I think we can probably agree to disagree on the starting points here, but some more comments...
My objection to the fixed unambigous border have two components.
1. In the common birds view, pictured as a realist type of external view, the decomposition of the universe as a whole is ambigous by the same token that there are almost an infinite number of ways to divide a line. All decompositions are equally plausible.
The argument here is clear and understandable, but there are several objections.
- First of all, the birds view lacks physical basis. I think it should have no place in our abstractions. All that's relevant are physical views.
If I understand you right, we seem to roughly agree on this point?
But I have another objection to the border problem.
2. Even in a physical inside view, manifested by a real observer including his experimental apparatous, the observers physical makeup including his experimental devices are always subject to change. They are part of the observers identity. A scientists builds maintains and develops his measurement devices, and you can certainly tell, from the device he has built, what he expects. The apparatous is the correspondence to a question. A scientists could easily build all kinds of wicked "apparatous", but still there is a reason for what certain devices are build and others are not, just like there are good questions, well worth asking, and bad questions.
So the observers identity, and thus boundary can in principle change during interactions. If you call this "an experimental problem" then ok, but in my opinion experimental problems are also subject to theoretical analysis. It is relevant.
Edit: this is also the basis for my objection to the uncritical use of probabiltiy theory. Clearly in such a scenario the entire notion of repeatable experiment and statistics becomes doubtful at best.
The most serious problems is when the observers complexity increases or decreseas. Which is related to an observer that is gaining, or loosing mass. In such a scenario there is loosely speaking a preferred boundary, I agree, but this boundary is still uncertain AND evolving.
I really don't see how the abstraction of a fixed boundary is realistic.
IMO, the very question someone asks, or the very experimental apparatous a scientist is constructing, is not random. It pretty much reveals what the observer knows and doesn't know. But what the observe knows is subject to revision all the time.
The reason why I think knowledge is not static, is that the only way to verify it is to put to to test in a real process. And real processes has duration, and during all processes knowledge changes.
The problem with QM, is thus that part of the prior information, that SHOULD be subject to revision if necessary, is hardcoded into hilbert spaces etc. This is a form of generalized background dependence that I think makes no sense at least in approximate cases, as we allready know.
meopemuk said:
However, if you consider each particular experimental setup, then the distinction between the two parts is clear. For example, in the famous double-slit experiment the scintillating screen (or photographic plate) is the measuring apparatus, and the electron constitutes the measured physical system.
I agree that from a FAPP point of view, the example you take, there is a reasonable clear boundary.
But, a more complete picture is the context of this screen. It's not a randomly placed and manufactured screen, the entire preparation of beams, screen postions etc contains massive information provided by the scientist. This context must be accounted for, in a evolving context, and not be hardcoded into the formalism.
You are from my point of view, trying to suggest that there is a clear separation between information, and the context of the information. Ie. a clear separation between the state of memory, and the memory hardware itself.
I suggest there is no such clear separation. Instead there is a feedback between the memory state and the memory hardware. We have this in general relativity, Einsteins equations is a feedback between the state relative spacetime, and spcetime itself.
Now forget about SPACETIME and instead consider the same reasoning that lead to GR, to a general information state, and the microstructure system that encodes the information state. SImilarly do Iexecpt a yet not identified correspondence to Einsteins equation, that is cast in terms of a measuremet theory and information, and is much more general rather than beeing constrained to a 4D manifold.
/Fredrik
