Hacker News Reader

> Either the universe is fully deterministic and free will and agency do not exist, and we should be much kinder to people who e.g. commit crimes (they were forced to do it by their brain state, after all)...

I don't think that's consistent, as the same reasoning applies to the people (i.e. the 'we' in your statement) who will, in some way or perhaps not at all, censure the perpetrator of said crime.

In other words, if full determinism removes agency from the perpetrator, then it also removes it from everyone else.

Additionally, the possibility of consequences (including those created and enforced by human beings, formal or otherwise) are also factors in the "deterministic" action of an individual, so for the benefit of society (assuming there are outcomes that society can improve via expectations and consequences) then it makes perfect sense for there to be consequences.

Bottom line though, whether the system is deterministic or not doesn't affect our experience of it.

> In other words, if full determinism removes agency from the perpetrator, then it also removes it from everyone else.

This is irrelevant to their claim. Normative statements don't require anyone actually have agency. After all, it's possible that everyone should do something that nobody can do.

If you want to say that fine then their claim is irrelevant too, then sure, but it's in no way inconsistent. I think it is relevant though. We can do things differently than we do them, like after reading this comment section; we just can't do them differently than we will do.

Worse still. Your guilty of the same thing you're accusing the poster of.

Are you sure about that? I don't think I'm taking issue with a normative claim here, I think I'm discussing the rationale offered for adopting it, which is where agency came into the discussion.

I agree that I am doing the same thing as the person I am replying to, just applying it to all participants (which, by the way, is not in any way an accusation.)

Perhaps we’re on our way to the inevitable outcome of being much nicer to criminals by pure coincidence of determinism :)

Nah. People who get het up about free will are just making some conceptual error, which I think goes like this;

1. Machines don't have free will.

2. If my brain is explained by conventional physics, it's a mechanism.

3. But I have free will!

4. So there's something wrong with physics.

But the mistake is in step 1, machines can have free will - at least, we do, so QED and STFU.

Why can't free will be deterministic? What does randomness have to do with anything? This is about thought.

Many would say you can't have free will in a deterministic world because free will is about selecting one option from many. If there is no alternative option -- because the universe is completely determined -- how can you have any choice? You were always going to do exactly what it was previously determined you were always going to do.

They want "agency", which means something like looking at the universe and interfering with it without being part of it, like a scientist with the universe in a petri dish, somehow external to any universe and not subject to any rules or causality.

It's a bizarre but apparently intuitive wish. But this isn't necessary. A robot can select one option from among many. We can predict the robots choice. A human can do the same, and we can't predict the human's choice so well, because humans are deep and complex and actually think, but it's still a mechanism, and so what? The error here is in thinking that "mechanism = robot" and "mechanism = amoral", neither of which are true - except for all the non-human mechanisms, which is the great majority of the mechanisms (processes, etc.) that we see, and we don't like being lumped in with them in case it makes us robots and removes responsibility for choices. But it just doesn't.

And yet, knowing that free will does not exist can change someone's whole life trajectory. So free will is a useful delusion. But we can't control if we get to have the delusion. It's not surprising if this like of reasoning seems familiar.

Free will does exist, deterministic agency exists, and reasoning that determinism removes moral responsibility is an error. What doesn't exist is this metaphysical type of god-like beyond-universe agency that people obsess over for no obvious reason.

Free will is not a delusion it is an illusion. The difference is significant. Illusions are normal sensory perceptions common across nearly all humans that just happen not to correspond to anything in physical reality.

> Many would say you can't have free will in a deterministic world because free will is about selecting one option from many. If there is no alternative option -- because the universe is completely determined -- how can you have any choice?

Because "select 1 option among many" is not what's meant by free will and not how anyone serious understands the term. If someone is holding a gun to your head and telling you to choose option X, which is an option that you would NOT otherwise choose, you still have a choice by your definition, because you can choose to die or choose to accede to those demands. No one would agree that you have a freely willed choice though. Clearly you're missing some important ingredient, and this extra bit is why free will is compatible with determinism.

Some people in those situations have absolutely chosen to get shot, which means that true determinism functions differently than a metaphorical gun. You might not be morally culpable for what you do while under threat of death, but that's a separate question from whether you could physically have done something different.

In a deterministic universe there isn't a gun to your head, instead what you will do was decided by the configuration of atoms immediately after the Big Bang (and presumably by the configuration of whatever the heck came before). In a deterministic universe I do what I do because particles hit into each other in particular ways over countless eons and those particle interactions eventually coalesced into what I call myself and the particles in my brain bounce in particular ways that interact to create an illusion of choice.

> You might not be morally culpable for what you do while under threat of death, but that's a separate question from whether you could physically have done something different.

The principle of alternate possibilities was debunked by the Frankfurt cases, and that was not the point of the gun scenario.

The point was that nobody would consider a person acting under threat of death to be making freely willed choices, so defining choice in the naive and reductive sense that was suggested is just incorrect because it cannot exclude this case. Therefore this naive and reductive definition cannot be what people mean by "choice" in the context of free will.

> In a deterministic universe I do what I do because particles hit into each other in particular ways over countless eons and those particle interactions eventually coalesced into what I call myself and the particles in my brain bounce in particular ways that interact

So an intelligent being was created with certain preferences and values, and as long as this being was able to deliberate and make choices in accordance with those preferences and values, that being was making freely willed choices. How this being was able to do this at the subatomic level is completely irrelevant, and bringing it up is, at best, a category error.

> The point was that nobody would consider a person acting under threat of death to be making freely willed choices, so defining choice in the naive and reductive sense that was suggested is just incorrect because it cannot exclude this case. Therefore this naive and reductive definition cannot be what people mean by "choice" in the context of free will.

I disagree with this characterization as it omits entirely the moral culpability central to Frankfurt's argumentation. Frankfurt cases are about moral culpability[1]. I did not refer to morality or culpability but only the base naive case given that the question was "why can't you have both?" and I think the epistemic force of the base case is nontrivial.

I think your error here is that you are presuming:

(a) If you are acting under threat of death, you cannot do otherwise but save your life. (b) When you define freedom as the ability to have done otherwise, you cannot be free in the case of (a). (c) All cases of human choice must be free. (d) Given (b) and (c), the definition must be incorrect.

As the other person in the thread pointed out, some people can choose death in the face of death or some other choice. It has happened many times throughout history. So saying "you must save your life" is outright false. Some people choose one option, some people choose another. This has no force on the definition of freedom.

Even if you say in that specific case a person is not free because they could not do otherwise, it does not follow that the definition of freedom can't handle the case. It may follow that the person is not blameworthy - which is what Frankfurt was getting at - but that's nowhere near saying that the theory can't handle the case at all. Some might say the person under threat of death could do otherwise, some might say the person under threat of death could not do otherwise. That's not fatal to anything, it just shows the definition is incomplete.

[1] Wikipedia covers it somewhat here: https://en.wikipedia.org/wiki/Frankfurt_cases#Frankfurt's_ob... but if you don't like Wikipedia for some reason, the intro to this 2007 paper https://digitalcommons.unf.edu/cgi/viewcontent.cgi?article=1... gives more or less the same description.

i've never understood why free will fits into discussions about deterministic vs random universes. in either situation, there isn't room for free will -- if things are deterministic your decisions are made by the state of the world at t-1, if things are random your decisions are made by coin flips.

the more i've thought about this over the years, the less clearly i've been able to define what precisely 'free will' would mean in terms of a physical system or causality

Agreed, there are to "steps" we need.

First, we need the universe to not literally just be a function of the previous state and the current time.

Second, we need (and I'm quoting myself from above to illustrate what I meant): "some kind of metaphorical "door" for our soul, or our true selves, or whatever you might call it, to propagate through meaningfully into our consciousnesses".

Lacking the first means there's no way for the second to exist. Supposing the first exists, we then need it to provide some kind of way for the "us" that exists outside of physics or as an emergent phenomenon out of the nondeterministic parts of physics to meaningfully shape what our brain's electrical states make us do.

If we have all of that, we finally have the beginnings of a system in which we can start to try to philosophically assign moral designations to actions and people. Which is think is about 50% of the reason people want free will to exist: so we can punish others for their actions.

(And just to clarify, to anyone not familiar with how people usually debate philosophy. It's all about your presupposed axioms. Things you and your partner agree to be true. This kind of discussion is one that presupposes relatively little, so don't think I'm trying to make political claims about how crime shouldn't be punished or whatever. I'm simply saying that if we presuppose only that the universe is not deterministic, that alone doesn't bring us forward enough to start building a system of ethics.)

Yesterday I complained that telepathy is defined solely as a negative, which means that it essentially can't exist because if it is ever found, it will be real, and as such, will have mechanisms and limitations and real characteristics, and therefore it can't be "real telepathy". Because the definition is intrinsically built on negation.

    def isItTelepathy(anything):
        return False

A lot of obfuscatory words are thrown at the definition, but it's what it amounts to.

Or, more fairly, this is not the only definition but it's a very popular one.

Free will has a similar problem. The definition a lot of people want to use is a negative one. It can't involve any predictable process. It can't involve any seen process. And so on. Well, when you feed a definition like the above into logic, you get some pretty broken answers.

Some people do give definitions that can be turned into real answers, but you can detect someone operating on the completely-negative definition by the way they get upset, sometimes even in a moralistic way, at the idea that you might have said something meaningful and insist on pushing their total-negation definition at you again.

There's an interesting conversation to be had on the matter in potentia but it requires so much effort sidelining the people who become offended if you actually proffer a definition and hold to it and perform actual logic on it that it's hardly worth it. So we pretty much get nowhere because we are essentially not allowed by about 80% of the people trying to have the conversation to have a concrete definition at all, not even for a moment, not even "for the sake of argument".

> in either situation, there isn't room for free will -- if things are deterministic your decisions are made by the state of the world at t-1, if things are random your decisions are made by coin flips.

Yes, but what does this have to do with free will as most people understand the term? They want to be responsible for their choices, but this is compatible with a deterministic universe.

> They want to be responsible for their choices, but this is compatible with a deterministic universe.

This is a controversial claim. How can someone be responsible for something they have no control over? I'd agree that's possible, but I don't think that's how most people understand it.

You a free to make choices in a deterministic universe. It was just always going to be the choice you made.

Introducing randomness just eliminates that determanism, not the ability to choose.

> This is a controversial claim. How can someone be responsible for something they have no control over?

The Frankfurt cases showed that control over alternate possibilities is not necessary for responsibility. If someone holds a gun to your head to make you do something that you were going to do anyway, that's not really compromising your free will is it?

Edit: it's also not really as controversial as you might think, experimental philosophy has demonstrated that most lay people's moral reasoning aligns with Compatibilism: https://philarchive.org/rec/ANDWCI-3

It's exactly as controversial as I think because I think it's as controversial as we have measured. https://survey2020.philpeople.org/survey/results/4838

Lay people are incoherent and childishly naive. I don't really care what they think.

> The Frankfurt cases showed that control over alternate possibilities is not necessary for responsibility.

This is a controversial claim. There are many popular counterarguments.

> It's exactly as controversial as I think because I think it's as controversial as we have measured.

I wouldn't consider 60% agreement to be that controversial frankly, especially when the next most popular contender at 18% is the incoherent mess that is libertarian free will.

> Lay people are incoherent and childishly naive. I don't really care what they think.

You should, because 1) many trying to dismiss free will use lay people's alleged incoherent definitions, but that's the wrong way to question lay people because they don't think in denotational terms, and if you question them properly they have nuanced and coherent approaches to moral reasoning, and 2) making sense of free will is all about how lay people talk and reason about these concepts, and the point of the philosophical debate over free will is whether there is a coherent definition that can make sense of these arguments.

Furthermore, the paper I linked is directly relevant to your post because you made the claim that it's not how most people understand the term, and that paper casts some serious doubt on that position.

> This is a controversial claim. There are many popular counterarguments.

"Popular" quantified how exactly? Frankly this is even less controversial than the acceptance of Compatibilism, so I won't lose any sleep over it.

Touché.

That's the point. Free will is a metaphysical concept that posits the existence of a soul outside physics.

Randomness in physics is a peephole through which the non physical world can interact with physics. Think of it like a game where you get control the RNG output for some rolls.

The only way to satisfy extreme notions of free will is to introduce supernatural agency, e.g. a soul granted by God. This way your choices are neither physically pre-determined, nor random.

And in fact when you dig deep into discussions with some people, it becomes clear that defending and proselytizing the existence of a soul (or similar supernatural concept) is in fact their goal in discussing free will in the first place. They think it's a useful way to back into their preferred version of spirituality, or they think it's the logical resolution to the contradiction you've highlighted, or maybe both.

I assume that for some, the soul is an analogy for free will, which is why it’s so hard for them to separate the ideas

Yeah I share the same stance. Im not sure there is even a point in knowing or caring about freewill regardless of whether it exists or not or is just some completely meaningless abstract concept.

At the end of the day, im still going to act like me. From what I can imagine the most knowing the answer could ever accomplish is induce existential crisis in some people.

> ...we __should__ be much kinder to people...

This implies a choice. If there is no free will, all you can say is we will or we won't based on our current brain state. Sure, we don't know what that outcome will be, but it's already determined by our own current brain state.

> (they were forced to do it by their brain state, after all)

Any any punishment from cruel and unusual or no punishment at all was forced by the brain states of the people in a position to decide and carry out any punishment. Again, without any free will, any talk about what "we should do" is pointless: it's determined.

Your much kinder argument I also think falls completely flat because I really think you end up in a binary: either you are culpable for your actions or not. If determinism excuses your dastardly acts in part because of lack of agency, what possible argument remains for not simply pretending that no wrong ever happened? On the other side, if a society living in a deterministic world still receives benefit from deterrence and retribution, then there would be no difference than if we're living in a world with free agency: so determinism ceases to be an argument about the degree we punish.

I do actually agree with you at one point: either the world is fundamentally deterministic or it isn't. I do think something approaching free will may possibly be an emergent feature of our world, much like a single hemisphere has no roll, but putting two together and rolling emerges. But even that, at the root would be fundamentally deterministic.

I would also suggest this. Whether or not free will is a fundamental feature of the world we live in or simply an illusion due to our inability to follow our own brains from one state to the next: you better live your live like free will is a thing and not worry too much about how predestined your actions are: if free will is an illusion, that illusion is strong... the outcomes will be determined by states that include the influences of free will like assumptions, or the absence thereof. If free will is a thing: you're covered. Of course giving this advice presumes you have a choice in taking it or not... if there is no free will you either will or won't and the only questions arise out of our inability to project into what we call the future.

free will and agency are two very different terms. Free will is a philosophically charged term that is often incoherent or ill defined, agency is fairly straight forward, it's the capacity of any actor to realize their goals and act autonomously in their own interest. It's the ability to shape your surroundings rather than be shaped by them.

There's plenty of agency in a deterministic universe, a dictator with a nuke has a lot more of it than a slave in a cage. There's no need to be nicer to criminals on grounds that the universe is deterministic, which it is. (both in Newtonian physics and quantum mechanics). The nice guy is as deterministic as the criminal. None of them are "forced to do" anything by their brain state, there is no subject that is puppeteered around by your brain like some reverse dualism, you and I are our brains (and bodies). What matters is if you want criminals to have more agency, and if you don't, you punish them.

> What matters is if you want criminals to have more agency, and if you don’t, you punish them.

In a deterministic universe, “want” doesn’t mean much. That you would decide to punish a criminal is the only thing that could happen.

No, again this is the same fallacy. Wants, needs, desires and preferences mean exactly what they meant before you figured out that the universe is deterministic. The misconception is simply not understanding that these are conceptual properties of material agents, not metaphysical claims about causality.

You wanted an ice cream five minutes ago, reading a book about determinism didn't change that want. We actually do want to punish criminals and therefore we should do it. Whether we all run on cellular automata at the bottom levels of physics is just completely irrelevant to discussing human interaction.

Okay, I see what you are saying.

If I was a ball, I’d want to roll downhill. If I chose to roll uphill it might be like me choosing to not have the ice cream and then do anyway because I love ice cream.

> the universe is deterministic, which it is. (both in Newtonian physics and quantum mechanics).

Bold claim. Have there been some advancements in our understanding, that I'm not aware of?

Compatibilitism is the the theory that player characters (conscious people)and NPCs (philosophical zombies) are morally equivalent. That the inability to see someone's (even your own!) inner motivating state means that it's equivalent to free will in all practical senses.

In other words, "The difference between theory and practice is that, in theory, theory and practice are the same. In practice, they are not", and all the paradox that entails.

> Compatibilitism is the the theory that player characters (conscious people)and NPCs (philosophical zombies) are morally equivalent.

Compatibilism has literally nothing to say about consciousness. It's a completely orthogonal question. There's nothing incoherent about deterministic p-zombies or non-deterministic p-zombies, nor is there anything incoherent about deterministic conscious people or non-deterministic conscious people.

The idea of philosophical zombies has something in common with solipsism, which is where everybody else is a zombie, and hey, maybe the universe is a simulation too. But the argument against this is that if the universe (and its population) is perfectly simulated, then it's exactly equivalent to real, and that means it is real - or the zombie is a real person, if we're just discussing one "perfectly simulated" person. Otherwise, if the simulation is just very very good, we're talking about something less relevant to arguments.

A useful illusion of Free Will is that we can to some degree predict the future state of the world and influence the future state of the world. To make meaningful agent type activity you have to have a world full of causality, in other words --- and to have causality it is awfully nice to have some deterministic properties (or science might not work out so well). So! I for one am a hard-compatiblist. I think that not only is "useful free-will" aka Agency compatible with determinism, I think that determinism is necessary!

(concepts probably stolen or bad copies of Daniel Dennett models)

It's clear that huge swathes of physics are deterministic! Or at least anything non-deterministic is on a small-enough order that we can still build rockets and keep clocks on satellites and earth in sync.

And I agree that a world completely free of causality would be a useless world. But I think that a world that is completely closed under physics and completely deterministic isn't compatible with a "free will" that could support a system of ethics, which I think is where a lot of people then want to take the "does free will exist" argument once they figure out the first part. So it's a matter of threading the needle.