# How do physicists talk about spin of individual particles when the universe is massively entangled?

The following two things seem to be true:

(1) The universe is massively entangled because the wave function that describes the entire universe has measure 1 of being entangled. Further, given how entanglement spreads, it seems likely that the universe is massively entangled.

(2) On a canonical presentation, entanglement of properties implies that one cannot specify the property of one particle without referring to the properties of the others entangled within the system. the example is the EPR state. One can't talk about the spin of either particle without referring to the spin of the other.

But it seems like we are able to talk about the spin of particles independently in the lab. After a measurement we may say that one particle is x-spin up without reference to any other particles. But if that's true, how is the truth of (1) and (2) maintained? How do we manage to talk about the state of a particle as if it wasn't entangled with the rest of the universe? How is that even possible given that entanglement implies that we must refer to the rest of the universe in our discussion of that property? Thanks for any answers! I'm sure I'm just really confused about something...

• entanglement requires some interaction. Isolate the experiment well enough, and you can keep a region clean enough to stop that, at least for the tiny particles you are studying Commented Apr 16 at 0:42
• @WillO Of course it doesn't! But it seems like for the eclipses we use approximations that are good enough. Spin doesn't seem to be approximated though?
– zzz
Commented Apr 16 at 2:07
• This Veritasium video may help - Parallel Worlds Probably Exist. Here’s Why Commented Apr 16 at 3:33
• @mmesser314 Hey! Yes, the many worlds interpretation is an easy way out of this. But I was wondering about how collapse theories like GRW or even hidden variable theories like Bohmian Mechanics may get away with this… Thanks for the answer anyways.
– zzz
Commented Apr 16 at 3:35
• Some notions that appear meaningful but have no meaning: "the average velocity of the universe; the mass of the universe; the wavefunction of the universe". 1st falls to SR, 2nd falls to GR, 3rd is, similarly, a misunderstanding of QM. Commented Apr 24 at 13:36

(1) The universe is massively entangled because the wave function that describes the entire universe has measure 1 of being entangled. Further, given how entanglement spreads, it seems likely that the universe is massively entangled.

Massively entangled - yes, completely entangled - no. Photons are being created all the time and the ones that are in transit that have not directly interacted with anything yet are not yet entangled with anything.

(2) On a canonical presentation, entanglement of properties implies that one cannot specify the property of one particle without referring to the properties of the others entangled within the system. the example is the EPR state. One can't talk about the spin of either particle without referring to the spin of the other.

The measurement of a pair of entangled particles is a (futile) attempt to full identify its spin state by e.g. measuring both its vertical component and its horizontal component simultaneously. When we measure the spin state of a single particle by passing it though a Stern Gerlach device we could detect whether it is spin up or spin down (let's say it was up), but we have no idea of its horizontal component. We can try to determine its horizontal component by passing it through a second SG device that is orthogonal to the first and for example determine its horizontal component is left spin. Now we think we have a good idea that the particle has both up spin and and left spin, but we are wrong. This can be demonstrated by passing it through yet another vertical SG spin analyser and 50% of the time it will tell us the the particle is spin down! This means measuring the horizontal component altered the vertical component so each time it passes through an analyser its orientation is changed. This also infers that when we passed the particle through the first vertical analyser we altered its left/right spin component so we all we know about the original spin state of the particle is that it had a vertical component and an unknown horizontal component.

We can also demonstrate that measurement alters the state of the particle when analysing polarisation of photons in a simple experiment that you can even do at home. Get some polarising filters and place them at 90 degrees to each other. Almost all the light is blocked. This is because if the first filter lets photons with a greater vertical than horizontal component pass through, then they do not pass through the second filter. Now if we insert a third filter between the first two, that is orientated at 45 degrees to both of them, significantly more light passes through. This is because some of the light that had a greater vertical component passing through the first filter now has a greater horizontal component after passing through the middle filter. The filters measuring the polarisation, also alter the polarisation.

This means the statement "One can't talk about the spin of either particle without referring to the spin of the other." is more accurately stated as "One can't talk about the (exact state of) spin of either particle without (even when) referring to the spin of the other."

How do we manage to talk about the state of a particle as if it wasn't entangled with the rest of the universe?

When a particle like a photon is emitted from an atom, it born in a coherent and not entangled with the universe. When it interacts with other particles and imparts energy to them it decoheres and its entanglement spreads out to the particles it interacts with. What constitutes an interaction? Normally bouncing off a mirror or passing through a polariser in e.g. an interferometer does not count as an (measurement) interaction. If we were to mount the mirrors in such a way that we could could measure the recoil of a photon bouncing off it in order to determine "which way" path information, that would constitute an measurement interaction and it would decohere and spoil the interference pattern. Same goes if we try to mount to mount the source in such a way as to measure the recoil to try to determine the time of emission. In fact, until the photon interacts with other particles in such a way that its location can be determined, there is no way to determine the photon even exists. Once detected we can infer approximately when it was emitted and even which path it took, but the more information we have about it, the more it decoheres and more entangled it gets with the rest of the universe. Of course we never gain complete information about the particle such as its exact polarisation or spin as this is forbidden by Quantum Mechanics and in particular, the Heisenberg uncertainty principle.

• Hi! Thanks for your answer! I believe the universe is completely entangled. I read this in Penrose 2004 Road to Reality: a Complete Guide to the Laws of Nature. He writes “every universe must eventually become entangled with one another” p.591
– zzz
Commented Apr 16 at 11:56
• Wikipedia en.wikipedia.org/wiki/Penrose_interpretation says the Penrose interpretation "is a speculation" which not the same thing as an accepted theory. His interpretation requires that decoherence requires strong gravitational fields, so at least his theory is falsifiable when we eventually carry out experiments in space.
– KDP
Commented Apr 17 at 0:00
• “every universe must eventually become entangled with one another” is not the same thing as every particle in this universe is entangled with every other particle.
– KDP
Commented Apr 17 at 0:01
• oops, i miswrote. He wrote particles. Further this is not part of Penrose's interpretation. It's a mathematical fact that there is measure 1 on the wave function being entangled.
– zzz
Commented Apr 17 at 3:44
• @zzz Mathematical facts are not physics. Mathematical facts exist only in human imagination. They sometimes model reality effectively, but that can only be verified by experiment. What experiment do you propose to test your model? Commented Apr 24 at 12:28

I am going to describe quantum theory without any modifications of quantum equations of motion such as collapse. Quantum theory describes measurable quantities in terms of mathematical doodads called observables represented by Hermitian operators. To get predictions about an observable you use a Hermitian operator with unit trace called the state $$\rho$$. Quantum theory allows us to calculate the expectation value of an observable $$\hat{A}$$ is $$tr(\rho\hat{A})$$. In the Heisenberg picture the observables of system 1 evolve over time and become dependent on those of another system 2 only when system 1 and system 2 interact. You could describe the x spin of a system in terms of its spin x observables regardless of whether the system is entangled:

https://arxiv.org/abs/quant-ph/9906007

The relative state of a system after a measurement can be described in terms of the previous state, the measurement result and the observables of that system:

https://arxiv.org/abs/2008.02328

So the state of a system can be specified without referring to other systems and entanglement does not change this.

• Thank you! This is a fantastic answer. I can see how not assuming collapse resolves the issue I've raised. I was wondering, however, in dynamical collapse theories like GRW, how this is maintained.
– zzz
Commented Apr 16 at 12:59
• @zzz GRW tries to deal with this problem of widespread entanglement by eliminating most versions of a system and since entanglement is about different versions of one system being correlated with those of another they are no longer entangled. For a review of GRW see arxiv.org/abs/2310.14969 GRW and other single universe theories have problems with reproducing the result of quantum field theory arxiv.org/abs/2205.00568 Commented Apr 16 at 20:04
• But normal QFT is not any better. It's plagued with renormailization problems. It's unclear why in any principled sense there can't be a QFT for single universe theories!
– zzz
Commented Apr 17 at 3:45
• @zzz I don't think renormalization is particularly problematic arxiv.org/abs/quant-ph/0112148 all it really implies is that existing theories don't describe the world down to arbitrarily small scales which won't be true anyway cuz spacetime will be quantised in quantum gravity. If you're interested in learning about renormalization I suggest reading "Renormalization Methods: A Guide for Beginners" by McComb, "Quantum field theory for the gifted amateur" by Lancaster and Blundell or "The conceptual framework of quantum field theory" by Duncan. Commented Apr 17 at 7:21

I talked to a physicist at Columbia and it seems like the answer is this. This all depends on your preferred interpretation of quantum mechanics. In Bohmian Mechanics, we talk about particles having particular spin because of there position in the wave function on that region in configuration space. So, the wave function will still be massively entangled but particles will have definite spin depending on where they are in configuration space. In MWI, worlds cease to interfere with one another so there's a definite spin state of the particle. In GRW, and other collapse theories, states will never be truly localized... So, what's going to happen is that the particle will remain entangled with the rest of the universe. This is called the tail's problem. However, even though states are never fully localized, it seems like there are plausible ways of resolving it.

Important note: the state of the world in any of those theories after we make a "measurement" will be an eigenstate of some particle at some position having some spin– we cannot distinguish between particles!