Free Will Vs Determinism

[Straw Dog]

 

[Straw Dog]

I had a friend today tell me that she thought determinism was the same as astrology. “I did what I did because mercury is in retrograde”. Of course, I disagreed and counter-argued to a degree.

I can’t help but wonder though... given that determinism cannot be scientifically proven beyond a shadow of a doubt, with quantum mechanics, computational science, and uncertainty, it does seem like a philosophical faith. There is a measure of sound reason to believe it, but it’s still just a scientifically unverifiable belief. It’s almost like we replaced the loss of our religious or mystical faith with a philosophical faith, which we may defend in a similarly zealous manner.

 

[Skwim]

... given that determinism cannot be scientifically proven beyond a shadow of a doubt, with quantum mechanics, computational science, and uncertainty, it does seem like a philosophical faith.

I take it then that free will can be scientifically proven beyond a shadow of a doubt. That about it?

There is a measure of sound reason to believe it, but it’s still just a scientifically unverifiable belief.

What science is that? As far as I and everyone else who understands the doctrine contend, determinism isn't a scientific concept, but a philosophical one. You do understand the difference between the two don't you?

.

 

[Ayjaydee]

Several years ago, eight to be exact, I posted my reason for dismissing the idea of free will and adopting hard determinism. Because the topic of free will vs determinism hasn't been discussed in some time and a lot of newcomers have come on board RF I thought I'd bring it up again. The following is taken from my original post.



Discussions about free will usually center around an affirmation and/or a denunciation of it. Very interesting thoughts on both sides have come out of such conversations, many well thought out, others not so much. Whatever the case, there's frequently been a problem with what is meant by "will" and "free will," so much so that the issue can quickly become mired in misunderstanding. To avoid this I've found the following definitions to be on point and helpful.

Will is the capacity to act decisively on one's desires.

Free will is to do so undirected by controlling influences.


The notion of free will is important to many because without it would mean each of us is nothing more than an automaton, a machine that performs a function according to a predetermined set of instructions, which is anathema to the notion personal freedom. If people lack freedom of choice how can they be blamed for what they do, or be deserving of any praise laid on them? For Christians this has the added consequence of robbing the concept of sin/salvation of any meaning. So most people are loath to even entertain the idea of no free will. Free will is almost always regarded as a given.

Any exception to free will is regarded as temporary constraint. "I am free to to do this or that unless someone/thing comes and prevents it. Of course this isn't what the free will issue is about at all. Free will is about the idea that, aside from any external constraints, "I could have chosen to do differently if I wished." So I think another valid way way of looking at free will is just that: the ability to do differently if one wished. "I got a haircut yesterday, but I could just as well have had a hot dog instead."

Those who most ardently disagree with this are the hard determinists, people claiming that everything we do has a cause. And because everything we do is caused we could not have done differently---no, you could not have chosen to have a hot dog--- therefore it's absurd to place blame or praise. A pretty drastic notion, and one rejected by almost everyone. So whatever else is said about the issue of free will ultimately it must come down to this very basic question: Are we free to do other than what we chose or not? I say, No you are not. Free will is an illusion. But before going into why, we first need to get rid of the term "choice" because it assumes to be true the condition under consideration, freedom to do what we want. So no use of "choice" or any of its cognates.


Here's how I see it.

There are only two ways actions can take place; completely randomly, or caused. By "completely randomly" I mean absolutely and utterly random, not an action which, for some reason, we do not or cannot determine a cause. This excludes things such as the "random" roll of dice. Dice land as they do because of the laws of physics, and although we may not be able to identify and calculate how dice land, it doesn't mean that the end result is not caused. This is the most common notion of "random" events: those we are unable to predict and appear to come about by pure chance. The only place where true randomness, an absolutely uncaused event, has been suggested to occur is at the subatomic level, which has no effect on super-atomic events, those at which we operate. And I don't think anyone would suggest that's how we operate anyway, completely randomly: what we do is for absolutely no reason whatsoever. So that leaves non-randomness as the operative agent of our actions. We do this or that because. . . . And the "cause" in "because" is telling. It signals a deterministic operation at work. What we do is determined by something. Were it not, what we do would be absolutely random in nature: for absolutely no reason at all. But as all of us claim from time to time, we do have reasons for what we do. And these reasons are the causes that easily negate randomness.

So, because what we do obviously has a cause, could we have done differently? Not unless at least one of the causal determinants leading up to the event in question had been different. If I end up at home after going for a walk it would be impossible to end up at my neighbor's house if I took the exact same route. Of course I could take a different route and still wind up at home, but I would still be in the same position of not ending up at my neighbor's. To do that there would have had to be a different set of circumstances (causes) at work. But there weren't so I had no option but to wind up at home. The previous chain of cause/effects inexorably determined where I ended up. So to is it with our decisions. We do what we do because all the relevant preceding cause/effect events inexorably led up to that very act and no other. We HAD to do what we did. There was no freedom to do any differently.

What does this all mean then? It means that we can never do any any differently other than what we are caused to do. Our actions are caused (determined) by previous events and intervening outside events (also causes) and nothing else. Even our wishing to think we could have done otherwise is a mental event that was determined by all the cause/effect events that led to it. We think as we do because. . . . And that "because" can never be any different than what it is. We have no will to do anything other than what we're caused to do. In effect then, free will does not exist, nor does choice, etc..

This means that blame and praise come out as pretty hollow concepts. As I mentioned, if you cannot do other than what you did why should you be blamed or praised for them? To do so is like blaming or praising a rock for where it lies. It had no "choice" in the matter. Of course, we can still claim to have free will if we define the term as being free of external constraints,but that's not really addressing free will, and why free will exists as an issue. The free will issue exists because people claim "I could have done differently if I had wished." Problem is, of course, they didn't wish differently because . . . .

This, then, is my argument---a bit shortened to keep it brief---against free will as it stands in opposition to determinism.

Thoughts?

The blame or the praise are components of the outcome
Arriving at home is not the end of the sequence of events

 

[Skwim]

The blame or the praise are components of the outcome.

Effects from it, and still pretty hollow.

Arriving at home is not the end of the sequence of events

As far as the illustration goes it is.

 

[Ayjaydee]

Effects from it, and still pretty hollow.


As far as the illustration goes it is.

You structured the illustration to fit your premise

 

[Skwim]

You structured the illustration to fit your premise

I structured my illustration to show the inevitability of an effect from specific causes.

.

 

[Ayjaydee]

I structured my illustration to show the inevitability of an effect from specific causes.

.

Of course all results have causes. But all results are free of other possible causes. Your feelings toward praise or disdain will activate some and not others and will be variable from moment to moment

 

[Skwim]

Of course all results have causes. But all results are free of other possible causes.

No. A result (effect) can arise from more than one set of causes. I can arrive at home (the result) by taking one of several different paths. (sets of causes)

Your feelings toward praise or disdain will activate some and not others

What "some others" are you talking about?

and will be variable from moment to moment

Not at all clear of what you're talking about. My feelings or these "some"s?

.

 

[Ayjaydee]

No. A result (effect) can arise from more than one set of causes. I can arrive at home (the result) by taking one of several different paths. (sets of causes)


What "some others" are you talking about?


Not at all clear of what you're talking about. My feelings or these "some"s?

.

The causes will vary from moment to moment as will the decided reactions to them

 

[Skwim]

The causes will vary from moment to
moment as will the decided reactions to them

Just so you're aware, when discussing determinism there is no need to address reactions, plural. Focusing on a single reaction (effect) is more than adequate and makes everything a whole lot neater and easier. The single event then serves to describe the nature of all events, which, as it turns out, shows that every event in the universe is caused, including those of the will. In effect, the will is forced to conform to the causes leading up to its moment of response. IOW, it is not free to do any differently. You do X rather than Y because you can do no differently.

.

 

[LegionOnomaMoi]

This has nothing to do with the point I made. Evolution is a natural process with no intent but it does gather information about the environment and encode it in genomes. My point was that there was no magical "free will" needed to transfer information about the world and encode it somewhere else.

Fine. But you made your point in relation to or opposition to mine. And my point concerned the assumption of freedom of choice (the ability to do something such that one could have done otherwise) as a foundational assumption of empirical science. My point related to the underlying logic we use and have used for centuries in the sciences to build our understanding about the world through experiments. So whether or not information is translated and encoded via evolutionary processes is moot. What matters is the extent to which any sort of physical determinism that would seek to be based or make contact with actual physical theory or our understanding of the physical world can be made coherent. The point of my quotes and of the talk I linked to earlier (as this issue has become a mathematicized one in various attempts in foundation physics to exploit loopholes in Bell's inequality) is that we take our ability to freely determine certain experimental conditions as given in empirical science (such as e.g., that individuals were assigned "randomly" to particular clinical groups or that a particular sample of some material was actually a sample from the specified population or even taking the configuration of celestial bodies at a particular time to be time t_nought) and use this assumption to make inferences given the results.

hypotheses, designing and performing experiments that you think is incompatible with determinism?

Many things. It is the basis even for many of the terms used in the foundational concepts of probability theory and statistics as well as classical physics, hence Bohr's emphatic statement that in QM "The freedom of experimentation, presupposed in classical physics, is of course retained and corresponds to the free choice of experimental arrangements for which the mathematical structure of the quantum mechanical formalism offers the appropriate latitude." (Atomic Physics and Human Knowledge, p. 73). Kolmogorov encoded this kind of freedom in to the basic formulation of modern probability theory (others were doing this already, but as Kolmogorov's measure theoretic formulation remains the basis for all probability and statistics, even non-classical extensions, I defer principally to him here). It was already built into classical physical descriptions in a variety of ways, from the trivial (e.g., degrees of freedom) to the most fundamental, basic understanding of the nature of statistical mechanics and the 2nd law debated by Maxwell, Boltzmann, Gibbs, and other giants.
On one level the basic issue is one of generalizing from a sample. We never take into account the entire universe when we do experiements. Instead, we try to construct a mini-universe either by treating some system as some sort of idealized one in isolation or by isolating it sufficiently well for our purposes. Imagine something like tracks in bubble chambers from cosmic rays or your classic Stern-Gerlach apparatus using electrons or something. First we have something of the material or substance we wish to experiment with. Then we have some experimental setup. But how do we even get to the first stage? How do I take a cup of water or shine a light or do anything at all in a completely deterministic world and say that when I do such and such experiments with the water, this was due to the properties of the water not because of a variety of "flukes" that were predetermined and necessarily followed from initial conditions of the universe? I can't even claim populations exist from which I could sample because my basis for classification of populations from which I could sample is the assumption that the samples I have come into contact with were not determined in advance. The entire point behind eradicating biases and careful sampling and such is because we believe it is within experimental control to carefuly choose initial conditions, experimental settings, experimental proceedings, etc., (including random assignments to various control/test groups as well as conditions) such that we can say the results can be generalized from beyond the exact details of whatever it is we said and did.
Hence the link to the talk. In a variety of sciences, this issue crops up as something that is problematic mainly from a methods perspective and/or a modeling one. That is, in epidemiology or social neuroscience or any number of fields, there are statistical methods in common practice like latent path analysis or SEM and so on that have a kind of causality built into them such that when you feed data into your models/equations you will get values out that correspond directly or indirectly the the extent to which one or more factors can be said to be the cause of some others. The mathematical presuppositions and experimental issues are vast, and do touch on this issue of experimental choice because it is of key importance at the least that the way in which we plan and conduct experiments accounts for what we freely choose to vary and what we hold constant (that is, we control for biases of various kinds and errors more generally).
What was not, apparently, sufficiently recognized in the development of statistical methods in the sciences as well as experimental methods and the sciences themselves was the absolutely foundational role that experimental choice played in the most basic assumptions and physical laws. Even after Einstein et al. explicitly showed how our then-current understanding of QM entailed a fundamental problem in our knowledge of physical systems in EPR 1935, it wasn't until Bell 1962 use of Bohm's (1951) reformulation of EPR that it began to become clear how fundamental "free choice" is to determinism itself.
To the physicist, determinism tends to mean "given initial condition in the manner of preparation/specification X, the final state will be Y" which is to say that it means we can predict at least in principle with perfect accuracy how a prepared/specified system in (idealized) isolation will behave according to deterministic laws encoded mathematically. That's why we can talk about initial conditions at all. Because we are free to construct our mini-universes in which we idealize a system and its environment as being all that there is, and then declare that it starts and ends as and when we please. Without the ability to exercise free choice in setting initial conditions, no physical laws can ever be discovered nor generalizations made from experiments in the natural sciences.
Quantum mechanics changed the way in which we can think we do this. It was proved, repeatedly, that under the assumption locality one could (and this has been done) arrange for a system to communicate instantenously or interact instantaneously with an experimental apparatus kilometers or even lightyears away or even backwards in time. These kind of problems forced a lot of people to think very, very hard about what kinds of basic, trivial assumptions were being made in such experiments so as to yield potential loopholes. One clearly obvious one that can get us out of any troubles is superdeterminism. In otherwords, we can always recover any kind of causal, local mechanistic theory we wish so long as we allow for hidden variables in the form of superdeterminism (some kind of determinism in which the whole of any experimental design is in fact already determined by the conditions of the universe arbitrarily far into the past). The reason that those theoreticians, experimentalists, and so forth have not universally or even in large minority latched onto this view is because it negates the whole of science. It amounts to saying what the OP does: that the outcome of every experiment was determined by conditions of the universe eons ago and any experimental choices made so carefully in attempt to control for errors and biases in sampling and in experimentation are moot. What will be would have been and could not have been otherwise and no causal powers were exercised by the seeming making of any "free" choices on the part of experimenters.

But this is nonsense. Or at least if it isn't nonsense, it is antitathetical to the practice of science and to basic scientific assumptions. Hence the rejection of superdeterminism as a valid explanation for violations of Bell's inequality. It could never, actually, be logically acceptable to deny that the results of any experiment were either biased or unbiased because all experiments were equally biased by being wholly determined in advanced if physical determinism were true. All experimental choices would be illusory, and thus all inferences made using these choices (a must for all science) would be moot.

The point is not that quantum mechanics is where this comes into play. Far from it. It is at play in all of science. Every time we use statistical inference, sampling theory, experimental control groups, replication, etc., we are saying that it is possible for experimenters to determine how certain things happen in experiments while allowing others to vary in such a way that it can be said of the results that they were not just particular to what happened at one place at one time but generalizable a population of interest whence the sample was randomly sampled from. The reason QM came into play was because it made us (or should make us and has in vast areas of the sciences) realize again certain presuppositions and assumptions we make in empirical inquiry and the logical bases we have for key inferences (above and beyond e.g., p-hacking or the crisis of confidence in p-values and the replication crises and so on and so forth, let alone the problems posed by "big science" for reproducibility, but that's digressing).

How about a short summary of how you think free will can work if choices are not entirely the result of their antecedents but involve no randomness?

I already spelled out in a few lines how in neuroscience it is critical for randomenss to play a role in neural mechanisms and processes. I don't play the "either it is randomed or it is determined" game as it is a basic fallacy and quite wrong.
But for one thing, you have to keep in mind time scales and structures. I don't know how to sum up something this complex in a few lines. I do know how to show that it doesn't suffice to think purely in these terms for causality in general, but that's another manner and again randomeness doesn't necessarily play the role you seem to wish.

 

[LegionOnomaMoi]

You know, we're so far apart here I don't believe our exchanges are doing either of us any good.

Ok, let's try going back to something very simple and something you already stated was fundamental: the randomness vs. determined dichotomy. I've brought visual aids. This will be quite simple and will not involve free will directly at all yet, merely examine how you would categorize a phenomenon you can see below in terms of randomness vs. determined. Here's a clip:

Spoiler: Back story

Some back story you can skip: Mark Silverman was the first Western invited professor to what was then the new Hitachi research lab ARL, and wanted to do something like what is shown in the clip above for his students. He was denied, but later the idea was revisited. When Hitachi realized the result (a silent 5-min clip Single-Electron Build-up of an Interference Pattern) was a good idea, they repacked it and added effects. This was decades ago, so the added narration and editing seems quite dated today, but that actually works here.

Now, what you are seeing is a version of the famous double-slit experiment or two-slit experiment done with electrons:

Single electrons are emitted via an electron microscope such that the emissions are very slowly produced in order to ensure that no more than one electron leaves the source at a time. They are detected as the dots you can see in the clip one at a time.
Here's the important bit. Where each electron lands is completely random. In fact, the standard view (and this view is also common to various other interpretations of QM) holds that not only do the elctrons not follow a determined trajectory but but also that their very positions are indeterminate prior to detection (that is, they do not exist as particles until they are detected at the random spots you see in the clip.
Moreover, there is no path, let alone a determined trajectory, the individual electrons could follow that would yield the results you see. But not only can we predict perfectly what the resulting pattern will be, we can so perfectly determined the results in advance by the experimental design (which allows for single electron interference using a biprism like a double-slit) and the resulting pattern is so clearly non-random.
So, we have a very simple system, quintessential complete randomness, and the result is a very organized pattern determined in advance by a probability "function".
So, what causes the pattern? Well, whatever the causes they must include the electrons being detected as these form the pattern that was so clearly determined we can predict it in advance. But each electron not only shows up randomly in a random way but is so indeterminate that it cannot be said to have a determinate position or followed a determined path before detection. So the fundamental cause of the determined pattern is absolute randomness.
Yet you assert this is impossible: no true randomness, still less no determined randomness.

 

[Mock Turtle]

Is it possible that there is a greyscale of our actions between free will and determinism?

 

[ratiocinator]

@LegionOnomaMoi - You seem to think that lots and lots of words and quotes is a substitute for actually making a coherent logical case. I ploughed through some of it, searching in vain for an actual point or argument - I've better things to do with my time than sift though empty verbiage looking for a point that, judging from previous posts, probably isn't there.

So, we have a very simple system, quintessential complete randomness, and the result is a very organized pattern determined in advance by a probability "function".
So, what causes the pattern?

It a classic example of a combination of randomness and determinism. You can exactly simulate such situations with a source of randomness and a computer (an algorithm). Again - you've typed a lot of words but where is your actual point?

 

[Skwim]

So, what causes the pattern? Well, whatever the causes they must include the electrons being detected as these form the pattern that was so clearly determined we can predict it in advance. But each electron not only shows up randomly in a random way but is so indeterminate that it cannot be said to have a determinate position or followed a determined path before detection. So the fundamental cause of the determined pattern is absolute randomness.
Yet you assert this is impossible: no true randomness, still less no determined randomness.

Why must they be predictable in order to be determined?

Aside from the fact that the use of "random" and "randomly" here begs the question, why does it follow that from our inability to determine the appearance of an electron that such an appearance lacks a cause?

A rather silly statement wouldn't you say?


.

 

[ratiocinator]

Aside from the fact that the use of "random" and "randomly" here begs the question, why does it follow that from our inability to determine the appearance of an electron that such an appearance lacks a cause?

Why does it follow that there must be a cause? Are you really trying to say that you have some evidential or logical reason to exclude actual randomness? If so, go publish and claim your Nobel.

 

[Skwim]

Why does it follow that there must be a cause?

How else do things come into existence?

Are you really trying to say that you have some evidential or logical reason to exclude actual randomness? If so, go publish and claim your Nobel.

Gotta be the first, and I'm hardly that, cool as I may be.

.

 

[LegionOnomaMoi]

Why must they be predictable in order to be determined?

They need not. Here, the problem is more than they are just not determined. They are indeterminate. Thus, for example, even if instead of trying to predict where electrons would be detected, you waited until after the electrons had interacted with the prism (had gone through one of the "slits") and tried to see where it will be detected, it wouldn't be there. Indeed, even if you wait until after an experiment like this is finished, you can choose whether or not what you will see will be the pattern you see in the clip, in which case you cannot learn anything about where the electrons were before they were detected on the screen or how they got there, or instead you can learn which of the two "slit" the electron traveled through, but then the pattern will be gone (you will get two bright spots or lines instead of the interference pattern you see in the clip).
In short, the reason you can't predict how an electron will appear on the screen is not just because it isn't deterministic, but because the electron has no determinate position or trajectory until it is forced to be localized on the detection screen. It is indeterminate. Other experiments explore this more directly in versions of wheeler's delayed-choice experiment in which you actually do let the whole process finish but wait to "look". Whether or not you get the pattern in the clip depends crucially on whether you try to find out how the electrons arrived on the screen (e.g., where were they the instant before they hit the screen), because if you do this you will destroy that pattern and they will have never arrived on the screen that way even though the experiment is done. Even if you put a series of registers to see if an electron went through slit A or B and wait until the experiment is done to look, you will find that the pattern in the video is gone. If you look to see the pattern first, then there will be no information from any register or detector as to which path any electron followed to arrive on screen or where any electron was the moment before landing. These two properties of the electron (a determined position and a determined trajectory) do not both exist at the same time.
What does exist, and does govern the behavior of each electron and the overall pattern, is a probability "function" that depends crucially on the individual indeterminacy of the electrons to accurately describe what you see in the clip.

 

[Skwim]

They need not. Here, the problem is more than they are just not determined. They are indeterminate. Thus, for example, even if instead of trying to predict where electrons would be detected, you waited until after the electrons had interacted with the prism (had gone through one of the "slits") and tried to see where it will be detected, it wouldn't be there. Indeed, even if you wait until after an experiment like this is finished, you can choose whether or not what you will see will be the pattern you see in the clip, in which case you cannot learn anything about where the electrons were before they were detected on the screen or how they got there, or instead you can learn which of the two "slit" the electron traveled through, but then the pattern will be gone (you will get two bright spots or lines instead of the interference pattern you see in the clip).
In short, the reason you can't predict how an electron will appear on the screen is not just because it isn't deterministic, but because the electron has no determinate position or trajectory until it is forced to be localized on the detection screen. It is indeterminate. Other experiments explore this more directly in versions of wheeler's delayed-choice experiment in which you actually do let the whole process finish but wait to "look". Whether or not you get the pattern in the clip depends crucially on whether you try to find out how the electrons arrived on the screen (e.g., where were they the instant before they hit the screen), because if you do this you will destroy that pattern and they will have never arrived on the screen that way even though the experiment is done. Even if you put a series of registers to see if an electron went through slit A or B and wait until the experiment is done to look, you will find that the pattern in the video is gone. If you look to see the pattern first, then there will be no information from any register or detector as to which path any electron followed to arrive on screen or where any electron was the moment before landing. These two properties of the electron (a determined position and a determined trajectory) do not both exist at the same time.
What does exist, and does govern the behavior of each electron and the overall pattern, is a probability "function" that depends crucially on the individual indeterminacy of the electrons to accurately describe what you see in the clip.

Understood; however, indeterminate ≠ utterly random. Utter randomness being the issue at hand.

And as I said back at the beginning of the year,

"You know, we're so far apart here I don't believe our exchanges are doing either of us any good, so I'm going to let you have the last word if you wish."
​

a sentiment that still holds,

Have a good day.

.