No, if you change reference frames (in the special relativistic sense), the direction of causality might be opposite, i.e. collapse at B might suddenly cause the collapse at A.

Well sure, but that’s still causality.

We are mixing up correlation and causation in this thread. In quantum entanglement you have a correlation, not causation.

For example, if you go on a trip and bring your suitcase with you. You discover on opening it that you packed only one blue sock. You then gain the information that your other blue sock must be at home. You didn't "cause" your sock to appear at home at that moment, you simply gained that information from a correlation that you know must be true. The cause occurred when you were packing. In this case the correlation is entirely local though: you always had that information with you in your suitcase, you just didn't look at it until now.

In quantum entanglement the cause and effect were the entanglement itself. When one of the particles is later observed, you're using correlation to decide what the other end is. The difference is that in quantum theory this correlation can be demonstrated to be non-local -- the information wasn't with you the whole time and was only determined at the moment of observation. But you can't take any action based on it and the other side can't either as you have no way to know who caused the collapse. You can't cause any meaningful effect on the other end with the information you get from your observation or the act of your observation.

If we are using "cause" in the colloquial sense then yes I see what you're saying, it "causes" it. But it's important to be specific with domain terms here because entanglement does not violate causality, which has a specific definition.

Yeah this points to why I think information is actually the salient feature.

Thanks for taking the time to comment. How could I go about learning more about the specific definition of causality you’re referring to here?

The wikipedia page on causality is a good place to start for a clear overview, specifically the section on causality as a physical concept: https://en.m.wikipedia.org/wiki/Causality_(physics)

As the wikipedia article touches on, there's not an agreed global definition and it can be used differently in multiple contexts (hence why it's important to be specific).

This stack exchange answer gives a good list of terminology: https://physics.stackexchange.com/a/34675

And this article puts a far more technical description on terms: https://www.mprl-series.mpg.de/media/proceedings/3/9/Proc3ch...

Yeah, but is it left sock or right sock? Imagine, that you opened the suitcase, get the blue sock, put it on the left foot, then you discover that the RIGHT sock is missing, thus the RIGHT sock is at home. This information wasn't with you the whole time and was only determined at the moment of observation.

Sure, but nothing about the state of the other sock changed, you just gave it a name. Similarly, every time we name a far away galaxy we don't break causality.

Don't you think a good definition of causality (i.e. whether event A causally influenced event B) should be observer-independent?

Anyway, chrisfosterelli already elaborated on why thinking of the collapse at A "causing" the collapse at B is not a good idea from the perspective of quantum mechanics, either.

Location B will measure both up and down, but will only be able to communicate with the version of location A that measured the opposite value.

There’s no need for FTL if you accept MWI.

My understanding is that when particles are entangled with each other, they must be described together with one wave function rather than one wave function for each of them. So once you measure particle A, the wave function collapses both for A and B.

Right. That… seems like causality to me.