There's more -- if it is the observation of the cat that is supposed to determine
its state, does the cat qualify as a valid observer (that is, can it determine its
own state?)? If a cat cannot, how about putting a person into the box... is the
person alive or dead before someone else opened the box, or are observers themselves
subject to being in superpositions of states (when the observer opens the box,
the result is that the observer is in a superposition -- one watching a live cat,
and the other watching a dead cat)?
One comment on one observer seeing a dead cat followed by a live cat... The
live and dead states are orthogonal, and the time-evolution operator for the
system does not transfer amplitude from the dead state to the live state... only
the other way around. In other words, the cross-section for an event that shows
a live cat after seeing a dead cat is zero.
This is not true. Once an observation has been made, it can'tbe unmade.
The wave function can't uncollapse.
The collapse of the wave function is such a shonky concept, which lacks
so much in any decent mathematical treatment that "uncollapse" is just as
plausible as "collapse". For more details see: John Bell, "Against
'measurement'", Physics World, August 1990, pp. 33-40, and references
therein. Strictly speaking "wave function collapse" is not a part of QM.
It's an "add on".
Zdzislaw Meglicki, gus...@arp.anu.edu.au,
Automated Reasoning Project - CISR, and Plasma Theory Group - RSPhysS,
The Australian National University, G.P.O. Box 4, Canberra, A.C.T., 2601,
Australia, fax: (Australia)-6-249-0747, tel: (Australia)-6-249-0158
Isn't the point of Schroedinger's cat that it was used by Schroedinger to show
that Quantum physics did not work on large scales in the same way as it
did on the mircospoic scale. None of this mumbo jumbo about cats observing
their own states.
I suppose I should read Scrhoedinger's oringal article, if any body know the
reference I would be most grateful.
Justin Pearson, | So what do you thinks going on? |
University of Kent at Canterbury | |
U.K. | |
> DOC...@SLACVM.SLAC.STANFORD.EDU (Jon J Thaler) writes:
>> i1n...@exnet.iastate.edu (Neal Rauhauser) says:
>>> And its quite possible to open the lib, see the dead kitty, slam
>>> the lid, re-open it, and have a live cat again. Note that this is
>>> not a very probable occurence, but it is possible. Confused yet?
>> This is not true. Once an observation has been made, it can'tbe unmade.
>> The wave function can't uncollapse.
> The collapse of the wave function is such a shonky concept, which lacks
> so much in any decent mathematical treatment that "uncollapse" is just as
> plausible as "collapse". For more details see: John Bell, "Against
> 'measurement'", Physics World, August 1990, pp. 33-40, and references
> therein. Strictly speaking "wave function collapse" is not a part of QM.
> It's an "add on".
I don't know what point you are trying to make. Measurement repeatibility
is a feature of every interpretation of QM. You can use wavefunction
collapse, many worlds, or hidden variables, but none of them predict
that the cat, once observed to be dead, will ever again be seen alive.
Only Elvis has that property.
I like that, "shonky". What does it mean, precisely?
>"uncollapse" is not plausible
Barring any evidence for this statement, much less any definition of
"collapse" or "uncollapse", this doesn't mean very much.
I've always been against the notion of "wave function collapse" as some
kind of physical process caused by "observation" (a term which has no
clear definition anyway). If one insists on "collapse" one either has
to aceept "uncollapse" or concoct a tale about how time-reversal
invariance is thwarted by thermodynamic subtleties... actually I think
some story along these lines is needed to account for the curious
time-reversal asymmetry of what we call observation -- but I think that
dragging in "wave function collapse" only makes matters murkier.
I stand partially corrected. To the extent that 'alive' and 'dead' are
quantum eigenstates of some 'life' operator, the state of the cat can be
changed from one state to the other in much the same way that the polarization
of light can be rotated by a succession of linear polarizers.
Don't over-interpret Bell. Sure, wavefunction collapse is a poorly
defined process which isn't part of (or reconcilable with) the
smooth time-evolution of QM states. But what really bothers lots
of people is the fact that the reduction occurs on interaction
with the classical measuring apparatus. Where is the split?
"uncollapse" is not plausible
In the context of the above, there are a few points I'd like to complain about
when it comes to wave-function collapse -- particularly as formulated in
Cohen-Tanoudji et al and in Dirac's monograph:
1) The time evolution operator involving observation does not seem to be
unitary (same class of complaint as the time-reversal asymmetry).
2) If you wanted to construct a Hamiltonian (it would have to be non-Hermitian
to produce a non-unitary time-evolution operator) to represent the impact of
observation, it would have to account the effective scattering from the input state
to the output (observed) state... ie you'd have to do work. Yet, observing the cat
is essentially a passive process. Somehow, I'm producing a profound scattering of
states with essentially no energy (no interaction term in the Hamiltonian) coupling
the cat's state to my observation of its state.
3) The observations one can see in quantum experiments can be explained
by solving Schrodinger's equation for the system+observation process in question, and
looking at the observables. For example, you can look at the production of
bubble-chamber tracks (look in Schiff for a nice discussion), or you can look at
linear response theory (Kubo) to see how a system's quantum states interact with an
applied field in a way that can be examined at a macroscopic level (both these
techniques treat the measurement process as a part of the quantum system...).
(NOTE that the probability of seeing weird non-conservative events, or seeing a live
cat following seeing a dead cat is very small -- as calculated from the stand-point
If it is clear from 1) and 2) that the impact of collapse produces scattering of
states in a way that is completely divorced from dynamical processes (Schrodinger's
equation), and by 3) handles a problem that is already handled by looking at
cross-sections that follow from simple dynamic considerations, why would one NEED
It's an Oz-tralian-ism. It means the same as "dodgy" :)
Tee hee... Actually, it means something of questionable
reliability or doubtful character. It can be applied to people. We
have a comedy show here which features the "Dodgy Brothers" who go in
for things like Used Car Selling and the like. Hope this has enriched
you in some way. :)
We now return you to your regular programme.
Department of Physics, Monash University, Clayton, VIC, 3168, Australia.
Quantum Mechanics is God's way of telling you to mind your P_i's and Q_i's.
On that subject, it seems to me that most treatments of measurement
as compatible with Schrodinger evolution wind up sneaking the split
back in somewhere along the line - for example, Kurt Gottfried's
treatment as quoted in the Bell article. Basically, Gottfried's
cooked up a dish that's palatable to him; but many other people
find it less than filling. I haven't got Schiff here - but my
instinct is that his treatment is similar.
If I was the only person who was bothered by this, I would figure
that I just didn't understand it. I take comfort from the fact
that a lot of smart physicists don't seem to understand it either.
That may not be scientifically objective, but it's human.
Yes, you forgot to uncollapse its wave function.
Um, how exactly do I do that? It is rather urgent; my landlady wants
to know where her cat is.
Yes - Schrodinger's experiment calls for a randomly selected cat.
By using the same subject you are biasing the results because this
cat has learned from past experiments and is affecting the outcome.:-)
>>>> Look, I tried the cat experiment. On the third trial, the
>>>>cat was dead. On each of the subsequent 413 trials, it remained dead.
>>>>Am I doing something wrong?
>>Yes, you forgot to uncollapse its wave function.
> Um, how exactly do I do that? It is rather urgent; my landlady wants
>to know where her cat is.
There are various ways; I suggest you just switch the sign of the
Hamiltonian and wait for as many days as the cat has been dead. If your
landlady is impatient you may also want to multiply the Hamiltonian by a
Senior physicists, labeled by many younger ones as "the establishment,"
contend that an advanced degree in physics is an excellent education, even
if it leads to no job. Besides, physicists are at the pinnacle of science,
they argue, and superbly qualified to take jobs as applied scientists,
engineers or even plumbers. NY Times Tues. Mar. 10 1992, p. C1.