[tsimionescu]

> All interpretations need to induce a measure over observations (not "worlds") to produce meaningful results. Without that all you have is an abstract mathematical object.

Agreed, but the MWI in particular does so by applying frequentist probabilities over all versions of an observer, the so-called worlds. The argument goes that there is an apparent non-deterministic process from the point of view of every individual observer, and that we can compute its probability based on how many observers would see a particular state versus the total number of observers.

For some reason, many MWI adherents want to claim that this is not an additional postulate, that MWI only needs the Shrodinger equation, but it clearly is a postulate in addition to that equation, just as much as the collapse idea in other interpretations.

You are right about the CI though, it's not really a useful term.

[consilient]

> For some reason, many MWI adherents want to claim that this is not an additional postulate, that MWI only needs the Shrodinger equation, but it clearly is a postulate in addition to that equation, just as much as the collapse idea in other interpretations.

I think you're misinterpreting them (us). The claim is that MWI requires one additional postulate, whereas collapse interpretations need at least two: they both need a way to make distributions into probability distributions over observations, but collapse additionally requires some way of dodging Wigner's Friend type scenarios: an objective classical transition, extra state beyond the wavefunction, outright antirealism, etc.

[goatlover]

It's not really anymore than tables and chairs not being part of fundamental physics means they're additional postulates to physics. Because table and chairs emerge from the underlying physics. Same thing with measurement, observers and probability in MWI. Or to put it another way, the baggage of classical physics being quantized misleads us into thinking measurement and probability need to be fundamental postulates.