Albro and the Crypto-Positivist

Albro Swift wrote:
 Frank rejects the idea of two separate events (spin measurements) 
 possessing a joint correlation when they are spacelike-separated, 
 because there is no means of apprehending such a thing (barring
 superliminal communication).

Frank Wappler wrote:
 On the contrary, I suggested to analyze the two sets of recorded events
 through such correlation to define the alignment of the two separated
 detectors; that amounts to a measurement of the alingment of those
 separated detectors.

Given three Stern-Gerlach devices A,B,C, apply your method to determine 
that the angle between A and B is theta_AB and the angle between B and 
C is theta_BC.  According to conventional realism we could then infer 
that the angle between A and C is the sum

               theta_AC  =  theta_AB  +  theta_BC

but your procedure will not give this result.  Therefore, your definition 
of "angle" violates conventional realism.  Your "angles" do not add 
linearly, so they are not the angles of Bell (or Euclid). Please note 
that no one disputes we can reconcile QM with Bell's inequalities if we 
reject one or more aspect of conventional realism, and that's all you're 
doing.


Albro Swift wrote:
 ... by taking distant measurements and operating on them in
 various complex ways PRIOR to bringing the results together, it
 becomess necessary to define *everything*, from galaxy clusters to
 kitchen sinks, in terms of the simple observational results of each
 individual EPR trial.

Frank Wappler wrote:
 This sounds utterly incompatible with my idea of deriving measurements
 precisely from available observations (and the assumption that
 monotonous maps between ordered sets is a communicable/reproducable
 concept). Could you please explain?

Sure.  Following in the footsteps of everyone else who has ever given a 
moment's thought to Bell's inequalities, you've hit upon the idea of using 
an alternative definition of "alignment" to give a realistic explanation 
of EPRB experiments.  However, we can define more general types of EPR 
experiments involving the measurement of qualities other than spin, and 
in order to fully carry out this program it becomes necessary to re-define 
not only spatial orientation but also every other aspect of spatial and 
temporal relations, as well as momentum and energy, etc.  In each case 
the redefined concept is incompatible with the corresponding conventional 
realistic concept, notably with regard to transitivity.  As a result, if 
you follow this path you eventually arrive at traditional phase space and 
the Copenhagen interpretation, which simply doesn't trouble itself over 
conventional realism.  Of course, this still leaves unresolved the 
measurement problem and the apparent superluminal influences on the 
realistic side of the map.  The answer you're advocating (whether you 
know it or not) is just to ignore the conflict with conventional realism.


Frank Wappler trolled:
 How would we know?  I suggest we can know through calibration 
 procedures.

We can't "know through calibration procedures" because calibrations don't 
create knowledge, they are simply associations that we choose to impose on 
our knowledge as a way of organizing our information.  They are generally 
non-unique, so we have to make a selection on the basis of principles such 
as simplicity, unity, economy, etc.  You appear to be trying to make such 
a selection, but without the benefit of much knowledge as to the actual 
observations that need to be accommodated.

By the way, it's ironic that your deliberations have led you to elevate 
"calibration" to the status of "The Origin of All Knowledge", because there 
was another thread in this newsgroup not long ago in which some people 
argued that calibrations were evil and the source of nothing but Error, 
basically because calibrations require assumptions, and if we calibrate 
our measurements based on our assumptions we will tend to get results that 
agree with our assumptions.  So (they claimed) we can gain true knowledge 
only by practicing calibration-free science.  Of course, those people 
applied their anti-calibration philosophy with the same consistency and 
intellectual rigor that animates your positivism, viz, they deployed it 
only when polemically convenient.


Frank Wappler wrote:
 _I_ consider observations...not as predetermined.  ..._I_ find the 
 assumption not justified that _I_ could be sure about any of those 
 observations before _I_ collect them...

The question is not whether YOU can be sure of anything, but whether 
anything definite exists independent of your perceptions.  You say no, 
which makes you a solipsist, i.e., someone who completely rejects 
independent external reality.  Obviously considerations such as those 
leading to Bell's inequalities degenerate into meaninglessness (as 
does everything else) in a solipsistic context.


Frank Wappler wrote:
 It would be interesting to understand if you consider my rejection 
 of predetermined observations itself a rejection of one or more 
 aspect of conventional realism...

Of course it is.  Solipsism is the archtypical rejection of realism.  
As a means of explaining our knowledge of the external world, 
solipsism is recognized by philosophers and physicists alike as 
nothing but a reductio ad absurdum.  On top of this, you're further 
confusing yourself with the self-contradictory notion of "multiple 
communicating solipsists".


Frank Wappler wrote:
 I doubt that I could exclude that at least some of them were 
 signals from other observers, that they could in principle 
 collect observations about each other and exchange signals 
 _without_ my participation;...  Confronted with the prediction 
 "This hammer will hit that nail." I still try to consult with 
 my thumb...

You miss the point.  You are supposed to swing the hammer and
consult your thumb, and another "observer" 10 miles away swings a
hammer and consults his thumb, and you both write down on a piece of
paper the angle at which you were holding your thumb (relative to 
local vertical) and whether or not you hit it.  You each do this 
repeatedly, with arbitrarily selected angles of thumb positions, 
and write down a sequence of 1000 angles with the corresponding 
result (did or didn't hit your thumb on that trial).

Now, according to what you've said, you accept the reality and 
definiteness of both of these observers and their respective 
observations, independent of each other ("...they could in 
principle collect observations..._without_ my participation..."), 
so there are definite markings on their pieces of paper.  When 
the two observers put down their hammers and meet in some common 
location and compare results, the correlation is already determined, 
because the correlation consists of the union of the two separate 
realities that you already claim to accept.  On comparison they 
find the following distribution of joint outcomes on the combined 
nth trials:
                                 carpenter A

                           no hit           ouch!
                        ------------     ------------
              no hit    (1-cos(q))/4     (1+cos(q))/4
 carpenter B
              ouch!     (1+cos(q))/4     (1-cos(q))/4


where q is the difference between the angles of their thumbs on
the nth trial.  Clearly they each hit their thumb half the time,
independent of the angle (of either thumb).  However, the correlation
on the nth trial depends on the difference between the angles, which 
is not surprising in itself, but the correlation is a *non-linear* 
function of the difference, which IS surprising because there's no 
conventional realistic way of accomplishing such a correlation 
(assuming we have spacelike separated carpenters who never lose 
count, etc.)

The only way to alter the correlation between the marks made on those 
two pieces of paper is to either retro-actively alter the experienced 
reality of one or both of the observers, OR change the definition of 
the "relations" between those two realities.  But neither of those can 
be accomplished without rejecting the reality of at least one of the 
observers OR defining a new meaning for the "relation", in which case 
it will necessarily contradict the mass of other observations (such as 
sun sightings and any of a million other mutual observations that can 
be made to establish a common basis of anglular orientation), all of 
which give results that are consistent with each other but that will 
be inconsistent with your re-defined relation.

To reconcile QM with realism, you would need to explain, in realistic 
terms, how the markings on those two pieces of paper can give the 
results that QM predicts.  If you claim the angles written down 
need to be adjusted, then you need to explain (realistically) why 
the unadjusted angles as written nevertheless give consistent sun-
readings, to say nothing of why they are consistent with any of a 
million other observations that the two carpenters could make, 
individually or jointly.  (It would be pointless to change our 
definitions of the relations between observers ad hoc to resolve 
our EPR observations if in the process we render all previously 
well-behaved relations incoherent.)  In addition, you need to 
explain why the purported error in angular measurements was non-
linear, when the two protractors were fixed the entire time.  And, 
most difficult of all, you would need to keep a straight face while 
claiming that your explanation is realistic.

Ideas along these lines have been studied to death.  Angular relativity, 
etc.  But as I mentioned before, it isn't just angles that you have to 
accommodate.  It's all physical observables, and once you've accommodated 
them all in a consistent way, you will find that you've constructed 
phase space, not realistic space.

(Ironically, the modification of realism that captures QM's predictions 
while doing the least damage to conventional realism is actually the 
original suggestion in this thread, which you rejected out of hand, 
namely, the suggestion that quantum interactions are temporally 
symmetrical - as the relativisitic Schrodinger equation already 
implies.)