🚀 go-pugleaf

In article <mcirvin.718296445@husc8> mcirvin@husc8.harvard.edu (Mcirvin) writes:
>firth@sei.cmu.edu (Robert Firth) writes:
>
>>However, as I said, that's a minor point.  Your major point is that the
>>act of observing a system, and finding it seemingly unchanged from a
>>previous observation, can indeed affect the subsequent behaviour of the
>>system.  In other words, a "null observation" has an effect on the
>>thing observed.
>
>I was under the impression that, if the fields used to do the null
>observations are included in the model, the Zeno effect shows up without
>the need to postulate any sort of extra collapse event.  Does anybody know
>specifics?
>-- 
>Matt McIrvin, posting nonsense again 

This is true.

Papers by Axel Schenzle, in Phys Rev A and elsewhere, go into a lot of
detail about this.

That doesn't mean the collapse picture isn't the easiest way to deal with
things.  One of Schenzle's big points, if I've followed correctly, though,
is that collapse sometimes makes incorrect predictions.  I would rephrase
that and say that sometimes it is inappropriate to invoke collapse, i.e.,
when a measurement has not been made "strongly" enough.  The advantage to
his completely analytic approach is that there is never any such ambiguity.

If I remember correctly, though, his work still does turn stochastic at
some point.  He finds a certain probability of the "probe" field behaving
a certain way, and then his calculations at future times are really
calculations of correlation functions, i.e., what the probability of
an excited atom at t+tau is if there was no fluorescence at t, and so forth.

I could be misrepresenting, so insert loads of disclaimer.

Furthermore, as to the comment about this not working for radiative decay,
I think that depends.  The exponential decay follows from a coupling to
an infinite number of modes; if you modify that, e.g., by putting the
decaying nucleus into some sort of hypothetical cavity which limited the
number of modes into which decay was possible, and find a suitable probe
coupling, then at small times the "exponential" would actually have zero
slope as required, and behave quadratically.  In fact, some people have
claimed to me that this is true in general regardless of whether you're
in an open system or not, and that the Wigner-Weisskopf approximation used
to get exponential decay treats the density of states as a constant, and
that this approximation is never true if you ask about sufficiently short
times.  In a truly open system, I've had trouble understanding how this
latter claim could be correct.

Any one understand this stuff well enough to say something definitive?

I also know that Fearn & Loudon have worked on this problem, and I think
they disagree with Schenzle.  At one point I even saw them present something
saying that the quantum Zeno effect shouldn't happen.  Unfortunately, the
model they were playing with seemed different from the experiments that
have been done, in that the probe coupled with the INITIAL state rather
than the final state, and was basically able to kick the particle into
the other state.  (They used position measurements in a double-well structure.)

Some of these disputes are collected in the recent volume put out by
World Scientific, "Foundations of Quantum Mechanics," edited by Sinclair,
Scully, & Nieto, or some such permutation.


-- 
Aephraim M. Steinberg                    | "WHY must I treat the measuring
UCB Physics                              | device classically??  What will
aephraim@physics.berkeley.edu            | happen to me if I don't??"
                                         |       -- Eugene Wigner