What I think you mean is that one of the theory's predictions can not be tested (that there are many worlds), at least at this point. — Andrew M

Sure that's part of it. Are you aware of Sabine Hossfielder's recent book, Lost in Math? And all of the huge arguments going on about whether string theory is or isn't science? I know these are not about exactly the same issue as the 'many-worlds interpretation' but they're metaphysical ideas arising from science, rather than testable hypotheses as such.

I don't think 'the Copenhagen interpretation' is, or attempts to be, a scientific hypothesis. It is just a collection of aphorisms and philosophical reflections, principally by Bohr and Heisenberg, which are about what you can and can't say on the basis of the discoveries of quantum mechanics. One of the appealing features of this 'interpretation' is that it's not masquerading as science - it's actually quite a modest attitude - whereas the 'many worlds interpretation' is presented as scientific when I really doubt that it is. (Anyway I know that I have zero credentials in this matter, I will bow out at this point and attend to more pressing mundane issues.)

Its from an article called Empathy and Consciousness.
https://evanthompsondotme.files.wordpress.com/2012/11/jcs-empathy.pdf

Thanks. I have the preview of his latest on Kindle.

I don't think 'the Copenhagen interpretation' is, or attempts to be, a scientific hypothesis. It is just a collection of aphorisms and philosophical reflections, principally by Bohr and Heisenberg, which are about what you can and can't say on the basis of the discoveries of quantum mechanics. — Wayfarer

Maybe so, but the end goal is to have rigorously defined theories that can be experimentally distinguished. Part of that task is identifying the assumptions made by the various interpretations (like counterfactual definiteness, free choice, locality, observer-dependence, etc.). Researchers in quantum foundations use this data to come up with no-go theorems and experiments such as that in the OP.

Interestingly, the OP experiment is based on Deutsch's version of Wigner's friend which he proposed specifically to distinguish experimentally between Copenhagen and MWI. Caslav Brukner (mentioned in the OP article) discusses this in https://arxiv.org/abs/1507.05255.

So, according to the article, the notion of objective reality has not been unequivocally undermined, as your headline asserts. It might be the notion of freedom of choice or the idea of locality which have been undermined; the article only claims that it must be that one of the three is wrong. — Janus

Well, here 'freedom of choice' does not refer to 'free will'. Rather, it is a denial of Superdeterminism (the link is to Wikipedia article), that is the idea that choices were pre-determined at the beginning of the universe. It is actually stronger than 'simple' determinism (like the one that de Broglie-Bohm interpretation (dBB) accepts), because the 'history' of events in the universe is 'already written' at the beginning of the universe. Superdeterminism is a known 'loophole' of Bell's theorem.

According to Bell's theorem, one cannot accept the predictions of QM, counterfactual definiteness (the view that we can speak meaningfully of experiments that have not been performed) and locality (assuming that superdeterminism is not true). dBB does not accept locality.

Note, however, that even if one does not accept counterfactual definiteness, one can still accept an 'objective reality', hence the article is mistaken in that claim. For instance, MWI accepts an 'objective reality', which is the quantum state of the entire universe but does not accept counterfactual definiteness. Objective collapse theories accept an 'objective reality' but do not accept counterfactual definiteness because results of experiments are seen as random (according to these theories, wave-functions are real, physical objects that 'collapse' at the measurement).

Agree. MWI says there is an objective reality, but it is entirely in superposition, and measurement just entangles the measurer with the measured thing. It does not collapse any wave function. Hence there is no defined state of anything (like dead cat), and hence no counterfactual (or even factual) definiteness. — noAxioms

Yeah! BTW, I believe that other than the very problematic concept of 'many worlds', MWI has a serious problem, check: https://arxiv.org/abs/1210.8447. The usual claim that the 'preferred basis problem' is solved by decoherence. But it is not correct. Decoherence solves the 'preferred basis problem' (in fact, 'for all practical purposes' in my feeble understanding) only if you already assume that there is a well-defined factorization in the Hilbert space (which is the only 'reality' in MWI, AFAIK). Without well defined subsystems, the factorization is completely arbitrary (also, it should be added that, in fact, one has no, a priori, reasons to do a factorization in the first place).
I do not know how MWI-supporters handles this in a non-circular way.

I also add that MWI and RQM are close. The difference being that RQM does not accept the reality of the 'universal wavefunction' because, in RQM, wave-functions are well-defined in relation to a specific physical system (the 'observer' in this interpretation).



Very interesting quote by Bohr, thanks for sharing! At that time, he seems to have held a view similar to Carlo Rovelli (who believes that you can define a 'perspective' for every physical system).

Nonetheless, the main problem with this view is that if the wave-function is considered to be information or a 'mathematical tool' (as in my understanding Rovelli does), then it is difficult to understand how we can speak of 'information' related to a non-conscious observer. This is, in fact, Michel Bitbol's point.

Apparently, however, Bohr changed his views over time. Check, for instance, this paper by Bitbol that I linked before:

http://www.bourbaphy.fr/bitbol.pdf — boundless

In other words, according to Bitbol, Bohr's views are similar to Kant's philosophy. While RQM is right in saying that QM is about 'perspectives', according to Bitbol these perspectives are well-defined only for conscious observers.

I haven't read the paper yet, but I did a search for references to it on physicsforums and found nothing, including in more than one recent discussion on the Wigner's Friend thought experiment. So I suspect the experiment doesn't say what the MIT article claims it says. To have a situation where two observers can obtain contradictory measurements, rather than just measurements with differing levels of detail, would be too epoch-making to ignore. — andrewk

The abstract of the paper:

The scientific method relies on facts, established through repeated measurements and agreed upon universally, independently of who observed them. In quantum mechanics, the objectivity of observations is not so clear, most dramatically exposed in Eugene Wigner's eponymous thought experiment where two observers can experience fundamentally different realities. While observer-independence has long remained inaccessible to empirical investigation, recent no-go-theorems construct an extended Wigner's friend scenario with four entangled observers that allows us to put it to the test. In a state-of-the-art 6-photon experiment, we here realise this extended Wigner's friend scenario, experimentally violating the associated Bell-type inequality by 5 standard deviations. This result lends considerable strength to interpretations of quantum theory already set in an observer-dependent framework and demands for revision of those which are not.

And the paper itself is titled "Experimental rejection of observer-independence in the quantum world".

No interpretation is a cop out, but MWI cannot have those observers in different world branches since they communicate. Alice knows the polarity and tells Bob that she does. Bob knows that the particle is still in superposition and tells Alice so. That cannot happen if the two are in different branches.
— noAxioms
In the experiment, Alice can communicate to Bob that she has measured a definite polarity (without the polarity itself being revealed) while the lab she is in remains isolated (and Bob does not communicate back, which would presumably constitute a measurement entangling him with Alice). So there are actually three MWI branches here. One where Alice measures a horizontal polarization, one where she measures a vertical polarization, and one that is the superposition of those two branches where Bob detects interference (and knows that Alice has made a measurement). — Andrew M

I have to disagree about the restrictions to communication you convey above. Alice knows the result of a measurement, and that makes for 2 Alice's now, one for each result. Those two version of Alice, being in different worlds, cannot communicate or otherwise be aware of each other. But they behave exactly identically because they're keeping that knowledge a secret. To Bob, Alice is in superposition of knowing those two states, and Bob can thus communicate two-way with both Alice's since they, by acting identically, are completely coherent. They can make out if they want. In reality, humans are incapable of this coherence, which is why they never use humans to play the role of Alice or Bob.

Anyway, point is, Alice learning of the measurement results splits Alice, but does not split the universe, as is commonly assumed. Bob, being able to speak to both versions of Alice, is still in a common world. So yes to the three worlds if you count them that way: One for each Alice, and one for Bob. But there is obviously communication between the Bob world and both Alice worlds, but Bob cannot pass a message from one Alice to the other. With the communication, Bob's world is clearly not isolated from Alice's world, and hence doesn't really count as a separate world.

BTW, I believe that other than the very problematic concept of 'many worlds', MWI has a serious problem, check: https://arxiv.org/abs/1210.8447 . The usual claim that the 'preferred basis problem' is solved by decoherence. But it is not correct. Decoherence solves the 'preferred basis problem' (in fact, 'for all practical purposes' in my feeble understanding) only if you already assume that there is a well-defined factorization in the Hilbert space (which is the only 'reality' in MWI, AFAIK). Without well defined subsystems, the factorization is completely arbitrary (also, it should be added that, in fact, one has no, a priori, reasons to do a factorization in the first place).
I do not know how MWI-supporters handles this in a non-circular way. — boundless

I actually don't know the terminology that well, in particular 'factorization'.

So perhaps I don't understand the problem here.

I also add that MWI and RQM are close. The difference being that RQM does not accept the reality of the 'universal wavefunction' because, in RQM, wave-functions are well-defined in relation to a specific physical system (the 'observer' in this interpretation).

I'm an RQM guy myself, and yes, nothing is just 'real', things are only real in relation to something else, so how can the universal wave be real when there is nothing to which it is real in relation? The view would be self inconsistent if it were to be otherwise.

I don't think 'the Copenhagen interpretation' is, or attempts to be, a scientific hypothesis. It is just a collection of aphorisms and philosophical reflections, principally by Bohr and Heisenberg, which are about what you can and can't say on the basis of the discoveries of quantum mechanics. — Wayfarer

Close. It isn't philosophy at all. The 'interpretation', unlike other philosophical interpretations of QM, is just a scientific statement concerning what is known about a system. Hence it is, as far as Bohr, Heisenberg, and Schrodinger are concerned, just an epistemological statement, not a metaphysical interpretation. There are plenty who take that epistemological wording also as some kind of statement of reality, but Copenhagen was not intended to be used this way.
So the taking of a measurement changes what we know, solidifying some possibilities and eliminating others, and hence collapses the wave function of the possible states of the thing. The wave function is not real, it just represents possibilities for something unknown. There is a wave function of places where I likely left my car keys, with some more probable than others. When I find them (or even when I look certain places and don't yet find them), that wave function changes since my knowledge of the system has been changed.

The wiki table on interpretations lists Copenhagen as a non-local interpretation, and I don't understand that. My knowledge of a system doesn't change due to an event that happens elsewhere. But I suppose that my knowledge of a distant system (like the distant half of an entangled pair) changes immediately upon my measurement of its local sibling, so maybe that's why they list it as a non-local interpretation.

↪fdrake Indeed - I think that is very much the kind of understanding that has emerged from science in the last several decades. But I think it sits oddly alongside what you said previously. — Wayfarer

Well we know that physical processes interact with others, it is not so surprising that their interaction can effect all involved or produce novel phenomena. This maxim applies more generally than in quantum physics - ecosystems can couple and interact, so can social groups. The details depend on the specifics.

I think that nature is inherently relational only sits oddly with intuitions that relations are somehow derivative or less important then their relata; like the two poles of subject and object, where everything interesting lays between.

Notice that Bohr says that ‘the objective world of nineteenth century science’ has become untenable. I think he’s correct in saying that, but isn’t it very much the substance of his disagreements with Einstein? — Wayfarer

Quantum phenomena undermine some aspects of determinism, relativity undermines absolute space and time. Sub specie aeternitas this requires us think of nature in the wake of discovery in both fields, but also when things are big enough or slow enough where the weirdness of either or both does not apply. Objectivity there, I think, means something like invariant of perspective - observer/frame independence. We know nature is not either in many circumstances. The trick is then in giving an immanent account of 'perspectival' dependence without making humans a necessary constituent of nature.

It does not help that the usual words we use to discuss these topics are perception analogies.

I actually don't know the terminology that well, in particular 'factorization'. — noAxioms

Ok! I'll try to give an explanation.

In MWI, there is only a quantum system, the universe itself. Its quantum state is a vector in a Hilbert space.

Now, consider a complex quantum system, that is a quantum system like, say, a pair of particles. Let us call them P1 and P2. To each particle is associated a Hilbert space, say, respectively, H1 and H2. To the total system we associate the Hilbert space, H, which is the tensor product of H1 and H2. So, the quantum state of the total system is a ray in the Hilbert space H, which is 'factorizable' into H1 and H2, the Hilbert spaces related to each particle. Here, the factorization is well-defined by the two particles themselves.

In the case of MWI, however, the only real system is the universe itself. Without additional structure, you can't speak about subsystems. So, you need a factorization, i.e. a decomposition of the Hilbert space related to the universe (which, however, we can argue that is needed a-posteriori because we observe subsystems. A-priori, there is no reason to even do a factorization). But even if we make a factorization we see that such a factorization is arbitrary and in some factorizations literally nothing happens (I suggest to read Scwindt's pre-print. I do not claim to have understood it completely, but IMO it explains well the point...). Since, however, in principle, the factorization is arbitrary, we can choose such factorizations where nothing happens.

The point is that people claim that MWI is more elegant than other interpretations because there is only one real thing, the wave-function of the universe, which never collapses. On the other hand, this is a moot point because in order to explain the multiplicity we observe you need to factorize/decompose the Hilbert space associated with the quantum state of the universe in a given way. MWI by itself however simply cannot do that (as I said, I do not know if MWI supporters have offered a counter-argument).

The only way 'out' seems to introduce an additional structure in the Hilbert space of the universe. But, at this point, how is MWI really 'simpler' than, for instance, the de Broglie-Bohm interpretation (which is an example of 'way out' offered in the paper)? In the de Broglie-Bohm interpretation, you still have the universal wave-function but you also have the particles that define the factorization/decomposition (that is, the theory itself offers you the actual subsystems). The other 'way out' that offers the paper is the Copenaghen interpretation. Of course, there are others but the paper offer these two.

To summarize, MWI claims that the only real thing is the 'universal wavefunction'. But without introducing additional structure it seems that there is no way to explain the 'multiplicity' we observe - to do that we need actual subsystems that introduce a factorization/decomposition.

I'm an RQM guy myself, and yes, nothing is just 'real', things are only real in relation to something else, so how can the universal wave be real when there is nothing to which it is real in relation? The view would be self inconsistent if it were to be otherwise. — noAxioms

Yeah! Speaking of myself, I lean towards either RQM or (some versions of) Copenaghen.

Nonetheless, the main problem with this view is that if the wave-function is considered to be information or a 'mathematical tool' (as in my understanding Rovelli does), then it is difficult to understand how we can speak of 'information' related to a non-conscious observer. This is, in fact, Michel Bitbol's point. — boundless

Information and probability are dual notions; wherever you have a probability distribution you have an entropy. The connection between the two is particularly intimate for discrete random variables - like when there is a given probability of being in one of countably many eigenstates of an operator. Quantum entropy measures the degree of mixing in a state; how close it is to behaving in a singular eigenstate (unless I'm misinterpreting, I am both rusty and mostly uneducated here). Information measures are derivable from probability distributions, but the process of mapping a distribution to an entropy value is not invertible - so the two notions can't be taken as inter-definable. As in, if you have an entropy, you have a single number, which could be generated from lots of different quantum states and probability distributions.

I'm sure there are problems, but I think there are good reasons to believe that information is just as much a part of nature as wave functions.

consciousness?’ because natural objects and properties are not intrinsically identifiable ; they are identifiable only in relation to the ‘conceptual imputations’ of intersubjective experience." — Joshs

I'm tired of debating whether nature is inside our theories of it or not. Imagining that nature is all we have to say about it is a pointless retrojection; it has been around longer than theory and impresses itself upon us even when we have no account of it. Yes, you will probably say, science produces truths, but more profoundly those truths are indexed to our theories and that those truths attain their sense solely within our theories.

The questions we ask nature are thus muted after leaving our lips; nature's behaviour becomes covered by the silent fog of our understanding, we could never ask questions of it, only of our relation to it. Our theories' mediating status between us and natural phenomena becomes bloated and complacent, it takes the target of theorisation to be the theory produced and not what domain is actually theorised.

Don't put our understanding in the way of our understanding. Questioning and inquiry are themselves ways to relate to nature.

Wigner: "Hey, that cat is either dead or alive."
Friend: "Yeah, I know which one!"
Wigner: "No, you don't, it's either dead or alive."
Friend: "I'm telling you it's dead."
Wigner: "Oh, I see, indeed yes, it's dead."

The funny stuff aside; how can you "measure" a superposition or how can a measurement be in a superposition? I thought any observation causes the wave function to collapse in a single eigenstate and a measurement, I would think, involves an observation.

I'm a bit unclear on what the article means what it says:

Wigner can even perform an experiment to determine whether this superposition exists or not. This is a kind of interference experiment showing that the photon and the measurement are indeed in a superposition. — technology review

I thought that saying a system is in a superposition is a description of its possible states and not a statement of its actual state if you would measure it.

What am I missing here?

"Good Heavens, Holmes! They may have been the footprints of a gigantic hound after all!"

You can't use your everyday conceptions to predict quantum stuff. The usual objective picture is in large part a predictive tool. So there is a conflict. Why would anyone pretend otherwise?

https://www.google.com/amp/s/amp.livescience.com/63595-schrodinger-uncertainty-relation-temperature.html

Well, apart from any QM theoretical connections which I probably wouldn't understand anyway, since I can't do the math, I can see no reason to believe that super-determinism is the case.

Now I am not a physicist, but I would remind us all of a conceptual issue that might cause misunderstandings here.

We seem to have a situation where one observer sees a certain situation, and another sees a contradictory situation. A sees p, B sees ~p.

It's worth reminding ourselves that this is not new. The same thing can happen in relativistic physics where one observer will see events in a different sequence to another.

But of course what happens in relativistic is that a set of equations are used to translate between the observations. SO although A sees p and B sees ~p, A will also see that B sees ~p, and B will also see that A will see p.

That is, A and B agree that: A sees p, yet that B sees ~p.

Now it seems to me that objective reality has here not so much been undermined as redefined.

A corollary: this fits in with a view of language, logic and mathematics such that we choose a grammar for our descriptions that suits our purposes.

The trick is then in giving an immanent account of 'perspectival' dependence without making humans a necessary constituent of nature. — fdrake

But humans provide perspective, a point of view, which in an ineliminable pole in the knowledge of anything. The philosophical issue revolves around the attempt NOT to face that.

Information and probability are dual notions; wherever you have a probability distribution you have an entropy. The connection between the two is particularly intimate for discrete random variables - like when there is a given probability of being in one of countably many eigenstates of an operator. Quantum entropy measures the degree of mixing in a state; how close it is to behaving in a singular eigenstate (unless I'm misinterpreting, I am both rusty and mostly uneducated here). Information measures are derivable from probability distributions, but the process of mapping a distribution to an entropy value is not invertible - so the two notions can't be taken as inter-definable. As in, if you have an entropy, you have a single number, which could be generated from lots of different quantum states and probability distributions.

I'm sure there are problems, but I think there are good reasons to believe that information is just as much a part of nature as wave functions. — fdrake

Well, yeah this more or less what Rovelli says. In his pre-print 'Relational interpretation of Quantum Mechanics' (see here: https://arxiv.org/pdf/quant-ph/9609002.pdf), he writes (p.3):

Also, I use information theory in its information-theory meaning (Shannon): information is a measure of the number of states in which a system can be – or in which several systems whose states are physically constrained (correlated) can be. Thus, a pen on my table has information because it points in this or that direction. We do not need a human being, a cat, or a computer, to make use of this notion of information.

This certainly avoids the problem of introducing a 'special' role of consciousness. In fact, the 'observer' in RQM is not really different from a 'reference frame' in classical or relativistic physics, i.e. any possible physical system.

But, interestingly, for all practical purposes Rovelli's position is not very different from Bitbol's proposal, in the sense that Bitbol does not think that consciousness has an 'ontological role'. It does not 'change' reality (as IMO actually say the 'consciousness collapse interpretation' held by Von Neumann, Wigner, Wheeler etc). Rather, it just says all our knowledge is 'situated', in the sense that it comes from a certain perspective and we cannot neglect this. So what I observe is the universe 'seen by me' and from that perspective I am indeed sort of special: not in the sense that I am the 'creator' of what I observe but simply because 'I' am the 'point of view'. So, at least for what is observed in the 'perspectives' of human beings the two models are indistinguishable.

Note, that, however Bitbol's view goes a bit ahead than Rovelli's (note that Bitbol is strongly influenced by Kant and phenomenology). According to Bitbol, knowledge starts from conscious experience. This means that consciousness itself defines automatically a perspective. Hence, each of us 'observers' reality from a precise viewpoint. Note that, in this case, the 'perspective' is easily identified. We do not know how 'reality' is 'seen' from the viewpoint of 'a pen on my table' (in fact, we cannot even know what if such a perspective makes sense).
As I said, however, Bitbol does not claim that we 'create' reality. Rather the situation here is much like in Kant. We cannot know how reality is independently from our perspective. We just cannot 'neglect' it completely. Why? Because, conscious experience is the starting point of all inquiry.

To summarize, for Bitbol and Rovelli, QM tells us that knowledge is perspectival. But Bitbol sees a link between this and Kantian philosophy and phenomenology which (in a different context) also say that 'perspective', which, in this case, is conscious experience. So, according to Bitbol this is also true for QM: we cannot know the 'world as it is', but only from our 'situated' experience.

I do not know if Bitbol is right here but IMO he raises very interesting point. Also, Bernard d'Espagnat has a similar view, see: https://arxiv.org/pdf/1101.4545.pdf



Yeah, I agree. Bell's theorem then implies that one between predictions of QM, counterfactual definiteness and locality must be abandoned. Most physicists believe that it is counterfactual definiteness that is to be abandoned. Hence, the 'claim' of this experiment is not really different from the implication of Bell's theorem.

The wiki table on interpretations lists Copenhagen as a non-local interpretation, and I don't understand that. My knowledge of a system doesn't change due to an event that happens elsewhere. But I suppose that my knowledge of a distant system (like the distant half of an entangled pair) changes immediately upon my measurement of its local sibling, so maybe that's why they list it as a non-local interpretation. — noAxioms

Well, I think that probably different 'Copenaghists' would give different responses (after all, there is no agreement among them about the right interpretation of the wave-function). But, I suspect that this problem might be avoided using the same argument that (IMO) is used by RQM, that is, reasoning with 'perspectives'. After I make a measurement, I am sure about the outcome of the other measurement. But until I actually receive the confirmation of it, such an event (the measurement) is outside my perspective.
I do not know however if this argument is really enough to avoid non-locality.
(Note that, more or less, this is the reasoning that is employed to avoid the 'block universe' interpretation of Relativity. In that case, the point is that each 'observer' can define 'its' own plane of simultaneity, i.e. its own present. But if we believe that all these events are 'actually real', then it is not too hard to show that it would imply that we are in 'block world': https://en.wikipedia.org/wiki/Rietdijk–Putnam_argument).

As an aside, a note in that table says that the de Broglie-Bohm interpretation is compatible with relativity. This is IMO wrong. The point is maybe that we cannot observe any violation of relativity via the transmission of faster than light signals. But this does not mean that dBB is actually compatible with it (BTW, non-locality is not the only reason (or even the main) that dBB is not my own favorite interpretation. But I would digress too much here).

"Don't put our understanding in the way of our understanding. Questioning and inquiry are themselves ways to relate to nature."

Maybe that's because questioning and inquiry are ways to relate to themselves. Maybe the in-between IS where nature is.
Scientists who interpret what they do via Kantain thinking talk about a real world that we mirror through our constructions of it, or as you say, 'mediating between us and natural phenomena".
Postmodern thinking would say instead that the purpose and effect of knowledge is not to represent what is, but to adaptively interact with and thus transform what it theorizes about.
What science does, then , is pragmatically reorganize a world in ways that we can make more or less useful to us. What we get right or wrong, true or false , only makes sense in relation to our changing theoretical norms. As Heidegger says "A science's level of development is determined by the
extent to which it is capable of a crisis in its basic concepts."

Certainly scientists are under no obligation, nor would it be at all helpful for most, to inquire into their presuppositions when they do science. The nature of the way empirical questions are formulated isn't designed for such understanding, which is then left unexamined and implicit. But then one could argue that a field like physics is applied philosophy. Each new scientific advance implies an unexamined shift in philosophical underpinnings. Thats why a history of physics, from Aristotle and the Scholastics through Galileo,and Newton to Einstein parallels the history of philosophy.

As I said, however, Bitbol does not claim that we 'create' reality. Rather the situation here is much like in Kant. We cannot know how reality is independently from our perspective. We just cannot 'neglect' it completely. Why? Because, conscious experience is the starting point of all inquiry. — boundless

(will also serve as a reply to our Heideggerian friend @Joshs)

This more general theme is precisely what I was reacting to in the thread. There's a weak, and I think very agreeable account and a strong, and I think very important to undermine account.

The weak claim goes something like this; all knowledge derives from interpretation, if we know anything about anything, it has to be an interpretation of that thing. Grant for the sake of argument that experiences are also interpretations of stuff; if you pat a dog, the topography and consistency of their hair leaves its impression through your sense of touch. Interpretations are always simplifications and approximations, with a required background to understand them and a necessary set of presuppositions on which they develop. Another way of stating this is that the way we theorise is through the creation of concepts; and we use concepts to make sense of things. Moreover, experience informs the tailoring of concepts.

The strong claim goes like this: it starts with the weak claim; all knowledge derives from interpretation - I agree with all you said, but that is not the end of the story. Consider that every relation that we have with any other thing is an interpretation; interpretations are the trace of our experience and history in our living bodies. When asked the question; what does your knowledge and experience consist of? You may say it consists of interpretations. What of the things we experience in the world? Those too are interpretations with a certain thingliness associated with them. Whenever we interact with anything, all that is generated is an interpretation, and all thought consists in a chain of such interpretations evaluated with a logic of links you might call a theory. Whenever we encounter an object, that object is an interpretation from a certain perspective; it is a contextually circumscribed concept. As Josh put it above:

What we get right or wrong, true or false , only makes sense in relation to our changing theoretical norms — Joshs

I believe in the weak claim, I believe in the emphasis portrayed in the strong claim on the dogged pursuit of where our concepts come from. What I don't believe is the characterisation of interpretation, knowledge, or interaction induced through:

When asked the question; what does your knowledge and experience consist of? You may say it consists of interpretations. What of the things we experience in the world? Those too are interpretations with a certain thingliness associated with them. Whenever we interact with anything, all that is generated is an interpretation, and all thought consists in a chain of such interpretations evaluated with a logic of links you might call a theory

Specifically, I disagree with:

(1) (I don't like that) The substitution of interpretations for the targets of interpretations; (rather I believe) knowledge must always be knowledge of its target.
(2) The claim that the target of an interpretation, or knowledge or theory, is consigned to a realm beyond our experience because all experience is interpretation.

Instead I want to emphasise that interpretation is a relationship between its target and a perspective; reasoning consists in the development of this relationship, but its standards of relevance are dictated by the demands the target places on our inquiry; as Lakatos puts it nature can always 'Shout NO!'. We understand nature through theories and interpretations, we don't just have access to theories and interpretations, theories, interpretations and experiences are how we access their targets.

The emphasis I put on quantum phenomena occurring long before our accounts of quantum phenomena were written, and long before the first vestige of quantum mechanics was developing in the canon theoretical physics, is to ape a famous argument that attempts to undermine (2). The argument is usually called the 'arche-fossil' and it was advanced by Meillassoux is the first few chapters of his 'After Finitude, an Essay on the Necessity of Contingency'. The strong account, as I termed it, Meillassoux calls 'correlationism'; stated in one line, correlationism is the belief that we can never have knowledge about thought or being, only of the relationship between them.

Kant fits in here as a correlationist, but he still has the noumenon. The nounmenon, even if we can never have determinate knowledge about it, can be the subject of imagination. Whether you believe that the role of the noumenon in Kant is negative; in that it simply marks the boundary of our possible knowledge; or positive; in that it refers to an uninterpreted reality which subsists beneath our interaction with it, there is still something other than interpretation which we are dimly aware of. Whether this noumenon can 'Shout no!' to our theories or whether it casts no shadow upon our interpretations and experiences is the substance of my disagreement with the 'strong account'. I believe that nature can 'Shout no!' (as Lakatos puts it) when we ask it well formed questions, and I believe that the capacity for it to shout no is a necessary component of a good account of knowledge, interpretation, and the ontology of our being in the world.

The strong account, or strong correlationism for Meillassoux, does away with the relevance of the noumenon; what is the point in it if we can never have experiential or interpretive access to it? Our interpretations persist in the relation of thought and being and never take inspiration from that which is outwith the relation between the two. That is to say, never from nature; even if it 'Shouts no!', we're necessarily deaf to its cry, for its cry would be another interpretation.

Meillassoux summarises this position like:

We said above that, since Kant, objectivity is no longer defined with reference to the object in itself (in terms of the statement’s adequation or resemblance to what it designates), but rather with reference to the possible universality of an objective statement. “It is the intersubjectivity of the ancestral statement – the fact that it should by right be verifiable by any member of the scientific community – that guarantees its objectivity, and hence its ‘truth’. It cannot be anything else, since its referent, taken literally, is unthinkable — Meillassoux, After Finitude Chapter 1, Ancestrality

Meillassoux considers a class of statement called an 'ancestral statement' whose meaningfulness undermines the strong correlationist account. For an example, consider 'the age of the universe is approximately 13.8 billion years'. The precise logic Meillassoux ascribes to the interpretation of this kind of statement for those who believe in the strong account/correlationism is as follows:

Consider the following ancestral statement: ‘Event Y occurred x number of years before the emergence of humans.’ The correlationist philosopher will in no way intervene in the content of this statement: she will not contest the claim that it is in fact event Y that occurred, nor will she contest the dating of this event. No – she will simply add – perhaps only to himself, but add it he will – something like a simple codicil, always the same one, which he will discretely append to the end of the phrase: event Y occurred x number of years before the emergence of humans – for humans (or even, for the human scientist). This codicil is the codicil of modernity: the codicil through which the modern philosopher refrains (or at least thinks she does) from intervening in the content of science, while preserving a regime of meaning external to and more originary than that of science. Accordingly, when confronted with an ancestral statement, correlationism postulates that there are at least two levels of meaning in such a statement: the immediate, or realist meaning; and the more originary correlationist meaning, activated by the codicil.

The literal truth of the statement, not just 'for us' for Meillassoux entails numerous unpleasant things for a correlationist to square themselves with.

We would then be obliged to maintain what can only appear to the post-critical philosopher as a tissue of absurdities; to wit (and the list is not exhaustive):

• that being is not co-extensive with manifestation, since events have occurred in the past which were not manifest to anyone;
• that what is preceded in time the manifestation of what is;
• that manifestation itself emerged in time and space, and that consequently manifestation is not the givenness of a world, but rather an intra-worldly occurrence;
• that this event can, moreover, be dated;
• that thought is in a position to think manifestation’s emergence in being, as well as a being or a time anterior to manifestation;
-that the fossil-matter (the state of affairs pictured in the statement 'the universe is 13.8 billion years old') is the givenness in the present of a being that is anterior to givenness; that is to say, that an arche-fossil manifests an entity’s anteriority vis-à-vis manifestation.

Needless to say, there are problems here. Being the incorrigible simpleton that I am, when I believe the statement 'The universe is 13.8 billion years old', I'm actually believing in its literal truth, and not the correlationist transformation of the statement; 'The universe is 13.8 billion years old for us/our current theories/whatever'. Moreover, though I have given no argument of this, I believe that the ability to accept the literal truth of that statement and moreover that it has a meaning at all are indicative that thought really can 'aim for that which is outside of it', and that our inquiries are meaningful only when they can admit the possibility of nature 'Shouting no!'.

How does this relate to the discussion in this thread? Well, firstly, it relates to the equation of a quantum observer with a human. Secondly, it relates to the equation of an observer-quantum state system as a human-quantum state system, and lastly it relates to the equation of a 'perspective' in an observer-quantum state system with anything like the usual meaning of worldview or theoretical background we would give to it. The debate about these terms and their application in the philosophy here is a lot broader than quantum mechanics; though I do very much appreciate the explicit reference to Kant, @boundless.

There is a wave function of places where I likely left my car keys, with some more probable than others. When I find them (or even when I look certain places and don't yet find them), that wave function changes since my knowledge of the system has been changed. — noAxioms

I’m pretty sure that’s not true. It is simplistic. The point in respect of sub-atomic particles, is that they’re not in some undetermined location prior to measurement -they’re literally not in any definite location until the measurement is taken. Otherwise, there would be nothing to discuss.

* that being is not co-extensive with manifestation, since events have occurred in the past which were not manifest to anyone — fdrake

I don't know if you took a look at the essay about Wheeler's 'Delayed Choice' I pointed to, but it explicitly address this issue. It describes a version of the delayed-choice thought-experiment where the light source is a distant quasar, and the place of the slits in the screen is taken by distant galaxies, which bend the light as it travels.

The quasar could be very distant from Earth, with light so faint that its photons hit the piece of film only one at a time. But the results of the experiment wouldn't change. The striped pattern would still show up, meaning that a lone photon not observed by the telescope traveled both paths toward Earth, even if those paths were separated by many light-years. And that's not all.

By the time the astronomers decide which measurement to make — whether to pin down the photon to one definite route or to have it follow both paths simultaneously — the photon could have already journeyed for billions of years, long before life appeared on Earth. The measurements made now, says Wheeler, determine the photon's past. In one case the astronomers create a past in which a photon took both possible routes from the quasar to Earth.

The 'delayed choice' and it's associated 'weirdness' has been the subject of extensive reporting in the popular science media in the last decade.

'The universe is 13.8 billion years old', I'm actually believing in its literal truth, and not the correlationist transformation of the statement; 'The universe is 13.8 billion years old for us/our current theories/whatever'. Moreover, though I have given no argument of this, I believe that the ability to accept the literal truth of that statement and moreover that it has a meaning at all are indicative that thought really can 'aim for that which is outside of it', and that our inquiries are meaningful only when they can admit the possibility of nature 'Shouting no!'. — fdrake

It's not so much 'correlationism' as the incorrigible realism that each of us is born with. What you're not seeing is the role the mind/brain plays in the statement about the age of the Universe (or anything else, for that matter.) But what this type of realism assumes is just what Kant means by 'transcendental realism':

I understand by the transcendental idealism of all appearances the doctrine that they are all together to be regarded as mere representations and not things in themselves, and accordingly that space and time are only sensible forms of our intuition, but not determinations given for themselves or conditions of objects as things in themselves. To this idealism is opposed transcendental realism, which regards space and time as something given in themselves (independent of our sensiblity). The transcendental realist therefore represents outer appearances (if their reality is conceded) as things in themselves, which would exist independently of us and our sensibility and thus would also be outside us according to pure concepts of the understanding. (CPR, A369)

The transcendental idealist, however, can be an empirical realist, hence, as he is called, a dualist, i.e., he can concede the existence of matter without going beyond mere self-consciousness and assuming something more than the certainty of representations in me, hence the cogito ergo sum. For because he allows this matter and even its inner possibility to be valid only for appearance– which, separated from our sensibility, is nothing – matter for him is only a species of representations (intuition), which are called 'external', not as if they related to objects that are external in themselves but because they relate perceptions to space, where all things are external to one another, but that space itself is in us. (A370)

The problem is, empiricism has bet the house on the fact that what it conceives of as nature, the real world, and so on, is what is finally and fundamentally real. But QM continues to throw this into doubt. Hence the controversy!

Two people facing each other over a set table:

"You've got the knives on the left!"

"No, you have!"

It's not so much 'correlationism' as the incorrigible realism that each of us is born with. What you're not seeing is the role the mind/brain plays in the statement about the age of the Universe (or anything else, for that matter.) But what this assumes is just what Kant means by 'transcendental realism': — Wayfarer

I do agree that the claim 'the universe is 13.8 billion years old' is something produced by our theories and understandings. But I also believe that it is literally true. In terms of your quote:

For because he allows this matter and even its inner possibility to be valid only for appearance– which, separated from our sensibility, is nothing – matter for him is only a species of representations (intuition), which are called 'external', not as if they related to objects that are external in themselves but because they relate perceptions to space, where all things are external to one another, but that space itself is in us. (A370)

I believe it's important to emphasise that nature also informs our sensibility and strikes accord with it. The impulse Husserl expressed that phenomenology (a discipline heavily indebted to Kant) is 'relearning how to see' is basically correct; with the addendum that reality can teach us how to see better when we give ourselves the goal of understanding it.

You seem to want to paint me as a pre-critical realist; I'm not, I know that experience is 'theory ladened' and informed by what we know, what we've experienced, our habits and the structure of our bodies; but I will insist that it is also informed by the objects we experience.

The problem is, empiricism has bet the house on the fact that what it conceives of as nature, the real world, and so on, is what is finally and fundamentally real. But QM continues to throw this into doubt. Hence the controversy! — Wayfarer

The same thing happened with special relativity; and people make the same mistake that the relativity of simultaneity is property of a human perspective; more properly it's a property of motion that can't adequately be accounted for while assuming the speed of light is infinite, the relativity of simultaneity is still something natural, it just takes really fast relative motions to have meaningful effects.

With that in mind, that different observers of quantum states can disagree on precisely what state the system is in is also not a property of a uniquely human perspective; other systems can serve as observers, and systems can serve as observers to coupled systems of observers and other systems. These are facts of nature that need no appeal to human cognition to explain; only an appeal to the randomness of the quantum state. Of course we need to cognise it to understand it, it's a theory; but it still describes some natural phenomenon.

You've misread Kant here.

He's giving an empirically realist account in this context, in which phenomena are as they appear (or would appear if we were present), rather than being something which can be detached from how they would appear in our experiences. He's not trying to suggest things don't exist without us or are beyond our explanation-- indeed, his point is the exact opposite.

Kant doesn't fit the scientism bill, but this point is compatible with accounts scientism, which reduces the world to a set of things known to us. After all, that's what scientism thinks: that the while world is constituted by things which appear in our experience (as se out in our given theory).

It still describes some natural phenomenon. — fdrake

Wouldn't dispute it for a moment. Kant himself devised the nebular hypothesis, he was a scientist as well as a philosopher. But you go further when you speak of systems as if they're observers. They're not. Observers are beings, and the word 'being' is significant in this context.

But you go further when you speak of systems as if they're observers. — Wayfarer

Rovelli makes the same point in the paper boundless linked, as does Bohr in the quote I gave earlier. It's convenient to talk of these things in terms of human perspectives and related terms, because that's how the common relevant vocabulary works, but they go to pains to distinguish observer from human. Whether a system is a being or not is probably outside the scope of the thread, I would tentatively say that it generally is, but it's more adequately described as an interacting connection of flows; what matters more is its image in the category of becoming rather than of being.

Maybe the same could be said of a wave function; the only observables are the states (and derived quantities) and their frequency of occurrence through experiment, but I'd be quite happy to think of a wave function as a being or a property of a being.

Don't think it's relevant really, things induce quantum effects in other things; hence ionic bonding and the arche-fossil reference for when this relationship is anthropomorphised too much.

how can you "measure" a superposition or how can a measurement be in a superposition? I thought any observation causes the wave function to collapse in a single eigenstate and a measurement, I would think, involves an observation. — Benkei

That is my understanding. And the abstract of the paper doesn't say a superposition is detected. I suspect there is some over-interpretation of the experiment's implications going on here.

Anyway, point is, Alice learning of the measurement results splits Alice, but does not split the universe, as is commonly assumed. Bob, being able to speak to both versions of Alice, is still in a common world. So yes to the three worlds if you count them that way: One for each Alice, and one for Bob. — noAxioms

At the risk of being annoyingly meta, I think there are multiple interpretations possible of the many-worlds interpretation of QM.

One can take a 'splitting' interpretation, in which worlds branch off when a measurement is made, or one can take the interpretation that all the different worlds exist in superposition from the outset, each one having a definite state of everything, and what happens when we make a measurement is that we narrow our knowledge of which possible worlds we could be in.

Under the non-splitting interpretation, if Alice measures spin as V, then Alice knows she is in a V-world, whereas Bob doesn't know whether he is in a V-world or an H-world, but knows that Alice does know. There are two Alices and two corresponding Bobs, but the knowledge of the two Alices is different whereas that of the two Bobs is the same.

Those two version of Alice, being in different worlds, cannot communicate or otherwise be aware of each other. But they behave exactly identically because they're keeping that knowledge a secret. — noAxioms

But note that they're not actually identical since they each have a different memory of what they measured. What Bob can do is reverse Alice's polarity measurement while retaining the record that the measurement occurred, which is identical for both Alices. This means that the two Alices will merge without memory of the polarity result and with all records of the polarity result having been erased. That is, there will be only one world branch again, with multiple histories, and with the record that a definite polarity result was measured by Alice.

This is analogous to the double-slit experiment where the single particle detected on the back screen had two distinct path histories (one for each each slit).

Anyway, point is, Alice learning of the measurement results splits Alice, but does not split the universe, as is commonly assumed. — noAxioms

Yes.

So yes to the three worlds if you count them that way: One for each Alice, and one for Bob. But there is obviously communication between the Bob world and both Alice worlds, but Bob cannot pass a message from one Alice to the other. With the communication, Bob's world is clearly not isolated from Alice's world, and hence doesn't really count as a separate world. — noAxioms

Agreed. Alice's worlds are separate from each other, but they are in superposition in Bob's world.