Are we living in deterministic universe?

[vlaiv]

Most of us will simply dismiss this question with a simple answer: "No, it's probabilistic in nature", but please bear with me for the following argument.

Lately I've become really stressed out by my inability to comprehend probability, no, not the mathematical side of it, nor calculations involved, but rather nature of it, or in terms of proper definition what probability is in physical world rather than as mathematical (axiomatic) concept. I think proper expression would be that I don't understand interpretation of probability. But all of that is besides the point for the argument, I just wanted to make an introduction into how I came about what I'm going to layout here. So, in doing research on the topic of probability, I've revisited the idea of delayed choice quantum eraser experiment. I was reading on the topic, and especially interpretations of results, and then it struck me. Long time a go, as a young "philosopher", I invented the concept of "reverse causality".

It goes something like this: If I push the glass of the table and it falls to the ground and breaks into pieces - we would normally say that the glass was broken because I pushed it of the table. What prevents us from saying the following: Glass needed to be broken and that is the reason I pushed it of the table. Therefore I concluded that causality should be able to work both ways. It was kind of naive and cute way of experimenting with different points of view.

Now as I was reading about interpretation of delayed choice quantum eraser, widely excepted interpretation is that it is the position of signal photon that hits detector screen (it's place within either interference pattern or regular diffraction pattern), that what determines where idler will be detected later on. So I remembered my idea about "reverse causality" - or the idea that paired events could be interpreted as causing one another without really assigning time direction to it. After some further thought on that topic I realized that actual order of events is not fixed - at least that is what relativity it saying - it depends on observer's frame of reference. So the actual time has nothing to do with pair of events that are causal - it's not that one is leading to the other, it is just relative which way we are interpreting it or rather assigning a meaning - but they are just simply linked events - happening in pair.

Now this is where it gets interesting, we can extend the argument that there are cases where there are two causes that both must be satisfied in order to have an effect. So it's not simple case of events in pairs, but rather we are starting to branch out, and have multiple events that are linked together. By simple means of induction, if we accept that there is no something we could call "singular" event - one that is not in causal relation to any other event - we come to conclusion that all events must be linked in a web. And that web of events is "direction invariant".

This smells a lot of deterministic nature of universe.

What do you think?

[cgarry]

13 minutes ago, vlaiv said:

After some further thought on that topic I realized that actual order of events is not fixed - at least that is what relativity it saying - it depends on observer's frame of reference. So the actual time has nothing to do with pair of events that are causal - it's not that one is leading to the other, it is just relative which way we are interpreting it or rather assigning a meaning - but they are just simply linked events - happening in pair.

I thought that relativity says that the order of cause and effect is always preserved, no matter what your frame of reference is.   The exception is if you are travelling faster than the speed of light when the order of cause and effect is not preserved, but then you are into time travel anyway...  This is most definitely not my specialist subject though, so I too am interested in what other have to say.

[vlaiv]

2 minutes ago, cgarry said:

I thought that relativity says that the order of cause and effect is always preserved, no matter what your frame of reference is.   The exception is if you are travelling faster than the speed of light when the order of cause and effect is not preserved, but then you are into time travel anyway...  This is most definitely not my specialist subject though, so I too am interested in what other have to say.

Being far from expert myself, but not sure about that. I was under the impression that timing of events has something to do with finite speed of propagation of information.

Imagine following (akin to delayed choice quantum eraser). There is pair of entangled particles, split by enough of a distance that information exchange is impossible for two separate measurements. Now one particle is being measured by person A and one by person B. Whoever "does the measurement first" will "determine" outcome of second measurement (if we think in terms of causality). They can't communicate information about measurement to each other until they've both completed measurements. Now, there exist two frames of reference such that: One in which person A does the measurement first, and one in which person B does the measurement first. So we can equally say that measurement A caused measurement B to be known, or we could say that measurement B caused measurement A to be determined. I'm not sure that relativity forbids above scenario, but it clearly shows that since prior to any measurement there is no definite answer to measurement results (no hidden variables), so it is kind of joined pair of events, but some might see it in order A, B and some might see it in order B, A (at least I think so).

[vlaiv]

I made a little diagram to emphasize point from previous post:

 

Green circle is measurement A, blue circle is measurement B. yellow square is observer moving left (little yellow arrow), red square is observer moving right (little red arrow). Now long yellow and red lines represent "Now slices" of yellow, and red observer respectively. Time is up. Since we know that first measurement instantaneously determines outcome of second measurement, for yellow observer measurement B already happened, and causality is BA, for red observer measurement A happened so causality is AB. So we see that AB==BA, or "that scenario had to happen regardless which measurement we observe as first". Note that there is no faster than light information exchange.

[Art Gecko]

On 18/05/2017 at 19:58, vlaiv said:

I made a little diagram to emphasize point from previous post:

 

Green circle is measurement A, blue circle is measurement B. yellow square is observer moving left (little yellow arrow), red square is observer moving right (little red arrow). Now long yellow and red lines represent "Now slices" of yellow, and red observer respectively. Time is up. Since we know that first measurement instantaneously determines outcome of second measurement, for yellow observer measurement B already happened, and causality is BA, for red observer measurement A happened so causality is AB. So we see that AB==BA, or "that scenario had to happen regardless which measurement we observe as first". Note that there is no faster than light information exchange.

Hi vlaiv, fascinating thread! I'm not sure I see how your example shows any evidence of the effect coming before the cause in your reverse causality theory though? I agree that it doesn't matter which point you measure first, A or B, but if the end result is to have a triangulation point between the two, you still need both measurements to get that triangulation point.... cause still comes before effect, no?

[vlaiv]

27 minutes ago, Art Gecko said:

Hi vlaiv, fascinating thread! I'm not sure I see how your example shows any evidence of the effect coming before the cause in your reverse causality theory though? I agree that it doesn't matter which point you measure first, A or B, but if the end result is to have a triangulation point between the two, you still need both measurements to get that triangulation point.... cause still comes before effect, no?

It just shows that our concept of cause and effect is tied to time frame, and might be limitation of the way that our brain works. We are used to thinking that cause is followed by effect, and that cause is "the reason" for effect. In both examples, event A can be thought as a cause for event B, but it may as well be thought as the other way around, that event B is cause for event A.

In example with relativity and entangled pair, the way you perceive it depends on your frame of reference. There exists frame of reference where it is ok to say that A -> B, but also there exists frame of reference where it is B -> A. Now this could be understood as some sort of artifact of theory of relativity that is of no major consequence because there is no information exchange, but other example is not tied to theory of relativity and it exhibits same behavior. In delayed choice quantum eraser experiment vs regular dual slit experiment we have following scenario: In regular dual slit experiment, absence of information "which slit" is "causing" interference pattern, while in delayed choice variant, existence of interference pattern is "causing" absence of "which slit" information. "Causing" here is meant in terms what comes first in time reference frame (which in this experiment is just regular frame of reference, no relativistic effects).

Now, if we have A implies B and B implies A, in mathematics this is A equivalent to B , or in other words - these two events must occur together but cause and effect have no real meaning, and we can only say that one is the cause for the other if we tie them to certain time frame.

Ok, so when we get to this point of both A and B are, don't even know how to name that, but let's call it: coupled event - this is just expression that means they happen as opposed of other combination of events happening, without causal implications, we can take further step and say, if we have A1 and A2 - mutually excluding events, and we have B1 and B2 also mutually excluding events, and A1B1 is coupled event, and A2B2 is also coupled event, we might reason the following: That is all fine and dandy, but that does not imply determinism, we can still have randomness of choice - either A1B1 or A2B2 regardless if they are coupled events or not. Now the idea of nonexistence of "Singular" event comes into play. If we reason that nothing happens without a "cause" as we perceive it currently - events always must come in "pairs" (or multiplets?), now we can say that A1B1 is event in itself and by means of mathematical induction, there is Cx that is coupled to it. And we can go on and on, both in past and in future (it does not matter since these are all coupled events) and we conclude that there is huge "web" of coupled events stretching all the way back in past and in future. Each A1B1 and A2B2 have their own "web" of events that differs by virtue that A1 - A2 are mutually exclusive (and same goes for B1-B2), and we just saw A1B1 happen but not A2B2 - this means that A1B1 web is "valid" web of events, and these are events that happened or will happen, and if we have such a single path - determinism.

[vlaiv]

Just had a further thought on all of this. And it is also interesting (but also mind bending ).

Let's consider single radioactive atom. It decays at some point. Current view is that it is pure random event, without a cause - no hidden variables (maybe this is Singular event that just throws a wrench into wheels of above argument? ).

Now we have two different lines of reasoning. First is, yes, radioactive decay happens without a "reason", just pure random event. Or, we can reason that it is not singular event because it must be "future" paired. It will have consequence, it is what we call "cause". But we just saw from above argument that it does not matter what is the cause and what is the effect and this is where reverse causality comes into play, if we say "Radioactive atom had to decay in order for detector to detect it" rather than "Detector detected decay because atom decayed randomly" - we conclude that there was nothing random about decay of that atom

No randomness - only determinism.

I'm not claiming that our universe is deterministic, I'm just trying to follow a line of reasoning and look where it might take me, and so far, the way I see it, it is pointing to determinism. There might be a flaw in this line of reasoning, and this is precisely why I put forward this idea, so we can examine it and see if it holds.

 

[Art Gecko]

I think I see where you're coming from: No single event can be completely isolated from effecting another event, even if it first appears to be unrelated... and on a sliding scale, one event will have a knock on effect with another and so on... The effect becomes the cause of the next interaction.. Looking at a bigger picture.. even parallel chains of events will always intersect eventually and form a web of events that make the entirety of.. well, everything!

Am I with you so far?

I think what you call Determinism, I call 'structure' and I'm glad the universe has it! If it didn't, I don't think we'd be here to have this conversation

[vlaiv]

Well, yes, I was pointing to a structure but also to a true determinism in sense that there is no true randomness.

This clearly shows in case in my last post - decay of atom. If we try to interpret it using "conventional" wisdom, there is no trigger that determines exact time when decay will occur. Best that we can do is give it a probability in time period (half time). Our current understanding implies that there are no hidden variables that influence time of decay, no internal process in atom, nor external for that matter, that triggers decay, it just happens spontaneously, and there is no way of predicting it even in principle - hence random process.

What I'm trying to point out here is that if we follow certain logic that appears to be at play in two cases I've mentioned, decay of atom is not random at all. It had to happen at exact point in time and in exact way in order to be paired to what we perceive as future events (decay detector detecting decay).

So it is true determinism that I'm pointing to, and I'm well aware that it is extremely controversial way of thinking, as it leads to not having free will and all sorts of other implications.

Back on exact reasoning, try to think of it this way: We think that past is set, there is only one single history. For future we think that it is not set, there are options, it can go either way. What I'm trying to imply here is that if we accept that event A which we see as in the past is coupled with event B that is in the future in such way that saying A causes B is only matter of perspective, and it is OK to say that B causes A, then if A is set (already happened), and there is no possibility that past is different than from what already happened, and future event B is cause of A which is in the past, then there are no other options for future event than B, because if in future some Bx happens that Bx would cause Ax to have already happened - but we know it did not.

 

[andrew s]

On 5/18/2017 at 18:58, vlaiv said:

Oops can't delete the quote box

[andrew s]

Sorry about that empty post but I was going to comment sooner but decided I needed to think about it more.

I don't think you are not using the standard definition of determinism "Determinism is the philosophical idea that every event or state of affairs, including every human decision and action, is the inevitable and necessary consequence of antecedent states of affairs."

Clearly if there is a causal chain then running it backwards will always lead to a definite event that is what causal chains are. However, that does not men that the initiating event was not random. Trying to argue backwards as you do is not legitimate in my view. You would need to show that the potentially "random" event had a clearly defined antecedent.

I think all you are demonstrating is that the probability of an event becomes 1 once it happens even if the odds were vanishingly small before the event.

However, I don't think this settles the argument either way though. I don't know the answer to the question. 

Regards Andrew

[vlaiv]

51 minutes ago, andrew s said:

Clearly if there is a causal chain then running it backwards will always lead to a definite event that is what causal chains are. However, that does not men that the initiating event was not random. Trying to argue backwards as you do is not legitimate in my view. You would need to show that the potentially "random" event had a clearly defined antecedent.

I'm just trying to figure out interpretation of two experiments listed above. One that we might call thought experiment (involves entangled pair and relativity), and the other one - real, conducted experiment.

Now, let's just look at the thought one for now, as it is less complicated to think about (other involves "which path" information, and that gets kind of fuzzy). We do know that entangled pair does not have hidden variables (Bell's theorem). Prior to measurement it is in superposition of states and "collapses" to a measured value. Now for some reason we like to think of it as first measurement collapses the state, so second measurement just finds system in collapsed state. But which is the first measurement? That clearly depends on observer. Now in it self this does not rule out randomness, regardless of the fact that it is not clear when "the actual decision" of outcome is made, it is being made consistently with entanglement, but that does not tell us that this is only way it could have happened, it just shows that it happens consistently, and causes us problem to pinpoint time at which "decision" is made (note that I'm quoting the word decision, I'm not implying that there is actual decision making happening here, just a figure of speech that emphasize idea that there is an option / multiple possibilities and one will happen).

Now the other experiment is the one that caught my attention. This is probably because I'm not used to think about "which path information" being entangled to "interference pattern". I can't grasp single particle being entangled to itself, or two events being entangled. But possibly that is exactly what is happening here. Maybe if we observe particle going thru double slit and landing on a screen as some sort of system in entanglement, sorry, as I'm writing this I'm trying to actually wrap my head around to what would be entangled with what here and how does it play out, so its a bit hard to write clearly.

Maybe best approach here is just to state what is happening and try to interpret it. So it is the delayed choice quantum eraser experiment. Short description of it would be: It is akin to dual slit experiment, particle is shot at dual slit, it passes thru one, the other or superposition of these two paths, if information of which path is taken exists and is available particle interferes with itself and forms interference pattern on screen. Twist with this experiment is that due to experimental setup, we determine if information about which slit exists, well after particle has already hit the detector screen. So it is a bit reverse than normal dual slit in terms that marking of photon with which slit information (cause) causes absence of interference pattern (effect).

Now, one interpretation proposes that events in future influence events in the past. I'm not overly happy with such interpretation so let's leave it aside for now. Other interpretation could be that photon hitting the screen collapses system state in such way that there is no more "which path" information. I'm kind of ok with that interpretation, but it does raise a question. If we look at actual experimental setup we will notice that only by matching times at detectors, one being screen that gives interference pattern, and the other being detector for which path information we can see that there is correlation between existence of which path information and interference pattern. If we look at individual photon case, position on detector screen just simply does not hold any information whether that particular photon belongs to interference pattern or not. Did it interfere with itself or not. Only by looking at which part detector - later on (later in time), we can determine if that photon actually interfered with itself or not.

We can argue that some photons interfere with themselves, and some photons don't, but that just goes against regular dual slit which clearly shows that if we don't mark photons - they always interfere.

Only thing that is left here is to say that we have ambiguity of the moment actual system state collapse occurred. As I'm aware of it, current understanding is that system collapses when there is energy / momentum transfer that leads to decoherence, and only places I can see it happening with this experimental setup are detectors - either detector screen or "which path" detector. Now if we say that detector screen caused system state collapse, hm, as I see it we are entering hidden variable territory there - no information if photon interfered is present in position on detector screen, same position can be hit in both interference and non interference patterns, and if we say that which path detector caused collapse, we can't avoid reasoning that future events determine outcome of past events.

Only way out of this that makes logic to me is if system behaves deterministically and we don't have cause / effect relationship in only one direction but we have coupled events, and we can observe cause/effect in both directions (it is actually future events influencing past events, but without firm arrow of time, with coupled events time direction is irrelevant - and this is property of deterministic systems).

Now question is what is wrong with this line of reasoning? I'm quite open for a flaw in this reasoning to be pointed out to me. I don't see it as "trying to argue backwards".

 

[vlaiv]

After another round of thinking I think I figured out plausible explanation for both experiments that don't include deterministic universe.

First relativistic / entanglement one - problem was that we could not determine exact time of function collapse. Well we might be able to dodge the bullet here and say that collapse happens at the time which comes first in the frame of reference in which entangled pair was created. I don't really have any arguments for this, but it looks like sensible answer and elegant solution. This way we can say that any frame of reference will pick transformed time to be the time of wave function collapse, and indeed in some frames of reference it will look like events in future determining outcomes in the past, but since there is no information exchange, I'm sort of happy with that, it's just an oddity of relativity same as time dilatation and length contraction.

Now to second experiment. Here I came to conclusion that term "which slit information exists" is misleading. We tend to think that something does exist, or that something does not exist. We don't really like the idea that something partially exists or exists with a certain probability, thought, later is somewhat acceptable, but then again very wrong usage of term probability - like probability of extra terrestrial life. We say probability, but there is only certainty about it, either aliens exist or they don't, nothing probabilistic about it.

So what really happens in experiment?

Photon is fired on a slit, and after passing through the slit it is marked with which slit information. When signal photon is detected at the screen, wave function collapses, but it does not collapse into state that we might call "which path information is lost" or "which path information is still there", but rather, depending on position of signal photon on detector screen it collapses into a state which we could call "N probability that idler photon will interact in such a way to let us obtain information of which path and 1-N probability that idler photon will interact in such way that we will not be able to determine which slit information".

So after wave function collapse, and depending on signal photon position, we now have new distribution of probability for idler photon behaving as if coming from either slits, or coming from both slits at the same time.

Now to points I want to make here. First one, having conditional probability described above, it is easy to see that if we establish correlation between idlers and signal photons we will either see interference pattern or not, and conclude that interference pattern is coupled with which slit information.

Second, I believe more important point, one I seem to forget from time to time, is that photon never goes thru the single slit. It always goes thru both slits, sort of.

If you look at simple dual slit experiment, it is wrong to say photon went to one slit an was detected. Proper formulation is: in case there are no detectors, photon is in state "passed thru first slit" + "passed thru second slit" - this state can interfere with itself to produce interference pattern. If there is detector formulation of state would be "photon passed thru first slit and was marked as passing thru first slit" + "photon passed thru second slit and was marked as passing thru second slit" - this state can't interfere with itself hence no interference pattern.

Now for delayed choice quantum eraser, state would be: "photon passed thru first slit and was marked as passing thru first slit and created signal and idler pair" + "photon passed thru second slit and was marked as passing thru second slit and created signal and idler pair", but upon detection of signal photon on detector screen, idler would not collapse to pure state but rather: "signal photon landed on detector screen with probability N of belonging to interference pattern" + "signal photon landed on detector screen with probability 1-N of belonging to non interference pattern" sort of state. So although photon is marked after passing, it can still end up as not being marked with certain probability depending on where signal photon landed on detector screen.

[andrew s]

I need some time to think this through but will respond Andrew

[andrew s]

I have been doing some research on entanglement and simultaneity.  This is the example given in your early diagrams.

What I found was that as you said if measured in the frame in which the pair creation and the measurements A and B were all at rest and A and B measured simultaneously then there are other frames in which A proceeds B or B proceeds A.  

However, if you look at a space time diagram in all cases in any frame A and B are space like separated (no light beam can connect them). This means they can't be given an invariant  time order and so they can't  be directly casually connected. Note only invariant quantities are conserved across inertial frames in relativity.

Indeed as the results of the measurements at A or B are random, but correlated, I would conclude that provided the random nature of QM holds up we are not living in a deterministic universe.

Regards Andrew

[andrew s]

On the radio active decay the (random) decay C and it's detection D are time like separated. This means the direction D will always occur within the future light cone of the decay event C in all possible inertial frames. They are casually connected.

What you call reverse causality doesn't exist as far as I can tell in either QM or SR.

Again provided the decay is random as predicted by QM then we are not in a deterministic universe.

Regards Andrew

[Leexinxin]

From my personal point of view，Probability is an uncertain factor，We can not be able to discuss an uncertain thing, in the course of the discussion he has been changing.Because the views of the various changes are uncertain, but the starting and ending is determined.You use the reverse thinking to explore the starting point is the wrong idea, because the result has been doomed, is due to the process of occurrence of things to determine if the development of a thing has its own orientation, then the whole process is the decisive material.

Then about whether we live in a deterministic universe，You need to first understand that our present world is composed of objects and time, objects lead to things can produce, time lead to the process, the object we all understand that this is the world, but time? Time is uncertain, anything will lead to the deviation of time, the whole space will be deviated, if the world is a computer, then the data will occur can not control the changes, these are some minor, but occasionally these changes are Huge, some files are lost, and even crash, causing the world to deviate from the track.

I rarely go to this site if you have a deeper discussion and me

Contact my email：<email address removed>

looking forward for your reply.

 

[billyharris72]

My tuppence (and it is worth about that). Thoughts appreciated as I am trying to work this out on the fly.

I think we may have to distinguish here between what we might call ontological vs epistemological indeterminacy. Determinism in the way a physicist might talk about it (i.e. at the quantum level, particles behave as if they exist in a superposition until observation), is not the same as determinism in the way a philosopher might talk about it. In a physical account of indeterminacy, we could take this to mean that actual particles exist in a "real" superposition (the world is made of particles that are really indeterminate) or as a statement that the particles and waves are not anaolgous with the particles and waves we observe (they are not really particles or waves at all, but something that behaves like these things, if we include the rather counterintuitive idea of indeterminacy, which seems to be the only way we have of getting our heads around them).

While those would be two different propositions within the framework of realist philosophy (ontological indeterminacy vs a deteriminate but not fully knowable universe) there is no difference at all for physics, since it is concerned with observation, evidence and falsifiability. If we cannot know something then it is indeterminate - it doesn't make sense to talk about a particle that behaves as if it were indeterminate but which is actually determined by causes that we are fundamentally and logically unable to observe, because from the largely positivist standpoint of science these basically mean the same thing. This question might, however, make sense from a realist ontological perspective.

This is something that John Stewart Bell picked up on when he noted that in a completely determinate universe, hidden variables can reproduce all the predictions of quantum mechanics.

 

Quote

There is a way to escape the inference of superluminal speeds and spooky action at a distance. But it involves absolute determinism in the universe, the complete absence of free will. Suppose the world is super-deterministic, with not just inanimate nature running on behind-the-scenes clockwork, but with our behavior, including our belief that we are free to choose to do one experiment rather than another, absolutely predetermined, including the ‘decision’ by the experimenter to carry out one set of measurements rather than another, the difficulty disappears. There is no need for a faster-than-light signal to tell particle A what measurement has been carried out on particle B, because the universe, including particle A, already ‘knows’ what that measurement, and its outcome, will be.^[5]

I actually find most objections to this version of determinism quite facile - the idea that free will is incompatible with determinacy and that it cannot itself be "completely" determined (which makes no sense - and how would randomness make us any more free than deteminacy?) shows a lack of philosophical and psychological understanding. Whatever "free will" is, it is not something that transcends the same laws of causality (or indeterminacy) that hold for so-called "inanimate nature".

However, there is one very strong objection that I can think of. If the universe is completely determined in this way, we are left with a situation where we have "real" things that are unknowable (I can't grasp the whole web of cosmic causality that determines everything, including the thoughts I am having right now; if we could I would also know what I was about to think next). That would leave me in a position where:

1) I can say the world is completely determinate.

2) For any given observation, things will behave exactly as if quantum indeternimacy was real.

Which takes us back to the point about this not being any different from the unoiverse being indeterminate. The realist argument starts to sound a bit like a brain in a jar argument - there is no way to disprove this either, but we can dismiss it since (from an observation perspective) it is of no consequence.

Told you as I was makling this up as I went along, but I think this makes sense.

Billy.

[vlaiv]

Last post by Billy does raise some interesting questions.

I totally agree that it does not matter if universe is deterministic or random/probabilistic at basic level - neither provides mechanism for "free will".

So given the premise that we have no free will at all (everything there is / exists conforms to physical reality), I wonder why did we evolve sense / need for free will at all? Why do we value freedom if there is no choice to be made at all? Even when presented with situation where there is no obvious decision making trigger (emotional, logical, what ever), we tend to go with "gut feeling" - why do we need illusion of free will so much? Could it be just a side effect of consciousness? But then again, I can't really grasp why did we evolve consciousness at all if there is no free will - what evolutionary benefit would there be of being aware of oneself if unable to "make decisions" - simple instinct / reactions would be enough, those shaped by fitness function.

On the other hand, if there is indeed a free will, that would imply that there is agent that does not conform to physical world / laws of physics (not part of physical system that has states / state transitions whether those are deterministic or probabilistic) - "decision maker", and moreover, I would argue that physical world needs to be probabilistic in its nature, with agent being able to alter probability distribution - "miracle worker" role. If world is deterministic then there is no decision to be made at all - no multiple future states to choose between.

I know that we are leaving the boundaries of physics as science and entering the realm of philosophy here, but I do think there is value in interpretation and don't subscribe to "shut up and calculate" approach.

 

[vlaiv]

29 minutes ago, vlaiv said:

Last post by Billy does raise some interesting questions.

I totally agree that it does not matter if universe is deterministic or random/probabilistic at basic level - neither provides mechanism for "free will".

So given the premise that we have no free will at all (everything there is / exists conforms to physical reality), I wonder why did we evolve sense / need for free will at all? Why do we value freedom if there is no choice to be made at all? Even when presented with situation where there is no obvious decision making trigger (emotional, logical, what ever), we tend to go with "gut feeling" - why do we need illusion of free will so much? Could it be just a side effect of consciousness? But then again, I can't really grasp why did we evolve consciousness at all if there is no free will - what evolutionary benefit would there be of being aware of oneself if unable to "make decisions" - simple instinct / reactions would be enough, those shaped by fitness function.

On the other hand, if there is indeed a free will, that would imply that there is agent that does not conform to physical world / laws of physics (not part of physical system that has states / state transitions whether those are deterministic or probabilistic) - "decision maker", and moreover, I would argue that physical world needs to be probabilistic in its nature, with agent being able to alter probability distribution - "miracle worker" role. If world is deterministic then there is no decision to be made at all - no multiple future states to choose between.

I know that we are leaving the boundaries of physics as science and entering the realm of philosophy here, but I do think there is value in interpretation and don't subscribe to "shut up and calculate" approach.

 

I think I just reinvented Cartesian dualism

There is quite a bit on this topic : https://en.wikipedia.org/wiki/Free_will

[andrew s]

So what can provide a mechanism for free will?

Regards Andrew 

[vlaiv]

37 minutes ago, andrew s said:

So what can provide a mechanism for free will?

Regards Andrew 

In this view of "decision maker", future state of the system separates into two or more branches (excludes hard determinism with single future state), and the actual future state of system depends not only on the present state of system but on present state + decision. Now in "miracle worker" role (making less probable - more probable, thus miracle), actor changes probability density distribution on quantum level so that only subset of future states are allowed. So we are not choosing by selecting but rather by disallowing certain future states. Probably explained easiest by a single excited atom, in sphere, that emits a photon - there is uniform distribution of probability where that photon will hit the sphere. Now imagine that we changed probability distribution so that left hand hemisphere is forbidden - photon will impact somewhere on the right hand hemisphere (with certain surface probability distribution) - thus we introduced a decision into physical system.

Now one can imagine brain being space where probability distribution changes - sorts of probability field - that influences neurons firing / not firing due to electrons moving / potential difference. So our "mind" creates a field, and decisions impact probability distribution over that field that in turn drives neurons, fires off signals that drive our muscles and from a decision to raise the hand - we end up with our hand in the air.

Now, this is obviously wild speculation (if its event worthy of a such title, more likely utter nonsense ), but it shows an example of mechanism that might exist.

[billyharris72]

While I'd stick with the idea that it doesn't matter whether we say the universe is "really" deterministic or random / indeterminate, I would argue that the idea of free will is compatible with determinism. That was sort of my previous point - we don't need an independent decision maker that transcends the network of causality that determines everything else (and in fact that idea just does not make any sense). Adding quantum indeterminacy makes absolutely no difference to the verdict when it comes to free will (if free will can't come from determinism, neither can it come from randomness; I don't know anyone who equates acting randomly with being free).

Start by asking what is free will. If we want to define it in metaphysical terms, as something that transcends causality (or randomness) and is some kind of res cogitans, not subject to cause and effect but which does cause things in the "inanimate world" then we can't do science with that. Such a conception would not be admitted by anyone in relation to anything other than ourselves, or God. Basically we are now talking about a quasi-religious concept, and a specifically Judeo-Christian one. In the spirit of never discussing religion or politics I'll refrain from going further but will merely note that from a scientific perspective it begs the question and does not explain anything.

So what else might free will mean? What does it mean to say I do something of my free will? I can think of a few cases:

1) I do something because I want to do it. That is, I am motivated intrinsically to do something.

2) I do something I don't want to do because I am faced with a choice and choose the lesser of two evils (I get to make decisions but not to decide the conditions under which I have to make them). Again, there is something (yet to be determined) that constitutes "making a choice" and that is in same way instrinsic.

I think both of these conform to a concept of free will, and in both cases there is (to my mind at any rate) something about the motivating factor being intrinsic (i.e. coming from me). So, if we can say there is a situation in which the motivation "comes from" me but is still caused (i.e. determined by a chain of cause and effect that goes "outside" me) then we can still (in a limited and non-spooky sense) talk about free will.

And this turns out to be pretty easy to do. Nature is full of situations where simple elements combine into complex systems that have their own emergent properties. These emergent systems "fall back" upon the world and shape how it continues to develop. The systems develop in ways that are "autonomous", or work according to a logic that is "intrinsic" to the emergent system. Yet we would not say that these systems are not determined. It sort of depends on the reference frame we take. From the perspective of atoms there is only particle physics; from the perspective of a growing crystal (or a complex organic molecule such as DNA) there are other levels of organisation that emerge from the system itself, which fall back upon and "structure" cause and effect, but which are still caused. If you like, going from the atomic to the molecular level is like moving from physics to chemistry, and this moves already shows emergent autonomous systems (Ilya Prigogene is really good on this stuff).

It seems to me that we can conceive organic systems, from amoeba up to and including humans, in exactly the same way. These systems are caused, but also cause things to happen - it depends on the frame of reference we use. For a complex system such as a human being, "free will" would then refer to behaviours that are in some way mediated through (or at least registered on) our consciousness system, such that, from the perspective of the system, it makes sense to talk of that activity being emergent.

I'd accept that this is a limited conception of free will. For example, in a given situation there is not really an option to "choose differently", other than in a purely hypothetical sense. But in reality there really isn't an opportunity to choose differently - you choose and that's that; time moves on. And given that I am who I am, in the situation I am in, surely having free will (i.e. being self determining) implies that I would choose exactly the same thing again. If I didn't I either would not be me, or "I" would be random, which is not exactly free.

Billy.

[vlaiv]

Yes, defining free will seems to be a bit of a problem, and with following definition that I came up with (non testable, so perhaps not real definition, but it can be used as a basis for thought experiment), world is just not compatible with free will no matter how it behaves, as long as we accept definition of physical system as one with states, single current state and time evolution to a single future state, where time is irreversible, or simply put always "flows" in one direction.

I would define a free will as follows: given exact system state, and exact history trajectory, would in principle individual be able to take two different choices? If so - free will, if not - no free will. And if you think about it, given that physical reality that we currently perceive and understand - we cannot test this - no way that we can put system in exactly the same state with same history, but on theoretical level, if exact state and history are given - no reason to ever take a different choice. Person would still have same circumstances, same knowledge, same reasoning, same everything. Even if we were to ask: please choose random number between 1 and 100, in exactly the same state with same history, one would choose same random number (we don't really have random number generator in our head, it is consequence of thought process).

Now, there are couple of paths of reasoning that follow from the above. First we just simply accept that free will as we "feel" it, does not exist, and that would put us on trajectory of trying to answer the question: "Why did we evolve sense / need of free will, and what purpose it serves / or had served in our evolutionary past". At this point, I simply cannot see what would be the purpose for it, it might be just a evolutionary glitch, but I feel that is tightly bound to consciousness.

Other path that we might take is to question basic premise for above definition if we accept notion there is free will, and try to construct universe where free will exists - that would mean to reexamine number of states both past and future, and also our concept of time flow. Depending on our construct this might even be testable. For example, good basis for this would be many worlds interpretation of QM, while not providing mechanism for actual free will, it would provide mechanism for sensing / feeling a free will. Upon world split that is related closely to "decision" - each branch would carry a copy of "decision maker" that believes they made "decision" that conforms to that particular world branch.

Third option is of course to modify definition, and for example say that there is domain based free will, but not global free will - meaning that there exist states / histories separated into domains that don't interact other than by means of decision maker. This line of reasoning does indeed go deep into metaphysical and I personally don't see any objections in taking discussion in such a direction if care is taken not to wander off into religion / theology.