In a highly watchable talk in the Oxford University lecture mini-series on Cosmology and quantum Foundations Carlo Rovelli gives a lot of persuasive arguments about why the Many Worlds Interpretation is suspect. But he goes fast and furious sometimes. Sometimes constructing strawman arguments (I do not think anyone seriously thinks just literally interpreting mathematics in a given model of physics leads to necessarily great ontological truths, apart from the likes of characters like Tegmark perhaps) but I think generally even these points are well made and interesting to ponder. Rovelli describes his own current opinion as “Everettian” — which means not a traditional Many Worlds interpretation but rather Relative State interpretation.
There are many key slides in his presentation that I thought worthy of mentioning and which inspired this current post of mine.
In another slide Rovelli puts up a couple of threads, one is,
- “Why don’t we see superpositions?” — what a silly question! Because in textbook QM we do not see the state, we see eigenvalues. We see where is the position o the electron or it’s momentum, never it’s wavefunction.
- These (facts) are described by the position in phase space in classical physics; and by points in the spectra of elements of the observable algebra in quantum physics.
Which is cool, but then he riles the zen masters by writing:
- They can be taken as primary elements, and the quantum formalism built up from them.
First, I should point out this is not erroneous. You can build up a theory from elements that are such primitives as “points in the spectra of elements of the observable algebra”.
But I think this is misleading for purists and philosophers of physics. Just because one approach to calculating expectation values works does not make it’s mathematical elements isomorphic in some sense to elements of physical reality. So I think Rovelli un-does some of his good arguments with such statements. (I’m not the expert Rovelli is, I’m just sayin’ ya know …)
You might counter: “Well, if you are not willing to take your theoretical elements of reality direct from the best mathematical model’s primitives, then where are you going to define your ontology (granting you are wishing to construct a realist interpretation)?”
I would concede, “ok, for now, you can have a favoured realist interpretation based on the primitives of your observables algebra.” But I think you are always going to have to admit this will be temporary, only an “effective interpretation” that is current to our present understandings.
My point is that while this makes for great contemporary physics it does not make for good philosophy (love of both knowledge and truth). The reason is blatant. If all you have is a model for computing amplitudes then there is really only a small probability for hoping this is a dead accurate and “True” picture of the real ontology in our universal physics. You can certainly freely pin your hopes on this chance and see where it leads.
I, for one, think that such an abstraction as an “observable algebra” although nice and concrete and clean, is just too abstract to be wisely taken literally as the basis for a realist interpretation. Again, I’m “just sayin’…”.
There are many more good discussion points in Rovelli’s lecture.
The Wavefunction is a Computational Tool
This meme has always gelled with me. You can map a wavefunction over time, for example, you can visualize an atomic electron’s orbital. But at no single moment in time is the electron ever seen to be smeared out over it’s orbital. To me, as a realist, this means the electron is probably not a wave. But it’s temporal behaviour manifests aspects of wave-like properties. Or to be bold: over time the (non-relativistic) constant energy electron’s state is completely coded as a wave. I will admit in future we might find hard evidence that electron’s truly are waves of some weird spacetime foamy medium, not waves in an abstract mathematical space, but I do not think we are there yet, and I think we will not find this to be so. My guess would be electrons are extended topological geons, perhaps a little more gnarly than superstrings, but less “super”. I think more like solitons of spacetime than embedded strings.
The keyword there for philosophy is “coded”. The wave picture, or if you prefer, the Heisenberg state matrix representation, (either the Schrödinger or Heisenberg mathematical tool will do) is a code for the time evolution of the electron. But in no realist sense can it be identified as the electron. Moreover, if you are willing to accept the Schrödinger and Heisenberg pictures are equivalent then you have a doubled-up ontology. To me that’s nonsense if you are also a realist.
Believe it or not though, I’ve read books where this is flatly denied and authors have claimed the electron is the wavefunction. I really cannot subscribe to this. It violates the principle of separation of ontology from theory (let me coin that principle if no one has before!). A model is not the thing being modelled, is another way to put it.
On a related aside note: John Wheeler was being very cheeky or highly provocative in suggesting the “It from Bit” meme. It sounds like a great explanatory concept, but it seems (to me) to lack some unknown extra structure needed to motivate sound belief. Wheeler also talked about “equations written on paper cannot bring themselves into existence” (or something to that effect). But I think “It from Bit” is not very far removed from equations writing themselves into a universe.
EPR is Entanglement with the Future?
That’s not quite an accurate way to encapsulate Rovelli’s take on EPR, but I think it captures the flavour. Rovelli is saying that in a Relational QM interpretation you do not worry about non-locality, because from each observers (the proverbial Alice and Bob at each end of an EPR experiment, or non-human apparatus if you prefer to drop the anthropomorphisms) point of view there is a simple measurement, nothing more. The realisation entanglement was happening only occurs later in the future when the two observers get back together and compare data.
I’m not quite with Rovelli fully on this. And I guess this makes me a non-Everettian. There might be something I’m missing about all this, but I think there is something to explain about the two observers from a “Gods eye” view of the universe at the time each makes their measurements. (Whether God exists is irrelevant, this is pure gedankenexperiment.) If you are God then you witness effects of entanglement in the measurement outcomes of Alice and Bob.
The recent research surrounding the ER=EPR meme seem to give a fairly sound geometric or geometrodynamic interpretation of EPR as a wormhole connection. So I think Rovelli does not need to invoke anything fancy to explain away EPR entanglment. ER=EPR has, I believe, put the matter of the realist interpretation mechanism of entanglement to rest.
No matter how many professors shout out, “do not attempt to make mental mechanical models of QM, they will fail!” I think ER=EPR defies them at least on it’s own ground. (Ironically, Susskind says just such things in his popular Theoretical Minimum lectures, and yet he was one of the original ER=EPR co-authors!)
What About Superposition: Is Superposition=ER?
I am now going beyond what Rovelli was entertaining.
If you can explain entanglement using wormholes, how about superposition?
I have not read any good papers about this yet. But I predict someone will put something on the arxiv soon (probably have already since I just haven’t gotten around to searching.) In a hand-waving manner, superpositions are bit like self-entanglement. A slightly harder interpretation might be that at the ends of a wormhole you could get particle duplication or mirror-effects of a sort.
One might even get quite literal and play with the idea that when an electron slips down a minimal wormhole it’s properties get mirrored at each end. Although, “mirror” is not the correct symmetry. I think perhaps just “copied at each end” is better. Cloned at each end? Whatever.
Maybe the electron continually oscillates back and forth between the mouths in some way? Who knows. It does require some kind of traversable ER bridge, or maybe just that when the bridge evaporates in a finite time the electron’s information snaps but to one end, but not both ends. Susskind and Hawking both concur now that there is no black hole information loss right? So surely a little ol’ electron’s information is not going to get lost if it wanders into a minimal ER bridge.
Then measurement or “wave function collapse” is likely a process of collapse of the wormhole. But in snapping the ER bridge the particle property can (somehow) only get restored at one end. Voila! You solve Schrodinger’s Cat’s dilemma.
Oh man! Would I not love t0 write a detailed technical mathematical exposition of all this. Sigh! Someone will probably beat me to it. Meehhh … what do I care, I’m not doing physics for fame or fortune.
Someone will have to eventually worry about stability of minimal ER bridges and the like. Then there are Lorentzian wormholes and closed time-like curves to consider. That Bao, Pollack, Remmen (2015) paper I cited above talks about “no-go” theorems arising from admitting ER bridges, no-go for causality violation and no-go for topology change. I think what theoretical physics needs is an injecting of going past such no-go theorems. They have to be “goes”. Especially topology change. If topology change implies violation of causality then all the better. It only needs to have direct consequences at the Planck scale, then it’s not so scary to admit into theory, whatever the mess it might cause for modelling. The upshot is that at the macroscopic scale I think allowing the “go” for these theorems rather than the “no-go” will reveal a lot of explanatory power, maybe even most of the explanation for the core phenomenon of quantum mechanics. They mention concerns about violation of causality All of which I think is brilliant. I can see this sort of deep space structure explaining a lot of the current mystery about quantum mechanics, and in a realist interpretation. Awesome! And that I am not “just sayin” — it truly would be justifiably awesome.
* * *
Hmmm … had a lot more to say about Rovelli’s talk. Maybe another day.
* * *