Causal loops in quantum theory is a subject pretty close to my heart. That said, I would approach it bottom-up, rather from this cosmological standpoint. The bit I take issue with is:

That reasoning boils down to two propositions: the first is that all events must have causes; the second is that only in closed causal chains do all events have causes.

Out universe as we best understand it can support an infinite regression of causes. If you imagine running the universe in reverse -- a unireverse, you might say -- it looks (and is mathematically described as) something like a humongous black hole.

Just as anything falling into a black hole falls forever from its point of view (ignoring death and spaghettification), anything falling back in time to the big bang could also fall forever.

And since time began at the big bang, there is no need to terminate this outside of that freefalling frame. Causality only applies inside a timeline.

The transactional interpretation of quantum mechanics though does have causal loops, an idea entertained in quantum theory since Feynman's QED. In the latter, the creation and subsequent annihilation of short-lived pairs of particles and antiparticles looks like one particle stuck in a temporal loop forever, emitting energy to change its temporal direction.

In transactional QM, the complex conjugate of a wavefunction is (mathematically validly) interpreted as the wavefunction moving backwards in time. For instance, an electron fired from a cathode arrives at a point on an anode. In the forward part of the loop, the electron wavefunction might have spread to many points on the anode. It is only when part of the anode sends a positron backward in time, essentially saying 'Yes, I have a place for you' that the wavefunction becomes real along the trajectories between the cathode and the receiving point on the anode.

Essentially, information from the anode in the future causes the cathode in the past to send an electron to it. But only because the cathode sent an electron to it: a causal loop!

I've read this a few times now and I'm still not sure I grasp the thesis. That is, I'm not sure which you think are the key points. The most peculiar part is the sense in which no time passes for, and no space is traversed by, light: a feature of SR that has always amazed and alarmed me. As you put it, the tick of any clock in a physical reference frame is due to the finite time between trajectories of the Higgs re-emiting particles. (We've discussed this before and as you know I'm not so sure we should take this literally, but I'll put that to one side.)

But I'm not clear how this all fits together for you. Is it that there's no effective separation between our seeing and the thing we see, except for a fluke of Higgs fields and reference frames?

You have no obligation to respond just because you are mentioned, but there's no need to be rude. — SophistiCat

Is this a wind-up? I have never had so many notifications on a thread I have no interest in. And please don't give me lessons on manners. I answered one question directed specifically at me and you accused me of hijacking the thread. Manners are not your strong suit either. Stop @ing me. It's just weird apart from anything.

You can't actually excape the first person perspective. — Yohan

A neuroscientist can scan your brain and determine what decision you'll reach before you reach it, for some test scenarios. There's a good reason for this. The usual best solution to a problem is whatever has worked for similar problems before. Therefore recognising patterns and matching them to outcomes is often more efficient and efficacious than identifying the options and choosing one. The will is anchored by what the unconscious recommends. We are free to override it, but typically don't. The first person perspective is that we made a sensible choice. The third person perspective, in those sorts of problems amenable to these studies, is that you matched facts to mental maps, found a solution, then decided that this was the solution "rationally". One of these perspectives is highly untrustworthy :)

We could try for intentional non-procreation. Give it a try! — schopenhauer1

The best (and only good) anti-natalism argument.

Is it conceivable that there is a world where events have connections, but the connections are not mechanical? That is, for a given state at T0, more than one future state of the system is possible? — Echarmion

Yes, that's the Copenhagen interpretation of quantum mechanics.

Determinism is the theory that every event is the result it's causes, probabilistic thinking (or stochastic for that matter) is about our ability to know with certainty what that event will be given the causes. The two are not only mutually compatible, they're not even in the same subject area, I'm lost as to why they keep getting treated as mutually exclusive options for the 'way the world is'. — Isaac

Yes, the net is being cast very broadly in what constitutes randomness. Our ignorance, the complexity of the system, and the sensitivity of its laws to initial conditions are apparently all the same thing as wavefunction collapse. Seek and ye shall find.

And that's a fundamental problem. I.e. we cannot just improve our measuring apparatus in some way. Either we come up with new physics, or this stays, whether it's an actual ontological reality or not? — Echarmion

That's right, which is why I cited the measurement problem as a potential disproof of determinism. Unfortunately we don't have a testable theory of measurement to see whether it is random (Copenhagen-like) or deterministic (MWI-like).

There is no evolutionary mechanism by which traits that confer a fitness disadvantage are removed. — Echarmion

Yes there is: death!

A bayesian might well argue that far from the world appearing to be deterministic, it actually appears probabilistic. — Echarmion

I would wonder why he was arguing anything unless he was pretty certain he'd be understood.

I wasn't so much complaining about a derail. It just seems that you (or maybe just Olivier) are itching to have this discussion - so why not have a dedicated topic for it? That would invite wider participation. — SophistiCat

I'm not involved in this thread at all. Stop @ing me in it.

But I then say that the concept of an experience is inherently a relational one: someone has an experience of something. An experience being had by nobody is an experience not being had at all, and an experience being had of nothing is again an experience not being had at all. This indubitable experience thus immediately gives justification to the notion of both a self, which is whoever the someone having the experience is, and also a world, which is whatever the something being experienced is. — Pfhorrest

Does this resolve beyond the cogito? I can doubt whether I exist, which implies the existence of a subject (self) and an object (self) because they are one and the same. But doubting the world exists doesn't imply a world beyond the self, right?

Can one of you guys start a thread on determinism/indeterminism, instead of hijacking other threads? (I have the damnedest time making OPs, but I might contribute if there is one.) — SophistiCat

Excuse me, I did one post on the matter because I was explicitly asked for input. I am not hijacking the thread.

I understand why you're evading the question, but even you must see this is horrifically inconsistent. On the one hand, things seem deterministic but, because of small error, the thing itself is non-deterministic. On the other hand, we might not be able to say anything about the thing itself, but if you can use stochastic methods it is, again, non-deterministic. Your argument is a black box: it doesn't matter what goes in, non-determinism comes out. This classifies it as a religious belief for me.

Just said that stochastic phenomenon look like a duck. — Olivier5

Okay, so irrespective of the actual physics, if we use stochastic methods to model them, they're random. So when I said:

if you are aware that stochastic methods are used to model deterministic processes and yet insist that anything modelled stochastically is random, you do give up the right to be taken seriously — Kenosha Kid

you really were insisting that anything modelled stochastically is random, irrespective of whether it is.

When I see a phenomenon that displays a behavior resembling randomness (eg the Galton box and its results plotted against a Gauss curve), I say it looks like randomness — Olivier5

No, I understand that, that is, as you say, very simple. Simplistic, even.

You said the above paper 'looks like a duck', i.e. "displays a behaviour resembling randomness". Can you walk us through that reasoning please?

I'm not really insisting that, if it looks like a duck, walks like a duck and quacks like a duck, it must be a duck. I'm just saying that this is the conclusion I will draw, personally, because I see no good reason to assume it's an elephant instead. You on the other hand, when you see that it looks like a duck, walks like a duck and quacks like a duck, you conclude that it ought to be an elephant... That is your call, not mine. That sounds pretty odd to me from an empirical epistemologic perspective but you are entitled to your opinion... — Olivier5

So, let me get this straight... You are aware that stochastic methods are employed to model both random and intractable, statistical systems. And you're not insisting that something modelled stochastically is automatically therefore random. So presumably you're telling me that the above paper is using stochastic methods to model something that looks random, i.e. walks like a duck. Walk us through that.

I can do so very easily. If it looks like a duck, walks like a duck and quacks like a duck.... Why would I not infer that it's a duck? — Olivier5

As you will, but if you are aware that stochastic methods are used to model deterministic processes and yet insist that anything modelled stochastically is random, you do give up the right to be taken seriously.

And they are also used for all sorts of calculations about random events. — Olivier5

Indeed. Point being you can't read 'stochastic' and infer 'non-deterministic'. For instance, Brownian motion is modelled stochastically. That is not to say that collision events are random, simply that the particles involved have unknown initial states that, were they known, could not be feasibly tracked over time, but nonetheless are statistically predictable.

And the result of all this is that we cannot predict the exact time an individual atom will decay. We can only give probabilities for timeframes, correct? — Echarmion

Yes, that is the measurement problem. At a given time, we know how much of the wavefunction should be |decayed> and how much |undecayed> but we don't know which we'll see when we measure.

Traits might also genetically linked, so that a trait that actually does nothing to improve inclusive genetic fitness becomes dominant because it's linked to other traits that do. — Echarmion

That would still be selected for. Sickle cell disease is an example. It confers a survival disadvantage in and of itself, but ends up making the odds of survival greater. If sickle cell disease conferred no survival benefit due to immunity from malaria, it would have been eliminated from the genome due to its survival disadvantage.

We could imagine that the motion is probabilistic, but with such a narrow Amplitude (is that the right word) that the inaccuracies wouldn't matter for everyday purposes. — Echarmion

Which is back to the non-determinism of the gaps: anywhere where it might show its face it is obliged to hide in tiny error bars.

Stochastic
Having a random probability distribution or pattern that may be analysed statistically but may not be predicted precisely.

Well, that's what I'm saying. Nature behaves as if there was some randomness in there... — Olivier5

Stochastic methods are used for all sorts of intractable deterministic, statistical applications.

But it's dependent on knowing the laws of the universe, which is equally esoteric. — Olivier5

No, it's not. Being able to approximate the laws of the universe well is dependent on the universe -- or at least the part of it being modelled -- behaving deterministically. The universe can behave deterministically without intelligent life modelling it. Intelligent life cannot model it scientifically if it isn't there.

Now I'm curious, do expound. — Olivier5

Perhaps in another thread, but it's basically a mish-mash of transactional quantum mechanics and many-worlds interpretation: determinism forwards and backwards in time.

It's non local, in particular. Which means you can never isolate any sub-set of events from the rest of the universe in any calculation. — Olivier5

That's true more generally, e.g. the EPR paradox is an example of ultra-nonlocal behaviour. Ultra-nonlocal behaviour was actually the subject of my PhD. That sort of behaviour is not problematic in transactional QM.

Interestingly, I've just found this paper: https://www.nature.com/articles/s41598-019-56357-3#ref-CR29

This argues that nature is fundamentally stochastic. This is not necessarily random, though. Stochastic modelling is suitable for the sorts of unknowable or intractable deterministic processes previously discussed.

I am really just talking about the probabilistic quality of QM, the fact that e.g. at what exact time an atom of a radioactive element decays appears random. — Echarmion

So, yes, radioactive decay is an example of a quantum field theory, the electroweak theory. But it isn't characterised by atoms or hadrons either spitting out or not spitting out components at random. The system evolves deterministically through both paths simultaneously, both decaying and not decaying, until the wavefunction collapses/universe branches/superposition decoheres/whatever else happens to yield singular observables.

This is presupposing that every attribute we have right now confers a survival advantage. — Echarmion

That doesn't follow. It just needs to have conferred a survival advantage to our ancestors.

There'd be nothing mysterious about the unpredictable outcomes. They'd just be things that the natural laws make difficult to predict — Echarmion

What do you mean, difficult to predict? For instance, would it make the vertical component of motion of a ball on an inclined plane difficult to predict?

I think Kenosha’s point is that despite the above, indeterminism sorta has the burden of proof here, because so far everything that we have been able to know has turned out to behave deterministically, so we should expect that to continue to be the case as the limits of our knowledge push further and further out. If something seems unpredictable at the moment, it’s probably just because of shortcomings on our measurements or theories, not because it’s inherently random. — Pfhorrest

Yes. The Universe appears to behave deterministically... We can write down formulae to predict outcomes and find them reliable: x in, y out. (With or without error bars.) If the universe were non-deterministic, it would have to seem deterministic, and thus any evidence of determinism would automatically be worthless. For that reason, this kind of constrained non-determinism is unfalsifiable: no evidence can disprove it. This is the same problem as the God-of-the-gaps argument.

However we should, in this non-deterministic universe, expect some behaviour that cannot be generalised well. There should be mysteries as to why we cannot predict outcomes. These could falsify determinism, and the measurement problem may well be a good example. But since unfortunately we have equally good/bad deterministic and non-deterministic theories for this, neither backed by knowability or tractability, the jury is out for now.

The use of probabilities could be down to measurement errors, chaotic systems, accuracy at scale, informational constraints, ...etc. Why would you see it as evidence of those fields not being fundamentally deterministic? — Isaac

Ah. Yes, having basically the same conversation with the same person on another thread. I didn't really get anywhere with it, but good luck.

I watched a thing by Jim Al-Khalili about something like that a long while back, but not having much understanding of the basics I didn't really come away with anything more than a very general picture. I didn't get the impression that biochemicals were going to suddenly start reciting Shakespeare or forming an impromptu dance troop any time soon though, so I think we're still safe to presume they'll continue to have the effects we've so far discovered them to have! — Isaac

Yeah, I think it's pretty mundane stuff. Certain quantum phenomena like tunnelling and the exclusion principle have small effects on simple chemicals and atoms. Charge transfer is ultimately a quantum phenomenon.

I haven't caught up on this convo. Is this the usual "<magic thing> is possible because quantum mechanics"?

That's not where the uncertainty comes from, the way I understand it. The uncertainty is fundamental. Not all values of the system can be known at a time, and the values that are not known can only be expressed as probabilities. — Echarmion

I think we're speaking at cross purposes; possibly I misunderstood your earlier point. Reading back, what I think you were referring to was random fluctuations in fields, i.e. creation and annihilation of particles whose lifetimes are consistent with the uncertainty principle. At any given time during an electron's trajectory, for instance, there is a finite probability that a virtual photon will be created by some distant charge that will scatter the electron. Since this is always true, the probability of it happening approaches 1. This is the gist of the quantum field theory of the electric force.

Sure, we can create plausible theories to explain how certain vestigial or otherwise weird anatomies came about. But that doesn't establish that the end result was selected for. Only that there wasn't sufficient pressure to select for a different result. — Echarmion

Yes, plausible. Compared to hypothesising that entire organs came into being for no reason, or that a magical man made it for no reason. If we go that route, it becomes a mystery why we don't all have organs that not only serve us no purpose, but served no purpose for any of our ancestors, in addition to the mystery of how. The resolution of that mystery is natural selection.

Bear in mind the starting point for this tangent was the claim that life has evolved characteristics that could not have been selected for. That still remains unshown.

Meanwhile, your argument, if applied to e.g. the appendix, would lead one to look for the benefits the appendix provides to modern humans to explain its existence. — Echarmion

???? My argument is that characteristics that benefited our ancestors can be passed down to us whether they benefit us or not. Evolution would not be a deterministic process if organs disappeared the moment they became useless.

Why do we need an explanation in the first place? Explanations are tools for specific ends, not an inherent necessity. — Echarmion

Fine, don't seek explanations then.

Exactly, and hence determinism is a rather esoteric idea. — Olivier5

Determinism is not dependent on being able to rewind the universe.

And likewise, you don't like the idea of randomness and you try to erase it from your POV, when I see it everywhere around me. To each his own metaphysics... — Olivier5

As I said, you don't know my position. It would probably surprise you.

Isn't this a misconception? Not being able to measure position and speed of a particle does not necessarily mean it ain't at an exact position and speed at a given time. — Heiko

It's not a misconception that you cannot. There is a version of QM called Bohmian mechanics in which particles do have exact position and momentum simultaneously. It is not well liked for other reasons.

There's been a recent paper in which quantum entanglement between two large (i.e. compared with atoms) objects were entangled over a distance, but yes generally it is thought that large objects have too many degrees of freedom to support coherent superposition. Also, cells are relatively hot, and heat also kills superposition.

But cells are still tiny, and confinement amplifies quantum effects. Quantum biochemistry is a thing but unfortunately not one I know a lot about.

What is not compelling to me is the story that, if the universe was to magically rewind at the time of the Big Bang and unfold again, every single thing will happen exactly the same as it did the first time around, like when you play the same movie over again — Olivier5

And yet when you rewind a VHS and play it again, you do expect exactly the same movie, not a random one. ??? That's real. Magically rewinding the universe is not.

What this comes down to, then, is taste. Irrespective of how much experience you have that doing the same thing in the same way produces the same result, you don't like the idea of predetermination so say the universe is random. Fine. But the universe isn't obliged to cater for your taste.

And QM is not a gap. Randomness is systemic in it, and it applies supposedly to the entire universe. — Olivier5

Yes, it is. We don't know whether the universal wavefunction is Copenhagen-like or MWI-like or something else. It remains to be seen.

Indeterminism says that some things are predetermined to a degree, but not necessarily everything and not necessarily to a perfect degree. — Olivier5

Yes, this is indeterminism-of-the-gaps. As our technology improves, error reduces, and this alleged non-determinism of nature is obliged to retreat. It is not a compelling or useful story.

In no situation can we possibly know all relevant information about a case, and even if by miracle we did know everything relevant about a case, we couldn't be sure of it. — Olivier5

If this were true, science wouldn't work, technology wouldn't work. One has to be able to know a sufficient amount of information in order to guarantee regularity of outcome and if we can do that -- which we clearly can -- then we can know what we need to know. This regularity of outcome is itself determinism.

We can never measure anything exactly, there's always a margin of error. — Olivier5

Error is not the same as non-determinism. Particle physicists can measure the magnetic moment of a neutron to within $10^-29$. That doesn't mean that because there is error, that quantity is not determined, and tomorrow someone could measure it as 1000 it's measured value. It just means there are technological limits to measurement. We would still see neutrons respond the same way to the same magnetic field, which is determinism. Error is not a path to non-determinism: non-deterministic behaviour is.

We can never be sure that any of our scientific theories is true. — Olivier5

You quote Popper like he's an authority, but demand proof? As I've said before, you don't need to know about general relativity to know that a ball on an inclined plane will roll downhill. That is also determinism.

The first two conditions will never be met. The third one would require infinite energy and time so it will never happen... None of these conditions will ever apply in our human lives. So much so that determinists appeal to various demons in their demonstrations, like the Laplace's demon. — Olivier5

That adds nothing. The above is only important IF we also wanted to know the exact position of every particle at some future date. To know whether a ball will roll downhill or uphill, none of that information is relevant.

No, I'm French. — Olivier5

Okay cool. Sorry, it's come across several times like you're willfully misunderstanding everything. Obviously there's a slight language barrier and I should have considered that. My bad.

It's not complicated to abandon an hypothesis, especially when it makes no pragmatic difference whatsoever. — Olivier5

Can I ask... Is English your first language?

Obviosly you're guessing. — Punshhh

Oh? You get your knowledge that

The Lib Dems only talk about UBI in passing, they wouldn't add it as a manifesto commitment, they would lose half their base. — Punshhh

from divine revelation, I suppose. If you're going to be an idiot, I'll leave you to it.

Isn't that essentially Einstein's argument of the hidden mechanics? There is no evidence, right now, that the uncertainty can be resolved. — Echarmion

What I mean is that as we measure, say, the spin of a neutron to ever greater precision, the degree of freedom of non-determinism to show its face gets ever smaller.

Lots of organs are weird and inefficient. The human eyeball is a common example, as are various vestigial limbs found in species. — Echarmion

Ah, but that doesn't mean they weren't selected for. We have an appendix that is useless to us, but we are descended from grass eaters.

Again this isn't "proof". I obviously have no idea how the universe " really" works. — Echarmion

It can't be proven. My point was just that you have an extremely simple explanation for consensus -- determinism -- or a really complicated and dubious one.

Why would random fluctuations in the wave function — Echarmion

There are no random fluctuations in the wavefunction. In even the probabilistic interpretations of QM the wavefunction evolves deterministically under a wave equation until measurement.

Prima facie, the way the universe actually appears to work is absurd, at least to our everyday notions. — Echarmion

Well QM is certainly absurd according to everyday notions, whether it proves deterministic or not.

Doesn't that show that what appears to be a determined, "mechanical" apparatus can turn out to be anything but? — Echarmion

Definitely. I'm not arguing for a deterministic view of QM though. Whether QM is Copenhagen-like or MWI-like is not within our grasp atm. I'll go where the evidence goes. But by the same token, it can't be used as proof that the universe is macroscopically or microscopically non-deterministic.

That proves that the universe isn't so random as to prevent these kinds of predictions — Echarmion

Or indeed any kind of prediction at the moment. It is unfortunate and perhaps not uncoincidental that the realm of physics where determinism is in doubt is the one where we cannot make predictions. We can't solve the equations even crudely for measurement apparatus. That's a huge problem.

But what you're talking about here is Popper's indeterminacy of the gaps. This presumably fundamental randomness of the universe is weirdly constrained to whatever our peak technological capability ends up being.

But we know that not every attribute of every organism is actually selected for. — Echarmion

Such as? There's the supposed junk DNA, but they are not physical characteristics that can be provided for. If there is a characteristic that benefits us, there has to be an environment in which that benefit can be exploited. Or God, I suppose. Something, anyway, that persists useful characteristics from generation to generation for hundreds of thousands of years, despite evolution apparently being purely random. ???

We don't check these against an objective reference point somewhere. It's only when new information does not fit the pattern at all that we re-evaluate and then only to find a new solution that is "good enough". — Echarmion

Very true, science is not divine revelation. But we do have consensus. A non-deterministic theory of nature not only had to explain why you experience the same phenomena under identical circumstances, but why everyone else does so too. So far, no one has reported that a ball on an inclined plane had a 50/50 chance of rolling up.

I'll take that as a no. — Olivier5

Do as you please, but don't expect kudos for it.

So... Did you find that magical experiment yet? — Olivier5

I found that no amount of explanation is going to demonstrate how that question demonstrates zero understanding.

One has to know the true laws of nature in order to know if they are determinist of not. — Olivier5

Again, not relevant. Newton's laws are sufficient to wang a probe around the solar system for a few decades and land it in your back garden. Tackling this with GR would again be intractable. But even more broadly, Galileo was able to make perfectly good predictions about balls rolling down inclined planes without knowledge of general relativity, or whatever true law GR is an approximation to.

Otherwise you're only testing an hypothethis, which you can reject if the experience fail, without rejecting determinism. Because another determinist hypothesis may still work. — Olivier5

Yes, that's true. I could write down a mathematical formula that describes how a ball released from a box will accelerate upward into the sky at 9.8 m/s/s then find that my theory is wrong. But, as I've already explained, that is irrelevant. What would be relevant is if I released the ball and sometimes it stayed still, sometimes it flew upward, sometimes it disappeared completely, sometimes it quoted Shakespeare. If that happened, with all relevant information about the ball available to me and no conditions in the setup that made me either predict those outcomes deterministically or lead me to believe I could not make a prediction that would discern deterministic from non-deterministic behaviour -- and no such possibilities spring to mind -- I would not expect a new deterministic theory to explain it and would likely consider determinism well and truly falsified.

The tractability of a problem is relevant insofar as I can have some expectation of what outcomes would be considered consistent or inconsistent with determinism. For instance, the Schrödinger's cat thought experiment is intractable because I cannot solve the time-dependent many-body wavefunction of a radioactive isotope, radioactivity detector, hammer, glass vial of poison gas, cat and box. I can have no real expectations of an outcome, and no means of establishing whether repeated experiments are at all comparable.

Irrespective of the particular theory of gravity, I do have the means to compare what happens to a ball released from a box. It is a tractable problem. What the final state of the ball is is not important. The phenomenon could be completely non-deterministic and still consistent with a deterministic theory if I only check once. What is deterministic is being able to repeat the same experiment under the same conditions (to whatever precision is relevant) and expect the same result.

You are doing this already btw. Whenever I hear this sort of thing I can't not chuckle, because there you sit, hitting keys on your keyboard with utmost expectation and surety that when you hit the key T a T (or t, as you select) will appear on your screen in the expected place and when you hit send it will with complete fidelity appear on everyone else's screen precisely where you intended, as you explain to us that the universe is not deterministic. I would wager that, were it a R in place of a T, you would blame yourself for a typo rather than blame the complex infrastructure we have for its lack of fidelity or the Universe for its lack of determinism. Your sheer persistence in arguing for a non-deterministic world is a testament to how deterministic you really know the world to be. And there's quantum mechanics involved in that process between you hitting a key and me reading your T to boot.

You are not God, are you? — Olivier5

If the relevant information required omniscience, it would not be a good test for determinism. Gimme a holler when you acquire the ability to retain information.

For example some cheating could well be at play. — Olivier5

Through gritted teeth... all relevant information