 1.
In other words, I think it is a on/off thing just like when you declare a person brain dead.
The measurement of “complexity” is possibly an even thornier can of worms than questions of “free will.” It has very specific well-defined meanings in certain domains, none of which would apply to this discussion. The closest common definition that would fit this discussion is the Creationist one of “specified complexity,” which, I assure you, you don’t want to associate yourself with in the slightest.
Sounds like, to you, “free will” has an awful lot of “you know it when you see it” to it. Might as well try to suggest a universally-acceptable definition of “pornography.”
Cheers,
b&
(Squeezing a word in). Coel is making a good point here. When we look at different life forms that are increasingly complex (bacteria–>protists–>worms–>fish–>reptiles–>mammal)there are new behaviors and options for behaviors among the more complex species that are not present in even rudimentary forms in lower species. My point here, to add to the argument Coel is making, is that ’emergent properties’ do appear with added levels of complexity. More complex life forms are not simply fancier versions of simpler life forms.
As I said, I’m likely to accept that there is a continuum and if I were interested enough, I’d try to erect a scale involving the factors other compatibilists here have discussed: measures of a creature’s awareness of its surrounding, the extent to which it pursues its own goals, etc.
Now, why not try to answer my question. Why do you not jettison (or even qualify) the credit you extend to Mozart, Darwin, Feynman, or your car mechanic for all the emergent properties of their complex brain function? I never implied that it was nihilistic to require evidence for an assertion – I implied that if you do indeed label all the products of people’s brain function as “illusions” then you are pretty much a nihilist. I doubt that that you do – and hence my question – why not?
I don’t consider the term free will as a piece of art or as a part of physics, why in the world should I?
If we define free will as simply brain functions, then I don’t get why we don’t just call it that.
Neurons firing in the brain. Plain and simple.
Jesper,
“If we define free will as simply brain functions, then I don’t get why we don’t just call it that.”
I don’t know why you keep saying that.
No one is “defining” free will as “simply brain functions” because that wouldn’t tell us what we are talking about.
You seem to be mixing up “definition” with “explanation” (or “a result of”).
That some part of human behavior can be explained as being the result of “brain functions” doesn’t mean that term ought to become a “definition.” If someone said “I’m going outside to brain function” what would that tell you? Nothing much. But they could say “I’m going outside to rake leaves in the yard” then that narrows things down, even THOUGH the explanation still happens to be found in their brain functions.
Similarly, to say “3 women in Cleveland lived in a basement, not of their own free will” tells us a lot more than “3 women in Cleveland experienced brain function in a basement.”
It indicates we are dealing with a scenario in which those women could not exercise the choice to leave when they wanted to. And it immediately suggests their being held captive and/or under coercion/physical threat. People immediately understand this is the type of thing being described, in a way that “brain functioning” just doesn’t do, right?
Then, we can always get more specific about what specifically these women wanted/willed/desires do to, that they could not and therefore what was stopping them from exercising such choices.
So, really, I don’t know why you keep thinking that we ought to re-define free will as “simply brain functions” when that simply doesn’t capture what’s being talked about. You could generalize virtually anything someone does – writing a check, running, planning a vacation, singing a song, tap dancing, being afraid, happy, etc – as resulting from “brain functions.” But
of course we have all those other words to narrow down the type of behaviors or scenarios we want to talk about.
Vaal
Vaal
Despite our many posts back and forth, I still don’t know exactly what you guys mean when you talk about free will and who/what has it, and I honestly think we’re going in semantic circles so I’ll make this brief:
What test can you think of that would disprove free will?
Jesper,
I find it exceedingly odd that you (and some others) can be given all manner of everyday examples of the use of the term “free” and “free will” – examples in which you’d surely understand what someone is talking about normally…but then act completely baffled by such talk in the context of this discussion.
Be that as it may…
“What test can you think of that would disprove free will?”
(Putting aside the awkward demand to “disprove”…where it’s more strictly scientific to ask “How would I test for it?”)
I already addressed that the last time you asked, with a specific example.
“Free Will” is just a blanket term for all the specific situations in which people can be said to have the ability to choose one course of an action that would fulfill their desire over another. To test this, you’d have to test single instances or claims for free willed choice making.
I gave the example of saying I chose to eat Cheerios in my kitchen yesterday instead of Corn flakes, and that it was a free willed choice. To say it was a free willed choice entails I could have chosen not to eat the Cheerios, or eat the Corn Flakes instead, had I desired.
To test this, put me in similar conditions to the ones in which I claim I can make such a choice, and observe if I choose to not eat Cheerios, and/or choose to eat Corn Flakes instead. If I can, I’ve provided testable empirical support for my ability to freely choose those actions. If I can not stop myself from eating the Cheerios in front of me, or can not choose the Corn flakes, then my claim is undermined. You can say “you’ve given no empirical evidence that you had that ability for free choice.” (It’s a bit rash to go from that to directly concluding no such ability exists. But as I said last time, IF you get null results
then you can go on to test other hypothesis you might come up with for why
I found it impossible to choose anything other than Cheerios, and you may find a mechanism that suggests it will never be possible for me to do so, which would mean
I was wrong to think I had the ability to “choose otherwise – choose other than eating the Cheerios.”
Can you point out any flaw in this, vs only re-expressing your befuddlement, I hope?
(Because I’m at a loss as to how to explain it any more clearly).
Vaal
Vaal.
Here’s the problem with your test.
Outcome 1: You eat the cheerios = free will.
Outcome 2: You eat the cornflakes = free will.
Outcome 3: You eat both the cornflakes and the cheerios = free will.
Outcome 4: You eat neither = free will.
No matter what the outcome you will conclude that free will exists simply because nobody is forcing you to do anything.
The game is rigged.
Jesper,
“Here’s the problem with your test.
Outcome 1: You eat the cheerios = free will.”
No, because in that case I will have failed to support my claim I was free to choose not to eat the Cheerios, or choose to eat the Corn Flakes instead.
Hence I would have failed the test, designed to provide empirical support for my claim to have that capability.
(Which is the scientific way of approaching such questions. If someone claims he has some ability, be it to run a marathon in a certain time, or read people’s minds, then you don’t ask for “disproof” of the claim; you ask for evidence, empirical support, for the claim, e.g. through tests like the above.
Failing the tests is a failure to support the claim).
Vaal.
Vaal
I simply don’t understand what you mean and how it is supposed to reflect reality.
Semantically you can make any outcome fit your disposition be it free will or no free will because this purely is a semantic exercise that’s going in circles. We’re back to the hypothetical outcome-generator.
Maybe it’s because you guys regard the matter as a continuum and I don’t.
Frankly I just don’t understand how you come to the conclusion that anything called free will exist in our causal deterministic universe, and given the subjective nature of the predicament I suggest we leave it at that.
After all, so far it is a matter of opinion, not fact.
> At what point between rock and plant does free will start
> and what unit would you use to describe it?
At very same point that sentience begin.
Pico-erg… since action/stasis is akin.
imo
They link it to choice, especially conscious choice. Whose mechanisms we haven’t mapped yet, but that we DO have some understanding of.
So you’re looking for mechanisms in the brain that are responsible for the actions we take and you wan’t to label them as free will mechanisms?
They consider the mechanisms that are important to free will to be, in particular, the ones that involve conscious choice. So they look for those in the brain and that’s how they distinguish them.
Alright, so consciousness is a requirement for free will.
How do we then decide what constitutes free will in non-humans?
Not necessarily. It depends on what is meant by consciousness. CHOICE is a requirement for free will, and in humans the most pertinent choices are intentional and conscious. That may not need to be the case for non-humans.
There’s a lot to unpack here, more than can be properly unpacked in a small set of comments.
Alright, help me out here then.
What processes of the mind are not considered free will?
Um, you mean you can’t get that from my telling you want processes ARE considered to be involved in free will?
I’m being a bit short, but for the most part you don’t seem to be interested in actually engaging the topic, but instead to just spin out lots and lots of questions and challenges and hoping that something will stick.
I’m trying to figure out what you guys think constitutes free will and what does not.
Just ignore my posts if you find my questions annoying.
In short: I’m having trouble understanding how free will is defined as a consciouss choice, but consciousness is not required.
I’m a bit at loss here.
Um, free will is defined at its base as choice, and in humans as conscious choice. Which is what I said. Consciousness may not be required for it, but the obvious choices that we care about in humans are conscious, and are through conscious deliberation. Again, for non-humans that might not be required.
Conscious choices are the obvious examples, but that doesn’t make them the only examples or consciousness necessarily for the right sort of choices. Now, what are the right sort of choices? Again, that’s far more detailed than can be said in a small set of comments.
So free will is considered the right kind of choices.
I just don’t understand how you objectively decide what choices are free will, and what choices are not.
It’s a long process of deep examination coupled with empirical investigation … and it isn’t actually COMPLETE yet. Hence why you aren’t getting set answers to questions that are actually really deep and important questions in the philosophical field. A lot of your questions, here and elsewhere, are considered challenges to the theory, and not merely questions.
BTW, I’m not a compatiblist …
Well, so far I wouldn’t call it a theory.
I consider it a semantic philosophical exercise.
But, then, you don’t seem to know what the positions actually are, so your judgement seems .. uninformed.
Consider me uninformed then.
I guess I’m not sophisticated enough to understand what is meant when anyone says “free will”.
I think it’s got to do with the brain, but a brain is not required, so yes I’m none the wiser.
In humans, it has to do with a brain. In other things, it might not. I’m not sure why you’re having such a problem with that.
I apologize for being a bit snippy over this, but the positions are pretty clearly spelled out in introductory books and courses on this, and the constant arguments from the FWDs accusing FWCs of holding positions they clearly don’t hold is frustrating to me.
Well, I haven’t accused any other readers for holding positions they don’t.
My mission simply is to figure out what is meant by the term, but I’m afraid It’s a futile attempt thus far.
Luckily, we don’t have to agree on everything. 🙂
My longer comment in this thread should at least sum it up without getting into the details. Try that and see if it helps.
I’ve followed these free will threads pretty closely si I’ve probably read it already, but I’m willing to give it another go.
What post exactly are you referring to?
Comment 46.
Got it.
It didn’t clarify to me what the term free will is supposed to mean other than when we make a choice that is more or less constrained by other factors.
Example 4 is supposed to be the maximum display of free will compared to the other options, but we’re still struggling to define it and how to measure it.
I still fail to see how the term free will is supposed to clarify and distinguish matters of the mind from other matters of the mind.
It all seems a bit too arbitrary defined for my taste.
The important thing is that free will matters are actions taken based on reasons. I’m not sure what other “matters of mind” you have in … heh … mind that is confusing you. And it isn’t arbitrary either, especially since for all of those 4 cases you can see how the reasons the person taking the action has for taking that action matter to how we consider and react to the choice.
Again, I have no idea what’s confusing you, but you might simply need to get it in a more systematic form and neutral form than what you get here.
What’s confusing me is the term free will.
I get that some decisions are more or less constrained than others, but I don’t get how you classify some decisions as free will, and others as not free will.
Okay … ignore the term “free will”. Treat it like a technical term in the debate. Does that eliminate any of your confusions?
My beef with the term is that it leaves me in an oxymoronic state where it’s neither here nor there.
But from now on I’ll think “more or less constrained choice compared to others choices”.
Why we call it free will is beyond me.
Historical reasons. That sort of constraint was what people thought was important and meaningful and jived with our experiences. Out of that grew dualistic free will. Now, those who think that we don’t have dualistic free will still want to maintain that those experiences and ideas are still important, and figure that it will just confuse people to have to invent a new term for it while saying “You don’t really have free will”. That sounds like them saying “Those choice examples don’t matter, except they do” which you can see would be confusing [grin].
Note, however, that FWDs like Jerry Coyne and Sam Harris talk, at least, like there isn’t any meaningful difference between the 4 cases I talked about.
They don’t differentiate between situations?
Or are we talking about morality now?
Saying that because there’s no free will that means that people don’t deserve credit for what they do doesn’t follow. That’s not determinism, that’s fatalism.
“INcompatibilists here do not jettison (or even qualify) the credit they extend to Mozart, Darwin, Feynman, or their car mechanic..”
Firstly, we are not “incompatibilists” we are not “compatibilists”, meaning that it is compatibilists who have the explaining to do.
But to answer your question:
We don’t jettison credit for the pragmatic reason that giving credit becomes part of the input into the brains of those individuals – and vicariously into the brains of other individuals – which produces output from those brains (behaviour) that increases the general well-being of everyone.
And it’s all based on determinstic cause and effect!
Neat hey?
Really? You extend credit to Mozart, Darwin, and John Lennon “giving credit becomes part of the input into the brains of those individuals”? And this increases well being of dead people? Or I guess it is all for the benefit of other individuals – not yourself. ‘Cause deep down, you know “intellectually” that their artistic and intellectual achievements are really “illusions”? Does that make you a closet nihilist?
Firstly, you forgot the mechanic! 😉
Secondly, where did I say the products of those individuals are illusions? What can that even mean? The Theory of evolution is there. Relativity is there. Imagine is there. My car is fixed. They are not illusions!
Thirdly, do you even have a coherent argument?
Imagine is there.
Yup, and when John Lennon composed it, he made (probably thousands of) decisions – all accompanied by whatever molecular mental processes that occur in the process of creating a song. Yet, no one – least of all our host – qualifies their admiration (or disdain, as the case may be) for the composer by pointing out that the composer’s actions were preordained by some deterministic chain of molecular processes in their brains. It isn’t as if Lennon had a choice in the matter, so ‘big deal’. On this web site, only “free will” gets singled out for such a qualification.
I could have been clearer in that comment. Basically, I think the attribution of “free will” to a person is little different than what we do all the time when ascribe the fruits of any of their thoughts to that that individual – and the people here who insist that free will is an illusion are just as likely to express their opinion about whether an individual is a great composer, mathematician, scientist, or (lest I forget 😄) car mechanic, as are the compatibilists. But if knowledge of the molecular mechanisms of thought somehow makes the agency of the person in whom the thoughts are formed an “illusion”, they why do incompatibilists form judgments at all … and I think they do form such judgments whether or not they ever communicate their judgments to others. In other words, as you might have gathered, I don’t buy your argument thst people (incompatibilists included) make such judgments to increase everyone’s well-being.
Freewill doesn’t get singled out for qualification. How can you qualify something that you say doesn’t exist? There is no freewill. Period. No qualification.
And, yes, we admire John Lennon and his song “Imagine”. But was determined. It couldn’t have been otherwise. But our admiration for both John Lennon and Imagine is not lessened by that fact. It is still a great song. He is still a great musician. We can’t imagine writing that song ourselves. The lyrics and the music resonate with us.
Is a rainbow any less beautiful when you know what produces it? Or does it just add another dimension.
But pray tell: where is freewill in all this? Do we appreciate John Lennon and his song only because he freewilled it into existence? He could have not produced it but he did? Good old John Lennon? He was talentless but he managed it anyway? He suffered to produce that song butnhe produced it anyway? His lyrics were written with blood, sweat and tears? His concerts were a drag but he performed them anyway? Good old John Lennon? Despite the sheer misery it caused him and the lack of any joy in producing his music, he freewilled it into existence anyway.
You’re not making any sense.
Thanks for clarifying for us which straw man has your preference. There are so many out there.
In addition to really poorly defined verbage, the whole free will argument seems to entirely depend upon a presuppostion by all parties that each person is an individual will in the first place — which is arguably not at all obvious.
Likewise, if a computer is or isn’t conscious/aware/willful, what about a computer network? (each modern computer, in fact, is constructed from several processing units, which means even a single computer is still a newtwork of potentially free-willed processors). Can we similarly break down our bodies into sub-units that might better make each of us a small community of wills, rather than an unified single individual will? Going the other way, can a traditional human community of many individuals be said to have free will as well?
I don’t think we understand what we mean by “will” or “conscious” or “individual” to even have this conversation let alone reasonably conclude something here.
“there is no fundamental difference between what goes on in human brains from what goes on in any other computational device.”
That, my friends, is b.s.
The brain is organized in a manner fundamentally different from artificial computing devices. Indeed, I’d be leery of calling the brain a “computational device”, even metaphorically.
Given this fundamental flaw in the argument, all the rest of it falls to the ground in tatters.
No, that’s not crucial for the argument. All that matters is that the brain is “wired” to give a single output for a single input or set of inputs, and so are computers. They are both deterministic devices.
The difference in organization of brains and computers is totally irrelevant.
I agree.
When the wiring is constantly rearranging itself under inputs, is itself an input to itself, and does not have finite start and stop states (or finite states at all for that matter), it’s hard to see how it can be represented by a theory of computation that requires strict distinction between inputs, processes and states.
I’m not saying I’m certain it’s impossible, I’m saying that to be certain that it is possible requires a pretty wide leap of faith in the current understanding of both computation and the brain.
That’s because you’re misunderstanding the theory.
The theory cares not how you’ve obfuscated the different functions. It cares what the de-obfuscated state of those functions is.
Self-programming code is just a very complexly recursive algorithm. And, even if the architecture in question is playing fast and loose with what it considers code and what it considers input, the logical separation is quite clear. In many such cases, what you’re misconstruing as input is really just the program copying itself to the tape and reading itself back in from the tape.
It may be messy. It may be too messy for you or any other human to unravel. But it’s not magical.
Cheers,
b&
Look, I’m just going to answer this last comment of yours and pack it in for this thread:
You now refer twice to magic. I contend that you’re the one who is really relying on magic to support your worldview. Let me explain.
Science is about actually figuring out how things work. not making grand claims about the nature of things. You make sweeping statements about the grand nature of the brain, claiming that we already know everything we need to know to make the right determination about this. When I detail challenges about these claims, your come back is that I don’t understand all the stuff that you understand. Like an astrologist or a homeopath explaining to the scientist asking for actual descriptions of how this shit works, that they don’t understand the grand theory of the memory of water or the zodiac’s deep meaning.
All the evidence that we actually have about how humans (and other living things) actually behave shows that there is something about how we form intents based on our representation of the world, it’s regularities, information we have about the future and about what things mean to us, that points to the fact that there’s something that we have not unravelled there. There is an explanatory gap. However much you claim that there isn’t, you’re certainly not explaining to me how physical entities encode meaning in physical quantities under the constraints of conservation laws that fully determine the meaningful content of messages. You have to explain that in order to claim full, “non-compatibilist” determinism. By the way, I don’t have a view as to whether it will be possible one day to have artificial consciousness, I certainly don’t believe that it’s impossible. I do think that it’s not with finite-state machines writing 1s and 0s on a semi-infinite tape that we’ll get there.
The view that I spin in this thread is that when the universe managed to create agents that contain operative (future) representations of itself, thus creating abstract (literally “formal”) causal objects that are not determined by physical conservation laws going from the past to the future (which are the only ones we know about) then the future becomes undetermined until it is computed **in real time**.
To claim that this comment thread was already predetermined from the big bang’s initial conditions without the *meaning* of this comment thread having any causal role in its existence is a naïve understanding of what causes and information are. I’m not even sure that this is your position, given that you keep avoiding issues and instead take a posture of the authority figure out to teach me something.
As far as I can tell from my questions and arguments that you keep ignoring, accusing me of magic thinking instead, and the banal analogies that you use, your views draw from a narrow field of knowledge consisting of information and computing theory. If all you have is a hammer, everything looks like a nail.
I see no curiosity in your argument, just a conviction that you have it all figured out already.
That’s it for me, I’m done on this thread.
Thanks for initially making the effort to engage in my arguments, I did get some interesting clues from you.
Do we know everything there is to know about consciousness? No; obviously; of course not.
Do we know everything there is to know about physics? Again, no way.
But: do we know everything about physics we need to know to understand the fundamental forces and properties at work in consciousness? Absolutely, unquestionably, and with as much certainty as that the Sun will rise in the East tomorrow morning.
And, since we have the map sufficiently filled in to know that with such great confidence, we can similarly rule out magic. Magic would include anything not currently explained by human-scale physics…
…and it’s exactly those sorts of super-physical explanations that you and Stephen Barnard have been suggesting.
It’s not a case of, “We don’t know, so therefore anything is possible.” It’s a case of, “Here’s the broad outline and a hell of a lot of details filled in, and here’re the specific large-scale properties of the parts that are still fuzzy.”
One of those particular details is that human brains are Turing-equivalent devices. We don’t know the specifics of what’s in the state table or the initial configuration of the tape or the like, but we know without doubt that that’s exactly what the human brain is logically equivalent to. And we may never fully understand it; Turing’s good friend, Godel, gave us some very good reasons to suspect that we may be logically incapable of fully understanding ourselves.
But that doesn’t at all mean that we are therefore something else entirely different.
Cheers,
b&
The brain is similar enough to computational devices that the analogy works.
The real error lies in people who think that the brain is organized in a way where there is a single sort of “main decider” driving the entire thing, like a pilot in a plane. That idea is the real reason why people think that free will exists. However, the brain is modulated (just like sophisticated computer systems are modulated); there’s no single decider running the show.
Another analogy would be that your brain is something like Congress where there are a slew of competing desires due to having to appeal to different constituents. In this sense, Congress is also “modulated”. And the “you” that feels like it has free will is more like a press secretary who is explaining to the public why Congress arrived at the decision that it did.
For example, there are two different “modules” for desire and enjoyment. The common wisdom states that how much you enjoy something dictates how strongly you pursue it. But that’s actually not the case; you can like something a lot but not really go after it or vice versa, since those two components are two different modules in the brain. Which of those two modules is responsible for “free will”?
Thanks for that link to desire and enjoyment. Definitely food for thought…especially for figuring out how I could get myself to do more productive things than type away at WEIT….
b&
“The brain is similar enough to computational devices that the analogy works.”
I’m amazed at how people who claim to be scientific advocates, and who will argue against e.g. creationists that “scientific knowledge is always tentative, driven by hard facts and evidence, godamit!!”, have such a difficult time actually telling the difference between true verified statements and tentative theoretical statements.
When you say the above quoted sentence, what you really mean is to preface that statement with “according to at least one of the current theories of the brain, it is similar” etc.. The scientific and other empirical data that we have to date is very far from sufficient to substantiate such a claim.
To confuse a reasonably sound theoretical framework for performing experiments, with the actual truth of the universe strikes me, I’m sorry to say, faith-based reasoning.
If you can elaborate on what it is that brains do that’s fundamentally different from what computers do, I’d really appreciate it. So would every information theorist on the planet.
Cheers,
b&
If you accept the premise that the mind is equivalent to a Turing machine, then you must accept the premise that a Turing machine can be conscious, which is absurd.
What on Earth should be so absurd about a conscious computer? What’s so special about consciousness that only humans can be conscious?
Do remember, this is a theoretical discussion. The question isn’t whether the PC or tablet you’re reading these words on could be conscious, but whether, in principle, a computer could be built with consciousness.
b&
I’m not sure you understand what a Turing machine is.
Well, it’s obvious that this conversation is going nowhere fast.
You’ve yet to do more than handwave your objections, while I’ve striven to provide plenty of detail and at least oblique references.
Even if you really do know your shit, you’re doing everything possible to suggest otherwise.
Cheers,
b&
“the “you” that feels like it has free will is more like a press secretary who is explaining to the public why Congress arrived at the decision that it did”
Good analogy.
Reblogged this on HUMAN RIGHTS & THE SIEGE OF BRITAIN POLITICAL JOURNAL and commented:
Daisy May: Can we have some popcorn tooo my favourite is Tuna flavour meowwww prrup prrup purrrrrrrrrrr
I think some computers may have free will. At least I’ve owned several that would spontaneously do things that had no apparent cause and which I certainly didn’t request/ command.
We need to separate questions of agency from questions of contra-causal freedom. Once we do, our situation will be much clearer, and we won’t need to invoke compatibilism. First, do humans have contra-causal freedom? No. Second, are humans agents (i.e., can they respond to reasons)? Yes.
This clearly shows us how we are like (and unlike) thermostats, without getting tangled up in the murky definition of ‘freedom’ as used in ordinary language.
I’ve learned a lot from some of the many excellent comments here, but I’ve got to add something here: The idea that physics is deterministic is a bit of a simplification.
I’m a physicist, not a philosopher, so I’m reminded of Feynman’s quote to the effect that if you think you understand QM, then you really don’t. In terms of what we do know, QM is basically probabalistic (we cannot say for sure what one particle will do, although we can often predict the relative probabilities of the possible outcomes very accurately). The fact that on the right scale things do seem to be deterministic is really just an average over large numbers. Schrodinger described this beautifully almost 70 years ago.
The uncertainty principle is worth mentioning too: It’s not just that we don’t know enough to be able to (precisely) predict the future of a single particle, but that it’s not possible to know enough to do so. If you think that there are some circumstances in which you could, then you’re really arguing that the last 100 years of physics are wrong. Now, that’s a perfectly fine position, but it does come with a heavy burden of proof.
So, does any of that apply to the concept of “free will”?
Behavior cannot be deterministic in one sense – if it’s governed by the laws of physics, then running precisely the same “experiment” twice (if that were possible) could give (at least slightly) different results. But probabalistic outcomes are not free will; Jerry’s comments on Tse go through that well.
Anyway, returning to the question: I think my answer is that yes, you could make a computer that has “free will” to the same extent that we do. But this series of posts has definitely made me question what I’d mean by that! So, I might also agree that “free will” is the wrong term.
So, say we accept that there is some influence at the quantum level on us at the classical level that could mean that even if we knew all the variables to predict an outcome, inevitably there would be something that possibly could change that outcome.
I still think this is determinism because the unpredicted outcome is another type of “hidden” variable. If you could actually “roll back the tape” you’d need to account for that piece that cannot be accounted for but that’s a constraint on our abilities I suspect…..
Or I could be full of crap. I’m neither a physicist nor a philosopher. 🙂
I like your choice of the phrase “hidden variable” very much. The usual interpretation of Bells theorem is that there cannot be any (local) hidden variables. Although it has been disputed, it is quite possible that no deterministic theory can reproduce QM.
There are also big questions about the role of the observer: One version of the Copenhagen interpretation is that the wavefunction doesn’t collapse until a conscious observer is involved.
The role of a conscious observer in “collapsing the wave function was flirted with for a short time. But no serious quantum physicist pays it any credence anymore. Only New Agers and pseudoscientists.
We’ve covered this before.
QM, being probabilistic, does not lend any credence to freewill. “Decisions” based on the throw of a dice cannot constitute freewill.
In any case, there is no evidence that QM influences brain function.
So, I assume you didn’t read the bit where I said, “But probabalistic outcomes are not free will; Jerry’s comments on Tse go through that well.”
Also, I think your statement, “In any case, there is no evidence that QM influences brain function.” is false. In one sense, you are trivially wrong: All macroscopic events are the sum over many QM ones. But, as just one example in the post you were responding to, I had mentioned Peter Tse; whether you believe his reasoning or not, he makes a case that some receptors are sensitive to single atoms and QM effects, which seems to contradict your “no evidence” claim.
I will enjoy my popcorn too, since I’ve sworn free from philosophy and so “compatibilism”.
But I note that Turing solved this one too, since he proposed the Turing tests for (human like) sentience. If blind questioning deems the computer a human, there is a non-arbitrary point for sentience.
Maybe free willies swing from that point?
It is (the) “Turing test”.
And that is likely something one can generalize, if one wants to claim that animals have free will. In as much as ‘easy’ it is to emulate an animal’s responses.
Anyone who’s worked with thermostats in an experimental setting will concur that thermostats don’t have a free will, they have a free won’t. A wicked vicious one.
And the popcorn is Ben’s subtle way of pointing to the halting problem as an experimental indication that brains are Turing machines: present a body-equipped brain with a constant supply of fresh popcorn, and there will be no halting. Until thermodynamic death, or until the popcorn supplies run out (whichever comes first).
Dennett talks about Avoiders. It is elegant, if slightly complicated. Maybe it is the way things happen.
But, I wonder if “Free Will” is simply little more than a sensation that brains have — a feeling associated with the internal negotiations between physically competing modules. This is a sensation that ultimately presents itself as a sort of post hoc awareness different parts of the brain have about how an event that took place.
Nodding to Dennett’s Avoiders, it might be true that brains simulate multiple “futures” spontaneously. The accuracy of the simulation being a function of data points, I suppose. The physical, neuronal “futures” cascade from an event/action. But the “future” simulation is not clearly linked to the Act *decision*. Even then, with such a link, one can only point to a post hoc match:
A consequence ?== a feeling of having made a decision ?== a simulation that took place.
We want to know whom to blame, as Dennett says. This is the *only* reason why the Free Will debate exists! Actions have consequences, but even proponents have problems deciding what caused this or that “Free” decision!
Sheesh. That went long.
” if humans have free will, the only possible conclusions are that all other computational devices”
There is a little more to it than just computational devices: consciousness, the ability to make choices, self-reflection, maybe a few more. But yeah, I easily can imagine robots and animals having free will. Since when is Homo Sapiens unique?
Arguing that free will doesn’t exist because only humans are supposed to have it is rooted in christian theology. Weird that you atheist determinists can’t get rid of it.
Who’s arguing that only humans have it?
It would be somewhat oxymoronic for an incompatibilist to argue that humans have free will sinse we don’t think free will ( as usually defined ) exists.
*since
That’s not at all how I phrased the matter in my post that Jerry quoted.
In particular, I most emphatically did not claim that humans are somehow special. Rather, I set up the juxtaposition that one can either see that humans do the same thing as everything else that thinks, or adopt some form of supernaturalism. I then observed that the most common uses of the term, “free will,” implicitly or explicitly separate humans from other thinking entities and embrace supernaturalism.
So your choice is either to wrestle the supernaturalistic term, “free will,” from the supernaturalists, or to join them and convince them that the term that they use to separate humans from the rest of the natural world really should be used to indicate that there’s no difference.
Cheers,
b&
A comprehensive failure of comprehension.
I suggest you read the post again.
Perhaps the difference between compatibilism and incompatibilism is not a real difference at all, but a matter of which level of reduction is emphasized.
Incompatibilists may acknowledge that there are meaningful differences between the organization of matter that is a rock, and the organization of matter that is a brain, but ultimately all interactions reduce to ones of mindless matter.
Compatibilists may acknowledge that ultimately all interactions reduce to ones of mindless matter, but rocks and brains belong in different categories because of what those systems are capable of at a macro scale.
You do realise that you’ve managed to avoid mentioning freewill altogether.
Taking the high middle ground hey?
No, this is not an Argument to Moderation. It is not really an argument about the existence of free will at all (which partly explains why I made no mention of it). It is mostly an observation about the nature of the debate.
Perhaps this is obvious and I’m wasting everybody’s time explicitly making this observation, but based on what I’ve seen of in/compatibilist arguments, they are not actually opposite positions. The argument seems to be more about what level of reduction deserves priority.
“Fixture”…lol.
The Church-Turing thesis says nothing about biology and brains and minds. This is a huge overreach. Whatever brains do, they don’t do it like Turing machines, and it’s by no means settled that what they do can be reduced to a Turing machine. In fact, that is unlikely to say the least.
I’m not sure you understand Church-Turing. Simply put, it states, “Every effectively calculable function is a computable function.”
If you really think that C-T says nothing about minds, then the only possible conclusion is that you think that minds can calculate things that are not computable by a Turing-compatible device.
What, specifically, do you think minds are capable of calculating that no Turing-compatible device could even theoretically compute?
Another way of putting it…imagine an incredibly advanced computer, so powerful and so sophisticated that it could construct an atomic-level simulation of several humans sitting together in a closed room. What would several actual humans sitting in a practically-identical room be able to do that the computer-simulated humans couldn’t?
Cheers,
b&
I’m very familiar with the Church-Turing Thesis. My professional background is theoretical computer science, specifically AI and the theory of computation.
Regarding what brains do that Turing machines don’t, it’s consciousness.
Your conjecture about a super powerful computer that could emulate a brain down to the finest physical level is just that — a conjecture — and one that I very much doubt could ever be realized in practice.
I’m puzzled.
You claim to specialize in the theory of computation and presumably would therefore have familiarity with Turing devices, and yet you’re now objecting to the possibility of conscious computers solely on the physical practicality of building a mechanical device with a sufficiently long (but still less-than-infinite) tape?
Very well.
Let’s try some other possibilities.
Imagine not a digital computer running a familiar type of simulation, but a steampunk-style machine with as many mechanical neurons and ganglia and what-not as a human brain. Would it be conscious? How ’bout if we start using techniques from nanotechnology to miniaturize it? What if we abandon the requirement for atomic-scale simulation and simply go for cellular-level simulation? (You do know that we can already model insect brains at least at that level, don’t you?)
And you still haven’t identified what it is about consciousness that makes it incomputable.
Cheers,
b&
There are many examples on uncomputable functions. The set of computable functions is countable. The set of real numbers is uncountable, so most real numbers are uncomputable. Other examples are Kogolmorov complexity and the halting problem. Is it unreasonable to suppose that conscousness is another?
Okay, now I really do know that you either don’t know what you’re writing about or you’re really, really bad at communication.
Irrational numbers have infinite decimal expansions. Of course they’re not computable. And your other examples are very close cousins to irrational numbers and are uncomputable for very similar reasons — namely, that they would require infinite resources. (Incidentally, that requirement for infinite, or even transfinite, resources is at the heart of any physics-related analysis of the C-T thesis.)
Either you’re claiming that a human could reasonably tell you the last digit of π or solve the Halting Problem — which would be an inevitable consequence of your suggestion — or you’re just trolling.
Methinks I’ll go with the latter.
Cheers,
b&
You have a distinct impulse toward insulting language. I’ll ignore it.
You think that the mind is a computable function. I think you have no scientific, philosophical, or mathematical justification for that. If you do, show it. Are we supposed to just take your word for it? It’s not my duty to refute it. Extraordinary claims require extraordinary evidence.
I’m a materialist, but philosophical materialism does not require universal metaphysical computability.
I’ve repeatedly presented exactly that.
But…again again again.
The laws underlying the physics of everyday life are completely understood. Those laws permit no mysterious, unknown forces to act upon the brain. We could, therefore, create a computer model of a brain to within the necessary resolution. One of two possibilities would ensue: either it would behave exactly as expected, thus demonstrating that the brain is a Turing-equivalent device; or it would be incapable of doing everything that a brain would do. The latter would constitute irrefutable proof that everything we know about human-scale (and beyond) physics is radically worng.
I have at least as much confidence that our understanding of physics at these scales is complete and correct as I do that the Sun will rise in the East tomorrow morning. To within any rounding that the human mind is capable of grasping, that amounts to 100% confidence (even if that number has been subjected to rounding).
If you have anything concrete to suggest as to what type of physics could be operating in the human brain that would permit it to do something that a Turing Machine couldn’t, even in theory…well, there’s at least a Nobel in it for you. For that matter, what you’re suggesting might be possible is logically indistinguishable from the Christian soul — and it’s my deep contempt for such childish nonsense that you’re likely picking up on.
Cheers,
b&
You’re dealing in metaphysics. You posit a grand super-duper computer than can emulate physical reality, including mind. As a materialist, which I assume you are, you must believe that this super-duper computer must be embodied in something physical. But it can’t be our physical world, because that’s what’s being emulating. It must be some greater super-duper physical world. And so on, ad finitum. Nonsense.
And don’t play the God gambit on me. I’m an atheist, and I’d carry a card if they gave one out.
We’ve circled back to this….
I’m sorry, but somebody claiming expertise in information theory who insists that something can only be considered Turing-equivalent after a real-world physical model of it has been constructed is no more credible than a “biologist” who is “withholding judgement” on Evolution until macroevolution has been observed in the lab.
Take that as an insult if you must, but I just can’t take you seriously at this point.
b&
All uncountable sets, including but not limited to the real numbers, are necessarily noncomputable. Mind states could plausibly be uncountable, and therefore not computable.
You’re now peddling infinite — nay, transfinite — minds, which is full-throttle woo. Whether Christian or Newage or whatever is irrelevant; this is Chopra territory at best and, as such, deserves no further comment.
b&
Believing that human brains are like computers sure seems to involve a lot of imagining of things that don’t exists.
I find that ironic.
It can only possibly seem that way if you’re unfamiliar with information theory. The field is essentially the physical realization of logical systems, and, as such, has a far more solid theoretical foundation than anything in physics. You can have far more confidence, for example, in the insolubility of the Halting Problem than you can in gravity. For, while consistent physics could be constructed with different gravity than the one we’re familiar with, no universe can even theoretically be constructed in which there is a general solution to the Halting Problem.
C-T hasn’t been formally proven yet, but it bears some very strong hallmarks of being equivalent to the physical conservation laws. That is, if you can violate C-T, chances are superlative that either you can use said violation to power a perpetual motion machine or that you’d need a perpetual motion machine to violate C-T. If that actually turns out to be the case, then betting that there’s no fundamental difference between brains and computers is every bit as safe as betting that there aren’t any perpetual motion machines.
Cheers,
b&
I’m not specialist for sure, but even I’m aware that C-T only applies to a small subset of functions i.e. total recursive ones. In fact, the C-T argument is rather tautological, since basically the set of these functions is defined by C-T as Turing-computable.
Surely you’re not claiming that these functions are the only functions that run the universe in general?
There are many uncomputable functions routinely computed by the universe in real time (leaving aside us mere conscious beings), so I’m not sure what you are proving re the computability of consciousness by invoking C-T.
In fact, one could argue that C-T would hint the other way, i.e. that it claims that Turing machine can **only** compute total recursive functions. That, plus the halting problem, makes invoking C-T to prove that consciousness can be implemented in a Turing machine rather risky.
What am I missing here?
I do believe what you’re missing an understanding of C-T, or perhaps recursion. C-T is simply the hypothesis that anything that can be computed by any means can be computed by a Turing-equivalent device with sufficient resources (and, by extension, all Turing-equivalent devices can emulate each other to within the limits of their resources).
And, in case it’s the recursion part you’re hung up on, keep in mind that any algorithm that doesn’t require recursion can just as easily be made recursive by wrapping it in a loop that says, “If you don’t yet have the answer, go back to the top.” An algorithm that doesn’t require recursion will simply never have a need to go back to the top.
Cheers,
b&
Thanks for clarifying. Helps me drill down to the core of what I see as the problem. I thinks it boils down to two questions:
1- Can the real universe ever “go back to the top”? Not if time is directional, which, at least for all intents and purposes, it is. It seems that this weakens your claim that all algorithms (which are not the same as functions, are they?) can be made recursive as applied to the physical world.
2- Isn’t the definition of “can be computed” highly correlated with what a Turing machine can do? If yes, then we are again in tautology-land.
It seems to me that when we talk about what computation is (and this is still a topic that I have not researched well enough), we have to base a lot on the concept of information. Yet information is a timeless layer of abstraction that sits above the time-bound actual real universe. Doesn’t blindly assuming that what we can prove in the world of information can be applied without constraint to the physical universe lack some reasonable measure of caution?
Ah — that clarification from you is important. Because, you see, I once labored under a similar misconception.
You see, there simply isn’t any such thing as an ideal Platonic abstraction of “information.” All there is is time-delayed communication.
You might think of a pattern of ones and zeroes on a hard disk as information. Backing up a step, you might even think of that same pattern of ones and zeroes simply existing “out there” somewhere. But it doesn’t: it only ever actually comes into existence when it’s physically realized, just like anything else. That might happen in your imagination…which, again, is a series of electrical impulses in your brain that correspond with that pattern, no logically different from the ones and zeroes on the hard disk.
But, even once we get past the Platonism, we’re still left with the problem of meaning.
If you’re familiar with cryptography, you should know about a One-Time Pad. And the reason that a One-Time Pad is unbreakable encryption is that all possible decryptions are equally logically valid. What you have on disk that’s a series of magnetic fields that represents the ASCII encoding of “ATTACK AT DAWN” could just as well represent some other encoding of “DRINK OVALTINE” or any of a nearly boundless number of other possibilities.
And, even aside from questions of encoding…well, that series of magnetic domains on the disk are irrelevant unless they’re read out and communicated back to somebody. They’re really nothing more than, as I mentioned, time-delayed electromagnetic communications, most emphatically subject to all of the limitations that Shannon discovered with respect to signal and noise.
So, while it’s certainly a very useful abstraction to think of information as a Platonic ideal…it turns out that it’s not.
Hope that helps….
Cheers,
b&
Well, the “let’s think of information as a string of 1s and 0s on a hard disk” analogy has always left me cold. I just don’t think that this is a way to think about empirical questions. It also assumes a lot of stuff: discrete bits, sequential processing of data and so on (see my other comment to you in this thread).
But beyond that, I think I must have misled you: it’s precisely because I don’t believe that there is a platonic realm out there that I say what I say. We deal in information theory as though there were a platonic realm. The whole point that I’m making is that information **as we theorize it** to discuss things like “algorithms going back to the top” and such like, or even recursive functions, when it’s clearly not possible for the real world to go back to a previous state, even with new data, because the universe doesn’t make a difference between states and data, **that way** of talking about information assumes a platonic realm. And there isn’t one.
I’m not doubting that you have a much deeper personal understanding of what you say about time-delayed communication than what you propose below. But just changing the representational paradigm and imagining a physical way to shift 1s and 0s in time doesn’t begin to address this problem of the fact that the universe is not a finite-state machine.
On your second point about MEANING, indeed quite a thorny subject.
How do you address what seems to me the fundamental problem that results from what you outline: any given encrypted configuration can indeed have, in theory, practically unbounded legitimate meanings. Which means that there is no relation, at all as far as I can tell, between Shannon entropy/information, signal/noise laws, which are subject to all sort of constraints in terms of quantities, and meaningful contents, which are practically unconstrained. I can’t see where the causal power of the meaning of a message is constrained by any physical conservation law.
Strict causal determinism boils all the way back down to conservation laws. Without conservation laws, no determinism. Yet there is a realm which, by way of its representative faculties (meaning), can derive intents which will have direct causal effects through agency. These (intents) are representation objects that have no physical quantities **describing** them. There are therefore no conservation laws that constrain the nature of the derived intent on the basis of the Shannon information quantities that carried the input representation that gave rise to that intent.
In that case, if we can’t describe with conservable quantities the effect of the message, how can we justify strict causal determinism where representational systems are embedded in physical agents?
I would submit that “meaning” is essentially meaningless. There is no quantum of meaning; it’s entirely arbitrary.
There are statistical correlations that you could certainly use to bootstrap a model of what humans think of as meaning, but there most certainly is no absolute frame which dictates that the discovered correlations are the only valid ones.
This shouldn’t at all be disturbing. We’ve known for…what, a century, now?…that there are no privileged frames of reference in the physical world, either. And we see parallels everywhere else we look, as well; Godel established pretty emphatically that no system of even modest complexity can ever be complete, even when it can do things that some other system can’t (and vice-versa).
Cheers,
b&
Ben,
you’re again answering points that I’m not making. Which leads you to restate what I said in a different way, not answering it. I was asking you because you seem to know about these things.
If you say that “meaning is arbitrary”, “there is no quantum of meaning”, then you are merely bolstering my point. My claim is: meaning is not reducible to quanta with conservation laws; agent systems with devices to manage, transform and reproduce information (brains and genes) actually process meaning and exert their causal agency from meaning or representations of the world they live in.
What I’m challenging here is the integrity of the deterministic causal chain. I’m proposing a fairly strong reason why the integrity of the causal chain shows a gap between raw physical information about physical objects, which constrains the local behavior of particles through conservation laws, and the causes generated by representational-intentional systems like brains and more generally phenotypes.
To prove that there is no gap, you have to show how a change in the meaning of a message entailing different effects (decisions by an agent) is reducible to physical events **in a way that is subject to general laws**.
Being subject to general laws means that I need to be able to infer the actions resulting from the decision only from the change in the physical substrate, without recourse to representations/meaning.
and
The points I’ve been addressing have been attempts to answer your misconceptions about the latter.
Free will does not hide in quantum indeterminacy, period, full stop, no matter how you want to re-phrase and obfuscate the question. Quantum indeterminacy is demonstrably random. The brain simply isn’t a quantum-scale device, to begin with, though it flirts with the edges. At absolute most — and there’s absolutely no reason to suspect that this is the case; it’s just the theoretical limit — at absolute most, the brain could use it as a source of randomness. We can be overwhelmingly confident that there is no quantum computing going on in the brain — though that wouldn’t get us away from Turing-compatible computing. Most likely, anything quantum-related in the brain, if there is any, is either quantum efficiencies (on the same order as what chloroplasts use for photosynthesis) or there’s some sort of redundancy or other method used to minimize quantum interference (such as what modern IC engineers have to deal with).
That’s it.
Everything that could operate at the scale of the brain is known and accounted for, and none of it could result in something more-than-Turing. Everything hypothesized but not yet known is explicitly ruled out in exactly the same way as the Luminiferous Aether has been explicitly ruled out.
If you’re still not clear on this, you need to start by re-reading Sean Carroll’s essay. And when you’re done re-reading it, if you still have doubts, you need to remember that the discovery of the Higgs and more recent cosmological observations have only piled more nines on to the end of that conclusion.
Cheers,
b&
This is confused. The “Church-Turing thesis” just simply proposes, as a hypothesis, that the “intuitively effective” procedures for calculating functions coincide exactly with those “allowed” by the functions calculable by a Turing machine (or the recursive functions, or Post production systems, etc.). This does not exhaust all the functions, even from natural numbers to natural numbers. (The characteristic function of the halting set for TMs is one that is not so.)
So the question is, is that hypothesis correct? It appears so, since there are no known mechanisms (dualistic or otherwise) that contradict it. (See my MS thesis and my paper “Learning to Hypercompute?” in the volume _Computing Nature_.)
Could there be “noneffective” procedures? Well, yes, but those are also computational but they are limited in the sense that they are heuristics – they are fallible. Turing points out that if something is expeceted to be infallible it cannot also be intelligent. (There’s no limitation on being able to *some* cases of an uncomputable task class.)
I might add, incidentally, that “Church-Turing” is wrong historically (it is really more accurate to speak of Post-Kleene, since they defended it as a hypothesis, not as a mere definition (like Church did). Turing himself had no clear view that I can find.
Yes, exactly.
Another way of putting it that might help clarify things…C-T proposes that, if it can be computed by any means, then a Turing machine can compute it; and, as well, if a Turing machine can’t compute it, then neither can anything else.
It also probably helps to remember that a Turing machine in this context is better thought of as a mathematical equation than a physical machine. There is an equation that corresponds with every Turing Machine (and vice-versa); the Turing Machine is simply a thought experiment that helps one imagine what you would do to realize the equation in the physical world.
Cheers,
b&
Instead of going on about Turing machines, you could substitute “the fastest, largest, most powerful digital computer in existence” with no loss of generality to your argument.
Thank you for confirming that you don’t know what a Turing Machine actually is. At the very least, I can concentrate my efforts elsewhere.
Cheers,
b&
P.S. Do you also argue that, since there’re only about 60 orders of magnitude between the size of the Universe and Planck Length that more than about 60 digits of π are meaningless? b&
I had to construct them in grad school — abstractly, of course. The infinite tape sure comes in handy.
I should amend that. A Turing machine has vastly greater power that “the fastest, largist, most powerful computer in existence” by virture of its infinite tape.
I indeed still harbor some large gap of understanding on the question of computation, as I point out a couple of times in the thread.
I seem not to be alone in this predicament, however, if this thread is any indication. I think every person who engaged me had a different definition of what C-T claims.
I am aware of the existence of both my knowledge and my ignorance. I try to remain close to the questions, which are the only things that provide the proper kind of energy to enlighten.
I hope you’ll spare me the “but some things we know for certain !” lesson. I’m aware of that. Alas, the questions at hand are not in this category, as this thread makes amply clear.
With respect to your points above, while I may be confused in some respects, I was merely pointing out that C-T does not claim to cover all possible types of functions. I also point out (I think elsewhere) that in the mathematical functions that we use in the sciences, we constantly make use of approximations in order to make them computable (as far as I understand: literally computable, not practically computable), such as ignoring variables because of the scale we’re interested in, or plugging in empirical data to renormalize certain functions, and so on. Conditions at limits, both finite and infinite, I also find problematic in terms of matching real processes. We use countless ways to approximate and ignore real factors in order to compute. I’m puzzled by the fact that the universe itself doesn’t get to ignore these factors, yet still manages to compute.
Such observations make me wonder whether computation is as well understood as we think, at least as a natural process. If we define computation as something like processing information, then we have to have a believable understanding of what information is, and I’m sorry, but Shannon isn’t it. I try to read as much authoritative stuff as I can, and I have yet to see this addressed in a way that is consistent with the claims that are made here about the inherent (and consistent across scales) computability of all things physical according to our current definitions of computability. I’ll certainly try and see if I can read the material that you refer to. It sounds germane.
Cheers.
By the way, as you say, “the field is the physical realization of logical systems”, absolutely, and is therefore strongly constrained by a realizability ethos, it seems to me.
To claim to prove that brains really, actually are computers in the real world, by requiring us to imagine a device that cannot exist to our knowledge, in a context where we are unable to describe how brains actually go about computing anything, is pushing that ethos a little far, isn’t it?
Not at all. This is just a variation on explaining Newtonian mechanics using spherical cows. While doing so might not give you the cheese recipe you were looking for, it absolutely identifies real-world limits that are very useful to know.
Of course, if you could offer an example of something that a brain can do that a computer, even in theory, cannot, that would be an excellent bit of empirical data that the theory is problematical….
Cheers,
b&
I believe that understanding what it means to actually compute anything is a required prerequisite to any discussion about what the brain can or cannot compute.
What I’m interested in is what the actual universe can compute. After all, that’s the constraint. An interesting conundrum, in terms seeing what the universe can do (not necessarily our brains, but it’s a start) that a computer can’t do, is to compute the actual classical outcome of the collapse of a quantum wave function, in an actual single collapse event (not on average – the universe doesn’t compute events on average).
Again, to be clear, I’m not claiming that this is what our brains do, I just observe that this is what the universe does, and computers can’t. Even in theory. Or at least, having a theory of whether computers can do that would require to have a theory of how the universe does that. Right now, we don’t even have a shred of a clue.
And this is not “mystery mongering”, it’s actually trying to make sense of the empirical clues that we have from the universe and mathematics. I think that this poses strong questions to our understanding of what we mean by “computable” in the real world.
If computable means taking a set of discrete objects (in the broadest sense of the term – given QFT, it’s hard to tell for sure what such fundamental discrete objects out there could be) with enumerable properties (real physics is all about ignoring the infinite amount of information that we consider negligible for the purposes of a given computation – but the universe doesn’t get to ignore that stuff) and determining a new set of enumerable properties for these objects, then we have to observe that computation in that sense is not how the universe actually works, in detail.
I’m not sold on the quantum brain, so that’s not what I’m saying here. But I believe that we are making very big assumptions about what it means to actually compute physical stuff. It seems highly probable to me that our understanding of the relationship between information, computation, mathematics, time (or emergence/thermodynamics/whatever) and physical events is full of holes. In particular, I find that our view of causality is very newtonian, and overextends the conservation laws to domains where it doesn’t apply.
Sorry, lot of loosely connected stuff, not much concise evidence. Working on it 🙂
First, considering that our physics is still incomplete, I hardly see how it’s the fault of computing that it can’t compute what physics doesn’t have answers to.
Next, computers most emphatically can model quantum indeterminacy, even if they can’t predict the specifics of an individual quantum event. And if the Many-Worlds Interpretation (which I personally have problems with, but I’ll admit is quite popular amongst physicists) is correct, then your complaint simply isn’t applicable; all possible outcomes are equally valid, and there’s no reason to privilege the outcome we observe as opposed to the other outcomes that also happen but that we don’t observe.
Last, again, the physics of the everyday world is completely understood, and there’s no room in there for quantum woo. It’s exceedingly unlikely our brains are doing any sort of quantum computation, though they may well make use of quantum efficiencies. Whatever quantum effects there are in our brains — if any — would be more than familiar to a modern IC engineer, and we know for a fact that there’s nothing spooky in there that gives computer chips free will. There absolutely certainly isn’t anything quantum in our brains that could even theoretically be applicable to free will, either.
Cheers,
b&
“The fault of computing”?
I thought we were having a conversation about something else…
If you agree that there are things the universe does that we don’t know yet, then how can you make definitive claims about what is or isn’t computable?
As for the rest, I’m sorry, but you are answering points that I specifically said I was not making:
1- Indeterminacy in actual events, not on average (which is what computers can model-probabilities, big deal)
2- I’m not making claims about the quantum brain.
Sorry, but if you prefer to only address the points that I’m explicitly not making, then you should consider what it is that makes you uncomfortable in those that I’m actually making.
Because we know everything relevant and necessary to answer the question.
Consider: we don’t have a perfect inch-scale map of the Earth. Does that mean that it’s therefore possible that a cheeky adventurer still someday might stumble across a large mesa in the middle of the jungle where dinosaurs still roam the land?
…no?
Then why on Earth should we need to know everything about the universe in order to know what types of things the brain does and doesn’t do?
There’s no quantum weirdness that affects the brain (though, again, quantum efficiencies aren’t out of the question). Dark matter and dark energy have even less of an influence on the brain than neutrinos, and none of those even offer the hypothetical possibility of influencing thought in a coherent way.
The brain is exactly what it appears to be: a very complex and powerful and messy meat computer. The complexity and power and messiness no more give it magical properties than an entry in the IOCCC could offer a general-purpose solution to the Halting Problem, even if it does twist your noodle.
b&
By the way, while I packed it in for the general arguments, I did intend to get back to your challenge: offering an example of something that a brain can do that a computer, even in theory, cannot. Here goes:
Being strangely both repulsed and excited while reading the rat scene in American Psycho, realizing as I keep reading that I’m twitching my leg in unease, while at the same time being envious of the nice suit wardrobe that Bateman possesses. After that, having difficulty having sex with my girlfriend for a while, while thinking a lot of things that I cannot even say here because, while they actually are in my brain, I just can’t bring myself to utter them.
Please go ahead and account for this as a set of Turing machine algorithms turning the physical substrate of a few hundred discrete shapes of black (letters) on white (the paper of the book) exposed to my retina for a few dozens of seconds one day **deterministically** condition my behavior in bed with my girlfriend weeks later. I’m saying that I want the causal chain from photons bouncing off the page to my retina, to these behaviors, **without accounting for meaning**, which, as you indicated, is an quantity-less, non-physical concept that can be ignored. Explain to me where the one-time-pad is and how it decodes the photons bouncing off my retina.
If you can do that, you will really have made a difference today.
Thanks.
I have repeatedly explained that we don’t (and may never have) have a complete understanding of the human brain, but that that fact no more allows for a ghost in the machine than our lack of a complete understanding of physics allows for perpetual motion machines.
But let me offer a brute force approach that should be convincing, even if it’s obviously impractical for humans to construct.
And that, simply, would be to build a complete physical model of you and your environment, to whatever resolution is necessary. Probably no more than individual neurons, but, if it’ll make you feel better, all the way down to the atomic level — or, Hell, even down to Planck scales.
The model or whoever creates it need not understand its own functioning any more than you understand your own functioning; it simply must accurately map the physics of your body and your immediate surroundings, in sufficiently fine detail.
If you deny that this would constitute a Turing-equivalent model of your consciousness, feelings and complexities and everything, then you reject physicalism and embrace supernaturalism.
And, again. The practicality of constructing such a model is irrelevant. Turing machines themselves cannot be physically constructed, as they require infinite tape; that’s part of the definition.
Hope that helps. If not…I doubt there’s anything further I could offer.
Cheers,
b&
Repeating what I’ve heard a thousand times is not going to help, sorry.
You still cling to your ghost in the machine straw man and simply won’t engage the question that I’m asking: using what I thought was a well-thought-out one-time-pad analogy, how do you provide a physical, conservation-law-bounded, description of the meaning of the words on the page and by what kind of process do these few hundred shapes on a page cause behaviors in me without using the meaning of the words in the account. Just the photons bouncing off the page to my retina.
I’m now convinced that is something that you can’t do, since you won’t do it, even schematically (it doesn’t require a deep description of the brain to do so, I can live with a description of the process with idealized substrates). It’s disappointing, but enlightening at the same time.
The only thing that you choose to do is to trot out the worn out “if I could copy you with 100% resolution, then I’d have a perfect copy of you”. Well, duh, dude. Big deal.
Showing that something is physical only proves that the brain is a Turing machine if you believe that a Turing machine is something that can simulate any physical process whatsoever. That’s tautological, since it’s precisely that assumption that we are discussing, on the Turing front.
Finally, if all you can prove is that the brain is a physical thing, that information is embedded in a physical substrate and that there’s no soul in there that is not embedded in its material configuration, you’re talking to the wrong guy: I know that already, and it’s a trivial piece of knowledge. I wonder why you consider that it’s such a profound insight.
What I’m addressing is the open causality, from the pure point of view of physical conservation laws (because I use the working hypothesis that deterministic causality reduces to conservation laws) of information.
Information is, even etymologically, form. It’s putting something in a form. There are no physical conservation laws about the form of things that I know of. There are constraints, a form does have conserved properties, both mathematical and physical, but no general deterministic laws completely describe and specify a form, unlike other physical quantities like momentum, spin and so on, which obey conservation laws of various kinds. You can radically change the form/shape of something with very minimal and uncorrelated changes in the physically determined properties of the substrate: proteins are a good example. Radically different forms can arise from differences in physically negligible quantities; in fact words are also a good example e.g. “matters” vs. “mattress”, “cat” vs. rat”. Even slight variations in the configuration of forms that just happen to be around a form can change the meaning of a given form – that’s context.
Yet the forms can cause effects in agents that store forms, relate forms to other forms (representation), intend other forms and use forms and laws to get information about future forms. Yet deterministic causality only works in one time-direction. So once you include real-time information about the future in your causal process, you can only resolve the actual outcomes at the present time. That is synonymous to “undetermined” in my book. “Hyper-local determinism” sounds good too.
By the way, what I say above would still hold if the “processor” were a Turing machine. I was wrong to insist about the Turingness of things, since it’s not central to my point. My views about Turingness are secondary. Sorry if that created confusion, I think it did. I’m really only interested in the determinism question here. I think that my indiscriminate mixing of two topics may have made it easier to misunderstand my point.
So if you intend to come back with yet another set of broad statements trivially proving that physical things are physical things as though I were arguing for the magical nature of the soul, I will really be done this time 🙂
Thanks for your time nonetheless.
Actually, that’s not tautological; that’s the heart of the entire matter.
First, as I keep pointing out, the laws underlying the physics of everyday life are completely understood. I’ve repeatedly linked to this article, but I very, very strongly suspect you haven’t read it yet, so let me make the link as plain and obvious as I possibly can and urge you in no uncertain terms to actually go and read it:
http://www.preposterousuniverse.com/blog/2010/09/23/the-laws-underlying-the-physics-of-everyday-life-are-completely-understood/
There. You’ve read it, no?
Now, unless you’ve got a problem with what Sean wrote, you should now understand that, yes, indeed, the laws underlying the physics of everyday life really are completely understood. And, as such, we have everything we need save sheer brute force computing power to create a computer model of things, at least at the scales we’re interested in. Imagine a video game (most of which already have very sophisticated physics emulation engines), but big and powerful enough to encompass an entire person.
That such can at least logically (if not practically in 2013 by humans) be done means that whatever happens in that simulation is computable in the Turing sense of the word.
The next step I’ve simply been assuming you understand, but you might not. The whole point of Turing’s work with Turing machines was to demonstrate the interchangeability of them. Construct two radically different computers; so long as both are Turing-complete (meaning, they can perform a certain minimum set of logical functions, given sufficient resources (memory, time, etc.), anything the one can do, the other can do and vice-versa. The one may well (and often is) far more efficient at the job than the other, but both are equally logically capable of performing the exact same computations.
So, the job at hand isn’t to figure out how the computation that goes on in your brain gives rise to consciousness and what-not; all we need to do is determine if it’s logically conceivable that it could be done at all.
And, since we can take the brute-force method and create a physical simulation of the brain, we know that, yes, indeed, it can be done.
Now, are there more efficient or easier-to-understand or more useful computer programs that could do the same thing as a human brain? Almost assuredly. But, again, that’s not the point; the point is that, yes, the brain is a Turing-equivalent machine, if for no other reason than we can make a model of a brain in a computer.
The only possible objection would be that there’s something we don’t understand about physics that can’t be represented by the laws we currently understand and which also can’t be represented by any computer simulation, no matter how sophisticated. The Church-Turing thesis says that’s not the case — but, more importantly, we know that’s not the case for human cognition, because the laws underlying the physics of everyday life are completely understood.
Which is why any claims to the contrary, whether born of ignorance or delusion, are nothing but woo.
I really hope that clears up any remaining misconceptions you might have, because I can’t imagine what possible gaps could be left.
Summary:
A) The laws underlying the physics of everyday life are completely understood. 2) That understanding means that it is theoretically possible to create computer models exactly mimicking any everyday phenomenon, including human cognition. iii) Because all everyday phenomenon are computable, C-T holds at least for that domain and we can therefore be confident that consciousness is entirely an emergent physical property of the mechanical brain.
Cheers,
b&
Like I said: if all you can do is to repeat that there are only physical things in the universe, you’re only saying things that are trivially true and that nobody contests.
That you consider that to be a profound insight says more about you than it does about what I’m saying, which you continue to ignore. None of what you say above has any bearing on the question that I’m asking.
If all you’ve got is a hammer, everything looks like a nail.
Bye for now.
What’s ironic is that you think brains can do things that we don’t know exist.
I wish I were in your brain right now in order to get a clue as to what you are imagining that I think and that doesn’t exist.
Sorry, I thought you were joking, and I thought i was joining in on the routine.
Should have used a smily i guess.
You aren’t making any sense whatsoever. I see a lot of obfuscating language with some random assertions emerging from the noise. You’re right, it’s a lot of loosely connected stuff. It’s incomprehensible, and your conclusions are overwhelmingly probably wrong in my eyes.
Right. Not any sense whatsoever. I had a room full of monkeys typing that comment, that’s why.
I’m not sure exactly what you consider my conclusions to be. Is it that there are holes in our understanding of how maths/computing/physics actually fit together as a real explanation of reality and real conservation laws?
If that’s the conclusion that you find “overwhelmingly probably wrong”, then you are one of these large numbers of people that we have at every epoch who actually thinks that real science is over because we already understand how it all works. Now it’s all about filling in the details. Good day Lord Kelvin 🙂
> Believing that human brains are like computers sure seems
> to involve a lot of imagining of things that don’t exists.
Each enters the world as a blank-slate;
then programmed… by kin, kith, Culture, State.
Each becomes… what each is taught to be.
Within social-constraints… mind is free.
imo
Indeed. The Church-Turing Thesis applies to sequential computability, whether the computation itself is performed sequentially or parallelly. Many types of concurrent computation does not enjoy a simple notion of computability. For example, many concurrent computations have neither beginning nor end, by design.
For a concrete example that upsets the intended conclusions (with the standard notion of computability), there’s race conditions where we get one result if one process arrives at some point first and another if another process arrives at some other point ahead of the first process.
Race conditions have nothing to do with Turing computability. They’re simply two different inputs.
A parallel computer, of course, has the practical possibility of being far more efficient at all sorts of real-world stuff than a simple, literal implementation of the exact mechanism Turing described — but, then again, that applies to all real-world computers save a few hobbyist novelties.
But even the most complex of parallel computers are still Turing-complete devices, and can be emulated by any other Turing-complete device with sufficient resources. Not necessarily as fast or as usefully, of course, but certainly logically.
Cheers,
b&
– Race conditions are not inputs to anything.
– Efficiency is not relevant to computability.
– Practical realizability is not relevant to computability.
As for the remaining parts of your last paragraph, may I suggest that you search for and read about, say, concurrent Church Turing?! The paragraph indicates that you have failed to appreciate what might constitute non-sequential computability.
Erm…I think you might be a bit confused as to what a Turing machine actually is — and it’s concurrency that’s entirely irrelevant. Just because you have two physical Turing-like machines set up with the one’s output wired up to the other’s input and vice-versa doesn’t mean that what you’ve got is actually a super-Turing machine. What you’ve really got is an obfuscated Turing machine, in which the physical mechanics of one or both machines constitute part of the (logical) input.
Remember that Turing machines are not physical constructs, but logical ones. The input is the sequence of bits fed into the machine in that particular order. In the simplest case of concurrency, the one input would be if the first physical Turing-like machine is slow. You could then reset the idealized Turing machine and re-run it with the input if the first Turing-like machine is fast. In other words, the timing of the operations of the two physical Turing-like machines simply constitute variations on the actual input that would actually be fed to a real, idealized, non-physical Turing machine.
Or, once again. You could offer an example of something that a concurrent Turing-like machine might be able to compute that a real Turing machine couldn’t….
Cheers,
b&
You continue talking about sequential computability.
You need to think about Milner instead of Turing.
I would be absolutely floored to discover that Robin Milner thinks that concurrent multiprocessing computers are not Turing-equivalent. A quick perusal of Wikipedia superficially confirms this; there is no mention of Turing at all on relevant pages, save for his receipt of the Turing Award.
Perhaps you could point me to one of his papers where he discusses how merely splitting up computation and IO into physically separate devices somehow mystically invalidates the bedrock math that underlies all computer science…?
Also, last I checked, all multiprocessing computers still use the Von Neumann Architecture, which most emphatically is Turing-equivalent. If you’ve got references to a multiprocessing computer that doesn’t use the Von Neumann Architecture, I would most gratefully appreciate a reference.
Cheers,
b&
Hello Ben
Have you found refs to the use of GPUs (eg from Nvidia, the Tesla) and DSPs (eg TI’s TMS 320) in recent multi-processor systems?
These are not the usual architecture, but of course they will also include ordinary CPUs for some of their control.
I have lost track of what is happening with dataflow machines. But again, quite a different architecture, at least the ones at Manchester Uni, UK.
Stream processors (such as GPUs) actually have an architecture that (at least, superficially) even more closely resembles that of an idealized Turing Machine than a Von Neumann device. That aside, they’re unquestionably Turing-equivalent. Your GPU can no more solve the generalized Halting Problem than the CPU.
Cheers,
b&
Let me try one last time.
The major problem is that you wish for a mathematical statement to say something about a non-mathematical concept. This is non-sensical.
I granted you that, perhaps, you meant some mathematical model of computation that was intended to capture the biology in question, and noted that this would entail comparing sequential and concurrent computation. This, I noted, is problematic because the notion of computability you are talking about is not applicable to all forms of computation within this scope, and I have given three specific examples:
1) Computation over open systems (race conditions).
2) Non-finite computation.
3) Interaction (Milner).
Neither of these can be reduced to a simple input-output relationship, as required for the standard, sequential notion of computability. Insisting that we only talk about this type of notion is begging the question at hand.
If you look to the work of Gurevich, Blass et al on Abstract State Machines, you will find lots of discussion of and results on step-by-step simulation as a proposed constructive abstraction over standard, or sequential, computability.
If you wish to read Milner, I’m fairly certain you will find pertinent discussion in his “Functions as processes”, but you need to be aware that it will not contain idle chit-chat of the the kind we engage in here. His main purpose with that article was simply to show that the interaction perspective is not less expressive than the algorithmic perspective, leaving open the question we are discussing.
If I remember correctly, there was also a paper at CONCUR 2012 you might be interested in called “Turing meets Milner”. I can’t fully recall the details just now, though.
If you search for “concurrent Church Turing”, I suspect you will find that most of what either you or I would want to say in a forum like this has already been said many times over.
As for 1), again, you’re confusing implementation architecture with logical architecture. You can quite easily construct an equivalent Turing machine for any multiprocessor architecture you like. The Turing machine, of course, will have but the one single processor, as well as one single input. And it will be logically identical to the fancy multiprocessor computer. At this point, it becomes obvious that what you’re getting confused about with concurrency questions simply represents multiple different inputs to the Turing machine. In one, the input is the output if the left-side CPU finishes first; the other, if the right-side CPU finishes first.
Non-finite computation is a non-starter, unless you’re going to go wooish on us and suggest that minds are infinite.
And, once you understand 1), your interaction concerns should be instantly recognized as a variation on the same theme.
Think about it for a moment: if simply breaking processing up into two physical CPUs meant that the computer in question was somehow more than a Turing machine, that would be a trivial solution to the Halting Problem, and the Incompleteness Theorem could be disproven by linking together two systems. Shirley, you can’t think that neither Turing nor Godel nor anybody since them hasn’t thought of such a solution and quickly realized the futility thereof?
Cheers,
b&
Your hypothetical Turing machine that emulates brain states would require non-finite computation. As I’m sure you’re aware from chaos theory, many dynamical systems (and presumably the brain) are sensitive to initial conditions. To emulate the brain down to the finest grain, and to have that emulation faithfully reflect the temporal evolution of an actual, physical brain being emulated, would require all independent variables to be represented to infinite precision.
Your notions of an infinitely divisible universe are now well over a century out of date. Let me introduce you to my friend. Granted, he’s been dead a few score years, but he still laughs at your Zeno-inspired naivet.
Cheers,
b&
Your friend Max Planck was quite a woo meister. Look it up. I’m afraid he can’t get you out of this predicament.
The fundamental equation of quantum mechanics — the Schrödinger equation — is a partial differential equation with real variables. Any exact emulation of a physical system that uses it would, in the general case, require infinite precision.
Let me get this straight, but only because you seem so determined to dig this hole and pull the dirt in over your head.
Is it or is it not your position that quantum mechanics permits infinite divisibility of matter and / or energy; and therefore human brains are capable of infinite computation; and therein lies the source of our free will?
b&
I’m not taking a position on free will, for or against. I’m saying that your Turing machine thought experiment is incoherent and (since you used the word first) naive.
Se, we’ll leave free will out of it.
Is your position therefore that quantum mechanics permits infinite divisibility of matter and / or energy, and therefore human brains are capable of computation of infinite granularity and / or scope, unlike a Turing-equivalent machine?
b&
No, that’s not my position. I’m addressing the your notion of computability, which I find wanting.
I’ve tried. I really have. I keep pressing you for details, for specifics, and all you’ve offered has been Chopra-esque handwaving and whoppers of statements on the same scale as those made by evolution deniers.
All I can take away from this is that you have some reason why you think that human brains aren’t Turing-equivalent devices, and that you will not, under any circumstances, offer any clear explanation as to what makes you think anybody else should take such an astonishingly earth-shattering proposition seriously. Along the way you’ve hinted that brains are capable of infinite calculation, that Planck limits don’t apply, and that Turing-equivalence means that you must be able to actually build a physical Turing machine complete with infinite tape.
You’ve done nothing to peddle woo, in other words. Unless you stop peddling woo and start offering clear and specific arguments based not in Chopra’s fantasies but the actual real world, I see no reason I should continue in this discussion.
If you’d like the last word, it’s all yours.
Cheers,
b&
You assume that everything that is physical is computable. That’s a leap of faith, and one that is questionable to say the least. I’ve told you how exact emulation of a physical system, using the best quantum physics we know, would require infinite resources, but that’s not good enough for you. Instead, you resort to ad hominem accusations of woo peddling.
I suspect you aren’t as insulting in person as you are when you can’t come up with a good rebuttal in an online conversation.
I have made no such assertion. I have repeatedly indicated that everything relevant (and to scales significantly beyond relevance) that happens in human brains is computable, and given ample (pointers to) evidence and logic to indicate how such a conclusion is as sound as one that the Sun will rise in the East tomorrow.
You, on the other hand — even in this post — continue to imply or even outright claim that, since we don’t know everything, anything is possible. And that is the very purest essence of woo there is.
The fact that this is all centered around questions about human thinking drives the point home even further; all your objections to the impossibility of constructing an adequate model of the brain because “Quantum!” apply equally well to any silicon-based computer. Unless, of course, it’s the divinely-infested soul that distinguishes the one from the other.
I promise, I really will leave it at that, even if you continue to misrepresent my position, unless you can offer something that isn’t woo to back up your position.
Cheers,
b&
Apparently I didn’t have the last word. Maybe this time.
So you admit at least the possibility (driven there against your will by reasoned argument) that there may be physical things that are not computable. I take that as a major concession.
“As for 1), again, you’re confusing implementation architecture with logical architecture.”
No, I’m not. You are. I have consistently been talking about models of concurrency. You have been talking multi-core architectures, seemingly unaware of what models of concurrency are. You are evidently also unaware of what open system means.
Then enlighten me. How does one have concurrency without at least simulated multiple processors?
And, further, how can a concurrent system be built that cannot be logically represented by a classical Turing machine?
And open systems are entirely irrelevant. That just means a longer tape as input for the Turing machine.
That you keep insisting that rearranging the logical organization of a finite-state machine, or that complex interactions between multiple finite-state machines and / or their environments is somehow problematical for mapping those machines into an equivalent Turing machine really very strongly suggests to me that you’ve got some sort of very fundamental misunderstanding of what a Turing machine actually is.
Maybe you could clarify things a bit. Is it your position that a concurrent system is capable of computation that a Turing machine is incapable of? A simple yes / no to that answer would go a very long way indeed to pinpointing just where this has all gone off the rails. And, if “yes,” an example of such a computation would be even more beneficial.
Cheers,
b&
You are making an alarming number of assertions about things you seemingly admit to not comprehending.
So, let me be quasi-realistic: imagine the home of a tinkerer and inventor in the greatest of Hollywood traditions, meaning among other things that there’s a large but exhaustible supply of parts. Our friend designs and builds a stationary robot that has the ability to assemble a range of different helper robots, including following its own design. The other types of robots are all autonomous and mobile. The robots are aware of and respond in certain ways to the pets that are also in the house, and come with intruder- and other emergency-detection routines. The transformer outside the house is glitchy and may periodically emit enough electro-magnetic interference to interrupt the shared communication protocol between the robots and may also occasionally fail to supply power for varying lengths of time, which prevents the self-guided recharging of the robots, among other things. The robots have a power-save mode they may enter into that they can sustain for a very long time and that the other robots may wake them up from. All components of the robots may fail. There’s a small number of physical communication lines to the outside for reporting the emergencies, sending updates, etc.
Will the pets and plants be alive when our friend returns from a 3-months holiday, barring natural catastrophes? Will all parts of the floor have been washed within a day prior the return and at no more than a week apart, barring power outages of more than 3 days? What happened to the burglar that showed up halfway through week 2, and proceeded to fall and brake both legs? Will any of the helper robots have bumped into each other? The clothes-drying line requires exactly three robots of given types to operate: will any of the laundry our friend left have been exposed to more than a light drizzle without having been rewashed? Have the robots done all this work in an environmentally-conscious fashion?
I’m sorry, but I’m utterly at a loss as to what this has to do with computability. If your point is that it’s easy to construct levels of complexity that humans would be hard-pressed to simulate with a Turing-equvalent device, that’s neither surprising nor even remotely relevant.
Cheers,
b&
I would like to distance myself from my previous posting just above this one.
I need to pull out of this exchange, and will not be be able to respond any further. I should not have made that posting in these circumstances. Sorry.
Ben
Ah was just responding to your request for multi-processor systems that are not the usual mixed data/instruction architrcture.
DSPs are usually Harvard architecture.
Dataflow are event driven & very parallel.
I made no claims re: Turing etc.
Just that there are many examples of non von Neumann multi-processor systems, eg the SKA analysis down here is using novel architectures based on GPUs.
Yes, you’re right — I haven’t been paying that much to the architectures of components and embedded systems and the like.
Of course, as you note, all the ones ever made are no more than Turing-equivalent. (Making something less than Turing-complete, of course, is trivial and uninteresting.)
b&
There’s only been a good model of parallel computing since 1980, with Robin Gandy’s paper on the subject. This work is finally cleaned up (as of 2003 or so); my MS supervisor Wilfried Sieg worked on this. There’s nothing hypercomputational there. In fact, Sieg is more skeptical about hypercomputation than I am!
Ben Goren: “I’m sorry, but I’m utterly at a loss as to what this has to do with computability. If your point is that it’s easy to construct levels of complexity that humans would be hard-pressed to simulate with a Turing-equvalent device, that’s neither surprising nor even remotely relevant.”
You had asked to be enlightened on models of concurrency and concurrent computation, and on open-system computation. My intention with the outlined scenario was to provide some context that we would allow us to discuss that. I was going to develop several of the underlying principles in subsequent postings before returning to the question of computability and why it is problematic outside of sequential, closed-system computation.
Apologies for the distraction it turned out to be.
Incidentally, I set the scenario up to be at the level of what could be simulated for a decent or above Master’s thesis.
Perhaps we could cut to the chase before circling back.
Are you suggesting that concurrent systems are capable of hypercomputation or that otherwise cannot even logically be modeled by a Turing Machine? Or is your point simply that concurrent systems create levels of complexity that make it impractical (but not impossible) to model with a Turing Machine?
The latter I’ll agree with in an heartbeat. Hell, I’ve written code, myself, that I’d have a hard time unravelling; no need to add all the nightmares of concurrency on top of it. And especially the types of concurrency that, say, older versions of VB supported!
But, at the end of the day, the fact remains that, if a concurrent system can compute it, so can a Turing Machine (no matter how inefficient or ugly the Turing Machine representation might be); and if a Turing Machine can’t compute it, neither can a concurrent system.
If you’ve got examples (especially proven ones) to the contrary…well, perhaps you could start by pointing me to the recipient of the Turing Award from the ACM who made such a remarkable discovery? (And, presumably, said person also has a Fields Medal, as well.)
Cheers,
b&
I do not have a problem with computability being computability. I have a problem with computation being equated with computability.
I’m sorry, but I just can’t make sense of that. Computation is the action of computing things. Only that which is computable can be computed. Anything which is computed is obviously computable.
Assuming that wasn’t just a thinko on your part where you meant to use some other word, I simply can’t imagine how that sentence could even theoretically be meaningfully parsed in the English language.
Maybe you could offer your definitions of the two words?
Cheers,
b&
I have not read all of the responses to this and the previous post on “free” will but there another question which may have been asked and answered. That question is what role does the concept of emergence play in the possible evolution of contingently “free” will. To draw on an off used parellel can one predict all the properties of water from the individual properties of oxygen and hydrogen atoms? When you go to higher levels of complexity do you encounter unpredictable properties such that at the complex level of the human brain could you make the leap from stimulus and response as such that it would be considered mostly or at least partially free of the constraints entailed by its component parts, mechanisms and even memories both imparted by genes and/or memes?
“When you go to higher levels of complexity…could you make the leap from stimulus and response…such that it would be considered mostly or at least partially free of the constraints”
You are talking about freedom here rather than freewill, but how is this freedom from deterministic cause and effect achieved? In other words, what does this freedom consists of? Is it simply a flip of the coin or roll of the dice? I guess that IS a sort of freedom from cause and effect. But it’s not freewill. The “will” part still has to be accounted for. What does “will” consist of? Is it the “will” that achieves freedom from deterministic cause and effect? If so, what does this “will” consist of?
I thought the Church-Turing thesis was: a mathematical function is computable mechanically by an intuitive algorithm if and only if that function is recursive. Do you think that the essence of that thesis is that what goes on in living brains is like what goes on in other mechanical devices?
I’ve never before heard even a hint of recursion being part of C-T. Rather, the simple version of C-T is that anything which is computable can be computed by a Turing-equivalent machine, with the corollary that any device which can emulate a Turing machine can, within its physical limits, emulate any other such device.
Note that said emulation need not be efficient or fast, and that a machine with more memory (etc.) cannot be completely emulated by one with less, and so on.
It might also help to keep in mind that a “true” Turing machine is a theoretical device with, amongst other things, an infinite supply of memory.
Cheers,
b&
In my statement of the thesis, I actually deliberated about whether to write “recursive” or “Turing computable”. As you probably know, the set of Turing computable functions is provably equivalent to the set of recursive functions: they are the same set.
Thanks for that clarification. As you might be able to figure out from some later posts of mine, I was assuming that the suggestion was algorithms that needn’t be recursive were somehow excluded from C-T.
b&
The term you are looking for is “recursively enumerable” I think.
As I understand it a recursive set is a set where you can write a program that generates all members of the set and you can also write a program that can tell you if an element is in the set. For example you could write a program to list all even numbers and also write a program to decide if a given number is even.
A recursively enumerable set is a set that you can write a program that will generate all members but you may not be able to write a program that will always tell you if an element is in the set. The halting problem is recursively enumerable but not recursive. You can in principle set up a program that will eventually list all programs that will halt. But you cannot write a program that will take a given program and always tell you if it will halt. It will essentially be searching through an infinite list for something that isn’t there an so will never halt.
C-T tells us that all universal Turing machines will agree on which sets are enumerable and which are recursively enumerable. Anything that one UTM can do another UTM can also do. Anything one UTM cannot do the other cannot do either.
*
Curses! I wrote something, got bogged down, and deleted it.
It is worth the read to follow the “fatalism” link — to Richard Carrier’s blog — J. Quinton provides in his response to himself @ #18, both for the opportunity to read the language used in Morrissette vs US [1952] (establishes the legal precedent for free will jurisprudence), and for Carrier’s subsequent remarks. The comments following are interesting, also. And brief.
My inference from Carrier’s opinion is that any resolution to “free will” determined by either philosopher’s or the scientific community (assuming the two things are even possible) will have to be quite something indeed to accomplish more than tweaks to existing legal precedent to any significant extent, or to significantly alter common public apprehension of how people make choices.
Disclaimer: Of course a different set of circumstances obtains in instances specific to religious — or other — applications. Anything from any source that contributes to the diminishment or disappearance of dualism is welcome.
As always late to these threads, due to time zones.
So, if humans have free will, the only possible conclusions are that all other computational devices, whether biological or mechanical, also have free will — a proposition which the overwhelming majority of proponents of “free will” will outright reject.
I do not know where Ben Goren gets the idea expressed in the last sentence. At least I have always argued that the ability to make choices that some of would call “free will” comes in grades. Yes, an autopilot has more than a brick but still so little and for so narrow an area compared with a sane human being that it hardly registers. No, one does not have to be able to draw a clear line for something to be a useful concept. There is usually no clear line between “one species” and “two species” in a speciation process either. Still you would probably be counted among those who argue that species is a useful concept, right?
Have you considered a post addressing the comment that inspired Ben Goren’s? That might be more interesting. Because just like Coel I cannot believe that incompatibilists make no distinction between a rock falling onto a person and a rock being thrown at the same person by another person, or between a kleptomaniac stealing something, a sane person being blackmailed into stealing something and a sane person voluntarily stealing something.
And if incompatibilists do make such distinctions, then they are merely compatibilists in practice who have taken a rather inexplicable dislike at the use of a perfectly harmless word. Semantics in the most pejorative sense because there would be no conceptual or practical difference.
Of course we distinguish between a falling rock and a rock that’s been thrown. As far as “compatibilists in practice…” Just because you use Newtonian physics in your everyday life, must you disavow quantum mechanics? Must you declare Newtonian physics the true and accurate picture of the world? Or could you use Newtonian principles where appropriate while still acknowledging that quantum mechanics is real? Could you allow that what works for you in your day to day life might not be the most accurate description of reality?
That comparison does not work because compatibilists and incompatibilists agree on what happens in reality. The questions are whether using the words “free” and “choice” has supernatural implications, and whether saying that we do not have more freedom of choice than a brick is a helpful interpretation of what is going on.
Free and choice have no supernatural implications, but free will does. And of course we can do more things than a brick, but “more than a brick” =\= free will.
Free and choice have no supernatural implications, but free will does.
It does not for me. Nor am I as sure that it does for the majority of humans as some others here appear to be.
As I have pointed out before, in my native language, “freiwillig” (literally free-willedly) is the word for English “voluntary”.
Ok, I finally get it. I understand CFW. It’s straight up Orwellian double speak. All brain activity and all human decisions are entirely determined, entirely algorithmic, entirely constrained by physics, chemistry, and biology – this we call Freedom of the Will. Also, war is peace…black is white…and up is down. It’s frankly chilling.
Are you being serious?
Vaal
Absolutely. Everything you and the other CFW proponents have said on all these threads ultimately comes down to that.
That’s really quite incredible that you could draw such a conclusion. And I thought you were starting to catch on somewhat.
That said, your characterization is not an argument (and in fact bears no relationship to anything I have argued) so what actual argument do you have against CFW?
Of necessity, we’ve broken down the term “free will” and first have discussed the use of the term “free” in a determined world.
Are you still rejecting any use of the term “free?” Or…what?
An example again:
“Linda” is held captive, chained in a neighbor’s basement. She has all sorts of desires, things she would do such as return to her family, if she were only freed of the chains.
“Susan” is visiting a neighbor’s house as well, but she is not held captive. She can go home to her family any time she chooses, and does so.
In normal everyday language, we would describe Linda as “not free” and in comparison Susan is “free.” And Linda would be described as being held in the basement “against her will” whereas Susan is not. Hence Susan would normally be described as “visiting or leaving the house of her own free will” while Linda, due to not having the same physical options to choose those actions despite wanting to, would be described as NOT being in her neighbor’s basement of her own free will.
Now, are you truly saying that this can only be “Orwellian double speak,” as if you really don’t know what is meant by those descriptions, or that they do not describe any particular real differences in the situation of each woman?
Seriously?
What ARE you saying?
Vaal
What I’m saying is that the use of free will in all your examples is as a figure of speech. One may say, “I was higher than a kite.” This does not mean the person’s body was physically higher up in the atmosphere than a kite. When you say, “I came of my own free will,” we all understand that you mean you were not coerced by another person, but not that you actually have free will. We can still use the word free while acknowledging that there is no free will. You’re playing word games, insisting something like “the ability to move in space is free will” or “an airplane autopilot has free will.” These are not typical usages of free will.
“When you say, “I came of my own free will,” we all understand that you mean you were not coerced by another person, but not that you actually have free will.”
But that’s absurd since what is MEANT by ‘free will’ in that instance IS exactly what you understood: that I was not coerced by another person!
And it’s how people often use that term and it’s what they MEAN when using that term “came here of my own free will = not having been coerced.
So when you say “but that’s not actually free will” you are just begging the question, against what is commonly MEANT by that term in such a case.
Are you saying words aren’t defined by how people commonly use the terms?
Vaal
… so what actual argument do you have against CFW?
I think he is saying that since you are just arguing semantics there can be no actual argument against CFW.
When most people ask if free will exists they are explicitly asking if there is any new physics or supernatural phenomena that frees us from determinism.
You answer “no” and then call that position free will. Some find that deeply annoying. It seems to them that you are just jerking them around with words. To them it would be far far better if you simply said “No Virginia there is no free will.”
If you speak of the existence of Santa you should make sure to make it clear if you are speaking metaphorically. Love and generosity may exist along with hope and devotion.
But. Santa. Does. Not. Exist. If your answer to the question about physics and the supernatural is “no” then ditto free will.
We cannot simply redefine free will anymore than we can redefine Santa.
“When most people ask if free will exists they are explicitly asking if there is any new physics or supernatural phenomena that frees us from determinism.”
No, that is not only how people think of free will. If it was, it wouldn’t explain the differences in how people apply the term “free willed choice.”
Look at how people who believe we have free will use it in describing situations:
Go back to my Linda-held-captive by a neighbor vs Susan visiting a neighbor “of her own free will” example. Linda would commonly be described as NOT being there of her own free will, while Susan who can leave as she please would be described as being there of her own free will. Linda does not have a free willed choice; Susan does.
Now, what is “free will,” commonly used here, actually being adduced to describe? Is it that Linda has suddenly become bereft of “new physics” and “supernatural powers” whereas Susan it is claimed operates on different physics and has supernatural power?
No. Of course not. What “free will” is being used to describe here is the REAL WORLD PHYSICAL DIFFERENCES in their situations; the fact that Linda has desires she can not fulfill, choices she is constrained from making, while she is held captive, whereas Susan’s situation is different: she can exercise the types of actions, fulfilling those desires, that Linda can not. Those are real world capacities and differences, and that is the difference between the two women that motivate people’s description of one person’s action being free willed and the other not being free willed.
That explains and predicts to what scenarios people will use the terms “freely chosen” or “of their own free will” or not, in a way that simply bandying terms like “new physics” or “the supernatural” does not.
The more you look at how people use the term “free” you will see it’s used to describe real world differences that exist whether determinism is the case or not.
Vaal
(And, as I’ve argued before, it’s a mistake to mix up some people’s explanation for where they got the ability to make free willed choices – e.g. a God giving it to them – with what “free willed choice” actually means in terms of real world scenarios we employ such terms to describe).
You are again arguing words. Mere semantics.
Both Linda and Susan experiences themselves as free agents and probably believe themselves free in an acausal sense. If they ask about free will they probably aren’t asking if they are being held hostage. They are asking an existential question about the nature of that sense of causality free choice.
The difference between no free will and compatibilism is mere semantics. Word usage. You may make an argument about what word usage is best or more useful but there cannot be an argument about which is correct. They are the same.
Compatibilism seems to be a very bad intuition pump in that it polarizes even people who basically agree.
“You are again arguing words. Mere semantics.”
But we first have to get clear on what we mean by “free will” before we go answering the question of “do we have free will?”
Right?
If you are going to dismiss nailing down what one means by “free will” as only being semantics in some superficial sense, then the incompatibilists who deny free will exists couldn’t make their case either, since, hey “it’s all semantics.”
But it’s not “merely” semantics in the sense of “well you could just call anything you want ‘free will.’
Rather, not only are we trying to be clear on what we mean by “free will,” we are arguing that it generally captures the “free will” that is of concern to most people.
Arguments and examples have been put forth for this, showing how the terms “free” are commonly, validly applied in ways that do not contradict determinism, and simply characterizing it as “semantics” is no argument.
Vaal
But we first have to get clear on what we mean by “free will” before we go answering the question of “do we have free will?”
Right?
I do and you do but clearly “we” don’t. It’s just that we will end up saying the same thing differently.
If you are going to dismiss nailing down what one means by “free will” as only being semantics in some superficial sense, then the incompatibilists who deny free will exists couldn’t make their case either, since, hey “it’s all semantics.”
Incompatabilists have nailed down free will just fine. They just did it differently than you. That does not make either definition wrong. It just means you are speaking different languages and so can continue to argue even when you agree.
Arguments and examples have been put forth for this, showing how the terms “free” are commonly, validly applied in ways that do not contradict determinism, and simply characterizing it as “semantics” is no argument.
It isn’t intended as an argument. There isn’t any argument to make since you have just chosen different ways to say the same thing. That is what a semantic argument means.
Your usage of “Free will” is nonstandard in the history of the free will debate but you are free to use it. It will just confuse and annoy some. Especially those who agree with you.
Vaal, free will as you’ve defined it is totally obvious and uninteresting. If free will is basically just the ability to go places, then of course it exists. When someone asks the question, Do humans have free will? They are not asking “Can a human move from point A to point B if they not chained to a wall?”
“Your usage of “Free will” is nonstandard in the history of the free will debate but you are free to use it. “
That’s simply wrong and has been shot down here before.
As Stanford Encyclopedia of Philosophy explains:
““Free Will” is a philosophical term of art for a particular sort of capacity of rational agents to choose a course of action from among various alternatives.”
That IS what I’ve been calling “free will” and it is what is standard held to be what “free will” is all about. But there have always been competing accounts for free will as the SEP goes on to point out:
“Which sort is the free will sort is what all the fuss is about. “
So, while we are broadly speaking about whether we can choose among alternative courses of action, no ONE account of free will is automatically privileged as “the account.’ As you seem to by trying to do here.
Compatibilist notions of free will have been around since at least the Greek Stoics, competing with incompatibilist and libertarian accounts pretty much since the concept has been debated. Compatibilism also has been featured in religious thought as well – the conclusion that although God knowledge means things are determined, and that we can not act against our determined nature, we nonetheless can locate authorship and responsibility for our actions in ourselves.
Further, far from compatibilism being “non-standard,” the recent, largest survey every of professional philosophers:
http://philpapers.org/surveys/
shows that a whopping majority, 59%, hold to compatibilism, vs 12.2% saying “no free will.”
Finally, I’m the one who has actually adduced real world examples of how people use the terms “free will” – “She signed the contract of her own free will” “She went back to his place of her own free will” etc is commonly understood to MEAN “without coercion.”
And I’ve brought different examples showing how it is the presence of physical restrictions and/or coercion that motivates people’s labeling of scenarios as being “free willed” or not.
So your claims are simply unsupported assertions (wrong ones), whereas I have actually given examples and argument in support of my claims.
Vaal
Yes, seriously, vaal, what are you contributing to this discussion but doublespeak?
You accept that the brain is deterministic but, out of that, somehow, comes freewill! But you can’t say what this freedom and willing consists of other than deterministic cause and effect.
What are we to conclude?
BillyJoe,
You keep dismissing my use of the term “free” by pointing out I’m using it within a deterministic system, as if that invalidates it’s use.
Just how consistent are you going to be about that, since everything is determined?
What stance are you actually taking about the use of the word “free” in a deterministic world?
Are you saying that all uses of the word, or concept of “free” or “freedom” are invalid, and incompatible with determinism? Are you purging that word from your vocabulary and when people say “the dog is free from his leash, the man is free from jail, the victim has been freed from her captor” etc…that these descriptions all can make no sense given determinism?
What in the world are you actually trying to argue? It’s really you who is caught in bizarre double-speak at this point, since your denials of my examples of “freedom” are making a hash of normal language.
Vaal
(And if you argue that any use of the term “free” is incompatible with determinism, it will show a deep misunderstanding on your part about the implications of the definitions and use of those terms…but be my guest.
On the other hand, if you are going to admit to valid uses of the term “free” given determinism…then you have no business dismissing my use of the term “free” because things are deterministic).
You keep missing the point Vaal. When the dog is free of the leash, it is free of the leash. When the girl is free of her captors, she is free of her captors. When you say you have free will, what are you free of? Most people expect free will to mean free of the determinism of the universe. We all say things like “it’s a free country,” or “you are free to come and go as you please,” etc. In each case, what you are free of is pretty clear. When you say we have free will, it is not clear what you mean we are free of. And what most assume you mean is wrong.
“When the dog is free of the leash, it is free of the leash. When the girl is free of her captors, she is free of her captors. “
Ok. So you agree that it is valid and justified to employ the term “free” even GIVEN determinism.
Therefore, the one objection you can’t bring, which you guys keep bringing, is that “but we reject your use of the word free on the grounds that you are talking about a deterministic system.”
You can’t have that double standard.
With that out of the way:
“Most people expect free will to mean free of the determinism of the universe.”
There are two issues that I’ve tried to keep separate, that you keep confusing.
1. Whether “free will” as I’m employing the term “free” is valid and coherent.
And then, if it’s granted that what I mean by “free will” is coherent and not woo-woo within determinism, whether:
2. What I mean by free will is in step with how that term and concept is used and understood more commonly.
I’ve argued for why it’s “yes” for both 1 and 2, but have been trying to stick to #1
for now, since you guys keep rejecting the coherence of “free” on the grounds of determinism. And doing so very inconsistently.
So…as per #1. I’m using “free” in “free will” in the same sense in which a a girl is “free” of her captor: that if she is “free” she can exercise CERTAIN desires and actions she was CONSTRAINED from doing while being captive. These are real world differences in someone’s capabilities, given “being captive” vs “not being captive.”
Right? Yes? No?
Once you answer, hopefully we can move on to #2 again.
Vaal.
#2, it’s not clear which usage is the most common in every day language. When my friends say they did something of their own free will, I don’t know whether they think their free will is contra-causal or deterministic.
I think it’s prudent to be more clear with our language and just avoid the dualistic language altogether. What’s wrong with saying they did something of their own volition (as defined in psychology)?
Jerry had a good point: Do robots and thermostats have free will, too?
Vaal, if a lawyer had a promotional offer of no-charge estate planning, you could very legitimately refer to that as a “free will.”
The question is not whether or not there are contexts in which any of the words have well-understood meanings, either separately or collectively.
The question is what the meaning is in the specific context of the topic at hand.
Whether an individual is chained to a wall has no bearing on what control that individual has over thought processes — or even, for that matter, whether it makes sense at all to be able to control thought.
That’s the context in which this discussion is taking place. Whether or not I give you a present of a puppy named after the 42nd President is irrelevant.
Cheers,
b&
Chance,
But you can look at what situations your friends would described as “free willed choice” to understand what motivates them to use such a description. And you can prod your friends use of the term “of my own free will” with more questions that can uncover assumptions, inferences they are making that even they may not be aware of.
In other words, free will is like morality.
When someone is a theist he may think that morality “exists” but that it has a magical explanation, and without that explanation it doesn’t exist.
But what do so many people who deconvert realize? They realize “Hey, that thing called morality, my reasons for doing good, I thought had a magic explanation for it’s existence? It turns out I still have it – it wasn’t magical after all.”
It doesn’t mean the things they thought of as “morality” didn’t exist; rather, they realize it does, they just had the wrong reasons for how or why it exists.
Same with free will. Free Will concerns issues like “Do I have a real choice?” “Could I have done otherwise?” “Am I the author/cause of my choice?” “am I responsible for my choices?” etc. There are secular people who say “yes” and religious people who say “yes,” but their EXPLANATIONS for why we have free will may differ (e.g. “God gave it to us” vs “it’s a natural emergent property of more complex brains that can model possible courses of actions to meet our desires, etc).
We can prod people’s notions of free will to uncover the assumptions that actually anchor it in the real world.
I’ve already given examples of a common use of “free will” to describe physical differences in “captive” vs “not being captive.”
Say your friend agreed he was sitting in front of you, chatting of his “own free will.” He may even be a dualist. But ask him: Ok, if, when you desired to leave I chained you to the chair…would you still call that being here of your own free will?”
He’s going to answer “no,” isn’t he? Then ask, well what is the actual difference that occurred between those two scenarios. What can you point to that makes “sitting of my own free will” vs “NOT sitting of my own free will?”
It’s obviously: the scenario of being physically chained, stopped from doing something he wants to do. Not “going from being dualistic to non-dualistic” or “going from a-causal to caused.”
And you can keep prodding this intuition further even with a God-head. When someone says “God gave me free will” start asking about scenarios in which God would have removed, or impeded his free will. Note how pretty much any example will have to do with physical constraints being introduced. In fact, that’s one of the most common retorts from theists for why God doesn’t intervene to stop people murdering and raping “It would interfere with our FREE WILL.” They actually DO understand, intuitively, that the “free” in “free will” derives from lack of constraints on their ability to do as they wish…just what I’m talking about in compatibilism.
Again, it’s realising they were doing a bad job of thinking about their assumptions and intuitions that allows theists to say after de-converting “I still have what I was concerned about losing, it’s just now I realize the basis isn’t magical.”
Cheers,
Vaal
OMG, vaal, you’ve just been equivocating all along and I just haven’t realised it! To be honest, I was looking for somethng deeper.
Here is a reference to the equivocation fallacy: http://en.wikipedia.org/wiki/Equivocation
Being not free to leave because you are held in chains and being free to leave because those chains have been removed has absolutely nothing to do with the freedom we’re talking about in relation to how the brain operates.
But I suppose we could make an analogy:
The brain is chained to deterministic cause and effect. But, unlike physical chains, the chains in the brain cannot be broken. The brain cannot be freed from its deterministic chains. There is no freedom in the brain. There is no freewill.
BillyJoe,
No, you’ve been equivocating.
There are generally two objections that you and others keep bringing to the compatibilist position;
1. It can’t be free if it’s determined “It’s deterministic, where’s the freedom in that?”
2. THAT’s not what free will really is!
I’ve been dealing with both objections separately; you keep mashing them together in confusion.
To break the first objection we have to concentrate on whether “free” both makes sense given determinism and whether it is commonly used in a way that is in contradiction determinism. If this can be shown to be the case, then the objection “But
it makes no sense to call it free because it’s deterministic” is an INVALID critique.
So I show how we commonly use the word “free” in a way that does not violate determinism at all – that it’s commonly used to describe real physical differences in situations – not “free from all constraints or causes” but “free from SPECIFIC constraints and causes – e.g. the water is flowing freely after the drain was unplugged, the dog is free of the chain, the women are free of their captors.
You guys end up having to AGREE that these are common, valid usage of the term “free” even given determinism. Therefore, that breaks the back of the claim that “it’s not really free if we are talking about a deterministic system.”
But…as soon as we start applying “free” to human choices, in which JUST LIKE THE WORD IS COMMONLY USED I’m not saying “free of all causes” but only describing freedom from SPECIFIC causes, you suddenly hit the re-set button and start talking about the brain as a deterministic system and saying “But your brain is DETERMINISTIC, where’s the freedom in that?”
So you switch back to a critique that has ALREADY been shown to be an invalid inference (that determined system invalidates the concept of freedom).
So I’m showing how I am consistent in the application of the word “free” as applied to human choice, given determinism; whereas you keep special pleading only in the case of the physics of human choice-making.
Further, I keep showing by real world examples (something lacking in your arguments and examples you keep ignoring) that the “freedom from physical constraint or coercion” is actually consistent with how people employ the term “freely chosen” and “free willed” to describe real world scenarios.
So…the shoe’s on the other foot, actually.
Vaal
Reading now over the entire comment stream (had only skimmed it earlier) two observations on the incompatibilist position stand out:
1. Once more the issue of incompatibilists believing that if my brain cells make a decision following deterministic laws that it can therefore somehow (?) not me making that decision. Problem is, I am my brain cells. The argument is just like saying that a pocket calculator does not make a calculation because it is a chip inside it that does the calculation.
This is kind of a very odd inversion of the fallacy of composition; instead of fallaciously arguing that the whole must always have the properties of the parts, it fallaciously argues that the whole cannot possibly have the capabilities of the sum of its parts. Maybe it should get its own formal name one day.
2. Many incompatibilists here desire a black/white or yes/no answer and are extremely uncomfortable with the idea of free will coming in degrees, to the point of projecting their black/white thinking onto compatibilists and assuming that the latter cannot possibly believe that free will has shades of grey.
Again, why must somebody who finds it easy to understand that there is no clear line between red and purple although red and purple are useful concepts, or between different species-ness and same species-ness although species are a useful concept, insist on black and white thinking in one of the most complex matters we could deal with, the working of brains?
For your #1 – I don’t know where you got that idea. I don’t think anyone is saying that. What we are saying is that if your brain makes a decision following deterministic laws, it is not truely free.
For #2 – I think it is a combination of history and common usage. No one else defines free will as a continuum. You’re making it far more complicated than it needs to be.
Actually #2, is an idea best developed by D. Dennett, perhaps the most prominent compatibilist, and the best concept of free will presented to date in my opinion. Just google ‘degrees of freedom’ and ‘Dennett’. It makes all the sense in the world to view free will as a continuum. And as such, yes, even computers and plants have degrees of freedom. It’s not uniquely human, only the degree of choice-making is uniquely human for the time being, with computers catching up (and surpassing our degree of freedom?) real fast.
Seems you also do not accept that the self is an illusion, even though you’ve exactly described why it is an illusion.
The deterministic brain in the skull of the body labelled “Alex”, produces output such as the post printed above. That’s it. There is no actual Alex in there controlling the deterministic brain. That is the illusion produced by that deterministic brain.
Sounds odd to put it that way doesn’t it? That’s why we use the more compact dualist language. We just need to be careful not to be seduced by it.
As to the other question:
If freewill existed, there certainly could be a spectrum form virtually no freewill to full freewill. It’s a mute point, though, if the evidence is that it doesn’t exist.
However, the point was that, if computers can have freewill, and if non-human animals can have freewill, you are talking about something fundamentally different to the vast population out there who believe in freewill.
Do you also believe that the pacific ocean is an illusion because it is only lots of water molecules?
Say what?
Of course the ocean (lots of water molecules) is not an illusion.
And neither is the brain (lots of neurones)an illusion.
Come on now! We can’t compare a brain to a computational device. A device can be programed to make a certain kind of choice or even a series of choices depending upon the situation. But a brain has much more variability, versatility, and creativity. A brain can reprogram itself over and over again for reason or not. The degree of ability to reprogram is enough to make similarity comparisons rediculous.
No actually we can compare brains to computers. Computers can be programed to reprogram themselves.
When brains reprogram themselves they do so for reasons. That is there must be rules governing how they reprogram themselves. Computers are very good at following rules even if following those rule involve changing the rules they follow.
http://en.wikipedia.org/wiki/Self-modifying_code
Cheers,
b&
Yes, it pays to read up on your subject before pontificating about it. In other words, it pays for a gosling to become a goose before trying to quack.
Wait…that doesn’t quite work!
Omg the popcorn picture got a big laugh from me.
I agree with Ben & Jerry on this one. Mmm… Ben & Jerry’s…anyone?
Yeah, the compatibilists are just being silly with semantics, and the religious are being completely nonsensical as usual. “Free will” is dead.
No one behaves as if he or she has no free will. Therefore we all live as though we have free will. How is this indistinguishable from having free will?
I think you got the argument wrong.
Here it is:
There is no freewill.
But there is the illusion-of-freewill.
This illusion-of-freewill is so good that we act AS IF we do have freewill.
It’s like the checker board illusion.
If you didn’t know it was an illusion, you would go through your whole life thinking that those squares were differently coloured. The illusion is that good.
Same with freewill.
It should seem obvious that there is no fundamental difference between what goes on in human brains from what goes on in an electron.
So, if humans have free will, the only possible conclusions are that all other particles and fields, also have free will — a proposition which the overwhelming majority of proponents of “free will” will reject. Another logically possible path to take is go with a ghost in the machine.
Proponents of both hard determinism and compatabilist free will, I think, inveterately abhor and rely on semantics, respectively.
Your last sentence doesn’t scan.
I think the word “both” shouldn’t be there.
Sorry I meant to say that hard determinists see evidence for determinism and if determinism is true then free will must not be possible. There is no room for semantics or equivocation.
Compatibilists see evidence for, among other things, the appearance of free choices being made by humans. Compatibilitsts then try to expain these choices as being free will or, at least, the illusion of being so real that it is indistinguishable from free will. But in order to make that consistent with determinism requires more assumptions which appear like semantics in some cases. One consequence of this is that compatibilists have a lot more explaining to do.
I just read every comment (so far) on this post. I did not have a clear idea of what people mean by free will when I started but I think I am beginning to understand what compatibilists mean by it. So the compatibilist position is winning, with me.
This is a case where I don’t think the “this is all semantics” argument works. Anyone who wants to nitpick can find fault with any definition, because definitions use words to define words and are ultimately circular. That is, the definition of “free” uses words and if you looked up some of those words their definitions would contain the word “free”. So if people are going to communicate they have to accept some common terms without insisting on perfect verbal definitions of them, especially when examples are given.
I still have some confusion over whether self-awareness and the ability to reprogram oneself are considered necessary parts of having free will. I guess I would prefer a definition in which those were separate concepts – to remove my confusion. In that case, I think I could categorize just about any behavior as having or not having free will, in the compatibilist sense. However, I suspect compatibilists would like to include those qualities to conform to the tradition of privileging humans as the foremost practitioners of free will, in which case there will be grey areas. Maybe there is “weak free will” (for thermometers) and “strong free will” (for humans).
Well, Jimmy, now the hard part…
What do you think the compatibilist means by freewill that is not equivocation or a semantic difference from what everyone else means by freewill?
Self-awareness should be a requirement of free will. If a carbon atom has free will, it does not care. It is not conscious of the fact that it has free will, therefore one should argue that it does not make sense to suggest that the carbon atom has free will. Humans are self-aware and some of us choose to understand that we may or may not have free will. If free will is true or not, it has has meaning to us and therefore makes sense to attribute free will to things which have self-awareness. At present, nothing in the universe is self-aware except us, therefore the only things that free will should apply to are humans, regardless of whether free will is true or not to any part of the universe.
Disclaimer: This will be very long-winded, somewhat complicated, full of platitudes and at its heart incompatabilist. On the bright side, it is at least reasonably complete.
First, statements given without argument: materialism is true and dualism is false. The natural world is fundamentally causal. It might be entirely deterministic or with some probabilistic components. Complicated systems are in practice difficult or maybe even in principle impossible to model, but are entirely made up of low-level, causal behaviors. Nothing new emerges in complicated systems, per se. Nonetheless, it will often be more productive to use higher level descriptions to understand such systems, because it isn’t feasible to model them at the atomic level. The principle that the pressure in a body of water is constant across all horizontal planes and only a function of depth is one such useful simplification.
Second, every sort of human mind, thermostat, clockwork and computer are included in the above – made up by tiny pieces which adhere to the regular law of physics – in principle reducible to atomic interactions, though better understood at a much higher level – Tom likes playing with trucks because he’s a boy, that sort of stuff. It is important to keep in mind that these sorts of simplifications are in some sense wrong, even when useful, in the same sense that Newtonian physics is wrong. Saying that the thermostat closes the circuit when the room temperature drops beneath 15 degrees Celsius is wrong, by leaving out all the lovely details about how the temperature isn’t constant across the room and how the inertia in the switch means the threshold temperature is different for turning on and turning off. There is always something lost when collapsing the details into higher level simplifications.
Third, leaving the dualist nonsense out of this, the entire free will discussion is fundamentally not about materialism and what actually happens in the natural world, but about one such higher level of description applied to minds and behaviors. This doesn’t mean it is entirely without substance – higher level descriptions can be both outright and provincially wrong just like they can be useful. Taking a contentious example, because those are the best for sidetracking a discussion, the high level claim that -gender is entirely a social construct, with no innate behavioral differences between the sexes- is outright wrong. Similarly, it is fair game to discuss whether or to what extent the free will model of language is apt or nonsensical.
Fourth, the idea of agency is one of these higher level descriptions and also an important part of human social interactions – such a big part, actually, that it tends to bleed into places where it does not at all belong, such as chalking bad weather up to displeased gods. Humans, and organisms in general, are wired to manipulate their environments to achieve reproductive success, and an important part of their environments is their fellow human beings. Physical violence, threats, exchanges of favors and so on and so forth are all parts of the arsenal for controlling other people’s behavior. This is where the concept of free will enters the picture in the first place. If someone steals food which belongs to someone else, what people think is the appropriate reaction depends strongly on the circumstances. If the person is at the point of starvation, it is tempting to say that he had no real choice. If he is a child too young to understand that it is theft, people will say that he did not know that it was the wrong thing to do. If the food was stolen from someone who won’t suffer noticeably from the loss, this is a migitating circumstance.
Fifth, the distinction between acting on one’s volition as opposed to under coercion, is scientifically nonsensical. Pointing back to (first) and (second) above, it is altogether obvious that no person ever has any more choice than does a thermostat. The contrafactual -could have done otherwise- is again a definite non starter without abandoning materialism. The idea of being -free of coercion- is just as meaningless. In the language of mathematics, the output of a function is determined by the combination of the input values and the function itself. Given a person’s response function, i.e. his mind, which he is in no sense free to determine himself, all external circumstances are a hundred percent coercive, evoking the response it does by necessity with no volition involved whatsover. There isn’t, materially, anything more coercive about giving money at gunpoint than drinking milk when thirsty.
Sixth, the human concept of agency has nothing to do with materialism. It is rather a part of the mental algorithms for finding the correct response to other people’s behavior, such as deciding between punishment and forgiveness. This algorithm is obsessed with something rather different from coercion. Primarily, it asks whether nearly everybody would have done the same thing in the same situation – such as giving away the bank’s money to an armed robber. If so, it gets a free pass and doesn’t warrant punishment. This system is constantly on the look-out for people who try to abuse the trust of the community to get away with more than their fair share and people whose violent tendencies put the people around them at risk, etcetera. It has very little to do with free will, and everything to do with what you can expect from people in the future given their behavioral history, and this sort of social control is of extreme importance to facilitate social behavior in groups of organisms with unaligned preferences.
Seventh, this segues nicely into why people outside of narrow circles of armchair philosophy take issue with the idea of free willed autopilots. Agency isn’t applied merely on the basis of a contrafactual – it could have not rained today, though it does rain, if some inputs to the climate were different or if the algorithm determining weather itself were different, doesn’t cut it. Agency requires that the entity in question, which means mostly other people though to some extent also animals such as dogs and cattle, are thought to comprehend reward and punishment and and make decisions in light of expected reactions. An airplane autopilot does not fit this bill at all, regardless of how incomprehensible the nitty-gritty of its function might be. It also explains why raving mad people mostly drop out of this moral system – though they do indeed make decisions, and would likely have acted differently if [insert contrafactual of choice here], their utter disregard for social control place them in much the same category as a tiger, morally, and they tend to wind up behind lock and key in much the same way.
In conclusion: dualist free will is false and compatabilist free will maps very poorly to the human concepts of moral, choice and responsibility. He could have done otherwise does not mean that he was physically unrestrained from opening the safe – he could have done otherwise means that a well behaved person would have done something else in this situation, and failing to do so relfects poorly on the person’s moral character. Actually, the point of -free will-, scare quotes intended, is not that people act free of coercion, but the exact opposite – that sane people’s behavior is very well constrained by their expectations of punishment and reward. A compatabilist conception of free will which includes thermostats and computer programs without explicit mimics of human social behavior into its fold has completely failed to understand what the human concept they want to preserve in the face of determinism is all about. It is my opinion, that using free will to refer to anything but the dualist nonsense should be discontinued. There is indeed a human concept of responsibility to be preserved in all of this, but compatabilism is a very false start.
Well-written indeed. Does not means that it is correct.
.. materialism is true and dualism is false. The natural world is fundamentally causal… totally agree.
–Nothing new emerges in complicated systems.. totally not agree.
Systems are not atomic, causal indeed, but not entirely classical newtonian that you surmised, not only all systems are probabilistic, most of them are near chaotic, which causal correlation may changed, all and each most likely deterministically though.
A lot of people who called themselves incompatibilist (note: I do not agree with this classification, I am not a compatibilist) think very simply, in the beginning is jesus … er causal, then all are causal.
And these facts are not weird, the core of physics is similar. Atomic view is not the -real- reality. Deeper than atom it is probabilistic. And it seemed on the other end as well, the dark-force and dark-matter as the majority of reality (which we yet not know much).
Levels of understanding, the difference between physics and chemistry, biology, psychology, sociology as I mentioned before is important because of the way human brain understand things. In -real-reality of course no levels are necessary. But we are far away from that status at present.
We still have no real causal determinism between a lot of totally deterministic systems (like weather I mentioned before), and most probable even never will be.
So, saying (like Coyne did), that there is no dualist religion-based free-will is OK, but saying that people has no free-will and this should have impact on the law of the land (!) on criminals and penal system, is not only totally WRONG, but also socially-irresponsible as in teaching creationism as science!!
I will not say to let all legal matter to lawyers, but changing penal laws on the basis that people are not responsible to their action based on a jump from atoms are determinism so Alfs and Toms are determined too is clearly too much of not-so-good-things! (as bad as jesus is/was .. therefore .. )
And lastly, I think this enjoyment in saying compatibilistic incompatibilisms is becoming too much Sophisticated-Tongue-Twisters (TM) for me.
Long time ago (1970s!), I read a sci-fi novel that the title I no longer remember (seemed like a russian novel, english translation of course).
A computer with many sensory inputs of its surrounding, within a big lab warehouse (it is intended as a defensive security device), monitors temperatures, movements, pictures, smells and all. It is given motoric capabilities as well, can close all doors, windows, operate pumps, generators, even own on-off switches and all, and having a movable robotic units under its control (some of them like dr. ock’s arms).
So the computer acts like a big alarm system, sensing, reporting, asking decisions, and operate on the instructions by the human handlers.
Everything is fine, until one day a somewhat lazy programmer thinking that it would be better if the computer can do everything automatically, rather than waiting for human response. What kind of instruction should be safe to give?
He started just with one: curiousity, fill up your memory banks with new things, new information and new action-reaction chain of events (=determinism rules). Compact the info, store them, and learn more. Learn the world by yourself.
The moment this new subroutines put into executive mode, it created priorities and acted. First, it wanted to know more about cats (which a lot are moving around inside), what are they, what are their action-reaction to things. Most of information gathered are vague, very soon the computer knew that a lot needs to be learned on the insides of the cat as well.
After reasonably satisfied with the dissections of the cats (about 30 minutes after the dissection of the first cat), then we have those two-legged mammals which somewhat bigger but lots of similarities that of course need to be explored in detail …..
Suffice to say that this novel is categorized “gory”. There are lots of plot holes in the story, one at the end of the novel, the computer just started to notice the door (I remember thinking why cats can be more interesting than doors? mystery ..).
Of course it found that the environment outside the lab warehouse is much more interesting … it also found more and more designs of mobile platforms …
Oh by the way, before the programmer plugged in the new app, he connected the computer to the cloud .. limitless memory banks …
And the novel is from this man’s POV, so we know all the gory thingies third party, and that most likely the computer is saving his life for something in its future plans ..
(this, in addition to “the 9 billion names of god”, is my inspiration to study sciences).
Any suggestion for possible title? (yes, it has a bit anti-cat tone)
Well, let me start by linking to a discussion of psychological studies that attempt to determine what the term “free will” means in everyday discourse (I skimmed it and noted that it hit the high points in the studies, but don’t take it as an endorsement of the position):
http://www2.gsu.edu/~phlean/papers/Nahmias_Psychology_of_Free_Will_prepublication.pdf
Surprise, surprise … people are inconsistent [grin].
Next, I wanted to list these cases earlier, but it didn’t seem relevant. Now, with the continuing confusion over the definition of free will, it might be helpful. So, take these cases:
1) Someone steals a million dollars because they have a chip in their head that someone else remotely activates to take over their actions and run their body by remote control to get them to do that.
2) Someone steals a million dollars because they are a kleptomaniac and so feel an overwhelming urge to steal.
3) Someone steals a million dollars because someone threatens to kill their family if they don’t.
4) Someone steals a million dollars simply because they feel like it and want it.
Now, both Free Will Libertarians and Free Will Compatiblists argue that it is just obvious that these are, in fact, meaningfully different cases, and that it would be absurd to suggest that they aren’t. Even from the perspective of how to react to it, it is clear that the right action should be different to all of these. In the first case, take the chip out. In the second case, treat the kleptomania. In the third case, lock up those who made the threat, not the person who stole. In the last case, lock up the thief. So to deny that there are in fact ANY differences in these cases and that they are all the same is absurd … and that’s seen as being the consequence of Free Will Determinism.
FWLs agree with FWDs that determinism means that these 4 cases are, in fact, essentially the same case, but that since it is so absurd that these things be considered the same that determinism must be false. FWCs consider the idea that determinism is false to be equally absurd as the idea that all 4 cases are essentially the same, and so want to preserve determinism while still acknowledging that these 4 cases are indeed meaningfully different. That’s the heart of the various positions.
I’d also like to touch on an experimental test, suggested by Vaal. To test compatiblism as opposed to FWD, what you need to do is test whether or not decisions are made due to a set of decision-making processes, or just due to set neural activations that may or may not interact with decision-making processes. Thus, in the Cheerios/Corn Flakes example, what you’re testing for is if the decision was made, essentially, for reasons, and reasons that the agent had at the time as opposed to a post hoc justification. So, something like asking someone to list their reasons for making a choice before they do, and seeing if they then make that choice and if they’d give the same reasons afterwards. Of course, because we can change our minds and can act spontaneously, this is actually a difficult experiment to run, but that would be the heart of it: if we act for reasons, then reasons are relevant to our actions, and then we can evaluate actions based primarily on reasons, and that preserves what’s importantly different in the 4 cases above while still allowing determinism to be true.
Interesting linked article. Makes me think that no one is not plagued by the illusion of free will. Whether for or against, the properties of free will are real for everyone.
Coel completely on point as usual.
Here’s my two minute summary:
Incompatibilists see “free will” as a binary metaphysical issue where you either always have it (libertarian/theological) or never have it (hard determinist).
Compatibilists see free will not as a binary metaphysical issue, but a behavioural continuum. Sometimes we have it, sometimes we don’t.
The reason I think the compatibilist model is the correct one is that it is the model used in jurisprudence. When a judge asks “Did you sign that contract of your own free will?” she is not asking “Is determinism true?”. She is asking if – in this situation – you acted free of certain specific constraints.
It seems odd to me that we have a significant chunk of atheists insisting that the religious demagogues use the ‘correct’ incompatibilist model and our legal systems use the ‘wrong’ compatibilist model.
Whether freewill is binary or a continuum is irrelevant.
The materialists who are not compatibilist have concluded that there is no freewill. Therefore whether freewill is binary or a continuum cannot be the deciding factor as to who is correct between compatibilists and those who are not compatibilists.
You need to present evidence for freewill.
Good luck with that.
But, according to you, freewill is real because judges believe in freewill???
Excuse me if I don’t share your enthusiasm.
Really, that must be about the stupidest argument I’ve heard yet!
Billy, you still haven’t grasped that the term “free will” has more than one meaning, and that this meaning is what the discussion is actually about.
They’ve concluded that there is no dualistic-freewill. So have the compatibilists.
However — and this is the crucial bit — both compatibilists and incompatibilists agree on the *existence* of the *behaviour* that compatibilists *call* “free will”; it’s just that the incompatibilists prefer not to use that term.
This dispute is *entirely* about how to think about certain behaviours (such as a child choosing an ice-cream flavour) in a deterministic world, and what words to use.
Nope, we’re all agreed on what exists, what behaviours occur, and we’re agreed that it is deterministic. What we’re not agreed on is what language to use to describe and understand it.
This is the frustrating thing about these debates, the incompatibilists don’t attempt to understand what compatibilists are saying, and thus aren’t even understanding what the debate between the two is about.
The incompatibilists always think that it is a continuation of the debate between determinists and duaists. IT ISN’T!
I think at this point most of us agree that the dispute is entirely about how to think about certain behaviors and what words to use to describe them. We think your words are confusing and misleading. You feel the same way about our words. We’re each trying to convince the other side to use different words. That’s the whole argument at its core.
Well no, I don’t think the words you use for such things are confusing and misleading, since I have no idea what wording you want to use for everyday situations.
What words do you want to use instead of: “choice”, “decision”, “attempt”, “option”, “plan”, “threaten”, “test”, “compel”, “consider”, “coerce”, et cetera?
If your answer is “we’ll use those same words, just as compatibilists do” then you *are* compatiblists is all but the shouting, you *are* making a compatibilist interpretation of those words in a deterministic world.
If you don’t want to use those words then here is a starter for you, please re-write this in the language you want to use, avoiding the word “choose/choice”:
Dad to child: “Have an ice cream, you can choose the flavour”.
Child: “I want strawberry!”
Child, after licking the ice cream: “But I don’t like strawberry”.
Dad: “Well, you chose it!”.
As I mentioned to Vaal, context is key.
To use the same example: if a lawyer has a promotional offer of no-charge estate planning, that’s a free will. And we can all agree on the validity and utility of that use of language. But it has no bearing on the philosophical question of the term, free will.
In the philosophical domain — a domain, which it should surprise no regular here to learn, I find generally incoherent and useless — the question applies to the ultimate control an individual has at the most fundamental possible level. And, frankly, in any such context, the question can’t even get out of the starting gate, for it is a self-contained oxymoron. Only Minitrue could describe freedom as something which inevitably unfolds according to the laws of physics, and a will not subjugated to a coherent order is as useless as Bob Dole without his Viagra.
Now, once one drops the useless baggage of the philosophical term, there are all sorts of real-world situations in which both terms independently — “free” as well as “will” — can be quite useful. I am somewhat free to speak my mind, with the caveats of First Amendment Zones and NSA wiretapping and the like. It is through force of will that I crank out that last extra push-up in a morning exercise routine.
The problem only arises when a philosopher puts the two terms together in an inappropriate and meaningless context. Make no mistrake: there is nothing free about willpower, and true freedom cannot be constrained by the will.
Fortunately, nobody seems much interested in being predictably random, though, so the problem of free will is only one that keeps philosophers and theologians (but I repeat myself) up at night.
Cheers,
b&
So your suggestion is to ditch the specific phrase “free will” but to keep all other words the same and to interpret them in a compatiblist way? OK, I’m game for that.
“Compatibilist” is too much of a philosophically-loaded term for me to feel comfortable buying into it.
But, yes. Ditch “free will,” but don’t let that stop you from examining the interactions between physical determinism and decision-making processes, emergent behaviors and all.
Cheers,
b&
Ben,
There’s all sorts of question-begging, special pleading and inconsistency in your claims.
“To use the same example: if a lawyer has a promotional offer of no-charge estate planning, that’s a free will. And we can all agree on the validity and utility of that use of language. But it has no bearing on the philosophical question of the term, free will.“
Says who? First, the lawyer example is very much in the domain of the philosohpical free will debate, since that’s a type of free will
espoused by compatibilism.
Second, what is your basis for saying freedom in that sense has no bearing on the philosophy of Free Will? It has every bearing. Why? Because if you are going to reject a position on free will by saying “but that’s not FREE!” then we have to evaluate your argument on what you presume “free” to mean. Which means going back to looking at how that term is generally applied
in various ways to see whether it actually implies the contradiction with determinism you say it does.
And…it doesn’t. You acknowledge this when you accept all the example applications of “free” such as the lawyer example or others I’ve given.
So right after making a statement that common understanding of the word “free” can be compatible with determinism, you start waving away “free” will again based on…the implication it contradicts determinism!
“Only Minitrue could describe freedom as something which inevitably unfolds according to the laws of physics,”
So here you are just baldly stating freedom isn’t going to make sense if it’s in the context of unfolding laws of physics, but since common uses of “free” do not contradict determinism, one is left asking “So…where again do you get this basis for the claim that freedom can’t occur given physical determinism?”
“That’s not really free” isn’t an argument, nor is “that’s not free because it’s unfolding from deterministic physics” isn’t an argument either, since you (and we) already accept that common uses of the term “free” don’t contradict determinism. And if “free choice” only means not being restricted in ways that actually can occur in a deterministic world, then just like all the other applications of the word “free” it isn’t in contradiction with, or incoherent with, determinism/physics.
The only way around this is to hold to some special pleading version of “free” as applied to human choice-making, which is also begging the question.
Vaal
Vaal, if, when a lawyer draws up a will for you describing what to do with your estate after your die and doesn’t charge you for it, if that free(-of-charge estate) will is somehow related to your philosophical free will…
…then I officially give up. Whatever language you’re writing in, no matter the superficial semblance it might have to English, it’s entirely foreign to me.
Sorry, but you’re incomprehensibly incoherent to me, and I have no clue how to bridge such a divide.
b&
“But, according to you, freewill is real because judges believe in freewill???”
No, on three counts. Firstly you just aren’t grasping the main point I made here and that many others have tried to explain to you: that there are two different types of free will in the general discourse. So we are already going to have problems when you keep assuming that the free will I “believe in” is libertarian free will.
Second, it doesn’t seem like a great fit to talk about “believing in” something that is a conceptual distinction. It’s like saying “I believe in intelligence”. Intelligence, like compatibilist free will, is not a ‘thing’ that either can be disproved by philosophical argument, it’s a way of conceptually dividing up certain observed behavioural phenomena. Now, we could be in a situation where we don’t observe the phenomena and be able to say things like “Free will does not exist on Pluto” or “Intelligence does not exist on Pluto”. But the phenomena that compatibilists wish to differentiate DO exist and you have to admit it. A person will behave one way in a situation where there is no other agent, and in a different way when faced with an armed robber. So a lot of the back and forth here really boils down to the question of whether “free will” is a sensible term to use for distinguishing between these behavioural phenomena.
Which brings me back to my point: we already do use “free will” in this compatibilist sense! That is how our secular legal systems use it, it’s how we speak about “will” generally (e.g. “you dragged me to this party against my will”) and “free” generally (“I’ll be free to visit you after work”).
Of course, we recognise that the other way of talking about free will has a long theological history but who cares? Theologians get virtually everything wrong, but we don’t need to get eliminativist about every term they have some crackpot theory about. We aren’t tempted to say “Oh, well, theologians have historically claimed that the Earth is 6000 yrs old, so whatever this 4.5 billion yr old planet is it can’t be the Earth”.
All I can say is that anyone who wants to argue that there is no such thing as free had better strictly stick to philosophical argument or else to putting their own interpretation to the very limited understanding we presently have of brain function, rather than to venturing into Computational Theory to make their arguments – because if they do, Computational Theory will not bolster their argument at all, it will totally undermine it. The brain is indeed a PROGRAMMABLE STATE MACHINE (as are all Turing machines) but one of vast complexity – not just in the scale measured by the number of “gates’ (or neurons) that comprise it, but more so in its organisation – which is a parallel muti-unit hierarchical intercommunicating system with numerous internal feedback loops and an extensive associative memory capacity. With regard to its inputs (senses and perceptions) and its outputs (actions from “decisions”) if we ask the question “what MOST determines the outputs?” we must answer the processing of the state machine. If we then ask the question “is this processing totally deterministic(i.e. computable)?” the answer must be NO – the reason being that not all “programmes executed” are always computable in themselves and that algorithms executed are “self modifying code” (a process which in itself is not necessarily computable). On top of all this, one part of the system (which, in computing terms, we would call the “Executive” ) is “conscious” – in that it has a high level interpreted view of the processing and the system state. And this Executive “thinks it has free will” – well… it is essentially right – it is the author of its own programmed behaviours… which though part of an overall deterministic world are indeterminate, and only an outcome the processors processing itself.
It’s very simple.
If our choices are determined by prior conditions, then they are not freely made.
If our choices are the result of a roll of the quantum dice or some other source of randomness, then they are not freely made.
And if our choices are based on neither prior conditions nor randomness, then they are completely unpredictable (even by the person who made the choice) and would merit the explanation “I don’t know what came over me”. Therefore, such choices are also not freely made.
From where else could our choices come from?
Credit to Sam Harris, as this argument is paraphrased from memory directly from an endnote in The End of Faith (I think it’s the first endnote in the second-to-last chapter). Reading the endnotes from that book expanded my mind more than any other book I’ve read. I highly recommend doing so for those who haven’t already. Also of interest for this debate are The Moral Landscape and the short e-book, Free Will.
“If our choices are the result of a roll of the quantum dice or some other source of randomness, then they are not freely made.”
Well, that all depends. We shouldn’t at all be so negative about things that emerge from randomness – after all that is where all physics and chemistry arise, and hence everything else. When the emergent effects are rich in complexity- say life itself – does having a lower foundation arising from quantum indeterminacy diminish what life is? Certainty not. And free will and consciousness itself, emergent as it is out of computational indeterminacy, is not to be diminished either.
It’s actually buried deep in endnote 7. And it’s debatable whether he’s making the “I don’t know what came over me” comment about a third way between caused or random, or just expanding on random … and randomness is precisely the sort of case where we would be saying “I have no idea why I did what I did”, so that he’s just making it a dichotomy between the two — he even calls it a dilemma — makes more sense.
As to your comment, what we’d expect if it was neither determined by past experience nor random is that we’d see people acting mostly the way we’d expect given their past histories, but not always that way, sometimes doing things we didn’t expect “just because they felt like it”. Which is pretty much what we do see.
Kelton Barnsley,
If it were that “simple” it’s unlikely very smart people would have argued all sides of the issue for thousands of years, ongoing into today. (But of course anything can be “simple”
if we satisfy ourselves with simple answers and don’t go on to justify our own assumptions).
So, taking only one of your claims:
“If our choices are determined by prior conditions, then they are not freely made.”
Why should someone accept that claim? What’s the basis for it?
It appears you are simply equating “determined” with “not-free,” assuming the concept of “free” to be incompatible with determinism. But where do you get this from? Surely not common uses or understandings of the terms “free,” “freely,” “freed,” “freedom” etc. Those terms are commonly applied to describe specific real world differences in behaviors that exist whether determinism is true or not.
When a plumber says “I unplugged your toilet so the water is flowing freely” he’s not making some weird claim like “your toilet is now a contra-causal entity, magically free of all causation – I’ve made the water indeterministic!” No, what he obviously means to describe is the real existing difference in how the water can flow now, vs how it does not flow when the toilet is plugged. The difference being identified by the term “free” is true whether everything is determined or not, and hence is entirely compatible with determinism.
Same for saying “the dog is free of it’s chain.” No one means by this claim that the dog is therefore free of ALL existing constraints – only that it is no longer constrained in the specific manner of being chained.
Same with saying a person released from prison is now “free.” No one thinks this means “free from all causes, all constraints, as if they’d become magically excepted from the world of causation.” It simply describes the difference between being held captive in prison, including the things someone can choose to do when not held captive in a prison, vs being released from captivity. These differences exist to categorize whether the entire system is deterministic or not.
So the implication in your claim that “because X is derives from a deterministic system it is therefore not free” is simply false, since “free” commonly describes differences that exist within a deterministic system.
Hence, if we apply to word “free” to someone’s choice, it can be – and typically IS – used to describe only specific differences in constraints – e.g. I am sitting here of my own free will insofar as I have the capability to leave if I want (not physically constrained) or am not under physical threat to my well being by another person (coerced). Nothing about these applications of the word “free” to our choice-making conditions is incompatible with or incoherent within determinism.
So…what else have you got?
Or…can you give more detailed support of what you mean by your claim?
Vaal
Vaal, if religious claims were that simple, it’s unlikely that there’s still be an ongoing raging debate after all these millennia about the reality of certain imaginary friends.
And, yet, it really is that simple. The religious have an emotional investment in their imaginary friends, and so offer a neverending stream of excuse and obfuscation to buffer their irrationality from the incursions of reality.
Considering the oxymoronic nature of the term, “free will,” and its existence primarily as a religious and philosophical term of art — especially coupled with the fundamentally irrational and anti-empirical nature of both endeavors — I see no reason why “free will” shouldn’t summarily be dismissed out of hand along with other religious claims such as “soul” or “sin.” That doesn’t preclude us from discussing personality or worngdoing; it merely removes it from the incoherent and superstitious context in which it’s hopelessly mired in.
Cheers,
b&
Thanks, Ben. I couldn’t have said it better myself.
Vaal, I think you’re starting to sound like a broken record. We all understand that if I don’t have a gun to my head, then I am free to follow the dictates of my own mind. But I am not free to determine my mind, and when it seems that I am, I am not free to determine how I determine my mind. No matter how you look at it, my choices and everyone else’s are part of a larger network of cause and effect. Calling this truth “free will” accomplishes nothing but obfuscation.
Kelton,
I certainly get that I’m at the broken-record point; but that is because the criticisms of the compatibilist position are themselves broken records, and simply re-state the same unsupported critique over and over.
“We all understand that if I don’t have a gun to my head, then I am free to follow the dictates of my own mind.”
Ok, but you made a claim that tried to undermine the validity of “free” on the grounds that “it was determined.” Since that really can’t be the reason you reject calling a choice “free” (because it was deterministic) now you have to make better sense of your own position.
So what’s your new addition to your argument.
Seems to be this:
“But I am not free to determine my mind, and when it seems that I am, I am not free to determine how I determine my mind.”
So, it still seems the broken record of “If my choice was determined, it’s therefore not free.”
But, you’ve already accepted that “determined” does not entail “not free.”
So where’s the additional justification for your claim? I don’t see any.
You’ve already admitted that “free to follow the dictates of our own mind” is a valid understanding of freedom given determinism.
And that’s the type of freedom compatibilism is talking about. So compatibilism is, by your own admissions, coherent with common understandings of “freedom” and “free choice.”
But then if you want to keep claiming something like “But THAT’s not free will!”
what’s the basis for that assertion?
If it’s “because we can’t escape the chain of determinism” that criticism is already a non-starter. If it’s “that’s not what people think of when people talk about free willed choices” then you seem to be wrong there too, since examples like “not being coerced” ARE common understandings of free willed choice making.
Basically, you seem stuck in the incompatibilist marry-go-round of “If it’s not the impossible type of free will I’m imagining, it’s not really free will.”
Which is just question-begging, and frankly unsupported, whereas I keep giving real-world applications of the terms “free/freely/freedom/free choice/free will”
and showing how they are typically descriptions applied to real world conditions.
Vaal
“Ok, but you made a claim that tried to undermine the validity of “free” on the grounds that “it was determined.” Since that really can’t be the reason you reject calling a choice “free” (because it was deterministic) now you have to make better sense of your own position.”
Yes, the fact that choices are determined is the ground on which I claim that they are not “free”. Words mean things, and to claim that choices made by a deterministic system are “free” is to push that word’s meaning beyond the breaking point. You might as well claim that a chemical reaction which releases more energy than it takes from its surroundings can be “endothermic”. You’re in opposite-land.
“But, you’ve already accepted that “determined” does not entail “not free.””
No, I haven’t. That’s my whole point. Stop playing word games and just admit that you are re-defining “free will” to mean something else entirely.
Kelton Barnsley,
“Yes, the fact that choices are determined is the ground on which I claim that they are not “free”. Words mean things, and to claim that choices made by a deterministic system are “free” is to push that word’s meaning beyond the breaking point.
Ok, so you are making exactly the unsupported, mere assertion I keep explaining.
What is your basis for asserting that the term “free” entails conflict with deterministic systems? You haven’t given this basis, only repeated your assertion.
I’ve given plenty of examples showing how “free” as commonly used and understood (and as you’ll find it explicated in the dictionary) is NOT incompatible with determinism. You are simply ignoring these arguments.
But then we have this:
Vaal: “But, you’ve already accepted that “determined” does not entail “not free.””
No, I haven’t. That’s my whole point. Stop playing word games and just admit that you are re-defining “free will” to mean something else entirely.
Kelton: “No, I haven’t. That’s my whole point.”
What? But you’d just written this:
“We all understand that if I don’t have a gun to my head, then I am free to follow the dictates of my own mind.”
So, despite that you think everything is deterministic, you’ve accepted that use of the term “free” to describe un-coerced choice.
So, as I say, if you accept the validity of “free” in that context, you can’t go dismissing the validity of the term “free” on the basis that things are determined.
You’ve accepted at least one such use.
Or, did you want to retract, or clarify that statement?
Vaal
It is not appropriate to attack the concept of free will just because it has a strong traditional connection to religious thought. Nor is it appropriate to reject anything emerging from indeterminacy as always being sub-standard by definition. Yes – if all mental processes are wholly determinant, then free will can’t exist. But neither then can consciousness –which is also a mental process- exist as anything but a predetermined sequence of events and related predetermined mental impressions – in a Cartesian Theatre with nobody ever watching the movie. The behaviour of those who believe that no free will exists, but still choose to behave as if it does exist in their day-to-day life, becomes totally incoherent.
Free will is an emergent property of extremely complex computational processing where the processor is, in essence, a part of a programming system that is itself essentially indeterminate. The processing entity resides within the individual human mind where that computation takes place.
Those reasons you list are secondary.
The primary reason I reject “free will” is because it’s an oxymoron, a married bachelor who works in military intelligence for impoverished first-world nations.
That that conclusion is bolstered by all the other reasons only makes the imperative to abandon the concept that much more emphatic.
Yes, by all means, discuss the emergent nature of consciousness and the illusion of the self and the decision-making process and all the rest. Just please don’t muddy the waters with what has got to be one of the worst terms of art to come out of theology and / or philosophy. There’s nothing free about a will, and nothing willful about freedom.
Cheers,
b&
“Yes – if all mental processes are wholly determinant, then free will can’t exist. But neither then can consciousness…”
This is a mistake, one which I tried to expose earlier on this thread. I have reproduced the relevant section of my comment below:
“There is a very important distinction between consciousness and free will.
Unlike free will, the existence of consciousness stands on firmer ground, empirically, than any other phenomena.
Consciousness demonstrably exists, even if the contents of consciousness are an illusion (as in “The Matrix” or Orpheos’ cave). Free will does not. We think we feel like our actions are freely authored because we do not pay close enough attention to our experience most of the time. Take a few moments to observe the way thoughts arise in your consciousness, and you will realize that you have no more free will to determine what your next thought is than you have to determine what or who will next wander into your field of vision.”
And hkornstein, I never said that actions arising from indeterminacy or randomness were “substandard by definition” – just that they cannot possibly be construed as free will, since random/uncaused actions are precisely the type of actions which you cannot be expected to take credit/blame for.
Kelton, claiming that there is no such thing as free will but simultaneously claiming that consciousness DOES exist is like trying to both have your cake and to eat it. If you say that all mental processes are deterministic, hence there is no free will, you must then go on to explain why consciousness is not a mental process. Are you trying to infer some new form of dualism?
No – if mental processes are wholly deterministic mechanisms then consciousness is also a deterministic mechanism. Your “consciousness” is nothing but a Turing machine operating upon and interpreting a long tape of predetermined mental happenings and producing in turn further totally predetermined mental impressions that you choose to define as awareness. By this standard your iPhone is conscious.
Sorry for butting in here, but what on earth are you on about?
Are you claiming that conscioussness is an indeterministic process?
I wonder if suicide is considered the ultimate display of free will.
In the animated movie “Justice League: Crisis on Two Earths”, Owlman (a villainous Batman from a parallel world) tried to destroy the multiverse to prove that he had free will. He reasoned that if every choice a person makes is negated by one of their counterparts in another universe making the opposite choice, then the only “free” choice one could possibly make was to bring the whole shebang crashing down in flames.
Hehe… A quantum entanglement story. 🙂
Would you also speculate that a fly endlessly and fruitlessly flying into a window, trying to get, is the ultimate display of free will?
No, that just reminds me of this quote: “”A fool is someone who does the same thing repeatedly and expects different results.”
Ben,
The “smart people” I refer to who take various sides on free will are not aligned the way the religious are aligned against the secular; they have, and do, comprise the brightest minds from every angle of belief system, the brightest secular minds included.
So the analogy would be more apt to the complicated issues involved in morality and ethics, rather than a religious belief.
And…there are plenty of reasons, some of which I’ve already given, for why it’s not as “simple” to decide the matter of free will as Kelton claimed.
Vaal
Jesper “Are you claiming that conscioussness is an indeterministic process?”
Yep…. as consciousness is itself an extremely complex subfunction working within the the even greater complexity of the human brain it is subject to the same randomizing effects (non-computability) as with the whole. I see this as an accurate description of the human mind.
I know it sounds a rather confusing, but think of the implications of the alternative… that everything that you are aware of, have been aware of, and will be aware of, is nothing more the performance of a predetermined mental script which runs in reaction and conjunction with another predetermined life script. Indeed, this alternative really does beg the question of whether there exists anything OTHER than zombies….
“I know it sounds a rather confusing, but think of the implications of the alternative… that everything that you are aware of, have been aware of, and will be aware of, is nothing more the performance of a predetermined mental script which runs in reaction and conjunction with another predetermined life script. Indeed, this alternative really does beg the question of whether there exists anything OTHER than zombies….”
hkornstein,
THAT is an argument from consequences if I ever saw one.
It sounds to me like you basically are saying that a useful lie is to prefer rather than an uncomfortable truth.
Do you think that consciouness is somehow physically seperate from the rest of the organism/system?
I know determinism can be a bit scary, but after a long hard trip I’ve had no choice but to accept that fundamentally I am a meat-puppet dancing according to the laws of physics. That doesn’t exempt me from being responsible for my own actions, it simply tells me that we only get one shot at life and that we should do our outmost to make our short time here count.
It’s very difficult for me to square consciousness and qualia with materialism, even though I’m a materialist at heart. First, there seems to be no reason for them to exist. They are superfluous. Second, they have an immaterial quality. They may be, as some say, an epiphenomenon or an illusion, but that begs the question. They exist, and their existence requires an explanation that is so far beyond our grasp. I have no doubt that consciousness and qualia are embodied in and dependent upon material stuff, but there is a transcendence — not in a religious sense, but as abstraction beyond the strictly material.
It is a very tricky subject. Objectivity is in a way an oxymoron when we are talking consciousness and subjective experiences, because we still don’t know exactly how to define it and measure it.
We can’t even agree on what “it” is yet.
Our knowledge of the brain is still in its infancy imo.
Understanding how the brain works at the finest grain will do nothing to explain consciousness and qualia. It would be like looking at transistor state transitions in a server and deducing what’s going on at amazon.com.
Looking at the brain alone won’t do it, I agree, we also need to get a better understanding of the rest of the body.
I see no reason why scientific endavour shouldn’t, somewhere down the line, make us fully understand what consciousness and qualia is and what it entails.
My hunch, and it’s only that, is that the mystery of consciousness and qualia goes deep into biology. That’s where the transcendence and abstraction beyond naive materialism begins.
Well, at least you’re now wearing your woo on your sleeve….
b&
Completely false.
Once you’ve got a transistor-level simulation of an entire Amazon datacenter, you’re now free to slice and dice and play with anything and everything to your heart’s content.
Does it take work? Of course.
Would it be much easier to have all the design schematics of the hardware plus the source code for the software plus all the engineers gathered together in a room at your beck and call? Of course.
But is it necessary? Absolutely not.
There are these things called, “disassemblers,” for starters….
Time after time, you equate “difficult” with “impossible,” or “big” with “infinite,” or some other variation on the theme. Once you understand that “infinity” is most emphatically not a really, really, really big number that’s so big that you won’t live long enough to count that high, you might begin to make some progress.
Cheers,
b&
It’s tiresome to converse with you, and I find you to be gratuitously insulting, so I’ll pass, so as not to make a fool of myself.
This summary by Mr. Kornstein is very good.
Makes reading discussions between compato vs. incompato like discussion between scientists (based on reality) vs. religionist (based on jumping assumption without regard to reality and assume the other’s position based on own ignorance).
Funny that a person who used to attack others of something might actually doing the same on similar issue.
1. All brain / mental activities are natural, no gods, no demons, no duality. No question about this, xtianity is wrong.
2. While determinism is a good approximate, it is not absolute. Billiard balls’ movements relatively determined, but other systems are definitely not, weather systems, stock markets, human brain. It is again newton vs einstein.
3. You cannot say that because jesus is god-creator, then everything is created by him .. so you cannot say that because atoms (seems to, especially to your understanding) be determined, therefore everything is determined, determinism roolz. Both the above premises are over generalization, and the deduction is totally misdirected.
4. You cannot frame a debate by creating your own definitions (compato vs. incompato), while the premises are all skewed. Sophisticated Theologians (TM) do that.
Nice work Mr. Kornstein!
(an expert in one hing in most cases is an idiot in others, that’s why we need wisdom not to be a stubborn one)
So the only neutral disposition regarding free will is agnosticism?
Sounds vaguely familiar…
Instead of saying we have “free will”, why not say we are “deep actors”. An actor says his lines, but with a deep actor nobody, not even the director or the actor himself can reliably predict in advance what those lines will be.
An alternative description would be that we are creatures whose brains carry out Type Four computations (computations which are gnarly and whose results can’t be predicted in advance) – but I like “deep actor” better.
I want everyone reading this to notice the subtle wordplay at work in the following quotation from Vaal:
“Vaal: “But, you’ve already accepted that “determined” does not entail “not free.””
No, I haven’t. That’s my whole point. Stop playing word games and just admit that you are re-defining “free will” to mean something else entirely.
Kelton: “No, I haven’t. That’s my whole point.”
What? But you’d just written this:
“We all understand that if I don’t have a gun to my head, then I am free to follow the dictates of my own mind.”
So, despite that you think everything is deterministic, you’ve accepted that use of the term “free” to describe un-coerced choice.”
In the first paragraph, I’m saying that determinism does entail “not free” in the sense that determined actions are not free from prior conditions. But then, Vaal uses my statement that if “I don’t have a gun to my head, then I am free to follow the dictates of my own mind.”
It should be clear that in the second case I am using the word “free” only in relation to freedom from the situation of having a gun to my head. I think my use of the word “dictates” should make it clear that my actions are still NOT free of determinism in general.
Vaal, in these two contexts, it is clear that I am using the word free in relation to two different sets of constraints. You seem to think that I should never use the word “free” unless I am always referring to freedom from certain external constraints. But I do not object to the word free – I object to the term “free will”, as it implies that the will itself is free from any constraints external to itself. This is not the case, as our wills are the product of our genes, environment, the last tv show we just watched, what we ate for dinner, etc. You are confusing the use of a term in two different contexts with the inconsistent use of a term to mean two very different phenomena.
A final point: http://xkcd.com/1081/
Kelton,
I know you accept the use of the term “free” in one example, but not the other, (in the free willed sense). That’s the whole point I’m raising!
What I’m highlighting is that you haven’t provided any consistent basis for accepting “free” in the sense you accept it for examples like “not having a gun to my head,”
yet REJECTING that sense of “free” when it comes to human choice making related to free will.
In the first case you accept that “free” can be used compatibly in a deterministic system.
In the second case – “free” will – you REJECT that “free” can be used in a deterministic system.
I ask you why exactly, and you replied:
“Yes, the fact that choices are determined is the ground on which I claim that they are not “free”.
And…that’s it. Simply from the fact the choices are determined. But the fact choices are determined didn’t stop you from accepting “free” in the first instance…and you give no extra reason why you reject it in the second instance.
You simply claim that to say “choices made by a deterministic system are “free” is to push that word’s meaning beyond the breaking point.”
But you don’t justify that sudden exception at all. You go on to say:
“But I do not object to the word free – I object to the term “free will”, as it implies that the will itself is free from any constraints external to itself.”
No it doesn’t. Certainly Not in the compatibilist sense, the one you’ve been arguing against. In this case “free will” means “free to do as I will” not “my will is free from ALL external constraints.”
Just as you understand that it’s valid to say “I signed this document of my own free will” speaks to the condition of not being under direct threat by another agent – you should be able to understand THAT’s essentially what is meant by compatibilist Free Will. I am sitting here of my Free Willed choice means it’s what I want to be doing and I’m suffering not constraint on my ability to do otherwise. It doesn’t mean “I’m free of ALL constraints.” No one ever really THINKS like that when describing real life situations.
So if you accept that “freedom” normally can be describing only specific situations within a determined system (e.g. freely signing contrast, free from prison, dog is free from a leash…) then it’s inconsistent of you to say “No, you can’t use that sense of “free” for “free will.”
The only way you can justify this sudden exception is to claim something like “that may be coherent, but that’s not what is generally meant by the term Free Will, so you aren’t really showing ‘real’ free will exists.”
But…round again…that begs the question.
Compatibilism IS a long standing account of Free Will. Does it capture the essence of what people think about when thinking someone is making a free willed choice?
Yes! It does. For all the reasons I’ve given. And you’ve even already accepted it in the sense it’s often related to free will “I signed the contract of my own free will.”
BTW, great XKCD cartoon. I thought it was going to be the “someone’s wrong on the internet” one, which is apt for us types who keep these threads lingering.
Cheers and thanks, even if you think I’m out to lunch on this.
Vaal
WHEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE!
:-D!
One of the the important features that arises from the Church-Turing thesis is, that if we are assuming it applies to the human brain, then it is a corollary that the brain is a Turing machine and the Turing halting problem applies. This means that the outcomes of the algorithmic processes occurring in it are indeterminate (the outcome of an algorithm is unknowable before it is computed).
As Seth Lloyd writes in his “A Turing test for free will” paper: http://arxiv.org/pdf/1310.3225v1.pdf
“If she is part of the universe, Laplace’s demon must fail to predict her own actions. The computational complexity analogue of the halting problem shows that to simulate the decision making process is strictly harder than simply making the decision. If one is a compatibilist, one can regard these results as justifying a central feature of free will. If one is an incompatibilist, one can take them to explain free will’s central illusion that our decisions are not determined beforehand. In either case, it is more efficient to be oneself than to simulate oneself.”
In the words of Mark Twain “you pays your money and you takes your choice!”. Like Geddy Lee, Rush and Sartre I will choose free will.
While quantum uncertainty is not a basis for free will, David Deutsch in his book the “Fabric of Reality” discusses the implication of the Everett interpretation of QM for deterministic free will.
The idea is that two identical copies of a human being in recently de-cohered branches of the wavefunction may have different informational inputs resulting from an external quantum stochastic process, which causes their deterministic decision making algorithm to come to choose a different counter factual than the one chosen by the other copy. If he were an Everettian he would think “I could have chosen otherwise, as Other copies of me chose otherwise”
Consequently I can write “Many of those copies of chemicalscum are at this moment writing these very words. Some are putting it better. Others have gone for a cup of coffee.” to paraphrase Deutsch. By the way going for a cup of coffee in this case is the halting in the Turing halting problem.
While we are talking about both Seth Lloyd and David Deutsch, this debate between the two quantum computation experts on the “Fabric of Reality” book is both amusing and instructive.
http://meche.mit.edu/documents/slloyd_deutsch_debate.pdf
This is a reply to the comment above by khornstein: “Kelton, claiming that there is no such thing as free will but simultaneously claiming that consciousness DOES exist is like trying to both have your cake and to eat it. If you say that all mental processes are deterministic, hence there is no free will, you must then go on to explain why consciousness is not a mental process. Are you trying to infer some new form of dualism?”
Khornstein, I don’t know how consciousness arises (though I suspect it has something to do with brains), and I am comfortable admitting my ignorance on that front. As I wrote above, I don’t consider the hard problem of consciousness to be solved.
That said, the existence of consciousness is an empirical fact. Before you know anything else, you know that you are conscious. I’m not sure if you doubt this claim, or if you think that by not doubting it I should also have a full description of how consciousness arises. Either way, you are mistaken. If you want to claim that you are not conscious, I don’t know what you’re talking about. And if you have a solution to the hard problem, I’d love to hear it.
The existence of free will is not on the same footing, empirically, that the existence of consciousness is. Most people only think they have free will because they do not pay close enough attention to the way that thoughts, intentions, and desires arise. Here’s an exercise: think of a movie.
Got one picked out? Good.
I’m guessing that several movie titles appeared in your mind before you settled on one. You might have heard the titles read in your own voice, or you might have pictured the text, or you might seen movie scenes being replayed in your mind’s eye. You also might have waffled between two or more possibilities before settling, for reasons that are perfectly mysterious, on one or the other.
Now let me ask you: are you free to think of a movie which does not occur to you? No. You can only think of things that do occur to you. Where is the freedom in that?
This experiment, coincidentally, should also demonstrate to you that you are in fact conscious, in case you weren’t already convinced.
Clearly, consciousness exists without free will.
“Now let me ask you: are you free to think of a movie which does not occur to you? No. You can only think of things that do occur to you. Where is the freedom in that?”
I can not understand your problem. If there are N movie titles stored in my memory and I am asked to name one, about 5 names “pop up” immediately. I choose one that I fancy. Why? – because I’m lazy and the problem seems trivial. However, if I had chosen to, I could spend all the afternoon listing all the movies I could retrieve from my memory to get a wider choice. I could devise endless methods of narrowing down to a final choice. I could choose all titles beginning with the letter T. The options and outcomes combine to provide millions of possibilities. What then determines my choice… the actions of my mind – an unbelievably complex state machine and the programmes that it executes. If that machine were a simple computer it would ALWAYS retrieve the same film, only one selection algorithm would apply. THAT is deterministic. But that is not what I do…. I programme myself to answer the question. You can not rely on me to come up with the same answer twice, nor the same method of choice. Can I state in advance what method and what choice will be forthcoming? –NO. It is indeterminate on many levels (one being that the method may not be computable). I alone am the author of the process and the result – free choice. Now a computer could be programmed to simulate human “movie choice” by adding complexity and some randomness, but that would just provide an external impression of what I do, not a duplication of processes which provide not only the possibility of choice, but the possibility of invention of method. As for consciousness, it can not be deterministic in this case of free-will, because it is structured on the identical architectures and programme structures as my other mind functions.
You’re assuming a computer model that’s too simplistic to the task at hand.
One that would actually fit the scenario you describe would not only have the same list of movies, it’d have your same memories of what it was like watching them or their trailers (or whatever), and it’d have your current thoughts and mood (Hard day at work and only interested in something mindlessly entertaining?) and so on.
It’s quite unlikely that there’s any significant true randomness in the human brain. But, if there is, any choices that hinge on such randomness are obviously not significant to you, or else the randomness would be swamped by actual preference.
Actually, the scenario you describe is exactly a deterministic one. A chaotic one, perhaps, in which even minor changes in input can lead to significant and practically (though not theoretically) unpredictably different outcomes. (Did that cute girl / guy smile at you as you got on the elevator, and was an otherwise-stressful day instead filled with thoughts of potential romance, making your movie preferences swing towards romantic comedy? Or did he / she frown and work therefore had a black mood hanging over everything and you could really go for a zombie flick?)
But, honestly, I rather suspect that much of what we do is far less unpredictable than what most people suppose. Yes, we’re very, very, very complicated and sophisticated computational machines. But we’re also creatures of habit, which means that huge swaths of our behavior can be optimized away. The resulting picture might be a bit fuzzy, but it’ll be perfectly recognizable.
At that point, it’s just a matter of filling in more and more details, to whatever level of precision you require.
Cheers,
b&
“You’re assuming a computer model that’s too simplistic to the task at hand.”
And you Ben, are assuming a human mind that’s far too simplistic – and therefore one that can be compared to such a deterministic computer system. I do not claim that ALL mental activity is randomised, but that there is a strong influence of randomness arising from computational indeterminacy and the chaotic tendencies in such a complex self-programming system with its many feedback loops. And I do not denigrate the influence of randomness as unfitting for any definition of free will, as all complex and elegant emergent properties in this universe have a base in randomness.
As a computer systems designer myself, I ask the simple question – “where does the prime responsibility for the system behaviour lay?” and the simple answer is it is in the individual human mind. The follow-up question “is the system behaviour wholly determinant?” and I must answer NO, not by any standard we ever apply in computational systems.
I must say I do not understand the tendency of so many here to rush to the conclusion that there is no such thing as “free will” based on such scant evidence, especially when the limited understanding that we do have of brain function and computational theory points in the opposite direction.
Cheers.
But chaos is explicitly non-random! Chaos is the impracticality of predicting the outcome of certain processes because of the variability and influence of initial states. It has to do with computational efficiency and available resources; it has nothing to do with computability.
Consider our great friend π, for example. It is famously computable, as everybody loves to do it. However, you can trivially pick a number of digits of π that would require more physical resources to calculate than exist in the entire Universe. But that does not make π incomputable! It doesn’t even make calculating that many digits of π uncomputable.
The question of computability has nothing to do whatsoever with whether or not it’s even theoretically possible for an engineer to put together a Bill of Materials for the computer you’d need to do the job; it’s all about whether or not there exists an algorithm (or its equivalent equation) that produces the same results.
Even then, randomness still doesn’t enter into the picture. Not even strong randomness; not even quantum randomness. The question isn’t whether or not the algorithm could predict the truly random figures; the question is, if given the same truly random figures as input, would the algorithm come to the same conclusion as the process in question?
Take off your engineer’s hat and put on your theoretical mathematician’s hat instead. Implementation is irrelevant.
But, even still. Let’s look at it from an engineering perspective. The human brain has (with rounding) a hundred billion neurons. The latest Intel CPUs have (with similar rounding) a hundred million transistors. Build a supercomputer with a thousand such CPUs and you’ve got a transistor per neuron. You’d likely need more than a 1:1 mapping, but no worries. Build your CPU with ten thousand, a hundred thousand, a million such CPUs. Or wait for Moore’s Law to catch up; ten more eighteen-month doublings, just fifteen years, gives you a thousand-fold increase, meaning you’ve now got a single chip with as many transistors as a human brain has neurons.
Remember. The question isn’t, “Is hkornstein capable of leading a project to create a digital simulation of an human brain?” The question is, “Does there exist an algorithm that, no matter how unwieldily, given the same inputs, would produce the same results as an human mind?” And the answer to that is, clearly, unquestionably, a resounding, “Yes.”
Cheers,
b&
The engineering perspective makes matters *worse*, since the question of engineering appropriate idealizations looms. Even Hava Siegelmann goes to the trouble of refuting her own model (in effect) that way by showing that any noise whatsoever in the analogue nets produces a *subTuring* machine.
My recent paper (mentioned above) goes uses the work of Norton to approach the idealization in another way, however.
Erm…that’s not even worng.
Again, Turing equivalency has absolutely nothing whatsoever to do with actually building a device to simulate anything.
Next, humans most emphatically are not in any way, shape, or form, “subTuring.” We are unquestionably Turing-complete. Any even borderline incompetent programmer can construct a mental model of a Turing Machine with all the requisite operations and provide you with the correct answer; at absolute worse, an human can physically construct a replica of a Turing Machine or equivalent device and use that as a crutch.
But, most importantly…if it’s your positions that humans aren’t even as capable as a Turing Machine, that means that the Turing Machine is actually more than is needed for the emulation.
Really, it’s like we were discussing whether or not a pilot could survive supersonic flight, and you came back suggesting that a lone human on foot would have trouble even walking at two miles per hour. Say wha…?
Cheers,
b&
“Next, humans most emphatically are not in any way, shape, or form, “subTuring.” We are unquestionably Turing-complete.”
Er, Ben, I hesitate to disagree with you, but the infinite tape is, as you have also implied in this thread, rather important and there for a reason. We, on the other hand, have no infinite tape or any other infinite resources, unless you suppose that we can, at least in theory, have our descendants carry on some computation for us through unbounded generations.
When discussing whether or not a system (computer, language, whatever) is Turing-complete, it is never a requirement to have an infinite tape. The computer you’re reading these words on is most emphatically Turing-complete, and it, too, lacks an infinite tape.
If an infinite tape were a requirement for Turing completeness, then nothing would ever be described as “Turing-complete.” And yet, you can even buy programmable calculators that are Turing-complete. Indeed, all you really need for Turing completeness is conditional branching and an arbitrary number of variables; the rest can be bootstrapped from there.
Cheers,
b&
“The computer you’re reading these words on is most emphatically Turing-complete, and it, too, lacks an infinite tape.”
Yes, as soon as we leave mathematics and start building real devices, the best we can do is “Turing-complete up to where its resources run out”. That’s usually good enough, and we call it “Turing-complete” for short, to avoid unnecessary pedantry.
If I keep following this discussion, I’ll have to watch out for arguments that depend on flip-flopping between the senses.
The question is, “Does there exist an algorithm that, no matter how unwieldily, given the same inputs, would produce the same results as an human mind?” And the answer to that is, clearly, unquestionably, a resounding, “Yes.””
You might be quite surprised to hear that I would ALSO give a resounding Yes to your question – such a system would produce exactly the same results as the human mind.
But you are asking the wrong question… the question you need to ask is do the RESULTS of that elegant computers processing include the property of free will. My answer to THAT question would ALSO be “Yes”
You seem to assume that any computer system running any algorithm must end up being deterministic. But that is not true.
Back to Turing…. computability is related to a determination if a computer working with any particular algorithm can ever provide a final answer (ergo full determinacy- it is not a “flawed” program). Reaching this allows the Turing device (or any computer or state machine) to halt processing. The Turing machine is permitted to work for an indefinite time to reach this possible state. The algorithm run by the Turing machine might itself not be computable. Three things further magnify the chances of non-computability in the human mind – the code is self-modifying, the processing time is arbitrarily truncated, and it is a multi-processor system.
As for chaotic behaviour, it is especially likely to be found in complex systems with feedback and iterative processing (e.g. the mind)…. but yes, chaos actually HAS deterministic results though highly sensitive to variability. BUT coupled with inputs or feedback that are possibly non-computable magnifies any element of randomness in the total system. Then there is the problem of the precision of the processing versus the level of sensitivity in the chaotic behaviour.
What puzzles me is why the function of decision in the human mind is not MORE random than it is, not less. On this issue I can only guess that mostly, the mind is solving very simple information retrieval problems that are computable.
Of course we don’t know the minds exact algorithms yet, nor it’s exact structure…. but we know enough to very seriously question any chances of it being wholly deterministic. I do find it incredibly surprising that those on the other side of this argument can dare to claim that they speak with any certainty at all on this question.
Cheers
Just to clarify: are you suggesting that the brain does actually perform these computations but a Turing Machine couldn’t? Or are you simply suggesting that they’d both give up at the same point?
If the former, we’ve got a problem. If the latter, you’re flogging a red herring.
This point has been made repeatedly, and not just by me. From the perspective of the Turing Machine, any randomness (including quantum or whatever) would constitute input to the Machine. And, more to the point, there is no will in randomness; randomness is what you use when you don’t give a damn. “Heads it’s chocolate sprinkles; tails, rainbow sprinkles.)
Cheers,
b&
Hkornstein said:
Indeed. Although one might suggest that that applies to both sides, to all sides, as there seems to be more than a few. But, more particularly, one might wonder what is the basis for your certainty, and Ben’s, that there exists “an algorithm that, no matter how unwieldily, given the same inputs, would produce the same results as an human mind”. Seems to me that if there’s any merit to the entire science of complexity and emergence then it appears quite likely there are going to be processes going on in the mind, within consciousness, that simply can not be *reduced* to any particular algorithm.
The certainty comes from the Church-Turing Thesis, which states exactly that and which is on superbly solid ground from all directions.
There may well not be any algorithm more efficient than the one the brain is using. But a simple brute-force method for a logically equivalent (but less efficient) algorithm would be to simulate the brain to whatever level of physical detail is necessary — whether that be at the level of neurons, or molecules, or all the way down to Planck Length, or whatever.
Cheers,
b&
“I programme myself to answer the question.”
Perhaps, but the you that’s programming your future self was programmed by prior conditions, including your past self. Where’s the freedom?
Ben, I will try to reduce my case to it’s simplest elements.
First, I think you do not appreciate the full structure of the Turing solution that I have described (and which I would say mirrors that of the mind). It is not a single processing entity – it is multi-element interconnected system made up of specialist subunits which themselves have subunits down to individual “neurons” that are not unlike Finite State machines (though extremely complex ones). Such a system is not only behaving according to the algorithms embedded or inherent in the state machines all the way up to the subunits, it is working according to dynamics established by the overall system architecture- it is a dynamic system in its own right. All the other elements I’ve described ….. feedback, self programming , and timing truncation are at play here… The level of interconnectivity and complexity in this system makes asking the question “is the internet deterministic?” seem trivial.
NOW…
1. Such a system (because of possible non computability) breaks the “chain of causality” related to an external input state because of the randomness inherent in the overall system
2. If this is the case the system is partly “free” of these specific inputs
3. The system itself is the principal actor if we care to locate (or “assign responsibility”) for what behaviours do occur
4. Randomness is no reason to denigrate rich complex emergent behaviour
I’m off to Canterbury for a few days so I can’t be posting. Don’t take any silence as any indication that any counterargument (if you could possibly have one) has me stumped
“Randomness is no reason to denigrate rich complex emergent behaviour”
Again, I’m not denigrating “rich, complex emergent behaviour”. It’s clear that that exists. What I’m denigrating (or rather, criticizing), is both the incoherent notion of “free will”, and the compatibilists’ insistence on calling a-will-completely-constrained-by-determinism-and-chance “free will”.
It should be obvious that the compatibilist argument is nothing but double-speak. “You have free will as long as you don’t think too hard about the deep causes of your behaviour or pay too much attention to your own subjectivity.”
As I just noted in another reply, there there is as much difference between, “difficult,” and, “impossible,” as there is between the largest number you can imagine and infinity. That there may or may not be a random input physically located within the system is irrelevant; logically, that’s part of the input fed to the algorithm. It’s also conceptually irrelevant; if the decision-making process has reached such an indeterminate state that it can be swayed by random input, there’s no significant meaning in the conclusion that the randomness tips the scales to. And there’s certainly nothing willful about such an outcome.
It may be impractical to model an human brain (though I rather suspect we’ll see it done before today’s newborns graduate high school, assuming no collapse of civilization), but impracticality is entirely different from impossibility.
Cheers,
b&
“That there may or may not be a random input physically located within the system is irrelevant; logically, that’s part of the input fed to the algorithm.”
No, not at all- the randoness may be “in the system” but it is NOT in the input (or assignable to an input function)…. it is in the results of the execution of the algorithm(s) – for reasons I have endlessly mentioned. Causality is broken, input to output. The only argument left to you is “does randomised processing count as free” as you put it. I say yes, because the decision making process is attribuatable to the “self” and the results still rational and is complex. If i get up in the morning and choose a tie to wear, randomness doesnt make me hang an ironing board around my neck, i still choose one of my ties by what i say is “free will”
“The only argument left to you is “does randomised processing count as free” as you put it. I say yes, because the decision making process is attribuatable to the “self” and the results still rational and is complex.”
If a decision being made by randomness inherent in the system is all it takes to satisfy your definition of “free will”, then I suppose you think a slot machine has free will, too.
Randomized processing does NOT count as free. No combination of determinism and chance can account for free will. It is an incoherent concept that cannot be mapped onto any conceivable reality. Let’s just let the term go.
Yes. You keep mentioning them. And you keep being incorrect and / or irrelevant when you do so.
Physical computers can have insane numbers of inputs. A Turing Machine only has one. The two are logically equivalent, a point that you’re refusing to understand.
Imagine a black box flight recorder. All those inputs coming from all those different sources at all different times in a practically chaotic manner…and they all get recorded onto a single hard drive (or flash device or whatever). And, from that single stream of ones and zeroes recorded onto the device, you can reconstruct the exact sequence of inputs and exactly reproduce the computation that the flight computers did based on those inputs.
A Turing Machine simulation of a mind would be the same basic idea.
Start with an exact computer copy of you the moment you wake up, and feed it a black-box-style stream of all the inputs you experienced after waking (including any true randomness, if any, that exists inside your brain), and it’ll pick the exact same tie. Do the same thing the next morning with a fresh copy of your initial state and the series of inputs you experience, and the computer’s tie selection will again match yours. And if circumstances led you to try to hang the ironing board around your neck, the computer, based again on your initial state and subsequent input, would also try to hang its virtual ironing board around its virtual neck.
Cheers,
b&
Proposal to incompatibilists
Compatibilists and incompatibilists have been battling it out endlessly on these threads. In an attempt to move things forward, here’s a proposal.
Since the incompatibilists insist on it, let’s restrict the phrase “free will” to meaning “the will has no physical cause”, i.e. dualism.
For the compatibilist sense of the phrase, let’s use “eerf will”.
“Eerf”, a reversal of “free”, means here what “free” means in every other usage of “free”. In “free speech” it doesn’t mean that the speech is not physically caused, it means it is not externally constrained.
I had thought that incompatibilists might be persuaded that this is a valid interpretation of “free will”, but it seems not, they are so hugely hung up on semantics that this is a complete sticking point for them.
Well, we compatbilists are not going to get stuck on mere language, so let’s use “eerf will”.
Now, having settled that (and we all agree that there is no “free will”, but there is lots of “eerf will”), can we now proceed to more interesting questions:
(Q1) what, in a determinstic world, is the difference between a cat making choices and a bricking falling and bouncing off scaffold?
How do we understand the goal-oriented choice-selecting behaviour exhibited by mammals (products of Darwinian evolution) and in some things created by those mammals (chess-playing computers and aircraft autopilots)?
How do we understand these things, given determinism, and what language to we use for it? What do we mean by “purpose”, “goal”, “choice”, “decision” in a deterministic world?
(Q2) What is the best way of persuading dualists to ditch dualism and accept determinism? Incompatiblists think it is jumping up and down yelling “there is no free will” repeatedly.
Well, if you were wanting to persuade a vitalist to ditch vitalism, you likely wouldn’t try “those cats and bears and dolphins are not actually alive”, you’d rather try to persuade them that the notion of being “alive” does not involve any “elan vital”, and that a better understanding does without it.
Would emphasizing, “you have eerf will” be the better way of persuading dualists?
“Well, we compatbilists are not going to get stuck on mere language, so let’s use “eerf will”.
What excellent terminology. It also gives we compatibilists the opportunity of telling the incompatibilists to eerf off.
Ben, We have reached that point in debate where the two contestants end up endlessly (and tediously) saying to each other “you’re wrong”. Let me propose a few questions to see if we can break the deadlock…..and to prove of course that you are wrong 😉
Let me first propose a little thought experiment. Let us SIMPLIFY and EXTERNALISE the problem to test the validity of your thesis on computability and determinacy in the human brain.
PROBLEM: At some arbitrary time I ask you to sit down at a table and sort out six randomly sized oranges into ascending order of size from left to right. This is just one of an infinite number of puzzles I could give you.
The question I now ask is “will you come up with the identical solution methodology at every possible time that I will propose the problem to you?”
Now if you were a computer you would. You would COMPILE a solution to the problem that I give. The language of compilation could be C++ or Basic or Java or FORTRAN etc. etc. in any one particular implementation of the particular compiler language. Given this choice of language and implementation the code produced to solve the problem (the executable) will ALWAYS be the same. (This case is “sorting” problem solution) I would point out to you that the executable solution itself has a possibility of NOT BEING CORRECT due to possible flaws in the compiler. This is always true… we can never produce a perfect compiler, or prove or test that it is… the reason being explained by Turing non-computability. It is fundamentally important to recognise this fact. The human mind is a sort of compiler in that it too compiles a solution using a symbology.
Now let me ask you, as a human being, will YOU (as with the computer) always produce the same “executable” method of solution??? Are you deterministic even to that simple degree? And are you then claiming that that ANY SOLUTION that you produce, unlike any compiler ever produced, will always be correct????
An additional question, if you had different levels of “impatience” in feeling how much time you should use to come up with a solution when asked…. would ALL solutions you come up with be the same?
Of course not.
NOW regarding the requirement for all solutions to the problem to be the SAME in any instance you might say “well the input conditions were different, the time I allowed was different…. It was hot in the room, I just had an argument with my girlfriend etc. etc. etc.” But are these other conditions RELEVENT to the solution? NO, they just add an element of randomness – like a stop on a roulette wheel. And like all randomness they will be different every time you approach a problem
NOW a further question – is the quality of your human input measurements so precise and so consistent that you will ALWAYS judge the orange sizes the same?
NOW suppose again that your (possibly incorrect and somewhat randomised) solution was in turn fed to produce another solution to another problem concerning the oranges. (This string of processing is massively long in the brain) Do you still claim that the result in all cases is determinant?
Now I have just given you a minor subset of the issues that exist in the human mind as a computing entity. I could greatly expand on these (and I touched on these factors in previous posts).
Are you convinced now? If not I will once again give you a list of the main factors that break causality in human decision-making.
My suggestion – stick to the argument that an underlying level of randomness cannot be “free” and we can move on.
Cheers
Howie
Howie, not only are any of your questions / examples relevant, they reflect a quite naïve Hollywood-style misunderstanding of computation.
If we’re going to go to Hollywood…don’t think of Robbie the Robot flailing his arms whilst chanting “DOES NOT COMPUTE!” in a monotone.
Instead, think of The Matrix. And not of the human characters plugged into the machine through giant bundles of cables leading into the base of the skull, but of the pure computer entities themselves, such as the Oracle and the Architect. Shirley, you must not think that your orange-sorting thought experiment is any more applicable to the Oracle than to Neo?
I know, I know. Argument from Hollywood is hardly precise. But, again…your objections are all of the sort that a scriptwriter would suggest as plot devices in a cheesy movie without regard to reality, so I’m not sure what other direction to go.
Cheers,
b&
Ben, a lesser person than myself might easily accuse you of obfuscation because you cannot come up with a counterargument to the conclusions that necessarily follow from my thought experiment. And I assure you that my “orange sorting challenge” example very accurately encapsulates a number of the issues of computational theory that back up my overall argument – computability, the halting problem, solution time truncation, data resolution, compilation theory and multi-threading- but in a much more easily conceptualised everyday form. So I offer the question to you again, and hope for a better response.
If you answer that a super-supercomputer array exactly simulating the processes running in a human mind when solving the “orange sorting challenge” would produce identical sorts of behaviours I would agree … but as I have stated before, such a comprehensive simulation would also exhibit free-will.
We can then go on to the question of does randomness “count as free” and join Kelton in this line of discussion?
Cheers
Howie
BTW – Computing is my field of expertise (lots of degrees etc.) so I don’t really think that I exhibit “a quite naïve Hollywood-style misunderstanding” of the subject. Not that I am resorting to any argument from authority in the matter- just wanted to clear the air.
Yes, of course to the first part — that’s been the point I’ve been trying to get across. That the mind is a Turing-equivalent device, and could be simulated by any other Turing-equivalent device with adequate resources and proper programming and input.
But now you’re back to describing a clockwork mechanism as having free will, which is precisely the exact example of what 99 44/100% of free will advocates would use as something that does not have free will (whatever it actually is).
You can make the point, as I do, that, when most people are pointing to what they label as their free will, they’re pointing to their internal mental simulations of the likely outcomes of different choices and deterministically selecting one of those choices based on their analyses of those simulations. But doing so is exactly as dishonest as a naturalist feeling awe watching an eclipse over the Grand Canyon and saying that that’s Jesus.
Cheers,
b&
Your assertion that “the mind is a Turing-equivalent device” is just that — an assertion.
If you’re going to dismiss the Church-Turing thesis and the repeated explanations I’ve offered about the completeness of our understanding of the relevant physics as mere assertion, then I’m going to dismiss your objections as woo peddling.
Sorry. The ghost of Hitchens so compels me; I have no choice.
b&
>
You so much as admitted that there may be physical things that are not computable. At least, you said that the opposite was not your position.
I think it’s very likely that there are physical things that are not computable. So there. That’s my assertion.
I don’t recall ever admitting that there are physical things that are not computable. I’m sure that I’ve described plenty of example of incomputability, especially including infinite sequences and the Halting Problem — but none of those examples can possibly have physical manifestations. You can most emphatically solve the Halting Problem for any physical computing device — though doing so may well require in some instances more computing power than could be physically assembled from the stuff in the observable universe. But it would be a specific solution to that specific device; it would not be a general-purpose solution to the Halting Problem.
Pure Chopra.
b&
When you raised the ghost of Max Planck I knew there was something wrong. You seemed to envision a Turing Machine with not only an an infinite tape (an impossibility), but the capability of infinite precision (for chaotic systems)– another impossibility, accurate down to the Planck length, with perfect knowledge of initial conditions. On top of that, you assume a complete theoretical knowledge of how such a simulation would go forward, which is not available. And you dismiss any concern of efficiency or practicality of implementation.
That armchair musing is supposed to close the books on materialism and refute, once and for all, any question of free will.
It sounds like theology to me.
Again, you expose your complete and utter ignorance of the fundamentals of information theory.
Turing Machines are not physical devices. They are algorithms, described as if they were a particular type of physical device in order to make it easier to comprehend them.
By even hinting at that a physical limitation to constructing a particular proposed Turing Machine is even vaguely relevant to questions of computability, you demonstrate yourself unfit to participate in the conversation. It should be as embarrassing to you as if you were to claim that no more than several dozen digits of π can ever even theoretically be known, because that’s how many orders of magnitude there are between the largest and smallest scales of the Universe.
Cheers,
b&
You can have your infinite-taped omniscient Turing Machine in the sky. I’m skeptical.
‘You can make the point, as I do, that, when most people are pointing to what they label as their free will, they’re pointing to their internal mental simulations of the likely outcomes of different choices and deterministically selecting one of those choices based on their analyses of those simulations.
But doing so is exactly as dishonest as a naturalist feeling awe watching an eclipse over the Grand Canyon and saying that that’s Jesus.’
Ben:
Is the dishonest part (or perhaps “mistaken”, instead of dishonest?)[XXX] the internal mental simulations of likely outcomes of different choices? Or is it [YYY] that I delude myself when I think I consider two or more options for action, then believe I am actually exercising mental faculties when I finally select one out of all the options I recognize as existing on a selection menu? Or is the mistaken part some [ZZZ] not listed?
I encounter so many different definitions of various terms I’m always hesitant to assume I am on the same page as anyone else. For example, it seems to me that your definition of determinism means that any event that occurred prior to the next instant beyond the present could be mapped — in the sense that every atom in the universe occupied a particular place at any prior instant, and for any prior event to have been different than it was at the time requires a different arrangement, a rearrangement, of those atoms, which clearly is impossible to arrange sans time travel (not an option for purposes of clarity, and about as likely as d*g anyway from my limited understanding). If this paragraph is in error, I don’t mind being shown the way.
My confusion is due to the claim I think I hear made — and I could be completely missing the mark here — but I do not know how it can be asserted with certainty that the human brain does not select from a palette of choices merely because it is impossible to change the circumstances that existed the precise moment I told the waiter to make it strawberry after all, and nonfat frozen yoghurt instead of ice cream, this is final, sorry I’ve been a pain in the ass, and please walk away from my table before I change my mind for the sixth time.
It seems to me that, if the determinist position is that Richard thinks HE made a decision in his brain when there is not even truly a selection set (much less a HE whose brain has the ability to make choices from a supposed selection set), then this determinist position must be falsifiable. I am on the right track? If so, what is the test for the determinist position that proves it?
This is long, and the subject has me so flummoxed I am not sure I understand what I just wrote. Sorry if it is too incoherent to mess with.
[ZZZ]
The dishonesty lies in taking a term, “free will,” that is expressly used to refute the notion that humans are (amazingly sophisticated and complex) clockwork machinery and redefine it to now apply it to exactly that clockwork machinery. Indeed, it’s a perfect example of Orwellian newspeak. It’s like pointing at a person strapped to a giant marionette mechanism and declaring him to be free because he’s been drugged and conditioned in such a way that he enjoys participating in the show.
There is no coherent definition of the term, “free will.” Period, full stop, end of story. A will that does not faithfully proceed according to its nature is no will at all; and freedom, once constrained, is no longer free.
Do humans have wills? Most assuredly. It is the will which drives the runner through the last mile of the marathon, which compels the student to finish the degree, which keeps the soldier from fleeing the battlefield in terror.
And do humans have many freedoms? In most civilized countries, yes — though, to be sure, America is far less civilized today than when I was a child, what with all the “free speech zones” and the CIA universal wiretap and the TSA gate rapists and the militarized cops stealing property in the name of the war on drugs and what-not.
But describing any of that as “free will” makes no more sense than calling a man a “married bachelor” or describing a woman as “a little bit pregnant.” Just because the words have valid meanings in their proper contexts doesn’t mean that you can string them together however you like and have them still be meaningful. Hell, all the letters in the alphabet are useful, but stringing them together randomly, such as, “eoOhvCNSkzFdncruEolx,” doesn’t mean the result has any meaning, either.
Cheers,
b&
Ben:
“If you’re going to dismiss the Church-Turing thesis and the repeated explanations I’ve offered about the completeness of our understanding of the relevant physics as mere assertion, then I’m going to dismiss your objections as woo peddling.”
Ben again (and quite rudely: “By even hinting at that a physical limitation to constructing a particular proposed Turing Machine is even vaguely relevant to questions of computability, you demonstrate yourself unfit to participate in the conversation.”
Ben, with respect, your understanding of Church-Turing and its implications is seriously seriously seriously flawed. You seem to imply that Church Turing implies that all solutions produced by computers will be computable (will finally produce a “correct solution”) and will never vary from that correct solution. That is not the case. Church Turing essentially says that ALL computer structures are equivalent and will produce the same results, BUT these results can be INDETERMINATE, and in general you cannot test a programme to see if it will produce that computable result. AND the Turing machine is an ideal…. It is given infinite time and infinite memory, and infinite data resolution to work with. No REAL computer is like that.
Then there is Turing applied to compilers – software languages that produce the executable code of a programme solution… Turing leads to the fact that no compiler can be assured of even producing error free code. Much of Computer Science is focussed on trying to avoid the pitfalls of having what is equivalent of non-perfect Universal Turing Machines. But even if this was achieved the problem of computability would not go away.
Evolution did not produce the human mind in the direction of being an ideal Turing Machine, it is a “problem solver” just like normal physical computers (but with far less of the disciplines that engineers and programmers apply to making their problem solving machines and programmes).
What does this mean?
It means that the human mind regularly violates the necessary requirements of being a “good Turing machine” and the requirements that we impose when we build a REAL computer and a real computer language. (This will allow us therefore to simulate a human mind with more rigorously precise “Turing-like” machines-ie REAL computers- if we ever learn how the mind works). We (the human mind) truncate time to produce solutions, we work with data at precision levels that can produce error, we work with finite memory, and more “computationally dangerous” we create programmes to solve problems “on the fly” . That is why a computer solves the “orange sorting problem” the same way all the time, and a human being doesn’t.
What does this mean?… doing things the way it does our minds BREAKS CAUSALITY and allows a level of randomness to enter our solutions. This leads to free will. Randomness is what allows humans to explore mental constructs, to “think”, and to create. We explore a multi dimensional solution space in a way that quantum indeterminacy allows mutations that explore multidimensional fitness landscapes – that leads to Evolution. We should not denigrate the “blessings” of randomness.
You seem to be constantly uttering some mantra about Church Turing when you do not realise that Church Turing actually proves that free will is possible.
hkornstein, I’m afraid you’ve got it turned precisely umop-apisdn.
Incomputability only comes into play with computers with infinite resources, such as a real Turing Machine. If the computer is finite, as an absolute worst case, all one need do is play out every possible permutation that exhausts the computer’s finite resources and examine the computer’s state at the point those resources are exhausted. Incomputability comes into play when those resources cannot be exhausted.
Unless it is your claim that the mind is infinite and has resources that cannot be exhausted, then there is nothing incomputable about the mind. Yes, of course — I’m sure that one could intelligently design something more efficient than the mind. But the inefficiency of the mind doesn’t mean that it does things that more efficient computational devices cannot; quite the opposite.
Put another way, you’re claiming that a beat-up old Yugo can do things that a Formula 1 racer can’t because the Yugo is running on three cylinders and the F1 car has all twelve pistons available for maximum power. It just doesn’t even pass the sniff test.
And then when you and the others start to go off into quantum woo land…well, that’s when the comparisons to the guru of quantum woo, Deepak Chopra, start to come out. If you find such comparisons offensive, don’t spew quantum woo! There’s nothing about quantum phenomena that make them incomputable. Unpredictable, yes, but not incomputable.
And randomness leading to will? Really? You think that making decisions based on a throw of the dice is meaningful and the source of our will? That a runner near the finish line of a marathon tosses a coin to decide whether or not to take those final few steps, and that coin toss is what constitutes the heart and soul of her will?
And you expect people to take this stuff seriously?
Randomness can be quite useful in all sorts of contexts, but it is the antithesis of decision-making. Basing your decisions on randomness means leaving things to chance; there’s no will there, not even poetically.
b&
“Incomputability only comes into play with computers with infinite resources, such as a real Turing Machine.”
NO – incomputability comes about when ANY computer cannot solve a problem by executing an algorithm. Solving the problem terminates the algorithm – it has executed to completion. All computers are subject to this mathematical possibility. This means that the computer, a totally deterministic device itself can PRODUCE UNDETERMINISTIC RESULTS when executing certain NON DETERMINISTIC algorithms
“If the computer is finite, as an absolute worst case, all one need do is play out every possible permutation that exhausts the computer’s finite resources and examine the computer’s state at the point those resources are exhausted. Incomputability comes into play when those resources cannot be exhausted.”
This sentence is complete gobbledegook, but I will now try to instill some a meaning in it and expand on that meaning. This may help.
A Universal Turing machine does have certain infinite attributes – time to solve an algorithm, memory to use in storage during the algorithms execution, and unlimited numeric precision to do processing. Turing machines set the limits of what can be achieved by computational (algorithmic) devices – nothing can do any better. A REAL computer, of course, does not have these unlimited capabilities, but it is still possible to address problems by executing algorithms using them, because they are “Turing like”. REAL computers (you call these finite?) can not compute an algorithm that a Turing machine cannot compute. They are subject to the limits of the Turing Computer and quite a few more, because they do not have these “infinite” capacities. Computer Science is largely the effort of mathematicians, programmers and engineers to minimise and mitigate these additional limitations (but they can not eliminate these limitations) The by-product of all this is that REAL computers, most of the time, produce “good enough” solutions to a “good enough” accuracy. Within these parameters results are very highly predictable. But REAL computers, just like Turing Machines, must be allowed to run as long as necessary to complete – to halt.
NOW…. The human brain. This is NOT a computer, but like most physical systems it can be SIMULATED on a computer complex …if we only knew how the brain works and if we have enough computing capacity and memory to do the simulation (very far from easy). The brain is a sort of Finite State machine complex, something lesser than a Universal Turing Machine but with Turing like functionality especially when configured in large networks (not unlike neural ones). BUT BUT BUT the brain does not work with the rigorous standards of computer science imposed… it is HIGHLY subject to all the things I have already mentioned regarding time allowed for solution, data resolution, muti processor synchronism problems and race conditions etc etc etc. The brain “compiles” trial decisions to problems on the fly. All this creates randomisation effects. RANDOMISATION BREAKS CAUSALITY with respect to the inputs from the outside world.
Randomisation is a positive force to trigger trial solutions (e.g. to design) even to CREATE. As I said, it is not unlike the influence of mutation in the process of evolution. And you, as with any religious creationist in ongoing Evolution debates, scoff at the possibility of anything of complexity arising from a basis in randomness. Well it does…. our species and the capabilities that makes our species unique arises from the effects of randomness… quantum randomness to create us, and computational randomness to give us free will.
Cheers
Howie
Once again, logically, randomness can only ever be an input. That the input (if it even exists) is physically contained within the confines of the case (in the case of the computer with a hardware RNG) or the skull (in your hypothesized quantum woo random brain will generator) is irrelevant; they’re still inputs.
No. When randomness is used in the way you describe, all it’s doing is sampling the space of possible options because said space is too large to exhaustively search. Such randomness — including in your example of Evolution — is almost guaranteed to be sub-optimal; only by sheer chance can it happen upon the truly optimal solution an exhaustive search (or a more intelligent, less brute force approach) would otherwise find.
For example, you can approximate π remarkably well by tossing toothpicks at a circle; there’s your random search. You can, at least in principle, do even better with precise measurements; that’d be the brute force search. Or, you can skip to the chase and calculate it far faster and with perfect accuracy to whatever degree of precision you require.
Again with the quantum woo! Though there certainly have been quantum events (such as radioactive decay) in the history of genetic variation of life on Earth, quantum randomness is entirely unnecessary and nearly entirely absent, statistically speaking, from the evolutionary process. It simply isn’t a factor worth noting other than in a footnote.
Next thing, you’re going to come back with Rupert Sheldrake’s morphic fields to explain how all this quantum randomness knows to organize itself into a creative force for will. Please! Spare us!
b&
Gödel’s first Incompleteness Theorem states that in any logical system at least as powerful as Peano Arithmentic (a very simple system) there are theorems that are not provable. The system is said to be incomplete.
Suppose we have a Turing machine simulating brain states. It’s currently in state A and after some computation reaches state B, so we can say that A => B. That’s a theorem, and the simulation represents a proof — a purely logical progression from A to B.
Gödel’s theorem says, clear as day, that there exists a theorem X => Y that is not provable, hence not computable.
Let’s get this out of the way.
Do you assert that there is any aspect of the human mind that is (or might be) infinite in any respect? If so, what?
I don’t think there’s anything productive that can be added to this discussion before you’ve answered.
(For the record, if it’s not yet already painfully obvious, my own answer is an emphatic, “NO!”)
Cheers,
b&
Do you assert that there is any aspect of a Turing machine that is (or might be) infinite in any respect? If so, what?
My own answer is an emphatic, “YES!”
Now are you going to rebut my argument based on Gödel’s Incompleteness Theorem, or not?
My strong suspicion is that you don’t have a clue about incompleteness, and that you’re an amateur mathematician with a little knowledge of a dangerous thing.
We are absolutely agreed that Turing Machines are infinite.
But that’s not the question I asked.
I shall ask it again, and remind you again that I have answered it (in the negative) and already replied to your decidedly different variation on the theme.
Are human minds (possibly) infinite in any respect?
Again, there is no point in continuing this discussion if you choose to emulate Christian trolls when confronted with “difficult” Bible passages and insist on changing the subject rather than answering. I’ve been arguing in good faith. Are you?
Cheers,
b&
“Are human minds (possibly) infinite in any respect?”
I don’t think so. I have no need of that hypothesis. My problem is with your frankly embarrassing and naive concept of computability, and your wanton misuse of mathematics, or rather pop mathematics.
Now are you going to rebut my argument based on Gödel’s Incompleteness Theorem, or not?
I’ll take that as a “no.” If you later feel the need to amend that statement, do please make it explicitly clear.
With your answer in the negative, you’ve already rendered Godel irrelevant.
Here’s why.
Since there’s nothing finite about the mind, then it is therefore finite and therefore enumerable. That, right there, gives us the trivial solution to emulation; a simple, massive lookup table akin to a complete book of all games of Tic-Tac-Toe or any other game with a finite number of legal games. This initial state provided this set of inputs results in this terminal state. Hey-presto, the brain is perfectly computable. Yes, this brute force method is inefficient, but efficiency is irrelevant; the fact that there’s at least one solution tells us that there are an infinite number of equivalent Turing Machines that will do the trick, and one may quite reasonably posit that at least a few will be more efficient than this brute-force approach.
Your first objection, if you follow the pattern, will be that this requires an impractical amount of resources. Raise that and you’ll only continue to demonstrate the incompetence of not understanding the difference between the finite and the infinite and the definition of the Turing Machine. I would hope we’re past that, but I still have my doubts.
Your next objection would likely be more quantum random mind woo. This again is a demonstration of incompetence, for such randomness can only be an input. But, if it really makes you feel better, we’ll go whole-hog and postulate the Many Worlds Interpretation and dramatically expand the number of terminal states to include all results from all possible indeterminate quantum outcomes. And, again, this number is still finite, even if it’s much bigger than before.
The only remaining question is whether the same initial state with the same input will potentially result in more than one output. If so, there are two possibilities. First, and the only reasonable option, is that you’re still not including all the inputs as inputs or you’re otherwise confused about the logical construction of a Turing Machine.
The only other remaining option is incoherent dualistic woo of one form or another.
Cheers,
b&
Why do you have funny looking white question marks inside black diamond shapes replacing every punctuation mark? Am I the only one seeing this?
Ben, your texts are approaching crank level, so I’m bailing for now. Get a grip.
If equating finite with enumerable and therefore computable constitutes crankdom, then wind me up, because that puts me in some damned august company, from Russell through Godel to Turing.
b&
I’m afraid you just don’t understand incompleteness. The hypothetical finiteness of the brain has nothing to do with it.
By the way, speaking of arguing from authority, are you seriously comparing yourself to Russell, Gödel, and Turing?
I see we’re moving into the “IS NOT! IS TOO!” phase.
Incompleteness, simply put, states that there are true statements that any given system will be unable to express. From within that system.
Those limitations apply within that system entirely, and do not somehow magically extend to any other external system that examines it.
For example: “All but God can prove this sentence true.” That’s a brief, poetic — and, of course, therefore, necessarily imprecise — summary of a common proof of the Incompleteness Theorem. You and I have no trouble proving it true, but God cannot. And, of course, “All but Stephen Barnard can prove this sentence true.” I can do it; God could if it weren’t for that niggling little existence problem; but you can’t. “All but Ben Goren can prove this sentence true.” You can do it, but I can’t.
So, are there things the brain is incapable of comprehending? Absolutely, unquestionably.
But, so what? That’s exactly what Gödel tells us we should expect.
Your error lies in assuming that, because there are things that the brain is incapable of comprehending, therefore nothing else is capable of comprehending the brain.
And, sorry, but, yes, that really is a very novice mistrake.
I’ve offered you more ways that something other than the brain could be modeled by something other than the brain itself than I can remember at this point, and yet all your objections keep circling back to your own inadequacy to actually perform the task. But of course the human mind is unlikely to ever comprehend itself; that’s exactly the class of impossibility that keeps popping up in Turing- and Gödel limits!
The question isn’t whether you or I are personally up to the task, or even if some successor to humanity will. The question is whether or not it’s a logically (unrestrained by the material budget of the universe) coherent task. And the answer, overwhelmingly, is, “yes.” Whether through a number of brute force methods or more elegant and sophisticated refinement is irrelevant, too; all are as equivalent to each other as x + y = 0 is to x = y * -1.
So, it should be clear that by dismissing the finiteness of the brain when considering questions of incompleteness, you’re the one to demonstrate your lack of understanding. If the mind is finite — and it most assuredly is — then something also finite sufficiently bigger than the mind can completely map the mind. The mind’s inability to map itself — which would be a demonstration of incompleteness — is entirely irrelevant to the question of the computability of the mind.
I’ve got a bunch of real-world work to do to pay the bills today, so it’s unlikely I’ll have the time to continue your remedial education in information theory. This is likely to be my last substantive post for the day.
Cheers,
b&
P.S. Yes, I compared myself to those three men — but only by including myself with them in the set of people who understand the difference between the finite and the infinite, a set you have repeatedly demonstrated yourself to be excluded from, again in the very post I’m replying to. b&
You talk a great game there Ben. But I think you’re putting a little too much faith in Church, Turing, and algorithmic processing. Touching but misplaced. For instance, this from the Wikipedia article on the thesis:
And this from Penrose’s The Emperor’s New Mind:
I wonder how you’ve fared in disabusing Penrose and Copeland – among a great many other well regarded scientists and philosophers – on those points.
While it is one thing to argue that “Turing Machines are [can be] infinite”, it is quite another to prove that minds are in fact such machines. Which I don’t see that you have done so – certainly didn’t see your proof of that in that article. And, off-hand for another thing, if it takes a thousand years to emulate a thought with a Turing Machine that a human makes in 10 milliseconds then I rather doubt that that would qualify as consciousness or as “free will”. Time being, as they say, of the essence.
If you think that I’ve been arguing for the infinitude of the mind, you have completely misunderstood everything I’ve written, as well as Church-Turing.
Church-Turing does not state that something that can do the couple basic operations needed for Turing completeness is somehow magically capable of transcendent behavior.
Rather, it states that, given enough resources, any two Turing-equivalent devices can replicate each other’s work.
A smartphone is Turing-complete. A supercomputer is Turing-complete. To within the limits of their resources, anything the one can do, so can the other. In this particular case, the supercomputer can trivially emulate anything and everything the smartphone does. The smartphone can do a great deal to emulate much of what the supercomputer does, but it will rapidly exhaust its resources. However, if you provide the smartphone with access to at least as much memory as the supercomputer has (and, really quite a bit more), then the smartphone + insane amounts of memory can, indeed, do everything the supercomputer can…also, of course, assuming it has enough battery life, that its components don’t wear out first, and so on.
Church-Turing simply states that the brain (and anything else capable of computation) is no different. It doesn’t state that the brain actually has access to enough memory to outperform any and all other computational devices, but it does state that any computational device with sufficient resources can emulate the brain.
I am utterly unaware of any “putative non-algorithmic behavior in physical laws.” Indeed, as I keep linking to, the laws underlying the physics of everyday life are completely understood — and they are decidedly algorithmic.
Such language is almost invariably code talk for some form of quantum woo, which is not only equally invariably worng but also guaranteed to be utterly irrelevant.
Cheers,
b&
Copeland and Penrose are both woo peddlers according to Ben Goren, as is anyone who disagrees with him on any point. He has the crackpot tenacity of a true believer.
OK Ben….
That was just one insult too many……
First you said I possess “a quite naïve Hollywood-style misunderstanding of computation.” That is more than just a little grating when received by someone who holds multiple degrees in the subject, who has been a Senior Consultant for the worlds largest computer consultancy, and has headed major software development projects and software development centers.. and takes a pride in that past career.
Then you have compared my counterarguments as equivalent to the fabrications of several notorious and disreputable charlatans.
You are, as Stephen Barnard said of you, “gratuitously insulting”. You throw out these insults while assuming an air of pretended authority in a subject – a level of authority that you in no way deserve. You seem to do it most when you are thrown on the defensive in the debate on a point of fact.
You owe me an apology. You also owe an apology to numerous other posters on this thread at whom you have aimed similar insults … posters who had in fact made very reasonable points of debate.
Jerry has laid an edict here that discussion here should be carried out in a courteous manner and with intellectual integrity …
An apology please….
hkornstein,
Arguing from authority rarely makes one look good in a rational debate. You keep asserting the same woo-ey ideas that our decisions are largely the products of randomness and that this randomness somehow constitutes free will. You are wrong on both counts, and even if you were right on the first you still wouldn’t be right on the second. Feigning offense and demanding your opponent apologize for “insulting” you is no substitute for presenting a coherent argument backed up by evidence and/or logic.
I am not claiming I am right by reason of going to authority Kelton, I have already stated this VERY VERY CLEARLY in a previous post if you bother to have a look. I am only saying that Ben uses insult instead of argument… and I am only one of the posters here to express this. He needs to stop this, it is the worst sort of ad hominem debate and forbidden by Jerry
As for randomness being essential for free will I refer you to a summary of Daniel Dennetts arguments on this requirement which mirror my own arguments http://www.informationphilosopher.com/books/scandal/Dennett.pdf
By quoting Dennett I hope you don’t accuse me of going to authority again…. he just states this requirement very cogently
If you care to inspect the archives here, you’ll find Dennett has done nothing to persuade Jerry nor other incompatibilists, and that his arguments for intention via randomness have been very poorly received, to say the least.
Dennett does a lot of great thinking, but this isn’t one of his best examples.
Cheers,
b&
Ben: “If you care to inspect the archives here, you’ll find Dennett has done nothing to persuade Jerry”
My my, who is jumping on to the argument from authority bandwagon now? And who these other authorities are, who have so “poorly received” Dennett’s views on the subject of free will and its underpinning in randomness I can hardly imagine. From a strictly personal perspective I really must say that I don’t agree with any of Jerry’s arguments on free will and Dennett’s seem to me to be totally compelling. I don’t see why your posit that Jerry’s views should automatically trump Dennett’s should hold sway – especially given Dennett’s specific expertise in the field of both cognitive science and philosophy. Anyway, I would certainly liked to have been a fly-on-the-wall when they drove to that seminar together and argued the issue.
Dennett’s “Two Stage Valerian Model” is just what I have described in my own discussion of Turing Incomputability, the human mind and randomness. In particular, my “Orange Sorting Challenge” that you so studiously made such an effort to sidestep answering, is exactly such a two-stage valerian structure. So I guess Dennett must join me in being a woo peddler along with Deepak Chopra.
As for your very limited apology, I will deal with your future posts in kind, and not hold back from describing EXACTLY what I think of both your arguments and your level of understanding of the subject area.
Again, I don’t have time today for lengthy responses. But, as I’m sure I must have already attempted to explain, your “Orange Sorting Challenge” is merely an exercise in obfuscation, and is also irrelevant. It’s trivial to make any problem arbitrarily complex — complex enough that no human could possibly unravel it.
That’s not the question.
The question is whether the human mind is doing anything that a Turing Machine cannot. And, since I or anybody else can come up with endless ways to brute-force the matter (or your Orange Sorting Challenge, for that matter), it’s instantly obvious that, no matter how clever or confusing the brain, it’s still a subset of the set of Turing Machines.
And a rather small subset, too — our tapes aren’t all that big in the grand scheme of things.
Let me turn it around on you: if you can come up with even one single brute-force method of solving your intractable Orange challenge, no matter how many resources would be required, then you can, yourself, have confidence that it’s a case not of anything radical and supercomputational. And, remember — as you have a tendency to forget — that randomness is not computation. If necessary, take the even-more-computationally-intensive multi-worlds approach and compute the outcomes for each of the random possibilities.
You might want to start with Randall Munroe for inspiration:
http://xkcd.com/505/
Cheers,
b&
“And, remember — as you have a tendency to forget — that randomness is not computation.”
Once again, you display your profound ignorance of computation. Randomness is an essential part of many computations. There is, for example, simulated annealing, which uses a thermodynamic metaphor to solve, approximately, some otherwise intractable optimization problems. There is genetic programming that uses a Darwinian metaphor to design complex systems.
I suggest that you stop embarrassing yourself.
Please stop spreading such nonsense.
Randomness is often an essential INPUT in many forms of computation. It can provide an effective method of sampling a search space too large to effectively encompass with the computing resources at hand. And, of course, when cryptographically-strong randomness isn’t needed, there are many algorithmic methods of deterministically generating sequences that share enough distributional properties of true randomness for many needs.
But randomness itself is NOT computation.
And, yes. This really is introductory textbook-level material. I swear, you’re giving every appearance of doing your utmost to demonstrate your complete ignorance and perfect misunderstanding of the most elementary principles of computing science — like a so-called engineer trying to tell people that an object on top of a shelf somehow doesn’t have more energy than an otherwise-identical one on the floor.
Cheers,
b&
Sorry, but credentials mean nothing. See all those with advanced biology degrees who reject “macroevolution” with perverted Haeckelian parodies, for example. And I’ve been attacking your ideas — yes, naïve ones — and not you.
I wouldn’t be surprised if you’re quite competent, even more than competent, in the duties of your job. But when you make such novice errors as thinking that true randomness is computation, not input…well, what do you expect? It’s no different from a Creationist insisting that random mutations are incapable of creating new information — “not even worng.”
Cheers,
b&
Ben: “But when you make such novice errors as thinking that true randomness is computation, not input…well, what do you expect?”
Just one disproof of many regarding your totally absurd statement (http://en.wikipedia.org/wiki/Indeterminacy_in_concurrent_computation
Exactly who is the novice Ben. I’ve rarely met anyone who claimed so such knowledge of a subject who turned out to be so pig ignorant of it
“Exactly who is the novice Ben. I’ve rarely met anyone who claimed so such knowledge of a subject who turned out to be so pig ignorant of it.”
He should keep his day job playing the trumpet, instead of trumpeting his ridiculous arguments here.
Okay, this thread is getting mean and unenlightening. I’ll ask all of you to please knock it off. If not, I’ll close the thread.
Once again you’ve got it completely umop-apisdn.
The question isn’t whether or not there are things that brains themselves are unable to compute. We know there’re lots of things they can’t compute.
The question is whether there are things they are capable of computing, but that no sufficiently larger computer would be capable of computing.
And since brains are decidedly finite, we know the answer to the second question as well. It may well be (but likely isn’t) impractical to build a computer up to the task, but that’s a limitation of engineering, not of information science.
This might help you. At the bottom of that page, you’ll see a link to “non-deterministic Turing machines.” The section on equivalence opens with a perfect summary of what I’ve been trying to bang into your skull: “In particular, nondeterministic Turing machines are equivalent with deterministic Turing machines. This equivalency refers to what can be computed, as opposed to how quickly.” The rest of the section does a not-miserable job of explaining why and how.
I hate to argue by Wikipedia, considering how unreliable it is (even if it’s vastly superior to the Britannica). But you like doing so, so maybe seeing Wikipedia make the points I’ve been trying to make will finally help you get it through your skull.
Cheers,
b&
“Indeed, accusing us of hypocrisy.”
Ooops… I do see that it can be taken in an insulting context, but it wasn’t meant that way, just a bad choice of wording. Should have just stuck with incoherent. And I would never have used the word “nihilist” as that would also inferred willful intent (and as a compatibilist I do accept that my acts are willful)
Apology offered. Sorry!
Kurt Gödel, before anyone else, and perhaps more than anyone after, understood the limitations of computation. And he proved it.
He proved that any non-trivial logical system is incomplete: There are theorems (true statements) within the system that are not provable.
His trick was to embed within the system a representation of itself. With some rigorous mathematical prestidigitation he proved his case.
My hunch, and it’s only that, is that the self-referential aspect of Gödel’s reasoning is at the heart of the ineffable conscious experience.
So close, and yet….
Just because a particular system itself is incomplete does not automatically mean that all other systems are incapable of encompassing all of said system’s gaps. Those restrictions only apply to that particular system. There may well be some other system that encompasses everything that one lesser system can do, as well as certain tasks of interest that it cannot.
And in the case of finite systems, any sufficiently larger system can completely emulate a smaller system.
The brain is decidedly finite. Construct a computer of sufficiently greater capacity than the brain, and it is perfectly capable of doing all of that which the brain is capable of doing. That which the brain cannot do might or might not be possible for the larger system; the larger system will have its own limits, including the limits that even apply to infinite systems. We’re not asking the larger system to do more than the brain can do, only as much as the brain can do. And, since the brain is not infinite, there’s no magic Gödel spice that prevents some other larger system from emulating it.
Cheers,
b&
Following Jerry’s lead, I will not reply to any of your comments in this thread after this, because it might tempt me into saying something I’ll regret.
Yes… and I’ll follow your very sensible lead Steven, and for the same reason….there’s nothing to be gained by staying on the thread
Not wishing to end my participation on this thread totally on a chord of rancour, I thought instead that it’s better to make some general observations on the wider topics of algorithms, evolution, the human mind, free will and “most importantly” randomness. It’s going to be seen as woo by some, but I’m past caring about that insult. But what I assure you is that what I say will be technically totally correct.
I guess we are all mostly here at WEIT because we are fascinated by Evolution. Well, why not? – evolution made us – it also made the human mind. At least we agree on that one thing -we are all atheists, after all.
As an engineer myself I sometimes marvel at evolution. There’s that old saw – “Orgel’s Second Law” by Francis Crick – “Evolution is cleverer than you are”. And it’s true. An engineer only has to look at the outstanding “designs” in nature (the mind for example) to appreciate that.
And Evolution is an ALGORITHM. It is an endless iterative process (a DO loop) of selecting more adapted solutions and culling out lesser adapted ones in a population of trial solutions – and within this process allowing the merging of individual characteristics of the best solutions together by that other great invention of evolution – sexual reproduction (though asexual works, but far more slowly). When we think about it, evolution is so ridiculously simple and so terribly wasteful.. and yes- it is so randomly based –but it really works.
When an engineer wants to improve on his own design he often uses Evolutionary Algorithms to come up with the solution. He only has to model the dynamics of the environment and set off a population of trial solutions with the various parameters of design and let them “breed”. Aircraft wings, bridges, telephone networks… endless designs better adapted to do a job than an engineer can produce alone. (http://en.wikipedia.org/wiki/Evolutionary_algorithm)
Then there is the human mind. This is a wonder of design (in fact we don’t really understand how it works). It’s not the vast number of “gates” that make it impressive… it’s the architecture. As best we know it’s a hierarchical highly structured multiprocessor array whose various units are themselves highly specialised, passing pre-processed encoded information between themselves in a vast communication network… with an associative memory, capable of supporting both simple and symbolic information. The brain can thereby learn, model, and “self programme”. Symbolic based computation in the mind allows a new form of Evolution itself based on Language – an innate capability we have found in the mind. (Chomsky et al. )
All this from randomness.
Now for free will.
The universe is NOT totally deterministic. There are two elements of randomness at play – quantum indeterminacy and computational indeterminacy. Many processes that we say are determinant are emergent from these random properties – for example chemistry, physics and yes, the process of Evolution (if all chemical reaction was wholly determinant … mutation could not arise to allow the Evolutionary Algorithm to work). In computational theory we also find indeterminacy – Turing describes this and sets the limit of what can be achieved algorithmically. This combines with what we can achieve physically when we build algorithmic entities – be they finite physical computers or finite human minds. Turing incompatibilities, lack of Turing infinities all combine with other randomising effects from timing, concurrency, self-coding etc. etc.
Causality is broken to a degree in these structures and complex behaviours emerge.
Free-will, if it exists, and I say it does requires randomness
Don’t knock randomness…
PS: Apology to Stephan for mixing up his name and writing Steven. I’m slightly dyslexic (another mental source of indeterminism I’m sure)
Peace
“All this from randomness”
Adaptive evolution (which surely built our brains) is a nonrandom process. Mutations are random in the sense that there is no teleological force causing mutations that are more likely to be helpful to occur, and some mechanisms that lead to mutation might be truly random (like those resulting from chemical damage caused by ultraviolet radiation, since a photon interacting with an electron in a nucleotide base is a quantum mechanical event), but to observe the process of adaptive evolution which built us and then to conclude “All this from randomness” makes one resemble a creationist’s straw-man conception of an evolutionist.
“Free-will, if it exists, and I say it does requires randomness
Don’t knock randomness…”
Thanks for admitting that you are merely asserting that free will exists, and also that it requires randomness.
Once again: I am not “knocking” randomness. I have no animosity towards the idea of randomness. I just don’t see how randomness gets you free will. Does a roulette wheel have free will?
1. Randomness breaks causality
2. The “decision algorithm” works over two phases….. The first throws up a number of plausible alternatives( but this process has underlying randomising indeterminancies on which ones they are). The second decides based on decisional criteria which single alternative will be executed
The decider= the self
But as the whole thing is iterative YOU programme your own further processing criteria for further processing
Moral responsibility (within nature and nurture limits) still applies
This is REALLY my last post on this thread
But Dennett is still there to be read and make the case better than I can
PS: my apologies once again to Stephen, this time for writing “Stephan”…
Where in the process you describe does “free will” enter? If you program your future self’s actions, then wasn’t your present self programmed by your past self (along with environmental influences)?
Where in the process you describe does “free will” enter? If you program your future self’s actions, then wasn’t your present self programmed by your past self (along with environmental influences)?”
A very good question.
But you didn’t read my previous answer closely enough.
The answer is COMPOUND INTEREST
I said that processing criteria (we could arbitrarily call this the “will”) is itself processed and refined and redefined on an iterative basis. Even if the randomising effect on redefining the next iteration is a very very small percentage, the overall effect continuously compounds. The “will” becomes less determined by initial conditions and more and more unique…. and the agent that makes that uniqueness must be attributed to the SELF. Randomness is necessary in the process. And of course initial conditions, and nature and nurture, do shape and influence our selves…. remember that a compatibilist doesn’t say that determinism is totally inoperative. Much input that goes into the system (new things learned, experiences etc) only add to the overall mix
but when any other new and novel “two phase decisions” are made the selection criteria is finally attributable to the developed “will” that exists and continues to evolve.
Doesn’t this all make sense? … as we grow up from a child, we become “much our own person” -we have that free will
My answer to all further questions – “read Dennett”
Cheers and Adiós
Howie
1. I highly doubt that randomness compounds in the way you describe. If it did, the effect of randomness on our behaviour would eventually dwarf the effect of the deterministic operation of our brains, and we would cease to have anything resembling a mind at all.
2. The “self” as a single, identifiable entity does not exist. I’m no neuroscientist, but it’s my understanding that the brain is made up of multiple components which communicate with (and sometimes compete with) each other. The “self” which is responsible for an impulsive, fight-or-flight response is not the same “self” which is responsible for deliberate action after conscious reflection. This explains why we can in one moment desire to be in shape, but in the next moment want nothing more than to relax on the couch with a beer. The part of our brain which is responsible for our conscious experience of the world is not necessarily the part which drives all of our decision-making.
3. “Two phase decisions” – I don’t know what you mean by this, and putting it in quotes doesn’t help.
4. Putting phrases in all caps is no substitute for making sense.
I know you’ve said (repeatedly) that you’re done on this thread, but if you’re not I’d appreciate an answer to my previous question: Does a roulette wheel have free will?
If so, you and I are using two very different definitions of free will.
If not, I don’t see how randomness in the brain can generate free will.
Perhaps a clear definition of what free will is – i.e., what would have to be true about the world in order for free will to exist – would help solve this conundrum.
Sorry if my reference to “two phase” processes seemed totally obscure. I left a link to Dennetts (Valerian) explanation far up this thread, but I can imagine it’s easy to miss with so many comments flying about here.
The idea of a two phase process of decision making with a randomising element goes all the way back to William James….
(” My first act of free will shall be to believe in free will” – a bloody brilliant quote I’d say)
and there have been many variants since then. I tend toward Poppers and Dennetts. Here is a link with an overview.. there’s a bit too much background but just keep reading to get into the good stuff …http://www.ucl.ac.uk/~uctytho/DoyleOnDoyle1.pdf
And heres the Dennett two stage:
http://www.informationphilosopher.com/books/scandal/Dennett.pdf
I’d say they addresses the issues in your question pretty well. The randomness source some of these philosophers chose is quantum effect, I myself believe that computability issues are the randomising source.
As for randomness itself, one could easily make the assumption (as you do) that randomness degenerates functionality and complexity. But as with Evolution where random mutation is guided by a selective process (yes – a phase 2) so does the development of free will. Popper is big on that idea.
And no, a roulette wheel have free-will, free-will involves the determinations of the human mind and all its capabilities … especially creativity. We are a self programming entity in part. I have been banging on a lot about it in the Turing wars here.
The debate won’t end here, I’m sure we’ll meet again in debate when Jerry decides to “rattle our chains” again with another free-will teaser…
Cheers
Howie
I’m surprised this is still continuing.
What you repeatedly describe is simply a non-deterministic Turing machine:
http://en.wikipedia.org/wiki/Non-deterministic_Turing_machine
But it’s no more than a mental crutch to make some problems easier to grasp:
Perhaps you should rattle the chains of Wikipedia and / or
instead of WEIT…?
b&
No Ben, you might not have noticed but Kelton and I are now discussing particular points in Philosophy and not Turing machines.
I am still here because I did not want to be rude to Kelton and leave any of his comments hanging unanswered while he continues to make them. I have a lot of respect for Kelton, his general subject knowledge and his courtesy in debate… and therefore I continue to respond.
However, as previously stated, I am not responding to your comments.
The non-existence of free will seems to be self evident to many here. From a strict deterministic materialist perspective, it seems to have no place.
Nevertheless, the feeling of free will is palpable. To say it doesn’t really exist is like saying “Consciousness is an illusion” — a Deepity. At least the *feeling* exists, and where does that fit into materialism?
There are two ways out: Abandon deterministic materialism, or consider that there may be complexities in physics, possibly computational, that we don’t yet understand.
I won’t condider the God solution.
I totally agree with you Stephen. I truly cannot understand how an incompatibilist can say free will does not exist and then go about his/her life acting as if it did. It is more than just hypocrisy – it is total incoherence. More than that, I cannot see how such a person can function to any mental efficiency when they enter their “pretend there’s free-will mode”. If I were an incompatibilist I would just wake up in the morning and think “why the hell even get out of bed?”. Well why indeed?, if he stays in bed his decision was predetermined and he is fully justified. But NO, goes his thinking, “I must return to the real world and go to work etc etc etc) This popping in and out of this “why-bother… oops, no I’ll pretend now instead now” must occur all day long. But the very existence of this pop-in-pop-out thinking must necessarily alter his behaviour, if for only the reason that it takes up time of allotted to decision making. The effect – his “pretending mode” is not the equivalent of the REAL mode of the decision process of the person he’s pretending to be –someone who believes he has free-will. It’s a mental and philosophical quagmire.
You are absolutely right Stephen, I think that at the bottom many incompatibilists here take their stance for emotional reasons because they don’t like the religious historic attachment to the idea, and they don’t like religion. Or they don’t like allocation of “blame” in deciding about punishment. Or they think that to be “scientific” they must accept Laplace’s Demon at face value.
But they don’t have to do any of this, they can still hold their attitudes – by becoming a compatibilist.
Of course, with philosophy, you can spin enough of a story that you can justify anything. But with science you really must stick with empirical evidence. You don’t think that all that evidence is quite here yet, is that what you’re saying?
This is insulting to those of us who are incompatibilists. Indeed, accusing us of hypocrisy. We’re programmed, perhaps, to think we can make free choices, so we’re not nihlilists. If you can’t understand that, then it’s a shame.
Anyway, your language is an insult to not just your host, but every incompatibilist here. Apologize, please.
“Indeed, accusing us of hypocrisy.”
Ooops… I do see that it can be taken in an insulting context, but it wasn’t meant that way, just a bad choice of wording. Should have just stuck with incoherent. And I would never have used the word “nihilist” as that would also inferred willful intent (and as a compatibilist I do accept that my acts are willful)
Apology offered. Sorry!
BTW… I would say that apologies are due from Ben… I, and several other posters here have been very seriously and willfully insulted many many times with personal attacks without any such censure, and with no apologies offered when I requested one. It was this behaviour that led to things getting out of hand here.
hkornstein: Do you think it is hypocritical to say “dreams aren’t real” and then to go about continuing to dream at night? Because I suspect you both experience them and recognize that they are illusions.
Does it make life difficult for you to recognize that matter is comprised mostly of empty void and that what seems solid isn’t really so?
“Do you think it is hypocritical to say “dreams aren’t real” and then to go about continuing to dream at night? Because I suspect you both experience them and recognize that they are illusions.”
Interesting point!
But as a rational person I know that my dreams are not real, and I act on that knowledge – I ignore my dreams and don’t call in a soothsayer so that I can act on their “meaning”.
Now if I believe that my actions are predetermined and that any thing that I do was going to happen anyway…. and that my realisation of how things work is not an illusion – but true, that’s different – how as a rationalist can I not alter my behaviour to suit this fact? How can it not affect my actions? When an incompatibilist says that penalties for prison should be changed he is ACTING on his belief. But then how can he say that his getting out of bed in the morning is just an effect of thinking he has free will and NOT thinking of not altering his behaviour in some way. It’s inconsistent.
Howie
+1
It is a false choice to say that the alternative to calling soothsayers is to ignore dreams. I rather enjoy mine, usually, and don’t ignore them any more than I ignore movies I see even though I know that they are (usually) fictional.
It is perfectly reasonable for someone to understand that their experience of “free will” is illusory and still experience the illusion.
Yes, I accept that “self” is an illusion & yet I walk around as if this Diana person is one coherent thing.
Well, I think it is in fact “one coherent thing” – as is the case for all of us. And, speaking of such things, you may wish to peruse the Wikipedia article on emergence.
And while I will readily agree that it certainly appears that we’re not much less ephermal than mayflies – coherence not being a guarantee of longevity, I also think it borders on the criminal that so many “incompatibilists” insist on a position that seems not much more than a conjecture at best, and “articles of faith” at worst.
Ooops: “… not much less ephemeral than mayflies …”
I don’t think the feeling of free will necessarily implies the existence of free will, but the feeling of free will still requires an explanation that seems to be outside the scope of materialism as we think we understand it. It exists, it isn’t material (although it must somehow be embodied in the material), and from a deterministic materialist perspective seems entirely superfluous, as does all conscious experience.
I fail to see why all matters of the mind are superfluous according to deterministic materialism.
My beef with the term free will is that it doesn’t really clarify what we are talking about and that definitions vary from person to person.
Consciousness isn’t dependant on the existence of free will, and I simply don’t see why the brain, like the rest of the body, shouldn’t be purely materialistic in its functions and properties.
There may be tons of things about how brains work that we haven’t discovered yet, but how that leads to the conclusion that free will exists seems to be a very subjective matter, for now at least.
Let’s wait and see what turns up before we declare the existence of free will as a fact.
“I fail to see why all matters of the mind are superfluous according to deterministic materialism.”
Qualia, feelings, emotions, and in general all conscious experience seem clearly immaterial to me. They can’t be measured or observed by anyone other than the one experiencing them. They are often called epiphenomena, and perhaps they are, but that begs the question, merely putting a symbol on something that exists but is so far inexplicable.
The best we can do is to correlate them, imperfectly, with brain states, which is not satisfying. Why shouldn’t the brain states exist as states of deterministic material systems without the extra, nonmaterial, nonmeasureable subjective conscious experience?
Well, nothing about the human experience seems immaterial to me, so I guess we don’t quite see eye-to-eye on that one.
I’m perfectly fine with accepting the brain as a purely materialistic system( I thought that was pretty clear ), but I also think there’s plenty of things about that system and how it works that we don’t know yet. For example, at what point does the brain and body develop consciousness and how can we scientifically measure and determine those processes.
I don’t see why science shouldn’t, somewhere down the line, be able to explain those processes in much greater detail by accumulating much more knowledge, along with the development of better technology that will allow us to peer deeper into the workings of our bodies and brains.
It may be naive, but to me the sky is the limit, so to speak.
If I ever believed in theological free will, I was too young to retain the memory — or substance abuse obliterated the brain cells which stored it. It is, and has always been, just as preposterous as any other teleological claim.
Free will in a philosophical sense remains an ongoing debate that originated in text approximately at midpoint first millennium BCE. On it continues to rage. If resolution to it is near, for my part it can’t arrive soon enough. Or not. When I ignore it, no noticeable absence plagues me.
Free will as a legal term is located (among numerous sites) in the (unanimous) majority opinion in Morrissette v US 1952, the precedent case establishing mental diminishment standards and is the extant hurdle in the field for incompatible claims to overcome.
Free will in colloquial parlance has precedent originating ? (is it farfetched to expect its origin in thought predates writing?)
Is the present incompatiblist position regarding traditional legal & colloquial concepts of free will sufficiently evidence based to expect the deletion of the term “free will” from the legal realm? It is necessary for this to occur, I would think, prior to the extremely more problematic disappearance of the term from popular lexicon, is it not?
‘Let’s wait and see what turns up before we declare the existence of free will as a fact.’
Well, okay, but isn’t it most probable the corollary must be demonstrated with a higher standard of evidence than presently proposed before it is possible for the non-existence of free will — as normative legal and colloquial terms — to be declared as fact?
If the legal system is dependant on a definition of free will that may or may not violate the laws of physics, then maybe we should reconsider our approach to the concept of guilt and justice.
In other words; I do not fear that anarchy will prevail if we discover and prove beyond doubt that free will is an illusion. The jury is still out and maybe that’s a good thing if our societies indeed are houses of cards.
And as a side note; I find it slightly ironic that those who don’t consider free will as reality are expected to prove its non-existence beyond doubt. Again, what definition of free will is it exactly we are supposed to disprove?
It may seem reckless and irresponsible, but I prefer an unpleasant truth rather than a useful lie.
hkornstein,
Sorry for the delayed response. I’ve been a bit busy.
First, I’d like to say that I appreciate that you’ve enjoyed our exchange and that you value what I have to say. There are few things more satisfying than a fruitful meeting of the minds. I’ve also refined my thinking from this argument.
Anyway, I believe that in the above post you are confusing fatalism with determinism when you wonder how an incompatibilist would choose to get out of bed in the morning without temporarily believing in free will.
The fact is that I get out of bed in the morning not because I momentarily tell myself I have free will, but because I have work to do and things to experience and if I don’t get out of bed I won’t get to do those things. As Sam Harris points out in his enlightening little book “Free Will”, staying in bed is also a choice, and one that will require increasingly heroic efforts to maintain. If you stay in bed all day, you will be assailed by impulses and desires to get out of bed and do something. You will get hungry and/or restless. If you live with another person they will probably encourage you to get up and make something of your day.
The point isn’t how long you can choose to stay in bed; the point is that the whole time, your desire to stay in bed is competing with conflicting desires to get out of bed, and at some point (or never) the balance tips and your desire to get out of bed overrides your desire to stay in. There’s a wonderful poem cited in The Devil’s Dictionary by Ambrose Bierce (I don’t know who actually wrote the poem) along with the definition of decide which I think might illustrate my point:
A leaf was riven from a tree.
“I mean to fall to earth”, said he.
The east wind, rising, made him veer.
“Westward”, said he, “I now shall steer.”
The west wind rose with greater force.
“‘Twere wise to change my course!”
With equal power they contend.
“My judgement I suspend!”
The wind died down, the leaf, elate, cried:
“I’ve decided to fall straight!”
First thoughts are best? That’s not the moral.
Just choose your own and we’ll not quarrel.
For however your choice may chance to fall,
You’ll have no hand in it at all.
The fact that our choices are determined by the laws of physics, chemistry, and biology does not mean that our choices are meaningless, or that they lack consequences. Whether humanity has a long-term future among the stars depends very much on the actions of people living today. So if you want the future to be better – for yourself, your friends and family, or your species – it’s best to act like the future will be better by doing the things that will make it so.
KELTON: “First, I’d like to say that I appreciate that you’ve enjoyed our exchange and that you value what I have to say. There are few things more satisfying than a fruitful meeting of the minds. I’ve also refined my thinking from this argument.”
The same goes for me Kelton. I think most of us come here to WEIT to learn something new or to test our present understanding of things- to clarify that understanding. At it’s best, that’s what happens here.
As for the “getting up in the morning” issue and incompatibilism I realise my argument is a bit loaded… that incompatibilitsts don’t really think about their predetermined behaviours all the time, and alter their behaviours accordingly. But I cannot see how holding such a position cannot fundamentally alter the overall behaviour a person has and in effect the type of person that they are.
KELTON: “Anyway, I believe that in the above post you are confusing fatalism with determinism”
Well, the question is whether you really avoid seeing the world in a fatalistic fashion if you are an incompatibilist. Of course you can say that you are hard-wired to not be fatalistic. But what does that really mean? – “hard wired” itself means that you see yourself as a sort of robot, but one that has the realisation of its true condition in nature somehow “turned off”. And saying what we “feel” about a situation is more relevant than what we ARE in a situation is a rather strange stance for a rationalist. Or instead, perhaps you can say that you are hard wired to ACT as though you had free will. But again, this only means that you are functioning on a false premise – that your outlook is fundamentally incoherent.
KELTON: “….our choices are determined by the laws of physics, chemistry, and biology does not mean that our choices are meaningless, or that they lack consequences”
Indeed –consequences occur. But it seems to me to be viewing one’s own actions as part of some inevitable whole – on some unalterable, predestined wheel of life. In a deteterministic world where is REAL choice?, where is creativity? Where is the chance to be personally heroic? A determinist stance is a fundamental world view, a view of WHAT we are.
I don’t see how it is necessary to be an incompatibilist to take an enlightened stance against punishment, or against religious dogma, or to be a rationalist. Therefore the question that needs to then be asked is “given that the jury is still over these issues, is this alteration of attitudes, decisions, and behaviours really a good stance to maintain until the verdict is in?” I believe in the scientific view, so if strong emperical proof is produved that free-will doesn’t exist I’ll change my view on the matter. I have been accused of arguing from authority quite a bit, and I must say that this is a valid criticism of some of the things I have been saying. But on the other hand, when so many prominent philosophers, cognitive scientists and mathematicians argue strongly that free will does exist, and that their arguments are really very very cogent, why act as though the case was decided in the opposite direction?
Thanks for the very charming poem. In return I’ll close with a poem I also like, which is more of a compatibilist inclination:
Every choice you make
Right or wrong
Good or Bad
Yes or no
You take it
What is it you ask
A chance my friend
You can take it
You can give it
But its not certain to come your way
A chance to make things right
A chance to make things wrong
A chance can give it all
A chance can take it all
Remember
They are not given because you want them
They are given because you need them
They are not taken because you need them
They are taken because you want them
Never second guess or take for granted
A chance that comes your way
hkornstein,
“But I cannot see how holding such a position cannot fundamentally alter the overall behaviour a person has and in effect the type of person that they are.”
Actually, being honest with yourself about the nonexistence of free will does alter a person’s behaviour to some degree. As Sam Harris points out in “Free Will”, realizing that other people are not, in a deep sense, responsible for their actions can actually make a person more compassionate and forgiving, and less quick to judge others. It also counters the ridiculous notion popular among Randians and libertarians that people who are down on their luck financially deserve it because they were lazy, unintelligent, etc.
The realization that people aren’t responsible for their actions doesn’t remove the pragmatic need to lock up dangerous criminals, but it does remove the logic of killing people as punishment for their crimes.
“Well, the question is whether you really avoid seeing the world in a fatalistic fashion if you are an incompatibilist…”
I’m not sure what you’re talking about in this paragraph. Let me just clarify that fatalism is the belief that certain events (such as Arthur becoming the King of England) are guaranteed by some mystical force or being to happen, no matter what anyone chooses to do. In other words, on fatalism there are multiple free paths people can take to their destiny, but their destiny is inevitable. Thor is destined to fight the Midgard Serpent regardless of the choices he makes beforehand.
Determinism, on the other hand, is the belief that the present state of the Universe is determined by the prior state and the laws of physics. On determinism, it’s not that I was destined to write this post no matter what I did leading up to it; it’s that at every point in time, my actions were determined by prior causes. This chain of causality goes all the way back to before my birth, so in what sense am I responsible for any of it?
“I don’t see how it is necessary to be an incompatibilist to take an enlightened stance against punishment, or against religious dogma, or to be a rationalist.”
It’s actually the other way around – being a consistent rationalist should lead to the rejection of compatibilism, which is the belief that we have free will even though our actions are 100% determined by prior causes and/or random chance.
“given that the jury is still over these issues, is this alteration of attitudes, decisions, and behaviours really a good stance to maintain until the verdict is in?”
The verdict will never be sufficiently in for those who wish to keep on believing in free will. Arguing that we should keep on pretending we have free will until the workings of the brain have been fully described is like arguing that we should keep on pretending there’s a god until the workings of the cosmos have been fully described. There will always be a hole, however small, for a God – or a free will – of the gaps to hide in.
I’ve always thought it was absurd to call self or consciousness an illusion. If it’s an illusion, it’s an illusion being had by the conscious self. Do you see the problem?
Consciousness is not an illusion. Free will is.
The conceits of the legal system are irrelevant to the scientific problem of explaining the apparently immaterial aspect of conscious experience.
KELTON: “Let me just clarify that fatalism is the belief….. etc.”
Well, I see little difference between things ending up the same no matter what you do, or things ending up the same because your actions were predestined (i.e. your actions were merely the outcome of some unalterable causal chain). Both situations seem equally depressing to me, the later alternative perhaps a bit worse because, in that case, you’re a mere automaton while in the former situation you can at lease act somewhat heroically. The Stanford Encyclopaedia of philosophy includes “the view that we are powerless to do anything other than what we actually do” as fatalistic, which is pretty much how I see the incompatibilistic view of the world.
KELTON: “As Sam Harris points out in “Free Will”, realizing that other people are not, in a deep sense, responsible for their actions can actually make a person more compassionate and forgiving, and less quick to judge others. It also counters the ridiculous notion popular among Randians and libertarians that people who are down on their luck financially deserve it because they were lazy, unintelligent, etc.”
I don’t see being more compassionate because of some predetermined causality is anything to be proud of, while being more compassionate because we have freely decided that compassion is a morally better course of action really does merit my respect. And it’s wrong to consider compatibilists as necessarily libertarian in their approach to moral values. This whole subject of free-will seems to me to be riddled with political overtones that it absolutely does not deserve. It’s an issue of cognitive science and nothing else – it’s not an issue of Republican leanings versus Democrat leanings or anything like that.
KELTON: “The verdict will never be sufficiently in for those who wish to keep on believing in free will.”
I “believe in” free will Kelton, because I find the arguments in it’s favour totally persuasive. I think it’s an accurate model of reality based on sound mathematical and computational premises – e.g. the two-stage arguments of both philosophers of science and cognitive scientists. I believe that causality is broken when it comes to evolved agents such as ourselves because of randomness – quantum, computational, or environmental (take your pick) and because we programme our own natures as agents, to a certain degree. Thus we are free, for better or for worse. I find Sam Harris’s arguments especially weak – his examples of “unconscious sources of alternatives” being far too trivial and incomplete, his referencing delays in our being conscious of processed choice lacking in understanding of computational necessity. Nope –I’m not convinced – but for very good reasons. You’ve done a lot better than Sam Harris in testing my arguments Kelton, and for that I thank you.
“Both situations seem equally depressing to me…”
It doesn’t matter what you think is depressing. It matters what’s actually true. The fact that you find a lack of free will depressing does not constitute evidence for free will, any more than the fact that many people find the idea of a godless universe depressing constitutes evidence for the existence of a deity. From my point of view, since free will doesn’t exist, we don’t have it even if we think we do. But if we refuse to acknowledge the truth, then we’re twice imprisoned – once the sum of all the genetic and environmental influences which determine our actions, and again by our own self-deception. I’d rather be free of the latter, even if I can’t be free of the former.
“I don’t see being more compassionate because of some predetermined causality is anything to be proud of, while being more compassionate because we have freely decided that compassion is a morally better course of action really does merit my respect.”
The purpose of morality is not so you can feel proud of how compassionate you are – that’s more like magnanimity. All we need to establish a moral dimension to our actions is to notice that suffering and happiness exist, and that there are better and worse ways to decrease suffering and promote happiness in this world. Also, we don’t “freely” decide moral truths any more than we freely decide historical or scientific truths. To the extent that we use reason to decide our opinions on morality, we are allowing our decisions to be constrained by evidence, and to the extent that we appeal to tradition or personal biases, we are constrained by those influences. Free will is not a prerequisite for morality, any more than the existence of Zeus or Yahweh is.
“And it’s wrong to consider compatibilists as necessarily libertarian in their approach to moral values. This whole subject of free-will seems to me to be riddled with political overtones that it absolutely does not deserve.”
I didn’t say that compatibilists were necessarily libertarian. In fact, I imagine that most libertarians believe in libertarian free will, and are therefore non-compatibilist. But it is wrong not to notice that conservatives in the U.S. have made a religious fetish of free will and self-reliance to a much greater extent than liberals have. This is why this debate is not merely academic. Acknowledging the illusory nature of free will would have real-world consequences, which I believe would be good on balance. It is the belief in free will which justifies capital punishment and the huge level of income disparity in the U.S.
“I “believe in” free will Kelton, because I find the arguments in it’s favour totally persuasive. I think it’s an accurate model of reality…”
I abandoned my (deeply cherished) belief in free will because I realized that it is an incoherent non-model of reality which is not compatible with any conceivable world. It is at this point that I realize I’m not sure whether I’m talking to a compatiblist or a believer in libertarian free will. If you’re a compatibilist (as you seemed to indicate in the last post), then you and I both agree that our actions are fully constrained by our circumstances and chance, but you choose to call this truth “free will” while I do not. If you believe in libertarian free will (as your arguments about non-computibility might indicate), then you have the task of coherently defining free will, and explaining how we can get behind the causes of our choices and control them, and then how we can get behind the causes of the choices we make to control our choices, and so on ad infinitum. It can’t be done.
Also, give Sam Harris more credit. I find his arguments in Free Will and the Moral Landscape compelling, and I certainly haven’t presented anything that he didn’t describe more eloquently in his books.
Well, I hope nobody minds this little diversion from philosophy back into the computational world – it has to do with “two phase” decisions. The idea is really crucially important in philosophy too, certainly all the philosophers that I am convinced by on the subject of free will (Dennett, Popper, William James, Martin Heisenberg ) use it as a means to describe the mechanism behind free will. The arguments seem both very solid and personally appealing to me given that they have a beautifully clean mathematical and computational precision to them.
Anyway.
Two Phase is nothing more than doing a process by executing it in two specific sequential parts (or places). For example, in a computer system it is having a solution handled by two separate processing systems where the first hands “choices” to the second or it may be only ONE processor executing two different algorithms in succession. Maybe you’ve heard of “Two Phase Commit” a process used to decide whether to “drop the bomb” (BTW- it’s used in a lot more applications than ending the world). A much wider use of this technique is in search and optimisation, and yes – this really has a bearing on free will. So why are all these compatibilist philosophers finding two phase essential to their stance? And why do they claim that randomness is essential to free will?
Well let me describe the detail.
Phase 1 involves randomness, and the generation of solution candidates. The randomness throws up different alternatives (which themselves could possibly be randomly selected) Randomness can even provide “mutations” of existing alternatives.
If you want to think of it that way Phase 1 is a way of throwing up options in a possible “solution space”. You can see why this is becoming very Darwinian.
Phase 2 is the selection from the alternatives. It is principally a deterministic process looking at a particular set of criteria. (in evolution the criteria is called Natural Selection). Phase two provides “guided randomness” in that it fits a filter to the quality of proposed solutions (in optimisation the we say this leads to Hill climbing in the solution space)
Now what does this all have to do with free will? Many (myself included) compatibilists argue this is the mechanism of human thought. And like all algorithms it is an iterative process. The solutions produced are endlessly recycled providing new and altered solutions and options. We also actually reprocess the “Phase Two selection criteria” itself in that way. What is the processing entity? – the individual human mind – (the two phases can be either sequential, of different sub-processors in the brain – I favour the two processor idea)
Most important: CAUSALITY IS BROKEN by randomness. (Without randomness EVERYTHING is deterministic) The randomness in Phase 2 can arise from quantum indeterminacy, computational indeterminacy, or non problem-specific triggers or perhaps a combination of all three.
This does NOT mean we are totally freed of initial conditions or new inputs – No, nature and nurture have their effects. But what we are is NOT totally causal, we programme ourselves, in a two phase fashion in our own solution space . Two-phase leads to a freedom – executed by the self – i.e. free will.
Anyhow, if anyone can destroy this model of how free will arises a lot of us compatibilists will be crying in our beer.
Howie
I’m really quite surprised you’re still banging on this so long after Jerry asked it be put to rest.
But, to your closing remark:
Your model for probabilistic computation is exactly that of a Non-Deterministic Turing Machine. Non-Deterministic Turing Machines are perfectly equivalent to regular Deterministic Turing machines. The only significant difference between the two is the efficiency with which problems are solved, with each having areas where the one has the advantage over the other. Right tool for the job, and all that, but either will get the job done eventually.
I have repeatedly linked to and quoted from the relevant introductory Wikipedia article in this thread; I shan’t do so again. You haven’t even pretended to acknowledge the fact that said article exists, let alone attempted to address it.
Nevertheless, your first phase is therefore equivalent to a regular, plain-Jane Turing device. And I do believe that you would already be comfortable equating your second phase similarly. We are therefore left with one Turing device feeding its output into the input of another Turing device, and it is another introductory-level principle of computer science that two such devices are equivalent to some other third single device, no matter how complex the interconnectedness.
While your “two-phase indeterminate computation” is unquestionably an efficient approach for many practical applications, it is perfectly logically equivalent with a single, albeit perhaps less efficient, regular old unsurprising Turing Machine.
Therefore, either your free will is not a property unique to your “two-phase indeterminate computation devices” and is also shared by an equivalent Turing Machine — a proposition which makes it not free in any reasonably relevant sense I can imagine — or your will is more woo than will.
Once again, all you’re doing is mistraking obfuscation for magic. You have chosen textbook examples of things that are trivially demonstrated equivalent in introductory-level classes, and, because you apparently don’t understand their equivalence, you’re assuming that profound mystery must lie in that which you don’t understand.
You compared your model with Evolution, and there are certain parallels. Yet the whole point of Evolution is that it is an unguided, unintelligent process that produces an illusion of intelligent design; in contrast, you have come to the equivalent of the Creationist conclusion: that a mysterious and ill-defined force of conscious intention magically imbues and guides the entirely mundane naturalistic processes of cognition.
Just as no biologist would serious consider any proposition that random mutation is the essential driving force behind evolution — it’s the selection that’s what matters, and the mutations don’t even have to be especially random — no computer scientist would seriously consider any proposition that randomness is the essential driving force behind computation. Randomness can help sample an impractically large search space, winnowing down the options to something more manageable, but it’s the actual computation that matters — and, in most cases, the randomness doesn’t even have to be all that random. Indeed, the special emphasis you keep putting on quantum randomness I personally find indistinguishable from the Creationists who insist that that’s how Jesus guided the evolution of, in particular, humans. Quantum mechanics may be weird and wacky, but it’s not some magical mystical faery dust that imbues the universe with conscious interconnectedness. (Unless, of course, you’re trapped in a Chopra novel.)
It’s also worth noting that, in both evolution and computation, the ideal best results come from not a random sampling, but a thorough census. We see how the randomness of evolution often results in sub-optimal solutions; just look at the recurrent pharyngeal nerve. The same happens in computation, especially your two-phase proposal. It’s guaranteed that many options will not be explored by your model, and it’s only a matter of luck if the ideal option is ever even considered. That is, the choice you most would want to make if you had all the options available to you might be something that you’re never even aware of because it was randomly overlooked. And we’re supposed to believe that freedom and / or intention lies in (the equivalent of ) periodically randomly closing your eyes and chanting “LALALALA I CAN’T HEAR YOU” because of input overload?
If your beer isn’t now being diluted with tears, your only hope is remedial information theory classes.
Cheers,
b&
BEN: “I have repeatedly linked to and quoted from the relevant introductory Wikipedia article in this thread; I shan’t do so again. You haven’t even pretended to acknowledge the fact that said article exists, let alone attempted to address it.”
Ben, I have not addressed this point that you have raised about non-deterministic Turing principally because I thought it totally irrelevant. But I will now as you have asked twice………
What is the difference between a Non-Deterministic Turing Machine (NTM) and a normal (deterministic) Turing Machine(DTM)? The answer -a NTM from any practical perspective, is indeterminate almost all of the time and a DTM is indeterminate only sometimes. So if you want to talk about NTMs you only support my case for indeterminacy, not break it down. Now if you had closely read the Wiki article that you quoted, you would have noted the sentence “However, it is possible to simulate NTMs with DTMs”. So…. there is no point in raising the issue of the NTM, as the DTM covers both cases and keeps things ever so very much simpler (endless branches etc.).
Ultimately non-computability in a “perfect machine” is not a problem of the machine, it is a problem of the nature of mathematics, and more specifically it is generally reflected in algorithmic solution. You cannot prevent indeterminacy from happening. Turing Machines are a computing IDEAL though they may seem primitive, and no computer can do better. But real-world computers do worse with respect to indeterminacy, for reasons I have tediously mentioned – lack of infinite resource or solution time being a rather important cause. Your argument that all randomness can be relegated to inputs (which is false, by the way) would still allow causation to be broken and thereby still allow the possibility of free will. And finally, if NONE of this were true(but it IS true), the fact that the human mind is in part self-programming, creates the situation where program error or inconsistency can itself cause randomising effects. If we can’t create algorithmic code that is always free of error when we are conscious and trying, how can you expect us do so when the process is unconscious? This self programming is exactly what Robert Kane describes as “Self-Forming Actions” in his particular version of the two-stage model, and combined with randomness encompasses the mechanism of free will.
If all this were not enough, even if Turing machines were somehow “perfect computers” (and they are not) how can you posit that the brain is a perfect computer also… is the selection pressure of Evolution guided to produce good computer architecture? ( a creationist in this ridiculous case would claim that god is actually John Von Neumann).
As I originally said Ben, you had better stick to philosophical argument when trying to argue against free will, you are on a hiding to nothing if you try to evoke computational theory to attempt to make your case.
BEN “ Once again, all you’re doing is mistraking obfuscation for magic. You have chosen textbook examples of things that are trivially demonstrated equivalent in introductory-level classes”
Well, if my thinking is so simplistic, it’s still rather nice to be in the company of such other uneducated fellows as Popper, Dennett, Penrose, Wolfram and Martin Heisenberg. I somehow seem to have fallen into the trap of accepting the arguments of a major cognitive scientist like Heisenberg, two leading philosophers of science and cognition, and several leading mathematicians instead of your claims. I suppose I’m just a sucker for authority.
BEN: “That is, the choice you most would want to make if you had all the options available to you might be something that you’re never even aware of because it was randomly overlooked.”
Well haven’t you ever overlooked doing something you should have done or should have thought? Most people do that all the time. I never said that two-step in the case of free will was an optimising process – it’s not. Otherwise we’d all be perfect, and we’re not that, are we Ben? Reread what I said. I said that two step solutions were common in computing, one example being optimisation, one two phase commit, etc etc. I said they were also common in nature, one other example being Evolution.
“If your beer isn’t now being diluted with tears, your only hope is remedial information theory classes.”
Actually Ben we haven’t been talking about Information Theory here at all, we have been talking about Computational Theory. And perhaps we should let the other readers here decide exactly who needs the remedial classes in either of these two subjects.
Cheers Ben
Howie