Friday, July 19, 2013

You probably have no free will. But don’t worry about it.

Railroad tracks. Image source.
Anybody who believes in reductionism and that the standard model of particle physics is correct to excellent precision must come to the conclusion that free will is an illusion. Alas, denial of this conclusion is widely spread, documented in many attempts to redefine “free will” so that it can somehow be accommodated in our theories. I find it quite amusing to watch otherwise sensible physicists trying to wriggle out of the consequences of their own theories. We previously discussed Sean Carroll’s attempt, and now Carlo Rovelli has offered his thoughts on free will in the context of modern physics in a recent Edge essay.

Free will can only exist if there are different possible futures and you are able to influence which one becomes reality. This necessitates to begin with that there are different possible futures. In a deterministic theory, like all our classical theories, this just isn’t the case - there’s only one future, period. The realization that classically the future is fully determined by the presence goes back at least to Laplace and it’s still as true today as it was then.

Quantum mechanics in the standard interpretation has an indeterministic element that is a popular hiding place for free will. But quantum mechanical indeterminism is fundamentally random (as opposed to random by lack of knowledge). It doesn’t matter how you define “you” (in the simplest case, think of a subsystem of the universe), “you” won’t be able to influence the future because nothing can. Quantum indeterminism is not influenced by anything, and what kind of decision making is that?

Another popular hiding place for free will is chaos. Yes, many systems in nature are chaotic and possibly the human mind has chaotic elements to it. In chaotic systems, even smallest mistakes in knowledge about the present will lead to large errors in the future. These systems rapidly become unpredictable because in practice measurement always contains small mistakes and uncertainties. But in principle chaos is entirely deterministic. There’s still only one future. It’s just that chaotic behavior spoils predictability in practice.

That brings us to what seems to me like the most common free will mirage, the argument that it is difficult if not impossible to make predictions about human behavior. Free will, in this interpretation, is that nobody, possibly not even you yourself, can tell in advance what you will do. That sounds good but is intellectually deeply unsatisfactory.

To begin with, it isn’t at all clear that it’s impossible to predict human behavior, it’s just presently not possible. Since ten thousand years ago people couldn’t predict lunar eclipses, this would mean the moon must have had free will back then. And who or what makes the prediction anyway? If no human can predict your behavior, but a computer cluster of an advanced alien civilization could, would you have free will? Would it disappear if the computer is switched on?

And be that as it may, these distractions don’t change anything about the fact that “you” didn’t have any influence on what happens in the future whether or not somebody else knew what you’ll do. Your brain is still but a machine acting on input to produce output. The evaluation of different courses of action can plausibly be interpreted as “making a choice,” but there’s no freedom to the result. This internal computation that evaluates the results of actions might be impossible to predict indeed, but this is clearly an illusion of freedom.

To add another buzzword, it also doesn’t help to refer to free will as an “emergent property”. Free will almost certainly is an emergent property, unless you want to argue that elementary particles also have free will. But emergent properties of deterministic systems still behave deterministically. In principle, you could do without the “emergent” variables and use the fundamental ones, describing eg the human brain in terms of the standard model. It’s just not very practical. So appealing to emergence doesn’t help, it just adds a layer of confusion.

Rovelli in his essay now offers a new free will argument that is interesting.

First he argues that free will can be executed when external contraints on choice are absent. He doesn’t explain what he means with “external constraints” though and I’m left somewhat confused about this. Is for example, alcohol intoxication a constraint that’s “external” to your decision making unit? Is your DNA an external constraint? Is a past event that induced stress trauma an external constraint? Be that as it may, this part of Rovelli’s argument is a rephrasing of the idea that free will means it isn’t possible to predict what course of action a person will take from exclusively external observation. As we’ve seen above, this still doesn’t mean there are different future options for you to choose from, it just means prediction is difficult.

Then Rovelli alludes to the above mentioned idea that free will is “emergent”, but he does so with a new twist. He argues that “mental states” are macroscopic and can be realized by many different microscopic arrangements. If one just uses the information in the mental states - which is what you might experience as “yourself” - then the outcome of your decisions might not be fully determined. Indeed, that might be so. But it’s like saying if you describe a forest as a lot of trees and disregard the details, then you’ll fail to predict an impending pest infection. Which brings us to the question whether forests have free will. And once again, failure to predict by disregarding part of the degrees of freedom in the initial state doesn’t open any future options.

In summary, according to our best present theories of nature humans don’t have free will in the sense explained above, which in my opinion is the only sensible meaning of the phrase. Now you could dismiss this and claim there must be something about nature then that these theories don’t correctly describe and that’s where free will hides. But that’s like claiming god hides there. It might be possible to construct theories of nature that allow for free will, as I suggested here, but we presently have absolutely zero evidence that this is how nature works. For all we know, there just is no free will.

But don’t worry.

People are afraid of the absence of free will not because it’s an actual threat to well-being, but because it’s a thought alien to our self-perception. Most people experience the process of evaluating possible courses of action not as a computation, but as making a decision. This raises the fear that if they don’t have free will they can no longer make decisions. Of course that’s wrong.

To begin with, if there’s no free will, there has never been free will, and if you’ve had a pleasant and happy life so far there is really no reason why this should change. But besides this, you still make your decisions. In fact, you cannot not make decisions. Do you want to try?

And the often raised concern about moral hazard is plainly a red herring. There’s this idea that if people had no free will “they” could not be made responsible for their decisions. Scare quotes because this suggests there are two different parts of a person, one making a decision and the other one, blameless, not being able to affect that decision. People who commit crimes cause pain to other people, therefore we take actions to prevent and deter crime, for which we identify individuals who behave problematically and devise suitable reactions. But the problem is their behavior and that needs to be addressed regardless of whether “they” have a freedom in their decision.

I believe that instead of making life miserable accepting the absence of free will will improve our self-perception and with it mutual understanding and personal well-being. This acceptance lays a basis for curiosity about how the brain operates and what part of decision making is conscious. It raises awareness of the information that we receive and its effect on our thoughts and resulting actions. Accepting the absence of free will doesn’t change how you think, it changes how you think about how you think.

I hope that this made you think and wish you a nice weekend :o)

105 comments:

Allah Gold said...

I have no choice. I have to worry about it.

Thomas Dent said...

I don't know what you mean by 'possible futures'. Only one thing will happen in the future, not many things at once, so what is 'possible' and what is not?

Tommaso Dorigo said...

I had no choice but to read this piece in full. And I am glad that I had no choice ;-)

Dull jokes aside, nicely put.

Cheers,
T.

Sabine Hossenfelder said...

Thomas,

What I mean with different possible futures is given the initial conditions at a moment t the state at t'>t is not entirely determined. Yes, in the end only one thing will happen (unless you believe in many worlds), but at the time t there was no telling what that would be. Best,

B.

Phil Warnell said...

Hi Bee,

I think between properly understood implications of the nature of the quanta and the limits of chaos there is amble room for what could be accounted for as a free will in a conscious decision making being. Further I think the very existence of a thing called hope stands as testament to that. That is I’m convinced the exact nature of the future is impenetrable and thus for conscious beings it not only forces us to make guesses about the future, yet also leaves us with uncertainty about the past, which in turn not only has effect on what we experience as the now, yet also that future we struggle to navigate. With the rest of nature there is no reasoned decision which can be recognized which account for the way things are and yet within the scope of human influence there are many things which can be directly attributed to being resultant as such, whose reasoning for them can be assigned only to hope and little else. So you could say I find our capacity to hope is what has our will to be free, as certainly does hopelessness has it not to be.

”On Bohm’s theory, there is, right now (that is : before those upcoming measurements get carried out) an objective physical matter of fact about what the future act of h’s is going to be: and (moreover) h knows with certainty, what that act is going to be; and (moreover) no other observer in the world (no matter how adept they may be at measuring or calculating) can possibly know (right now) what that act is going to be.

And so h, under these sorts of circumstances (even though the complete physical theory of the world here is a deterministic one), has what you might call an inviolably private will.”


-David Z. Albert, “Quantum Mechanics and Experience” (page 188)


Hopeful regards,

Phil

Sabine Hossenfelder said...

Tommaso, Thanks for the kind words :o)

Phil Warnell said...

Hi Bee,

I think between properly understood implications of the nature of the quanta and the limits of chaos there is amble room for what could be accounted for as a free will in a conscious decision making being. Further I think the very existence of a thing called hope stands as testament to that. That is I’m convinced the exact nature of the future is impenetrable and thus for conscious beings it not only forces us to make guesses about the future, yet also leaves us with uncertainty about the past, which in turn not only has effect on what we experience as the now, yet also that future we struggle to navigate. With the rest of nature there is no reasoned decision which can be recognized which account for the way things are and yet within the scope of human influence there are many things which can be directly attributed to being resultant as such, whose reasoning for them can be assigned only to hope and little else. So you could say I find our capacity to hope is what has our will to be free, as certainly does hopelessness has it not to be.

”On Bohm’s theory, there is, right now (that is : before those upcoming measurements get carried out) an objective physical matter of fact about what the future act of h’s is going to be: and (moreover) h knows with certainty, what that act is going to be; and (moreover) no other observer in the world (no matter how adept they may be at measuring or calculating) can possibly know (right now) what that act is going to be.

And so h, under these sorts of circumstances (even though the complete physical theory of the world here is a deterministic one), has what you might call an inviolably private will.”


-David Z. Albert, “Quantum Mechanics and Experience” (page 188)


Hopeful regards,

Phil

Arun said...

I suppose from the scientific rather than philosophical point of view, the challenge is to come up with experiments where "free will" is the most economical hypothesis to explain the outcomes.

Theophanes Raptis said...

Jihadists and Salaffists too, believe in the "Kismet", the book of God where everything is written and thus predetermined. This is why it does not matters to them whether they or you or all of us die in the effort to bring the glory of Allah afront! "Allahu Akbar" my dear...

http://www.youtube.com/watch?v=IjZach8EaCg

Arun said...

Free will can only exist if there are different possible futures and you are able to influence which one becomes reality.

"you are able to influence" is not clearly defined, but from the rest of the essay it seems that what you have in mind is something that violates the laws of physics, and so, you're shooting fish in a barrel.

Let us see the next place "you" and "influence" is used.

It doesn’t matter how you define “you” (in the simplest case, think of a subsystem of the universe), “you” won’t be able to influence the future because nothing can.

The subsystem of the universe - if it interacts with whatever it is the future of which you are considering, then the quantum mechanical amplitudes do depend on the state of this subsystem. Therefore I can only think "influence" means something beyond this dependence, i.e., something unphysical.

I do think though, that the onus is on the one who asserts that "free will" exists is to come up with a definition that is some reasonable correspondence with the popular notions and is also in accordance with the laws of physics.

Phillip Helbig said...

Excellent summary, and good to mention the criminal angle at the end. Many discussions are like "if there is no free will, there is no responsibility, so we can't put people in jail" etc. Completely bogus, since modern criminal codes aren't based on revenge anyway: they are based on deterrence, protecting society from known offenders and rehabilitation, all of which are independent of free will. No free will doesn't mean "a criminal must commit a crime" but means "given some circumstances, a criminal will commit a crime", but society can (and should) set up circumstances, such as laws, to make such crimes less likely. (Of course, if one believes that no free will means that the criminal bears no responsibility, then by the same token one can't criticize the judge for putting him behind bars, since the judge has as little free will as the criminal.)

Phillip Helbig said...

"I think between properly understood implications of the nature of the quanta and the limits of chaos there is amble room for what could be accounted for as a free will in a conscious decision making being. Further I think the very existence of a thing called hope stands as testament to that."

Weasel words. Unless you can demonstrate free will in practice, this is just meaningless. Note also that there are lab tests which indicate that people will take a certain decision seconds before they consciously decide it.

"That is I’m convinced the exact nature of the future is impenetrable and thus for conscious beings it not only forces us to make guesses about the future"

Yes, but just since we can't know the future in no way implies that we have free will.

Historically, it was often religion which denied free will ("if it happened, it was God's will", though this is ambiguous). This was often used as an excuse (bad karma, the Fates etc). More enlightened times introduced the idea of individual freedom and responsibility. It turned out to be wrong, but that doesn't mean that the alternative is using the lack of free will to excuse anything.

http://www.lyricsfreak.com/r/rush/freewill_20119963.html

Some versions have, instead of "If you choose not to decide, you still have made a choice.", "If you choose not to decide, you cannot have made a choice." Take your pick. :-)

Come back when you can play the guitar solo. :-)

Andy Ford said...

If in fact free will does not exist, and there is only one future, then please explain the purpose of science, exploration, human curiosity and most specifically the reason for life's existence into your theory. Sadly I find it falls apart. If indeed we have no free will then we have been pre-programmed by (unspecified) to explore, utilize and develop scientific techniques and ultimately, make discoveries as was pre-ordained by (unspecified)? I personally feel that, like the continual conflict found in every facet of life, free will is just another process allowing for the continuation of existence. For example, we use conflict in every part of our language, our perception of the world, the cells in our body (death, regrowth), we even use it in our exploratory procedures like science, philosophy and psychology. Because of conflict, progress, evolution, discovery, life, death, hope, war, destruction, emotion are possible. Something static, like a single, pre-determined future removes the above and indeed invalidates all we know while also destroying itself. Now, perhaps the conflict engine I speak of is actually a malfunction of human perception (though I will note animals, plants and even planets, suns and galaxies seem to operate under this conflict engine). It is possible that your theory is correct. However, I might state that if your theory was correct, then a man or woman could safely take a fully loaded gun, aim it at their temple and pull the trigger. Unless, of course this pre-determined future allowed for that individual to die at that time, they would survive.

To me, it seems like the absence of free will in science makes scientific examination easier. It also is impossible to prove or disprove determinism or free will in this current state of human understanding. Furthermore, I believe that pursueing scientific discovery with the premise of a pre-defined future is irresponsible as it functions at a shortcut, but not an accurate one. Also, if a pre-defined future does indeed exist, then any discoveries wouldn't be the person who invested their time and effort into the process but rather credited to the static futures invariability. All the brilliant minds in the world are essentially mindless drones, serving as mouthpieces or slaves to a master with absolute and unquestioning power. Also, in historical reference, most individuals arguing that free will is a fiction are usually feeling conflicted or powerless against actions they took or actions taken against them. It is an easy way to convince yourself that any responsibility you might have in said trauma is null and void because you had no choice in the matter. On the line of criminals, if a person with a gun went to rob you on the street, there are a number of viable choices, with a large selection of possible outcomes. The most predictable choices a person might make are the only ones most people are willing to accept as choices, because the other choices likely result in death....

I did like your article, if I didnt agree with 90% of it.

the.excession said...

"-unless you want to argue that elementary particles also have free will"

This has already been done. John Conway and Simon Specher did this a couple of years ago.
http://paw.princeton.edu/issues/2009/07/15/pages/6596/index.xml

"Scientists" have always compared brain/mind/freewill/etc to the "coolest thing around". From looms and telephone-switchboards to the last decades over-use of the computer analogy.

I'm guessing that we are, as always, over-valuing our achievments/understandings and that in the future the notion of reductionism will be smiled at as a primitive mis-understanding.

And of course, I just had to write this... ;)

Physicalist said...

Hi Bee,

Thanks for the nice, thought-provoking post.

Rovelli seems (to me) to be in the ballpark of a standard compatibilist account of free will, but as you say, he’s placing too much emphasis on questions of predictability. Questions about what we know are misplaced in this context; freedom is about what the facts actually are – regardless of whether we do or can know those facts.

But his notion of freedom being absence of constraint is pretty much on target.

The compatibilist says that we don’t need to worry about determinism. The relevant question for freedom is whether you do something because you wanted to, or whether your desires were irrelevant in that case.

When Ghandi fasted in protest, he did so freely. When someone is lost in the desert without food, she isn’t fasting freely. If I give someone money because I want to, I’m acting of my own free will. If I do it because a gun is held to my head, it’s not a free action.

It should be obvious that this distinction between free and unfree actions is independent of the question of determinism. On the compatibilist account, to say that I have free will is (to the first approximation, anyway) to say that I can do what I want.

This compatibilist account of free will goes back at least to Aristotle, and is endorsed by the majority of philosophers. (See, for example, the SEP entry on compatibilism for more: http://plato.stanford.edu/entries/compatibilism/.)

The position that you seem to be rejecting is the “libertarian” account of freedom, which is indeed incompatible with determinism, and is rejected the compatibilists as well. But we need to be careful not to deny the compatibilist notion of freedom when we reject the libertarian’s account.

For example, you mention “the fact that “you” didn’t have any influence on what happens in the future.” I’m not sure what work the quotes around “you” are supposed to be doing, but obviously I do have an influence on what happens in the future. I’m a real physical system that makes a real causal impact on the world. I go through decision-making processes, and these processes certainly influence the future.

Perhaps this was just a typo (it seems you got this point right in your previous post discussing Carroll), but I see too many scientists slide from denying libertarian freedom to denying compatibilist freedom (presumably because the don’t grasp the distinction).

All the best.

Uncle Al said...

"Anybody who believes...that the standard model of particle physics is correct...must come to the conclusion that free will is an illusion." The standard model is a
violently parameterized jury rig. SUSY is hornswoggle. Parity violations and symmetry breakings are fundamental failures of rigorously derived, defectively postulated theory. MSSM is hallucinatory curve fitting, SUGRA is empirical crap, neutrinos are not Majorana. Identify dark matter with a particle, or is it self-tying mass-equivalent spacetime knots? Science 338(6108) 783 (2012)

I exist. The ovum might have been templated, but the spermatozoan was free will - one chance in a billion, by the hemocytometer.

Phillip Helbig said...

"This has already been done. John Conway and Simon Specher did this a couple of years ago."

Yes, discussed quite a bit in the blogosphere. It works only because their definition of free will is not what most people mean by the term.

vmarko said...

But in principle chaos is entirely deterministic. There’s still only one future. It’s just that chaotic behavior spoils predictability in practice.

Bee, this is wrong! Heisenberg's inequalities put a conceptual lower limit on the precision of initial conditions. Chaos can (and does) blow up these uncertainties, and after a finite time these errors outgrow the size of the system. While chaos would be entirely deterministic if the initial conditions could exist with infinite precision, it is dictated by nature that this precision is always finite. Determinism in general has been falsified by experiment, and saying that there is only one future (chaotic or not) is just wrong.

What you are arguing here is similar to saying that superluminal communication is possible since Newtonian mechanics doesn't put an upper bound on velocity. But Newtonian mechanics has been falsified by experiment, so your argument has a false assumption inside.

Best, :-)
Marko

vmarko said...

If no human can predict your behavior, but a computer cluster of an advanced alien civilization could, would you have free will?

As per my previous comment...

The computing power is irrelevant. Given a chaotic system, the time to first bifurcation is a fixed constant (and can be calculated using Lyapunov coefficients). After that, a physical system makes a choice (of course not needing any consciousness, deliberation or such stuff, it's just a random outcome). New information is being created at that point.

And there is no way one can predict this new information, regardless of the computing power one has at ones disposal.

It is a common misconception that, given a big enough computer, you can calculate predictions to a time arbitrarily far into the future. Heisenberg's inequalities put a hard cut on that time, and this is completely independent on any computational power.

Best, :-)
Marko

Michael F. Martin said...

Somehow W.V.O. Quine satisfied my curiosity about free will vs. determinism. Quine's observation is that the two are compatible so long as we are willing to acknowledge that our "free" will is part of chain of causation, with links both upstream and downstream outside our control:

"One is free, in the ordinary sense of the term,
when one does as one likes or sees fit; and this is not altered by the fact, if fact it be, that what one likes or sees fit has had its causes. / The notion that determinism precludes freedom is easily accounted for. If one's choices are determined by prior events,
and ultimately by forces outside oneself,
then how can one choose otherwise? Very well, one cannot. But freedom to choose to do otherwise than one likes or sees fit would be a sordid boon. "

ppnl said...

If one just uses the information in the mental states - which is what you might experience as “yourself”...


Free will is a red herring. It is that word "experience" that is the mystery. Our sense of free will comes from the fact that we experience or thoughts, our mental calculus. Free will may be an illusion. But the fact of experience is not. How do you explain the fact of experience?

PTMR said...

I concur, there is no free will.

Giotis said...

Wow! Just what the doctor ordered! Que Sera, Sera …

No responsibility, no guilt, no struggle for anything; just enjoying life and whatever comes our way.

Without free will we’re free at last!

Aaron Sheldon said...

I feel there is an essential element missing from your post. Namely that individual free will is a source of random noise. We all have free will up to the extent that nature is thermodynamically ambivalent about our choices. As long as the choices a person faces are roughly thermodynamically equally probable then the choice is free, as nature (oddly) really has no say in it.

However, this implies that on aggregate human societies have no free will, and are fundamentally governed by thermodynamic constraints. It is only under the most rare conditions that the global choices human society faces are actually thermodynamically ambivalent.

For example anything we do concerning anthropogenic climate change is fundamentally thermodynamically controlled. Whether or not human society curtails climate change depends on which outcome is thermodynamically more favourable.

Aaron Sheldon said...

Determinism is emergent. It is an inevitable property of statistical aggregates.

uair01 said...

Don't feed the troll.
Not even if she's called Sabine Hosenfelder.
:-)

Zephir said...

In dense aether model the Universe is random. Does it mean, it's indeterministic just before it? Of course not but the randomness doesn't exclude (the existence of) fluctuations which differ with their complexity and dimensionality, including the highly intelligent and hyperdimensional Boltzmann brains. The fact that our consciousness is essentially drive with randomness from general perspective doesn't mean, it cannot be predictable easily from less temporal/global perspective.

Arthur Schopenhauer: Man can do what he wants but he cannot want what he wants.

Zephir said...

BTW The extreme randomness is indistinguishable from extremely intelligent deity in our limited low-dimensional perspective. Try to imagine the shadow of rod regularly rotating rod in 3D projected at the 2D plane. This shadow will appear shortened (i.e. more similar to scalar particle) and it will move randomly. This is just the perspective in which our animal pet do perceive various human activities. The do perceive us in the same way, like we are observing the motion of ants inside of their nest: i.e. like chaotic motion without any logical connections. But does it still mean, that the motion of ants is not driven with anything from sufficiently general perspective? Or are we just incapable to perceive it in the same way like the motion of 3D rod by its 2D shadow?

Robert L. Oldershaw said...


I have a couple of comments on this interesting issue.

1. Much human behavior is reasonably predictable and this predictability is heavily used in marketing, politics, social interactions and life in general, both in the cases of groups of people and of individuals. The basic daily patterns of activity and specific choices can be predicted accurately for many people, although not perfectly.

2. Free will need not be a "dumbbell idea" with all the weight at the extremes, i.e., all or nothing; excluded middle. As I have argued before, through trial and error and feedback loops an individual can learn and change. There are many dramatic true stories of real people doing this. I think this is probably true even for a strictly deterministic universe, as Spinoza pointed out.

3. Because of #2, there is limited but significant accountability for our actions even if "free will" is very limited. If we did not improve our actions beyond an age of , say, 16 then it would be a far more chaotic world. We are not robots; we learn and change; and we participate in the process.

4. Neuroscience is rapidly making progress on the issue of "free will", in spite of the often desperate need to believe that we have complete "free will". For example, experiments have shown that people makes choices about when to take an action before they are aware of making that choice.

Such experiments will eventually remove some of the emotion from our thinking about "free will" and its limitations.

Have a nice weekend, and drink your strong coffee responsibly.

John Merryman said...

For one thing, free will is an oxymoron. To will is to determine. We distinguish between good and bad decisions in order to do what we think is the right decision, not to randomly chose.
The future is not determined though, because even though laws are deterministic, total input cannot be known prior to the occurrence of the event, since the lightcone is only completed by the event. You can argue this is simply epistemic and the ontological reality is deterministic, but there is no ontologically real past or future, only what is physically happening, so since determinism is not relevant without a timeline, it is necessarily epistemic and that cannot be known.

Babak said...

I agree free will too may be an illusion, but I think it's worth pointing out that [in]determinism itself is hard to pin down as concept. So while it may be true there is only one future, it may not be knowable (predictable) in the present. For example, "When will this computer program stop?", or "Is this 3 body system stable?"

The first is an example of mathematic indeterminism. It turns out there are classes of programs that the only way you can find out whether (and when) they'll stop is to actually run them and hope they stop some time soon. This is known as the Halting Problem in computer science, and has been shown to be a recasting of Godel's Incompleteness theorem in mathematics. While physics has yet to work in mathematical indeterminism into its theories, computers and the programs they run are indeed in the physical world.

The second is an example chaotic indeterminism. Here it's not a question of the error in measurement that makes the future unpredictable; rather it's that even with perfect measurement (with few exceptions) we can't characterize a 3-body system as stable or unstable.

Plato Hagel said...

Who wrote this post? Where did your idea come from and what things did you gather in order to complete the thought process?

So in that sense, as if a matter forming congregation of things macroscopically(as ideas become your reality and ours), it can be ascertain that such a gathering could have reached such a conclusive result? Oh! differences of opinion as to possible futures?:)

I think one would get the jest now that one can focus out of such a myriad of things that such a illusion could become so real?:)

You see, for your purposes then such a leaning and attribute as to make such a statement has defining property in the case of who Sabine is. You see?:)

Best,

John Merryman said...

As individual points of reference, we experience reality as a sequence of events and so the elemental model is linear cause and effect. Therefore it must be deterministic, as each event is followed by another.
This vector of time is emergent though. The physical reality is constant flux and time is a measure of change. It does not move along a vector from past events to future ones, rather events form and disperse, so it is these configurations going future to past. Potential precedes actual, not the other way around. The earth is not traveling some dimension from yesterday to tomorrow. Tomorrow becomes yesterday because the earth spins. So what is future is not fundamentally determined. Like the computer programs, we have to let it run to find out, because there is no simpler program we could run ahead of time to find out everything. The total input is incomplete prior to the event.
Even past events are constructs of relative understanding in the first place and this changes with time as well.

John Merryman said...

Human history is a fight to write the narrative. The gods are laughing.

JimV said...

I have no argument with most of your post, but think I might be able to add something to the discussion on this point:

"To begin with, it isn’t at all clear that it’s impossible to predict human behavior, it’s just presently not possible. Since ten thousand years ago people couldn’t predict lunar eclipses, this would mean the moon must have had free will back then. And who or what makes the prediction anyway?"

When writing computer games, I find it is useful and often even necessary to include a random element in calculations. Without this, not only are the games boringly predictable but they can actually fall into traps which only random actions will get them out of (example available on request).

Although the pseudo-random number generators which are available in simple computer languages are not very robust, they can be improved by a simple means: count CPU cycles between external inputs (such as mouse clicks and key presses), and use the current result as a fresh seed for the random number generator. Since CPU cycles depend on the frequency of the alternating current of the power source, and this varies (typically by 0.5% of the basic line frequency of 50 or 60 hz, but this is enough to affect cycle counts), the result is totally unpredictable, even if a robot were tapping the keys at set intervals.

All of which is to say, that if I could figure that out for my Applesoft version of "Breakout", I'll bet a few billion years of evolution could do the same, and have wired similar algorithms into our behavior, for cases in which we are not sure what to do. Whether it has or not, in principle it could.

Sabine Hossenfelder said...

Hi Marko,

You have several confusions in your argument. First, the time-evolution of the state in quantum mechanics is entirely deterministic, it's the measurement which (in the standard interpretation) isn't. Second, determinism hasn't been "falsified" because it can't be falsified. There's always the possibility that something determines measurement outcomes, we just don't know what. (Hidden variables, etc). It's just that to our best present knowledge the measurement outcome is indeed non-deterministic. Third, there's nothing wrong with my statement about chaos. It is irrelevant whether it's a limit in practice or in principle. The point is that as long as there's one state and a time-evolution on that state, the future is determined. Whether you or anybody else can determine it is an entirely different question. Best,

B.

Sabine Hossenfelder said...

Hi again Marko,

"After that, a physical system makes a choice (of course not needing any consciousness, deliberation or such stuff, it's just a random outcome)."

I have no clue what you mean with that. Random outcomes aren't choices. And I also don't know what randomness you refer to. Best,

B.

Sabine Hossenfelder said...

PPNL,

"Experience" is neural activity. I don't know why people think there's something mysterious about it. Best,

B.

MarkusM said...

Remarkably, John Conway who together with Simon Kochen proved to Free Will Theorem does not know what free will is. All that he can say is what it is not.
His 6 lectures are incredible - a MUST SEE:
http://www.princeton.edu/WebMedia/lectures/

In my opinion free will is a totally fishy concept that as longs as it lacks a proper mathematical definition is of no use to physics.

Best.

John Merryman said...

B,
The past has, by definition, been determined. The future has, also by definition, not been determined.
The question then gets back to the nature of time and can there really be a form of blocktime, or is that our projection of causality in a linear fashion, vs causality as an necessary dynamic?
Is there some form of four dimensional "fabric of spacetime," or is there just an eternally occurring physicality, constantly reconfiguring? If it is is the second, than the future will always be potentialities, not yet actualities, thus not yet determined, ie, past.

John Merryman said...

There is a difference between, "the future is determined," and the future will be determined.
No matter how omniscient one might be, they would still have to run the equation to find the answer.
If God knows the future, then God has a finger on the scale.

11 said...

If we live inside an entangled raspberry shaped multiverse which includes anti-material and material entangled copy universes, then there is still overall symmetry after the big bang and a real base for MATERIAL AND HUMAN CHOICE MAKING as multiversal wavefunction collapse.
Future Benjamin Libet experiments could give us more clarity.
see:
Wavefunction collapse and Human choice-making inside an entangled mirror symmetrical Multiverse.
http://vixra.org/pdf/1103.0015v1.pdf


MarkusM said...

"To begin with, it isn’t at all clear that it’s impossible to predict human behavior, it’s just presently not possible."

Interestingly, Niels Bohr speculated (see his 10th volume)
that to determine the initial state of a biological system which is essential for predicting its behavior, one had to do such a heavy load of measurements, that one would destroy it.

In a sense, this would imply that biological matter is another "quality" of quantum matter.

Best

11 said...

Dear Markus,

Yes biological matter is an other than quantum matter,in the sense that biological matter wavefunction collapse is subjected to a retarded collapse choice which give humans the possibility to reject such a choice within a time span of about 300 msec. ( B. Libet)see:
Wavefunction collapse and Human choice-making inside an entangled mirror symmetrical Multiverse.
http://vixra.org/pdf/1103.0015v1.pdf

Leo Vuyk

BG said...

Bee, you say:
"In a deterministic theory, like all our classical theories, this just isn’t the case - there’s only one future, period."

This sentence corresponds to the standard interpretation of a deterministic theory, which is the following: at the initial time of the universe something chooses the initial conditions of the universe with infinite precision, and then the deterministic law of motion univocally determines the future of the universe.

There is however another possible interpretation. Note first of all that in a deterministic universe choosing the initial conditions of a trajectory is equivalent to choose the trajectory among the set of dynamically allowed trajectories. The interpretation I propose is the following: no trajectory is chosen with infinite precision at the initial time of the universe, but there is a set of actual trajectories which is continuously refined during the evolution of the universe by choices performed by the universe itself. For example: at the time t the past choices have restricted the set of actual trajectories to a suitable set A, which is compatible with the actual past but which includes many different futures. At that time the universe (for example a human being) makes a free choice, and so reduce to B < A the set of actual trajectories. It is obvious that this second interpretation is fully compatible with free will.

vmarko said...

Hi Bee,

First, the time-evolution of the state in quantum mechanics is entirely deterministic, it's the measurement which (in the standard interpretation) isn't.

But the state in QM is a probability amplitude, it doesn't "determine" any outcome of any experiment.

QM provides one with deterministic evolution of probabilities, not observables. So the whole theory is probabilistic, not deterministic.

Second, determinism hasn't been "falsified" because it can't be falsified. There's always the possibility that something determines measurement outcomes, we just don't know what. (Hidden variables, etc). It's just that to our best present knowledge the measurement outcome is indeed non-deterministic.

Do you really take hidden variables seriously? :-)

Local realism has been falsified by the violation of Bell's inequalities. So there are three choices --- give up realism, or give up locality, or give up both.

If you give up realism, you also give up determinism, since initial conditions for a physical system do not exist beyond finite precision.

If you give up locality, you end up with nonlocal equations of motion, which do not have a well-defined Cauchy problem, and do not guarantee the existence of a unique solution for given initial conditions. So you have to give up determinism in this case as well.

If you give up both realism and locality, you're even worse off.

Hidden variables can exist only in the loopholes for the experiments that violate Bell's inequalities, but if you want to hide determinism in there, good luck! :-)

Third, there's nothing wrong with my statement about chaos. It is irrelevant whether it's a limit in practice or in principle. The point is that as long as there's one state and a time-evolution on that state, the future is determined.

If by "state" you mean classical phase-space point (classical chaos), the physical system is never in "one state" only. The state of the system doesn't exist as a point in phase-space but only as a finite volume, of size determined by Heisenberg's inequalities. That volume can be evolved deterministically only up to a bifurcation point (if the system is chaotic), when this phase-space volume gets split into two disconnected pieces, each with its own future evolution. There is no way to predict in which of these two states the system will end up in the future.

In other words, the Cauchy theorem guarantees the uniqueness of a solution for EOM, provided that there exist initial conditions with infinite precision (a point in phase space). But these just don't exist! The physical system can never be in just one point of phase space, it is always smeared out across a finite volume.

If by "state" you mean the QM state, i.e. the wavefunction --- then no, the future is not determined. Rather, the probabilities for the future are determined. That does not imply determinism.

Random outcomes aren't choices. And I also don't know what randomness you refer to.

Consider a ball at rest on top of a hill, in an unstable equilibrium. Since the ball's position and momentum are determined only up to finite precision, it cannot stay in equilibrium, but will roll down to one of the two sides of the hill. In other words, prepare a ball in that initial state, wait some time, than measure the ball's position. There is no way for QM to predict the result of that measurement. The ball will roll down in a random direction. In this sense, the ball (or Nature, if you will) makes a choice about which way the ball will go. God plays dice, remember? :-)

Maybe the word "choice" is a poor choice of words. ;-) But that's just semantics.

Best, :-)
Marko

Sabine Hossenfelder said...

Hi John,

"The future has, also by definition, not been determined."

You can define whatever you want, but this is clearly not a definition that I think many people would agree on. The past/present/future distinction relies on the possibility of time-slicing and the future is everything 'past' now, where the direction (which one is past which one future) is fixed by the increase of entropy. Psychologically, the difference between past and future is that you can remember the past but not the future, but that doesn't tell you anything about whether the future is determined or not. Best,

B.

Sabine Hossenfelder said...

BG,

"the standard interpretation of a deterministic theory, which is the following: at the initial time of the universe something chooses the initial conditions of the universe with infinite precision,"

I don't understand what you mean by choice. In the 'standard interpretation' the initial conditions are not chosen, they just exist. You can measure them more or less precisely, but that's a different issue entirely. The only context in which you 'chose' them is when you're trying to model a system and aren't sure which are the correct initial conditions. But that's not a choice that alters the real system, it's a choice that allows you to describe it more or less exactly. Best,

B.

Sabine Hossenfelder said...

Hi Marko,

"QM provides one with deterministic evolution of probabilities"

QM provides one with a deterministic evolution of the state, combined with a probabilistic evaluation of observables for that state and the corresponding reduction of the state.

"Do you really take hidden variables seriously?

Yes, but that's a digression.

"So there are three choices --- give up realism, or give up locality, or give up both."

To begin with, I'm comfortable giving up realism. But besides this, all you have to do is give up free will, the free will that you need for choosing detector settings independent of the preparation of the state. It's called super-determinism. Technically, it's a form of non-locality in Bell's sense, but I find it quite misleading to refer to it as non-locality. It's spatial correlations, but there's no non-local propagation of information.

"The ball will roll down in a random direction. In this sense, the ball (or Nature, if you will) makes a choice about which way the ball will go. God plays dice, remember? :-)

Maybe the word "choice" is a poor choice of words. ;-) But that's just semantics."


Yes, it's just semantics. As I've said many times before, I don't see the point in fighting about words. If you want to call non-determinism "choice", fine by me. But it's not a choice in the sense that I've defined it in my post, that it is possible for a subsystem that is called "you" to influence a non-determined future outcome.

As I wrote here, I think what you need to make sense of "choice" and "free will" is a time evolution that's neither deterministic nor just undetermined. It seems that our disagreement goes back to our different definition of "choice". Best,

B.

Sabine Hossenfelder said...

Hi Phil,

"Not being able to predict" is a necessary but not a sufficient condition for "hope". In that sense I would agree that free will allows for hope. But free will does not necessarily imply hope and neither is the existence of hope a proof for free will. I would argue that hope is basically the process of putting odds on favorable future outcomes that we are not able to predict with certainty. Clearly lots of psychology involved in placing these odds, but I don't think that you need free will (in the sense that I defined it) for this. Best,

B.

Neo Borocillina said...

"People who commit crimes cause pain to other people, therefore we take actions to prevent and deter crime, for which we identify individuals who behave problematically and devise suitable reactions."

Are not fare away (dark) times in which, in democratic coutries, a psychiatrist could force an *unwilling* subject to a neurosurgical intervention although the presumebly ill "unwillig subject" had not committed any crime at all. It was for "his own good". If I'm an in important psychiatrist and you are simple countryman why should I respect your free will? At the end of the day I have no free will, you have no free will, but I know how your brain works far better than yourself, and my competence can make you be in much more harmony with the society, so let's give all the decision-making power about our brain to neuroscientists!
Is this a purely sci-fi vision vision?
I hope so, but something very similar has been already tried (tragically), and it seems to me, that it was because it was supposed that:
1) mind is no more than physical brain;
2) there is no free will to take into account.
I do not think that the ethical problem is so easy handle.
Thank you very much for your very interesting post.

BG said...

Bee,

"In the 'standard interpretation' the initial conditions are not chosen, they just exist."

Ok, it is not a problem of terminology. The problem (for me) in the "standard interpretation" is that you assume that something exists (namely, precisely defined initial conditions) without the possibility to measure it. This is the point of view of scientific realism, but it is not a scientific truth.

I repeat, a possible alternative interpretation is that what actually exists is not a single initial condition or, equivalently, a single deterministic trajectory, but rather a set of possible trajectories which are compatible with our knowledge. In this alternative point of view, what actually happens in a measurement is not an increase of our knowledge about the only existing trajectory, but rather "a choice" of the universe, which chooses the trajectories corresponding to the obtained result and excludes the others.

Do you agree that these two interpretations are equivalent form the mathematical and the predictive point of view?
Bruno

susskind said...

"Now you could dismiss this and claim there must be something about nature then that these theories don’t correctly describe and that’s where free will hides. But that’s like claiming god hides there."

No, it isn't. We've got lots of empirical evidence -- even though it may be misleading -- that favors the hypothesis of our having free will. I don't know of any empirical evidence that supports the existence of God (or Gods).

I think we can interpret our complete ignorance about, for instance, the nature of consciouness as pretty good evidence for thinking that our current theories are only very partial descriptions of reality.

Plato Hagel said...
This comment has been removed by the author.
Plato Hagel said...

Bee:Psychologically, the difference between past and future is that you can remember the past but not the future, but that doesn't tell you anything about whether the future is determined or not

I know you have a Deja Vue post that explains this, but I would be more inclined to believe that if one looked deeply that the probabilistic outcome would have already been revealed? That you have already seen some future glimpses that you are unaware consciously, which means that they already reside in your thought process some where.:)

No!Not just google pics or paragraphs taken from you and displayed, but truly, a course you had already set by choosing what you are doing now?:)

Plato Hagel said...

How can free will exist? Does our views of the universe allow for such things as the future to unfold? If information always existed, then that potential already resides in the universe? Why not?

Now, of course the Big Bang wasn’t like that because it was a great explosion, but it was very much like that in all other respects. You could have a box that could expand in such a way that keeps the entropy constant. So you have an adiabatic expansion, as it’s called: this expansion that keeps the entropy constant. That’s the kind of picture that people have for the Big Bang.See: The Cyclic Universe: A conversation with Roger Penrose

John Merryman said...

Bee,
"You can define whatever you want, but this is clearly not a definition that I think many people would agree on. The past/present/future distinction relies on the possibility of time-slicing and the future is everything 'past' now, where the direction (which one is past which one future) is fixed by the increase of entropy. Psychologically, the difference between past and future is that you can remember the past but not the future, but that doesn't tell you anything about whether the future is determined or not."
I'll try to refrain from going too far off topic on the issue of time, though it is the basis for determinism, but it's been my contention that the major problem with physics is it instinctively treats time as sequence from past to future, rather than the process by which the future becomes past. This leads to enormous conceptual issues that are ultimately unnecessary. It was the subject of my last year's FQXI contest entry: http://fqxi.org/community/forum/topic/1304
Regards,
John

John Merryman said...

Reductionism is handy. Except when you miss something important.

vmarko said...

"QM provides one with deterministic evolution of probabilities"

QM provides one with a deterministic evolution of the state, combined with a probabilistic evaluation of observables for that state and the corresponding reduction of the state.


Ok, we are saying the same thing. I am just talking in Heisenberg picture (fixed state, evolving observable operators), and you are talking in Schrodinger picture (evolving state, fixed observable operators).

all you have to do is give up free will, the free will that you need for choosing detector settings independent of the preparation of the state. It's called super-determinism.

Oh, the superdeterminism! Let me summarize it --- the universe is deterministic, and its initial conditions are such that some biological systems on some planet 13 billion years from the initial moment are constantly performing precisely those experiments whose results will fool them into thinking that the universe is not deterministic.

This point of view sounds ridiculous. The amount of fine-tuning required for this is so unbelievable that it makes the cosmological constant problem look trivial!

Superdeterminism is a perfect candidate for use of the Occam's razor: is it simpler to think that the universe is deterministic and has initial conditions that conspire to make us believe it isn't, or is it simpler to think that the universe just isn't deterministic, and that all our experimental results are legit rather than "framed" against us?

If you believe in superdeterminism, you might just as well give up doing science, since all experiments are framed against you finding the true laws of physics. I just don't believe it.

As I wrote here, I think what you need to make sense of "choice" and "free will" is a time evolution that's neither deterministic nor just undetermined. It seems that our disagreement goes back to our different definition of "choice".

Oh, I see, your idea of a free will function is interesting! Terminology aside, my point of view on free will is essentially the usual two stage model. I think that it is even compatible with your free will function, though I didn't have time to think about it seriously yet. But IMHO the laws of physics, being indeterministic, certainly allow for the existence of free will (in the framework of the two-stage model).

Best, :-)
Marko

John Merryman said...

To will is to determine. What is "free" will in the first place? Are we/our decision making process, not a factor in the equation????

Uncle Al said...

Regarding vmarko's ball: Work out the average time an infinitely sharp pencil, balanced perfectly upright, will take to fall. Consider a pendulum of length 1 and mass 1. Specify the orientation by the angle, x(t), measured from the position with the bob vertically below the suspension point. The potential energy is

V = g * (1-cosx) = 2g * (sin(x/2))^2

Set the initial conditions, x(0) = 0, x'(0) = 2g The energy

E = 1/2 (x')^2 + 2g * [sin (x/2)]^2
dx/dt = sqrt {4g [1 - [sin (x/2)]^2}

Then,

dt = sqrt{1/4g[1 - sin(x/2)]^2)}dx

The integral of dx from x = 0 to x = a diverges as "a" approaches pi.

What does this say about the determinism of a canonical ensemble of independent pencils balanced on their points, regarding the time interval to initiate each fall and the direction of each fall? "Many worlds" is not a worthy rebuttal. (Tapping a table covered with newly minted pennies balanced on their edges with random plane orientations does not result in 50_50 heads:tails. The coin punching process is not edge-symmetric.)

tytung said...
This comment has been removed by the author.
tytung said...

As I mentioned in a previous blog entry here, although free-will fundamentally does not exist according to our current best theories, it exists at an emergent level, same as the level where the sense of self emerges.
So it IS important how "you" is defined.

Daniel W said...

Sabine,
Someone must exist to have free will. Who do you think exist? You? And what is that according to our best theories?

Sabine Hossenfelder said...

Susskind,

"We've got lots of empirical evidence -- even though it may be misleading -- that favors the hypothesis of our having free will."

Like what?

Sabine Hossenfelder said...

Neo,

"in democratic coutries, a psychiatrist could force an *unwilling* subject to a neurosurgical intervention although the presumebly ill "unwillig subject" had not committed any crime at all. It was for "his own good"...

Nothing like that follows from what I said. Look up "slippery slope" on the list of logical fallacies. Best,

B.

Sabine Hossenfelder said...

Hi Bruno,

No, I don't agree. People have tried this time-symmetric reformulation of quantum mechanics in which a measurement affects a state in the past, and for all I can tell it's very controversial for not to say philosophically problematic. Ask yourself what counts as a measurement, who measures, and what is an increase in knowledge and then take into account that backwards lightcones from spatially separated regions overlap. Be that as it may, it's something that people have tried and the reason I didn't mention it is simply that I forgot. I should also mention Scott Aaronson's take on the question. Best,

B.

Sabine Hossenfelder said...

Hi Marko,

Thanks for the reference, I'll have to think about the two-stage model.

"Oh, the superdeterminism! Let me summarize it --- the universe is deterministic, and its initial conditions are such that some biological systems on some planet 13 billion years from the initial moment are constantly performing precisely those experiments whose results will fool them into thinking that the universe is not deterministic.

This point of view sounds ridiculous."


Yeah, and I don't know where you got it from. All you need for superdeterminism are spatial correlations between the prepared state and the measurement apparatus. Which, if you consider that they, in all experiments that we do, have been in causal contact, isn't much of an assumption. At least in my opinion.

I think you're forgetting that, as I said earlier, I'm not a realist. The prepared state doesn't have to be a classical "thing" that just "conspires" to appear quantum mechanical. Let it be a quantum state. The point is that you can't use Bell's theorem to rule out a superdeterministic theory. Best,

B.

Sabine Hossenfelder said...

Tytung,

As I mentioned previously, and again in this blogpost, appealing to emergence doesn't make your will any more free than it is fundamentally. Best,

B.

Sabine Hossenfelder said...

Daniel,

Someone must exist to have free will. Who do you think exist? You?

Define "exist". Best,

B.

susskind said...

Well, I thought it was obvious from everyday life. It is one of the most obvious facts that people have the ability to make free decisions. In the day to day life meaning of free -- e.g. you can decide in which party you vote, if you are not subject to coercion.

Obviously, you can always define "free" in such a way that this is denied. You can argue that your decision is not free because you've been indoctrinated, for example.

Someone might say you cannot take take any decision "freed" from the cognitive limitations associated with the fact that you're a human being.

But these objections just illustrate that there is no concept of "free" in an absolute sense, i.e. regardless of a given context.

But I think I can press my point still a little further. Take this post. In order to stand for the meaning of what you wrote, you've got to assume you have free will. This must be so for you to be capable of evaluating and interpreting the consequences of present scientific theories. If you don't have free will, your argumentation is meaningless since then it is nothing more than the unfolding of the initial conditions of strictly psychological dynamical processes in your brain---essentially unconstrained by any conceptual, logical, or rational criteria.

I would add that the fact that we cannot derive from fundamental physics anything relevant about our cognitive capacities makes any extrapolations therefrom to the question of free will rather meaningless.

After all, we have tested such theories with great success, no doubt, but in a rather limited region of what we are used to call reality.

Sabine Hossenfelder said...

Susskind,

"Well, I thought it was obvious from everyday life. It is one of the most obvious facts that people have the ability to make free decisions."

The only thing that's obvious is that most people have the feeling that they have a free will.

"Take this post. In order to stand for the meaning of what you wrote, you've got to assume you have free will. This must be so for you to be capable of evaluating and interpreting the consequences of present scientific theories. If you don't have free will, your argumentation is meaningless since then it is nothing more than the unfolding of the initial conditions of strictly psychological dynamical processes in your brain---essentially unconstrained by any conceptual, logical, or rational criteria.

That's plainly wrong. I have written this post without any "free will". My brain has just executed some processes that act on input it has gathered in the past and made the output available for other people. Conceptual, logical, and rational criteria are present in this process because they've proved beneficial in natural selection.

Frankly, I have the impression you haven't thought about the question of free will very much. Best,

B.

susskind said...

Sabine, You could have written the post without having any free will. I obviously agree. I wrote that you need to assume you have free will in order to "stand by what you wrote". Probably it is because I haven't thought enough about this question that I don't understand what means evaluating data, arguments and theories without free will. Presumably you can. I'd be quite happy to hear about it, if that's not asking too much.

Sabine Hossenfelder said...

Susskind,

" I wrote that you need to assume you have free will in order to "stand by what you wrote". Probably it is because I haven't thought enough about this question that I don't understand what means evaluating data, arguments and theories without free will."

No, I don't need to. As I explained I don't believe in free will. So I guess you can count my post as a falsification of your hypothesis ;)

What I mean is the following. Forget about the human brain, think about a computer. A computer has some software that it uses to process input and it produces output. Input could be things that you feed into the computer, but a more advanced version might go around and collect information themselves. Output is the result of whatever the computer is programmed to do. (Eg do a calculation, show a webpage, send data over a wireless, etc).

The brain does the same thing, just that the program is complicated and hasn't been "designed" by some programmer, but evolved by natural selection. The brain is also much more complex than presently existing computers. But these are differences in degree. You have as much free will to make decisions as my computer has when I submit this comment. It takes the data, does what the program says, and produces output, and that's it.

The only relevant difference is that humans are self-aware and thus try to construct theories about their own behavior. That means, you try to predict what you will do given the input that you are consciously aware of. That prediction isn't great which leaves you with the impression that there are different possible outputs. It is almost certainly the case that this imperfect self-prediction has an evolutionary advantage, otherwise it wouldn't have survived. Thus the persistence of the free-will illusion.

Best,

B.

Plato Hagel said...

People seem to have trouble with reductionism....30 sec to 1:30 after that, is a culminating effort to artistically define.

susskind said...

Sabine, maybe, you could well be right. I don't know. I'm open in this regard. But as far as I can see, this is a tough nut to crack, and I've never heard of anyone having succeded in cracking it.

" No, I don't need to. As I explained I don't believe in free will. So I guess you can count my post as a falsification of your hypothesis ;) "

But wouldn't you agree that your belief in the content of what you wrote could be as much an illusion as that of believing in free will? Since, anyway, you had no choice. ;)

Funnily enough, John Searle once pointed out to an asymmetry that exists between the notion of free will being an illusion and other type of illusions. Paraphrasing him, when you take a decision you have to act on the pressuposition that your action is free. Go to a restaurant, the waiter asks which dessert you wish to order and you just answer: Look I'm a determinist, I will just wait and see what I order. Still, everyone will grant you that the decision may anyway be determined.

Plato Hagel said...

Perhaps by analogy this is more pleasing.....think of it as an aspect of expression not so detailed by the computerized process, but an effort to see beyond the structures with which you measure things.

The Higgs boson, places an aspect toward "will building" that is more then just matter forming?

It's not just about reductionism anymore, and it opens up a whole new area with which to think about the world we live in, the things that can be created.

BG said...

Hi Bee,

I realize that I am not able to explain my point, because the interpretation I propose has nothing to do with the time-symmetric formulation of quantum mechanics. I do the last attempt to explain myself, because I do not want to annoy you.

In great synthesis, the mathematical elements of a deterministic theory are the following:
1) a suitable set of initial conditions;
2) a law of motions (usually a differential equation)
3) a set of trajectories in a suitable configuration or phase space, satisfying the law of motion.

Due to the deterministic character of the law of motion, there is a one-to-one correspondence between the set of initial conditions and the set of trajectories.

That's for the mathematics. Let me now discuss the possible interpretations of this mathematical model, in particular, what I have called the "standard interpretation", and the "alternative interpretation".

Standard interpretation: just one initial condition, together with the corresponding trajectory, exists in the real world. The others initial conditions and the others trajectories are only mathematical entities, but they do not exist in the real world. We do not know exactly what the existing initial condition is, because our knowledge of the physical world is limited, and it is compatible with many initial conditions. However, our knowledge of the existing initial condition/trajectory increases with time evolution, e.g. during measurements. For example, when we toss a coin and obtain the outcome "head", our knowledge of the real initial conditions increases because we can exclude the initial conditions/trajectories corresponding to the outcome "tail" (recall that we are discussing of a deterministic universe, and the initial conditions of the universe determine also the result of our toss).

Alternative interpretation: all the initial conditions compatible with our knowledge of the empirical world at a suitable time exist. When we toss a coin and obtain "head", we "destroy" the initial conditions/trajectories corresponding to the results "tail" (I prefer to say that the universe refines the "choice" of its trajectory to the trajectories corresponding to the outcome head). Note that this destruction does not affect the past, in the same way in which, in the standard interpretation, the augmented knowledge of the existing initial condition does not affect our memory of the past. If you want, you can imagine this destruction of a set of trajectory as the deterministic analogous of the wave function collapse.

So, you see that the difference between the two interpretations is not a matter of experimental physics, but simply consists of a different attribution of the notions of knowledge and existence to mathematical entities such as initial conditions and trajectories. In the standard interpretation a single trajectory exists and measurements increase our knowledge about it, while in the alternative interpretation all the trajectories compatible with our knowledge at a suitable time exist, and measurements destroy the trajectories compatible to the non-obtained outcomes.

Of course the standard interpretation does not admit free will, while the alternative interpretation does. I do not favor one interpretation against the other, because they cannot be decided with an experiment, and therefore the choice between them is a metaphysical matter. For this reason, as a physicist, I do not care it. I discuss about this only because it seems to me that you claimed that absence of free will in a deterministic universe is a logical necessity. My opinion is, on the contrary, that existence of free will also in a deterministic universe is a metaphysical issue, analogous to the choice between the standard and the alternative interpretation.

Best, Bruno

Plato Hagel said...

the Higgs not only has a temperature when it kicks in, it has a low temperature when it kicks out. Pg 51. The Cyclic Universe: A Conversation with Roger Penrose

One is looking for a natural process that operates within the expression of things cosmologically. So you might look at the sun in just one way, and our relationship with the sun, but there is a balance struck, as something similar in relationship to the total universe?

Neo Borocillina said...

"Nothing like that follows from what I said. Look up "slippery slope" on the list of logical fallacies."

I'm sure that "in everyday life" you would never ever compare the decision making process of a drunk man with that of a person simply belonging to a different ethnic group but:
"Is for example, alcohol intoxication a constraint that’s “external” to your decision making unit? Is your DNA an external constraint?"
This is a typical "slip" when considering mind as mere brain and "free will" as non-existent.

Best regards.

Neo Borocillina said...
This comment has been removed by the author.
Christine said...

Sabine,

Correct me if I am wrong, but it appears to me that your argument relies entirely on the procedural paradigm:

input -> process -> output.

If that is the case, I do not think this paradigm is a sufficient condition for your argument on the absence of "free will" (as you define).

The reason is that we could move to another paradigm for describing physical laws, as, for instance, by the use of a declarative (nonprocedural) logic, in which you do not need a process flow. (It could mimic the physical expectations based on the procedural paradigm).

One could argue that describing a universe without any "flow" is useless and contradicts what is observed. The point here is, however, that the universe could be fundamentally declarative and not procedural, and you could retrieve a sense of causality, or whatever feature that a physical law seem to require in order to express evolution of a state, by setting "boundary conditions" to this purely declarative paradigm. The name "boundary conditions" here is, however, not a good one, as you may jump to "initial conditions". That is not what I refer.

Those "boundaries" are not inputs. They are some kind of "matching conditions" to produce a specific "running" (physical evolution) which you match to what is observed. So the "logic" is the other way around: the observation is not because of some condition -- but because there is some observation, you have a condition.

In fact, the nature of such conditions could be fundamentally much more complex/unspecified/incomplete than we could infer by a realistic modeling of any physical system, including the brain. First, they could not necessarily be specified at some initial time or at some spatial boundary of your declarative model. They could involve other parameters, relations or properties, such as the masses of fundamental particles, etc, or even fuzzy conditions. Second, the problem might even be worse, as BOTH the procedural and declarative paradigms (or all others we, humans, constructed, such as functional/lambda calculus, etc) should, as far as we understand, be subject to Gödel's incompleteness theorems, since those constructions are fundamentally axiomatic systems.

So we live in a universe whose laws are describable by mathematical frameworks that are inherently incomplete. Either this is how our universe is, or our mathematics is incomplete in an otherwise "complete" universe, or both.

The problem I have with your argument (apart from the philosophical objections I mentioned on FB and privately), is that, given the above considerations, there is in fact a large room for "free will" than otherwise. Evidently, what I point out is not a "proof", nor even a flavor of one, but it appears to me these considerations are important and you seem to miss them. Yet, by all means, I do sustain that such questions cannot be unanswered by physics (nor by philosophy, as it is not a case for finding explanations or proofs), but to be regarded as a metaphysically unanswerable question.

Best,

Christine

Sabine Hossenfelder said...

Susskind,

"Go to a restaurant, the waiter asks which dessert you wish to order and you just answer: Look I'm a determinist, I will just wait and see what I order."

That's a confusion of identity, much like the strange idea that a murderer cannot be blamed for his action if he doesn't have free will, which implicitly assumes there's two identities, one acting and one observing the action. It's the same in this example. "You" are the one making the decision, so you cannot at the same time "wait" for somebody or something else to make the decision instead of you. As I wrote in my post, "you" in any sensible meaning of the word still have to make the decision.

I mentioned this in my paper, there's a mental disorder called depersonalization which does exactly this mental "split", resulting in the feeling of "watching oneself act". Best,

B.

Sabine Hossenfelder said...

Christine,

Please note that in my post I was very clear that what I said is "according to our best present knowledge" of the laws of nature. That is to say the standard model and general relativity and you know what the math says. There's initial states, there's a Fock space, there's a Hamiltionian, there's Borns rule, and that's basically it. I am all open to the possibility that these laws are incomplete and a more fundamental description allows for free will. (After all, that's exactly what I proposed in my paper.) So in that sense I certainly agree with you. Best,

B.

Sabine Hossenfelder said...

Neo,

Sorry, I can't make sense of your comment. All I said was that absence of free will doesn't justify inflicting pain on humans any more than existence of free will does. As we learn about the functions of the human brain that does of course open new treatment options for certain conditions, but the question which of these can be useful under which circumstances is difficult for social and medical reasons, again regardless of whether or not humans have free will. Best,

B.

susskind said...

" As I wrote in my post, "you" in any sensible meaning of the word still have to make the decision."

Yes, Sabine, I agree. But while doing so you must act in the pressuposition that your choice is free. Even though, afterwards, you may sit on your armchair and argue that, in the light of present physical theory, you are unable to see how your decision could have been free. And I find this " personality shift" slightly awkward. I understand you find it natural. But that is fine.

Personally, I prefer to describe the situation as follows. In the hypothesis that people have free will, it seems to be very hard even to start to imagine the nature of its underlying mechanism. We are stuck for the time being without being able to make much in way of progress regarding whether people have free will or not. We should also keep in mind that we could be banging our heads against fundamental limitations of human's biogically evolved cognitive capabilities . Maybe this is a problem we cannot crack in much the same way as an ant cannot conceive of an Hamiltonian.

Daniel W said...

Sabine,
The definitions are up to you. I think your beliefs about our world a wrong and I think it would serve you well to clarify a lot of your reasoning. Therefore I think you should consider your own existence in terms of "our best theories".

Maybe you can't do that. Then you should realize it's not even interesting to bother about the existence of free will since free will requires the existence of a person.

Or maybe you realize that a person is something approximate or vague in the context of "our best theories". But then it doesn't make sense to speak of free will in the rationalistic fashion you do it.

Neo Borocillina said...

"All I said was that absence of free will doesn't justify inflicting pain on humans any more than existence of free will does."

Pain is not an illusion of course. Let's suppose I have a cancer, this cancer make me suffer a lot, and makes my family suffer too. Tons of pain. I could try to cure that cancer, but, for my personal own reasons, I do not want. in many countries I have the constitutional right to refuse any treatment. Why? Of course because my will of not caring myself is free. In fact if a psychiatrist evaluates that my judgment is not free at all (because of the disease for example) I can be forced to be treated. On one side we have tons of pain (mine, and of my relatives and friends), and pain is measurable using functional magnetic resonance imaging techniques, pain is "scientific", on the other side there is "individual free will", no measure, no free_will term in the big Lagrangian describing the physical processes of my brain, but at the end of the day, in liberal countries, free will wins.
Why? Because of a strange an irrational attitude? I do not think so, I think that is because in human history there is a big bunch of examples of the fact that absence of free will did really justified inflicting pain on humans. It has been already experienced.

Best regards.

p.s.
René Descartes believed that animals could not feel pain and (wikipedia) "Researchers remained unsure into the 1980s as to whether animals experience pain"... but the vast majority, using their illusory common sense, was absolutely sure long before scientific confirmation, that animals do really suffer. Often, in the macroscopic world, commons sense works.

Christine said...

Sabine,

Yes, I understand that you wrote based on "our best present knowledge of the laws of nature". I had the impression, though, that you thought that this knowledge were already sufficient to reach general conclusions on free will, otherwise what would be the point of your text? My opinion is that it is not sufficient, and perhaps it will never be. My previous comment illustrates this somewhat superficially. In any case, although we do not need to converge, it seems we are reaching a clarification on our positions on this matter. That is good enough, I think.

Best,
Christine

Plato Hagel said...

Sabine:The other factor is that a rejection of free will has been claimed to be correlated with lowered moral responsibilityThe Free Will Function

I see through out comments here that you are reasserting yourself because of this survey and it's meaning. So you are saying then that it is possible to have a choice, but no free will?

This paper is dedicated to the attempt to make sense of free will and to argue that one does not need to resort to metaphysics to make place for free will, though the argument will remain far
from claiming that free will is real.
The Free Will Function

The status of observer in the sense with which you may have labelled depersonalization disorder is not relevant here. It may in a sense by it' use, display the metaphysical bend that one is striving here to discard?

It may be a internal struggle then for perceptions about our cosmos and our place in it, versus, the universe as a box and not seeing outside as a possibility? Would we say then that such an observation is a disorder with which we may grasp the context of the universe without our influence on it?

Best,

Franis Engel said...

I like to think of it this way: because physiologically we have veto power to detour or negate what's in the works that we're already preparing for...we haven't got free will. We've just got it backwards. Instead, we've got "Free Won't."

Neo Borocillina said...

Plato Hagel:
"I see through out comments here that you are reasserting yourself because of this survey and it's meaning. So you are saying then that it is possible to have a choice, but no free will?"

It is surely a limit of mine, but I do not see a huge difference between the F(ti) function described in the (as usual interesting) article and a particular markov chain like, for example, a markovian decision process.
I am not sure that a passionate libertarian would really be happy with such a description.

What I do not completely understand is why determinism is supposed to have so much space in the discussion about free will.
Determinism is an "emergent property" itself..like is presumed to be "will": we *want* to make predictions, in some well defined situations, and in those situations a deterministic approach works. Nothing more.

Plato Hagel said...

Hi Neo,

However you configure the space by analogy then, "instead of a cosmological view" you resort too, "your" mathematics describes your approach.

Okay.

Process control(policy), defined by box, and how such a process is exemplified. Perhaps, windows as shown in Bee's post above, that exemplify the construct of the room? This leaves no room for such determinism, yes?

There is just the room, and no starting point.:)

Best,



Plato Hagel said...

A phenomenology, in the most purest sense?:)

Kyle Vitasek said...

Hossenfelder, I've been debating various points of your post w/a friend who holds closely to your train of thought.

We are attempting to get at 'pre-commitments' that we both hold, yet he maintains that his dependence on pure science means he has no pre-commitments.

I willingly admit that my pre-commitments are that there is some order to the universe as i have observed in science and in philosophical readings as well as faith readings.

Can you help us find a common starting point for our debate of your position?

Murphy said...

The exact same reasoning you give in justifying the concept of "different possible futures" also justifies the concept of free will. Although one might believe in the idea that there is "really" only one possible existence past and future, because of physical consistency, nevertheless we use all sorts of concepts like "possibility" and "probability," and we *imagine* multiple possible future states. Your very definition of free will relies on this, so you can't dismiss it. And when we act towards certain imagined futures and away from others, free of certain types of identifiable constraints and influences, we say this is free will. Free never needed to mean, "free from the binds of existence itself" and there is no rational reason to define it thusly.

Don Foster said...

Dr. Bee,

Despite the nodding acceptance here of folks who likely know the science, you should remember that the fabled “calendar riots” of 1752 were provoked by a few missing days on a newly drawn calendar. What will people do when they discover that science has quietly stolen their free will?

Free will aside, is contemporary physics suggesting that the universe is a clockwork progression from first tick to final tock? That every baby’s burp, every bit of turbulence from a bee’s wing, every tone from a jazz saxophone and every impulse from some wet, neural network’s complexity -- this could be a very long list -- that all these phenomenon are causally determined by the state of the wave function of some thirteen-plus billion years ago, back when the universe was still, more or less, getting it’s clothes on?

Such a precise billiards game this envisions, what incredible bank shots. We begin with a scalding broth and end with intricately coiled DNA. (Is it possible to somehow map that early state upon the present?) I can’t help but think there must be more to the story, wonder if physics is actually settled enough to make such a pronouncement.

Perhaps the chemistry of the future is more organic. Walking through any woodland scene one is struck by the degree of material consensus, the diversity of elemental chemistry held in dynamic repose by an intricate weave of systemic relationship -- a myriad of meta-level relationships with radii of correspondence that encompasses an enormous swath of time and space.

Does some general, QM wave function have the bandwidth to encode this profusion of extended relationship and make it deterministically relevant? It seems that, while the quantum world may provide an elegant supporting framework, it is upon the rude boards of these meta-level stages that the real play is enacted and an outcome determined.

The world steps forward with choices made in a myriad of present moments, each one a physical demarcation of future path. While these turnings are largely iterative and routine, I would like a physics that can accommodate and validate the possibility of innovation. I would prefer a narrative in which there is challenge and response, where we are breasting the waves and events evolve upon some lively edge of the present moment. I don’t believe we would have come to this discussion in a clockwork universe.

Regards,

Kakaz said...

I may agree that IF You have a complete knowledge about one mental state and complete information about one environment THEN free will would be an illusion and You may easily predict someone behaviour.

But let me ask how to get such kind of complete ( or nearly complete) knowledge? I suppose it is not possible to get even knowledge about external environement - not to mention someone present mental state.
As we have to use "some macroscopic quantities" there is too many degrees of freedom in our head to describe by them. And "macroscopic mental states" do not give us enough constraints to compute someone behaviour.
So free will is not the effect of "microscopic physics" nor "emergent phenomenon" but rather "emergent lack of constraints" following from the fundamental lack of knowledge. And it is hard to imagine if it is possible to brake such limit.

Rob Steele said...

I would love to hear your take on Jonathan Edwards' Freedom of the Will.

Gang Sun said...

Allah Gold said:"If in fact free will does not exist, and there is only one future, then please explain the purpose of science, exploration, human curiosity and most specifically the reason for life's existence into your theory."
Of course, there is no purpose. Purpose is only illusion. However, there is cause and effect: doing science does satisfy many human needs, and it causes the scientists to do science. On the other hand, pretending that you have a 'purpose' may be a good life strategy, though.

Archie Meijer said...

We can't know anything other than that when we do certain things we call experiments it produces certain observable results. Then we can extrapolate theories to make predictions.

Even if we had free will that would still be doable.

Take the instrumentalist view, that science and the descriptions it yields are just tools that we use and we can't actually determine anything about what is fundamentally real and free will survives.

hush said...

Appreciate Sabine's approach.

Envy her.

Under any label "free will" encompasses, not even minimum requirements to meet demands of/from this unknown have been established.

She attempted this. This puts her ahead of the game all free will proponents of this unknown want to play.

Bravo.

In fact the minimum requirements from the demands for dark matter, energy or quantum gravity makes the search for hamburgers or free will utterly ridiculous.

If you are tired of being a proponent and target of ridicule
follow her suggestion - let go for now.

Michael Mansberg said...

I think that the fact that "I have free will" is axiomatically obvious. Only slightly less so than the fact that "I exist".

Alex Hayden said...

Well, what if DNA has something to do with wiggling out of the block universe rigidity, giving rise to consciousness, which tend is tied to an ability to have free will, choice, or intentional volition over one's environment?

Even though there is the block universe argument, billions of years were put into refining DNA, the cell, the genome, and the abilities of humans in their environment. You could argue that it's all about humans attempting to increase their order around a world/universe that is becoming disordered.

However, I think more research attention needs to be given to DNA, consciousness, entropy, and how all of these things connect in relation to QM and relativity.

Tim Pierce said...

If we have no free will, why do we perceive that we do?

And in that case, why do we perceive at all and what exactly do we mean by 'we'?