Scivillage.com Casual Discussion Science Forum

Full Version: You don’t have free will, but don’t worry. (Sabine Hossenfelder)
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3 4 5 6 7 8 9 10
https://backreaction.blogspot.com/2020/1...worry.html

EXCERPTS (intro): . . . Last week, I explained what differential equations are, and that all laws of nature which we currently know work with those differential equations. These laws have the common property that if you have an initial condition at one moment in time, for example the exact details of the particles in your brain and all your brain’s inputs, then you can calculate what happens at any other moment in time from those initial conditions. This means in a nutshell that the whole story of the universe in every single detail was determined already at the big bang. We are just watching it play out.

[...] A lot of people seem to think this is a philosophical position. They call it “materialism” or “reductionism” and think that giving it a name that ends on -ism is an excuse to not believe it. ... We do not guess, we know that brains are made of particles. And we do not guess, we know, that we can derive from the laws for the constituents what the whole object does. If you make a claim to the contrary, you are contradicting well-established science...

So, the trouble with free will is that according to the laws of nature that we know describe humans on the fundamental level, the future is determined by the present. That the system - in this case, your brain - might be partly chaotic does not make a difference for this conclusion, because chaos is still deterministic. Chaos makes predictions difficult, but the future still follows from the initial condition.

What about quantum mechanics? In quantum mechanics some events are truly random and cannot be predicted. [...] These random events in quantum mechanics are not influenced by you, regardless of exactly what you mean by “you”, because they are not influenced by anything. That’s the whole point of saying they are fundamentally random. Nothing determines their outcome. There is no “will” in this. Not yours and not anybody else’s.

Taken together we therefore have determinism with the occasional, random quantum jump [...nothing...] resembling this intuitive idea that we can somehow choose which possible future becomes real. The reason this idea of free will turns out to be incompatible with the laws of nature is that it never made sense in the first place. You see, that thing you call “free will” should in some sense allow you to choose what you want. But then it’s either determined by what you want, in which case it’s not free, or it’s not determined, in which case it’s not a will.

Now, some have tried to define free will by the “ability to have done otherwise”. But that’s just empty words. If you did one thing, there is no evidence you could have done something else because, well, you didn’t. Really there is always only your fantasy of having done otherwise.

In summary, the idea that we have a free will which gives us the possibility to select among different futures is both incompatible with the laws of nature and logically incoherent. [...] Look at the writing of any philosopher who understand physics, and they will acknowledge this.

But some philosophers insist they want to have something they can call free will, and have therefore tried to redefine it. [...] So, yeah, if you want you can redefine “free will” to mean “no one was able to predict your decision.” But of course your decision was still determined or random regardless of whether someone predicted it. Others have tried to argue that free will means some of your decisions are dominated by processes internal to your brain and not by external influences. But of course your decision was still determined or random, regardless of whether it was dominated by internal or external influences. I find it silly to speak of “free will” in these cases.

I also find it unenlightening to have an argument about the use of words. If you want to define free will in such a way that it is still consistent with the laws of nature, that is fine by me, though I will continue to complain that’s just verbal acrobatics...

[...] Another objection that I’ve heard is that I should not say free will does not exist because that would erode people’s moral behavior. The concern is, you see, that if people knew free will does not exist, then they would think it doesn’t matter what they do. This is of course nonsense. If you act in ways that harm other people, then these other people will take steps to prevent that from happening again...

There have been a few research studies that supposedly showed a relation between priming participants to not believe in free will and them behaving immorally. The problem with these studies, if you look at how they were set up, is that people were not primed to not believe in free will. They were primed to think fatalistically...

[...] Just because free will is an illusion does not mean you are not allowed to use it as a thinking aid. If you lived a happy life so far using your imagined free will, by all means, please keep on doing so.

[...] The result of that thinking is determined, but you still have to do the thinking. That’s your task. That’s why you are here. ... Why am I telling you this? Because I think that people who do not understand that free will is an illusion underestimate how much their decisions are influenced by the information they are exposed to. ... After watching this video, I hope, some of you will realize ... you need to ... pay more attention to cognitive biases and logical fallacies... (MORE)

Jeez, another scientist making claims outside of their field. Any neuroscientist would tell you that there is zero evidence against genuine free will, especially once Libet was debunked. Likewise, a neuroscientist would know that every synapse relies on quantum excitation, a random process, but decisions are not caused by single synapses.
NOTE: Due to a second post made within the span of an hour being automatically merged with a former post by the forum system, these comments were originally included in the the start of the thread. Since a "merge" was going to happen anyway. It has been edit-relocated now that the hour limit has expired, to shorten the length of the first post.
- - - - - -

CYNICAL SINDEE: With regard to language: incompatibilists themselves pre-set what "will" and the adjective "free" means so that the combination is incompatible with determinism. That includes incompatibilists who believe in free will, not just those who don't (like the above).

Setting aside issues of undetermination and any potential problems with claiming there are grand formulae that could simulate or predict all the changes at a large scale and be inclusive of particular, local sub-set developments without need of auxilliary theories and equations which could fudge that idealized portrayal...

...Then... "will" is indeed deterministic from the standpoint of it involving a cognitive system making selections according to its own tendencies and agendas. It is "free" when it can make its decisions or pursue its goals without being severely obstructed by a situation or manipulated like a puppet by another similarly equipped or deliberating agent slash organization (perhaps under threat).

A legal definition of "free will" in the justice system is accordingly practical in that context or should fall out of the demands of everyday life, rather than these metaphysical affairs (i.e., reasoning abstracting principles, generalizations, formulas, or "rational objects" from phenomenal events/entities and granting some Platonic reality or prior-in-rank existential status to them).

Which is to say: No one -- barring self-persecuting fatalists -- wants to be blamed for something they were forced to do against their will or usual nature -- whether via other intelligent agents or accidental/contingent circumstances. And you can bet that even those on the political left (of which many scientists slot) would want racists or suspected racists, sexists, etc to be penalized and potentially brainwashed rehabilitated to conform to the southpaw academic establishment in the future. (I.e., they don't want every shade of "punishment" discarded due to _X_ ultimately not being accountable for what he/she does.)

The idea of "free will" itself does change individual human behavior on a spectrum ranging from light to heavy -- just as a religious orientation or a political one like Marxism will. A concept like free will impacts personal ability, accomplishment, and reform by driving one beyond what would otherwise be a diminished cut-off point. Needless to say, other aspects of advanced civilization such as the idea of "human rights" of various degrees are likewise intellectual inventions -- they're not discovered as residents of non-artificial wilderness.

As John Guidry possibly suggests below, the elimination of individual responsibility would gradually erode or radically alter the justice system, morality, and social customs -- that aforementioned "people will [still] take steps to prevent" you from committing harm.

There could even be a tumbling into more cruel pre-Enlightenment systems as the postmodern or enforced cultural diversity trend of diminishing Western ideas and influence persists. (Contrast that with heterogeneity "naturally" arising locally in people interactions without ideological academic and administrative regulatory enforcement). Ironically, this societal self-mutilation is circularly a Western invention itself (fueled by guilt?), so at the end of the road the system that promoted cultural diversity dies and the latter may no longer be maintained. "Social engineering" inherently treats entire populations as lab rats, since the experimental process by its very nature is uncharted territory (and an ideology's predictions or certainty about the outcomes is purely a priori dogma.)



Free WIll, Determinism, and the Criminal Justice System
https://www.orlandocriminaldefenseattorn...ce-system/

EXCERPTS: [...] As a criminal defense attorney, I am anxious to see whether or not folks who believe we have no free will are willing to dismiss all charges against my clients who may have (God forbid) raped their wife or killed their dog (sometimes pets evoke more emotion than spouses, I’m just saying).

[...] our criminal laws are founded on the notion that if a person is not acting by his free will, the law cannot hold him “accountable for his choices”. There are plenty of other examples of Florida criminal laws that would benefit my clients, should everyone agree that free will is an illusion. For example, confessions cannot not be entered into evidence unless they are made of the defendant’s “own free will”. The term “free will” is contained right there in the definition of numerous legal concepts. Other criminal law concepts would lose their meaning as well, like “premeditation”. Is it realistic to speak of premeditation if freewill doesn’t exist? Is a robot on an assembly line in China premeditating the building of an iPhone? The mere fact that a robot takes several distinct steps to complete a task doesn’t render its actions ‘premeditated’. Such concepts should be purged from our criminal justice system if we’re all just biological robots.

[...] Everyone wants to hold criminals responsible for their actions. This “responsibility” has its foundation in the belief that we all have the free will to choose right from wrong. What if free will is just an illusion, how would that impact the criminal justice system? Free will creates the moral structure that provides the foundation for our criminal justice system. Without it, most punishments in place today must be eliminated completely. Its no secret that I’m a firm believer in free will, but I’m also a firm believer in arguing against it when it helps my clients. That’s what we lawyers do (call me a hypocrite if you like, I can take it). Now, let’s delve into the issues and practical effects of eliminating free will.

We only punish those who are morally responsible for their action. If a driver accidentally runs over a pedestrian-there will be no criminal charges in the death of the pedestrian. This is what we call an “accident”. However, if a husband runs over his wife after an argument, that same pedestrian death now constitutes murder. It was the driver’s “intent” that made one pedestrian death a crime, and the other not. But, what if we examine the husband’s brain, and an MRI discovers a frontal lobe defect that could explain his deviant behavior? Is he still guilty of murder? If such a defect “caused” the husband’s actions, our criminal justice system has laws in place that would label the husband “Not Guilty By Reason of Insanity”. That being said, what happens if “causation” runs deeper than a mere frontal lobe problem? (MORE - details)
I take a largely agnostic view on freewill. At times it feels I have it and can do otherwise than what I choose to do. And at other times I feel caught up in causes from beyond myself. So if freewill exists, it only occurs sometimes with thoughtful deliberate action. In any case, it seems to make us happy and makes for a more moral society assuming it does exist.
I've always kind of liked Sabine Hossenfelder and I've even agreed with many of the things she says. But sorry, I think that she's wrong about this.

(Oct 12, 2020 05:16 PM)C C Wrote: [ -> ]https://backreaction.blogspot.com/2020/1...worry.html

EXCERPTS (intro): . . . Last week, I explained what differential equations are, and that all laws of nature which we currently know work with those differential equations.

I sense some possible circularity there. "Laws of nature" are defined by mathematical physics in such a way that differential equations accurately express them. So triumphantly pulling out the conclusion that all laws of nature are equivalent to these differential equations would just seem to reveal a premise that the physicists themselves baked in at the beginning.

One might want to argue that using differential equations in this role is justified by success in using them to make predictions. But that sort of argument introduces the induction and underdetermination problems.

Quote:These laws have the common property that if you have an initial condition at one moment in time, for example the exact details of the particles in your brain and all your brain’s inputs, then you can calculate what happens at any other moment in time from those initial conditions. This means in a nutshell that the whole story of the universe in every single detail was determined already at the big bang. We are just watching it play out.

Circular if physics decides initially that it's going to model physical reality with one-to-one mathematical functions. The question then would be whether the posited one-to-one correspondence between initial conditions and future states is a feature of reality itself or just a feature of the mathematics that physicists have chosen to model reality. The old map vs territory distinction. Just because maps are flat and printed on paper doesn't mean that those things are going to be true of the territory. Yet maps do capture something of the territory and can indeed be informative. I'm inclined to think of physics' relation to reality in much that way.

Quote:[...] A lot of people seem to think this is a philosophical position.

And it certainly seems to be.

Quote:They call it “materialism” or “reductionism” and think that giving it a name that ends on -ism is an excuse to not believe it.

We have to call it something. The tradition is to call doctrinal positions '-isms'. Whether or not we should believe them depends of whether they make sense, contradict experience or are justified convincingly.

Quote:... We do not guess, we know that brains are made of particles. And we do not guess, we know, that we can derive from the laws for the constituents what the whole object does. If you make a claim to the contrary, you are contradicting well-established science...

Her invocations of "we do not guess, we know" doesn't seem to derive from science at all. It's half-baked lay epistemology.

Quote:So, the trouble with free will is that according to the laws of nature that we know describe humans on the fundamental level, the future is determined by the present. That the system - in this case, your brain - might be partly chaotic does not make a difference for this conclusion, because chaos is still deterministic. Chaos makes predictions difficult, but the future still follows from the initial condition.

Chaos may be deterministic, but doesn't it also say that even infinitesimal differences in initial conditions can lead to dramatically different evolutionary histories? If that's correct, then the question would seem to hinge on the more metaphysical issue of whether reality is always precisely defined to any level of precision, or whether it's kind of fuzzy and indistinct when we try to be too precise. If there is any fuzziness at all, any superimposed states, quantum probabilities or wave function collapses, then chaos might destroy the one-to-one correspondence between past and future states that Sabine is assuming.

Quote:What about quantum mechanics? In quantum mechanics some events are truly random and cannot be predicted. [...] These random events in quantum mechanics are not influenced by you, regardless of exactly what you mean by “you”, because they are not influenced by anything. That’s the whole point of saying they are fundamentally random. Nothing determines their outcome. There is no “will” in this. Not yours and not anybody else’s.

Here is where we need a better definition of "free will". When we say that we acted freely, we mean that we chose the action and that it wasn't imposed on us by some outside force. Free will isn't synonymous with behaving randomly. It isn't the same thing as epileptic seizures. A free action is an action that's produced by our own motivations, informed by our desires, our knowledge and all kinds of mental states like that.

Where did those mental states come from? From earlier mental states and from the environment. So any reasonable concept of free will will have to acknowledge temporally short-term determinism. I don't think that most champions of free will would want to argue with that. Why did I behave as I did? Because I wanted to. Why did I want to? Because of my desires and my understanding of the situation. Why did I have those desires and that understanding? Because of my history.

Where champions of free will start to object is where people like Sabine argue that our mental states now as we will ourselves to take some action were all determined by the state of the universe long before any of us were born. Sabine said it herself up above: "This means in a nutshell that the whole story of the universe in every single detail was determined already at the big bang." It's a form of creationism, except with the idea of divine purpose removed.

My alternative would be that determinism is probably quite accurate if we are talking brain states in milliseconds. If we know the prior state we can accurately predict the later one. As I argued above, our concept of free will depends on this being so. (Since I determine my own actions when they are freely willed.) But knowing my brain states ten years ago wouldn't be much help in predicting what I am going to do today. Knowing the state of the universe 15 billion years ago wouldn't enable any prediction that the Earth or human beings would someday exist, let alone me personally, let alone this particular action.   

Quote:Taken together we therefore have determinism with the occasional, random quantum jump [...nothing...] resembling this intuitive idea that we can somehow choose which possible future becomes real.

If we combine the probabilistic nature of quantum mechanics with chaos, we seem to have a decent argument against temporally long-range determinism. The past wouldn't seem to determine the future in any detail. There seems to be a fuzzy imprecision built into the evolution of future states from past states. Hence Sabine's differential equations metaphysics may not be the best model of what's happening.

Quote:The reason this idea of free will turns out to be incompatible with the laws of nature is that it never made sense in the first place. You see, that thing you call “free will” should in some sense allow you to choose what you want. But then it’s either determined by what you want, in which case it’s not free, or it’s not determined, in which case it’s not a will.

As I just explained, proponents of free will needn't be pushed into embracing the caricature that she's so busily attacking. They will be perfectly happy saying 'I did X because I wanted to. Why did I want to? Because I have particular desires and understood the situation in particular ways.

In other words, free will not only doesn't deny temporally short-range determinism, it depends on it. What free will does deny is that the state of the universe long before any of us were born, even the initial state of the universe at its origin event, has determined everything to come and is pulling everyone's strings as if we were merely puppets.

That's just bad metaphysics.
As usual on such subjects, nice post, Yaz.

(Oct 12, 2020 09:29 PM)Yazata Wrote: [ -> ]In other words, free will not only doesn't deny temporally short-range determinism, it depends on it. What free will does deny is that the state of the universe long before any of us were born, even the initial state of the universe at its origin event, has determined everything to come and is pulling everyone's strings as if we were merely puppets.

That's just bad metaphysics.

I'm not even sure how the most supernatural notion of free will is supposed to exist and act without a significant degree of determinism in the world. I mean, if the world were capriciously random, how would we even distinguish our choices? Would we even have a concept of the self-consistent? Could we conceptualize free will in a world where our choices have no meaning because their outcomes don't necessarily correspond with our intentions?
Yazata Wrote:Chaos may be deterministic, but doesn't it also say that even infinitesimal differences in initial conditions can lead to dramatically different evolutionary histories? If that's correct, then the question would seem to hinge on the more metaphysical issue of whether reality is always precisely defined to any level of precision, or whether it's kind of fuzzy and indistinct when we try to be too precise. If there is any fuzziness at all, any superimposed states, quantum probabilities or wave function collapses, then chaos might destroy the one-to-one correspondence between past and future states that Sabine is assuming.

It sounds pretty but as the story goes...

"Chaos theory is sometimes pointed out as a contradiction to Laplace's demon: it describes how a deterministic system can nonetheless exhibit behavior that is impossible to predict: as in the butterfly effect, minor variations between the starting conditions of two systems can result in major differences. While this explains unpredictability in practical cases, applying it to Laplace's case is questionable: under the strict demon hypothesis all details are known—to infinite precision—and therefore variations in starting conditions are non-existent. Put another way: Chaos theory is applicable when knowledge of the system is imperfect whereas Laplace's demon assumes perfect knowledge of the system, therefore chaos theory and Laplace's demon are actually compatible with each other."

https://en.wikipedia.org/wiki/Laplace's_...aos_theory

And like she said, you still don't have free will over random events.

I don't know. She sounds pretty convincing to me.
(Oct 12, 2020 10:46 PM)Secular Sanity Wrote: [ -> ]It sounds pretty but as the story goes...

"Chaos theory is sometimes pointed out as a contradiction to Laplace's demon: it describes how a deterministic system can nonetheless exhibit behavior that is impossible to predict: as in the butterfly effect, minor variations between the starting conditions of two systems can result in major differences. While this explains unpredictability in practical cases, applying it to Laplace's case is questionable: under the strict demon hypothesis all details are known—to infinite precision—and therefore variations in starting conditions are non-existent. Put another way: Chaos theory is applicable when knowledge of the system is imperfect whereas Laplace's demon assumes perfect knowledge of the system, therefore chaos theory and Laplace's demon are actually compatible with each other."

https://en.wikipedia.org/wiki/Laplace's_...aos_theory

And like she said, you still don't have free will over random events.

I don't know. She sounds pretty convincing to me.
Backing up just one paragraph in your reference:

Due to its canonical assumption of determinism, Laplace's demon is incompatible with the Copenhagen interpretation, which stipulates indeterminacy. The interpretation of quantum mechanics is still very much open for debate and there are many who take opposing views (such as the Many Worlds Interpretation and the de Broglie–Bohm interpretation).
https://en.wikipedia.org/wiki/Laplace's_...ersibility


IOW, due to the advent of QM, Laplace's demon is fundamentally moot. There is no possibility for perfect knowledge, even in principle, according to modern physics. Granted, that still leaves random inputs from quantum indeterminacy. But that's only a problem for free will if a choice relies on a single quantum event. Just like a collection of stochastic results can form a very clear interference pattern of bars, perhaps the over 1 quadrillion indeterministic firing of synapses can result in a truly free choice.

Ultimately, there is no answer to this question in science alone. Believing someone who asserts there is is unscientific hubris.
(Oct 12, 2020 10:46 PM)Secular Sanity Wrote: [ -> ]
Yazata Wrote:Chaos may be deterministic, but doesn't it also say that even infinitesimal differences in initial conditions can lead to dramatically different evolutionary histories? If that's correct, then the question would seem to hinge on the more metaphysical issue of whether reality is always precisely defined to any level of precision, or whether it's kind of fuzzy and indistinct when we try to be too precise. If there is any fuzziness at all, any superimposed states, quantum probabilities or wave function collapses, then chaos might destroy the one-to-one correspondence between past and future states that Sabine is assuming.

It sounds pretty but as the story goes...

"Chaos theory is sometimes pointed out as a contradiction to Laplace's demon: it describes how a deterministic system can nonetheless exhibit behavior that is impossible to predict: as in the butterfly effect, minor variations between the starting conditions of two systems can result in major differences. While this explains unpredictability in practical cases, applying it to Laplace's case is questionable

It seems to require an additional premise.

Quote:under the strict demon hypothesis all details are known—to infinite precision—and therefore variations in starting conditions are non-existent.

Yes, that's where the crux lies. It isn't necessarily a problem of what is known, though it might be for physicists trying to predict things. It's more a problem of what is knowable in principle. What is reality like? Is reality really absolutely precise like an infinitely high def photographic image? Or is reality kind of fuzzy and indistinct when examined too close? In other words, do Sabine's physical variables really have infinitely precise values? (Even though there might always be limitations on the accuracy of a physicist's measurements.)

Quote:Put another way: Chaos theory is applicable when knowledge of the system is imperfect whereas Laplace's demon assumes perfect knowledge of the system

The metaphysical assumption seems to be sneaking in that reality is infinitely precise and that any imprecision is the result of defects in the physicist's knowledge. But do we really know that reality is an infinitely sharp picture?

That's where quantum mechanics enters the argument and joins chaos. QM (at least as a layman like myself understands it) seems to say that some physical variables can behave probabilistically, can have more than one superimposed value simultaneously, or can even be undefined and lack a value.

That supplies the fuzziness. Chaos serves to amplify some of the implications of that fuzziness from the quantum microscale onto the classical mesoscale. Which would serve to subvert the kind of one-to-one correspondence between prior events and subsequent events that Sabine's differential equations metaphysics seems to want to assume.

I guess that I'm arguing for a sort of compatibilism, arguing that free will is compatible with temporally short-range determinism (it even requires it, if we are to say "I did it because I wanted to"). That move requires introduction of some better definition of 'free-will'. Free-will seems to me to require that prior mental states (intentions) need to have a tight short-range temporal/causal linkage to subsequent mental states (volitions). We want to distinguish intentional willed actions from spastic convulsions. On the free-will thesis, our actions are determined in some large part by our minds (which might indeed be described as an inner process in entirely naturalistic neurophysiological terms).

I suspect that the precise one-to-one sort of linkage between earlier and later states of the universe weakens dramatically as the temporal gap between causes and effects increases. Put another way, if we reran the Big Bang with precisely the same physical "laws" and initial conditions as close as physically possible, things might turn out very different than what we see today.

This seems to be very consistent with what everyone (even physicists) experience in their own lives. And that's a point in its favor.
(Oct 13, 2020 04:07 AM)Yazata Wrote: [ -> ][...] I guess that I'm arguing for a sort of compatibilism...


I've been a compatibilist for quite a while. I had a reflexive aversion to incompatibilism since it subsumes both the acceptance of free will (libertarianism) and the rejection of free will (hard determinism, hard incompatibilism, etc). Incompatibilism seems devoted to making free will as complicated as possible by making it dependent upon metaphysical baggage. Compatibilism is more like: "Who cares? What difference does that ___ make?"

Two common definitions from a dictionary and Wikipedia:

"Free will is the power of making free choices unconstrained by external agencies." .... "Free will is the ability to choose between different possible courses of action unimpeded."

Where is the shrill, incompatibilist-sounding demand that free will requires magic or something, that choices cannot be outputted by even classic, utterly predictable clockwork mechanisms?

It's the human body processing information and making its selections, not everything in the environment but the body doing that. Rocks don't have the capacity to choose or decide or scheme to control. Nor do chairs or clouds or lakes or the Moon. They don't even simulate/represent the circumstances of an external environment. They don't have interests.

Has anyone in their entire life NEVER been allowed to do something that they actually chose and wanted to do because there was ALWAYS an "external agency" preventing them from doing it? Is that not some kind of animism, to have this pervasive determinism paranoia that stupid, non-conscious agencies ranging from the earliest state of the cosmos to "universal laws" to local, non-living surroundings are out to deprive an autonomous organism of the capacity to generate its own volition? That its own structure and operation is not essential for making that possible?

We know what lack of "free will" is in everyday life when we're forced to do something we don't want to do. That's not fictional, we can cognize such an occurrence and later even point to the times and places it transpired. Since "lack of free will" acquires its very significance via denoting its missing _X_, the latter is substantiated in turn. Neither status is continually the case (to do what we want or not be able to do what we want) -- there would be no need of conceptually discriminating either without the opposite (at best the latter might fall out of a speculative possibility).
Pages: 1 2 3 4 5 6 7 8 9 10