Conceptual overlap between LQG and String/M-theory?
Gene Partlow
While we hear mainly about the rivalry between Loop Quantum
Gravity (LQG) and String/M-theory (S/M), and occasional name
calling thereby, we read more rarely (eg: Brian Greene et al)
that there are hints that both approaches may be slowly homing in
on the SAME theory, but from different perspectives or frames.
That is, we have two blind men groping the same elephant (and
very respectfully, I'm sure).
What I want to know is: What are the qualities of both efforts that
suggest that both ideas are actually two parts of the SAME gem. I
know that both LQG and S/M make fairly impressive contact with things
already known...eg: black hole entropy, etc? I am more interested in
what parts, if any, of the machineries of each approach resembles (or
hints at) aspects of the machineries of the other approach ...what is
the actual degree of overlap in both sets of core ideas?
Further, if there is any such overlap, are any theorists focusing some
effort on exploring and expanding the points of contact in the actual
idea/tools of both LQG and S/M, and with what results?
Hopefully,
Gene Partlow
[PS: I would have included the rarely mentioned Twistor Theory as
a third 'contender'/blindman, but I wanted to keep it simple]
[PPS: Theorists now suspect that gravity may 'leak out' of a brane
universe via extra dimensions. If the same thing happens with
time, it would go some way toward explaining why we're late for
appointments so often! No need to thank me for this insight...
just spell my name right when you nominate me for the door
prize at Uppsala.]
John Baez
Gene Partlow <star...@earthlink.net> wrote:
>While we hear mainly about the rivalry between Loop Quantum
>Gravity (LQG) and String/M-theory (S/M), and occasional name
>calling thereby, we read more rarely (eg: Brian Greene et al)
>that there are hints that both approaches may be slowly homing in
>on the SAME theory, but from different perspectives or frames.
We can hope so. And who knows - it might even be the RIGHT theory.
>What I want to know is: What are the qualities of both efforts that
>suggest that both ideas are actually two parts of the SAME gem.
The simplest point is that string theory says everything is
made of little strings, while loop quantum gravity says everything
is made of little loops. Viewed with sufficiently smeared
eyeglasses, these don't look all that different. Of course the
details are extremely different, but one might hope that these
differences mainly arise from the fact that 1) string theorists
haven't worked as hard at doing everything in a background-free
manner, and 2) loop theorists haven't thought as hard about
describing matter, or using higher-dimensional versions of loops
(which show up as "membranes" in string theory).
>Further, if there is any such overlap, are any theorists focusing some
>effort on exploring and expanding the points of contact in the actual
>idea/tools of both LQG and S/M, and with what results?
Various people including Lee Smolin and I have attempted to
make the connection a bit more precise - with only very modest
success, at least in my case. We've both published some papers
on this business, but Smolin's best idea so far has been to start
an institute where lots of people talk about both loop quantum
gravity and string theory: the Perimeter Institute. I'm going
there for 3 weeks starting on Tuesday March 4th, and I'll report
back on what it's like. If we unify string theory and loop
quantum gravity, you'll be the first to know!
Here's a nice new interview of Smolin, in which he talks a bit
about the relationship between loop quantum gravity and string
theory, as well as many other things:
http://www.edge.org/3rd_culture/smolin03/smolin03_index.html
Fizz Fann
> In article <c504f3da.0302...@posting.google.com>,
> Gene Partlow <star...@earthlink.net> wrote:
>
> >While we hear mainly about the rivalry between Loop Quantum
> >Gravity (LQG) and String/M-theory (S/M), and occasional name
> >calling thereby, we read more rarely (eg: Brian Greene et al)
> >that there are hints that both approaches may be slowly homing in
> >on the SAME theory, but from different perspectives or frames.
>
> We can hope so. And who knows - it might even be the RIGHT theory.
Can you say how many loop quantum gravitists or string theorists think
they are likely to be the same theory? Is there a general consensus
that their converging, or is it more like a few people at the edge who
see faint glimmers of a correlation?
I want to ask a more general question as well, but I add the cave at that
apart from being a physics major, I don't know much about physics! The
question is which of the theories of everything deals the best with the
weird stuff that happens when you measure something quantum
(Schrodinger's cat ect.)?
What I remember is that there's a "quantum jump" when you make a
measurement and the system does something discontinuous. The idea
was supposed to be that when a more final theory comes along, the
discontinuous part will look like a smooth normal process in the new
theory and that will solve part of the mystery.
I suspect that the answer will be that if their two sides of the same elephant,
then LQG and String Theory answer this question in the same way. I
guess I just want to know whether the cat is alive or dead! :-)
Alex.
Moataz H. Emam
> on the SAME theory, but from different perspectives or frames.
> That is, we have two blind men groping the same elephant (and
> very respectfully, I'm sure).
Being a student of string theory, I knew next to nothing about LQG. One
fine day in a conference in Chile I had a discussion over lunch with a
student of LQG who knew nothing of string theory. I was talking about
coherent closed string states and how these 'average' out to curved
backgrounds in string theory and other fields. The other student's eyes
began to widen and then told me that if he had just walked in the door
now, he would have thought I was talking about the loops in LQG. He
explained the similarities to me in plain words.
Understand that these discussions were far from mathematical. We were
having lunch and discussing things generally. I am sure that once one
went down to details, the similarities would immediately disappear, but
it was somewhat obvious that there ARE similarities to begin with. In
physics, one is used to the concept of having apparently unrelated
theories actually describing the same phenomena and are related by some
duality transformation.
--
Moataz H. Emam
Department of Physics
University of Massachusetts
Urs Schreiber
> Being a student of string theory, I knew next to nothing about LQG. One
> fine day in a conference in Chile I had a discussion over lunch with a
> student of LQG who knew nothing of string theory. I was talking about
> coherent closed string states and how these 'average' out to curved
> backgrounds in string theory and other fields. The other student's eyes
> began to widen and then told me that if he had just walked in the door
> now, he would have thought I was talking about the loops in LQG. He
> explained the similarities to me in plain words.
This reminds me of the following, somewhat heretic, point: As
everybody knows, the big motivation for LQG is to have a
background free theory of quantum gravity, as opposed to
perturbative string theory where strings propagate on a
supposedly classical background manifold. But a closer
examination of the target space of a string sigma-model seems
to reveal (see references below) that the classical background
metric that enters the single string's action functional
really just defines the macroscopic limit of an otherwise very
non-classical spacetime that is already present in string
perturbation theory by way of the vertex operator algebra. To
say this differently: According to the references given below,
to any given classical background metric perturbative string
theory can and does associate a definite noncommutative
geometry that has the classical background as a large-scale
limit. Intuitively this is because the classical metric that
enters the sigma-model cannot be operationally probed by means
of strings.
In this sense, shouldn't the "background" metric of
perturbative string theory better be regarded as a "boundary
value" that fixes the classical limit of the non-classical
spacetime that the single string propagates in? Viewed this
way, there is not much of a difference to the way a classical
spacetime is treated in the LQG framework: Given a fixed
classical background metric we can try to find states
(superpositions of spin networks) that approximate this
classical background. To do so we have to "fix" the classical
background that we want to approximate.
Hm, maybe I am just arguing about words. I guess what I really
want is to talk about the following references :-):
The most self-contained presentation of the program of
constructing a non-commutative string spacetime from the
vertex operator algebra seems to be
Fedele Lizzi anmd Richard Szabo, Duality Symmetries and
Noncommutative Geometry of String Spacetimes, hep-th/9707202 .
It is based on earlier work by J. Froehlich et al:
J. Freohlich and K. Gawedzki, Conformal Field Theory and
Geometry of Strings, hep-th/9310187
A. Chamseddine and J. Froehlich, Some Elements of Connes'
Noncommutative Geometry and Space-time Geomtetry, in Yang
Festschrift, eds. C. Liu and S. Yau, International Press
Boston (1995), 10-34 .
Moataz H. Emam
> I want to ask a more general question as well, but I add the cave at that
> apart from being a physics major, I don't know much about physics! The
> question is which of the theories of everything deals the best with the
> weird stuff that happens when you measure something quantum
> (Schrodinger's cat ect.)?
> What I remember is that there's a "quantum jump" when you make a
> measurement and the system does something discontinuous. The idea
> was supposed to be that when a more final theory comes along, the
> discontinuous part will look like a smooth normal process in the new
> theory and that will solve part of the mystery.
Unfortunately, it doesn't work that way. Both theories use quantum
mechanics. In other word, neither of them 'explains' QM or provides more
insight in its inner workings. Before I studied string theory, I had the
vague hope that it would provide more info on the 'hidden workings' of
QM, but of course it doesn't. The string is quantized in pretty much the
same way a particle is quantized. No more info, no hidden variables. So
far, Bohr's view, as opposed to Einstein's, is still the winner :-)
Uncle Al
> In article <c504f3da.0302...@posting.google.com>,
> Gene Partlow <star...@earthlink.net> wrote:
> >While we hear mainly about the rivalry between Loop Quantum
> >Gravity (LQG) and String/M-theory (S/M), and occasional name
> >calling thereby, we read more rarely (eg: Brian Greene et al)
> >that there are hints that both approaches may be slowly homing in
> >on the SAME theory, but from different perspectives or frames.
[snip]
> Here's a nice new interview of Smolin, in which he talks a bit
> about the relationship between loop quantum gravity and string
> theory, as well as many other things:
>
> http://www.edge.org/3rd_culture/smolin03/smolin03_index.html
"It's only since the middle 1980s that real progress began to be made
on unifying relativity and quantum theory. The turning point was the
invention of not one but two approaches: loop quantum gravity and
string theory. Since then, we have been making steady progress on both
of these approaches. In each case, we are able to do calculations that
predict surprising new phenomena. Still, we are not done."
That's news to me! What empirical predictions constrain M-theory
and/or loop quantum theory? What are the "surprising new phenomena?"
--
Uncle Al
http://www.mazepath.com/uncleal/eotvos.htm
(Do something naughty to physics)
John Baez
Fizz Fann <thephy...@yahoo.com> wrote:
>ba...@galaxy.ucr.edu (John Baez) wrote in message
>news:<b3kq2r$s90$1...@glue.ucr.edu>...
>> In article <c504f3da.0302...@posting.google.com>,
>> Gene Partlow <star...@earthlink.net> wrote:
>> >While we hear mainly about the rivalry between Loop Quantum
>> >Gravity (LQG) and String/M-theory (S/M), and occasional name
>> >calling thereby, we read more rarely (eg: Brian Greene et al)
>> >that there are hints that both approaches may be slowly homing in
>> >on the SAME theory, but from different perspectives or frames.
>> We can hope so. And who knows - it might even be the RIGHT theory.
>Can you say how many loop quantum gravitists or string theorists think
>they are likely to be the same theory?
Well, they sure aren't the same theory NOW.
So I assume you mean: "...are likely to become the same
theory when we mess around with them enough to get them to work?"
>Is there a general consensus that they're converging, [...] ?
No, nothing that strong! A better description would be "a few people
hope that they might converge if we work at it".
First of all, most string theorists don't know enough about loop
quantum gravity to have an opinion on this convergence issue.
In fact, most of them don't even care!
Loop quantum gravitists are the main ones who care about
this issue. Within that community, about half the people hope
that string theory and loop quantum will converge if we work at it.
The rest wish that string theory would shrivel up and die.
Don't forget, there are roughly 10 times as many string theorists
as loop quantum graviters. So, loop quantum gravity people feel
they need to pay attention to string theory, but not so much vice versa.
It has wisely been said: "We don't know yet what the Theory of
Everything will look like... but we know what it will be called:
string theory."
If string theory and loop quantum gravity converge, the end result
will probably be called string theory, or maybe M-theory, simply
because there are more string theorists.
>I want to ask a more general question as well, but I add the caveat that
>apart from being a physics major, I don't know much about physics! The
>question is which of the theories of everything deals the best with the
>weird stuff that happens when you measure something quantum
>(Schrodinger's cat etc.)?
None of the most popular Theories of Everything attempt to
deal with this issue. For that, they believe, you need a
Theory of Everything Else.
The reason is that most people working on Theories of Everything
believe that these puzzles can be resolved not by new physical
theories but by improving how we *think about* quantum mechanics.
I share this view.
On the other hand, there are some people who think quantum theory
needs to be changed to deal with these puzzles, most notably Roger
Penrose. If you wish to learn more about this, use Google and other
resources to look for stuff about "objective reduction".
>What I remember is that there's a "quantum jump" when you make a
>measurement and the system does something discontinuous. The idea
>was supposed to be that when a more final theory comes along, the
>discontinuous part will look like a smooth normal process in the new
>theory and that will solve part of the mystery.
I don't think there is anything abnormal or discontinuous that
happens when a measurement is made. I claim that according
to quantum mechanics, the laws governing measurements are exactly
like the laws governing any other process.
A lot of people agree with me... but a bunch don't. You will find
that these issues elict endless argument. I'm sick of these arguments,
so I won't bother to defend my views. I did plenty of arguing about
this in my youth.
John Baez
Uncle Al <Uncl...@hate.spam.net> wrote:
Lee Smolin says in
http://www.edge.org/3rd_culture/smolin03/smolin03_index.html
>>"It's only since the middle 1980s that real progress began to be made
>>on unifying relativity and quantum theory. The turning point was the
>>invention of not one but two approaches: loop quantum gravity and
>>string theory. Since then, we have been making steady progress on both
>>of these approaches. In each case, we are able to do calculations that
>>predict surprising new phenomena. Still, we are not done."
>That's news to me! What empirical predictions constrain M-theory
>and/or loop quantum theory? What are the "surprising new phenomena?"
Well, he didn't say the phenomena could be detected using currently
available apparatus! But, we might be getting close.
Superstring theory predicts the existence of superpartners, and I
seem to recall that if the (not-at-all-understood) spontaneous breaking
of supersymmetry occurs at an energy scale low enough to fix the
flaws of the SU(5) grand unified theory, evidence for supersymmetry
should be detectable by the Large Hadron Collider sometime around 2007.
For details, you'd have to ask an expert.
In the case of loop quantum gravity, we expect a kind of "graininess"
of spacetime at the Planck scale, and while this might seem hopelessly
beyond current experiment, astronomical measurements are beginning to
put some constraints on what this graininess can be like - see Lieu
and Hillman's paper:
Richard Lieu, Lloyd W. Hillman
The phase coherence of light from extragalactic sources - direct
evidence against first order Planck scale fluctuations in time and space
http://www.arXiv.org/abs/astro-ph/0301184
However, caveat emptor:
D. H. Coule
Planck scale still safe from stellar images
http://www.arXiv.org/abs/astro-ph/0302333
and see also my comments earlier on s.p.r.
Fizz Fann
> >Can you say how many loop quantum gravitists or string theorists think
> >they are likely to be the same theory?
>
> Well, they sure aren't the same theory NOW.
> So I assume you mean: "...are likely to become the same
> theory when we mess around with them enough to get them to work?"
Ok, I guess I didnt fully get it before. I had thought that you were
speculating that they might be something like fractions and decimals -
two different mathematical systems which describe the same things.
Is it better to say that there are paramaters in the two theories
that might be tweaked to bring them to the situation like fractions
and decimals, but at the moment they're different, and the hope is
that where they can be made to match, we have the right theory?
> >Is there a general consensus that they're converging, [...] ?
> No, nothing that strong! A better description would be "a few people
> hope that they might converge if we work at it".
Ok. I'll join the group of hopers, because if I ever try to understand
this stuff properly, I'll only have to learn one of them!
> Don't forget, there are roughly 10 times as many string theorists
> as loop quantum graviters. So, loop quantum gravity people feel
> they need to pay attention to string theory, but not so much vice versa.
That sounds like a bad situation. The drive to unify the two theories
is a sociological effect instead of being driven by the science?
> If string theory and loop quantum gravity converge, the end result
> will probably be called string theory, or maybe M-theory, simply
> because there are more string theorists.
Ok, so I have a question then. It seems that string theory can tell us
about quarks and electrons too, but I got the impression (mostly from
the name) that loop quantum gravity was just about gravity. I take it
now that I was wrong on this, but I suspect everyone else who isnt
a specialist is under the same impression.
> >I want to ask a more general question as well, but I add the caveat that
> >apart from being a physics major, I don't know much about physics! The
> >question is which of the theories of everything deals the best with the
> >weird stuff that happens when you measure something quantum
> >(Schrodinger's cat etc.)?
> None of the most popular Theories of Everything attempt to
> deal with this issue. For that, they believe, you need a
> Theory of Everything Else.
That also sounds very bad. Do theorists of everything wake up covered
in sweat in the middle of the night thinking their wasting their
life?
> The reason is that most people working on Theories of Everything
> believe that these puzzles can be resolved not by new physical
> theories but by improving how we *think about* quantum mechanics.
>
> I share this view.
I know you said that you don't like to talk about this, so I don't
want to pester you too much about it, and I'm sorry for sickening
you with old questions. I read through the Omnes book that you
recommended and it gave me a headache, but I think I see the
ideas involved - thanks.
> I don't think there is anything abnormal or discontinuous that
> happens when a measurement is made. I claim that according
> to quantum mechanics, the laws governing measurements are exactly
> like the laws governing any other process.
>
> A lot of people agree with me... but a bunch don't. You will find
> that these issues elict endless argument. I'm sick of these arguments,
> so I won't bother to defend my views. I did plenty of arguing about
> this in my youth.
Ok I dont want to attack your views, but just to find out what they
are, because you mentioned getting over prejudices before would help
it become clear and that made me interested. I gather from everything
that I've seen that the two possibilities for not-collapsing views
are the "Many Worlds" idea and the "Bohmian" one that Ilja Schmelzer
mentioned before, and that the one you prefer is the many-worlds one.
> But you said that
> ... most people working on Theories of Everything
> believe that these puzzles can be resolved not by new physical
> theories but by improving how we *think about* quantum mechanics.
So do you mean that most of the people doing string theory or loop
quantum gravity use the Many Worlds Idea as well?
And when you say "believe", do you mean that the Many Worlds idea
hasnt been improved enough yet to solve the problems?
If these questions sicken John too much, I'd be happy to hear from
anyone else whose prepared to tell me.
Alex.
Fizz Fann
> Fizz Fann wrote:
> > What I remember is that there's a "quantum jump" when you make a
> > measurement and the system does something discontinuous. The idea
> > was supposed to be that when a more final theory comes along, the
> > discontinuous part will look like a smooth normal process in the new
> > theory and that will solve part of the mystery.
>
> Unfortunately, it doesn't work that way. Both theories use quantum
> mechanics. In other word, neither of them 'explains' QM or provides more
> insight in its inner workings. Before I studied string theory, I had the
> vague hope that it would provide more info on the 'hidden workings' of
> QM, but of course it doesn't. The string is quantized in pretty much the
> same way a particle is quantized. No more info, no hidden variables. So
> far, Bohr's view, as opposed to Einstein's, is still the winner :-)
So you mean that the wavefunction collapse still happens in string theory
and in LQG? Does that mean that we shouldn't view these theories as
actual "theories of everything", but that, after they are finished, we
expect another, _more final_ theory which doesn't have this feature?
On the subject of Bohrs view, which I am trying to understand, it
seems that he thought of quantum mechanics as being an algorithm
for predicting probabilities, rather than a true picture of the world.
He said something like that their is no quantum world - their is only
an absract quantum description. Is that the way that string theorists
see their own calculations? I had the impression that it was really
considered to be the final, perfect true picture of the world.
Regards,
Alex.
News Admin
on 4/3/03 7:30 pm, Urs Schreiber at Urs.Sc...@uni-essen.de wrote:
> "Moataz H. Emam" wrote:
>
<about formal similiarities between LQG and string theory>
But we'd like the classical 'background' to be dynamic, wouldn't we? And
we'd like the dynamics to emerge from the theory, rather than the other way
round, yes? So putting the background metric in as a boundary condition,
though it may give you an interesting insight into its small-scale quantum
structure, doesn't really achieve the goal of having a quantum theory of
spacetime that reduces to the classical theory on the large scale. It just
reduces to particular classical solutions, instead.
N'est-ce pas?
<snip refs.>
Tim
Jeffery
thephy...@yahoo.com (Fizz Fann) wrote in message
news:<b2b24cf0.03030...@posting.google.com>...
> Ok, so I have a question then. It seems that string theory can tell us
> about quarks and electrons too, but I got the impression (mostly from
> the name) that loop quantum gravity was just about gravity. I take it
> now that I was wrong on this, but I suspect everyone else who isnt
> a specialist is under the same impression.
No, you were right before. String theory is an attempt to unify
gravity with the other forces, electromagnetism, weak force, strong
force. Loop quantum gravity does not attempt any such thing. It's an
attempt to come up with a quantum version of general relativity.
You are still attaching too much significance to the phrase "same
theory". Even the people who are hoping they are the "same theory",
when they say "same theory", they don't mean as much the "same" as you
seem to think they mean. A better way to think of it is that they are
both attempting to model different parts of some underlying truth, but
they modeling different parts of it. They are both attempts to
quantize gravity but they are not the same thing. Some optimistic
people are hoping that they are trying to explain different parts of
the same thing. So you're still going to have to learn both of them
but you should spend more time studying string theory.
It's common in physics for there to be two main rival strategies to
approaching a problem, and then one becomes successful, and the other
just sort of withers away. For a while, you have the vast majority of
physicists studying one main type of theory, and then a small minority
studying an alternative approach, which I think is good because it's
good to have more than one way of approaching the problem, but if the
main way is more successful, you should acknowledge that. QCD explains
strong force interactions, and Regge theory was an alternative
approach. Supersymmetry explains the hierarchy problem, and
technicolor was an alternative approach. Superstring theory is an
attempt to quantize gravity, and loop quantum gravity is an
alternative approach. Dark matter explains the rotation of galaxies,
and MOND was an alternative approach.
> So do you mean that most of the people doing string theory or loop
> quantum gravity use the Many Worlds Idea as well?
Absolutely not. Very few people ascribe to the Many Worlds Principle.
Most physicists ascribe to the Copenhagen interpretation.
Modern physics, such as string theory, doesn't solve the problems of
quantum mechanics, which is from the 1920's. Actually, there really is
no problem with quantum mechanics. There is no paradox associated with
Schrodinger's Cat. If you study quantum mechanics, like relativity, it
makes perfect sense, if you just let go of the classical view of the
world. I suggest you better understand the physics of 80 years ago
before you try to understand the physics of today.
Jeffery Winkler
Aaron Bergman
"News Admin" <ne...@news.demon.net> wrote:
> But we'd like the classical 'background' to be dynamic, wouldn't we? And
> we'd like the dynamics to emerge from the theory, rather than the other way
> round, yes? So putting the background metric in as a boundary condition,
> though it may give you an interesting insight into its small-scale quantum
> structure, doesn't really achieve the goal of having a quantum theory of
> spacetime that reduces to the classical theory on the large scale. It just
> reduces to particular classical solutions, instead.
You get classical solutions plus perturbations around them. So, you get
perturbative dynamics, about what you'd expect from a theory defined
perturbatively.
Aaron
--
Aaron Bergman
<http://www.princeton.edu/~abergman/>
<http://aleph.blogspot.com>
Urs Schreiber
"News Admin" <ne...@news.demon.net> wrote in message news:<BA8FAAFF...@timsilverman.demon.co.uk>...
> But we'd like the classical 'background' to be dynamic, wouldn't we?
Yes!
> And
> we'd like the dynamics to emerge from the theory, rather than the other way
> round, yes?
Yes!
> So putting the background metric in as a boundary condition,
> though it may give you an interesting insight into its small-scale quantum
> structure, doesn't really achieve the goal of having a quantum theory of
> spacetime that reduces to the classical theory on the large scale. It just
> reduces to particular classical solutions, instead.
I know that my original posting was somewhat "heretic", as I had
written, and I expected my statement to be refuted immediately (if
anyone cared) with arguments like: "There is no back-reaction of the
single string with the background spacetime." (Namely in this sense
the backgound is fixed in the technical sense: We do not vary the
background fields in the sigma-model action. I seem to have learned
from John Baez that every object in an action that is not varied
counts as fixed background.)
To address your points: First, how is the dynamics of spacetime
encoded in LQG? Obviously there are the kinematical states, for
instance being spanned by a basis of spin network states, and on these
the dynamics is imposed by imposing the various constraints of the
theory, the quantized version of Einstein's equations.
In string theory one instead finds that the background metric has to
satisfy the (higher- order corrected) classical equations of motions.
But I think this is not in contradiction to the fact that perturbative
string theory defines a non-classical (quantum) spacetime with
large-scale limit given by the background metric. That's because we
are dealing with coherent states.
Consider as an example the single-particle non-relativistic quantum
harmonic oscillator. Pick any coherent state |alpha(t)> with as yet
undetermined t-dependence and enter the Schroedinger equation with
this state. Of course youl'll find that this is equivalent to the
classical harmonic oscillator equation of motion for alpha(t). I think
what happens in string theory with the background fields is completely
analogous. They are coherent states of gravitons (and possibly other
quanta) and we find (by demanding the vanishing of the conformal
anomaly) that the expectation values have to obey the classical
equations of motion. In accord with the harmonic-oscillator example
this does not necessarily imply that we are dealing with a classical
(non-quantum) spacetime. The papers that I cited argue that string
spacetime is instead highly non-commutative on small scales already in
perturbative string theory.
In any case, the analogue of the LQG constraints that encode the
dynamics by picking out physical states from the set of kinematical
states are the "beta-function background equations". This is like the
analogue of the Schroedinger equation for coherent states is the
classical equation of motion for the expectation value.
So the string background certainly is dynamical in this sense and the
dynamics is determined by the theory (the conformal anomaly has to
vanish for consistency). (Even more so than in LQG, by the way, where
the dynamics is actually chosen by hand. I like how Lubos Motl has
argued in his recent papers that what is powerful in LQG is the
kinematics, while what is perhaps rather problematic is the dynamics.
In any case, the interesting results of LQG, like black hole entropy
calculations, rest entirely on the kinematics of the theory. Usually
one cannot even solve the Hamiltonian constraint. If one can,
approximately, the result seems to be problematic: Einstein-Hilbert
dynamics at Plack scale but not on large scales, instead of the other
way round.)
Of course if you look at a single string propagating on some
background, then this background is "fixed" in the sense that it's
equations of motion do not contain the energy momentum of that single
test string (kind of like a Born-Oppenheimer approximation). I think
this is what is meant by saying that string spacetime is
non-dynamical. I do not argue with that. But this is a point that
cannot be compared with LQG, since this can so far not handle
matter-gravity interactions at all.
Fizz Fann
news:<325dbaf1.03030...@posting.google.com>...
> So you're still going to have to learn both of them
> but you should spend more time studying string theory.
That seems to be the general consensus in the physics world.
> It's common in physics for there to be two main rival strategies to
> approaching a problem, and then one becomes successful, and the other
> just sort of withers away. For a while, you have the vast majority of
> physicists studying one main type of theory, and then a small minority
> studying an alternative approach, which I think is good because it's
> good to have more than one way of approaching the problem, but if the
> main way is more successful, you should acknowledge that. QCD explains
> strong force interactions, and Regge theory was an alternative
> approach. Supersymmetry explains the hierarchy problem, and
> technicolor was an alternative approach. Superstring theory is an
> attempt to quantize gravity, and loop quantum gravity is an
> alternative approach. Dark matter explains the rotation of galaxies,
> and MOND was an alternative approach.
So did any if these theories turn out to have the same relationship
that is beng hypothesised to exist between string theory and loop
quantum gravity?
> > So do you mean that most of the people doing string theory or loop
> > quantum gravity use the Many Worlds Idea as well?
> Absolutely not. Very few people ascribe to the Many Worlds Principle.
> Most physicists ascribe to the Copenhagen interpretation.
This is what I learned in physics class, but its still not clear
what the Copenhagen interpretation is because different people
say different things. The idea I got was that the Copenhagen
interpretation had the wavefunction collapse which was a
discontinuous event which might look like a smooth event in some
later, better theory. Is that wrong?
> Modern physics, such as string theory, doesn't solve the problems of
> quantum mechanics, which is from the 1920's. Actually, there really is
> no problem with quantum mechanics. There is no paradox associated with
> Schrodinger's Cat. If you study quantum mechanics, like relativity, it
> makes perfect sense, if you just let go of the classical view of the
> world.
Thats the remark that I want to understand. John Baez said we have
to get over prejudices and change the way we think about quantum
mechanics and now you are saying that I have to let go of my classical
view of the world. Please help me understand this, because people
who use "altenative" interpretations (which seem strange but not
illogical) are saying that the standard thing to do is to put your head
in the sand. I tried to ask John Baez and he said he didn't want to
talk about it. I respect that, but I hope somebody will be willing to
talk and tell me what is the view of the mainstream and whether lots
of people have their head in the sand.
So here's the problem. Bohr says that "their is no quantum world. their
is only an abstract quantum mechanical description". The ieda that I get
from reading things on the web and the Omnes book and so on is that
Copenhagen says that their IS a classical world, and we must not let go of
it, but that the quantum description breaks down (collapse) when the
microscopic systems interact with the classical world. The Copenhagen
interpretation seem to be compleetly in conflict with the view that string
theory can be a theory of everything if it has only a quantum part and no
collapse and no classical part. So I dont see how string theorists can
say they use the Copenhagen interpretation.
> I suggest you better understand the physics of 80 years ago
> before you try to understand the physics of today.
That is what I am trying to do. I have solves the Schrodinger equation
for a few potentials and looked at the ket formalism and so on. Now I
want to see how mainstream people understand it.
In particular, I have a question that mixes relativity and quantum
mechanics and I would like to know what the mainstream answer is.
It is that I can affect whats on Anderomeda by measuring things
here, but I cant control the consequences. Someone on Anderomeda
can see the consequences and do things that have consequences in
my past, but he cant control those consequences either. What is
their to stop my measurements from having consequences in my
past that cause a paradox? Because uncontrolable consequences
is not no consequences. I am happy if somebody can show me
equations for this because I will work through them to understand.
I cant find an answer to this question on the web or anywhere. I
only get Star Trek references when I look for causality loops
with google.
Ilja Schmelzer says there is a preferred frame and I understand
his answer. He sounds sensible. Charles Francis says that things
are even more relative than Einstein said and I am trying to
understand if this helps to solve the problem.
Both of them seem to say that the mainstream people don't have a
good answer. The don't want to talk about it or they deny reality.
I want to see if that's true.
I hope somebody can help.
Alex.
John Baez
Fizz Fann <thephy...@yahoo.com> wrote:
>ba...@galaxy.ucr.edu (John Baez) wrote in message
>news:<b4622j$lr6$1...@glue.ucr.edu>...
>> >Can you say how many loop quantum gravitists or string theorists think
>> >they are likely to be the same theory?
>> Well, they sure aren't the same theory NOW.
>> So I assume you mean: "...are likely to become the same
>> theory when we mess around with them enough to get them to work?"
>Ok, I guess I didn't fully get it before. I had thought that you were
>speculating that they might be something like fractions and decimals -
>two different mathematical systems which describe the same things.
No, certainly not in the current state of these two theories!
>Is it better to say that there are paramaters in the two theories
>that might be tweaked to bring them to the situation like fractions
>and decimals, but at the moment they're different, and the hope is
>that where they can be made to match, we have the right theory?
Holy moly, no - nothing that simple! Let me try to be clear:
Right now these theories look UTTERLY DIFFERENT...
... EXCEPT for a few hints that suggest to the starry-eyed optimists
among us that maybe, just MAYBE, as we tinker around with both theories
in our desperate struggle to get them to work, they may start looking a
bit more similar.
>> >Is there a general consensus that they're converging, [...] ?
>> No, nothing that strong! A better description would be "a few people
>> hope that they might converge if we work at it".
>Ok. I'll join the group of hopers, because if I ever try to understand
>this stuff properly, I'll only have to learn one of them!
Right!
>> Don't forget, there are roughly 10 times as many string theorists
>> as loop quantum graviters. So, loop quantum gravity people feel
>> they need to pay attention to string theory, but not so much vice versa.
>That sounds like a bad situation. The drive to unify the two theories
>is a sociological effect instead of being driven by the science?
It's not *just* a sociological effect; there are lots of purely
scientific reasons to want to unify these two approaches to quantum
gravity. But sociological effects are always important. Physicists
are, after all, people.
Don't forget: physicists are often expected to pull in grants to
get promotions. Especially in the absence of clear experimental
data, this puts physicists under pressure to go along with fashions.
Everyone working on loop quantum gravity knows their life would be
easier if they switched to string theory. If one is too stubborn
to do that, one can still hope that string theory and loop quantum
gravity will merge somehow.
But please, don't think this sociological explanation is the whole
story! There are always *many* overlapping explanations for why
people do what they do.
Actually, I'd rather talk about physics than physicists....
>> If string theory and loop quantum gravity converge, the end result
>> will probably be called string theory, or maybe M-theory, simply
>> because there are more string theorists.
>Ok, so I have a question then. It seems that string theory can tell us
>about quarks and electrons too, but I got the impression (mostly from
>the name) that loop quantum gravity was just about gravity. I take it
>now that I was wrong on this, but I suspect everyone else who isn't
>a specialist is under the same impression.
First of all, string theory doesn't really tell us much about
quarks and electrons that we didn't already know. Not now, at least!
Right now, there are zillions of different string theories - that is,
as far as currently testable predictions go. They probably all merge
at ultra-high energies, but they all say different things about the
particles we see in the lab - and none of them is known for sure to
match what we DO see.
In fact, in a recent paper Lenny Susskind guessed that there may be
on the order of a googolplex of these different theories (technically
called "string theory vacua").
But you're right that all these string theories seem to involve
stuff besides just gravity: various forms of matter.
And you're right that most (but by no means all) work on loop quantum
gravity has focussed on trying to understand how gravity and quantum
mechanics might fit together, without including the complications of matter.
(String theorists think that this can't possibly work, because
*their* theories seem to *need* matter. Loop quantum graviters
retort that string theory needs forms of matter that aren't
like any we've seen! And before you know it, the string theorists
and loop quantum graviters are throwing rocks and bottles at each other;
read old articles on s.p.r. if you want to see that old fight.)
>> >I want to ask a more general question as well, but I add the caveat that
>> >apart from being a physics major, I don't know much about physics! The
>> >question is which of the theories of everything deals the best with the
>> >weird stuff that happens when you measure something quantum
>> >(Schrodinger's cat etc.)?
>> None of the most popular Theories of Everything attempt to
>> deal with this issue. For that, they believe, you need a
>> Theory of Everything Else.
>That also sounds very bad.
Not to me. At least so far, the argument over interpretations of
quantum mechanics has had shockingly little to do with the actual
use of physics to predict the results of experiments.
>Do theorists of everything wake up covered in sweat in the middle
>of the night thinking they're wasting their life?
Of course! Don't you???
Anyone who *never* worries that they're wasting their life, probably is.
>> The reason is that most people working on Theories of Everything
>> believe that these puzzles can be resolved not by new physical
>> theories but by improving how we *think about* quantum mechanics.
>>
>> I share this view.
>I know you said that you don't like to talk about this, so I don't
>want to pester you too much about it, and I'm sorry for sickening
>you with old questions.
Okay. They're very important questions; as soon as I learned
quantum mechanics I got really excited about them, and I thought
about them for years. I'm just into other stuff now. So, you
should definitely keep thinking and reading about this stuff.
>I read through the Omnes book that you
>recommended and it gave me a headache, but I think I see the
>ideas involved - thanks.
When you get a headache like that, it's because you're growing
new neurons, and new connections between existing neurons. So,
it's a good thing.
>> I don't think there is anything abnormal or discontinuous that
>> happens when a measurement is made. I claim that according
>> to quantum mechanics, the laws governing measurements are exactly
>> like the laws governing any other process.
>>
>> A lot of people agree with me... but a bunch don't. You will find
>> that these issues elict endless argument. I'm sick of these arguments,
>> so I won't bother to defend my views. I did plenty of arguing about
>> this in my youth.
>Ok I don't want to attack your views, but just to find out what they
>are, because you mentioned getting over prejudices before would help
>it become clear and that made me interested. I gather from everything
>that I've seen that the two possibilities for not-collapsing views
>are the "Many Worlds" idea and the "Bohmian" one that Ilja Schmelzer
>mentioned before, and that the one you prefer is the many-worlds one.
There are not just *two* possibilities. There are many -
at least one per physicist who doesn't believe in collapse!
Usually more, because most people change their minds. And
even more if the many-worlds interpretation is right.
I certainly don't like the Bohm stuff. I feel this stacks an extra
layer of machinery on quantum mechanics which doesn't help us predict
the results of experiments any better than we could before.
I am much more fond of Everett's work on relative states, and ideas
descending from that, like Griffiths and Omnes' "decoherent histories"
and Zurek's "environmentally induced decoherence". I feel the best of
line of work simply amounts to using the math of quantum mechanics
to predict what we actually observe in various situations where
believers in collapse would have trouble deciding whether or not
the "wave function collapsed".
The "many-worlds" idea is (in my opinion) one of the less clear
ways of following up on Everett's work, because it easily leads one
into false puzzles like "when do worlds split?", "how many are there?",
and the famous "pointer basis problem".
(If you don't understand any of the buzzwords I'm using, you
can probably find them in Omnes' book. I don't have the energy
to explain them here, but they are definitely fun to think about.)
>> But you said that
>> ... most people working on Theories of Everything
>> believe that these puzzles can be resolved not by new physical
>> theories but by improving how we *think about* quantum mechanics.
>So do you mean that most of the people doing string theory or loop
>quantum gravity use the Many Worlds Idea as well?
I don't know. I don't usually ask my colleagues about their
sexual orientation, and I don't usually ask them about their
interpretation of quantum mechanics. For the most part, these
don't visibly affect their work on string theory or loop quantum
gravity.
Oz
Fizz Fann <thephy...@yahoo.com> writes
>It is that I can affect whats on Anderomeda by measuring things
>here, but I cant control the consequences. Someone on Anderomeda
>can see the consequences and do things that have consequences in
>my past, but he cant control those consequences either. What is
>their to stop my measurements from having consequences in my
>past that cause a paradox? Because uncontrolable consequences
>is not no consequences. I am happy if somebody can show me
>equations for this because I will work through them to understand.
Ah, well, if the experts won't give you an answer, I'll give you mine.
Having said that, I'm almost certainly wrong.
I have an excellent record on being wrong.
So best if you don't read this.
I am not quite sure exactly what your problem is.
I am not quite sure why you worry about andromeda in a quantum exchange.
Typically a non-entangled particle interaction on earth shouldn't have
any effect on andromeda. An entangled one doesn't have any effect over
chance unless someone bothers to check it out, in which case they cannot
communicate faster than lightspeed.
I do sometimes wonder if the energy density of the vacuum isn't some
reflection of all the distant wavefunctions of all the particles in the
universe. Be that as it may, it does set some bound on actions that may
happen on earth that might affect andromeda. In effect the noisy vacuum
overwhelms any small signal so it becomes undetectable (even in theory).
An undetectable effect is no effect at all.
Or perhaps you didn't mean any of this at all.
PS Personally I hold we live in a quantum world, and the classical world
is just a good approximation.
--
Oz
This post is worth absolutely nothing and is probably fallacious.
Note: soon (maybe already) only posts via despammed.com will be accepted.
Jeffery
> jeffery...@mail.com (Jeffery) wrote in message
> news:<325dbaf1.03030...@posting.google.com>...
[unnecessary quoted text deleted by moderator]
> > It's common in physics for there to be two main rival strategies to
> > approaching a problem, and then one becomes successful, and the other
> > just sort of withers away. For a while, you have the vast majority of
> > physicists studying one main type of theory, and then a small minority
> > studying an alternative approach, which I think is good because it's
> > good to have more than one way of approaching the problem, but if the
> > main way is more successful, you should acknowledge that. QCD explains
> > strong force interactions, and Regge theory was an alternative
> > approach. Supersymmetry explains the hierarchy problem, and
> > technicolor was an alternative approach. Superstring theory is an
> > attempt to quantize gravity, and loop quantum gravity is an
> > alternative approach. Dark matter explains the rotation of galaxies,
> > and MOND was an alternative approach.
> So did any if these theories turn out to have the same relationship
> that is beng hypothesised to exist between string theory and loop
> quantum gravity?
No, they did not. I don't think string theory and loop quantum gravity
will be unified either, although we may end up using loop quantum
gravity to do non-perturbative calculations that would be very
difficult in string theory.
The history of physics is very complicated. String theory itself
originated in an off-shoot of Regge theory which is now a defunct
pre-QCD way of trying to describe the strong force. I'm sure loop
quamtum theory will contribute to our view of the Universe, but I
predict that basically superstrings and M-theory will become our view
of the Universe, although it will continue to evolve into something
totally different than the current view.
> > > So do you mean that most of the people doing string theory or loop
> > > quantum gravity use the Many Worlds Idea as well?
> > Absolutely not. Very few people ascribe to the Many Worlds Principle.
> > Most physicists ascribe to the Copenhagen interpretation.
> This is what I learned in physics class, but its still not clear
> what the Copenhagen interpretation is because different people
> say different things. The idea I got was that the Copenhagen
> interpretation had the wavefunction collapse which was a
> discontinuous event which might look like a smooth event in some
> later, better theory. Is that wrong?
The Copenhagen interpretation basically says that quantum mechanics is
totally random. For instance which slit an electron goes through in a
double slit experiment in random. Of course you need a discontinuous
event to separate the past from the future. Normally, in physics, you
want theories to be symmetric, but quantum mechanics was so symmetric
that wave functions could propogate both forward and backward in time.
Obviously, you don't want them to travel backwards in time, and so you
you use Green's function to separate the forward and backward
traveling wavefuncdtion. The discontinuity, meaning the observation,
is what distinguishes the past and the future, and you don't want to
get rid of it, because you want to be able to dintinguish between
them, since you don't want the future to affect the past.
> So here's the problem. Bohr says that "there is no quantum world. there
> is only an abstract quantum mechanical description". The idea that I get
> from reading things on the web and the Omnes book and so on is that
> Copenhagen says that their IS a classical world, and we must not let go of
> it, but that the quantum description breaks down (collapse) when the
> microscopic systems interact with the classical world. The Copenhagen
> interpretation seem to be compleetly in conflict with the view that string
> theory can be a theory of everything if it has only a quantum part and no
> collapse and no classical part. So I dont see how string theorists can
> say they use the Copenhagen interpretation.
Obviously, classical physics isn't true, although in daily life, you
can treat it as true for practical purpose. For Newtonian mechanics is
obviously not true, since it doesn't take into account special
relativity, yet we use Newtonian mechanics all the time to solve
practical problems because in daily life when you're traveling much
slower than the speed of light, the error introduced by ignoring
relativistic effects is neglible. Similarly, of course quantum
mechanics is true, and all theories in physics have to include, but
despite that, at the scale of size that we live at, the classical
world seems true to us.
Also, our only information about quantum mechanics which governs the
tiny world of subatomic particles, is the data from experiments.
Therefore we just think up theories to fit the experimental results.
I'm guessing this is what Bohr meant when he said, "there is no
quantum world, there is only a quantum mechanical description",
referring to our theories.
Here you can read different ways of interpreting quantum mechanics.
http://www.upscale.utoronto.ca/PVB/Harrison/BellsTheorem/BellsTheorem.html
Hundreds of huge books have been written on this subject but it
probably can't be resolved experimentally. Even on the above website,
you see how philosophical as opposed to scientific it gets, meaning
you can't resolve the debate between these various interpretations of
quantum mechanics by observation and experiment. But that's not
important. What's important is that quantum mechanics itself
accurately predicts experimental results.
Now another part of the problem is that some people want to make
quantum mechanics as weird as possible but they enjoy the sensation of
being freaked out by quantum weirdness. This is especially done by
writers of popular science books since that's what sells books. I take
the opposite view which is that we should try to make it seem as least
weird as possible. You have to drop some of the assumptions of
classical physics, but then once you do that, it makes absolutely
perfect sense. Then don't even get involved in the philosophical stuff
that can't be resolved by experiment. The main point I would make is
that what "really exists" is the wavefunction, which can be identified
with the probability of detecting the particle. Just imagine that
what's there when you're not measuring is the wavefunction, and not
the particle itself. Despite that, particle physicists go ahead and
imagine them as point particles anyway, because this fact is so
understood by them. However, if you're dwelling on quantum mechanics,
you think in terms of the wavefunction. The observation is simply what
separates the past from the future. I would explain the so-called EPR
paradox by saying that somehow one particle can instantly affect a
distant particle, but this does not violate special relativity because
it would not be possible for a person to send a signal faster than
light.
> > I suggest you better understand the physics of 80 years ago
> > before you try to understand the physics of today.
> That is what I am trying to do. I have solves the Schrodinger equation
> for a few potentials and looked at the ket formalism and so on. Now I
> want to see how mainstream people understand it.
>
> In particular, I have a question that mixes relativity and quantum
> mechanics and I would like to know what the mainstream answer is.
> It is that I can affect whats on Anderomeda by measuring things
> here, but I cant control the consequences. Someone on Anderomeda
> can see the consequences and do things that have consequences in
> my past, but he cant control those consequences either. What is
> their to stop my measurements from having consequences in my
> past that cause a paradox? Because uncontrolable consequences
> is not no consequences. I am happy if somebody can show me
> equations for this because I will work through them to understand.
Nothing ever travels into the past. I don't know where you heard that.
You can do an EPR-type experiment, and however far apart the electrons
are, if you measure one to be spin up, the other will be spin down.
This would be true if one electron is here, and the other is in the
Andromeda galaxy. So what? Nothing ever travels into the past. There
is no paradox. Also, from a practical point of view, there is no way
you could keep two electrons entangled over such an incredibly long
distance.
> I cant find an answer to this question on the web or anywhere. I
> only get Star Trek references when I look for causality loops
> with google.
That should tell you something. There is no such thing as "causality
loops". They exist only in Star Trek, and have nothing to do with
physics. If there's one thing you have to get into your head, there is
no such thing as time travel. It's intrinsically impossible, no matter
what.
> Ilja Schmelzer says there is a preferred frame and I understand
> his answer. He sounds sensible. Charles Francis says that things
> are even more relative than Einstein said and I am trying to
> understand if this helps to solve the problem.
>
> Both of them seem to say that the mainstream people don't have a
> good answer. The don't want to talk about it or they deny reality.
> I want to see if that's true.
I think you should only listen to mainstream people. Earlier, you used
the word "alternative". I hope by that, you didn't mean crackpot.
Don't read or listen to anyone who you suspect is not "mainstream"
because it will only fill your head with nonsense.
Jeffery Winkler
Arnold Neumaier
> Now another part of the problem is that some people want to make
> quantum mechanics as weird as possible but they enjoy the sensation of
> being freaked out by quantum weirdness. This is especially done by
> writers of popular science books since that's what sells books. I take
> the opposite view which is that we should try to make it seem as least
> weird as possible.
Yes. The hallmark of true understanding is that there is nothing
weird left. So to work towards understanding quantum mechanics
means to try to make it seem as little weird as possible.
> You have to drop some of the assumptions of
> classical physics, but then once you do that, it makes absolutely
> perfect sense. Then don't even get involved in the philosophical stuff
> that can't be resolved by experiment. The main point I would make is
> that what "really exists" is the wavefunction,
Actually, what really exists is the density matrix. The wave function
is already a (rank one) approximation to the density matrix that makes
strict sense only at zero temperature, while we all know that the real
world has positive temperature. It is just a useful idealization that
makes some problems more tractable, of the same nature that Fourer
modes make the analysis of periodic functions tractable although real
periodic functions are never exactly sinusoidal. In the same way, one
can decompose density matrices into wave functions (although not
uniquely).
> which can be identified
> with the probability of detecting the particle.
Not quite: Probabilities are squares of absolute values of inner
products of wave functions. In terms of the density matrix rho(x,x'),
one can interpret the diagonal elements rho(x,x) as the particle
density of the corresponding quantum field, which can be interpreted
as the probability density for the result of ideal position
measurements. Not as probability of measuring x, since - due to the
continuous spectrum - the probability of measuring a particular number
is always zero.
Arnold Neumaier
Oz
Arnold Neumaier <Arnold....@univie.ac.at> writes
>Actually, what really exists is the density matrix. The wave function
>is already a (rank one) approximation to the density matrix that makes
>strict sense only at zero temperature,
In a few simple words could you give an idea of the difference between
density functions (whatever they are) and the wavefunction. Obviously
you have stated it above, but a few words of clarification would be
nice.
>while we all know that the real
>world has positive temperature.
Hard to argue with that. I can only imagine that you make some
correction/allowance for disturbance.
>It is just a useful idealization that
>makes some problems more tractable,
--
Tim S
on 11/3/03 6:32 am, Urs Schreiber at Urs.Sc...@uni-essen.de wrote:
>
> "News Admin" <ne...@news.demon.net> wrote in message
> news:<BA8FAAFF...@timsilverman.demon.co.uk>...
>
>> But we'd like the classical 'background' to be dynamic, wouldn't we?
>
> Yes!
>
>> And
>> we'd like the dynamics to emerge from the theory, rather than the other way
>> round, yes?
>
> Yes!
>
>> So putting the background metric in as a boundary condition,
>> though it may give you an interesting insight into its small-scale quantum
>> structure, doesn't really achieve the goal of having a quantum theory of
>> spacetime that reduces to the classical theory on the large scale. It just
>> reduces to particular classical solutions, instead.
>
<snip>
>
> To address your points: First, how is the dynamics of spacetime
> encoded in LQG? Obviously there are the kinematical states, for
> instance being spanned by a basis of spin network states, and on these
> the dynamics is imposed by imposing the various constraints of the
> theory, the quantized version of Einstein's equations.
>
> In string theory one instead finds that the background metric has to
> satisfy the (higher- order corrected) classical equations of motions.
> But I think this is not in contradiction to the fact that perturbative
> string theory defines a non-classical (quantum) spacetime with
> large-scale limit given by the background metric. That's because we
> are dealing with coherent states.
<snip gravitational waves as coherent string states.>
OK, I understand what you're saying (I think!) But...
>
> Of course if you look at a single string propagating on some
> background, then this background is "fixed" in the sense that it's
> equations of motion do not contain the energy momentum of that single
> test string (kind of like a Born-Oppenheimer approximation). I think
> this is what is meant by saying that string spacetime is
> non-dynamical. I do not argue with that. But this is a point that
> cannot be compared with LQG, since this can so far not handle
> matter-gravity interactions at all.
However, in this case, we aren't interested in seeing the string as a piece
of matter; we're only interested in its gravitational aspects. So the
comparison in LQG wouldn't be the gravitational effect of having some matter
around, but rather -- I guess -- the propagation of some kind of excitation
in the 'geometry'. Which I can't be any more precise about because I have no
feel whatsoever for how dynamics works in LQG. :-(
Does this make any sense?
Tim
Fizz Fann
> thephy...@yahoo.com (Fizz Fann) wrote in message
> news:<b2b24cf0.03031...@posting.google.com>...
> > jeffery...@mail.com (Jeffery) wrote in message
> > news:<325dbaf1.03030...@posting.google.com>...
> > So did any if these theories turn out to have the same relationship
> > that is beng hypothesised to exist between string theory and loop
> > quantum gravity?
> No, they did not. I don't think string theory and loop quantum gravity
> will be unified either, although we may end up using loop quantum
> gravity to do non-perturbative calculations that would be very
> difficult in string theory.
Ok. I guess that I'll keep hoping that they are eventually unified, though.
Its a nice idea and it would be a terrible situation to have two
theories of everything which were different but neither could be tested.
> The Copenhagen interpretation basically says that quantum mechanics is
> totally random. For instance which slit an electron goes through in a
> double slit experiment in random.
I'm getting very confused here indeed. Almost everything I've read says
that the way to see it is that the particle goes through _both_ slits
as though it were a wave, but that the radnom bit is where it finally
lands up, when it is a particle. In another thread, Charles Francis told
me that orthodoxy isn't Copenhagen because Copenhagen has the idea
of wave-particle duality and orthodox idea doesn't have that. Would you
agree with him?
> Of course you need a discontinuous
> event to separate the past from the future. Normally, in physics, you
> want theories to be symmetric, but quantum mechanics was so symmetric
> that wave functions could propogate both forward and backward in time.
> Obviously, you don't want them to travel backwards in time, and so you
> you use Green's function to separate the forward and backward
> traveling wavefuncdtion. The discontinuity, meaning the observation,
> is what distinguishes the past and the future, and you don't want to
> get rid of it, because you want to be able to dintinguish between
> them, since you don't want the future to affect the past.
I'm getting even more confused. Some things that Ive found on the
web talk about advanced and retarded Greens functions but I
understand that that is in some relativtistic theory which doesnt
have measurement in it at all! Also I cant find anything that talks
about obervations distinguishing between past and future. And it
does'nt make sense to me because surely there were observations
in the past and there will be more in the future.
> Obviously, classical physics isn't true, although in daily life, you
> can treat it as true for practical purpose. For Newtonian mechanics is
> obviously not true, since it doesn't take into account special
> relativity, yet we use Newtonian mechanics all the time to solve
> practical problems because in daily life when you're traveling much
> slower than the speed of light, the error introduced by ignoring
> relativistic effects is neglible. Similarly, of course quantum
> mechanics is true, and all theories in physics have to include, but
> despite that, at the scale of size that we live at, the classical
> world seems true to us.
Wait. Why is it obvious that classical mechanics isnt true and obvious
that quantum mechanics is true? Why cant there be a future theory that
quantum mechanics is an approximation to, or a statistical version of,
so that the things in quantum mechanics looks like pressures in gases
which is a statistical concept?
> Here you can read different ways of interpreting quantum mechanics.
>
> http://www.upscale.utoronto.ca/PVB/Harrison/BellsTheorem/BellsTheorem.html
Weird. What do people in general think of Bohms idea of the implicate
order?
> Hundreds of huge books have been written on this subject but it
> probably can't be resolved experimentally. Even on the above website,
> you see how philosophical as opposed to scientific it gets, meaning
> you can't resolve the debate between these various interpretations of
> quantum mechanics by observation and experiment. But that's not
> important. What's important is that quantum mechanics itself
> accurately predicts experimental results.
I think so. A lot of the problem is that if you don't let people
talk about measurement and collapse the way lots of people seem
to think, then its hard to say that it predicts any results at all!
That's where the philosophy comes in.
> The main point I would make is
> that what "really exists" is the wavefunction, which can be identified
> with the probability of detecting the particle. Just imagine that
> what's there when you're not measuring is the wavefunction, and not
> the particle itself.
Ok, but then it sounds like I can change what it is that really exists
by measuring something.
> The observation is simply what
> separates the past from the future.
Thats not so simple for me. Can you explain a bit more this idea of
separating the past and the future?
> > In particular, I have a question that mixes relativity and quantum
> > mechanics and I would like to know what the mainstream answer is.
> > It is that I can affect whats on Anderomeda by measuring things
> > here, but I cant control the consequences. Someone on Anderomeda
> > can see the consequences and do things that have consequences in
> > my past, but he cant control those consequences either. What is
> > their to stop my measurements from having consequences in my
> > past that cause a paradox? Because uncontrolable consequences
> > is not no consequences. I am happy if somebody can show me
> > equations for this because I will work through them to understand.
> Nothing ever travels into the past. I don't know where you heard that.
> You can do an EPR-type experiment, and however far apart the electrons
> are, if you measure one to be spin up, the other will be spin down.
> This would be true if one electron is here, and the other is in the
> Andromeda galaxy. So what? Nothing ever travels into the past. There
> is no paradox.
I must be very bad at explaining the question because it seems nobody
except Ilja can understand it, even though I have asked it many times
in many different ways. Let me try again and Ill put points so that
you can stop me if I say something wrong.
1. What I do here on earth can have consequences on Anderomeda right
away in my frame of reference in the sense that what axes I use to
measure affects the results in somebodys notebook on Anderomeda. This
is true even though I cannot deliberately send a signal because I cannot
control the influence I have. (I understand that I cannot deliberately
send a signal and that he cannot ever deduce anything about my decisions
based on his results in his notebook. I have read this many times and
been told it many times and if I am told it again then it means that I
have yet again failed to communicate the question.)
2. The person on Anderomeda can take his notebook afterwards and
accelerate to a different frame of reference. In his new frame of
reference, his notion of simultanity is different and for him
I am now younger than I was when I did the experiment in (1) above.
This is possible because him and me are separated by a "spatial
distance" instead of a "time distance" in special relativity.
3. Young me and old him can do another EPR experiment during which
he affects the result in my notebook. He can use the notebook that
he wrote in (1) in order to decide whether to measure x or y.
4. When I grow up, I can make my choice about which axes to use based
on the results from the EPR experiment that I did when I was young. Or
I can decide to not bother with the second experiment when I grow up
because (for example) I got more ups than downs when I was young.
5. No rule mentioned so far (except Iljas preferred frame) puts
_extra_ constraints on the results of the experiment to prevent me
from getting more ups than downs in the experiment I did when I
was young.
I affect his notebook. He affects my past. Others have objected to
me using the word affect. So I will say it like this. Sometimes,
when I measure x he gets up but if I had measured y instead he would
have gotten down. This is the result of Bells inequalities being
vilated. In those cases, I say that I have affected his results.
> Also, from a practical point of view, there is no way
> you could keep two electrons entangled over such an incredibly long
> distance.
Ok so use a shorter one. Anderomeda was an example because its clearly
a spatial distance. Any other spatial distance is ok too.
> There is no such thing as "causality
> loops". They exist only in Star Trek, and have nothing to do with
> physics. If there's one thing you have to get into your head, there is
> no such thing as time travel. It's intrinsically impossible, no matter
> what.
I am not trying to travel through time. If you tell me that theirs
an extra law which I hadn't heard of before that makes sure in these
cases that I do'nt get more ups than downs then I will be happy. But
I want to know what that law is.
> > Ilja Schmelzer says there is a preferred frame and I understand
> > his answer. He sounds sensible. Charles Francis says that things
> > are even more relative than Einstein said and I am trying to
> > understand if this helps to solve the problem.
> >
> > Both of them seem to say that the mainstream people don't have a
> > good answer. The don't want to talk about it or they deny reality.
> > I want to see if that's true.
> I think you should only listen to mainstream people. Earlier, you used
> the word "alternative". I hope by that, you didn't mean crackpot.
> Don't read or listen to anyone who you suspect is not "mainstream"
> because it will only fill your head with nonsense.
Maybe they will, but so far, only a person who has admitted being not
mainstream has understood the question and given me an answer. The
answer is strange and maybe he put nonsense in my head, but until
somebody else can answer the question then it seems like the only
answer is also the best answer.
Regards,
Alex.
Gene Partlow
> what happens in string theory with the background fields is completely
> analogous. They are coherent states of gravitons (and possibly other
> quanta)
Hi,
I couldn't resist jumping back into this string/LQG theme, because
even though I know about as much about them as Curly (soytenly!),
you touched on a favorite idea I had posted about way back and got
no results on...namely what are the properties of a field of COHERENT
gravitons? Are they qualitatively different from a presumably in-
coherent field of gravitons? Especially, what would be the motion
of a test particle which found itself in such a coherent field?
Just thought I'd meddle some more,
Gene
Fizz Fann
news:<Zt2ZDaE9...@btopenworld.com>...
> Fizz Fann <thephy...@yahoo.com> writes
> >It is that I can affect whats on Andromeda by measuring things
> >here, but I cant control the consequences. Someone on Andromeda
> >can see the consequences and do things that have consequences in
> >my past, but he cant control those consequences either. What is
> >their to stop my measurements from having consequences in my
> >past that cause a paradox? Because uncontrolable consequences
> >is not no consequences. I am happy if somebody can show me
> >equations for this because I will work through them to understand.
> Ah, well, if the experts won't give you an answer, I'll give you mine.
> Having said that, I'm almost certainly wrong.
> I have an excellent record on being wrong.
> So best if you don't read this.
:) I couldnt resist. Its fun to talk and think about this.
> Typically a non-entangled particle interaction on earth shouldn't have
> any effect on andromeda. An entangled one doesn't have any effect over
> chance unless someone bothers to check it out, in which case they cannot
> communicate faster than lightspeed.
Here is the problem. Everyone dismisses the effect because its not
controllable. It like saying that a blind man cant shoot anyone because
he cant see what he is doing. But he can pull the trigger and it does
something -something different from what would happen if he didn't. He
cant control it but that does'nt mean that he doesn't do anything. What
I mean is that when I choose to measure along some axes, the person on
Andromeda gets a result which would be different if I had made a
different choice.
Heres a different way of saynig it. I cant control the effect that I have,
but I could if I knew in advance what result I would get if I measured
along (say) the x-axes. I know I don't know this, but if I did then I could
do something like this. If I know I would get up then I will measure along
the x-axes and he will get down because he always measures with the x-axes.
If I know I would get down, I will measure along the y-axes and then he
has a fifty fifty chance of getting down. These are predictions of quantum
mechanics. Overall he gets down three fourths of the time and he
always measures along the x-axes. That is signalling faster than light.
The reason I cant do it is because I lack knowledge and am unlikely to
guess the right combination of measuring x or y. It is not because there
is no effect or an effect that we should pretend isn't there.
Now we must admit that although it is unlikely for me to guess the right
sequence by chance, it is still possible. And it is possible for him
to accelerate to some high speed where his relativity frame makes him
simultaneous with my past. And then it is possible but unlikely for him
to guess the right sequence of combinations to send young me a signal,
and the signal that he sends can tell my younger self about the signal
that he received from me in his past which is my future. And if that
happens, my younger self can decide to cause a paradox by not sending
him a signal when he grow up.
The thing is that from what Ive been told by everone except Ilja
Schmelzer who says their is a preferred frame, this paradox is not
impossible but just very unlikely. In order to stop this happening,
their must be some extra law or something like that which nobody
has mentioned so far. Ilja says this is obvious but nobody else
sees the problem. Instead I am told to change the way I think or
not to talk about it or to let go of my prejudices.
> I do sometimes wonder if the energy density of the vacuum isn't some
> reflection of all the distant wavefunctions of all the particles in the
> universe. Be that as it may, it does set some bound on actions that may
> happen on earth that might affect andromeda. In effect the noisy vacuum
> overwhelms any small signal so it becomes undetectable (even in theory).
I just said Andromeda because it is faraway. Maybe the particles would
break down or bump off something before they get their but it doesnt
matter. Jupiter would work just as well.
> An undetectable effect is no effect at all.
Yes, but the person on Andromeda can set off a nuclear bomb
when he gets a particular result and that is very detectable.
> PS Personally I hold we live in a quantum world, and the classical world
> is just a good approximation.
Certainly I agree that their are quantum effects and that classical physics
does not explain them. If you are saying more than that then I do not know
what it is that you are saying.
Alex.
[Moderator's note: spelling of "Andromeda" corrected throughout. - jb]
Ilja Schmelzer
>> Both of them seem to say that the mainstream people don't have a
>> good answer. The don't want to talk about it or they deny reality.
>> I want to see if that's true.
> I think you should only listen to mainstream people.
> Don't read or listen to anyone who you suspect is not "mainstream"
> because it will only fill your head with nonsense.
This is a reasonable recommendation for laymen.
On the other hand, where is something going wrong if mainstream
scientists do the same. If alternatives to some mainstream positions
are rejected not because somebody has shown their flaws, but simply
because of ignorance.
Ilja
--
I. Schmelzer, <il...@ilja-schmelzer.net> , http://ilja-schmelzer.net
Ralph Hartley
Jeffery wrote:
> Here you can read different ways of interpreting quantum mechanics.
>
> http://www.upscale.utoronto.ca/PVB/Harrison/BellsTheorem/BellsTheorem.html
>
> Hundreds of huge books have been written on this subject but it
> probably can't be resolved experimentally.
Let's call interpretations of QM that cannot be distinguished by experiment
"proper" interpretations (as opposed to actual modifications of QM a la
Penrose).
The choice between proper interpretations is a matter of taste (and/or
philosophy). I would no more try to win someone over to the "Everett"
interpretation (which I prefer most of the time) than I would try to
convince them to share my preference for the color green.
Nor need anyone choose once and for all time. There are some things I would
never paint green, and for some purposes I *will* say that the wavefunction
collapses.
> I would explain the so-called EPR
> paradox by saying that somehow one particle can instantly affect a
> distant particle, but this does not violate special relativity because
> it would not be possible for a person to send a signal faster than
> light.
The key to avoiding a paradox is that if my measurement can be said to
"affect" something that happens at a distant point, it can be said with
*equal* accuracy that the event at the distant point was "affecting" me.
There is no way to tell which is "cause" and which "affect"
So if I "affect" something in a distant galaxy, and something there
"affects" something in my past, I can avoid the paradox by describing as
both me and the event in my past "affecting" something in the distant
galaxy. Someone else (who e.g. sees the wavefuction collapse as
instantainious in *his* frame) might say that the distant event "affected"
both me and my past. But there is no single frame in which instantaious
wavefunction collapse implies that I am "affecting" my past.
I put scare quotes on "affect" because is some interpretations no such
thing happens at all (the "affect" is not objectively observable, only the
correlation is).
> Also, from a practical point of view, there is no way
> you could keep two electrons entangled over such an incredibly long
> distance.
It's not *that* much harder than communicating at all.
If you give me a means of sending photons for some distance, with a
reasonable chance of receiving them, I'll give you a way to entangle
electrons over the same distance. (Radio wouldn't do, you can't detect the
individual photons.)
Photons keep their polarization over very long distances. Given enough
photon pairs each with even a little entanglement, and a classical
communication channel, the entanglement can be "purified" and an electron
spin can be "teleported" (Yes, it is a bad name for the process, but I
think we're stuck with it).
Ralph Hartley
John Baez
Jeffery <jeffery...@mail.com> wrote:
> > It's common in physics for there to be two main rival strategies to
> > approaching a problem, and then one becomes successful, and the other
> > just sort of withers away.
I don't think this is a very helpful way of viewing the history of
physics if we want to guess what might happen in quantum gravity.
I prefer your other description:
Jeffery <jeffery...@mail.com> wrote:
>The history of physics is very complicated.
Here's what I think:
I think we don't understand quantum gravity very well yet. Neither
string theory nor loop quantum gravity have generated a single verified
experimental prediction. Both are enmeshed in deep difficulties.
For example, if you read Lenny Susskind's new paper about the
anthropic principle in string theory (hep-th/0302219), I guarantee
your hair will stand on end. A googolplex of different low-energy
theories, all of them describing different bubbles of metastable
"fake vacuum", each of which looks like a different exponentially
expanding universe with its own laws until the fake vacuum decays? Aargh!
Is this the best we can do? I hope not. I hope we're just
fumbling around in the initial stages of understanding quantum gravity.
So, if we want to guess what might happen with this subject, I think it's
good to think about other episodes in physics where people were struggling
to understand difficult things. These episodes are very complicated and
interesting - much more so than a simple "battle between two theories"
where one eventually wins.
A good example is the theory of hadrons. At some point Zweig came
up with the idea that these particles were made of "aces". Gell-Mann
said they were made of "quarks" - but only in an abstract mathematical
sense; he was careful to avoid saying quarks were actual particles.
Feynman said hadrons were made of "partons", which he thought of as
actual particles. All these ideas contained a bit of the truth...
but they are all quite primitive compared to modern QCD.
To add to the complication, there was also Regge theory, an attempt to
understand the scattering of hadrons without assuming they had constituents.
In an attempt to get a nice formula for the S-matrix, Veneziano stumbled
into string theory. This later became a theory for how a quark and
antiquark were stuck together by an elastic piece of string to form
a meson. This also has a piece of the truth, since the string acts
a bit like a flux tube of the strong force in the QCD model of confinement.
There was also the Skyrme model, which described mesons as topological
solitons... and probably lots of other things I'm forgetting right now.
In short: a big interesting stew, from which the truth eventually
emerged. That's how I see work on quantum gravity now.
Fizz Fann
ba...@galaxy.ucr.edu (John Baez) wrote in message news:<b4m5ee$ef6$1...@glue.ucr.edu>...
> In article <b2b24cf0.03030...@posting.google.com>,
> Fizz Fann <thephy...@yahoo.com> wrote:
>
> >ba...@galaxy.ucr.edu (John Baez) wrote in message
> >news:<b4622j$lr6$1...@glue.ucr.edu>...
>
>
> ... EXCEPT for a few hints that suggest to the starry-eyed optimists
> among us that maybe, just MAYBE, as we tinker around with both theories
> in our desperate struggle to get them to work, they may start looking a
> bit more similar.
Ok. I get it. I'm told that has never happened with two different
theories before, though, so I guess it would be a surprising thing.
> >That sounds like a bad situation. The drive to unify the two theories
> >is a sociological effect instead of being driven by the science?
>
> It's not *just* a sociological effect; there are lots of purely
> scientific reasons to want to unify these two approaches to quantum
> gravity. But sociological effects are always important. Physicists
> are, after all, people.
>
> ... But please, don't think this sociological explanation is the whole
> story! There are always *many* overlapping explanations for why
> people do what they do.
Ok. I've looked at a few posts from before and it looks like string
theorists behave as though there theory has already been proved to
be right even though it hasnt. Do you think that that is because
they get more funding and fame ect than loop quantum gravitists do?
And do you think LQG people would behave the way string theorists
do now if they got more funding than string theory?
> Actually, I'd rather talk about physics than physicists....
Ok. Could'nt resist.
> First of all, string theory doesn't really tell us much about
> quarks and electrons that we didn't already know. Not now, at least!
> Right now, there are zillions of different string theories - that is,
> as far as currently testable predictions go. They probably all merge
> at ultra-high energies, but they all say different things about the
> particles we see in the lab - and none of them is known for sure to
> match what we DO see.
So that makes string theory more testable than LQG? Because we can
at least check if it gives us electrons that are smaller than camels.
> In fact, in a recent paper Lenny Susskind guessed that there may be
> on the order of a googolplex of these different theories (technically
> called "string theory vacua").
So each vacua gives a different string theory. Could it be that one or
more of these things is loop quantum gravity? Or looks like it after
some transformation?
> (... And before you know it, the string theorists
> and loop quantum graviters are throwing rocks and bottles at each other;
> read old articles on s.p.r. if you want to see that old fight.)
Thats no way for respectable scientists to behave. :)
> >> None of the most popular Theories of Everything attempt to
> >> deal with this issue. For that, they believe, you need a
> >> Theory of Everything Else.
>
> >That also sounds very bad.
>
> Not to me. At least so far, the argument over interpretations of
> quantum mechanics has had shockingly little to do with the actual
> use of physics to predict the results of experiments.
But it might at some point. Like the way thinking about relativity
of motion is a philosophical thing at small speeds but a physical thing
at big speeds. Maybe when electric feild is strong the interpretations
of quantum mechanics are different. Im not saying that its like that,
of course. Just maybe.
> >Do theorists of everything wake up covered in sweat in the middle
> >of the night thinking they're wasting their life?
>
> Of course! Don't you???
Well I think I might more if I spent my life trying to understand the
world but knew I had made an assumption that could easily be wrong and
based everything else on that.
On the other hand, Ive heard that string theory and LQG are worth
studying as mathematics even if their not real physics. So its a
shame to wake up in sweat thinking that your wasting your life
helping one subject instead of another. Its like waking up in sweat
thinking that maybe the wrong orphan got your charity donation.
> When you get a headache like that, it's because you're growing
> new neurons, and new connections between existing neurons. So,
> it's a good thing.
:) I've had lots more since then.
> >Ok I don't want to attack your views, but just to find out what they
> >are, because you mentioned getting over prejudices before would help
> >it become clear and that made me interested. I gather from everything
> >that I've seen that the two possibilities for not-collapsing views
> >are the "Many Worlds" idea and the "Bohmian" one that Ilja Schmelzer
> >mentioned before, and that the one you prefer is the many-worlds one.
>
> There are not just *two* possibilities. There are many -
> at least one per physicist who doesn't believe in collapse!
> Usually more, because most people change their minds. And
> even more if the many-worlds interpretation is right.
You mean that the physicists in other worlds will have different
interpretations? Will there be a googolplex of them? :)
> I certainly don't like the Bohm stuff. I feel this stacks an extra
> layer of machinery on quantum mechanics which doesn't help us predict
> the results of experiments any better than we could before.
I still have to learn more about Bohm. It seems that lots of people
treated him unfairly and tried to suppress his ideas. Aparrently
Oppenheimer said that if we cant disprove him we have to agree to
ignore him and thats what happenned. Sociology at work!
> I am much more fond of Everett's work on relative states, and ideas
> descending from that, like Griffiths and Omnes' "decoherent histories"
> and Zurek's "environmentally induced decoherence". I feel the best of
> line of work simply amounts to using the math of quantum mechanics
> to predict what we actually observe in various situations where
> believers in collapse would have trouble deciding whether or not
> the "wave function collapsed".
Ok. I got more headaches from reading this stuff about histories and
decoherence. Theirs something I dont understand about these ideas
though which hopefully somebody can tell me if John does'nt want to
(everybody's been good and patient with me so far - thanks.) The
question is that the theory says that you can say that certain operators
have well-defined values in families and histories and so on. The
operators are the observables of normal quantum theory I understand.
But in normal quantum theory they are defined in terms of measurement.
They have eigenvalues which are the results of the measurement and
the equation X|psi>=x|psi>. And the eigenvectors are the ones which
are the state of the system after measurement when that eigenvalue
was gotten during the measurement. But the measurement is the collapse
and these theories of decoherent histories have no collapse and no
measurement. So what do the operators and observables mean and how
did they get into the theory?
If I see the mathematics right, then the only things in a non-collapse
theory are the hilbert space and the state vector and the hamiltonian
observable. I think the Zurek idea has the enviroment to do something
like collapse so you can have other operators but the decoherence
theories do'nt have something like that that I can see.
> The "many-worlds" idea is (in my opinion) one of the less clear
> ways of following up on Everett's work, because it easily leads one
> into false puzzles like "when do worlds split?", "how many are there?",
> and the famous "pointer basis problem".
I think these are problems too. Somebody mentioned that perhaps the
theory you like is Saul Youssef's theory of quantum probability. Do
you think that that theory is compatible with the decoherence theories?
> >So do you mean that most of the people doing string theory or loop
> >quantum gravity use the Many Worlds Idea as well?
>
> I don't know. I don't usually ask my colleagues about their
> sexual orientation, and I don't usually ask them about their
> interpretation of quantum mechanics. For the most part, these
> don't visibly affect their work on string theory or loop quantum
> gravity.
Is it rude to ask somebody about their interpretation of quantum
mechanics? If it is I am sorry if I offended anybody. But I thought
that my questions were about science and not about sex. :)
Thanks,
Alex.
Arnold Neumaier
Oz wrote:
>
> Arnold Neumaier <Arnold....@univie.ac.at> writes
>
> >Actually, what really exists is the density matrix. The wave function
> >is already a (rank one) approximation to the density matrix that makes
> >strict sense only at zero temperature,
>
> In a few simple words could you give an idea of the difference between
> density functions (whatever they are) and the wavefunction. Obviously
> you have stated it above, but a few words of clarification would be
> nice.
For a single particle, a density matrix is an integral operator
rho with kernel rho(x,x'), so that it acts on wave functions psi as
(rho psi)(x)= \int dx' rho(x,x') psi(x'),
with the properties that
<psi|rho|psi> >= 0 for all psi
trace rho =
Examples are the rank 1 density matrices defined by
rho(x,x')=phi(x)phi(x')^* with <phi|phi>=1
which have the property
<psi|rho|psi>=|<phi|psi>|^2 >= 0.
This gives the connection to the squared probability amplitudes,
and shows that rank 1 density matrices describe pure states.
Note that a phase in phi cancels, so that rho is unambiguous
while the wave function is defined only up to a constant
complex factor of absolute value 1.
Arbitrary convex linear combinations of density matrices are also
density matrices; they correspond to so-called mixed states.
All states in statistical physics are mixed states; pure states
only figure in a Fourier decomposition into eigenmodes of the
Hamiltonian, and are convenient abstractions. As the temperature
goes to zero, the canonical states of statistical mechanics
converge (in the nondegenerate case) to a rank 1 density matrix
formed from the ground state of the Hamiltonian.
You can read about this in books about (quantum) statistical
mechanics. I'd recommend the book
L.E. Reichl, A Modern Course in Statistical Physics,
2nd ed., Wiley 1998. (But the first edition is ok, too).
It gives a very broad view of statistical physics and its applicatiosn.
Arnold Neumaier
Urs Schreiber
Tim S schrieb:
> <snip gravitational waves as coherent string states.>
Actually it need not be waves. A black hole would also be a coherent
string state, for instance.
> However, in this case, we aren't interested in seeing the string as a piece
> of matter; we're only interested in its gravitational aspects. So the
> comparison in LQG wouldn't be the gravitational effect of having some matter
> around, but rather -- I guess -- the propagation of some kind of excitation
> in the 'geometry'.
I think you are saying that outside any matter distribution spacetime is
vaccuum spacetime. That's in a sense why you can study BHs in LQG
without describing "matter".
Does LQG even say anything about the interior inside the event horizon?
The singularity? The texts that I have seen treat the horizon as a
boundary of the spacetime being described, iirc.
BTW, one can in principle vary the metric in the string's sigma-model.
People studying cosmic strings do that all the time, obtaining an
energy-momentum tensor of the cosmic string which couples to GR. By
related methods in hep-th/9907030 the back-reaction of a highly excited
fundamental string on the gravitational field had been estimated. The
result is that the string contracts under its own gravity thus doing
away with an apparent paradox in earlier calculations of the
string/black-hole correspondence.
This are of course classical, or at best semi-classical, calculations.
This raises the question: What happens when we quantize the
beta-function background equations of the string?
In Polchinski II, p. 259 it says:
"[...] conformally invariant [worldsheet] theories correspond to string
backgrounds that satisfy the classical equations of motion. One might
then guess that the proper setting for quantum string theory would be a
path integral over all background field configurations - that is, over
all two-dimensional quantum field theories. This last is more
speculative [...]"
I remeber that somebody once mentioned on spr how CFTs can be understood
as vacua in the space of all field theories, or something like that. But
I cannot find the posting right now.
Oz
Fizz Fann <thephy...@yahoo.com> writes
>>Oz:
>> Typically a non-entangled particle interaction on earth shouldn't have
>> any effect on andromeda. An entangled one doesn't have any effect over
>> chance unless someone bothers to check it out, in which case they cannot
>> communicate faster than lightspeed.
>
>Here is the problem. Everyone dismisses the effect because its not
>controllable. It like saying that a blind man cant shoot anyone because
>he cant see what he is doing. But he can pull the trigger and it does
>something -something different from what would happen if he didn't. He
>cant control it but that does'nt mean that he doesn't do anything. What
>I mean is that when I choose to measure along some axes, the person on
>Andromeda gets a result which would be different if I had made a
>different choice.
And vice-versa of course. So what? You have to measure it either up or
down, and you will get a result that is either up or down. You will get
that whether or not someone on earth (or andromeda) measured it 'first'.
What you decide to do as a result of the measurement is up to you,
locally.
As to whether someone on earth has 'affected' you, well of course they
have. If you make the decision based on detecting something about a
particle from earth then 'earth' affects you. This is *your* decision.
Particles moving from A to B are surely an expression of A affecting B.
So what?
>Heres a different way of saynig it. I cant control the effect that I have,
>but I could if I knew in advance what result I would get if I measured
>along (say) the x-axes. I know I don't know this, but if I did then I could
>do something like this.
But you don't know. That;s the point.
Your argument thus falls over immediately.
>Now we must admit that although it is unlikely for me to guess the right
>sequence by chance, it is still possible.
So? It's equally likely you get it wrong.
>And it is possible for him
>to accelerate to some high speed where his relativity frame makes him
>simultaneous with my past. And then it is possible but unlikely for him
>to guess the right sequence of combinations to send young me a signal,
>and the signal that he sends can tell my younger self about the signal
>that he received from me in his past which is my future. And if that
>happens, my younger self can decide to cause a paradox by not sending
>him a signal when he grow up.
Unfortunately not. No external information is travelling ftl.
That's the point.
>The thing is that from what Ive been told by everone except Ilja
>Schmelzer who says their is a preferred frame, this paradox is not
>impossible but just very unlikely. In order to stop this happening,
>their must be some extra law or something like that which nobody
>has mentioned so far. Ilja says this is obvious but nobody else
>sees the problem. Instead I am told to change the way I think or
>not to talk about it or to let go of my prejudices.
Ilja is a smart cookie. However I think he is talking about correlations
within the entangled wavefunction, but then I'm often wrong.
>> An undetectable effect is no effect at all.
>
>Yes, but the person on Andromeda can set off a nuclear bomb
>when he gets a particular result and that is very detectable.
Indeed. But unfortunately half the time the particular result is wrong.
>> PS Personally I hold we live in a quantum world, and the classical world
>> is just a good approximation.
>
>Certainly I agree that their are quantum effects and that classical physics
>does not explain them. If you are saying more than that then I do not know
>what it is that you are saying.
That classical results are really macroscopic quantum results.
It happens that macroscopically things look simpler, and are thus easier
to model.
Ilja Schmelzer
> Don't forget: physicists are often expected to pull in grants to
> get promotions. Especially in the absence of clear experimental
> data, this puts physicists under pressure to go along with fashions.
> Everyone working on loop quantum gravity knows their life would be
> easier if they switched to string theory.
That means, something is heavily wrong with current science. If even
the second (in number of scientists) research direction feels such a
pressure for conformity with the leading direction that "everyone
... knows their life would be easier", that's really hard.
Moreover, in a purely theoretical domain, where people need only a PC.
Some conformity pressure may be necessary if you have to spend lots
of taxpayers money for accelerators and so on, which we don't want to
waste for crank science. But in almost pure theory?
Urs Schreiber
> you touched on a favorite idea I had posted about way back and got
> no results on...namely what are the properties of a field of COHERENT
> gravitons? Are they qualitatively different from a presumably in-
> coherent field of gravitons? Especially, what would be the motion
> of a test particle which found itself in such a coherent field?
If you want to study coherent states of gravitons it would be helpful to
do that in a moderately consistent framework like, well, strings for
instance. (I don't know how much sense one can make of coherent graviton
states in canonically quantized gravity.) And in string theory we know
that a coherent graviton state corresponds to a spacetime which
approximates a classical solution of the EFE. Hence, in the appropriate
point-particle limit of strings, the motion of a test particle in such a
background is just the usual geodesic motion with respect to the
corresponding classical metric.
Incoherent graviton background would be modeled in perturbative string
theory by appropriate non-exponentiated insertions of graviton vertex
operators into the string path integral.
Oz
>Ok. I guess that I'll keep hoping that they are eventually unified, though.
>Its a nice idea and it would be a terrible situation to have two
>theories of everything which were different but neither could be tested.
Currently we have two, and we know both are wrong.
So I think two that explain all we currently know correctly would be an
improvement.
Urs Schreiber
> So each vacua gives a different string theory. Could it be that one or
> more of these things is loop quantum gravity? Or looks like it after
> some transformation?
Note that there is, for historical reasons, a little semantical
ambiguity with the word "theory" when strings are involved. The
googleplex of "theories" that Susskind talks about are really "different
vacua" of a single underlying dynamics. Think of how FRW cosmologies for
different values of the curvature parameter k are not really different
theories but different solutions to one underlying dynamics. However, to
somebody not focusing on cosmological dynamics but, say, on field theory
on a fixed cosmological background, the different k-values may rather
appear to correspond to different theories. Something similar happens in
string theory: For a cosmologist it may appear as if contexts with, for
instance, different numbers of supersymmetries, dimensions, cosmological
constants, types of elementary particles and gauge interactions, etc.
are different theories of the universe. But from the string perspective
(if it should be, approximately, correct) these just correspond to
different points in the solution space of the underlying single
"theory".
So to address your question: I think it is unlikely that, even if LQG
and strings are someday recognized as two aspects of the same thing,
that then their relation would be such that LQG applies in some of the
string vacua and not in others. Rather, by design, LQG would be expected
to be related to the non-perturbative aspects of the underlying theory.
Jeffery
> I'm getting very confused here indeed. Almost everything I've read says
> that the way to see it is that the particle goes through _both_ slits
> as though it were a wave, but that the radnom bit is where it finally
> lands up, when it is a particle. In another thread, Charles Francis told
> me that orthodoxy isn't Copenhagen because Copenhagen has the idea
> of wave-particle duality and orthodox idea doesn't have that. Would you
> agree with him?
If you want to say that one electron goes through both slits, go
ahead, but that's a somewhat misleading way of phrasing it. I meant
which slit it ended up looking like it went through is random.
>
> I'm getting even more confused. Some things that Ive found on the
> web talk about advanced and retarded Greens functions but I
> understand that that is in some relativtistic theory which doesnt
> have measurement in it at all! Also I cant find anything that talks
> about obervations distinguishing between past and future. And it
> does'nt make sense to me because surely there were observations
> in the past and there will be more in the future.
The reason you use the Green function is to make sure it only travels
forward in time.
>>
> Wait. Why is it obvious that classical mechanics isnt true and obvious
> that quantum mechanics is true? Why cant there be a future theory that
> quantum mechanics is an approximation to, or a statistical version of,
> so that the things in quantum mechanics looks like pressures in gases
> which is a statistical concept?
We have overwhelming experimental evidence for quantum mechanics.
> Thats not so simple for me. Can you explain a bit more this idea of
> separating the past and the future?
Because the future can't affect the past.
>
> 2. The person on Anderomeda can take his notebook afterwards and
> accelerate to a different frame of reference. In his new frame of
> reference, his notion of simultanity is different and for him
> I am now younger than I was when I did the experiment in (1) above.
> This is possible because him and me are separated by a "spatial
> distance" instead of a "time distance" in special relativity.
This is wrong. Both you and your friend are traveling into the future.
You may disagee about whether two events are simultaneous but your
present self is never going to be within his past light cone. Your
former self might fall within his past light cone but your present
self never will. You can look at the problem in any reference frame
you want to, but your present self and his present self will always be
on the same horizontal line in the Minkowski diagram. From the way you
phrase it, I'm almost wondering if this is some garbled
misunderstanding of the famous so-called twin paradox. When people say
that one twin accelerates away from Earth, and the stay at home twin
is "younger", they just mean they've aged less, and look younger than
the other twin, from the point of view of the other twin. They are not
somehow in the past of the other twin.
Jeffery
Arnold Neumaier
Andy Young wrote:
>
> On 15 Mar 2003, Arnold Neumaier wrote:
> > Actually, what really exists is the density matrix. The wave function
> > is already a (rank one) approximation to the density matrix that makes
> > strict sense only at zero temperature, while we all know that the real
> > world has positive temperature. It is just a useful idealization that
> > makes some problems more tractable, of the same nature that Fourer
> > modes make the analysis of periodic functions tractable although real
> > periodic functions are never exactly sinusoidal. In the same way, one
> > can decompose density matrices into wave functions (although not
> > uniquely).
>
> in an ensemble of particles, and seems poorly defined in a quantum system.
In quantum statistical mechanics, an ensemble of particles is nothing
else than a mixed multiparticle state; so there is no contradiction
with your remark.
Equilibrium quantum systems at positive temperature are well-studied
in statistical mechanics; their density matrix is given (for monatomic
gases, say) by
rho=Z^{-1}exp(-kH/T-mu N),
where k is the Boltzmann constant, H is the Hamiltonian, N the number
operator, T the (constant) temperature, mu the (constant) chemical
potential, and Z is a constant (called the partition function) depending
on T and mu. One can expand this density matrix into a mixture of
common eigenstates of H and N, and this is indeed the way to actually
compute the thermodynamic properties of the quantum system.
What you might have had in mind with the statement 'seems poorly defined'
is the nonequilibrium case. Here the temperature becomes a field, and
one finds various (and more or less foggy) views in the literature
about how to make the notion precise. But for applications to small
quantum systems it is always assumed that the system is in contact
with a 'heat bath' which is in equilibrium and hence has a well-defined
constant temperature, and this defines the temperature of the small system.
The dynamics is then given by a Lindblad equation, which is a generalization
of the von Neumann equation for the density matrix, with additional
dissipative terms involving the temperature and modeling the loss of
energy to the heat bath. Unlike the von Neumann equation, the
Lindblad equation does not preserve the rank of a density matrix,
and hence turns even an initially pure state into a mixed state.
Since real systems are always coupled to the environment, which is
at least locally at equilibrium, this means that every small real
quantum system must be described by a density matrix.
There are no problems at all with the definition of temperature in this
situation.
> Although I sympathize with the motivation of the comment (that a pure
> state is some type of an idealization, and real quantum systems are mixed
> states), I disagree with concluding that no pure state could be justly
> defined. A pure state only seems like an idealization because it is
> easier to compute as a linear superposition of other states.
A pure state is always an idealization; but it is very often used because
it is much easier to work with it, and as mentioned above, one can in
many cases reduce the general situation to that of pure states.
But the latter is no longer possible when one studies nonequilibrium
phenomena or when energy loss matters. (In the latter case one can do
_some_ things in a kind of pure states, by giving up instead the
Hermiticity of the Hamiltonian and the definite inner product,
which is done, e.g., to compute resonances.)
Arnold Neumaier
Kevin A. Scaldeferri
Ilja Schmelzer <schm...@wias-berlin.de> wrote:
>Moreover, in a purely theoretical domain, where people need only a PC.
>Some conformity pressure may be necessary if you have to spend lots
>of taxpayers money for accelerators and so on, which we don't want to
>waste for crank science. But in almost pure theory?
A PC(*), a salary, a few grad students and postdocs, a secretary, journal
subscriptions, a travel budget, overhead. It doesn't look too hard for
one professor to need half a million or more a year.
(*) or maybe a $50,000 Sun server
--
======================================================================
Kevin Scaldeferri Calif. Institute of Technology
The INTJ's Prayer:
Lord keep me open to others' ideas, WRONG though they may be.
John Baez
Urs Schreiber <Urs.Sc...@uni-essen.de> wrote:
>Gene Partlow schrieb:
>> you touched on a favorite idea I had posted about way back and got
>> no results on...namely what are the properties of a field of COHERENT
>> gravitons? Are they qualitatively different from a presumably in-
>> coherent field of gravitons? Especially, what would be the motion
>> of a test particle which found itself in such a coherent field?
I could work this out... but only if you paid me.
>If you want to study coherent states of gravitons it would be helpful to
>do that in a moderately consistent framework like, well, strings for
>instance.
I think that's overkill, at least when you're just getting started!
To get going you can just use linearized gravity, in which the
description of a coherent graviton beam is very similar to the
description of a laser beam in quantized Maxwell theory.
Linearized gravity is a free quantum field theory, so you can
describe coherent states using ideas have been standard for
many decades. Briefly, for any solution of the linearized Einstein
equations, there will be a "coherent state" in the Fock space for
gravitons. This describes a Poisson distribution of gravitons,
all having your given solution of the linearized Einstein equations
as their wavefunction. It's just like how people describe a
laser beam - but with gravitons taking the place of photons.
Only when your "gravitational laser beam" is so strong that
the nonlinearities in Einstein's equation become important, will
you need a better description.
For example, the gravitational radiation LIGO is trying to detect
is well described by the linearized approximation - at least by
the time it gets to us.
As for Partlow's original question, unless we take into account
*quantum* aspects of the particle, the motion of a particle in
a coherent graviton beam would be identical to its motion in the
corresponding classical solution of the (linearized) Einstein equations.
So, to get something more exciting, we should consider a quantum system
(e.g. a test particle, or atom) in a coherent graviton beam.
>(I don't know how much sense one can make of coherent graviton
>states in canonically quantized gravity.)
There's a minor industry devoted to just this, because people
want to understand how canonical quantum gravity is related to
the classical theory. Thomas Thiemann has a four-part paper
entitled "Gauge Field Theory Coherent States", Sang Pyo Kim
has a paper on the "Coherent State Representation of Semiclassical
Quantum Gravity", Ashtekar et al have one on "Coherent State
Transforms for Spaces of Connections", and there's also one by
Ashtekar and a different et al called "Quantum gravity, shadow
states, and quantum mechanics".
By the way, about your remark that strings give a "moderately
consistent framework" for quantum gravity, it's worth noting
that so far, superstring theory has been shown perturbatively
renormalizable only up to two loops. I suppose this counts as
"moderately consistent", but not much more.
A review article by Smolin, due to appear on the arXiv today
or tomorrow, will discuss this and many points of comparison
between strings and loops.
Charles Francis
In message <b2b24cf0.03031...@posting.google.com>, Fizz Fann
<thephy...@yahoo.com> writes
>jeffery...@mail.com (Jeffery) wrote in message
>news:<325dbaf1.0303...@posting.google.com>...
>
>> thephy...@yahoo.com (Fizz Fann) wrote in message
>> news:<b2b24cf0.03031...@posting.google.com>...
>
>that the way to see it is that the particle goes through _both_ slits
>as though it were a wave, but that the radnom bit is where it finally
>lands up, when it is a particle. In another thread, Charles Francis told
>me that orthodoxy isn't Copenhagen because Copenhagen has the idea
>of wave-particle duality and orthodox idea doesn't have that. Would you
>agree with him?
I should clarify this. I am using Bub's terminology here, one which I
agree with quite strongly. Many physicists do not distinguish orthodoxy,
or Dirac-Von Neumann, from Copenhagen. In fact Copenhagen contains
something of a confusion of Heisenberg and Bohr's ideas - I once saw it
described as a smorgasbord from which you take the bits you like. The
orthodox Dirac-Von Neumann interpretation was distilled from Copenhagen,
but instead of saying that the particle has wave like property and
somehow goes through both slits, in Dirac-Von Neumann the question
"which slit" becomes meaningless and impossible to ask.
>Ok, but then it sounds like I can change what it is that really exists
>by measuring something.
You can change the relationships found in nature by measuring things. In
QM you cannot help but change those relationships. That is not quite the
same as changing what really exists.
>> > In particular, I have a question that mixes relativity and quantum
>> > mechanics and I would like to know what the mainstream answer is.
>> > It is that I can affect whats on Anderomeda by measuring things
>> > here, but I cant control the consequences.
The mainstream answer is basically that you can't affect things on
Andromeda in any meaningful way. I think most mainstream physicists stop
discussing it at this point, finding nothing else meaningful to say. I
agree with that basic position, but also I admit I find it intensely
dissatisfying.
>> >Someone on Anderomeda
>> > can see the consequences and do things that have consequences in
>> > my past, but he cant control those consequences either. What is
>> > their to stop my measurements from having consequences in my
>> > past that cause a paradox? Because uncontrolable consequences
>> > is not no consequences.
The affects are not "in the past", but "outside the light cone". One way
to explain EPR (a way which most physicists find unacceptable btw,
though I think we should be open to possibilities) is that your
measurement does affect the past, namely the spin orientation at time of
emission, but only in a way which cannot be detected until results from
both experiments are collated. I don't think it this simple, or that
this fits in with the orthodox interpretation of qm which I prefer. I
think that the quantity spin really is meaningless except in the context
of measurement, and hence that it is meaningless to talk of spin at time
of emission. This is a mainstream answer. On the other hand it is
profoundly dissatisfying, because something is going on and we cannot
actually describe it properly. That is more or less what Einstein meant
by saying that qm is not complete.
So that is actually the point at which I depart from the mainstream.
Most physicists seem to be satisfied that all that can be said has been
said. I am not. I agree with much of what has been said, but I don't
think that is a good point to stop research.
>I must be very bad at explaining the question because it seems nobody
>except Ilja can understand it, even though I have asked it many times
>in many different ways. Let me try again and Ill put points so that
>you can stop me if I say something wrong.
Again I will interject strange interruptions, both with a view to giving
you an orthodox interpretation, and highlighting how difficult this
stuff is to describe.
>1. What I do here on earth can have consequences on Anderomeda right
>away in my frame of reference in the sense that what axes I use to
>measure affects the results in somebodys notebook on Anderomeda.
Except that you cannot say so. These consequences are outside the light
cone, and therefore they are meaningless to you.
>2. The person on Anderomeda can take his notebook afterwards and
>accelerate to a different frame of reference. In his new frame of
>reference, his notion of simultanity is different and for him
>I am now younger than I was when I did the experiment in (1) above.
>This is possible because him and me are separated by a "spatial
>distance" instead of a "time distance" in special relativity.
Yes, but outside the light cone you cannot make a clear distinction
between past and future, but it doesn't matter because there are no
causal affects outside the light cone.
>I affect his notebook. He affects my past. Others have objected to
>me using the word affect.
I think it is more accurate to object to the word past. Also although
there is a correlation in the results, so that your result affects his
result, because you cannot control your result, you do not affect his
notebook. There is an effect, but you are not the cause of it.
Regards
--
Charles Francis
Tim S
> Tim S schrieb:
>> <snip gravitational waves as coherent string states.>
> Actually it need not be waves. A black hole would also be a coherent
> string state, for instance.
Mm, yes... assuming that this very non-perturbative situation actually makes
sense and can be calculated.
>> However, in this case, we aren't interested in seeing the string as a piece
>> of matter; we're only interested in its gravitational aspects. So the
>> comparison in LQG wouldn't be the gravitational effect of having some matter
>> around, but rather -- I guess -- the propagation of some kind of excitation
>> in the 'geometry'.
> I think you are saying that outside any matter distribution spacetime is
> vacuum spacetime. That's in a sense why you can study BHs in LQG
> without describing "matter".
Actually, it meant I hadn't taken in what you were saying, even though your
words were right there in front of me on the screen. :-) You're absolutely
right that the absence of matter and hence back-reaction is a bit of a sad
lack.
> Does LQG even say anything about the interior inside the event horizon?
> The singularity? The texts that I have seen treat the horizon as a
> boundary of the spacetime being described, iirc.
But isn't that because of what they're trying to do -- i.e. give a
statistical mechanics interpretation of the area-entropy relationship -- so
the event horizon is what they're interested in? There's nothing special
about the interior of a black hole per se, is there?
> BTW, one can in principle vary the metric in the string's sigma-model.
> People studying cosmic strings do that all the time, obtaining an
> energy-momentum tensor of the cosmic string which couples to GR. By
> related methods in hep-th/9907030 the back-reaction of a highly excited
> fundamental string on the gravitational field had been estimated. The
> result is that the string contracts under its own gravity thus doing
> away with an apparent paradox in earlier calculations of the
> string/black-hole correspondence.
>
> This are of course classical, or at best semi-classical, calculations.
> This raises the question: What happens when we quantize the
> beta-function background equations of the string?
Indeed.
> In Polchinski II, p. 259 it says:
>
> "[...] conformally invariant [worldsheet] theories correspond to string
> backgrounds that satisfy the classical equations of motion. One might
> then guess that the proper setting for quantum string theory would be a
> path integral over all background field configurations - that is, over
> all two-dimensional quantum field theories. This last is more
> speculative [...]"
Thus sort of glossing over the whole "quantum gravity" aspect of quantum
gravity. Summing over backgrounds...modulo diffeomorphisms? Subject to what
constraints? Making what contribution to the partition function, exactly?
With what action?
Tim
Fizz Fann
news:<b5502p$6m6$1...@ra.nrl.navy.mil>...
> Jeffery wrote:
> > Hundreds of huge books have been written on this subject but it
> > probably can't be resolved experimentally.
> Let's call interpretations of QM that cannot be distinguished by experiment
> "proper" interpretations (as opposed to actual modifications of QM a la
> Penrose).
>
> The choice between proper interpretations is a matter of taste (and/or
> philosophy).
But it is possible to make demands like that it should not require us
to change the way we think about logic and that it should not be
vague, for example the way it's not clear what a measurement is and
what it isn't. I would say that if the only answer that can be given
to a question about the results in a hypothetical experiment (which
could conceivebly be performed) is "change the way you think until
you stop asking that question" then there's something wrong.
> I would no more try to win someone over to the "Everett"
> interpretation (which I prefer most of the time) than I would try to
> convince them to share my preference for the color green.
Right, but it seems that in this case its more like is Schrodinger's
cat alive or dead. So you might say that you prefer to think its
alive and whether it is or not isn't important - its only a matter
of taste. If you were inside the box looking at the cat
(you are Wigners friend) then someone outside would say you were in
a superposition and its just a matter of his personal preference
to say whether you are actually in a well-defined state (like in
Bohms theory) or whether you are in a superposition.
> Nor need anyone choose once and for all time. There are some things
> I would never paint green, and for some purposes I *will* say that
> the wavefunction collapses.
But do you mean it's like a tool that you use like approximating a
curve with a straight line, when you believe inside that its really
not so?
> > I would explain the so-called EPR
> > paradox by saying that somehow one particle can instantly affect a
> > distant particle, but this does not violate special relativity because
> > it would not be possible for a person to send a signal faster than
> > light.
> The key to avoiding a paradox is that if my measurement can be said to
> "affect" something that happens at a distant point, it can be said with
> *equal* accuracy that the event at the distant point was "affecting" me.
> There is no way to tell which is "cause" and which "affect"
This is the language game again. Dont talk about this and don't say that
and then you can't ask the question. Here is the situation. In order to
explain the results of the experiments and the violation of Bells
inequalities we need to say theres let us say a "relationship" between
my decisions about which axes to measure along and the results of the
faraway experiment. You don't want me to point at one and say its affecting
the other but to be democratic and say their all equal and we shouldn't
blame one for what the other one does. It just happened that way and
neither is cause and neither is effect.
But the problem is that one of the things in this is my decision. And
the other thing is the result of a measurement - an "up" or "down".
You are saying that its just as good to say that the result (up or down)
caused me to make the decision that I did. That it somehow affected
the events in my brain, which might (for all the people on Andromeda
know) be a huge machine made of big cogs and gears doing computations.
The gears and cogs obey classical physics so you do'nt need to use
little correlations to say what they will do. They will turn and
push each other. The results of the way they push each other (which
are my decisions) have a "relationship" to the results of a faraway
quantum measurement. Now I say that it is very wrong to say that
the results of the measurement caused the motion of the cogs to be
affected because the cogs were obeying classical physics. That is,
that we can clearly say that if theirs a relationship then the cogs
are not going to be bossed around by a faraway measurement result
but the measurement result might be bossed around by the cogs. So
it _is_ better to say that my decisions are affecting the faraway
results than the other way around.
> So if I "affect" something in a distant galaxy, and something there
> "affects" something in my past, I can avoid the paradox by describing as
> both me and the event in my past "affecting" something in the distant
> galaxy.
Well I am saying that you can't do that. Decisions that you make might
just as well have been made by a machine which is so big and heavy
that you only use classical physics to describe what it does and then
you cant say that those decisions have been affected by faraway
quantum measurement results. But the events that unfold on those
faraway planets are in a very definite and real sense affected by
the results of the measurements there because the decision to
blow up the planet might be made based on the result of a measurement
like this. That is, your decisions affect their results but their
results dont affect your decisions.
> Someone else (who e.g. sees the wavefuction collapse as
> instantaneous in *his* frame) might say that the distant event "affected"
> both me and my past. But there is no single frame in which instantaneous
> wavefunction collapse implies that I am "affecting" my past.
There is not, but one can point to certain things, like the decisions
of experimenters and say that they are not affected by faraway things.
Then we know what is independent and what is dependent. And the answer
is that your decisions can always be considered to be independent but
the results of the measurements cant. If the correlations are there
then everyone must agree on that fact no matter what frame they are
in. The purpose of mentioning the frames that I did was to draw attention
to those particular correlations. Affecter versus affected can be decided
based on which one is a decision and which one is a measurement result.
So the question remains.
> I put scare quotes on "affect" because is some interpretations no such
> thing happens at all (the "affect" is not objectively observable, only the
> correlation is).
I have the very strong feeling at the moment that the whole orthodox
understanding of quantum mechanics is an unspoken agreement not to
talk or think about it along with the collective hallucination that
there is an orthodox viewpoint which makes sense.
I said in an other thread that I am happy to use the word "consequences"
instead of affect if it bothers people. Wittgenstein said that
people who think about philosophy end up distorting the usage of normal
words (such as affect) until those people start saying very strange things
and need therapy to help them use those words in the normal way again.
I think that has happened with "reality" for quantum physicists too.
Regards,
Alex.
PS. You made me doubt myself for a second. The language game is a
powerful tool for confusing people.
[Moderator's note: I am sure that Hartley is not trying to confuse
Alex, but rather to clarify things. So, let's be nice. - jb]
Fizz Fann
> Fizz Fann <thephy...@yahoo.com> writes
> >Everyone dismisses the effect because it's not
> >controllable. It's like saying that a blind man cant shoot anyone because
> >he cant see what he is doing. But he can pull the trigger and it does
> >something -something different from what would happen if he didn't. He
> >cant control it but that doesn't mean that he doesn't do anything. What
> >I mean is that when I choose to measure along some axes, the person on
> >Andromeda gets a result which would be different if I had made a
> >different choice.
> And vice-versa of course. So what? You have to measure it either up or
> down, and you will get a result that is either up or down. You will get
> that whether or not someone on earth (or andromeda) measured it 'first'.
No no. I measure along x or y. And I throw away the result of my
measurement because I don't care about it. But my choice - x or y -
affects the results that the faraway experimenter gets.
Let me anticipate what you are going to say. That I can't deliberately
send a signal and he cant know anything about what I did based on
the results that he gets. Fine. I've read that a million times by
now and I'm not saying otherwise. But I am saying that even still,
the faraway experimenter would get different results if I made
different choices.
> As to whether someone on earth has 'affected' you, well of course they
> have. If you make the decision based on detecting something about a
> particle from earth then 'earth' affects you. This is *your* decision.
> Particles moving from A to B are surely an expression of A affecting B.
>
> So what?
No. The particles don't go from earth to Andromeda. They are generated
halfway between Earth and Andromeda and one goes to Earth and one goes
to Andromeda. During the entire time the experiment happens, there is
not enough time for anything to get all the way from here to there, but
my decisions here affect what happens there.
> >Here's a different way of saying it. I cant control the effect that I have,
> >but I could if I knew in advance what result I would get if I measured
> >along (say) the x-axes. I know I don't know this, but if I did then I could
> >do something like this.
> But you don't know. That's the point.
> Your argument thus falls over immediately.
No; that your point. My point was different, which was why I said
"I know I don't know this, but ..."
It's the part after the "but" which is the point.
> >Now we must admit that although it is unlikely for me to guess the right
> >sequence by chance, it is still possible.
> So? It's equally likely you get it wrong.
It is overwhelmingly likely that I will get it wrong. But not impossible.
That's the point. According to the theory, it's not impossible. And
if it's possible, I can keep trying it until I eventually succeed.
And if it has a finite probability, then I will eventually succeed. I
will successfully send a signal to a man on Andromeda, and then he can
try to signal into my past and keep trying until _he_ succeeds.
> >And it is possible for him
> >to accelerate to some high speed where his relativity frame makes him
> >simultaneous with my past. And then it is possible but unlikely for him
> >to guess the right sequence of combinations to send young me a signal,
> >and the signal that he sends can tell my younger self about the signal
> >that he received from me in his past which is my future. And if that
> >happens, my younger self can decide to cause a paradox by not sending
> >him a signal when he grow up.
> Unfortunately not. No external information is travelling ftl.
> That's the point.
I have no idea what you mean by "external information". Perhaps you
could say something more enlightening than "Unfortunately not."
For example, which of the sentences does the "Unfortunately not" apply to:
1. it is possible for him
to accelerate to some high speed where his relativity frame makes him
simultaneous with my past.
2. it is possible but unlikely for him
to guess the right sequence of combinations to send young me a signal
3. the signal that he sends can tell my younger self about the signal
that he received from me in his past which is my future
4. if that
happens, my younger self can decide to cause a paradox by not sending
him a signal when he grows up.
> >The thing is that from what Ive been told by everone except Ilja
> >Schmelzer who says their is a preferred frame, this paradox is not
> >impossible but just very unlikely. In order to stop this happening,
> >their must be some extra law or something like that which nobody
> >has mentioned so far. Ilja says this is obvious but nobody else
> >sees the problem. Instead I am told to change the way I think or
> >not to talk about it or to let go of my prejudices.
> Ilja is a smart cookie. However I think he is talking about correlations
> within the entangled wavefunction, but then I'm often wrong.
I think he and I are talking about the same thing. Hes the only person
who understood the question right away. For everyone else I had to
struggle to re-explain it again and again.
> >> An undetectable effect is no effect at all.
> >Yes, but the person on Andromeda can set off a nuclear bomb
> >when he gets a particular result and that is very detectable.
> Indeed. But unfortunately half the time the particular result is wrong.
Right, but some of the time it's right and it's not ok to have a
theory with paradoxes in it just because they only happen some of the
time.
> >> PS Personally I hold we live in a quantum world, and the classical world
> >> is just a good approximation.
> >Certainly I agree that their are quantum effects and that classical physics
> >does not explain them. If you are saying more than that then I do not know
> >what it is that you are saying.
> That classical results are really macroscopic quantum results.
> It happens that macroscopically things look simpler, and are thus easier
> to model.
I don't know what a calssical result or a quantum result is. There are
results to experiments. Sometimes they can be explained with classical
mechanics and sometimes they can be explained by quantum mechanics.
Best wishes,
Alex.
John Baez
Urs Schreiber <Urs.Sc...@uni-essen.de> wrote:
>I think you are saying that outside any matter distribution spacetime is
>vacuum spacetime. That's in a sense why you can study BHs in LQG
>without describing "matter".
Actually, the paper by Ashtekar and Krasnov and myself also treats charged
black holes. For this we use the Einstein-Maxwell equations instead
of the vacuum Einstein equations. We also do black holes coupled
to a dilaton field - not my idea, but some people like to study them,
and it's not much extra work.
>Does LQG even say anything about the interior inside the event horizon?
>The singularity?
Martin Bojowald and Abhay Ashtekar are writing a paper about this.
They don't give a full-fledged loop quantum gravity treatment
of what happens inside a quantum black hole, but they've made
a good step in this direction.
The good news is that the singularity seems to go away -
just like it does in Bojowald's work on the big bang!
>The texts that I have seen treat the horizon as a
>boundary of the spacetime being described, iirc.
In these papers (like the one with Ashtekar, Krasnov and myself)
we are trying to understand the area-entropy relation *without*
understanding the essentially dynamical issue of what happens
when some matter collapses to form a black hole - i.e. whether
a singularity forms or not when we take quantum effects into account.
These dynamical issues are very hard in loop quantum gravity, and
only a few people, like Bojowald, have had the guts and skill to
make progress so far.
>I remember that somebody once mentioned on spr how CFTs can be understood
>as vacua in the space of all field theories, or something like that. But
>I cannot find the posting right now.
It sounds like you're talking about how CFTs (some of them)
describe *string theory* vacua.
Kevin A. Scaldeferri
Fizz Fann <thephy...@yahoo.com> wrote:
>> Unfortunately not. No external information is travelling ftl.
>> That's the point.
>I have no idea what you mean by "external information". Perhaps you
>could say something more enlightening than "Unfortunately not."
>
>For example, which of the sentences does the "Unfortunately not" apply to:
>
>1. it is possible for him
> to accelerate to some high speed where his relativity frame makes him
> simultaneous with my past.
True
>2. it is possible but unlikely for him
> to guess the right sequence of combinations to send young me a signal
False
>3. the signal that he sends can tell my younger self about the signal
> that he received from me in his past which is my future
False
>4. if that
> happens, my younger self can decide to cause a paradox by not sending
> him a signal when he grows up.
False
The light cone (causal) structure is invariant under Lorentz
transformations. If the spacetime does not admit closed timelike
curves (and it seems we are talking about Minkowski space here, so it
doesn't) than no amount of travelling at high speeds will change that.
A.J. Tolland
> A review article by Smolin, due to appear on the arXiv today
> or tomorrow, will discuss this and many points of comparison
> between strings and loops.
This article irritated me. I prefer Steve Carlip's review
article, which struck me as less biased. Two things in particular irked
me:
1) He trumpets the mathematical consistency of Loop Quantum
Gravity, and then mentions (section 5.2, item 8) that "all of the forgoing
results have been extended to quantum GR with the standard matter fields".
This seems like a dubious claim, given that we do NOT know that the
Standard Model QFT is consistent. I don't see how any model containing it
can possibly be known to be mathematically consistent. In fact, I can't
even see why the gravity sector of such a model should be mathematicallly
consistent: If there is coupling between the matter and gravity fields,
then there should be effects analagous to "virtual matter loops".
Can anyone set me straight on this?
2) He elevates background independence to the level of physical
law, without acknowledging the ambiguity of this phrase, and without
acknowledging that it is a principle which will remains open to
experimental study. He apparently means "spacetime background
independence", so why not just say "diffeomorphism invariance of the
dynamics"? Or alternately, why not demand an explanation of the numerical
value of Newton's constant? Is there a reason this number should not be
regarded as a non-dynamical classical field?
--A.J.
Ralph Hartley
> 2. it is possible but unlikely for him
> to guess the right sequence of combinations to send young me a signal
>
> 3. the signal that he sends can tell my younger self about the signal
> that he received from me in his past which is my future
>
> 4. if that
> happens, my younger self can decide to cause a paradox by not sending
> him a signal when he grows up.
By that argument, you get paradoxes even without QM. It works if you
communicate by guessing.
You send him a message by just thinking it. He receives it by correctly
guessing what you thought.
He relays the message to your past self, again just by thinking it. Your
past self receives the message by guessing it.
Your past self decides to cause a paradox by not sending the message.
Of course, it isn't likely the the recipients will correctly guess the
message, but sometimes they will. Who keeps track of it all to prevent a
paradox from happening?
You really *do* have to send information to cause a paradox, and that means
something you control.
Ralph Hartley
From dae...@irje.ucar.edu Fri Mar 21 18:26:22 2003
Return-path: <dae...@irje.ucar.edu>
Envelope-to: ba...@ren.math.uwo.ca
Received: from irje.ucar.edu ([192.52.106.5])
by ren.math.uwo.ca with esmtp (Exim 3.35 #1 (Debian))
id 18wVu1-0001DV-00
for <ba...@ren.math.uwo.ca>; Fri, 21 Mar 2003 18:26:21 -0500
Received: (from daemon@localhost)
by irje.ucar.edu (8.9.1a/8.9.1) id QAA07159
for ba...@ren.math.uwo.ca; Fri, 21 Mar 2003 16:26:20 -0700 (MST)
Received: from mscan3.ucar.edu (mscan3.ucar.edu [128.117.64.125])
by irje.ucar.edu (8.9.1a/8.9.1) with ESMTP id QAA07156
for <physics-...@irje.ucar.edu>; Fri, 21 Mar 2003 16:26:19 -0700 (MST)
Received: by mscan3.ucar.edu (Postfix)
id C9F01DAB74; Fri, 21 Mar 2003 16:26:19 -0700 (MST)
Delivered-To: physics-...@ncar.ucar.edu
Received: from localhost (localhost [127.0.0.1])
by mscan3.ucar.edu (Postfix) with ESMTP id BB973DAB71
for <physics-...@ncar.ucar.edu>; Fri, 21 Mar 2003 16:26:19 -0700 (MST)
Received: from horus.isnic.is (horus.isnic.is [193.4.58.12])
by mscan3.ucar.edu (Postfix) with ESMTP id 00EA4DAB71
for <physics-...@ncar.ucar.edu>; Fri, 21 Mar 2003 16:26:14 -0700 (MST)
Received: from newsmaster.cc.columbia.edu (IDENT:cu4...@newsmaster.cc.columbia.edu [128.59.59.30])
by horus.isnic.is (8.12.8/8.12.8/isnic) with ESMTP id h2LNQC4f059937
for <sci-physic...@moderators.isc.org>; Fri, 21 Mar 2003 23:26:13 GMT
(envelope-from ne...@columbia.edu)
Received: from newsmaster.cc.columbia.edu (localhost [127.0.0.1])
by newsmaster.cc.columbia.edu (8.12.8/8.12.3) with ESMTP id h2LNQAk6029533
for <sci-physic...@moderators.isc.org>; Fri, 21 Mar 2003 18:26:10 -0500 (EST)
Received: (from news@localhost)
by newsmaster.cc.columbia.edu (8.12.8/8.12.3/Submit) id h2LNQ7J1029532
for sci-physic...@moderators.isc.org; Fri, 21 Mar 2003 18:26:07 -0500 (EST)
To: sci-physic...@moderators.isc.org
Path: not-for-mail
From: Peter Woit <wo...@cpw.math.columbia.edu>
Newsgroups: sci.physics.research
Subject: Re: Conceptual overlap between LQG and String/M-theory?
Date: Fri, 21 Mar 2003 18:26:03 -0500
Organization: Columbia University
Lines: 38
Message-ID: <b5g72f$sqq$1...@newsmaster.cc.columbia.edu>
References: <c504f3da.0302...@posting.google.com> <206f2305.03031...@posting.google.com> <c504f3da.03031...@posting.google.com> <3E782E29...@uni-essen.de> <b5dgos$i82$1...@glue.ucr.edu> <Pine.SOL.4.44.0303201739240.6208-100000@pub-708c-8>
NNTP-Posting-Host: park.math.columbia.edu
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Trace: newsmaster.cc.columbia.edu 1048289167 29530 128.59.192.12 (21 Mar 2003 23:26:07 GMT)
X-Complaints-To: postm...@columbia.edu
NNTP-Posting-Date: 21 Mar 2003 23:26:07 GMT
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.3) Gecko/20030313
X-Accept-Language: en-us, en
In-Reply-To: <Pine.SOL.4.44.0303201739240.6208-100000@pub-708c-8>
X-Spam-Status: No, hits=-7.4 required=5.0
tests=EMAIL_ATTRIBUTION,INVALID_MSGID,IN_REP_TO,NOSPAM_INC,
REFERENCES,SPAM_PHRASE_00_01,USER_AGENT,
USER_AGENT_MOZILLA_UA,X_ACCEPT_LANG
version=2.41
Status: OR
A.J. Tolland wrote:
> This article irritated me. I prefer Steve Carlip's review
>article, which struck me as less biased. Two things in particular irked
>me:
>
> 1) He trumpets the mathematical consistency of Loop Quantum
>Gravity, and then mentions (section 5.2, item 8) that "all of the forgoing
>results have been extended to quantum GR with the standard matter fields".
>This seems like a dubious claim, given that we do NOT know that the
>Standard Model QFT is consistent. I don't see how any model containing it
>can possibly be known to be mathematically consistent. In fact, I can't
>even see why the gravity sector of such a model should be mathematicallly
>consistent: If there is coupling between the matter and gravity fields,
>then there should be effects analagous to "virtual matter loops".
> Can anyone set me straight on this?
>
>
His article probably didn't irritate you half as much as he and many
others are
irritated by endlessly seeing people write things like "string theory,
the only known consistent theory of
quantum gravity..." and I think that's the point. I don't remember
seeing anything
in his article about "mathematical consistency" of any theory and he
certainly
isn't claiming that LQG is a mathematically consistent TOE.
The only part of his argument that went into any detail about issues of
consistency
had to do with precisely the often heard "only consistent theory" claim,
which
is based on the belief that string perturbation theory is finite at each
order. It is
not so well-advertised that this until recently was really only known to
one-loop,
now is known to two loops.
Ralph Hartley
> 2. it is possible but unlikely for him
> to guess the right sequence of combinations to send young me a signal
>
> 3. the signal that he sends can tell my younger self about the signal
> that he received from me in his past which is my future
>
> 4. if that
> happens, my younger self can decide to cause a paradox by not sending
> him a signal when he grows up.
By that argument, you get paradoxes even without QM. It works if you
communicate by guessing.
You send him a message by just thinking it. He receives it by correctly
guessing what you thought.
He relays the message to your past self, again just by thinking it. Your
past self receives the message by guessing it.
Your past self decides to cause a paradox by not sending the message.
Of course, it isn't likely the the recipients will correctly guess the
message, but sometimes they will. Who keeps track of it all to prevent a
paradox from happening?
You really *do* have to send information to cause a paradox, and that means
something you control.
Ralph Hartley
[Moderator's note: sorry, in my first attempt to post this
article it had another article attached at the end. - jb]
Peter Woit
> This article irritated me. I prefer Steve Carlip's review
>article, which struck me as less biased. Two things in particular irked
>me:
>
> 1) He trumpets the mathematical consistency of Loop Quantum
>Gravity, and then mentions (section 5.2, item 8) that "all of the forgoing
>results have been extended to quantum GR with the standard matter fields".
>This seems like a dubious claim, given that we do NOT know that the
>Standard Model QFT is consistent. I don't see how any model containing it
>can possibly be known to be mathematically consistent. In fact, I can't
>even see why the gravity sector of such a model should be mathematicallly
>consistent: If there is coupling between the matter and gravity fields,
>then there should be effects analagous to "virtual matter loops".
> Can anyone set me straight on this?
His article probably didn't irritate you half as much as he and many
others are irritated by endlessly seeing people write things like
"string theory, the only known consistent theory of quantum
gravity..." and I think that's the point. I don't remember seeing
anything in his article about "mathematical consistency" of any theory
and he certainly isn't claiming that LQG is a mathematically
consistent TOE.
The only part of his argument that went into any detail about issues
of consistency had to do with precisely the often heard "only
consistent theory" claim, which is based on the belief that string
perturbation theory is finite at each order. It is not so
well-advertised that this until recently was really only known to
one-loop, now is known to two loops.
[Moderator's note: let us focus on physics, rather than comparing
how irritated we are about various things. - jb]
John Baez
In article <b2b24cf0.03031...@posting.google.com>,
Fizz Fann <thephy...@yahoo.com> wrote:
>ba...@galaxy.ucr.edu (John Baez) wrote in message
>news:<b4m5ee$ef6$1...@glue.ucr.edu>...
>> Right now these theories look UTTERLY DIFFERENT...
>>
>> ... EXCEPT for a few hints that suggest to the starry-eyed optimists
>> among us that maybe, just MAYBE, as we tinker around with both theories
>> in our desperate struggle to get them to work, they may start looking a
>> bit more similar.
>Ok. I get it. I'm told that has never happened with two different
>theories before, though, so I guess it would be a surprising thing.
Ever heard of Heisenberg's matrix mechanics versus Schrodinger's
wave mechanics? The two competing approaches to quantum mechanics?
>I've looked at a few posts from before and it looks like string
>theorists behave as though there theory has already been proved to
>be right even though it hasn't.
Only *some* string theorists do this - not nice ones like
Robert Helling and Aaron Bergman.
>Do you think that that is because they get more funding and
>fame etc than loop quantum gravitists do?
Actually, I think some string theorists do this because when
they first got very excited about the theory in the mid-1980s,
they made some very bold claims about how they were close to
having everything all figured out. It's quite hard to back
down from such claims without looking silly and losing grant
money. And the latter is a real issue, because money for
theoretical particle physics has become harder to get ever
since the cancelling of the superconducting supercollider in
1994. Without many new experiments to demonstrate concrete
progress, theorists feel more of a need to boast about how
they *must be* on the right track, to convince themselves and
also the people in charge of the money.
>And do you think LQG people would behave the way string theorists
>do now if they got more funding than string theory?
Probably. Of course, the main thing to do with more money is
hire more people. If loop quantum gravitsts got more money,
there would be people working on loop quantum gravity. Right
now there are roughly 10 times as many string theorists as
loop quantum gravitists: this would account for roughly 10
times as many new results, 10 times as much fame, and 10
times as much boasting even without any difference in how
good the theories are or what the people are like.
>> First of all, string theory doesn't really tell us much about
>> quarks and electrons that we didn't already know. Not now, at least!
>> Right now, there are zillions of different string theories - that is,
>> as far as currently testable predictions go. They probably all merge
>> at ultra-high energies, but they all say different things about the
>> particles we see in the lab - and none of them is known for sure to
>> match what we DO see.
>So that makes string theory more testable than LQG?
Actually, this is what makes string theory rather *hard* to
test. When your theory has lots of adjustable features, it's
hard to test, not easy.
Nonetheless, most string theorists expect supersymmetry to
be restored at energies that are low enough to give some
testable predictions, like a low-mass Higgs, and a fairly
light superparticle. Both these are things that might be
detected at the forthcoming Large Hadron Collider around 2007.
These are really predictions of the minimal supersymmetric
Standard Model, rather than string theory per se. But, they
are more testable than anything loop quantum gravity currently
has to offer - except for the dispersion of light of distant
quasars due to discreteness of spacetime at the Planck scale.
>Because we can at least check if it gives us electrons that are
>smaller than camels.
Actually string theory is so adjustable that some people think if
we can get it to predict electrons at all, we can probably get
their mass to come out however we want. But this is still a big
"if".
>> In fact, in a recent paper Lenny Susskind guessed that there may be
>> on the order of a googolplex of these different theories (technically
>> called "string theory vacua").
>So each vacua gives a different string theory.
["Vacua" is the plural of "vacuum", so you meant "each vacuum".]
The situation is rather subtle:
There are 5 main superstring theories together with something
else called 11-dimensional supergravity. The general belief,
backed up by lots of indirect evidence, is that there's
fundamentally one theory which has all these 6 theories as
limiting cases. People call it "M-theory", but nobody knows
exactly what this theory is. String theory bigshots spend a
lot of time trying to figure this out.
So, people hope these 6 theories merge into one at high
energies. But at low energies - like what we can study with
a particle accelerator - they sort of splinter apart into lots
more *different* theories!
A bit more precisely: the 6 theories I mentioned reduce
to many different quantum field theories at low energies,
basically depending on the shape of spacetime and the values
of certain fields. Each of these different field theories
predicts different sorts of particles that we'd see in the lab.
An allowed choice of the shape of spacetime and the values of
certain fields is called a "string theory vacuum".
So, it's not quite true that "each vacuum gives a different
string theory", as you said.
Instead, each string theory has lots of different vacua.
Or if you're an optimist, there's one M-theory and it has
lots of different vacua. But either way, each different
vacuum gives a different prediction about what sorts of
particles we should see in the lab, and how these particles act.
>Could it be that one or more of these things is loop quantum
>gravity? Or looks like it after some transformation?
People have suggested this possibility: perhaps loop quantum
gravity with a specific choice of matter fields. But the real
answer to your question is that nobody knows.
>> (... And before you know it, the string theorists
>> and loop quantum graviters are throwing rocks and bottles at each other;
>> read old articles on s.p.r. if you want to see that old fight.)
>That's no way for respectable scientists to behave. :)
I take it you haven't spent much time with actual scientists.
>> >Do theorists of everything wake up covered in sweat in the middle
>> >of the night thinking they're wasting their life?
>> Of course! Don't you???
>Well I think I might more if I spent my life trying to understand the
>world but knew I had made an assumption that could easily be wrong and
>based everything else on that.
Well, don't forget: no matter *what* you do, it's possible that
you've made a horrible mistake of some sort and basically wasted
your life. :-)
But this isn't physics anymore, so please don't reply to this comment.
>On the other hand, I've heard that string theory and LQG are worth
>studying as mathematics even if they're not real physics.
Right. That's definitely true. That's why I'm doing this stuff
as part of being a math professor, and that's how I console myself
when I wake up in a sweat thinking that nobody in my lifetime will
ever understand quantum gravity, and I should have spent my life on
something else.
>> The "many-worlds" idea is (in my opinion) one of the less clear
>> ways of following up on Everett's work, because it easily leads one
>> into false puzzles like "when do worlds split?", "how many are there?",
>> and the famous "pointer basis problem".
>I think these are problems too.
Umm, I didn't exactly say they are "problems". I said they
are "false puzzles" - i.e. the result of having your head on backwards.
>Somebody mentioned that perhaps the
>theory you like is Saul Youssef's theory of quantum probability. Do
>you think that that theory is compatible with the decoherence theories?
I never like to answer questions like "Is Dr. X's interpretation
of quantum mechanics compatible with Dr. Y's?" The only way to
answer a question like that is to lock Dr. X and Dr. Y in a room
for a week, force them to discuss this subject, and see if they're
both alive when you open the door.
>> I don't know. I don't usually ask my colleagues about their
>> sexual orientation, and I don't usually ask them about their
>> interpretation of quantum mechanics. For the most part, these
>> don't visibly affect their work on string theory or loop quantum
>> gravity.
>Is it rude to ask somebody about their interpretation of quantum
>mechanics?
Heh. Not exactly, but if you ask real live physicists, they often
won't want to answer this question.
John Baez
In article <b5b1j8$kil$1...@sue.its.caltech.edu>,
Kevin A. Scaldeferri <ke...@sue.its.caltech.edu> wrote:
>In article <i3g1y16...@wias-berlin.de>,
>Ilja Schmelzer <schm...@wias-berlin.de> wrote:
>>Moreover, in a purely theoretical domain, where people need only a PC.
>>Some conformity pressure may be necessary if you have to spend lots
>>of taxpayers money for accelerators and so on, which we don't want to
>>waste for crank science. But in almost pure theory?
>A PC(*), a salary, a few grad students and postdocs, a secretary, journal
>subscriptions, a travel budget, overhead. It doesn't look too hard for
>one professor to need half a million or more a year.
And don't forget the Rolls Royce with a personal chauffeur!
I should emphasize that most of us aren't quite as spoiled as
the hypothetical professor Scaldeferri describes. For example,
I have 6 grad students, but I don't have money to support them:
they have to scrape out a miserable living as teaching assistants.
I don't have money for any postdocs. I don't have a secretary
or journal subscriptions - though of course the math department
and the university do, and someone has to pay for that. I don't
usually apply for grants, so I don't have a travel budget, and
the university doesn't get to skim off "overhead". Instead,
I try to act so cool that people pay for me to visit them.
So far that strategy seems to be working. And remarkably, my
department isn't pressuring me to get grants! This means I can
work on whatever I want, as long as I keep publishing in good
journals, keep getting good teaching evaluations, keep getting
good letters of recommendation from bigshots, keep cranking out
students with PhDs, etcetera.
Still, my life would have been more cushy if I'd switched to
string theory.
But that's okay. I'm perfectly happy being the weirdo I am,
wasting my time on the internet instead of hobnobbing with
Witten. I'm incredibly lucky to have such a fun life. In
particular, it was just dumb luck that I wound up teaching
math rather than physics. The pressure to get grant money
seems much higher in physics - it's often used as a partial
criterion for hiring and promotion. One reason is that there's
more money to get. To run experiments, of course, grants are
necessary. But even in theory, grantsmanship seems more
obligatory in physics than in math. To get grants, you have to
keep selling your research program to your peers (the anonymous
referees of the grant proposal). In a situation where theory
is not checked by experiment, this can lead theorists to work
on subjects that are fashionable, instead of what they really
want.
Urs Schreiber
John Baez schrieb:
>
> In article <3E7703F0...@uni-essen.de>,
> Urs Schreiber <Urs.Sc...@uni-essen.de> wrote:
>
> >I think you are saying that outside any matter distribution spacetime is
> >vacuum spacetime. That's in a sense why you can study BHs in LQG
> >without describing "matter".
>
> Actually, the paper by Ashtekar and Krasnov and myself also treats charged
> black holes. For this we use the Einstein-Maxwell equations instead
> of the vacuum Einstein equations. We also do black holes coupled
> to a dilaton field - not my idea, but some people like to study them,
> and it's not much extra work.
Ah, right. Good that you are there to correct all my false statements!
> >I remember that somebody once mentioned on spr how CFTs can be understood
> >as vacua in the space of all field theories, or something like that. But
> >I cannot find the posting right now.
>
> It sounds like you're talking about how CFTs (some of them)
> describe *string theory* vacua.
Yes, this comment was not supposed to be related to black holes or LQG
or anything, but to how the conformal-ness of CFTs used in string theory
is thought to be a first approximation to the real thing being described
by more general worldsheet field theories. I know that Aaron Bergman or
Robert Helling or somebody else once said something about this, but for
some reason I cannot find it by searching the google archiv.
Oz
Fizz Fann <thephy...@yahoo.com> writes
>No no. I measure along x or y. And I throw away the result of my
>measurement because I don't care about it. But my choice - x or y -
>affects the results that the faraway experimenter gets.
1) Undetectably by him.
2) So does a message saying 'fire the bomb'.
Thomas Larsson
news:<Pine.SOL.4.44.0303201739240.6208-100000@pub-708c-8>...
> This article irritated me. I prefer Steve Carlip's review
>article, which struck me as less biased. Two things in particular irked
>me:
This irritation is understandable, given that Smolin's paper looks like a
declaration of open war, despite the polite language. It is hard to see
how it can be construed otherwise.
> 1) He trumpets the mathematical consistency of Loop Quantum
>Gravity, and then mentions (section 5.2, item 8) that "all of the forgoing
>results have been extended to quantum GR with the standard matter fields".
>This seems like a dubious claim, given that we do NOT know that the
>Standard Model QFT is consistent. I don't see how any model containing it
>can possibly be known to be mathematically consistent.
Nature is experimentally known to behave in a way which to some approximation
is described by the SM. Is this not a reason to expect that some model
containing the SM, or reducing to it in some appropriate limit, is
consistent?
>In fact, I can't
>even see why the gravity sector of such a model should be mathematicallly
>consistent: If there is coupling between the matter and gravity fields,
>then there should be effects analagous to "virtual matter loops".
> Can anyone set me straight on this?
> 2) He elevates background independence to the level of physical
>law, without acknowledging the ambiguity of this phrase, and without
>acknowledging that it is a principle which will remains open to
>experimental study. He apparently means "spacetime background
>independence", so why not just say "diffeomorphism invariance of the
>dynamics"?
What's the difference?
All principles of physics are of course open to experimental study.
However, diff invariance is a very striking feature of Einstein's general
relativity, which has been subjected to experimental tests. In fact, diff
invariance is a principle repeatedly emphasized and discussed at length
by the deepest thinkers on gr-qc, such as Isham and Rovelli. It is seldom
mentioned on hep-th, though.
However, something that constantly confuses me when LQG'ers talk about
diff invariance is that I never know whether they mean spacetime diffs or
just spatial diffs. If LQG is manifestly invariant under spacetime diffs
it is of great interest to me, since it would tautologically mean that it
carries a quantum representation of the spacetime diff group. However, I
fail to see how this can be true, at least in the Hamiltonian formation,
since the splitting of spacetime into space and time introduces a causal
background. This might be less severe than the metric background of string
theory, but it nonetheless violates Einstein's spirit.
Fizz Fann
> Fizz Fann <thephy...@yahoo.com> writes
> >No no. I measure along x or y. And I throw away the result of my
> >measurement because I don't care about it. But my choice - x or y -
> >affects the results that the faraway experimenter gets.
> 1) Undetectably by him.
>
> 2) So does a message saying 'fire the bomb'.
Agreed. What is your point? What does this have to do with my original
question?
Alex.
DickT
> how irritated we are about various things. - jb]
Well, speaking of physics, and LQG, Hanno Sahlmann and Thomas Thiemann
have just proved that the Ashtekar-Isham-Lewandowski representation of
the quantum algebra of LQG is itself the irreducible representation of
an algenbra of bounded operators which they define and compare to the
Heisenberg algebra of quantum mechanics. See gr-qc/0303074. Now
correct me if I'm wrong, but it seems that if they can go ahead and
prove this representation is unique, then LQG will have about the
strongest rigorous math-phys definition since quantum mechanics
itself. No?
Dick Thompson
Lee Smolin
[Moderator's note: I forwarded to Lee Smolin the comments by
Tolland, Woit and Larsson on his review article. Here is his
reply. - jb]
Dear John,
Thanks, please pass the following along to the news group:
I apologise of course if the article irritated anyone; that
was not the intention. But I would ask people to read carefully,
as I was aware of most of these issues and did choose words
carefully so as to say things that are correct. For example:
1) I did try to be very clear to distinguish spatially diffeomorphism
invariant from spacetime diffeomorphism invariant in the detailed list of
claims. I do not claim that solutions to the hamiltonian constraint are
spacetime diffeomorphism invariant, as the full set of constraints
only generate spacetime diffeomorphisms on solutions. On page 30 I
explicitly mention that
" There are also some unresolved issues concerning the role of the four
dimensional diffeomorphism group in the Hamiltonian theory. This
comes into the details of the regularization of the Hamiltonian
constraint and the relationship between the hamiltonian and path
integral quantizations. A set of related issues have to do with the
relationships between the different forms of the quantum hamiltonian
constraint arrived at by different regularization procedures and
different operator orderings."
2) I also clearly distinguish in point 4 that I am quoting results
that are "physical" in a setting where the time gauge is fixed.
3) I also clearly distinguished different kinds of background
independence, that is one reason I distinguished loop quantum gravity I and
II. LQG I is independent of metric and fields but still dependent on
topological and differential structure, while LQG II depends on no fixed
topological or differential structure.
I think I was clear to indicate the relationship between my use of
diffeomorphism invariant and the more general term background
independent.
4) As to matter, I apologise if I was imprecise. No claims about
consistency etc beyond those mentioned explicitly were made.
I stand by the claim, detailed in Thiemann's review and papers
cited there, that all the standard kinds of matter can
be incorporated and that the key results do extend to include
matter.
However, to be sure there is no misunderstanding,
in the next revision I will be more specific and replace claim
8 with the following:
"Matter may be added at to both the hamiltonian and spin foam
formulations. For the hamiltonian theory it is known how to extend
the definition of the spatially diffeomorphism invariant states to
include all the standard kinds of matter fields, including gauge
fields, spinors, scalars and Kalf Ramond fields.
These states are also invariant under ordinary Yang-Mills and
Kalb-Ramond gauge transformations. The forms for the
matter field terms in the hamiltonian constraints are known
precisely. The spin foam models have been extended to include
gauge and spinor degrees of freedom\footnote{To my knowledge whether
loop quantum gravity suffers from the fermion doubling problem is
an open question.}. Inclusion of matter fields does not
affect the finiteness and discreteness of the area and volume
observables. "
Of course there are quantum effects including matter, but the
theory including matter is uv finite because there are no excitations
with wavelength smaller than the planck scale, as described. But
the questions about the existence of a good low energy limit remain,
with and without matter, as I believe I made clear.
Of course, like all hypotheses about physics, the various forms
of background independence and diffeomorphism invariance are subject to
experimental study. I believe I discuss several ways to test them
experimentally.
Finally, my review is in no way a declaration of war. The language is
polite (thanks for noticing) because I do genuinely respect the people
working on the different approaches. I mean what I say: I have worked on
both loop quantum gravity and strings, and I expect to keep working on
both, and I say that both should be supported. I happen to believe that it
is important to state clearly what are the genuine results and distinguish
them from the open conjectures. My hope is that doing so will lead to more
progress on both sides. For example, I suspect (of course I don't know)
that more progress would have been made on some of the open conjectures of
string theory such as perturbative finiteness past two loops were more
people not under the apparently widespread impression that the problem has
been solved or is trivial. On the loop quantum gravity side we have
always been very public about what the open problems are. For example,
before Dreyer's paper we were very clear that a parameter had to be fixed
to match the 1/4 in the black hole entropy. It was also always us (and not
some outside critics) who always, in most review talks and papers,
reminded people that having exact solutions to all the constraints was not
sufficient because the theory could be well defined as a quantum field
theory and still not have a low energy limit that reproduces classical
general relativity.
Thanks,
Lee
Fizz Fann
Ralph Hartley <har...@aic.nrl.navy.mil> wrote in message news:<b6g34o$mo2$1...@ra.nrl.navy.mil>...
> Fizz Fann wrote:
>
> > 2. it is possible but unlikely for him
> > to guess the right sequence of combinations to send young me a signal
> By that argument, you get paradoxes even without QM. It works if you
> communicate by guessing.
>
> You send him a message by just thinking it. He receives it by correctly
> guessing what you thought.
>
>
> Your past self decides to cause a paradox by not sending the message.
>
> Of course, it isn't likely the the recipients will correctly guess the
> message, but sometimes they will. Who keeps track of it all to prevent a
> paradox from happening?
I'm not sure whether or not you really do'nt understand or are just
having fun. I'll explain it anyway.
A paradox happens when we have something like A implies not A.
If we have a sequence like "A implies B; B implies C; C implies D;
D implies not(A)" then we have a paradox, or rather, we have a paradox
if A is true, since not A can be true and the above sequence can still
work. By "A implies B" we mean that B will be true if A is true. To
make this more tight, we definitely have a paradox if the above is
true and it is also true that "not(A) implies not(B); not(B) implies
not(C); not(C) implies not(D); not(D) implies A". Agreed?
In the case of the violation of Bells inequalities, we have the following
situation. The faraway experimenters results depend on my decisions. That
is at least sometimes it is the case that if I measure x he will get up
and if I measure y he will get down. So with the A, B, C, and D above,
for my question, we have:
A = "I measure along x"
not A = "I measure along y"
B = "He gets up"
not B = "He gets down" (by convention, at this point he always measures x)
C = "When he grows up and accelerates to a frame in which he is simultaneous
with my younger self and does an EPR experiment with my younger self,
he measures along x."
not C = " ditto ditto .. measures along y"
D = "My younger self gets up" (at this point I always measure x)
not D = "My younger self gets down"
The violation of Bells inequalities means that at least sometimes
(it is not clear when, but at least sometimes!) A implies B and not
A implies not B. That is, his results depend on the way I align my
measuring device.
B implies C because the faraway experimenter can do whatever he likes
and he has decided that this is the course of action that he will take
if he gets up as his result, but if he gets _down_ as a result then
he would measure along y. So B implies C and not B implies not C.
C implies D and not C implies not D, at least sometimes, again by virtue
of the violation of Bells inequalities.
D implies not A because I decided before this whole enterprise began
that I would measure along y if my younger self had gotten up as a result
and I would measure along x if my younger self had gotten down as a result.
So D implies not A and not D implies A.
Now there is a sometimes in the A<=>B part and a sometimes in the C<=>D
part. Maybe those sometimeses happen to be exclusive, so when A implies
B, then C does not imply D. But a law which imposes this constraint is not
present in the laws of quantum mechanics or anything else which has been
mentioned so far (as far as I know).
Your system of communicating by guessing does not even approximately
achieve the same result because in that system we do'nt have the
A implies B or the C implies D parts at all, so there can be no paradox.
> You really *do* have to send information to cause a paradox, and that means
> something you control.
If you want to be 100% certain that you get to cause a paradox then you
have to send information. I am not saying that you can cause a paradox
with 100% certainty (there are _two_ sometimeses in the above scenario).
I am saying that there is nothing in quantum mechanics which I have seen
which prevents A<=>B and C<=>D simultaneously above. A<=>B and C<=>D
simultaneously would cause a paradox without sending information in your
sense and without any nonlocal effects being controlled.
My question was: is there an extra law which has'nt been mentioned which
guarantees that when A<=>B, it is not the case that C<=>D? If there were,
would it mean that the extent to which Bells inequalities were violated
in the above hypothetical experiments would have to be numerically
different from the usual degree of violation when nobody is deliberately
trying to cause a paradox? The last question could in theory be
experimentally checked.
If Ilja is watching, parhaps he could say whether this is the case for
the preferred frame solution.
Best wishes,
Alex.
Fizz Fann
My apologies for replying twice. After looking at my first reply, I
realised that I had'nt actually answered your question properly,
or rather, directly.
Ralph Hartley <har...@aic.nrl.navy.mil> wrote in message news:<b6g34o$mo2$1...@ra.nrl.navy.mil>...
> Fizz Fann wrote:
>
> > 2. it is possible but unlikely for him
> > to guess the right sequence of combinations to send young me a signal
>
> By that argument, you get paradoxes even without QM. It works if you
> communicate by guessing.
>
> You send him a message by just thinking it. He receives it by correctly
> guessing what you thought.
It is the sender, not the recipient, who does the guessing in my scenario.
In your system, what the sender does has no consequences for the recipients
future actions. With qm nonlocality, the senders actions really do affect
the future life of the recipient, in the sense that had the sender done
something different, the recipient would get different results.
Regards,
Alex.
Urs Schreiber
John Baez schrieb:
> Urs Schreiber wrote:
> >(I don't know how much sense one can make of coherent graviton
> >states in canonically quantized gravity.)
>
> There's a minor industry devoted to just this,
Thanks for pointing this out. I see the same point emphasized in the
Smolin paper you mention:
> A review article by Smolin, due to appear on the arXiv today
> or tomorrow, will discuss this and many points of comparison
> between strings and loops.
I assume that you are referring to hep-th/0303185. I have a couple of
comments and questions:
- On p. 15 the author says that one difference between LQG and strings
is that in string theory spacetime is taken to be smooth down to
arbitrarily small scales. As I have entered this thread with the remark
that this is indeed _not_ the case I would like to hear other people's
opinion about it.
I had devoted some key strokes to the fact that while the background
metric that enters the sigma model is indeed smooth on all scales, this
metric is not operationally effective at small scales, since "string
uncertainty" takes over. I.e. smaller scales require higher probe
energies which make the string probes become excited and expand thus
making a local smooth manifold structure invisible.
This heuristic picture has been made precise in the course of an
extensive program (I had given some references) in which it is studied
how the vertex operator algebra of perturbative string theory (which
encodes all information about string interaction) gives rise to a
non-commutative geometry approximating the classical geometry defined by
the metric tensor which enters the sigma model.
- After reading the review I am a little confused about what LQG says
about its classical limit. On p. 26 it says about the LQG program:
"[...] at one point a problem appeared with the action of the
Hamiltonian constraint, which suggested difficulties with the recovery
of general relativity in the low energy limit. This problem was
addressed and solved".
This seems to be unambiguous. But then on p. 27 it reads: "The main open
issues concern whether and how general relativity, coupled to quantum
matter fields, is recovered [by LQG] in a suitable low energy limit.
[..] the question of whether or not the theory has a good low energy
limit is open for general states". This seems so say the opposite.
Indeed it makes appear questionable the former claim on p. 24: "On the
basis of these results, it can be claimed that loop quantum gravity I is
[...] the correct quantization of general relativity [...]."
How can that be claimed if it remains open whether general relativity is
recovered in the appropriate limit?
I realize that there must be some fine print that I am not fully aware
of. There is the "Kodama" state, an exact quantum solution with
apparently good (semi-)classical limit which seems to play a pivotal
role.
- I was surprised to read on p. 35 that "No concistent stable string
backgrounds are known which are time dependent." Why is that? What about
any time dependent (non-stationary) solution to the corresponding
supergravity? Again, I assume I am missing the fineprint. Only related
thing I know is from Polchinski's book (I, p. 117) that static
backgrounds are nice, among other things, because they allow to have a
general d-1 dimensional CFT for the remaining spatial dimensions.
(BTW, I have recently invested some thought on implementing a "time
gauge" for the superstring (as opposed to the usual light-cone gauge),
i.e. a gauge in which on a static background we have X^0 = X^0(tau) <=>
partial_sigma X^0 = 0. I made some progress, my main problem being that
I do not see if and how the resulting spacetime Hamiltonian commutes
weakly with the spatial constraint algebra on spacelike hypersurfaces
(except for Minkowski background). I could not find much on time gauge
in string theory in the literature, except for some application in
classical cosmic strings. Does anyone know any helpful references?)
- Then, on p. 38, T-duality in string theory is briefly described with
the words: "In all these cases there is a symmetry in which one
exchanges winding and vibrational modes". This is false, and I assume
the author knows better: T-duality exchanges winding and _momentum_
degrees of freedom.
I wish I could also make some comments or ask some questions about the
remaining part, which is largely concerned with string dualities and
M-theory.
Gene Partlow
> then there should be effects analagous to "virtual matter loops".
>
> 2) He elevates background independence to the level of physical
> law, without acknowledging the ambiguity of this phrase, and without
> acknowledging that it is a principle which will remains open to
> experimental study. He apparently means "spacetime background
> independence", so why not just say "diffeomorphism invariance of the
> dynamics"? Or alternately, why not demand an explanation of the
>numerical value of Newton's constant? Is there a reason this
>number should not be regarded as a non-dynamical classical
>field?
>
> --A.J.
I don't want to derail the theme of this string of posts, but you
tweaked my interest here. Newton's constant is in some sense
connected with the "coupling beween the matter and gravity fields",
and some theorists have speculated that G may be variable. For
nefarious reasons of my own (a half-vast toy model involving an in-
finite sequence of branching, proliferating universes with some
'constants' changing over time) I suspect that G may be _constant_
not only in our universe but all universes...a transuniversal in-
variant. Thus, my question...
Is it conceivable that a true constancy for G (as somehow more
fundamental than some other parameters) might follow from a
"diffeomorphism invariance of the dynamics", or am I just trying to
equate apples and rutabagas? ;-|
Gene Partlow
Kevin A. Scaldeferri
In article <b2b24cf0.03032...@posting.google.com>,
Fizz Fann <thephy...@yahoo.com> wrote:
>A paradox happens when we have something like A implies not A.
>
>If we have a sequence like "A implies B; B implies C; C implies D;
>D implies not(A)" then we have a paradox, or rather, we have a paradox
>if A is true, since not A can be true and the above sequence can still
>work. By "A implies B" we mean that B will be true if A is true. To
>make this more tight, we definitely have a paradox if the above is
>true and it is also true that "not(A) implies not(B); not(B) implies
>not(C); not(C) implies not(D); not(D) implies A". Agreed?
>
>In the case of the violation of Bells inequalities, we have the following
>situation. The faraway experimenters results depend on my decisions. That
>is at least sometimes it is the case that if I measure x he will get up
>and if I measure y he will get down. So with the A, B, C, and D above,
>for my question, we have:
>
>A = "I measure along x"
>not A = "I measure along y"
>
>B = "He gets up"
>not B = "He gets down" (by convention, at this point he always measures x)
>
>C = "When he grows up and accelerates to a frame in which he is simultaneous
> with my younger self and does an EPR experiment with my younger self,
> he measures along x."
>not C = " ditto ditto .. measures along y"
>
>D = "My younger self gets up" (at this point I always measure x)
>not D = "My younger self gets down"
Naively, it does appear that something like this might lead to a
paradox. However, this is still a vague description of the situation.
Can you specify the precise quantum states and measurements which
would have to be prepared for this problem to actually arise?
Ilja Schmelzer
> In article <i3g1y16...@wias-berlin.de> ,
> Ilja Schmelzer <schm...@wias-berlin.de> wrote:
>> Moreover, in a purely theoretical domain, where people need only a PC.
>> Some conformity pressure may be necessary if you have to spend lots
>> of taxpayers money for accelerators and so on, which we don't want to
>> waste for crank science. But in almost pure theory?
>
> A PC(*), a salary, a few grad students and postdocs, a secretary, journal
> subscriptions, a travel budget, overhead. It doesn't look too hard for
> one professor to need half a million or more a year.
That's not what he needs to make theoretical physics. That's what the
sufficiently rich society gives him.
Ilja
--
I. Schmelzer, <il...@ilja-schmelzer.net> , http://ilja-schmelzer.net
[Moderator's note: What say we go back to talking about physics? -TB]
Ralph Hartley
> Ralph Hartley <har...@aic.nrl.navy.mil> wrote:
>>Fizz Fann wrote:
>>>2. it is possible but unlikely for him
>>> to guess the right sequence of combinations to send young me a signal
>>You send him a message by just thinking it. He receives it by correctly
>>guessing what you thought.
>
> It is the sender, not the recipient, who does the guessing in my scenario.
So? What difference can that make to anything objectively observable? The
probability that the message will be received correctly is the same in both
cases.
You seem to think that even if A and B are independent, because *sometimes*
A and B both happen, then it makes sense to say that "sometimes A->B". What
you end up with is a concept of causation that can't cause anything.
> In your system, what the sender does has no consequences for the recipients
> future actions. With qm nonlocality, the senders actions really do affect
> the future life of the recipient, in the sense that had the sender done
> something different, the recipient would get different results.
But who's to say the causality isn't in the opposite direction? In your
example where your settings seem to affect the distant result, it isn't
your settings themselves that have the effect, it's the fact that they were
based on a correct guess.
But the distant results affect whether your guess is right or not. Not by
changing your guess (which is based on classical cogs and wheels and is
presumably out of reach of quantum effects), but by changing the correct
answer (whether they also affect your *results* is a matter of
interpretation, and can't be proven either way).
Your argument seems to be:
(1) I guess what value I would get if I measured in the vertical direction.
(2) If I correctly guess down, I measure in the vertical direction, get
down, and the other guy gets up.
(3) If I correctly guess up, I measure in the horizontal direction get left
or right, and the other guy gets up and down with equal probability.
(4) So if I guess correctly, the other guy gets up with probability 3/4
which is greater than 1/2.
(5) Sometimes I guess correctly, so sometimes I affect the distant result.
Point (5) is debatable, but I'm giving up on arguing with your strange
concept of "causation". Fortunately, I don't need to, since (3) (and
consequently (4) is wrong.
(3A) If the other guy gets up, and I measure in the vertical direction, I
*will* get down. So if I *correctly* guess up, he *must* get down,
regardless of what measurement I do, because if he got up and I guessed up
my guess would *not* be correct!
It might be *possible* to think of some definition of "what I would have
gotten" that could be up even if he got down, but it would not agree with
the normal concept at all. The concepts you use to the describe a situation
can *never* change any objective outcome, so if paradoxes can't happen with
"what I would have gotten" defined my way, they can't happen with your way
either, unless your way is inconsistent.
(4A) Even if base my actions on a correct guess, he still gets up 1/2 of
the time.
Ralph Hartley
ps. The tricky bits with this experiment happen when measurements are made
in directions other that just vertical or horizontal. But I've had about
enough.
Thomas Larsson
Lee Smolin <smo...@fake.email.address.com> wrote in message news:b5qr90$5dn$1...@glue.ucr.edu...
>
> 1) I did try to be very clear to distinguish spatially diffeomorphism
> invariant from spacetime diffeomorphism invariant in the detailed list of
> claims. I do not claim that solutions to the hamiltonian constraint are
> spacetime diffeomorphism invariant, as the full set of constraints
> only generate spacetime diffeomorphisms on solutions. On page 30 I
> explicitly mention that
>
> " There are also some unresolved issues concerning the role of the four
> dimensional diffeomorphism group in the Hamiltonian theory. This
> comes into the details of the regularization of the Hamiltonian
> constraint and the relationship between the hamiltonian and path
> integral quantizations. A set of related issues have to do with the
> relationships between the different forms of the quantum hamiltonian
> constraint arrived at by different regularization procedures and
> different operator orderings."
>
> 2) I also clearly distinguish in point 4 that I am quoting results
> that are "physical" in a setting where the time gauge is fixed.
>
> 3) I also clearly distinguished different kinds of background
> independence, that is one reason I distinguished loop quantum gravity I and
> II. LQG I is independent of metric and fields but still dependent on
> topological and differential structure, while LQG II depends on no fixed
> topological or differential structure.
>
> I think I was clear to indicate the relationship between my use of
> diffeomorphism invariant and the more general term background
> independent.
>
First I want to say that I like the philosophy behind LQG a lot, in particular
the idea to quantize gravity per se, without introducing a lot of unobserved
stuff like supersymmetry or extra dimensions. It is also very appealing (to
someone who grew up scientifically in the 80s when the Zeitgeist was a lot more
algebraically oriented) that some loop quantum gravitists have recently started
to apply algebraic methods, see e.g.
http://www.arxiv.org/abs/gr-qc/0303074
Irreducibility of the Ashtekar-Isham-Lewandowski representation
Authors: Hanno Sahlmann, Thomas Thiemann
http://www.arxiv.org/abs/gr-qc/0302090
On the superselection theory of the Weyl algebra for diffeomorphism
invariant quantum gauge theories
Authors: Hanno Sahlmann, Thomas Thiemann
http://www.arxiv.org/abs/gr-qc/0302059
Diffeomorphism covariant representations of the holonomy-flux star-algebra
Authors: Andrzej Okolow, Jerzy Lewandowski
So maybe studying representations of the diffeomorphism algebra is not
such a fringe activity after all...
On closer scrutiny, I am less enthusiastic about the details of these papers.
They do not study the spacetime diff group, but rather spatial diffs times
Yang-Mills transformations. I suppose that this is the constraint algebra of
Hamiltonian gravity in Ashtekar variables, and as such it equals the spacetime
diff algebra once the equations of motion are taken into account (at least the
analogous thing happens for the constraint algebra of ADM gravity). However, in
the algebraic framework the equations of motion go out of focus, so it is unclear
to me if they really study the right algebra. Even if it is equivalent to
spacetime diffs classically, there might be quantization problems.
Another complaint is that the claim that the Ashtekar-Isham-Lewandowski rep is
the only known rep. Reps of the diff group (space or spacetime - mathematicians
don't care) have been classified; the irreducible modules are tensor densities
or, in the totally skew case, closed forms. This was proven for the algebra of
polynomial vector fields by
A.N. Rudakov,
Irreducible representations of infinite-dimensional Lie algebras of Cartan type
Math. USSR Izv. 8, 836-866 (1974).
Globally these modules should correspond to spaces of sections of tensor bundles
(i.e. suitable tensor products of tangent, cotangent and determinant bundles in
case I got the terminology wrong). This is probably dealt with in
A M Vershik, I M Gelfand and M I Graev,
Representations of the group of diffeomorphisms
Russian Math Surveys 30:6 (1975) 1 - 50
Unfortunately I have lost my Xerox copy, so I can't give any details.
In view of this classification, it seems unlikely that anyone can come up with a
genuinly new representation, unless I am missing some fine print. The physicists
above study of course a different group, namely diffs coupled to Yang-Mills
transformations rather than just diffs, but I don't see that changing anything.
The relevant modules should be tensor fields valued in G modules, or the closely
related Yang-Mills connection.
Of course, Rudakov's classification only applies to proper reps. Projective reps
is a completely different game...
It might be mentioned that I'm not the only one who has proposed that
understanding spacetime diffs is important to quantum gravity. This was also
the main motivation behind the following papers
E Ramos, C H Sah and R E Shrock,
Algebras of diffeomorphisms of the N-torus,
J Math Phys 31 (1989) 1805 - 1816
F Figueirido and E Ramos,
Fock space representations of the algebra of diffeomorphisms on the n-torus,
Int J Mod Phys A6 (1991) 711 - 806
However, like myself at the time, these authors were unable to say anything
significant about representations. That only became possible after the
breakthrough in
S Eswara Rao and R.V. Moody,
Vertex representations for N-toroidal Lie algebras and a generalization of the
Virasoro algebra,
Commun. Math. Phys. 159, 239--264 (1994).
Fizz Fann
> ps. The tricky bits with this experiment happen when measurements are made
> in directions other that just vertical or horizontal. But I've had about
> enough.
Agreed. This whole process is beginning to sicken me. If you don't
feel that I've addressed any of your points with this post, what
say we just drop the whole thing? It seems that others are able
to see the paradox which I was talking about even if you can't.
I do believe that you are intelligent and could contribute to the
discussion, though, so it would be a shame if you just walked away
in disgust.
> >>You send him a message by just thinking it. He receives it by correctly
> >>guessing what you thought.
> > It is the sender, not the recipient, who does the guessing in my scenario.
> So? What difference can that make to anything objectively observable? The
> probability that the message will be received correctly is the same in both
> cases.
You are right that the probability of the receiver getting any particular
result doesn't change if the sender guesses blindly. However, the
receiver gets an individual result, and not a probability distribution.
The thing which is observable is the individual result, and it is this
which is affected by the actions of the faraway experimenter.
Indeed, the faraway experimenter gets to decide whether or not "what I
would get if I measured along x" is the same as "what I would get if I
measured y", more or less. Certainly each of those things is observable
since I can measure along x and observe "what I would get if I measured
along x", for example. It is indeed as you said; he needs to make
measurements other than vertical/horizontal to do this.
> > In your system, what the sender does has no consequences for the recipients
> > future actions. With qm nonlocality, the senders actions really do affect
> > the future life of the recipient, in the sense that had the sender done
> > something different, the recipient would get different results.
> But who's to say the causality isn't in the opposite direction? In your
> example where your settings seem to affect the distant result, it isn't
> your settings themselves that have the effect, it's the fact that they were
> based on a correct guess.
No; it's the settings themselves which have the effect. However the
effect is uncontrollable. This isn't just a bizarre idea which I
came up with in a bar; Bell and Eberhard showed that the predictions of
quantum mechanics can only be explained if the results of a measurement
depend on the settings of the faraway detector. Those predictions were
tested in experiments and the nonlocal effect was confirmed.
> But the distant results affect whether your guess is right or not. Not by
> changing your guess (which is based on classical cogs and wheels and is
> presumably out of reach of quantum effects), but by changing the correct
> answer (whether they also affect your *results* is a matter of
> interpretation, and can't be proven either way).
The problem is that you are considering the distant results are though they
were written more firmly in stone than the decisions of the experimenters.
My point about the gears and cogs was precisely intended to show the
opposite: the experimenters' decisions are more "solid" than the results
of the measurements.
> Your argument seems to be:
> [ parody of my argument ]
This has forked into two arguments. One is about whether a paradox can
be caused. Other posts in this thread deal more completely with that
(see the ones involving A=>not(A)=>A etc). The other is about whether
one experimenter could cause a detectable signal for another if he had
access to certain extra information.
Here's my argument for the second point. Firstly, it has been shown
that the results of a measurement depend on the settings of a faraway
detector. That means that, at least sometimes, it is the case that, for
at least some set of axes, 1, 2 and 3:
If experimenter 1 measures along axis 1 and experimenter 2 measures
along axis 2 then experimenter 2 will get "up" as a result
AND
if experimenter 1 measures along axis 3 and experimenter 2 measures
along axis 2 then experimenter 2 will get "down" as a result.
If this was never the case, then nobody could ever say that the
results of experimenter 2 depended on the setting of experimenter 1's
measuring device.
(It would make the discussion much more direct if you could say whether
or not you agree with that.)
If you agree with that, then perhaps you might also agree that
_if_ experimenter 1 was aware (or correctly guessed) that the
situation was as described above, then he could choose to
measure along axis 1 rather than axis 3, thereby _causing_
experimenter two to get "up" rather than "down".
Now your point was that the causation could be in the opposite
direction. You were taking the result of experimenter 2 as though
it were writ in stone before experimenter 1 made his choice.
What I am saying is something like: the two events (choice of 1
and result of 2) are separated by a spacelike distance, so we
cannot point at one and say it comes before the other and is
therefore the one doing the causation. However, we can say that
one is governed by deterministic classical physics and is
therefore not free to obey the whims and dictates of the distant
event. Luckily, only one of the two is like this (the choice
of experimenter 1), so the result of experimenter 2 can be
pointed out as the dependent thing. That is, it is the thing
which is less firmly writ in stone and whose value can be
said to be dependent on the faraway choice of experimenter 1
(among possibly many other things).
The above paragraph makes no mention of correct guesses, and
that is because it is designed to show which of the two things
(choice of 1 or result of 2) is dependent and which is
independent. After one has accepted that, one can then turn
to the issue of asking about what happens if experimenter 1
correctly guesses the situation each time. It is not right to
look at the notion of correct guesses and say that it undermines
the argument given in the paragraph above, because that argument
doesn't use the notion of guesses at all.
> (5) Sometimes I guess correctly, so sometimes I affect the distant result.
I'm not saying that at all. I affect his result anyway, but sometimes
I cause him to get down and sometimes I cause him to get up. The point
of the correct guesswork was that if I were aware of (or
correctly guessed) all the "what would he get if I measured this"es
then I could choose my measurements in such a way as to make him
get up more often than down.
> (3A) If the other guy gets up, and I measure in the vertical direction, I
> *will* get down. So if I *correctly* guess up, he *must* get down,
> regardless of what measurement I do, because if he got up and I guessed up
> my guess would *not* be correct!
Again, your problem is that you are considering his result as though
it were more firmly writ in stone than my decision about what to
measure. You say: "if he got up and I guessed up my guess would *not*
be correct!" The point is that it is not as simple as "he got up" or
"he got down". Whether he gets up or down depends (at least
sometimes) on what measurement I make. So the guess which I make is
not just as simple as "what did he get?", but it is more like "what
would he get if I measured along this axis?"
> The concepts you use to the describe a situation
> can *never* change any objective outcome, so if paradoxes can't happen with
> "what I would have gotten" defined my way, they can't happen with your way
> either, unless your way is inconsistent.
I do believe that discussing this with you has been useful for me, and
has made me clarify certain points in my mind, but with all due respect,
you have omitted one possibility in your above sentence.
Best wishes,
Alex.
Fizz Fann
[Note to moderator: This is a repost. I posted this one week ago
and it did not appear and I did not receive a notice from a
moderator saying anything about it. According to:
http://math.ucr.edu/home/baez/spr.html
one should wait a week before assuming that the post was lost.
I have also changed the title since the original title wasn't
appropriate.]
ke...@sue.its.caltech.edu (Kevin A. Scaldeferri) wrote in message
news:<b5sqe3$f9l$1...@sue.its.caltech.edu>...
>
> Naively, it does appear that something like this might lead to a
> paradox. However, this is still a vague description of the
situation.
> Can you specify the precise quantum states and measurements which
> would have to be prepared for this problem to actually arise?
Ok. Here is an attempt at a more precise description of everything
involved in the construction (with diagrams!). I have changed the
axes involved in the measurements to make the point more clear and
to give more of an intuition about how it is that the choice of
axis affects the result of the faraway experiment. I have also
made it so that there are N particle pairs involved in each experiment.
To bring the description more into line with typical descriptions of
such scenarios, I have changed the names of the experimenters
to Alice and Bob, rather than "me" and "him".
All of this may be either weirdly phrased or badly expressed
because I've only been looking at this stuff for about two
months. I will be glad to clarify any point which may be unclear.
There are two experimenters (Alice and Bob) and two experiments,
E1 and E2, performed in frames F1 and F2 respectively. The
experiments are EPR experiments similar to those used to test the
violation of Bell's inequalities.
Alice and Bob are separated by a large spacelike distance, so
large that a light signal from the event of the birth of one
would not reach the other until the other had died.
Experiment E1, in frame F1:
---------------------------
Alice is born in frame F1, and, according to her notion of
simultaneity, Bob is old and is not moving relative to Alice,
so that he is also in the inertial frame F1. Midway between
Alice and Bob is a source of pairs of spin-half particles. The
source is also in the inertial frame F1. Each of the particle
pairs which it creates is in the singlet spin state described by:
(|up>|down> - |down>|up>)/sqrt(2)
The measurement of a particle's spin is performed
with a Stern-Gerlach device. The SG device has inside it a
magnetic field which points along a particular axis. The
device may be rotated to change the axis of the field, and,
once an axis has been chosen, the spin of a particle along
that axis may be measured by observing in which direction the
particle is deflected upon entering the region of the magnetic
field. Since the particles are spin-half particles, the possible
results of a measurement are plus or minus 1/2, which we can
call "up" and "down" respectively.
Because the particles are in the "singlet" state, if one has
been measured to have spin up along a particular axis, then
the other, with certainty, will be found to have spin down
along that same axis if a measurement along that axis is
performed. If one particle is found to have spin up along
a particular axis, and a measurement of the spin of the
second particle along an axis which is at an angle of theta
with respect to the first axis is performed, then the prediction
of quantum mechanics is that the probability that the second
particle will be found to have spin up along that axis is
sin^2(theta/2), and the probability that it will be found to
have spin down is cos^2(theta/2).
One of the particles from each pair goes towards Alice and the
other goes towards Bob. The source of particles has been there
for a long time, so that some of the particles which it emitted
in the distant past arrive at Alice while she is young and the
corresponding partner-particles arrive at Bob when he is old.
Both Alice and Bob are equipped with Stern-Gerlach devices.
In frame F1 which is shared by young Alice, old Bob, and
the particle source, the particles move along the z-axis,
on which Alice, Bob, and the particle source are located as
shown:
<-- particle particle -->
-z <---- Alice ------------- source ------------- Bob --------> +z
When Alice or Bob make a spin measurement of a particle, the
particle enters the magnetic field of the SG device along
the z-axis and is then deflected either in the direction of
or against the direction of the magnetic field.
The relevant axes along which the magnetic field may be aligned
are all perpendicular to the z-axis and are shown below:
+s +y +w
\ | /
\ | / <- 45 degrees
-x____|___+x
/|\
/ | \
/ | \
-w -y -s
Bob measures along the w or s axes, while Alice measures along the
x-axis each time. To see that Bob's measurements affect Alice's
results, consider the following simple argument.
Assuming counter-factual definiteness, denote the result that
Alice would get to an x measurement by R_x and the result that
she would get to a y measurement by R_y.
Suppose that Bob measures along the w-axis. His result wll
be either "up" or "down". If it is "up", then, since the
spin of Alice's particle is "opposite" to the spin of Bob's
(which is "up" along the w-axis), both R_x and R_y are
most likely to be "down".
The prediction of quantum mechanics is that, if Bob gets "up"
along w, then a measurement of y-spin by Alice will produce a
"down" with 85.3% probability, and a measurement of x-spin will
also produce a "down" with 85.3% probability (=cos^2(45/2)).
Also, if Bob measures along the w-axis and gets "down", then
Alice would get "up" for an x measurement (with probability
85.3%) and "up" for a y measurement (prob=85.3%). So R_x and
R_y are each probably "up".
Either way, if Bob measures along the w-axis, then R_x is
probably equal to R_y.
On the other hand, if Bob measures along the s axis and gets
"up", then R_x is "up" with probability 85.3% (=sin^2(135/2))
and R_y is "down" with probability 85.3% (=cos^2(45/2)). If
he measures along s and gets "down", then R_x is "down" with
probability 85.3% and R_y is "up" with probability 85.3%.
Either way, if Bob measures along the s-axis then R_x is
probably _not_ equal to R_y.
Since Bob can choose to measure along w or s, he gets to choose
(to some extent, determined by the probabilities) whether or
not R_x is equal to R_y, where these are the results that
Alice would get for a measurement of x- or y-spin respectively.
It is in this sense that Bob's choice to measure along w or s
affects the result that Alice gets for her measurements.
Although he cannot control the actual results, R_x or R_y,
that she would get, (these appear, according to quantum
theory, to be random) he gets to decide whether
R_x equals R_y (approximately).
During experiment E1, Alice (who is young) always measures along
the x-axis. Bob, who is old, has in his possession a notebook which
contains a list of N letters, each a "w" or an "s". It will become
clear during the description of experiment E2 where this list
came from. For the nth incoming particle, Bob measures along
the w-axis if the nth entry in his list is a "w" and along the
s-axis if the nth entry in his list is an "s". There are N
particle pairs in total, so the number of letters in the list
is equal to the number of particle pairs.
Alice measures always along the x-axis and gets either "up" or
"down" for her result. When she gets "up" as a result, she
writes "w" in her notebook and when she gets down she writes
"s" in her notebook.
Since Bob's choices to measure along s or w are affecting her
results, it is the case that:
Nonlocal effect 1:
------------------
At least sometimes it is the case that if Bob measures along
w then Alice will get "up" as a result and if Bob measures
along s then Alice will get "down" as a result.
------------------
If N is sufficiently large, then we can be fairly sure that
this has been the case a significant number of times.
After experiment E1 has finished, Alice possesses a notebook
with a list of N letters, each of which is an "s" or a "w".
Alice then boosts into frame F2 and grows older, becoming
old Alice. In her new frame of reference, F2, old Alice
is simultaneous with "young Bob", who has yet to grow up
and participate in experiment E1. Young Bob is at rest
relative to old Alice; that is, he shares the inertial
frame F2. When he grows up, he will boost into the frame
F1 in preparation for experiment E1.
For now, he is doing:
Experiment E2, in frame F2:
---------------------------
Midway between young Bob and old Alice is a second source
of EPR particle-pairs. It also emits particles in the
singlet state (|up>|down> - |down>|up>)/sqrt(2) multiplied
by a spatial wavefunction. The particle source is also in
the frame F2 and has been there long enough that Alice
and Bob receive particles from it while they are in the
frame F2. There are N particle-pairs involved in this
experiment, and Alice receives one particle from each
pair while Bob receives the partner particle.
Both Alice and Bob are equipped with SG devices and the
axes are as described in experiment E1.
Alice, who has the list which she prepared in experiment
E1, measures the nth incoming particle along the
w-axis if the nth entry in her list is a "w", and measures
along the s-axis if the nth entry in her list is an "s".
Young Bob always measures along the x-axis. Whenever he
gets "up" as a result, he writes "s" in his notebook.
Whenever he gets "down", he writes "w" in his notebook.
After the experiment, he has a list of N letters in
his notebook. This is the list which he will use in
experiment E1, when he grows up.
Assuming that the laws of physics do not depend on the
reference frame involved, the arguments used in the
description of experiment E1 above also indicate that in
this experiment, we have
Nonlocal effect 2:
------------------
At least sometimes it is the case that if Alice measures along
w then Bob will get "up" as a result and if Alice measures
along s then Bob will get "down" as a result.
------------------
As with nonlocal effect 1, if N is sufficiently large, this
should happen a significant number of times.
The Paradox:
------------
Considering the following statements:
A(n) = "During E1, Bob measures the w-spin of the nth particle"
not(A(n)) = "During E1, Bob measures the s-spin of the nth particle"
B(n) = "During E1, Alice gets 'up' as her nth result"
not(B(n)) = "During E1, Alice gets 'down' as her nth result"
C(n) = "During E2, Alice measures the w-spin of the nth particle"
not(C(n)) = "During E2, Alice measures the s-spin of the nth particle"
D(n) = "During E2, Bob gets 'up' as his nth result"
not(D(n)) = "During E2, Bob gets 'down' as his nth result"
Nonlocal effect 1 indicates that, for some values of n,
A(n)=>B(n) and not(A(n))=>not(B(n))
The prescribed behavior of Alice indicates that (for every n)
B(n)=>C(n) and not(B(n))=>not(C(n))
Nonlocal effect 2 indicates that, for some values of n,
C(n)=>D(n) and not(C(n))=>not(D(n))
The prescribed behavior of Bob indicates that (for every n)
D(n)=>not(A(n)) and not(D(n))=>A(n)
Now, if N is sufficiently large, then unless there is some
extra rule to prevent this from happening, there will be
some n for which both nonlocal effect 1 and nonlocal effect
2 are at work, and then we would have
A(n)=>not(A(n))=>A(n), which is a paradox.
To prevent this, as I understand it, there would need to be
a rule saying that for a particular value of n, nonlocal
effect 1 can be at work, or nonlocal effect 2 can be at
work, but not both.
A further question arising from such a law would be whether
such a constraint, which would effectively prevent nonlocal
effects in certain circumstances, would interfere with
the degree to which Bell's inequalities were violated in
such experiments. That is, suppose that a particular
fraction f of the N particle-pairs in experiment E1
would exhibit nonlocal effect 1 if there was no experiment
E2. Then when experiment E2 is added into the picture,
would the fraction of pairs in E1 exhibiting the effect
still be f?
Regards,
Alex.
Ralph Hartley
I think I'm starting to see where your problem is.
You think that things like "I would have gotten x" and "his results depend
on my choice" are statements of fact, that are objectively true or false,
and not just ways of looking at things, dependent on your point of view.
This might very well be so. There are interpretations of QM in which it is.
If you demand that a interpretation not force you to "change the way you
think about logic and truth" (paraphrasing your expired post), then you are
limiting yourself to one of those.
It doesn't matter to this argument, but I assume you also believe that
there is a *true* interpretation. It may be impossible to find out which
that is, since they are experimentally indistinguishable. Whether this
position is true, or even meaningful, or not is a philosophical issue,
which I will not discuss further.
However, the fact that QM violates Bell's inequality *does* mean that you
have to give *something* up.
I'm pretty sure that the thing you have to lose to satisfy your "demands"
is Lorenz invariance. All such interpretations have a "preferred" frame
(*which* frame is preferred is, unfortunately(?), unobservable). Bohmian
mechanics, for example, has this property.
It is *not* ok, in your interpretation, to figure out what affected what in
any frame other than the the preferred one. The frame will not matter as
far as predicting the results of experiments, so if you use the wrong one
you can't be proven wrong, but if you use two *different* frames they can't
*both* be the preferred frame, so one or more of them may give the wrong
answer, causing an apparent paradox.
Fizz Fann wrote:
> Ralph Hartley <har...@aic.nrl.navy.mil> wrote:
>>(3A) If the other guy gets up, and I measure in the vertical direction, I
>>*will* get down. So if I *correctly* guess up, he *must* get down,
>>regardless of what measurement I do, because if he got up and I guessed up
>>my guess would *not* be correct!
> Again, your problem is that you are considering his result as though
> it were more firmly writ in stone than my decision about what to
> measure.
But it is, if his measurement happens *before* your decision. What he got
affects the correctness of the guess you will make (though not your guess
itself). His measurement is already made, nothing you can do can affect it!
If the separation between the measurements is timelike (so his measurement
happened *absolutely* before yours), this is the *only* possibility. So
don't say that your settings affect his result even if his measurement came
first! If you did you would also have to say the your settings affect his
measurement even if you aready looked over his sholder and saw his results,
before making your settings.
> it is not as simple as "he got up" or
> "he got down". Whether he gets up or down depends (at least
> sometimes) on what measurement I make. So the guess which I make is
> not just as simple as "what did he get?", but it is more like "what
> would he get if I measured along this axis?"
This is valid *if* your decision was made *before* his measurement. His
measurement (which hasn't happened yet) cannot affect your guess or it's
correctness, which is already a done deal. The guess you already made can
affect the measurement he will make.
This is the only interpretation of the events if the interval was timelike
and your measurement came first. We are discussing the case where the
interval is spacelike, so which is correct is frame defendant.
If you think that one of the two statements "your measurement affected his
result" and "his measurement affected the correctness of your guess" is
true, and the other false, then you *must* think that there is a "true"
reference frame, since one statement is true in some frames, and the other
in others.
If you think there is a "true" frame, do all your logic in that frame (or
at least always in the same one). "Causation" will always be forward in
time, and no paradox will appear.
It doesn't matter at all in what reference frames the various experimenters
and pieces of equipment are at rest. What causes what is different in
different reference frames, so if you think that causation goes in a
particular direction, you must think that there is particular *fixed* frame
to use when determining that, and that all other frames are "wrong".
Someone accelerating from the "right" frame to a "wrong" one can't change
the direction of causation.
And *that's* why I think such interpretations are "ugly".
Ralph Hartley
[Moderator's note: I have (rather belatedly) changed the subject header
from "Conceptual overlap between LQG and String/M-theory?" to something
more descriptive of this discussion. - jb]
A.J. Tolland
On Wed, 26 Mar 2003, Gene Partlow wrote:
> Is it conceivable that a true constancy for G (as somehow more
> fundamental than some other parameters) might follow from a
> "diffeomorphism invariance of the dynamics", or am I just trying to
> equate apples and rutabagas? ;-|
This doesn't seem likely to me. I think you are probably chasing
rutabagas. Sorry.
--A.J.
Ralph Hartley
Fizz Fann wrote:
> Ok. Here is an attempt at a more precise description of everything
> involved in the construction (with diagrams!).
[Snipping most of the description]
OK. Just to summarize there are 4 measurement results, and 2 variable
measurement settings.
Lets call the results E1a, E1b, E2a, and E2b, (each can be "up" or "down")
and the settings S1 and S2 (which can have the values "w" or "s").
The measurement settings are completely determined from the experimental
results in the following way:
E1a = up -> S2 = w
E1a = dn -> S2 = s
E2b = up -> S1 = s
E2b = dn -> S1 = w
So there are 16 possible outcomes.
Lets call P(a,b,c,d,e,f) the probability that E1b=a, E1a=b, E2b=c, E2a=d,
S1=e, and S2=f.
Quantum mechanics (and the setup you described) gives us the probabilities
of each:
P(up,up,dn,up,w,w) = 1/4 * (1-0.853)*0.853 = 0.0313
P(up,up,dn,dn,w,w) = 1/4 * (1-0.853)*(1-0.853) = 0.0054
P(up,dn,up,dn,s,s) = 0.25
P(up,dn,dn,up,w,s) = 1/4 * 0.853 = 0.2132
P(dn,up,up,up,s,w) = 1/4 * (1-0.853) = 0.0367
P(dn,up,up,dn,s,w) = 0.2132
P(dn,up,dn,up,w,w) = 1/4 0.853*0.853 = 0.1819
P(dn,up,dn,dn,w,w) = 0.0313
P(dn,dn,dn,up,w,s) = 0.0367
and all the rest are 0.
Thus all the outcomes have a fixed, known, probability, and they add up to 1.
If you do this experiment many times, these are the outcomes you will see.
Nothing in the above numbers is any kind of paradox or logical
contradiction. If all your talk of "causation" and "what would have
happened" gives some sort of contradiction, you must have created the
contradiction, because there is none in the actual outcomes of the experiment.
> There are two experimenters (Alice and Bob) and two experiments,
> E1 and E2, performed in frames F1 and F2 respectively.
What do you mean by "performed in frames"? The relative motions of the
apparatus or experimenters don't change the results shown in the table above.
> Nonlocal effect 1 indicates that, for some values of n,
> A(n)=>B(n) and not(A(n))=>not(B(n))
>
> The prescribed behavior of Alice indicates that (for every n)
> B(n)=>C(n) and not(B(n))=>not(C(n))
>
> Nonlocal effect 2 indicates that, for some values of n,
> C(n)=>D(n) and not(C(n))=>not(D(n))
>
> The prescribed behavior of Bob indicates that (for every n)
> D(n)=>not(A(n)) and not(D(n))=>A(n)
>
> Now, if N is sufficiently large, then unless there is some
> extra rule to prevent this from happening, there will be
> some n for which both nonlocal effect 1 and nonlocal effect
> 2 are at work, and then we would have
>
> A(n)=>not(A(n))=>A(n), which is a paradox.
Which values of n?
Ok, suppose you have N trials, and observed the results as above. Lets call
the trials where effect 1 happened T1, and those in which effect 2 happened
T2, and the rest T0. Show me how you can assign trials to these three sets
so that T1 and T2 have trials in common.
In other words show me a trial in which A->not(A)->A. This isn't just
because you don't *know* which trails they were. There is just no
consistent way to make such an assignment.
> To prevent this, as I understand it, there would need to be
> a rule saying that for a particular value of n, nonlocal
> effect 1 can be at work, or nonlocal effect 2 can be at
> work, but not both.
Yes, there are such rules. They are the rules of logic. There is no way to
assign a trial to T1 and T2, just because for any fixed outcome, the result
of each experiment must be up or down.
That isn't to say that there isn't something funny going on here. The
violation of Bell's inequality *does* mean you have to give up some assumption.
What assumption do *I* discard?
It depends on what problem I am trying to solve, or what mood I am in. Some
choices are:
The wavefunction collapses instantly, "what if" questions are meaningful,
but only in one fixed frame.
The wavefunction collapses instantly in *any* frame, but "what if"
questions are meaningless, or at least relative. The measurement creates
the measured value, it doesn't have any "pre-existing" value.
No collapse at all. Bob and Alice become entangled with the particle states
they measure, and none of the outcomes (with positive probability) are ever
any more "real" than any others.
Any of these possibilities (and more) give exactly the same outcomes. You
might ask which I believe, but I don't have to choose.
One of the following statements must be true:
Quantum mechanics is correct, and no experiment can distinguish between its
interpretations, even in principle.
or
Quantum mechanics is not correct, so *none* of its interpretations are correct.
There is *no* possibility that Quantum mechanis is correct, but some
experiment will eventually tell us what its *true* interpretation is.
I remember in the Koran (which I read in translation a *very* long time
ago, so I may have this wrong) there is a story about five boys who were
placed in eternal(?) sleep in a cave. Some versions of the story say there
were five including their dog, and others that the dog would make six. But
God says, "Only I know how many there were, and I haven't told anybody.
Anyone who claims to know is a liar".
Ralph Hartley
Fizz Fann
> I think I'm starting to see where your problem is.
Oh goody.
> You think that things like "I would have gotten x" and "his results depend
> on my choice" are statements of fact, that are objectively true or false,
> and not just ways of looking at things, dependent on your point of view.
I am aware of, and have noted in the detailed description of the setup,
the assumption of "counter-factual definiteness", which is not merely
a bizarre and stupid "way of looking at things", or my "problem", as
you present it. It is also assumed in Bell's and Eberhard's proofs
that quantum mechanics is nonlocal. Stapp, who introduced the concept,
has the following to say about it:
"[the] condition can be converted from a requirement that a
contingent nonstatistical theory - or law of nature - exists,
to the requirement that it is possible to assume that an
unperformed measurement would have had some definite result
if it had been performed. Stated more forcefully, [the]
condition requires only that if the experimenters had actually
adjusted the mechanical devices to give the alternative
experimental setup, then these alternative experiments would
have had certain definite results."
The assertion that the condition is false is equivalent to the
statement that if the experimenters had actually
adjusted the mechanical devices to give the alternative
experimental setup, then those alternative experiments would
NOT have had definite results.
Perhaps this is what you believe, but it is more precise than
a mere "way of looking at things". One can precisely formulate
a question, stating the assumption in the formulation, and one
shouldn't expect to be told that the question shouldn't
have been asked because it is based on a problematical way of
looking at things.
Indeed, although I don't particularly want to defend CFD too
much, I will at least say that, from the point of view of a being
who is living inside the physical world, who makes decisions
and then sees the consequences, CFD will appear to be true.
You will not achieve much if you do not at least work with the
assumption that your actions have consequences. Those who reason
"If I eat, I will not be hungry and if I don't eat, I will be
hungry" will be hungry less often than those who reason "Whatever
will be, will be, and whatever will not be, is not a suitable
subject for consideration."
> This might very well be so. There are interpretations of QM in which
> it is. If you demand that a interpretation not force you to "change
> the way you think about logic and truth" (paraphrasing your expired
> post), then you are limiting yourself to one of those.
So you are either advocating the abandonment of the notions of logic
and truth as mathematicians and logicians understand it, or are claiming
that I understand it in a different way than they do. Please tell me
which is the case.
> However, the fact that QM violates Bell's inequality *does* mean that you
> have to give *something* up.
Agreed.
> I'm pretty sure that the thing you have to lose to satisfy your
> "demands" is Lorenz invariance. All such interpretations have a
> "preferred" frame (*which* frame is preferred is, unfortunately(?),
> unobservable). Bohmian mechanics, for example, has this property.
You are then stating that (you are pretty sure that) to keep
counterfactual definiteness one must drop Lorentz invariance. If
that were true, that would be something of a significant result,
I think. On the other hand, I don't believe it. If there were
an additional rule to choreograph nonlocal effects in different
frames as in the example of the causality loop, then it would
need to be noncausal, since the results obtained in frame F1 would
have to have a dependence on the settings of the device in
experiment E2. That is, there is the additional hidden assumption
that causality should hold in both frames in the question which
I hadn't noticed before. So perhaps the statement might be that
one must drop one of the following: counterfactual definiteness,
causality, or Lorentz invariance.
> It is *not* ok, in your interpretation, to figure out what affected
> what in any frame other than the the preferred one. The frame will
> not matter as far as predicting the results of experiments, so if
> you use the wrong one you can't be proven wrong, but if you use two
> *different* frames they can't *both* be the preferred frame, so one
> or more of them may give the wrong answer, causing an apparent
> paradox.
I'm pleased to see that you finally saw the paradox.
> > Again, your problem is that you are considering his result as though
> > it were more firmly writ in stone than my decision about what to
> > measure.
> But it is, if his measurement happens *before* your decision. What he got
> affects the correctness of the guess you will make (though not your guess
> itself). His measurement is already made, nothing you can do can affect it!
In the case of timelike separation, this is kind of true. You might say
that it's a way of looking at it. When he made his measurement, he got
a result, and we don't know what that result depended on at the time.
If nature is nonlocal, it can have depended on things far away. If
nature is noncausal, it can have depended on things in the future.
> If the separation between the measurements is timelike (so his measurement
> happened *absolutely* before yours), this is the *only* possibility. So
> don't say that your settings affect his result even if his measurement came
> first! If you did you would also have to say the your settings affect his
> measurement even if you aready looked over his sholder and saw his results,
> before making your settings.
That's an interesting point, and under such circumstances, we can see
precisely the effect that I was talking about. The degree to which Bell's
inequalities is violated is _lessened_ if we try to construct a paradox
using the noncausal idea that events in the future can affect the past.
I do not advocate this idea, so please don't try to convince me that
I'm stupid for mentioning it.
Instead, suppose for a moment that we consider the possibility that
Bob, who looks over Alice's shoulder as you said and sees her result,
can in some way be partially responsible for it by virtue of his future
actions. Let Alice measure along the w axis (midway between x and y)
and suppose that sometimes (sometimes!) it is the case that if Bob
measures along y then Alice would have gotten "down" while if Bob
measures along x then Alice would have gotten "up". Then Bob tries
to use these instances to cause a paradox as follows:
If he sees that Alice got "up", he then goes and measures along y,
and if he sees that Alice got "down", he then goes and measures along
x, trying to change the past and unravel the fabric of spacetime
or achieve whatever nefarious nihilistic ambitions he aspires to.
Well, what happens to the violation of Bell's inequalities as
detected by Alice and Bob? It is this violation which might
allow the paradox, since it gives Bob some influence over Alice's
results. So if the paradox is to be prevented, the degree of
violation must decrease, and indeed it does.
A version of Bell's inequality reads:
(x+,y+) <= (x+,w+) + (w+,y+)
Where (p+,q+) means the fraction of events during which Alice measured
along axis p and got "up" and Bob measured along axis q and got "up".
Well, whenever Alice measures w and gets "up", Bob has (by arrangement)
measured along y, so the overall fraction of events in which (w+,y+)
occurred has increased, at the expense of the number of events of
the type (w+,x+) or (w+, x-). The right hand side of the inequality
becomes bigger overall, so the degree of violation is lessened.
Bob's behaviour is _destroying_ part of the observed nonlocal effect,
in this case the noncausal part of it, by trying to use the
noncausal effect to cause a paradox. If he simply stops his abusive
behavior of correlating his choice of measurement axis with Alice's
results, then the nonlocal effect returns in its full glory.
Now, I'm not saying that this is the best "way of looking at it",
but it is at least not ridiculous to think that there is a noncausal
effect which Bob exerts on Alice's result, which disappears when
he tries to use it to cause a paradox, but which reappears when he
stops.
That's all fine, but it's only a way of looking at it. Nevertheless,
the way of looking at it predicts that if a certain type of behavior
is followed by the participants, then the degree to which Bell's
inequalities will be violated must decrease, and it turns out that
it does in fact decrease. One might say that it's only because
we're choosing to measure certain things and it is our choice of
measurement which is interfering with our ability to observe the
"true" extent of the violation of the inequalities, or one might
say that the nonlocal (or in this case noncausal) effects have
disappeared when their appearance would have caused a paradox.
Now, in the case of the causality loop, with the two EPR experiments,
there is no simple correlation of Alice's result with the axis
which Bob chooses to measure along, as there is in this
"looking over the shoulder" experiment. One cannot simply dismiss
the effect by saying that the choice of measurement axis is
such that we will not see the true extent of the violation. The
experimenters can follow the rules which I described during a
specified subset of the normal measurements involved in an
EPR experiment (or two, in this case), without in any way affecting
the proportion of measurements in which Alice measures w and Bob
measures y, for example. That is, they can do two "normal" EPR
experiments but get the nonlocal effects of the two to fight
against each other.
So the question is, will the extent to which Bell's inequalities
is violated in these two experiments be lessened? If we admit
counter-factual definiteness and say that the cases in which
paradoxes would occur are excluded, then it should be like the
above non-causal case: those nonlocal effects which allow paradoxes
must be excluded and Bell's inequalities must not be violated
quite as much as usual. So this is a testable experimental prediction
and perhaps it's not just a way of looking at it after all.
> If you think that one of the two statements "your measurement affected his
> result" and "his measurement affected the correctness of your guess" is
> true, and the other false, then you *must* think that there is a "true"
> reference frame, since one statement is true in some frames, and the other
> in others.
You were pretty close to understanding, but now you're drifting away.
You see, the nonlocal "relationship" is between the setting of a
measuring device and the result of a faraway measurement. It's not
between the result of a measurement and the correctness of a guess.
There is no guess and there is no correctness. There are only
settings and results. If we are saying that something is affecting
something else, each of those somethings must be a setting or a
result. For reasons I have already given, involving cogs and gears,
only results and not settings can be affected.
You are clinging very dearly to an image of the world in which a
three-dimensional hyperplane wafts through four-dimensional space,
bringing the things it passes out of the realm of mere possibility
and into the realm of actuality. That is why you are trying to
imagine that his measurement result can be somehow affecting my
setting, or some property of it (which you imagine to be its
"correctness"). But no; the setting was decided long long ago
by the initial configuration of some gears and cogs which obey
deterministic physics, and the measurement result is dependent
on who knows what, but whatever it is, it's nonlocal, and from
the indications of quantum theory and experiments, the measurement
result depends on the setting of my measurement device, among
other things. This is true even if somebody somewhere (God or
somebody like that) prefers a three-dimensional frame which
wafts past the measurement result first.
> And *that's* why I think such interpretations are "ugly".
It doesn't have to be so. The real problem is with viewing
causation as going from past to future. If you agree that
a system obeying deterministic physics (whose behavior
we can and do predict perfectly until t=+infinity) isn't "caused"
to do anything by a system obeying unpredictable physics, then
the resulting picture isn't a preferred frame. Causality goes
firmly along the direction of deterministic, predictable events,
but might hop backwards in time (in some frames) if the
unpredictable events have undeniable statistical relationships
to the deterministic systems.
Best wishes,
Alex.
Ilja Schmelzer
> You think that things like "I would have gotten x" and "his results
> depend on my choice" are statements of fact, that are objectively
> true or false, and not just ways of looking at things, dependent on
> your point of view. This might very well be so. There are
> interpretations of QM in which it is. If you demand that a
> interpretation not force you to "change the way you think about
> logic and truth" (paraphrasing your expired post), then you are
> limiting yourself to one of those.
A nice description of my position too.
Indeed, I don't plan to change the way I think about logic and truth
if confronted with such a trivial thing like some observation. I
prefer to give up some particular scientific theories as falsified.
> It doesn't matter to this argument, but I assume you also believe
> that there is a *true* interpretation. It may be impossible to find
> out which that is, since they are experimentally indistinguishable.
Exactly.
> Whether this position is true, or even meaningful, or not is a
> philosophical issue, which I will not discuss further.
SCNR: At least the position that it is true seems consistent.
Assuming it is false, we have some false philosophical interpretation
about our world. Given that philosophical interpretations about our
world may be false, therefore also true, assuming that interpretations
of QM cannot be false or true sounds strange. The philosophical
thesis that such philosophical claims are meaningless does not seem to
be meaningful too.
> However, the fact that QM violates Bell's inequality *does* mean
> that you have to give *something* up.
Indeed.
> I'm pretty sure that the thing you have to lose to satisfy your
> "demands" is Lorenz invariance.
And, indeed, I'm giving up Lorentz invariance.
If I chose giving up "the way I think about logic and truth" in case
of conflict between "the way I think about logic and truth" and
Lorentz invariance, it means Lorentz invariance would be something
like a religious dogma for me. This would be the end of science. If
somebody agrees to give up "the way he thinks about logic and truth"
to preserve some particular scientific theory against experimental
falsification, how else should we describe this?
Instead, I insist that preserving "the way I think about logic and
truth", whatever the experiment shows, is not a religious dogma, but
simply the scientific method. There is, indeed, no imaginable outcome
of an experiment which could force me to give up "the way I think
about logic and truth". In the worst case I would conclude only that
I have no theory which allows to explain the given experiment - which
is a quite common state of science.
> All such interpretations have a "preferred" frame (*which* frame is
> preferred is, unfortunately(?), unobservable). Bohmian mechanics,
> for example, has this property.
Yep.
> It is *not* ok, in your interpretation, to figure out what affected
> what in any frame other than the the preferred one. The frame will
> not matter as far as predicting the results of experiments, so if
> you use the wrong one you can't be proven wrong, but if you use two
> *different* frames they can't *both* be the preferred frame, so one
> or more of them may give the wrong answer, causing an apparent
> paradox.
Yep.
> If you think that one of the two statements "your measurement
> affected his result" and "his measurement affected the correctness
> of your guess" is true, and the other false, then you *must* think
> that there is a "true" reference frame, since one statement is true
> in some frames, and the other in others.
Yep.
> It doesn't matter at all in what reference frames the various
> experimenters and pieces of equipment are at rest. What causes what
> is different in different reference frames, so if you think that
> causation goes in a particular direction, you must think that there
> is particular *fixed* frame to use when determining that, and that
> all other frames are "wrong".
Now, causation (at least in general) has some particular direction:
The direction from a "free will" (or however randomly chosen) decision
(what to measure) to the outcome of the experiments.
> Someone accelerating from the "right" frame to a "wrong" one can't
> change the direction of causation.
That would be, indeed, strange.
> And *that's* why I think such interpretations are "ugly".
Is "ugly" a good reason to give up "the way I think about logic and
truth"? Not for me. "Ugly" is, even in the worst case, only a
challenge to find something more beautiful.
Fizz Fann
> Fizz Fann wrote:
> > Ok. Here is an attempt at a more precise description of everything
> > involved in the construction (with diagrams!).
>
> [Snipping most of the description]
>
> Quantum mechanics (and the setup you described) gives us the probabilities
> of each:
>
> P(up,up,dn,up,w,w) = 1/4 * (1-0.853)*0.853 = 0.0313
> P(up,up,dn,dn,w,w) = 1/4 * (1-0.853)*(1-0.853) = 0.0054
> P(up,dn,up,dn,s,s) = 0.25
> P(up,dn,dn,up,w,s) = 1/4 * 0.853 = 0.2132
> P(dn,up,up,up,s,w) = 1/4 * (1-0.853) = 0.0367
> P(dn,up,up,dn,s,w) = 0.2132
> P(dn,up,dn,up,w,w) = 1/4 0.853*0.853 = 0.1819
> P(dn,up,dn,dn,w,w) = 0.0313
> P(dn,dn,dn,up,w,s) = 0.0367
>
> and all the rest are 0.
Agreed. This is what happens when you use the bare formalism of
quantum mechanics along with the statements that E1a=up -> S2=w
and so on and follow them to their logical conclusion.
> Thus all the outcomes have a fixed, known, probability, and they add
> up to 1.
Agreed.
> If you do this experiment many times, these are the outcomes you
> will see.
Bell's inequalities were considered significant because if they were
violated, as quantum mechanics predicted, then a particular view of
the world, that of a world of local causes, was no longer tenable.
The significance of the violation was not to be found by staring
at the numbers on their own, so those who merely stared at
the numbers and never thought about what implications they held
for the way the world works never realised the significance
of nonlocality.
Now you are behaving exactly that way, looking only at the numbers
and refusing to consider the fact that we try to understand the
world in terms of concepts like "Event 1 happened, and it is the
case that if event 1 happens then event 2 will happen while if
event 1 doesn't happen, then event 3 will happen, so event 1
caused event 2 rather than event 3".
> Nothing in the above numbers is any kind of paradox or logical
> contradiction. If all your talk of "causation" and "what would have
> happened" gives some sort of contradiction, you must have created
> the contradiction, because there is none in the actual outcomes of
> the experiment.
Nowhere have I said that I expect a paradox to occur in nature.
I have said that if old Alice affects young Bob's result and
old Bob affects young Alice's result then the two could conspire
to try to cause a paradox and they must not be successful. Your
amazing display of mathematical prowess does not address the
question. Obviously there is no paradox in mathematics or
quantum theory and you are not going to find one by looking
for it in the mathematics.
There are two nonlocal effects. If both are at work simultaneously,
then there is a paradox. Therefore they are not both at work
simultaneously. Writing this fact down numerically does not clarify
the situation.
> > There are two experimenters (Alice and Bob) and two experiments,
> > E1 and E2, performed in frames F1 and F2 respectively.
>
> What do you mean by "performed in frames"?
I mean that the experimenters and apparatus involved were in those
inertial frames when the experiments were performed.
> The relative motions of the apparatus or experimenters don't change
> the results shown in the table above.
No, but they change the physical significance of the results.
> > Now, if N is sufficiently large, then unless there is some
> > extra rule to prevent this from happening, there will be
> > some n for which both nonlocal effect 1 and nonlocal effect
> > 2 are at work, and then we would have
> >
> > A(n)=>not(A(n))=>A(n), which is a paradox.
> Which values of n?
>
> Ok, suppose you have N trials, and observed the results as
> above. Lets call the trials where effect 1 happened T1, and those in
> which effect 2 happened T2, and the rest T0. Show me how you can
> assign trials to these three sets so that T1 and T2 have trials in
> common.
Such an assignment cannot be made. I said myself that it would be a
paradox. If you think I have claimed somewhere that such an assignment
could be made then you are mistaken.
> In other words show me a trial in which A->not(A)->A. This isn't just
> because you don't *know* which trials they were. There is just no
> consistent way to make such an assignment.
Again, you are mistaken. You think that I have claimed that there
is a paradox which doesn't lead to a logical contradiction. I don't.
The real question is whether _nature_ has made such an assignment of
trials to T0, T1, and T2 or whether making any such assignment
ourselves would be merely an exercise in reassuring ourselves that
it's possible to make sense of the world. Perhaps you believe the
latter, but it is reasonable to investigate the possible consequences
of the former.
> > To prevent this, as I understand it, there would need to be
> > a rule saying that for a particular value of n, nonlocal
> > effect 1 can be at work, or nonlocal effect 2 can be at
> > work, but not both.
> Yes, there are such rules. They are the rules of logic.
The rules of logic are those rules by which we _deduce_
that both effects are not active simultaneously. With
a different set of physical laws, it could have been the case
that both effects could indeed be present simultaneously, but
that Bob might have been prevented from behaving as described
by some other physical law. We recognise that, since we are
unwilling to sacrifice Bob's or Alice's ability to behave
as described, we must accept that both nonlocal effects
cannot be at work during the same trial. However, presumably
nature does not work through the same process of deduction
as we do. The reasons which nature has for not allowing
the two effects to occur simultaneously have nothing to do
with holding Bob's or Alice's freedom in high esteem.
> That isn't to say that there isn't something funny going on here. The
> violation of Bell's inequality *does* mean you have to give up some
> assumption.
>
> What assumption do *I* discard?
>
> It depends on what problem I am trying to solve, or what mood I am
> in. Some choices are:
>
> The wavefunction collapses instantly, "what if" questions are meaningful,
> but only in one fixed frame.
>
> The wavefunction collapses instantly in *any* frame, but "what if"
> questions are meaningless, or at least relative. The measurement creates
> the measured value, it doesn't have any "pre-existing" value.
I don't completely understand whether you are saying that these
are positions that you choose to adopt or assumptions that you
choose to abandon. In any case, they don't look like assumptions of
Bell's theorem to me.
Those assumptions were, as I understand it:
Local causality
Counter-factual definiteness
Hidden variables (later shown to be unnecessary by Eberhard)
Freedom of the experimenters to set the detectors whatever way they wish.
Incidentally, the idea that "The measurement creates
the measured value, it doesn't have any "pre-existing" value"
doesn't help. It does not allow you to save local causality
since the process which determines the newborn value must take
into account the situation at the distant detector in order
to give the predicted results.
> No collapse at all. Bob and Alice become entangled with the particle
> states they measure, and none of the outcomes (with positive
> probability) are ever any more "real" than any others.
Many worlds. I'm still unsure of what I think about this interpretation.
I guess I'll decide when somebody tells me how to define any physical
thing at all based purely on the Hamiltonian and the state vector. It's
hard to see how this interpretation deals with nonlocality when space
itself doesn't appear in the mathematics.
> Any of these possibilities (and more) give exactly the same outcomes. You
> might ask which I believe, but I don't have to choose.
I wouldn't ask you to choose from among those. I would ask you what
you think of nonlocality, though, which is not addressed by any of
those approaches. Each of those is a way of thinking about the
mathematical formalism - not a way of thinking about the physical
world which you see in front of you.
> There is *no* possibility that Quantum mechanis is correct, but some
> experiment will eventually tell us what its *true* interpretation is.
How do you know?
> I remember in the Koran (which I read in translation a *very* long time
> ago, so I may have this wrong) there is a story about five boys who were
> placed in eternal(?) sleep in a cave. Some versions of the story say there
> were five including their dog, and others that the dog would make six. But
> God says, "Only I know how many there were, and I haven't told anybody.
> Anyone who claims to know is a liar".
:) Perhaps if we pray, then God will tell us the correct interpretation
of quantum mechanics.
Best wishes,
Alex.
Ilja Schmelzer
> has the following to say about it:
>
> "[the] condition can be converted from a requirement that a
> contingent nonstatistical theory - or law of nature - exists,
> to the requirement that it is possible to assume that an
> unperformed measurement would have had some definite result
> if it had been performed. Stated more forcefully, [the]
> condition requires only that if the experimenters had actually
> adjusted the mechanical devices to give the alternative
> experimental setup, then these alternative experiments would
> have had certain definite results."
>
> The assertion that the condition is false is equivalent to the
> statement that if the experimenters had actually
> adjusted the mechanical devices to give the alternative
> experimental setup, then those alternative experiments would
> NOT have had definite results.
But realism covers also a lot of situations where CFD, as described
here, does not hold.
Assume you have a black box, with somebody inside. If the
experimenter presses a given button, then the guy inside makes some
random choice (free will, roulette, quantum experiment, or a
combination of them) to obtain the result.
AFAIU the phrase "has no definite results" in the description above,
this would be an example of non-CFD. But using such random methods
you cannot violate Bell's inequality. And Bell's theorem (which
assumes that some probability distribution over the possible states of
reality is given) covers this simple type of random hidden variables.
Ralph Hartley
>>
>> It depends on what problem I am trying to solve, or what mood I am in.
>> Some choices are:
>
> positions that you choose to adopt or assumptions that you choose to
> abandon.
They are (perhaps overbrief) summaries of possitions I might take, when
abandoning some of the assumptions.
> Those assumptions were, as I understand it: Local causality
> Counter-factual definiteness Hidden variables (later shown to be
> unnecessary by Eberhard) Freedom of the experimenters to set the
> detectors whatever way they wish.
And a few others as well.
> Incidentally, the idea that "The measurement creates the measured value,
> it doesn't have any "pre-existing" value" doesn't help.
Perhaps I phrased it poorly. I meant to deny counter-factual definiteness.
The position is that the concept of "what would have happened if I did
measurement other than the one I did", is either invalid, relative (e.g.
frame dependant) or subjective.
> It does not allow you to save local causality since the process which
> determines the newborn value must take into account the situation at the
> distant detector in order to give the predicted results.
Or the distant detector could take *my* results into account. That still
isn't purely local, but that view makes the limited nature of the
non-locality a bit more clear. All the non-locality is in the form of a
correlation, and the direction of causality is always ambiguous.
I preffer to define "causalty" as the ability to send a message (with
results predictably better than chance). I kow you agrre that by that
definition, causality is local.
>> No collapse at all. Bob and Alice become entangled with the particle
>> states they measure, and none of the outcomes (with positive
>> probability) are ever any more "real" than any others.
>
> Many worlds. I'm still unsure of what I think about this interpretation.
> I guess I'll decide when somebody tells me how to define any physical
> thing at all based purely on the Hamiltonian and the state vector. It's
> hard to see how this interpretation deals with nonlocality when space
> itself doesn't appear in the mathematics.
Space appears in the Hamiltonian. It is possible to have a Hamiltonian that
is non-local, but all the evidence seems to indicate that it *is* local.
The interpretation *permits* does not *reqire* locality.
The Everret interpretation (forget the term "many worlds", it isn't really
accurate) keeps (or at least is compatible with) all the assumptions above,
and the additional assumption that there is an independant reality, but
denys that "the one unique result I got" is an element of that reality.
It's *pretty*, because not only is it local, it's even linear.
Consider a world in which the wavefunction collapse was real but allways
happens one second *after* the measurement (I think Roger Penrose has
suggested that something like this really happens). What would such a world
look like? A little reflection shows that (at least from the inside) it
would look just the same. The experimenter would be in a superposition of
states for a second, but then the wavefunction collapses, and he only
remembers a unique outcome.
Now suppose that the delay between measurement and collapse is a *day*. Is
there any experiment the inhbitants of this world could do to tell that
this is happening? There is not. At least with respect to last weeks
experiments (it takes a day for the results get written up), there is no
way to tell the difference between immediate collapse of the wavefunction
and entanglement between the experimenter and the thing that was measured.
Bell's inequality, for example,is violated in exactly the same way.
Nor could any experiment tell if the collapse was delayed a billion years.
So at least as a limit, we can see that it doesn't really matter (in any
observable way) if the wavefunction collapse *never* happens.
>> There is *no* possibility that Quantum mechanis is correct, but some
>> experiment will eventually tell us what its *true* interpretation is.
>
> How do you know?
Because the different interpretations all use the same math to predict the
results of any experiment. So they all agree on any experiment.
>> I remember in the Koran
[which I did misquote but not in a way material to my point]
>> there is a story about five boys who were placed in eternal(?) sleep
>> in a cave. Some versions of the story say there were five including
>> their dog, and others that the dog would make six. But God says, "Only
>> I know how many there were, and I haven't told anybody. Any [man] who
>> claims to know is a liar".
>
> :) Perhaps if we pray, then God will tell us the correct interpretation
> of quantum mechanics.
Perhaps, but that would be out of charachter.
Ralph Hartley
Ralph Hartley
> Ralph Hartley:
>> Fizz Fann wrote:
>
>>> Ok. Here is an attempt at a more precise description of everything
>>> involved in the construction (with diagrams!).
>>
>
> Agreed.
>
>> Thus all the outcomes have a fixed, known, probability, and they add
>> up to 1.
>
> Agreed.
I thought you would. I hope you didn't think I was trying to put words into
your mouth implying that you wouldn't.
> Bell's inequalities were considered significant because if they were
> violated, as quantum mechanics predicted, then a particular view of the
> world, that of a world of local causes, was no longer tenable.
Just so. There is a short list of principles that many people find self
evident (and somewhat longer list that are so internalized they are hard to
even state), and the experiment says you have to give at least one of them up.
> Now you are behaving exactly that way, looking only at the numbers and
> refusing to consider the fact that we try to understand the world in
> terms of concepts like "Event 1 happened, and it is the case that if
> event 1 happens then event 2 will happen while if event 1 doesn't
> happen, then event 3 will happen, so event 1 caused event 2 rather than
> event 3".
I'm not refusing to consider that. The point is that those concepts may not
be valid. If they aren't, then it should be no surprise that trying to
understand the world that way leads to contradictions.
In fact, those concepts *are* valid, but only in a fixed reference frame,
it doesn't matter which. Actually, I think what you need is a bit weeker
than a reference frame. Any consistant rule for determining which of two
events happened "first" will do. This is good, because in GR reference
frames are only local.
> Nowhere have I said that I expect a paradox to occur in nature. I have
> said that if old Alice affects young Bob's result and old Bob affects
> young Alice's result then the two could conspire to try to cause a
> paradox and they must not be successful.
I think it's possible that your paradox comes from trying to apply causal
reasoning in inconsistant frames.
Since past affects future, and not the other way around, it must be that
old Alice affects young Bob and young Alice affects old Bob, or that young
Bob affects old Alice and old Bob affects young Alice (and regardless,
young Bob affects old Bob and young Alice affects old Alice, but that's not
so controversial).
> There are two nonlocal effects. If both are at work simultaneously, then
> there is a paradox. Therefore they are not both at work simultaneously.
Just so, but on Tuesdays I believe in an interpretation that has no
nonlocal effects at all.
>> The relative motions of the apparatus or experimenters don't change
>> the results shown in the table above.
>
> No, but they change the physical significance of the results.
What do you mean by "physical significance"? I tend to thank of something
that doesn't affect the results of any possible experiment as something
that doesn't *have* any physical significance. (Of course the motions of
the experimenters are measurable, and so are physical, but I don't see how
they are significant to this discussion).
Do you think there should be a rule "Causality is well defined, and is
determined in the (possibly changing) reference frame of the
experimenters"? I don't think there is any interpretation of QM in which
that is true. Your paradox might be viewed as a (somewhat convoluted) proof
of that fact (a proof that might also be undermined by the conditional
nature of the "causation").
>>> then unless there is some extra rule to prevent this from happening
The word that makes this sentence wrong is *extra*. That A=>not(A) and
not(A)=>A can't both happen is implied by your *definition* of "=>". QM
doesn't have to do anything special to avoid an event that is impossible by
definition.
> The real question is whether _nature_ has made such an assignment of
> trials to T0, T1, and T2
Maybe it does, and maybe it doesn't, but it certianly doesn't need a rule
to keep from assigning a trial to both T0 and T1, since they are defined to
be disjoint sets.
Does there need to be a law of nature that prevents any male barber from
shaving all men in his town that don't shave themselves, and no one else?
(Old paradox: Does he shave himself?)
> or whether making any such assignment ourselves would be merely an
> exercise in reassuring ourselves that it's possible to make sense of the
> world. Perhaps you believe the latter
No, as I said before, I am agnostic, but I believe very strongly in my
agnositicism.
>>> To prevent this, as I understand it, there would need to be a rule
>>> saying that for a particular value of n, nonlocal effect 1 can be at
>>> work, or nonlocal effect 2 can be at work, but not both.
>
>> Yes, there are such rules. They are the rules of logic.
>
> The rules of logic are those rules by which we _deduce_ that both
> effects are not active simultaneously.
But if your definition of "->" implies that A->not(A) and not(A)->A can't
both hold (I think it does), then they can't both hold *regardless* of what
the laws of nature are.
> but that Bob might have been prevented from behaving as described by
> some other physical law.
But we have lots of experimental observations of classical systems, that
lead us to belive that it *is* possible to behave that way.
Of course you *could* give up one of those "principles that are so
internalized they are hard to even state." But since there are (many)
perfectly good interpetations of QM (though perhaps none you find totally
convincing), that are consistant with the apparent freedom to act that way,
I see no reason to bother with them.
Ralph Hartley
Fizz Fann
Ilja Schmelzer <schm...@wias-berlin.de> wrote in message news:<i3gadey...@wias-berlin.de>...
> thephy...@yahoo.com (Fizz Fann) writes:
> >
> > The assertion that the condition is false is equivalent to the
> > statement that if the experimenters had actually
> > adjusted the mechanical devices to give the alternative
> > experimental setup, then those alternative experiments would
> > NOT have had definite results.
>
> But realism covers also a lot of situations where CFD, as described
> here, does not hold.
>
> Assume you have a black box, with somebody inside. If the
> experimenter presses a given button, then the guy inside makes some
> random choice (free will, roulette, quantum experiment, or a
> combination of them) to obtain the result.
One might also speculate that whether a particle is registered at
all depends heavily on the alignment of the SG magnet. Then it becomes
the case that sometimes for certain alignments which weren't used,
neither up nor down would have been observed. As you say, though, it
doesn't help you avoid nonlocality because whether a particle is
registered at a particular detector would have to depend on the
situation at the other detector. You can also count "no event observed"
as a result and then you're back to where you started.
> AFAIU the phrase "has no definite results" in the description above,
> this would be an example of non-CFD.
It's not really so simple. The little man in the box doesn't really
have any source of "pure" randomness to draw on which is known to be
more unspeakable than a pseudo-random number generator. That is, inside
the box there could just as well have been a computer outputting
/dev/random and pressing the button might just show you the most
recently output byte. As you know from Bohmian mechanics, quantum
events aren't really known to be any more "inherently random"
than a computer's random number generator, and the rest of physics,
including the "free will" of the experimenters, is just the same.
So I would say that even in this case, if I don't press the button,
there's still a "what I would have gotten" if I had.
Really the only way to definitely have no CFD is to have the random
number generator in the box seeded by the conditions inside my brain,
so that my decisions to press or not to press the button are entwined
with "what I would have gotten". In that case, you can't change my
decision without also changing the random number that I would have
got. There may be _several_ ways of changing my decision, and each of
these could have caused different seedings of the random number
generator. In that sense there is no simple "what I would have got
if I had pressed the button", but there _is_ a "what I would have
got if I had pressed the button because I was bored" or a "what
I would have got if I had pressed the button accidentally", and these
can be different things, since the internal workings of the box
can depend on whether I'm bored or clumsy (maybe not because they
can sense it, but perhaps because in order for me to be bored at
that time, something else in my past would have had to have been
different, and that thing can affect the internal workings of the
box too).
That is the only circumstance I can see that CFD can reasonably
be said to be unacceptable.
> But using such random methods
> you cannot violate Bell's inequality. And Bell's theorem (which
> assumes that some probability distribution over the possible states of
> reality is given) covers this simple type of random hidden variables.
Right, but a loophole in the theorem is hyperdeterminism, and that is
precisely the case in which CFD fails as I described above.
I don't believe it, of course. Spooky action at a distance is more
acceptable.
Regards,
Alex.
Fizz Fann
Ralph Hartley <har...@aic.nrl.navy.mil> wrote in message news:<b745be$958$2...@ra.nrl.navy.mil>...
>
> > Incidentally, the idea that "The measurement creates the measured value,
> > it doesn't have any "pre-existing" value" doesn't help.
>
> Perhaps I phrased it poorly. I meant to deny counter-factual definiteness.
> The position is that the concept of "what would have happened if I did
> measurement other than the one I did", is either invalid, relative (e.g.
> frame dependant) or subjective.
Now that's a terrible position to find yourself in.
> > It does not allow you to save local causality since the process which
> > determines the newborn value must take into account the situation at the
> > distant detector in order to give the predicted results.
>
> Or the distant detector could take *my* results into account. That still
> isn't purely local, but that view makes the limited nature of the
> non-locality a bit more clear. All the non-locality is in the form of a
> correlation, and the direction of causality is always ambiguous.
Right. So the thing is that in each case we can't quite say whether
the overall statistics appeared because I was affecting your results
or because you were affecting mine, but they indicate that at least
one if not both of those are true. If the world is really
Lorentz-invariant then we should expect that it's a little bit of
both.
But with my set up, here's what we can do: we can block one of the
options. Or rather, we can prevent the situation where Alice affects
Bob's results in E1 _and_ Bob affects Alice's results in E2. Not
so amazingly spectacular maybe, but we can also _measure_ the overall
extent of the nonlocality, by seeing how much Bell's inequalities are
violated. If we suppose that there are two contributions to the
nonlocal effect - one from Alice's effect on Bob's results and one
from Bob's effect on Alice's, then when we block one of those channels,
at least partially, statistically, we might expect to see less of
an overall nonlocal effect.
> I preffer to define "causalty" as the ability to send a message (with
> results predictably better than chance). I kow you agrre that by that
> definition, causality is local.
Right; I won't disagree there.
> >> No collapse at all. Bob and Alice become entangled with the particle
> >> states they measure, and none of the outcomes (with positive
> >> probability) are ever any more "real" than any others.
> >
> > Many worlds. I'm still unsure of what I think about this interpretation.
> > I guess I'll decide when somebody tells me how to define any physical
> > thing at all based purely on the Hamiltonian and the state vector. It's
> > hard to see how this interpretation deals with nonlocality when space
> > itself doesn't appear in the mathematics.
>
> Space appears in the Hamiltonian. It is possible to have a Hamiltonian that
> is non-local, but all the evidence seems to indicate that it *is* local.
> The interpretation *permits* does not *reqire* locality.
Space appears in the expression for the Hamiltonian. That is, we
write H=p^2/2m + V(x), but the operators p and x don't belong in
the Everett interpretation because they are defined in terms
of measurements and collapses and so on. The only operator in
Everett is H.
> The Everret interpretation (forget the term "many worlds", it isn't really
> accurate) keeps (or at least is compatible with) all the assumptions above,
> and the additional assumption that there is an independant reality, but
> denys that "the one unique result I got" is an element of that reality.
> It's *pretty*, because not only is it local, it's even linear.
Everett doesn't have an axiom that there is a spatial structure
anywhere. All there is is H and |psi>. No space, no particles,
no results of measurements.
> Consider a world in which the wavefunction collapse was real but allways
> happens one second *after* the measurement ...
> Nor could any experiment tell if the collapse was delayed a billion years.
We can speculate that only the present exists and the past never did,
but we can't speculate that only the future, a billion years from
now, exists and that we don't. If there were never any well-defined
structure corresponding to my body in physical space, then I would
not be having this correspondance, and the fact that a skeleton
consistent with my having once existed might be dug up in a future
better-defined world would not make my experience of now any more real.
> So at least as a limit, we can see that it doesn't really matter (in any
> observable way) if the wavefunction collapse *never* happens.
It does matter - if the only mathematical things which have any reality
at all are the Hamiltonian and the state vector, then one must accept
that there is no special spatial structure in the universe - that,
for example, the p-representation is just as good as the x-representation,
since both are added in by hand and have nothing to do with H or
|psi>. In qm, x is an operator whose eigenvalues and eigenvectors
are defined in terms of a measurement which you must already know
how to do, or it is a parameter appearing in the action which is
imported from classical mechanics, where you already know what it
means. You can't derive it from anywhere in quantum mechanics from
any more basic mathematical structure.
> >> There is *no* possibility that Quantum mechanis is correct, but some
> >> experiment will eventually tell us what its *true* interpretation is.
> >
> > How do you know?
>
> Because the different interpretations all use the same math to predict the
> results of any experiment. So they all agree on any experiment.
That's true for nonrelativistic quantum mechanics, but the different
interpretations might lead to mathematically distinct relativistic
theories. I get the feeling that it's been called philosophy for so
long that everybody assumes without justification that it will never
be testable. Thinking about whether the earth was flat or round, or
wondering what chemicals the stars were made of, were once idle
philosophical pursuits with no hope of experimental resolution.
> > :) Perhaps if we pray, then God will tell us the correct interpretation
> > of quantum mechanics.
>
> Perhaps, but that would be out of charachter.
It's probably been tried before too.
Best wishes,
Alex.
Fizz Fann
> Fizz Fann wrote:
> > Ralph Hartley:
> >
> > Agreed.
>
> I thought you would. I hope you didn't think I was trying to put words into
> your mouth implying that you wouldn't.
Not at all. I did think you were treating it as one might dismiss
nonlocality by saying that those are the predictions and that's
all there is.
> > Bell's inequalities were considered significant because if they were
> > violated, as quantum mechanics predicted, then a particular view of the
> > world, that of a world of local causes, was no longer tenable.
>
> Just so. There is a short list of principles that many people find self
> evident (and somewhat longer list that are so internalized they are hard to
> even state), and the experiment says you have to give at least one of them up.
Can you (to whatever extent you think it's reasonable) give a list of
some of the other assumptions you think are necessary for somebody
to accept the proof?
> > Now you are behaving exactly that way, looking only at the numbers and
> > refusing to consider the fact that we try to understand the world in
> > terms of concepts like "Event 1 happened, and it is the case that if
> > event 1 happens then event 2 will happen while if event 1 doesn't
> > happen, then event 3 will happen, so event 1 caused event 2 rather than
> > event 3".
>
> I'm not refusing to consider that. The point is that those concepts may not
> be valid. If they aren't, then it should be no surprise that trying to
> understand the world that way leads to contradictions.
Right; they may not be valid. In my opinion it's important for us to
continue using these concepts unless we're absolutely forced to
abandon them. Quantum mechanics doesn't force us to abandon them,
but it makes us accept very strange things if we want to keep them.
> In fact, those concepts *are* valid, but only in a fixed reference frame,
> it doesn't matter which. Actually, I think what you need is a bit weeker
> than a reference frame. Any consistant rule for determining which of two
> events happened "first" will do. This is good, because in GR reference
> frames are only local.
Right. This is part of what I'm saying. The rule for determining which
thing happened "first" that I would give (in cases where the usual
ideas break down) would be to say that if some physical process is
obeying deterministic physics (like gears and cogs) and can be predicted
perfectly for the next ten years, then the behavior of that process
seven years from now is "before" the behavior of a nondeterministic
system five years from now. That is, the settings in EPR experiments,
which can be controlled by gears and cogs, are "before" the results,
regardless of the frames in which the experiments are done.
I'm trying to make a consistent generalisation of the notion that
if X happens before Y then X can be said to affect Y but Y can't
be said to affect X. When X happens at an earlier time, it's
trivial. When X and Y are separated by a spacelike distance, though,
you need to do something else. Or rather, you need to do something
_if_ the statistical results can't be explained without admitting
a causal relationship between the two.
> > Nowhere have I said that I expect a paradox to occur in nature. I have
> > said that if old Alice affects young Bob's result and old Bob affects
> > young Alice's result then the two could conspire to try to cause a
> > paradox and they must not be successful.
>
> I think it's possible that your paradox comes from trying to apply causal
> reasoning in inconsistant frames.
Well insofar as a frame is considered as a three-dimensional plane
separating things set in stone from things yet to be decided, then
that is certainly true. We shouldn't change our minds about what
has been set in stone forever halfway through an argument.
> > There are two nonlocal effects. If both are at work simultaneously, then
> > there is a paradox. Therefore they are not both at work simultaneously.
>
> Just so, but on Tuesdays I believe in an interpretation that has no
> nonlocal effects at all.
Conspiracies or many-worlds. Those are the only ones I've seen, and
I'm coming around to the point of view that Everett can't be
formulated in such a way that the person reading the formulation
can derive the existence of the physical world without already
knowing about it and using his knowledge.
> >> The relative motions of the apparatus or experimenters don't change
> >> the results shown in the table above.
> >
> > No, but they change the physical significance of the results.
>
> What do you mean by "physical significance"? I tend to thank of something
> that doesn't affect the results of any possible experiment as something
> that doesn't *have* any physical significance. (Of course the motions of
> the experimenters are measurable, and so are physical, but I don't see how
> they are significant to this discussion).
It's in the same sense as in the EPR experiment. Sure, qm gives those
predictions, but the physical significance is changed by the realisation
that the two experimenters can be spacelike separated. In this case,
the positions and relative motions of the experimenters allow them
to orchestrate their measurements in the way described, which couldn't
happen without the prescribed motions.
> Do you think there should be a rule "Causality is well defined, and is
> determined in the (possibly changing) reference frame of the
> experimenters"? I don't think there is any interpretation of QM in which
> that is true. Your paradox might be viewed as a (somewhat convoluted) proof
> of that fact (a proof that might also be undermined by the conditional
> nature of the "causation").
Well I don't think causality should have anything to do with reference
frames at all. For me, "X caused Y" means something like "From the
knowledge that X is true and confidence I have in whatever physical
laws are applicable, it is possible to deduce Y". That is, causality
goes from things we already know to things we have to deduce, which
is why it so often coincides with going from past to future.
> >>> then unless there is some extra rule to prevent this from happening
>
> The word that makes this sentence wrong is *extra*. That A=>not(A) and
> not(A)=>A can't both happen is implied by your *definition* of "=>". QM
> doesn't have to do anything special to avoid an event that is impossible by
> definition.
I think logical reasoning about nature is being confounded with
processes by which nature operates here. Think of the twin paradox in
special relativity. It indicated there was something wrong with the
logic involved - ie that the person looking at the paradox wasn't
using the theory properly. *But* what happens is that nature has
an "extra rule" which wasn't used in the statement of the paradox -
that is that strange things happen when the twin travelling
away from earth starts accelerating back towards earth (as far as
that twin is concerned). The twin paradox only appears to be a
paradox if you don't know this rule. As you might say, special
relativity doesn't have to do anything to avoid an event which is
impossible by definition (any paradox), but there is an effect or
rule of nature which intervenes to prevent the paradox and save
the day.
> Does there need to be a law of nature that prevents any male barber from
> shaving all men in his town that don't shave themselves, and no one else?
> (Old paradox: Does he shave himself?)
Right; another nice example. Imagine there is only one barber in a
town full of cleanly-shaven men. We follow him around a bit to see
what he does and we try to formulate a theory describing his
behavior. We observe a correlation between people who he shaves
and people who don't shave themselves. We produce for ourselves
a theory based on this observation - he shaves those who don't
shave themselves. Then I say "Wait a minute! If that rule is true
and he shaves himself, then that would cause a paradox! So there
must be _another rule_ to describe what happens when he gets
shaved." That is, the observed correlation can't be all there is
to the story.
> >>> To prevent this, as I understand it, there would need to be a rule
> >>> saying that for a particular value of n, nonlocal effect 1 can be at
> >>> work, or nonlocal effect 2 can be at work, but not both.
>
> >> Yes, there are such rules. They are the rules of logic.
> >
> > The rules of logic are those rules by which we _deduce_ that both
> > effects are not active simultaneously.
>
> But if your definition of "->" implies that A->not(A) and not(A)->A can't
> both hold (I think it does), then they can't both hold *regardless* of what
> the laws of nature are.
Yes, but this is deduction that we can do. I know they don't hold
because I've deduced it with these means. But the means by which
I deduce this are not a description of how nature operates. The
movements of Bob's arms and legs are described by physical laws
which I am not mentioning, and it is a relationship between those
physical laws and the other putative laws which determine the
results of measurements which prevents the paradox in "practice".
Best wishes,
Alex.
Ralph Hartley
> Instead, I insist that preserving "the way I think about logic and
> truth", whatever the experiment shows, is not a religious dogma, but
> simply the scientific method.
But some of the ways you think about logic and truth may be unjustified or
completely wrong! How is clinging to them against all evidence "simply the
scientific method"?
> There is, indeed, no imaginable outcome
> of an experiment which could force me to give up "the way I think
> about logic and truth".
So you would cling to them against all *imaginable* evidence?
Do you think that the way you think about logic and truth is logically
necessary? Do you think that there can't be other ways to think that are at
least as valid. Logic doesn't give *anything* for free, you have to start
with axioms.
People have been forced to change the way they think about such things
before, and it seems likely that they will again. An experiment might not
do it, but an unexpected theorm almost certianly would.
When it comes to physics, you also have to consider that though logic may
be logic, there is no unique way to apply it to the real world.
Suppose General Relativity was discoverd before non-euclidian geometry.
Would you refuse to accept it because you would have to change the way you
thought about points and lines?
There is *nothing* logically inconsitant about other interpretations of QM.
That they don't admit absolute answers to some questions you would like to
ask, shouldn't be held against them. After all, accepting other
interpretations only reqires changing how you *think* about logic and
truth. Presumably, logic and truth wouldn't be changed at all.
Ralph Hartley
Ilja Schmelzer
thephy...@yahoo.com (Fizz Fann) writes:
> One might also speculate that whether a particle is registered at
> all depends heavily on the alignment of the SG magnet. Then it becomes
> the case that sometimes for certain alignments which weren't used,
> neither up nor down would have been observed. As you say, though, it
> doesn't help you avoid nonlocality because whether a particle is
> registered at a particular detector would have to depend on the
> situation at the other detector. You can also count "no event observed"
> as a result and then you're back to where you started.
The consideration of the case with insufficient detector efficiency is
another question. Detector efficiency is a real loophole. If there
is enough room for anwering "no detection" we can violate Bell's
inequality.
>> AFAIU the phrase "has no definite results" in the description above,
>> this would be an example of non-CFD.
> It's not really so simple. The little man in the box doesn't really
> have any source of "pure" randomness to draw on which is known to be
> more unspeakable than a pseudo-random number generator. That is, inside
> the box there could just as well have been a computer outputting
> /dev/random and pressing the button might just show you the most
> recently output byte. As you know from Bohmian mechanics, quantum
> events aren't really known to be any more "inherently random"
> than a computer's random number generator, and the rest of physics,
> including the "free will" of the experimenters, is just the same.
> So I would say that even in this case, if I don't press the button,
> there's still a "what I would have gotten" if I had.
Ok - that's part of definition of non-CFD, everybody who uses this
term should clarify this.
For the consideration of the violation of BI this great conspiracy is
irrelevant. This follows from another argument - the FTL phone
argument:
If we don't accept that violation of BI falsifies Einstein causality
based on the Great Conspiracy loophole, we should reject also that a
working FTL phone falsifies Einstein causality - the power of the
Great Conspiracy loophole is the same.
If non-CFD is associated with the Great Conspiracy loophole, we should
throw away it immediately for the same reason.
> Really the only way to definitely have no CFD is to have the random
> number generator in the box seeded by the conditions inside my brain,
> so that my decisions to press or not to press the button are entwined
> with "what I would have gotten". In that case, you can't change my
> decision without also changing the random number that I would have
> got. There may be _several_ ways of changing my decision, and each of
> these could have caused different seedings of the random number
> generator. In that sense there is no simple "what I would have got
> if I had pressed the button", but there _is_ a "what I would have
> got if I had pressed the button because I was bored" or a "what
> I would have got if I had pressed the button accidentally", and these
> can be different things, since the internal workings of the box
> can depend on whether I'm bored or clumsy (maybe not because they
> can sense it, but perhaps because in order for me to be bored at
> that time, something else in my past would have had to have been
> different, and that thing can affect the internal workings of the
> box too).
In some sense, the other information contained in your brain (are you
bored or not) are part of reality too, in this sense I would argue
that nonetheless we have CFD.
>> But using such random methods you cannot violate Bell's
>> inequality. And Bell's theorem (which assumes that some probability
>> distribution over the possible states of reality is given) covers
>> this simple type of random hidden variables.
> Right, but a loophole in the theorem is hyperdeterminism, and that is
> precisely the case in which CFD fails as I described above.
But hyperdetermism (the Great Conspiracy above) may be ignored.
> I don't believe it, of course. Spooky action at a distance is more
> acceptable.
But what I like for my FTL phone argument is that it shows that
hyperdetermism is not simply less acceptable, but really stupid.
Imagine a relativist who defends Einstein causality using the
hyperdeterminism loophole in an FTL translation on TV ;-).
Ralph Hartley
Ilja Schmelzer wrote:
> Ralph Hartley <har...@aic.nrl.navy.mil> writes:
>>It doesn't matter to this argument, but I assume you also believe
>>that there is a *true* interpretation. It may be impossible to find
>>out which that is, since they are experimentally indistinguishable.
>
> Exactly.
>
>>Whether this position is true, or even meaningful, or not is a
>>philosophical issue, which I will not discuss further.
>
> SCNR: At least the position that it is true seems consistent.
> Assuming it is false, we have some false philosophical interpretation
> about our world. Given that philosophical interpretations about our
> world may be false, therefore also true, assuming that interpretations
> of QM cannot be false or true sounds strange. The philosophical
> thesis that such philosophical claims are meaningless does not seem to
> be meaningful too.
I tried to resist breaking my own promise, but this pargraph broke my
record for hardest to parse.
Philosophical claims may be true, false, ambiguous, relative, meaningless,
or unknowable (this list is not exhastive, and the possiblities may not be
mutually exclusive).
Many are meaningless, some must be false, some may even be true.
I didn't say that interpretations of QM *cannot* be false or true, only
that they *neednot* be.
There may be an actual true number of angels that can dance on the head of
a pin, but that doesn't make arguing about it any less a waste of time.
I think I'm beggining to see John Baez' point.
Ralph Hartley
Ralph Hartley
> Ralph Hartley wrote:
>>The position is that the concept of "what would have happened if I did
>>measurement other than the one I did", is either invalid, relative (e.g.
>>frame dependent) or subjective.
> Now that's a terrible position to find yourself in.
I've been in much worse. Voluntarily.
At 2AM on Sunday I was deep in a previously unexplored cave, with 8 hours
of hard travel, including several kilometers of crawling, squeezing, and
climbing house-sized rocks, and with two vertical ropes (one over 40m) to
climb between me and the light of day, and I was already tired.
A little irreducible philosophical ambiguity isn't that bad.
> Space appears in the expression for the Hamiltonian. That is, we
> write H=p^2/2m + V(x), but the operators p and x don't belong in
> the Everett interpretation because they are defined in terms
> of measurements and collapses and so on. The only operator in
> Everett is H.
...
> Everett doesn't have an axiom that there is a spatial structure
> anywhere. All there is is H and |psi>. No space, no particles,
> no results of measurements.
I think you are being much to restrictive here. The Everett interpretation
works on any quantum system, so it doesn't *need* those things. That
doesn't mean it *denys* them. A good model of our world would be expected
to at least *admit* a description in those terms.
>>Consider a world in which the wavefunction collapse was real but always
>>happens one second *after* the measurement ...
>>Nor could any experiment tell if the collapse was delayed a billion years.
> We can speculate that only the present exists and the past never did,
> but we can't speculate that only the future, a billion years from
> now, exists and that we don't.
Bet?
> If there were never any well-defined
> structure corresponding to my body in physical space, then I would
> not be having this correspondence
So you aren't "well-defined" now, but you will have been. Do you think you
could tell the difference any more than you can feel the motion of the earth?
>>So at least as a limit, we can see that it doesn't really matter (in any
>>observable way) if the wavefunction collapse *never* happens.
> It does matter - if the only mathematical things which have any reality
> at all are the Hamiltonian and the state vector, then one must accept
> that there is no special spatial structure in the universe - that,
> for example, the p-representation is just as good as the x-representation
So? Do you have a problem with that?
For me, it is enough that the world *admits* a causal (by my
definition) spatial structure. Why should I care if it isn't unique?
In fact if there is any symmetry at all, I would expect that it
wouldn't be. I could imagine that there might be some sort of minimal
quotient, but it wouldn't bother me too much if there were were more
than one, inequivalent, spatial structure.
Certainly for non-interacting particles p is just as good as x, the
duality is a symmetry of the system (I agree that your worries about
normalization are a red herring, just consider a compact space). But
if you add a small local (in x) interaction, then the description in
terms of x will be local (by construction), but in terms of p it won't
be. (The p basis is special in being completely *global*, because the
FT of a delta function is a constant. That's why we use it sometimes.)
Ralph Hartley
Ralph Hartley
> Ralph Hartley wrote (even number of >s):
>>Just so. There is a short list of principles that many people find self
>>evident (and somewhat longer list that are so internalized they are hard to
>>even state), and the experiment says you have to give at least one of them up.
>
> Can you (to whatever extent you think it's reasonable) give a list of
> some of the other assumptions you think are necessary for somebody
> to accept the proof?
That when you do a measurement, there is an absolute, unique,
non-subjective outcome.
That the experimental method is valid at all. (I'm not saying I would give
that one up, but if you thought God created every event on a case by case
basis, would the proof even make sense?)
> Right. This is part of what I'm saying. The rule for determining which
> thing happened "first" that I would give (in cases where the usual
> ideas break down) would be to say that if some physical process is
> obeying deterministic physics (like gears and cogs) and can be predicted
> perfectly for the next ten years, then the behavior of that process
> seven years from now is "before" the behavior of a nondeterministic
> system five years from now.
Since gears and cogs operate locally, this just means that for lightlike
separations the causal order is the time order (which is the same for all
frames).
> That is, the settings in EPR experiments,
> which can be controlled by gears and cogs, are "before" the results,
> regardless of the frames in which the experiments are done.
Before the local result. To say that the setting always is before the
distant result is not consistent. For example, if the distant measurement
is absolutely (timelike separation) before the setting. QM doesn't seem to
care, you get the same statistics either way.
> I'm trying to make a consistent generalization of the notion that
> if X happens before Y then X can be said to affect Y but Y can't
> be said to affect X. When X happens at an earlier time, it's
> trivial. When X and Y are separated by a spacelike distance, though,
> you need to do something else.
OK, but whatever rule you use must agree with the time order for timelike
separations. And it must always give a consistent order, no loops.
*Any* rule that meets those requirements is equivalent, as far as
experimental results are concerned. In that sense, the rule is non-physical.
>>I think it's possible that your paradox comes from trying to apply causal
>>reasoning in inconsistent frames.
>
> Well insofar as a frame is considered as a three-dimensional plane
> separating things set in stone from things yet to be decided, then
> that is certainly true. We shouldn't change our minds about what
> has been set in stone forever halfway through an argument.
Right, and to get your paradox you did exactly that.
Tell me what order of events you used. I think it was Bobs E1, then Alice's
E1, then Alice's E2, then Bob's E2, then Bob's E1. The results and settings
could not have become "set in stone" in that order, because that *isn't* an
order.
> Well I don't think causality should have anything to do with reference
> frames at all.
Sorry! You can't have that! If you want causality to always be well defined
(and some other stuff that you apparently want), you *have* to choose a
frame. You can't change that frame without changing the direction of
causality, no matter who accelerates.
> For me, "X caused Y" means something like "From the
> knowledge that X is true and confidence I have in whatever physical
> laws are applicable, it is possible to deduce Y". That is, causality
> goes from things we already know to things we have to deduce, which
> is why it so often coincides with going from past to future.
That is *my* kind of causality. By that definition Bob's settings don't
affect Alice's results at all. Knowing Bob's setting was w and the physical
laws, it is not possible to deduce *anything* about Alice's result.
It is only if you also know "Bob would have gotten up if his setting was s"
that you can deduce anything about Alice's result. That statement, if it is
meaningful at all, is frame defendant. For it to have an absolute truth
value, there must be an absolute order of events.
>>>>>then unless there is some extra rule to prevent this from happening
>>
>>The word that makes this sentence wrong is *extra*. That A=>not(A) and
>>not(A)=>A can't both happen is implied by your *definition* of "=>". QM
>>doesn't have to do anything special to avoid an event that is impossible by
>>definition.
>
> I think logical reasoning about nature is being confounded with
> processes by which nature operates here. Think of the twin paradox in
> special relativity. It indicated there was something wrong with the
> logic involved - ie that the person looking at the paradox wasn't
> using the theory properly. *But* what happens is that nature has
> an "extra rule" which wasn't used in the statement of the paradox -
> that is that strange things happen when the twin traveling
> away from earth starts accelerating back towards earth (as far as
> that twin is concerned). The twin paradox only appears to be a
> paradox if you don't know this rule. As you might say, special
> relativity doesn't have to do anything to avoid an event which is
> impossible by definition (any paradox), but there is an effect or
> rule of nature which intervenes to prevent the paradox and save
> the day.
But here we know *all* the rules. For any experiment you describe, I can
tell you the outcome (or you could do it yourself).
You want more rules to decide the direction of causation. They would be
extra in the sense of not being needed to predict any results. If you do
formulate them, you must do so in a consistent way.
Please excuse me for calling rules that don't change any experimental
results "non-physical".
>>>>Yes, there are such rules. They are the rules of logic.
>>>
>>>The rules of logic are those rules by which we _deduce_ that both
>>>effects are not active simultaneously.
>>
>>But if your definition of "->" implies that A->not(A) and not(A)->A can't
>>both hold (I think it does), then they can't both hold *regardless* of what
>>the laws of nature are.
>
> Yes, but this is deduction that we can do. I know they don't hold
> because I've deduced it with these means. But the means by which
> I deduce this are not a description of how nature operates.
We *have* a description of how nature operates (if you count
interpretations, we have more than one, but they are equivalent, as far as
results are concerned).
If our description is consistent (as it appears to be) and it is correctly
applied, it never says that A->not(A)->A. If it did, we might be very worried.
The fact that you appear to show that leads one to suspect that you have
not applied it correctly (I think by changing direction of causation in mid
argument, but possibly by conditioning on two or more mutually exclusive
events).
Ralph Hartley
Fizz Fann
> Fizz Fann wrote (odd number of >s):
> > Ralph Hartley wrote (even number of >s):
>
> >>Just so. There is a short list of principles that many people find self
> >>evident (and somewhat longer list that are so internalized they are hard to
> >>even state), and the experiment says you have to give at least one of them up.
> >
> > Can you (to whatever extent you think it's reasonable) give a list of
> > some of the other assumptions you think are necessary for somebody
> > to accept the proof?
>
> That when you do a measurement, there is an absolute, unique,
> non-subjective outcome.
Ok. But it's an observation, rather than an assumption, that
when you do a measurement, there is an absolute, unique,
subjective outcome. In terms of those subjective outcomes,
that is, for the world that we actually see and live in, the world
will appear to be nonlocal.
> That the experimental method is valid at all. (I'm not saying I would give
> that one up, but if you thought God created every event on a case by case
> basis, would the proof even make sense?)
Yeah. I've been working with the hypothesis that that isn't the
case, but I could be wrong. I'll be very annoyed at God if it
turns out that he's been playing games with me. As Ilja pointed
out, resorting to arguments like these in order to explain
the observed nonlocality in nature is stupidly extreme (I think
he said "extremely stupid"), because those arguments would also
be consistent explanations for a working FTL telephone.
Don't you consider those assumptions (well-defined outcomes
of experiments and validity of the scientific method) rather
more difficult to abandon than locality? Bell described
those approaches as "romantic" and "unscientific", and I think
I'd probably agree with him that the part inside me which
finds them appealing is the part which longs for mysticism
and doesn't like the harsh mechanical reality of the world.
I very strongly suspect that others who favor these approaches
are the same.
> > Right. This is part of what I'm saying. The rule for determining which
> > thing happened "first" that I would give (in cases where the usual
> > ideas break down) would be to say that if some physical process is
> > obeying deterministic physics (like gears and cogs) and can be predicted
> > perfectly for the next ten years, then the behavior of that process
> > seven years from now is "before" the behavior of a nondeterministic
> > system five years from now.
>
> Since gears and cogs operate locally, this just means that for lightlike
> separations the causal order is the time order (which is the same for all
> frames).
Right.
> > That is, the settings in EPR experiments,
> > which can be controlled by gears and cogs, are "before" the results,
> > regardless of the frames in which the experiments are done.
>
> Before the local result. To say that the setting always is before the
> distant result is not consistent. For example, if the distant measurement
> is absolutely (timelike separation) before the setting. QM doesn't seem to
> care, you get the same statistics either way.
No no. I mean before both results. Because the machinery which determines
the settings can have been set up long ago in the intersection of the
past light cones of the events at which the results happen.
> > I'm trying to make a consistent generalization of the notion that
> > if X happens before Y then X can be said to affect Y but Y can't
> > be said to affect X. When X happens at an earlier time, it's
> > trivial. When X and Y are separated by a spacelike distance, though,
> > you need to do something else.
>
> OK, but whatever rule you use must agree with the time order for timelike
> separations. And it must always give a consistent order, no loops.
Why a consistent order? There are rules of deduction with loops -
consider Goedel's theorem. "This sentence is unprovable" - if
it were provable it would be both false and true (a paradox!),
therefore it's unprovable and therefore true. My system with
the causality loop is the same - if both nonlocal effects are
active at once, we have a paradox, therefore they're not both
active. You can't deduce that without looking at the loop.
> *Any* rule that meets those requirements is equivalent, as far as
> experimental results are concerned. In that sense, the rule is non-physical.
Nope. I've given a testable experimental prediction - that in
certain experiments, the extent of violation of Bell's inequalities
will be observed to be smaller than usual.
> >>I think it's possible that your paradox comes from trying to apply causal
> >>reasoning in inconsistent frames.
> >
> > Well insofar as a frame is considered as a three-dimensional plane
> > separating things set in stone from things yet to be decided, then
> > that is certainly true. We shouldn't change our minds about what
> > has been set in stone forever halfway through an argument.
>
> Right, and to get your paradox you did exactly that.
I don't consider a frame to separate things set in stone from
things yet to be decided.
> Tell me what order of events you used. I think it was Bobs E1, then Alice's
> E1, then Alice's E2, then Bob's E2, then Bob's E1. The results and settings
> could not have become "set in stone" in that order, because that *isn't* an
> order.
If causality is deduction, then loops can be used there as they
are in logic.
> > For me, "X caused Y" means something like "From the
> > knowledge that X is true and confidence I have in whatever physical
> > laws are applicable, it is possible to deduce Y". That is, causality
> > goes from things we already know to things we have to deduce, which
> > is why it so often coincides with going from past to future.
>
> That is *my* kind of causality. By that definition Bob's settings don't
> affect Alice's results at all. Knowing Bob's setting was w and the physical
> laws, it is not possible to deduce *anything* about Alice's result.
You're confusing "cause" and "affect" (no pun). Bob's settings *affect*
Alice's results - they do not cause them because Alice's results
can depend on other things apart from Bob's settings. Imagine a
locked room with the light switch on the outside. You don't know
whether the light is presently on or off, but you know that flipping
the switch will change its state. Your statement above would read
"By that definition flipping the switch doesn't affect the state
of the light at all."
> >>QM doesn't have to do anything special to avoid an event
> >>that is impossible by definition.
> >
> > I think logical reasoning about nature is being confounded with
> > processes by which nature operates here. Think of the twin paradox in
> > special relativity. It indicated there was something wrong with the
> > logic involved - ie that the person looking at the paradox wasn't
> > using the theory properly. *But* what happens is that nature has
> > an "extra rule" which wasn't used in the statement of the paradox -
> > that is that strange things happen when the twin traveling
> > away from earth starts accelerating back towards earth (as far as
> > that twin is concerned). The twin paradox only appears to be a
> > paradox if you don't know this rule. As you might say, special
> > relativity doesn't have to do anything to avoid an event which is
> > impossible by definition (any paradox), but there is an effect or
> > rule of nature which intervenes to prevent the paradox and save
> > the day.
>
> But here we know *all* the rules. For any experiment you describe, I can
> tell you the outcome (or you could do it yourself).
The point is not that we do know everything about special relativity.
We have a theory (quantum mechanics) in which you can't prefectly
predict the outcome of experiments; we only know about corelations.
To make the analogy to special relativity appropriate, forget for
a moment most of the details of the theory and recall only that for
two observers in constant relative motion, each observer will see the
clock of the other running slowly. Imagine that that is the only
correlation which has been noticed and that the rest of the theory
is unknown. Somebody who only knows this correlation could point
at the twin paradox and correctly observe that this can't be the
whole story.
> Please excuse me for calling rules that don't change any experimental
> results "non-physical".
Done.
> >>But if your definition of "->" implies that A->not(A) and not(A)->A can't
> >>both hold (I think it does), then they can't both hold *regardless* of what
> >>the laws of nature are.
> >
> > Yes, but this is deduction that we can do. I know they don't hold
> > because I've deduced it with these means. But the means by which
> > I deduce this are not a description of how nature operates.
>
> We *have* a description of how nature operates (if you count
> interpretations, we have more than one, but they are equivalent, as far as
> results are concerned).
We have no description of how nature selects a single result to
a quantum measurement which generalises to relativistic situations,
except possibly the preferred frame version of Bohmian mechanics.
Regards,
Alex.
Edward Green
> Bell's inequalities were considered significant because if they were
> violated, as quantum mechanics predicted, then a particular view of
> the world, that of a world of local causes, was no longer tenable.
Actually, I think it's somewhat less cosmic than you portend.
The raw fact of the matter is, a violation of a Bell's inequality
means, no more and no less, that a given set of conditional or
marginal probabilities, like P(A|B), P(B|C), P(C|A), are not
consistent with any possible joint distribution on events A,B,C -- in
this example.
This is a raw, mathematical, unphilosophical result, which frees us
from any twisted language about "what would have happened if ... ",
which we may leave undefined.
The question is, why should we care about such a condition? The
answer, after sufficiently deep introspection, is that this concrete
mathematical test turns out to capture pretty well what we want to
mean by "local causes" ... but like all slogans, there is no shortcut
for a particular understanding of the abbreviation ... in this case,
for the possibility that certain kinds of experiments are the result
of any conceivable local information blobs travelling to two separate
locales, and, without any further affect or cooperation between the
locales, determining the results of experiments at those locales, in
conjunction with some other locale variables.
Now, does this eliminate "locality"? Only in this very particular
sense. It leaves open the possibility that the distant locales are
"local" in some other sense in which we have yet to discover, and it
also leaves open the possibility of superluminal cooperation. Of
course, given the slogan like understanding of SR this is viewed as a
desperate move, but in fact superluminal phase velocities are fairly
comfortably known in various descriptions.
There are also "statistical loop holes", so called, which are usually
also regarded as the desperate measures of weak minds, but this
opinion stems from a misconstrual of experimental inference.
Fizz Fann
mind, and the like are not about physics and so are off-topic for
this group. Let's stick to the physics. -TB]
Ilja Schmelzer <schm...@wias-berlin.de> wrote in message news:<i3gr88e...@wias-berlin.de>...
> Ralph Hartley <har...@aic.nrl.navy.mil> writes:
>
> > However, the fact that QM violates Bell's inequality *does* mean
> > that you have to give *something* up.
>
> Indeed.
>
> > I'm pretty sure that the thing you have to lose to satisfy your
> > "demands" is Lorenz invariance.
>
> And, indeed, I'm giving up Lorentz invariance.
That seems rather drastic for me. What do you think of
the Lorentz-invariant Bohmian theory of quant-ph/0105040?
It seems that to some extent the authors did it just to
shut up the people who claim that Bohmian mechanics can't
possibly be made Lorentz-invariant. They seem to be good
at doing the impossible - in quant-ph/0208072 they disprove
the legend that Bohmian mechanics can't account for pair
creation. It seems Bohmian mechanics deserves the record
for having the largest number of incorrect claims of
impossibility surrounding it.
> Instead, I insist that preserving "the way I think about logic and
> truth", whatever the experiment shows, is not a religious dogma, but
> simply the scientific method. There is, indeed, no imaginable outcome
> of an experiment which could force me to give up "the way I think
> about logic and truth". In the worst case I would conclude only that
> I have no theory which allows to explain the given experiment - which
> is a quite common state of science.
I venture here to propose a theory of quantum physicists' behavior.
That once they have given up the way they think about logic and
truth, as they are apparently told to do when learning quantum
mechanics, they don't want to go back afterwards. They were told
that it's naive to keep thinking that the world they see around
them is really there - that quantum mechanics *forces* us to accept
that there's really only a function which has a non-zero value on
the world they see. They did as they were told because they certainly
didn't want to be naive. When it turns out that their quantum mechanics
teachers had lied to them, and that they're not forced to abandon
logic and truth in favor of quantum logic and amplitude, well it's
too late, because they've already abandoned them. And they still
consider it naive if you don't abandon them. They think "Well maybe
you don't have to abandon them just to explain nonrelativistic
things like particle diffraction after all, but it *must* be
necessary to abandon them *somewhere*. After all, it's naive if
you don't. These Bohmian people are naively trying to do the
impossible."
I make a physical prediction based on this theory - that further
claims of impossibility will be made about Bohmian mechanics, and
that they too will turn out to be incorrect. The reason is that
the majority of physicists seem to really believe that you actually
have to stop thinking logically about a real world with a definite
state (in the "naive realism" sense). You don't, but they think
you do, so they will claim that those who pig-headedly refuse to
adopt the vague quantum philosophy are going to run into trouble,
because they're basically trying to do the impossible. That's why
the no-hidden-variables theorems stood unchallenged for so long,
and it's why they're there in the first place, and it's why it's
so widely believed now that Bohmian mechanics can't be made
Lorentz-invariant and can't cope with pair creation.
> > All such interpretations have a "preferred" frame (*which* frame is
> > preferred is, unfortunately(?), unobservable). Bohmian mechanics,
> > for example, has this property.
>
> Yep.
Apparently not. With all due respect, might I accuse you of too
eagerly accepting the validity of an argument just because you
like the conclusion? The conclusion might, for all I know, be
correct, but if we accept or reject arguments just because we
like the conclusions we commit the same sin which allows people
to think of Bohmian mechanics as a naive attempt to return to
pre-quantum notions of "reality" and "logic". I believe it's
precisely that sin which allowed the flaws in the no-hidden-variables
theorems to go unnoticed. People liked the conclusions; they
reinforced their own notions of what was naive and what wasn't,
so the arguments must have been correct.
> > If you think that one of the two statements "your measurement
> > affected his result" and "his measurement affected the correctness
> > of your guess" is true, and the other false, then you *must* think
> > that there is a "true" reference frame, since one statement is true
> > in some frames, and the other in others.
>
> Yep.
Nope.
> > It doesn't matter at all in what reference frames the various
> > experimenters and pieces of equipment are at rest. What causes what
> > is different in different reference frames, so if you think that
> > causation goes in a particular direction, you must think that there
> > is particular *fixed* frame to use when determining that, and that
> > all other frames are "wrong".
>
> Now, causation (at least in general) has some particular direction:
> The direction from a "free will" (or however randomly chosen) decision
> (what to measure) to the outcome of the experiments.
This is exactly my point. The free will of the experimenters means
that little causation arrows point away from them and towards the
measurement results, no matter who prefers what frame.
> > Someone accelerating from the "right" frame to a "wrong" one can't
> > change the direction of causation.
>
> That would be, indeed, strange.
Very much so. I'm glad I don't do anything like that.
> > And *that's* why I think such interpretations are "ugly".
>
> Is "ugly" a good reason to give up "the way I think about logic and
> truth"? Not for me. "Ugly" is, even in the worst case, only a
> challenge to find something more beautiful.
So if it turns out that you don't, in fact, have to give up
Lorentz invariance, how willing would you be to take it back,
having once abandoned it, and invested a lot of time and
energy working on the hypothesis that it's false?
Best wishes,
Alex.
Fizz Fann
> > Space appears in the expression for the Hamiltonian. That is, we
> > write H=p^2/2m + V(x), but the operators p and x don't belong in
> > the Everett interpretation because they are defined in terms
> > of measurements and collapses and so on. The only operator in
> > Everett is H.
> ...
> > Everett doesn't have an axiom that there is a spatial structure
> > anywhere. All there is is H and |psi>. No space, no particles,
> > no results of measurements.
>
> I think you are being much to restrictive here. The Everett
interpretation
> works on any quantum system, so it doesn't *need* those things. That
> doesn't mean it *denys* them. A good model of our world would be
expected
> to at least *admit* a description in those terms.
Well the Everett interpretation really only works on *one* system,
viz the universal wavefunction. The question is the following:
which basic mathematical ingredients does Everett need in order
to be capable of describing the world as we see it? My assertion
is that space, particles and so on are things that we put in
ourselves, by hand. They don't come out of the Hamiltonian,
Hilbert space and state vector.
> >>Consider a world in which the wavefunction collapse was real but
always
> >>happens one second *after* the measurement ...
> >>Nor could any experiment tell if the collapse was delayed a billion
years.
>
> > We can speculate that only the present exists and the past never
did,
> > but we can't speculate that only the future, a billion years from
> > now, exists and that we don't.
>
> Bet?
I might bet that *we* can't speculate, but I won't bet that *you*
can't.
It seems you're a gambling man.
> > If there were never any well-defined
> > structure corresponding to my body in physical space, then I would
> > not be having this correspondence
>
> So you aren't "well-defined" now, but you will have been. Do you
think you
> could tell the difference any more than you can feel the motion of
the earth?
:) Can a non-existing person tell the difference between himself and
an existing one? The point in question is that on some putatively
well-defined world which pops into existence, some bones are dug up,
and you claim that that gives phenomenal existence to the thoughts
and experiences of hypothetical people who might have owned those
bones. I should point out that if you reverse-evolve the wavefunction
of that later world for a billion reversed years, you won't get
anything that looks like a single well-defined world with well-defined
people having well-defined experiences.
Interestingly, if you reverse-evolve the uncollapsed wavefunction,
but let it drag with it the bubble of probability surrounding the
point in configuration space representing the well-defined
world with its bones, you'll find the bubble of probability getting
smaller and smaller the further backwards you go in time (no
Liouville's theorem here!), and you'll find that underneath the
wavefunction of the early world, the little point of probability
that you got by reverse-evolving, is following a Bohmian trajectory,
so you'll have an essentially Bohmian picture of the early world with
each particle moving along its trajectory in space.
But anyway, this isn't really important because it's about whether
one thing can attain physical reality by virtue of its mathematical
relationship to another thing which has it. Let's just say that
if you didn't specify that the future world was well-defined then
there would be no way that you could single out an earlier world
and claim that it was too.
> >>So at least as a limit, we can see that it doesn't really matter
(in any
> >>observable way) if the wavefunction collapse *never* happens.
>
> > It does matter - if the only mathematical things which have any
reality
> > at all are the Hamiltonian and the state vector, then one must
accept
> > that there is no special spatial structure in the universe - that,
> > for example, the p-representation is just as good as the
x-representation
>
> So? Do you have a problem with that?
>
> For me, it is enough that the world *admits* a causal (by my
> definition) spatial structure. Why should I care if it isn't unique?
> In fact if there is any symmetry at all, I would expect that it
> wouldn't be. I could imagine that there might be some sort of minimal
> quotient, but it wouldn't bother me too much if there were were more
> than one, inequivalent, spatial structure.
The real question is: should you be concerned about being eaten
alive by mutant tortoises? If I held such views as yours it would
be a real concern for me.
At the moment, quantum mechanics describes a wavefunction
psi(x1,x2,...) which depends on the positions of all the
particles in the universe. We draw some comfort from the
fact that psi is relatively large when the values of the
x1, x2 etc are set in such a way that our atoms are in
various locations in our bodies as we see them now, and that
if we were to construct values of x1, x2 etc which described
a configuration of the universe in which we were being
eaten alive by mutant tortoises, but had all the same memories
as we do now (and are therefore somewhat surprised to
find ourselves being eaten alive), then psi would be
comparatively small in that little corner of the configuration
space. That is, psi(the way the world seems to us) has
a larger magnitude than psi(a world in which we are being
eaten alive).
So although the answer to the question "Are there
Everett worlds in which we find ourselves unexpectedly
being eaten alive by mutant tortoises?" is "Yes,"
we can at least add "but not very many."
We can say that if we accept that there is only one
spatial structure with a fixed number of particles
and so on. But if we think that the only things
out there which are ultimately real are the Hamiltonian
and Hilbert space and state vector, and say that
any spatial structure compatible with those is just
as real as any other, then the world becomes a
more frightening place.
In the same way that a system of one particle moving
in two dimensions has the same Hilbert space and
Hamiltonian as a system of 10^23 particles interacting
with complicated potentials in three dimensions, the
Hilbert space and Hamiltonian of our universe will
admit practially any systam and any spatial structure
at all.
That is, although psi(x1, x2, ...) will be small
when I choose x1, x2, and so on to describe us
being eaten alive, it will be possible to find
another set of coordinates (possibly a different
number of them, describing a universe with a
completely different number of particles interacting
with different potentials) such that psi(q1,q2,..)
will not be so small when I choose to set the values
of q1 etc to describe a world in which we are being
eaten alive.
You might think that you are unlikely to find yourself
living in such a spatial structure, but the copy of
you which lives there has the same memories as you
up until this point, and nothing saved him, and if
you really believe that his world is just as real
as yours, then you can no longer say that it occurs
with less quantum amplitude than yours does, so you
can't draw comfort from the same source as you did
when there was only one preferred spatial structure.
What I'm trying to get at with this rather convoluted
example is that saying that there's only the Hilbert
space and Hamiltonian and state vector and that
anything compatible with those things is just as real
as we are leads to unimaginably many worlds, and
removes the restriction that the sum of probabilities
of those worlds must be normalised according to
the rule that that integral of psi^2 should be one.
That particular restriction holds only when you
restrict yourelf to one preferred set of observables.
The situation with many worlds is slightly analogous
to the following: Take the number 1, and decompose it
into any sum of other numbers. Then interpret those
numbers whatever way you want (as positions of particles,
as spins and so on). Then that world which you imagine
ir really "out there" and our world is one of those.
Regards,
Alex.
Ilja Schmelzer
thephy...@yahoo.com (Fizz Fann) writes:
> Ilja Schmelzer <schm...@wias-berlin.de> wrote in message news:<i3gr88e...@wias-berlin.de> ...
>> Ralph Hartley <har...@aic.nrl.navy.mil> writes:
>>> However, the fact that QM violates Bell's inequality *does* mean
>>> that you have to give *something* up.
>>
>> Indeed.
>>
>>> I'm pretty sure that the thing you have to lose to satisfy your
>>> "demands" is Lorenz invariance.
>>
>> And, indeed, I'm giving up Lorentz invariance.
>
> That seems rather drastic for me.
Not for me. Last not least, some people (Lorentz, Poincare) have
preferred a preferred frame even without knowing about all these
quantum effects at a distance. What is drastic with going back to
common sense?
> What do you think of the Lorentz-invariant Bohmian theory of
> quant-ph/0105040?
"Opposite microcausal and macrocausal arrows of time" means giving up
standard common sense notions of causality without good evidence.
Standard causality (with one arrow of time) is one of the assumptions
which I use to prove the necessity of a preferred frame.
> It seems that to some extent the authors did it just to
> shut up the people who claim that Bohmian mechanics can't
> possibly be made Lorentz-invariant. They seem to be good
> at doing the impossible - in quant-ph/0208072 they disprove
> the legend that Bohmian mechanics can't account for pair
> creation.
Nonetheless, in the QFT context I prefer the Bohmian field theory
proposal.
> It seems Bohmian mechanics deserves the record
> for having the largest number of incorrect claims of
> impossibility surrounding it.
;-).
> I venture here to propose a theory of quantum physicists' behavior.
> That once they have given up the way they think about logic and
> truth, as they are apparently told to do when learning quantum
> mechanics, they don't want to go back afterwards.
A similar theory may be proposed for relativists, they also don't want
to go back to a preferred frame.
>>> All such interpretations have a "preferred" frame (*which* frame is
>>> preferred is, unfortunately(?), unobservable). Bohmian mechanics,
>>> for example, has this property.
>>
>> Yep.
>
> Apparently not. With all due respect, might I accuse you of too
> eagerly accepting the validity of an argument just because you
> like the conclusion?
You might accuse me, but I will defend myself. With classical realism
and classical causality (one-directional, without closed causal loops)
you can prove the existence of a preferred foliation.
> I believe it's precisely that sin which allowed the flaws in the
> no-hidden-variables theorems to go unnoticed. People liked the
> conclusions;
Knowing the examples of failure of no-go-theorems, I agree that it is
useful to check theories which propose to do the impossible. But, of
course, in this case the first what has to be done is to understand
which of the assumptions of the no-go theorem does not hold.
And then to decide if it is useful to give it up. In case of
causality I see no good reason to give it up. (To give up yet another
part of common sense to look less naive.)
>>> If you think that one of the two statements "your measurement
>>> affected his result" and "his measurement affected the correctness
>>> of your guess" is true, and the other false, then you *must* think
>>> that there is a "true" reference frame, since one statement is true
>>> in some frames, and the other in others.
>>
>> Yep.
>
> Nope.
Why?
>>> It doesn't matter at all in what reference frames the various
>>> experimenters and pieces of equipment are at rest. What causes what
>>> is different in different reference frames, so if you think that
>>> causation goes in a particular direction, you must think that there
>>> is particular *fixed* frame to use when determining that, and that
>>> all other frames are "wrong".
>>
>> Now, causation (at least in general) has some particular direction:
>> The direction from a "free will" (or however randomly chosen) decision
>> (what to measure) to the outcome of the experiments.
>
> This is exactly my point. The free will of the experimenters means
> that little causation arrows point away from them and towards the
> measurement results, no matter who prefers what frame.
And these objectively existing little arrows define the preferred frame.
>>> And *that's* why I think such interpretations are "ugly".
>> Is "ugly" a good reason to give up "the way I think about logic and
>> truth"? Not for me. "Ugly" is, even in the worst case, only a
>> challenge to find something more beautiful.
> So if it turns out that you don't, in fact, have to give up
> Lorentz invariance, how willing would you be to take it back,
> having once abandoned it, and invested a lot of time and
> energy working on the hypothesis that it's false?
Oh, I have already abandoned a lot of my ideas, even after having
invested a lot of time and energy. Starting with attempts to solve
the problem of four colors in my childhood. I'm quite happy that most
wrong stuff I have given up before publishing it on arxiv.org.
Regarding symmetry groups, including the Lorentz group - I like them.
If something can be made more symmetric, I'm happy to do it. For
example, for the equations which define the hidden preferred variables
I use a nice "relativistic" equation - the harmonic condition, which
is simply a relativistic wave equation.
Fizz Fann
> thephy...@yahoo.com (Fizz Fann) writes:
> > Ilja Schmelzer <schm...@wias-berlin.de> wrote in message
> > news:<i3gr88e...@wias-berlin.de> ...
> > >I'm giving up Lorentz invariance.
> > That seems rather drastic for me.
> Not for me. Last not least, some people (Lorentz, Poincare) have
> preferred a preferred frame even without knowing about all these
> quantum effects at a distance. What is drastic with going back to
> common sense?
I don't think it's quite as simple as returning to a once-clear
common sense view of the world. Classical physics was invariant
under Galilean transformations, and most if not all physicists
were aware of this. In that sense, there was no common consensus
on a preferred frame. The only really commonly understood preferred
frames were the Aristotelian one (ie the preferred frame is the
one where the ground is not in motion, and in which moving objects
will come to rest) and the luminiferous ether one, which was
bizarre precisely because it broke Galilean invariance and suggested
that observers moving at different speeds would see different physics.
Your frame is neither of these, so accepting it wouldn't simply be a
return to a previously well-accepted and understood idea.
> > What do you think of the Lorentz-invariant Bohmian theory of
> > quant-ph/0105040?
> "Opposite microcausal and macrocausal arrows of time" means giving up
> standard common sense notions of causality without good evidence.
> Standard causality (with one arrow of time) is one of the assumptions
> which I use to prove the necessity of a preferred frame.
Ok. I don't want to defend it, but I'll say that, since macroscopic
causality emerges from the microcosm in their model, you must rely
on your audience's willingness to reject the notion that causality
can possibly be emergent in order to convince them of your frame.
That is, it's not a case of "giving up standard common sense notions
of causality", but of allowing the possibility that common-sense
causality is a derivable rather than an axiomatic feature of the world.
I personally don't expect that a hypothetical audience would so
vociferously insist on causality-as-axiom rather than
causality-as-theorem that they would dismiss a candidate theory as
unacceptable based on it. (Of course, any audience for such theories
will probably remain hypothetical for the foreseeable future!)
> > It seems that to some extent the authors did it just to
> > shut up the people who claim that Bohmian mechanics can't
> > possibly be made Lorentz-invariant. They seem to be good
> > at doing the impossible - in quant-ph/0208072 they disprove
> > the legend that Bohmian mechanics can't account for pair
> > creation.
> Nonetheless, in the QFT context I prefer the Bohmian field theory
> proposal.
I'm still undecided on that. In QFT, it seems that things which
really are waves, like phonons and things which really are
particles, like electrons, aren't treated so very differently
by the mathematics. Simply making a Bohmian version of both
would still not sufficiently account for their extremely different
observable physical characteristics. Also, I'm not sure I'm
happy with Bell's "only fermions are Bohmian" field theory either,
since two fermions are a boson.
Another thing - there seems to be a widespread myth in physics lore
that you need field theory to combine relativity with quantum
mechanics. That's not entirely accurate, because it's the measurement
axioms which cause the incompatibility - Dirac theory and Klein-Gordon
theory, and, to some extent, electromagnetism, can be considered as
relativistic quantum theories even without being relativistic quantum
field theories. Of course, they have no measurement axioms.
> > I venture here to propose a theory of quantum physicists' behavior.
> > That once they have given up the way they think about logic and
> > truth, as they are apparently told to do when learning quantum
> > mechanics, they don't want to go back afterwards.
> A similar theory may be proposed for relativists, they also don't want
> to go back to a preferred frame.
Possibly true, but, like above, it's really not as simple as going
back to an old and well-accepted understanding of the world which
stood us in good stead. Preferred frames have been occasional
oddities which gave incorrect physical predictions each time.
> > >Some poor uncited soul wrote:
> > > > If you think that one of the two statements "your measurement
> > > > affected his result" and "his measurement affected the
> > > > correctness of your guess" is true, and the other false, then
> > > > you *must* think that there is a "true" reference frame, since
> > > > one statement is true in some frames, and the other in others.
> > > Yep.
> > Nope.
> Why?
Because I think that the statement "his measurement affected the
correctness of your guess" is vacuous rubbish. The entities which
are demonstrated by the violation of Bell's inequalities to have
a nonlocal influence operating between them are the choices of
measurement axes and the results of measurements. There are no
guesses and no correctnesses, and even if there had been, the
inequalities don't refer to them and they are not recorded anywhere
and aren't relevant to the observed nonlocal effect.
[Moderator's note: terms like "vacuous rubbish" have a tendency
to raise the temperature of the debate, and are best avoided. - jb]
The assertion that the above vacuous statement is true in some frames
is nonsense. The situation at hand is that choices of measurement
axes are affecting results, and _not_ the other way around. The
truth of _that_ statement is not frame-dependent, unless you wish
to deny the free will of the experimenter.
So I do think that one statement is true (the one which is, in fact,
true) and the other false (the vacuous one), but I don't think that
there's a preferred frame, because I don't believe the truth of the
statements is frame-dependent.
> > > > It doesn't matter at all in what reference frames the various
> > > > experimenters and pieces of equipment are at rest. What causes
> > > > what is different in different reference frames, so if you
> > > > think that causation goes in a particular direction, you must
> > > > think that there is particular *fixed* frame to use when
> > > > determining that, and that all other frames are "wrong".
> > > Now, causation (at least in general) has some particular direction:
> > > The direction from a "free will" (or however randomly chosen) decision
> > > (what to measure) to the outcome of the experiments.
> > This is exactly my point. The free will of the experimenters means
> > that little causation arrows point away from them and towards the
> > measurement results, no matter who prefers what frame.
> And these objectively existing little arrows define the preferred frame.
In the causality loop example, that can't possibly be the case, because
if you look at the whole situation from a single frame, you'll find
those arrows pointing backwards in time, because in the preferred frame,
at least sometimes the choice of measurement axis will be made after
the corresponding distant result has been obtained.
Best wishes,
Alex.
Ralph Hartley
Fizz Fann wrote:
> Ralph Hartley wrote:
>>Fizz Fann wrote:
>>>Can you (to whatever extent you think it's reasonable) give a list of
>>>some of the other assumptions you think are necessary for somebody
>>>to accept the proof?
>>
>>That when you do a measurement, there is an absolute, unique,
>>non-subjective outcome.
>
> Ok. But it's an observation, rather than an assumption, that
> when you do a measurement, there is an absolute, unique,
> subjective outcome.
No. Subjectively there is an absolute, unique, outcome. That doesn't make
the subjective outcome absolute or unique, it only means that it seems that
way.
In the Everett interpretation, the outcome of a measurement is not unique,
but subjectively it seems like it is.
> In terms of those subjective outcomes,
> that is, for the world that we actually see and live in, the world
> will appear to be nonlocal.
So? Lots of things are not as they seem. Get used to it.
>>That the experimental method is valid at all. (I'm not saying I would give
>>that one up, but if you thought God created every event on a case by case
>>basis, would the proof even make sense?)
>
> Yeah. I've been working with the hypothesis that that isn't the
> case, but I could be wrong. I'll be very annoyed at God if it
> turns out that he's been playing games with me.
Me too, but in that case it might be better to worry if he is annoyed at
me! I put that in as an example of an assumption that even I wouldn't give up.
> those arguments would also
> be consistent explanations for a working FTL telephone.
He is correct that if your argument explains such a thing, then it must be
wrong. The kind of non-locality QM requires does not.
> I'd probably agree with him that the part inside me which
> finds them appealing is the part which longs for mysticism
> and doesn't like the harsh mechanical reality of the world.
> I very strongly suspect that others who favor these approaches
> are the same.
Tisk tisk tisk ...
Interpretations of QM don't just throw out assumptions like that and say
"anything goes", they put something equally unyealding in their place.
Lets go back to my analogy with perspective. The perspective from a given
point is a harsh mechanical reality, it just isn't unique. Other view
points are just as good. There isn't anything particularly mystical about
perspective *or* quantum mechanics, it's really as romantic as arithmetic.
>>>That is, the settings in EPR experiments,
>>>which can be controlled by gears and cogs, are "before" the results,
>>>regardless of the frames in which the experiments are done.
>>
>>Before the local result. To say that the setting always is before the
>>distant result is not consistent. For example, if the distant measurement
>>is absolutely (timelike separation) before the setting. QM doesn't seem to
>>care, you get the same statistics either way.
>
> No no. I mean before both results. Because the machinery which determines
> the settings can have been set up long ago in the intersection of the
> past light cones of the events at which the results happen.
It matters if it *has* been set up that way, not just that it could have
been. For example, suppose the interval is timelike and the distant result
proceeds the local one, I look at the distant result and make my setting
based on it. Clearly, my setting is *after* the distant result.
>>but whatever rule you use must agree with the time order for timelike
>>separations. And it must always give a consistent order, no loops.
>
> Why a consistent order? There are rules of deduction with loops
But you don't seem to be following them.
> consider Goedel's theorem. "This sentence is unprovable" - if
> it were provable it would be both false and true (a paradox!),
No, it would be both false and provable. The distinction is what keeps it
from being a loop.
> therefore it's unprovable and therefore true.
The conclusion that it is true is based on our belief that the axioms are
consistent (which is also unprovable, otherwise the Godel sentence would be
provable, and the axioms would be inconsistent).
> My system with
> the causality loop is the same - if both nonlocal effects are
> active at once, we have a paradox, therefore they're not both
> active.
I'm still not very clear on when exactly each nonlocal effect happens. It
seems to depend on what *would* have happened, which is something that
depends on the time ordering. It also isn't clear that you don't have a
situation where effect 1 happens exactly when A and B occur, and effect 2
happens exactly when ~A and C occur.
>>*Any* rule that meets those requirements is equivalent, as far as
>>experimental results are concerned. In that sense, the rule is non-physical.
>
> Nope. I've given a testable experimental prediction - that in
> certain experiments, the extent of violation of Bell's inequalities
> will be observed to be smaller than usual.
Either you predict that the results will be those given by QM (in which
case all interpretations predict the exact same thing) or you predict
something different, and you think QM is wrong. I can't tell which.
> If causality is deduction, then loops can be used there as they
> are in logic.
Causality is not deduction.
Also, loops are not generally OK in logic. "This sentence is false" is the
classic example.
>>But here we know *all* the rules. For any experiment you describe, I can
>>tell you the outcome (or you could do it yourself).
>
> Imagine that that is the only
> correlation which has been noticed and that the rest of the theory
> is unknown. Somebody who only knows this correlation could point
> at the twin paradox and correctly observe that this can't be the
> whole story.
By misapplying part of SR you can get all sorts of paradoxes, pasting
together pieces of SR and non relativistic mechanics is inconsistent.
Likewise, misapplying part of QM, pasting together bits of QM and common
sense reasoning about causality, can do the same thing.
That doesn't mean QM isn't understood, any more than it means SR isn't. QM,
properly applied is consistent (and agrees with every experiment to date).
>>We *have* a description of how nature operates (if you count
>>interpretations, we have more than one, but they are equivalent, as far as
>>results are concerned).
>
> We have no description of how nature selects a single result to
> a quantum measurement which generalises to relativistic situations,
> except possibly the preferred frame version of Bohmian mechanics.
But there is no guarantee that nature *admits* such a description (or at
least none better than Bohmian mechanics, which is a bit contrived). Would
it be nature's fault if wavefunction collapse *never* happens, or if the
results of measurements are *truly* random?
Ralph Hartley
Aaron Denney
In article <b2b24cf0.03050...@posting.google.com>, Fizz Fann wrote:
> Ilja Schmelzer <schm...@wias-berlin.de> wrote in message
> news:<i3g4r4h...@wias-berlin.de>...
>> proposal.
>
> I'm still undecided on that. In QFT, it seems that things which
> really are waves, like phonons and things which really are
> particles, like electrons, aren't treated so very differently
> by the mathematics. Simply making a Bohmian version of both
> would still not sufficiently account for their extremely different
> observable physical characteristics. Also, I'm not sure I'm
> happy with Bell's "only fermions are Bohmian" field theory either,
> since two fermions are a boson.
I'm not sure boson is the best word. The wave-function remains the same
under exchange, but the exclusion principle should still hold.
--
Aaron Denney
-><-