Why is a photon called stable?
cogito123@attglobal.net by ncar.UCAR.EDU with ESMTP id HAA06003 for
1. A photon is refered to as a stable particle. Why?
2. Could there be some kind of "conservation of photons" law?
3. It seems to me that photons must always travel at exactly c, while light
(in a refractive medium, for example) does not. Shouldn't we define c as
the speed of photons?
John Baez
>1. A photon is refered to as a stable particle. Why?
A particle which decays into other particles while merrily zipping
through the vacuum is said to be unstable. Photons don't do that.
>2. Could there be some kind of "conservation of photons" law?
No; try turning on a light.
>3. It seems to me that photons must always travel at exactly c, while light
>(in a refractive medium, for example) does not. Shouldn't we define c as
>the speed of photons?
c is defined as the speed of light in a vacuum, or equivalently,
as the speed of photons in a vacuum. If someone leaves out "in
a vacuum", they're just being lazy.
Hans Aberg
>c is defined as the speed of light in a vacuum, or equivalently,
>as the speed of photons in a vacuum.
This should be:
c is defined as the speed of light in a vacuum, or equivalently,
as the speed of photons.
In a media, the photons are absorbed and re-emitted by the atoms, which
cause time delays which makes the light appear to travel slower than c.
But the photons still move at the same speed c.
Hans Aberg * Anti-spam: remove "remove." from email address.
* Email: Hans Aberg <remove...@member.ams.org>
* Home Page: <http://www.matematik.su.se/~haberg/>
* AMS member listing: <http://www.ams.org/cml/>
Charles Francis
>I have some question about photons.
>1. A photon is refered to as a stable particle. Why?
Left alone a photon remains the same. Only when it interacts with other
matter is it absorbed.
>2. Could there be some kind of "conservation of photons" law?
No. Actually the opposite. The number of photons is not fixed and is
inherently unknown.
>3. It seems to me that photons must always travel at exactly c, while light
>(in a refractive medium, for example) does not. Shouldn't we define c as
>the speed of photons?
c is most correctly defined as the maximum theoretical speed of
information. With this definition relativity would remain the same even
if it were found that the photon has a small mass and never travels at
c.
Regards
--
Charles Francis
c.h.thompson
cogi...@attglobal.net wrote in message news:3c497...@news1.prserv.net...
> I have some question about photons.
> 1. A photon is refered to as a stable particle. Why?
I didn't think anyone said this.
[Moderator's note: pretty much all particle physicists say this. - jb]
> 2. Could there be some kind of "conservation of photons" law?
In some cases you may have approximate conservation of "photon number", but
not individual photons.
> 3. It seems to me that photons must always travel at exactly
> c, while light (in a refractive medium, for example) does not.
> Shouldn't we define c as the speed of photons?
If you are going to model light as photons, then it still consists of photons
inside the refractive medium. I believe one story of why it goes slower
there is that it is repeatedly absorbed and re-emitted. However, I would
highly recommend reading some of the history of ideas on this subject before
crystallising your own! See, for example,
Whittaker, Sir Edmund, "A History of the theories of aether and
electricity", Nelson, London (1951), then, maybe
Hendry, John, "The Creation of Quantum Mechanics and the Bohr-Pauli
Dialogue", D Reidel Publishing Company 1984
and books and articles by the early quantum theorists, in particular
Schroedinger, de Broglie. A little later comes a very interesting article
by Lamb (the same who worked out the theory of the "Lamb shift"):
Lamb, Willis E Jr., "Antiphoton", Applied Physics B 60, 77-84 (1995)
Caroline
--
c.h.th...@pgen.net
http://users.aber.ac.uk/cat/
Hans Aberg
<cha...@clef.demon.co.uk> wrote:
>c is most correctly defined as the maximum theoretical speed of
>information.
This is probably the least accurate definition, as information is a more
general concept than that of known physical fields and quantities:
For example, send out two electro-magnetic signal in two opposite
directions containing the same binary information. Then locations a equal
distances from the origin will have (disregarding GR time effects) the
same information at hand at same times. So the information has "travelled"
at speeds higher than c.
Is it possible to duplicate this experiment with a teleportation wave,
thus making information travel at higher speeds than c? That is, the
original location sends out a mixed state which is manipulated at one
location, and examined for manipulations at the other.
-- The answer I got in this group was not clear on that. Perhaps SR
covariance says that c cannot be exceeded in this way.
But the point is that speed of information may be a wider concept than c:
Physics is the science concerning physically measurable quantities and
their relatations, and in order to fit the concept of information into
that picture, one has to first define it in that context.
Hans Aberg * Anti-spam: remove "remove." from email address.
* Email: Hans Aberg <remove...@member.ams.org>
* Home Page: <http://www.matematik.su.se/~haberg/>
* AMS member listing: <http://www.ams.org/cml/>
[Moderator's note: people may enjoy discussing why Aberg's trick
cannot in fact be used to transmit information faster than light. - jb]
nicolaas.vroom@pandora.be by ncar.UCAR.EDU with ESMTP id CAA09214 for
cogi...@attglobal.net
schreef in berichtnieuws 3c497...@news1.prserv.net...
> I have some question about photons.
> 1. A photon is refered to as a stable particle. Why?
> 2. Could there be some kind of "conservation of photons" law?
> 3. It seems to me that photons must always travel at exactly c, while
light
> (in a refractive medium, for example) does not. Shouldn't we define c as
> the speed of photons?
I have a similar question.
Photons propagate in circles/spheres
If you turn a light ON and OFF than
a sphere of photons propagate in space.
This sphere grows with increasing radius.
Now suppose my light only emits one photon
1. Will this photon propagate in a sphere ?
(ie it can be detected in all directions)
2. Will this photon propagate in a cone ?
(ie it can be detected only in certain directions)
I expect #2
3. Is there a maximum distance that we can detect
this photon ?
4. If we place a 1000 by 1000 grid of CCD's
will we always detect this one photon ?
(of course we can make the grid larger)
5. If not what is the reason ?
Ed Fredkin
> Someone wrote:
The someone is me. I apologise for any confusion.
>
> >1. A photon is refered to as a stable particle. Why?
>
> A particle which decays into other particles while merrily zipping
> through the vacuum is said to be unstable. Photons don't do that.
While it is obvious that a photon doesn't decay as does a muon, isn't
it true that in its reference frame it has a lifetime of zero?
Half-life is measured in a particles frame of reference. In the case
of photons they have a lifetime of 0. It seems that there are just 2
aspects of a photons subjective "life"; creation followed immediately
by annihilation. Of course, in our reference frame, annihilation can
take place billions of years later on the other side of the Universe.
>
> >2. Could there be some kind of "conservation of photons" law?
>
> No; try turning on a light.
A photon is emitted when an electron is accelerated. An electron is
accelerated when it absorbs a photon of charge interaction. When I
turn on a light, aren't electrons absorbing photons of charge
interaction and in turn emitting photons of light?
The idea that an electron simply emits a photon seems far fetched.
Doesn't it make sense to assume that some event causes an electron to
emit a photon? And if so, doesn't that event have to be the
interaction of the electron with some particle, such as a photon of
charge interaction?
If an electron can be deflected by an electric field without changing
its spin, doesn't that imply that such interactions with photons of
charge interaction must involve pairs of photons; one absorbed and one
emitted, in order to conserve angular momentum?
>
> >3. It seems to me that photons must always travel at exactly c, while light
> >(in a refractive medium, for example) does not. Shouldn't we define c as
> >the speed of photons?
>
> c is defined as the speed of light in a vacuum, or equivalently,
> as the speed of photons in a vacuum. If someone leaves out "in
> a vacuum", they're just being lazy.
Hans Aberg says:
"In a media, the photons are absorbed and re-emitted by the atoms,
which
cause time delays which makes the light appear to travel slower than
c.
But the photons still move at the same speed c."
Aberg's comment is exactly what I was thinking of.
c.h.thompson
Hans Aberg <remove...@matematik.su.se> wrote in message
news:remove.haberg-2...@du131-226.ppp.su-anst.tninet.se...
> ba...@galaxy.ucr.edu (John Baez) wrote:
>
> >c is defined as the speed of light in a vacuum, or equivalently,
> >as the speed of photons in a vacuum.
>
> This should be:
>
> c is defined as the speed of light in a vacuum, or equivalently,
> as the speed of photons.
>
> In a media, the photons are absorbed and re-emitted by the atoms, which
> cause time delays which makes the light appear to travel slower than c.
> But the photons still move at the same speed c.
Might I ask what evidence there is for this? All we can actually measure is
light going in and light coming out of a block of glass -- and we can't even
really "measure" that! We can only try and deduce things from the angle of
refraction and, perhaps, from interference effects if we split a beam and
send part through the glass, part not. What evidence is that that within
the glass the light is proceeding in little jumps, from electron to electron
or whatever?
Qunatum theory recognises that light is a wave as well as a particle. Why
is this apparently ignored?
Charles Francis
<c.h.th...@pgen.net> writes
>cogi...@attglobal.net wrote in message news:3c497...@news1.prserv.net...
>> 2. Could there be some kind of "conservation of photons" law?
>In some cases you may have approximate conservation of "photon number", but
>not individual photons.
Photon number is not definable as an observable quantity. This has to do
both with the fact that there can be an infinitely large number of
photons with energy tending to zero (net energy finite), and also to do
with the fact that observable for bosons the field commutes and as a
result physical quantities depend on the derivative of the field, not
the field itself. As a result only changes in the number of photons are
observable, not the actual number of photons.
Regards
--
Charles Francis
Charles Francis
writes
>I have a similar question.
>Photons propagate in circles/spheres
>If you turn a light ON and OFF than
>a sphere of photons propagate in space.
>This sphere grows with increasing radius.
>
>Now suppose my light only emits one photon
>1. Will this photon propagate in a sphere ?
No. The photon wave function will propagate in a sphere, but all that
really does is tell us that the photon is equally likely to be found in
any direction, not that the photon travels in more than one direction.
>(ie it can be detected in all directions)
The photon will only be detected at one point.
>2. Will this photon propagate in a cone ?
>(ie it can be detected only in certain directions)
Of course no practical light source can have perfect spherical symmetry,
so really the wave function propagates conically rather than
spherically, but that does not affect the fact that the photon will be
detected at a point, not spread out over an area.
>I expect #2
>3. Is there a maximum distance that we can detect
>this photon ?
As far away as you can place the apparatus. We can detect photons which
have propagated from within a few seconds of the big bang.
>4. If we place a 1000 by 1000 grid of CCD's
>will we always detect this one photon ?
Not in practice, but if you could build a perfect grid that would detect
every photon passing through then yes.
Regards
--
Charles Francis
Ed Fredkin
news:<remove.haberg-2...@du128-226.ppp.su-anst.tninet.se>...
> In article <ifFRGeCk...@clef.demon.co.uk>, Charles Francis
> <cha...@clef.demon.co.uk> wrote:
> >c is most correctly defined as the maximum theoretical speed of
> >information.
> This is probably the least accurate definition, as information is a more
> general concept than that of known physical fields and quantities:
>
> ... speed of information may be a wider concept than c:
> Physics is the science concerning physically measurable quantities and
> their relatations, and in order to fit the concept of information into
> that picture, one has to first define it in that context.
There may be more to information than communication. CPT symmetry
implies that a process going backwards in time (with C and P also
reversed) is good physics. That must mean that as a process evolves,
the information (necessary for the reversed process) must be
conserved. This really means that the total quantity of information
must be a constant. Further, the information that describes the state
prior to an interaction must be related to the information that
describes the state after the interaction by a bijective function.
What these 2 observations mean is that it is hard to understand the
present view of processes like the decay of a muon as the
informational equations don't make sense.
The only good informational models I can think of for inelastic
processes must involve additional particles, currently unknown as
involved in such processes, or some other kinds of lower modes for
information, also currently unknown. If someone understands all this
and can help me out, please do so.
Ed F
John Baez
Ed Fredkin <e...@fredkin.com> wrote:
>The idea that an electron simply emits a photon seems far fetched.
Maybe so, but this is one of the basic processes in quantum
electrodynamics:
\
\
~~~~~~~~
/
/
so there is no conservation of photons in this (well-established)
theory.
PS - Do you remember the time we ate in a Chinese restaurant in
Kendall square with our mutual friends Mark Smith, Tom Toffoli,
and Norm Margolus? I remember getting into an argument over
whether the world was fundamentally discrete. You may be pleased
to know that now I think it might be. But not a cellular automaton!
PPS - What are you up to these days? Do you still live on that
island you bought?
Ed Fredkin
> In article <386d7a77.02012...@posting.google.com>,
> Ed Fredkin <e...@fredkin.com> wrote:
> >The idea that an electron simply emits a photon seems far fetched.
> Maybe so, but this is one of the basic processes in quantum
> electrodynamics:
>
> \
> \
> ~~~~~~~~
> /
> /
>
> so there is no conservation of photons in this (well-established)
> theory.
Maybe so, but the diagram above is not a physically realizable event;
rather it is part of a larger diagram that could represent a real
event. While the single vertex can conserve both momentum and energy,
the total process represented by one vertex cannot have E^2 =
p^2c^2+m^2c^4.
Diagrams that represent physically realizable events, such as the one
below, contain virtual particles. If there is some kind of law of
conservation of photons, it would have to involve some other concept
of, as yet unknown, kinds of photons. The motivation has to do with
the idea that complete, real events need to conserve information
(because of CPT reversibility). For many such events, additional
particle seem to be needed.
\ /
\ /
~~~~~~~~
/ \
/ \
> PS - Do you remember the time we ate in a Chinese restaurant in
> Kendall square with our mutual friends Mark Smith, Tom Toffoli,
> and Norm Margolus?
I do have a vague recollection, but it's from long ago.
> I remember getting into an argument over
> whether the world was fundamentally discrete. You may be pleased
> to know that now I think it might be. But not a cellular automaton!
Well, I'm now is a much better position for trying to convince you.
Take a look at www.digitalphilosophy.org especially "Introduction to
Digital Philosophy"
>
> PPS - What are you up to these days?
I left MIT 2 decades ago, did 6 years as physics prof at BU, currently
on the CMU faculty, but only visit there now and then. I'm working on
what I call Digital Philosophy which is mostly about discrete models
of physics.
>Do you still live on that island you bought?
I hate to squelch such a charming rumor but… Moskito Island, which I
bought in 1968, has only seen me on visits. My wife and I go down a
few times a year, for a few days up to a few weeks. Somehow the place
appeals to physicists. It's where Wolfram saw his first Cellular
Automata. I'd like to say that others like K. Wilson, G. ‘t Hooft,
R. Feynman, C. Rebbi, E. Fredkin, C. Bennett, R. Landauer, T. Toffoli,
N. Margolus, J. McCarthy, M. Minsky, J. Cocke and many others all did
their best work on Moskito Island, but I'm not sure.
Hans Aberg
<c.h.th...@pgen.net> wrote:
>> In a media, the photons are absorbed and re-emitted by the atoms, which
>> cause time delays which makes the light appear to travel slower than c.
>> But the photons still move at the same speed c.
>Might I ask what evidence there is for this?
This was extensively discussed in this list some time ago (you may try a
repost on the subject). The main outcome is that it is a model that one
can compute predictions with that conforms with many observed phenomenon.
Modern physics does not try to say what really happens, only puts forth
theories that conform with observations.
>Quantum theory recognises that light is a wave as well as a particle.
>Why is this apparently ignored?
The wave behavior is a fundamental fact of the QM behavior of photons as
well. One experiment is to send the photons one-by-one onto a slit and
observe that the interference pattern remains, which would not happen if
photons were only particles. One can do similar things with electrons, so
electrons are QM waves as well.
So one has QM particle-waves that sometimes exhibit the behavior of a
particle, and sometimes that of a wave, in concordance with observed
behavior.
c.h.thompson
news:a2nssm$bke$1...@sue.its.caltech.edu...
> nicolaa...@pandora.be writes
> >Now suppose my light only emits one photon
> >1. Will this photon propagate in a sphere ?
>
> No. The photon wave function will propagate in a
> sphere, but all that really does is tell us that the
> photon is equally likely to be found in any direction,
> not that the photon travels in more than one direction.
>
> The photon will only be detected at one point.
How thoroughly has this been tested? It sounds plausible if the "photon"
concerned is produced by a nuclear event, in which case it may well be
produced as a very narrow beam, but what if it is ordinary visible light, or
radio frequency?
> >4. If we place a 1000 by 1000 grid of CCD's
> >will we always detect this one photon ?
I strongly suspect that you could sometimes detect more than one. I don't
know so much about how CCD's work but have studied some experiments
involving very low intensities of visible light. Some photomultipliers seem
to require the addition of local electromagnetic noise in order to produce a
"detection". The visible light will be spread out over the face of the
photocathode, arriving over a finite period of time (20 ns or so, in the
case in point). Whether or not it is detected, and when, depends on local
random factors. Clearly there is a possibility that if two detectors were
placed there instead of one, both could sometimes fire.
Charles Francis
<remove.haberg-2...@du128-226.ppp.su-anst.tninet.se>, Hans
Aberg <remove...@matematik.su.se> writes:
>In article <ifFRGeCk...@clef.demon.co.uk>, Charles Francis
><cha...@clef.demon.co.uk> wrote:
>>c is most correctly defined as the maximum theoretical speed of
>>information.
>This is probably the least accurate definition, as information is a more
>general concept than that of known physical fields and quantities:
>For example, send out two electro-magnetic signal in two opposite
>directions containing the same binary information. Then locations a equal
>distances from the origin will have (disregarding GR time effects) the
>same information at hand at same times. So the information has "travelled"
>at speeds higher than c.
This is probably the least accurate definition of speed I have
ever seen, since it accurately defines speed to be something which it is
not.
There are very good reasons for using the definition of c as the maximum
speed of information, because it remains an accurate definition in a
quantum context, and does not depend on such things as whether light is
travelling in a medium.
>Is it possible to duplicate this experiment with a teleportation wave,
>thus making information travel at higher speeds than c? That is, the
>original location sends out a mixed state which is manipulated at one
>location, and examined for manipulations at the other.
Look up the EPR experiment and Bell's theorem. It has been widely
discussed both here on s.p.r and in the literature as to why no
information can travel like this. There's probably something in the FAQ.
>But the point is that speed of information may be a wider concept than c:
>Physics is the science concerning physically measurable quantities and
>their relatations, and in order to fit the concept of information into
>that picture, one has to first define it in that context.
And one does define information in that context. Information is the
information about measurable quantities and their relations as found in
physics, and there is a maximum theoretical speed at which information
can be transmitted. As this is an absolute theoretical value, there is
no ambiguity or physical constraint to qualify it, and it is more
accurate that any attempt at experimental definition.
>[Moderator's note: people may enjoy discussing why Aberg's trick
>cannot in fact be used to transmit information faster than light. - jb]
What trick? He isn't transmitting information from A to B, but merely
from C to A and C to B where C is a point between the two.
[Moderator's note: Right. That's why it doesn't transmit information
faster than light. - jb]
--
Charles Francis
Maury Markowitz
> > cause time delays which makes the light appear to travel slower than c.
> > But the photons still move at the same speed c.
>
> Might I ask what evidence there is for this?
By "this" do you refer to photons in general, or photons travelling at c
"even" in a medium? From your following text it's not terribly clear what
"this" you are asking about. Oh wait, I'm wrong:
> send part through the glass, part not. What evidence is that that within
> the glass the light is proceeding in little jumps, from electron to
electron
> or whatever?
The simple observation that some of the light "goes right though" the
glass with time t = d/c seems like ample evidence. If light refracts due to
a classical-like system (a wavefront being slowed, or photons being slowed
for a hybrid system) you would certainly not expect this sort of behaviour.
There are any number of less common and more complex experimental results
that also suggest the same thing. Cherenkov radiation, certain detectors,
behaviour in BEC's, etc.
> Qunatum theory recognises that light is a wave as well as a particle.
Why
> is this apparently ignored?
Because quantum theory doesn't really say anything of the sort. "particle"
and "wave" are terms we still use to describe facets of the behaviour of a
greater whole, terms that are still used in this context for historical
reasons more than any other. Although I would say that the general
conceptual model for things like photons still has a particle-like "thing"
in most cases, I think it's safe to say that the same model doesn't include
wave-ish behaviours - that is, while most would be happy talking about a
photon as a particle, they don't ascribe inherent wave behviours to it -
it's not a little ball wiggling up and down.
Maury
eric alan forgy
On Thu, 24 Jan 2002, John Baez wrote to Ed Fredkin:
> PS - Do you remember the time we ate in a Chinese restaurant in
> Kendall square with our mutual friends Mark Smith, Tom Toffoli,
> and Norm Margolus? I remember getting into an argument over
> whether the world was fundamentally discrete. You may be pleased
> to know that now I think it might be. But not a cellular automaton!
OK, I'm sure you were aware of the subsequent fallout when you wrote this
:)
Is there something in particular recently that has made you take this idea
more seriously than perhaps you might have in the past that "the world was
fundamentally discrete"? Does it have to do with smooth manifolds being
used to set up the spin foam model, but after you have constructed the
spin foam states, then the fact that they were built from smooth manifolds
becomes immaterial and you could just as well have STARTED with a spin
foam?
I don't have Geroch's masterfully beautiful paper [1] at hand right now,
but I was able to grab this blurb from Zapatrin, "Polyhedral
representations of discrete differential manifolds," J. Math Phys 38 (5),
May 1997.
"... The environment of the present paper is formed by the following scope
of ideas and techniques.
The first belongs to Geroch [1] and asserts that even in classical
general relativity the notion of the space-time manifold is essentially
used only once: to set up the algebra of smooth functions. We can instead
start from the (commutative) algebra as the basic object of the theory.
[snip]
If we accept an algebra to be the starting object (called the basic
algebra), we can try to go beyond the class of commutative Banach algebras
representable by functions on manifolds. In particular, we can assume
these algebras to be noncommutative, which gives rise to the
noncommutative geometry. Another opportunity that looks very attractive
from the computational point of view is to assume the basic algebra to be
finite-dimensional and commutative."
I am writing my PhD thesis now with this philosophy lying at the heart of
it. I try not to bring it up here too often (I don't think I have until
now) in fear of being labeled a "crackpot" :)
I don't know many recognized physicists (Sorkin is a nice exception) that
openly admit to this suspicion, but I bet that many of them do secretly
think the world may be fundamentally discrete.
The trouble with having a serious discussion about this topic is that the
idea is so vague. What does it really mean for the world to be
fundamentally discrete? I don't think anyone can really make this idea
concrete mathematically (not yet anyway). At least not in a universally
acceptable manner. I admit I am trying to do just this, but unfortunately
I am not very good in mathematics and I was not given the gift of physical
insight like Einstein's, but I keep chugging along anyway with my meager
resources :)
So, having bared my soul, would you or anyone else care to elaborate on
their opinion on the subject?
Eric
Throwing all caution to the wind.
[1] R. Geroch, "Einstein Algebras," Commun Math Phys 26, 271-276
(1972)
John Baez
>ba...@galaxy.ucr.edu (John Baez) wrote in
>message news:<a2ocdp$98e$1...@glue.ucr.edu>...
>> In article <386d7a77.02012...@posting.google.com>,
>> Ed Fredkin <e...@fredkin.com> wrote:
>> >The idea that an electron simply emits a photon seems far fetched.
>> Maybe so, but this is one of the basic processes in quantum
>> electrodynamics:
>>
>> \
>> \
>> ~~~~~~~~
>> /
>> /
>>
>> so there is no conservation of photons in this (well-established)
>> theory.
>Maybe so, but the diagram above is not a physically realizable event;
>rather it is part of a larger diagram that could represent a real
>event. While the single vertex can conserve both momentum and energy,
>the total process represented by one vertex cannot have E^2 =
>p^2c^2+m^2c^4.
Right, but from this basic process it's easy to build fancier ones
that conserve energy and momentum but violate photon conservation,
and these happen all the time: the best-known is the annihilation
of a positron-electron pair into 2 photons:
s s
s s
s s
s s
s->-s
/ \
/ \
| |
^ v
| |
| |
>If there is some kind of law of
>conservation of photons, it would have to involve some other concept
>of, as yet unknown, kinds of photons.
Okay, fine. When someone finds these, I'll get interested.
>The motivation has to do with
>the idea that complete, real events need to conserve information
>(because of CPT reversibility). For many such events, additional
>particle seem to be needed.
Ordinary quantum field theory is CPT invariant and conserves
information without needing to postulate "as yet unknown kinds
of photons".
>> PS - Do you remember the time we ate in a Chinese restaurant in
>> Kendall square with our mutual friends Mark Smith, Tom Toffoli,
>> and Norm Margolus?
>I do have a vague recollection, but it's from long ago.
About ten years ago....
>> I remember getting into an argument over
>> whether the world was fundamentally discrete. You may be pleased
>> to know that now I think it might be. But not a cellular automaton!
>Well, I'm now is a much better position for trying to convince you.
>Take a look at www.digitalphilosophy.org especially "Introduction to
>Digital Philosophy"
I'll take a look, but I'll warn you right off the bat: I'm a
dogmatic proponent of orthodox views, a hidebound reactionary
of the worst ilk.
>I'm working on
>what I call Digital Philosophy which is mostly about discrete models
>of physics.
Have you talked to Steve Wolfram about this stuff lately? I gather
that his perpetually forthcoming book is supposed to describe a bunch
of models of this sort.
nicolaas.vroom@pandora.be by ncar.UCAR.EDU with ESMTP id CAA13223 for
Charles Francis <cha...@clef.demon.co.uk> schreef in nieuwsbericht
> In article <3C4D2984...@pandora.be>,
> nicolaa...@pandora.be writes
>
> >4. If we place a 1000 by 1000 grid of CCD's
> >will we always detect this one photon ?
>
> Not in practice, but if you could build a perfect grid
> that would detect every photon passing through then yes.
>
My target is to study the two slit experiment
consisting of two slits L,Left and R, Right.
But first I want to see what happens if
only one slit is open.
For my experiments I use a single photon generator.
In the first part of this experiment I use my
photon generator 100 times, only the L slit is open.
In my 1000*1000 grid I (will) detect 100 hits.
(The distribution of hits is not the subject
of this post)
In the second part of this experiment I ^only^
close the L slit and I open the R slit.
The setup/direction of the photon generator
stays exactly the same.
Again I use the photon generator 100 times.
Question will there be 100 hits ?
Suppose the answer is yes.
In that case is than the following conclusion
justified:
Each single photon has the capability to go through
anyone of the two slits L and R.
When both are closed the photon it will not go through.
When one is closed it will go through the other.
Suppose the answer is no.
Nick
Tony Smith
In a post to s.p.r. from Eric Alan Forgy on the thread
World is fundamentally discrete? (was: Why is a photon called stable?)
EAF said:
"...John Baez wrote to Ed Fredkin ... I remember getting into an argument
over whether the world was fundamentally discrete.
You may be pleased to know that now I think it might be.
EAF continued, asking:
"... What does it really mean for the world to be fundamentally discrete?
...".
David Finkelstein, whose web page at
http://www.physics.gatech.edu/people/faculty/dfinkelstein.html#Research
generally describes some of his current ideas,
including a Clifford algebra model that might be
describable as "fundamentally discrete",
has a weekly seminar that I informally attend from time to time,
and
this past week (on 23 January 2002) we talked a bit about
a "fundamentally discrete" physical world.
Since my model can also be regarded as "fundamentally discrete",
and is also based on Clifford algebras, as described in
http://www.innerx.net/personal/tsmith/HDFCmodel.html
and
http://xxx.lanl.gov/abs/hep-ph/9708379
our talk did not rise to the level of thesis-antithesis-synthesis,
but was just a discussion, centered on another Clifford-algebra
related thing, the hyperfinite II1 von Neumann algebra factor.
As John Baez has said:
"... the hyperfinite II1 factor
is a kind of infinite-dimensional Clifford algebra.
...[formed by taking]... the union of all
these algebras ... of 2n x 2n matrices [which are]... Clifford algebra[s]
... and ... complet[ing] th[at union of Clifford algebras]...".
At the seminar, we discussed such questions as:
If Physics is really Fundamentally Finite,
what is the point of
taking the union of ALL 2n x 2n matrix Clifford algebras
and completing it?
Why not just use a big (but finite) bunch of Clifford algebras,
or,
in view of Clifford periodicity, by which a very big Clifford algebra
can be written as the tensor product of a lot of small ones,
Why not just use a big (but finite) tensor-product of
a small Clifford algebra ?
IIRC, at David's seminar we all philosophically agreed
that we liked the idea of Fundamentally Finite Physics,
but
we also realized that there are some useful reasons to go to
Infinite Limits, such as the fact that
an Infinite Limit Structure allows us to make Continuous Structures
and related things for which calculation can be much simpler
than calculations using large Finite Structures
(for example, my calculations of particle masses
and force strength constants
http://www.innerx.net/personal/tsmith/HDFCmodel.html#calculations
use Continuous Structures that I view as useful approximations
to an underlying discrete fundamental physical world).
Another useful aspect of the existence of an Infinite Limit Structure
is that, if an Infinite Limit Structure exists, then you know
that there will be no pathological surprises as N gets bigger and bigger.
So, even if you never use the Infinite Limit Structure itself,
its very existence would give you confidence that everything is OK
for ALL of your N-level Finite Structures,
sort of like epsilon-delta proofs in math analysis assure you that
calculus really works, and any finite-difference approximations
will be consistent, not only with the "ideal" infinitesimal result,
but also with each other.
From the math-analysis point of view,
maybe you could say that Fundamentally Finite Physics is sort of
like finite-difference methods using computers,
while Infinite Limit Physics is sort of like the infintesimals etc
of NonStandard Analysis.
Also,
there is a practical drawback to doing Fundamentally Finite Physics:
You don't have a single (perhaps large) number N
that you are certain is big enough to be applicable
to describe ANY experiment that you might want to do.
In other words,
with Fundamentally Finite Physics, you have to pick a big-enough N
for EACH proposed experiment on a case-by-case basis,
while
if you use an Infinite Limit Structure,
you know that it is applicable for everything.
Maybe I should not have used the term "drawback", because in the
real world of experimental physics, you DO have to make case-by-case
decisions about your experimental setup,
so
in fact the Fundamentally Finite Physics is closer to real
experimental physics as done by humans here on Earth,
while
Infinite Limit Structure Physics is more like a Platonic Ideal.
--------------------------------
The above comments are by no means a definitive discourse
on anything, but maybe they contain the germs of some useful points.
--------------------------------
Further, in a related thread, John Baez asked Ed Fredkin:
"... Have you talked to Steve Wolfram about this stuff lately? ...".
In his book A New Kind of Science, excerpts of which are
(or have been) on the web at
http://www.wolframscience.com/
Stephen Wolfram says (at page 531):
"... Ultimate Models for the Universe. ...
could it even be that underneath all the complex phenomena
that we see in physics there lies a simple program which,
if run for long enough would reproduce every detail of our universe? ..."
In an interview published in the New Scientist
(25 August 2001, pages 44-47),
Stephen Wolfram was asked by Marcus Chown:
"... Have you discovered the simple program
that is generating the Universe? ...",
to which question Stephen Wolfram replied:
"... Not yet.
But I have found increasing evidence that it exists ...".
Tony 26 January 2002
Charles Francis
single photons you should find that only a very small proportion of them
go through either slits, the vast majority will be stopped by the screen
and you will not detect them.
In article <3C51245A...@pandora.be>, nicolaa...@pandora.be
writes
>
>
>Charles Francis <cha...@clef.demon.co.uk> schreef in nieuwsbericht
>> In article <3C4D2984...@pandora.be>,
>> nicolaa...@pandora.be writes
>
>>
>> >4. If we place a 1000 by 1000 grid of CCD's
>> >will we always detect this one photon ?
>>
>> Not in practice, but if you could build a perfect grid
>> that would detect every photon passing through then yes.
>>
>
>My target is to study the two slit experiment
>consisting of two slits L,Left and R, Right.
>But first I want to see what happens if
>only one slit is open.
>
>For my experiments I use a single photon generator.
>
>In the first part of this experiment I use my
>photon generator 100 times, only the L slit is open.
>In my 1000*1000 grid I (will) detect 100 hits.
>(The distribution of hits is not the subject
>of this post)
>
Regards
--
Charles Francis
c.h.thompson
Maury Markowitz <ma...@sympatico.ca> wrote in message
news:a2tion$s5g$1...@inky.its.caltech.edu...
> > > In a media, the photons are absorbed and re-emitted by the
> > > atoms, which cause time delays which makes the light appear
> > > to travel slower than c. But the photons still move at the
> > > same speed c.
> >
> > Might I ask what evidence there is for this?
>
> By "this" do you refer to photons in general, or photons
> travelling at c "even" in a medium?
The latter.
> The simple observation that some of the light "goes right though"
> the glass with time t = d/c seems like ample evidence.
Ah, I did not know this happened! Can you give me a ref to the experiments?
Keeping light at speed c even in glass is one of the possibilities I've
considered, but I thought the evidence was against it.
> If light refracts due to a classical-like system ...
I'm currently studying ideas on this from around 1800 onwards, from
Whittaker's Theories of the Aether. I'm a bit doubtful whether anyone has
yet got the right theory.
> There are any number of less common and more complex
> experimental results that also suggest the same thing. Cherenkov
> radiation, certain detectors, behaviour in BEC's, etc.
I've read a description of how Cherenkov radiation works, in Kleinknecht,
Konrad, "Detectors for Particle Radiation", Cambridge University Press,
1986. This was purely in terms of waves, with the light travelling slower
than the particle.
BEC's also fit very neatly into a classical picture, but these dont' involve
a solid medium, do they? Also, if we're dealing with standing waves, then
the chances are that the speed is not very relevant.
Hans Aberg
I think that you both got the trick wrong: C sends out a mixed state, say
a spin up and down. It will remain so until A distorts the state say by
measuring it. Then B measures it: If B finds a down (resp. up) state, B
will know that A found an up (resp. down) state.
If the scheme is changed so that A can force the mixed state measured to
be up (resp. down), then that information is "teleported" to B which will
measure down (reps. up), thus constituting the transport of information
that does not rely on transportation of any physical quantity. It relies
on the fact that the QM field is spread all over the universe
simultaneously like wave.
>Look up the EPR experiment and Bell's theorem. It has been widely
>discussed both here on s.p.r and in the literature as to why no
>information can travel like this. There's probably something in the FAQ.
I think that these deal with completely different matters. I have a vague
memory of this, but I have not re-looked it up when writing this. I think
you could not get out anything about the speed of the teleportation from
that stuff you mention.
I still think that one should speak about the speed of "information"
without explicitly defining which physical quantities it involves.
Hans Aberg
<cha...@clef.demon.co.uk> wrote:
>>Is it possible to duplicate this experiment with a teleportation wave,
>>thus making information travel at higher speeds than c? That is, the
>>original location sends out a mixed state which is manipulated at one
>>location, and examined for manipulations at the other.
>
>Look up the EPR experiment and Bell's theorem. It has been widely
>discussed both here on s.p.r and in the literature as to why no
>information can travel like this. There's probably something in the FAQ.
I have now looked up the FAQ:
http://math.ucr.edu/home/baez/physics/Quantum/bells_inequality.html
I am not sure what you mean here, because I recall that there were
experiments to the effect that there are no local hidden variables and QM
thus is correct (only global hidden variables -- the QM fields), at last
over a few meters distance.
Therefore one should be able to teleport information this way (using say
an electron pair in mixed up/down sate until measurement) at least over a
few meters, and how do you then deduce from QM that the speed does not
exceed c?
c.h.thompson
Hans Aberg <remove...@matematik.su.se> wrote in message
news:remove.haberg-2...@du131-226.ppp.su-anst.tninet.se...
> In article <3c4e7...@news1.vip.uk.com>, "c.h.thompson"
> <c.h.th...@pgen.net> wrote:
> >> In a media, the photons are absorbed and re-emitted by the atoms, which
> >> cause time delays which makes the light appear to travel slower than c.
> >> But the photons still move at the same speed c.
>
> >Might I ask what evidence there is for this?
[snip]
> >Quantum theory recognises that light is a wave as well as a particle.
> >Why is this apparently ignored?
>
> The wave behavior is a fundamental fact of the QM behavior of photons as
> well. One experiment is to send the photons one-by-one onto a slit and
> observe that the interference pattern remains, which would not happen if
> photons were only particles.
Can you give me a ref for the relevant experiment? The only ones that I
have studied have not been convincing. There have always been other
possible intrepretations, in which the observed low numbers of coincidences
were explained not by the absence of "actual" coincidences but by other
relevant experimental factors such as the way the detectors behave.
Perhaps I have never seen an actual test of the basic double-slit
experiment, but the following test for coincidences between beams emerging
from a beamsplitter involves the same assumptions:
Grangier, P, G Roger and A Aspect, "Experimental Evidence for a photon
anticorrelation effect on a beam splitter: a new light on single-photon
interferences", Europhysics Letters 1, 173-179(1986)
See Marshall, T W and Santos, E, Europhysics Letters, 3, 293-6 (1987) for a
"classical" rejoinder.
> One can do similar things with electrons, so
> electrons are QM waves as well.
Yes, but that's another story!
Ralph E. Frost
eric alan forgy <fo...@students.uiuc.edu> wrote in message
news:Pine.GSO.4.31.02012...@ux12.cso.uiuc.edu...
> The trouble with having a serious discussion about this topic is that the
> idea is so vague. What does it really mean for the world to be
> fundamentally discrete?
Holy, holy, holy.
Discrete, yet connected.
>I don't think anyone can really make this idea
> concrete mathematically (not yet anyway). At least not in a universally
> acceptable manner. I admit I am trying to do just this, but unfortunately
> I am not very good in mathematics and I was not given the gift of physical
> insight like Einstein's, but I keep chugging along anyway with my meager
> resources :)
Nested polyhedral lattices or networks make pretty good approximations and
might be close enough to be acceptable at the grade-school and high school
levels. College folks, however, would like graduate to work with more
sophisticated spin networks and variants.
zirkus
>In particular, we can assume
> these algebras to be noncommutative, which gives rise to the
> noncommutative geometry.
> I am writing my PhD thesis now with this philosophy lying at the heart of
> it. I try not to bring it up here too often (I don't think I have until
> now) in fear of being labeled a "crackpot" :)
A key point of this philosophy is:
Normally in physics you only need the local picture (with trivial
bundles in each open set). However, for a general noncommutative
algebra there may be no reasonable "open sets" so you have to develop
the global picture a priori. Are you familiar with the idea of
nontrivial bundles because they are needed for dealing with magnetic
monopoles, the Aharanov-Bohm effect, etc.
I can't remember the topic of your thesis but, if you want, tell me
when it is about to be completed and I will see if I can find anything
that might be relevant from the worlds of stochastic or noncommutative
differential geometry.
Also, don't worry if wizards might seem a bit irratible because, after
all, they are only wizards and not patient gods. For instance, many
years ago, the famous mathematician Raoul Bott wrote his masters
thesis on the mathematics of impedance matching but said that even to
this day he still has doubts about the mathematical rigor of his
thesis:
J. J. Lodder
> In a post to s.p.r. from Eric Alan Forgy on the thread
> World is fundamentally discrete? (was: Why is a photon called stable?)
> EAF said:
> "...John Baez wrote to Ed Fredkin ... I remember getting into an argument
> over whether the world was fundamentally discrete.
> You may be pleased to know that now I think it might be.
> But not a cellular automaton! ...".
> EAF continued, asking:
> "... What does it really mean for the world to be fundamentally discrete?
Nothing actually,
for such a statement is inherently not falsifiable,
and therefore not scientific.
It is always possible that below a 'fundamental' discrete theory
there lies an even more 'fundamental' deeper continuous one,
and visa versa.
Jan
Ed Fredkin
> In a post to s.p.r. from Eric Alan Forgy on the thread
> World is fundamentally discrete? (was: Why is a photon called stable?)
> EAF said:
> "...John Baez wrote to Ed Fredkin ... I remember getting into an argument
> over whether the world was fundamentally discrete.
> You may be pleased to know that now I think it might be.
> But not a cellular automaton! ...".
> EAF continued, asking:
> "... What does it really mean for the world to be fundamentally discrete?
> ...".
>
contents of space-time would all be discrete. We call this hypothesis
"Finite Nature". Logical consequences of the Finite Nature
assumption include no infinitesimals or infinities and no truly random
events.
Finite Nature differs from our current mathematical models in that a
totally discrete model is basically an Automaton (as is defined in
computer science). That means that a discrete model can be directly
and exactly implemented as a process by programming it on a computer.
This is true for all totally discrete models. This is a major point:
the mathematics of continuous variables encapsulates our knowledge of
physical processes, and it lets us make approximate predictions about
the behavior of simple systems. For some systems there are solutions
parametric in time. Finite Nature lets us make exact working models,
but only at the most microscopic levels. If the Finite Nature
assumption were to turn out to be true, then it should be possible to
derive, analytically, the laws of physics from the definition of the
process that runs in the automaton. However, with minor exceptions,
exact solutions that are parametric in time are unobtainable for
automata. To see exactly what happens as an automaton runs, you have
to run it step by step.
There are a number of experiments that could shed light on this
question. The primary experiment would be the finding that that there
are processes that violate rotational and translational symmetries.
This would be similar to the way that charge (C), parity (P), and time
(T) symmetries along with CP etc. are violated.
These issues are the subject of papers posted at
www.digitalphilosophy.org
In particular see, "Finite Nature", "A Physicist's Model of
Computation" and "Introduction to Digital Philosophy"
Ed F
zirkus
> I am not sure what you mean here, because I recall that there were
> experiments to the effect that there are no local hidden variables and QM
> thus is correct (only global hidden variables -- the QM fields), at last
> over a few meters distance.
According to the following paper, it is not yet clear how to interpret
the physical meaning of Bell's probabilistic assumptions, and thus
local realism in QT might still be possible:
Nicolaas Vroom
Charles Francis wrote:
>
> I'm afraid I don't understand your experiment at all. If you generate
> single photons you should find that only a very small proportion of them
> go through either slits, the vast majority will be stopped by the screen
> and you will not detect them.
>
> Regards
>
> --
> Charles Francis
The target of my experiments is to do the
true two slit experiments with single photons
but before I do that I first do a set of
3 pre experiments to learn more.
First I do an experiment without any screen
with slits.
I have only a single photon generator
and CCD detectors of 100*100 CCD's
I generate 100 single photons
How many hits will I get?
The place is not important.
I expect 100. (or 99)
Not more.
(otherwise this are not single photons.)
Second I place a screen with two slits
L and R between the generator and the CCD detectors
However only the L slit is open
The R slit is closed.
I generate 100 single photons
How many hits will I get?
Accordingly to your reply much less then 100
Assume we get 10 plus or min 1
Third exactly the same as above but now:
only the R slit is open
The L slit is closed.
I generate 100 single photons
How many hits will I get?
Accordingly to your reply much less then 100
Assume we get 10 plus or min 1
Fourth exactly the same as above but now:
L and R slit are open.
I generate 100 single photons
How many hits will I get?
20 plus or min 1 ?
10 plus or min 1 ?
(We first have to agree about the numbers
the position were comes later)
Did you ever do the experiment in this way
considering those 4 possibilities ?
My experience is only using a stream of continuous
light/photons through a grid.
Unfortunate I do not have the possibilities
to perform the above 4 experiments.
Nicolaas Vroom
p.ki...@ic.ac.uk
Charles Francis <cha...@clef.demon.co.uk> wrote:
>>1. Will this photon propagate in a sphere ?
> No. The photon wave function will propagate in a sphere, but all that
> really does is tell us that the photon is equally likely to be found in
> any direction, not that the photon travels in more than one direction.
>>(ie it can be detected in all directions)
> The photon will only be detected at one point.
But only if we have a detector of photons that returns information
about some localised region of space in which the photon was detected.
I could enclose an excited atom with a spherical photodetector
which cannot tell me anything about where on it a photon
might land. If the atom were to de-excite and emit a photon,
saying
> The photon will only be detected at one point.
is clearly wrong. Further, the only sensible description of the
photon in this case would use (or a the very least include) a set of
spherical modes centered on the atom. Of course to be properly
non-localizing, the photon detector would operate by absorbing
the photon and "converting" it into some some non-local excitation
(random e.g. a surface plasmon), and this excitation would then
be measured by some other process.
>>2. Will this photon propagate in a cone ?
>>(ie it can be detected only in certain directions)
> Of course no practical light source can have perfect spherical symmetry,
> so really the wave function propagates conically rather than
> spherically, but that does not affect the fact that the photon will be
> detected at a point, not spread out over an area.
This is both misleading and wrong. Wrong in the case I described
above, and misleading in suggesting there is anything intrisically
"pointlike" about photons.
--
---------------------------------+---------------------------------
Dr. Paul Kinsler
Department of Physics (QOLS)
Imperial College (ph) +44-20-75947520
Prince Consort Road (fax) +44-20-75947714
London SW7 2BW Dr.Paul...@physics.org
United Kingdom http://www.lsr.ph.ic.ac.uk/~kinsle/
Charles Francis
<c.h.th...@pgen.net> writes:
>Charles Francis <cha...@clef.demon.co.uk> wrote in message
>news:a2nssm$bke$1...@sue.its.caltech.edu...
>> nicolaa...@pandora.be writes
>> >Now suppose my light only emits one photon
>> >1. Will this photon propagate in a sphere ?
>> No. The photon wave function will propagate in a
>> sphere, but all that really does is tell us that the
>> photon is equally likely to be found in any direction,
>> not that the photon travels in more than one direction.
>> The photon will only be detected at one point.
>How thoroughly has this been tested?
It has been tested as thoroughly as it is physically possible to test. I
don't know all the experimental details of photon detection, but I do
know that single photons can be detected under carefully controlled
conditions. For example a Young's slits experiment can be done with very
low intensity light so that photons come through one by one. Then the
interference pattern on a photographic plate can be observed to build
up, point by point, each photon marking a single point on the plate, but
as more photons come through the statistical distribution of points
builds up the interference pattern.
Likewise using a Geiger counter you may only know that a photon (in the
form of gamma radiation) has entered the tube, not precisely where in
the tube it is. But you always know that a single photon has caused a
sudden rush of ionisation by the fact that the counter produces discrete
clicks at random intervals.
>It sounds plausible if the "photon"
>concerned is produced by a nuclear event, in which case it may well be
>produced as a very narrow beam, but what if it is ordinary visible light, or
>radio frequency?
It has nothing to do with the width of the beam. The problem with all of
these examples is that they generally occur as one photon among millions
of billions of photons. Certainly it is easier to detect a high energy
photon, but we are quite capable of detecting single photons of visible
light in a darkened room. In all cases the width of the beam only
describes the statistical distribution of photons, each photon is
detected in as small a region as your apparatus allows you to resolve.
>> >4. If we place a 1000 by 1000 grid of CCD's
>> >will we always detect this one photon ?
>I strongly suspect that you could sometimes detect more than one. I don't
>know so much about how CCD's work but have studied some experiments
>involving very low intensities of visible light. Some photomultipliers seem
>to require the addition of local electromagnetic noise in order to produce a
>"detection". The visible light will be spread out over the face of the
>photocathode, arriving over a finite period of time (20 ns or so, in the
>case in point). Whether or not it is detected, and when, depends on local
>random factors. Clearly there is a possibility that if two detectors were
>placed there instead of one, both could sometimes fire.
The problems have to do with experimental conditions. It is more or less
impossible to stop a Geiger counter or other detector from firing due to
background radiation. You then have to separate out the meaningful
events from the not meaningful. If two detectors fire, only one of them
has detected the photon, and the other is a random background event. But
really I think you have to devote your life to doing these kinds of
experiment to get the experiments good enough and the analysis good
enough to produce meaningful results.
Regards
--
Charles Francis
Charles Francis
> Some poor uncited soul wrote:
>> In a medium, the photons are absorbed and re-emitted by the atoms, which
>> cause time delays which makes the light appear to travel slower than c.
>> But the photons still move at the same speed c.
>Might I ask what evidence there is for this? All we can actually measure is
>light going in and light coming out of a block of glass -- and we can't even
>really "measure" that! We can only try and deduce things from the angle of
>refraction and, perhaps, from interference effects if we split a beam and
>send part through the glass, part not. What evidence is that that within
>the glass the light is proceeding in little jumps, from electron to electron
>or whatever?
We can calculate such behaviours of matter from fundamentals, just as
truly as Newton could calculate the arc of a cannon ball that was not
observed, (actually more truly as Newton may not have had the correct
figures for air resistance)
>Qunatum theory recognises that light is a wave as well as a particle. Why
>is this apparently ignored?
Principally because wave particle duality is not strictly quantum theory
but the interpretation of quantum theory, and it is an interpretation
which few who study such issues accept these days. There are other,
better, ways to interpret quantum theory, but most physicists try to
avoid interpretational issues altogether and stay with the maths.
Also because this explanation of photons going from electron to electron
is very much Feynman's take on qed, seems natural enough in his path
integral approach., and it has been popularised in Feynman's book on
qed. I happen to go along with Feynman pretty closely, and can think of
no one better qualified than Feynman or Dirac to interpret QED, since
they are the two most important figures in its development. Both
described particle interpretations, but both also acknowledged that
there were unresolved problems in interpretation.
It should be said that the mainstream view these days is not of
particles, but of fields, and that if you adopt such a view the
appearance of particles is only some sort of kink in the field. In which
case as the field is spread across the glass you may find it natural
that the kink moves more slowly across it.
I have not done either calculation (from particles or from fields), nor
do I know that anyone has. But I would be willing to lay odds that both
calculations will give the same result, and again that is a reason why
so many physicists say one should ignore the interpretation and
concentrate on the calculation.
Regards
--
Charles Francis
Charles Francis
eric alan forgy <fo...@students.uiuc.edu> writes
>On Thu, 24 Jan 2002, John Baez wrote to Ed Fredkin:
>> I remember getting into an argument over
>> whether the world was fundamentally discrete. You may be pleased
>> to know that now I think it might be.
>Is there something in particular recently that has made you take this idea
>more seriously than perhaps you might have in the past that "the world was
>fundamentally discrete"?
Since I have been in the habit of having guessing games with Toby
Bartels about what John thinks, I try and maintain the tradition by
suggesting that John's view seems to have shifted somewhat during
numerous discussions on s.p.r and sci.physics. It is not so much that
anyone has produced an argument that has convinced John, but that the
more one thinks about these fundamental issues during such arguments the
more one seems to convince oneself of the necessity of it.
>I don't know many recognized physicists (Sorkin is a nice exception) that
>openly admit to this suspicion, but I bet that many of them do secretly
>think the world may be fundamentally discrete.
>The trouble with having a serious discussion about this topic is that the
>idea is so vague. What does it really mean for the world to be
>fundamentally discrete? I don't think anyone can really make this idea
>concrete mathematically (not yet anyway). At least not in a universally
>acceptable manner.
The critical thing is the last phrase, to make it concrete
mathematically in a universally acceptable manner. I am quite sure that
once a genuinely rigorous mathematical construction has been produced
and accepted it will be possible to describe it in intuitive terms quite
easily. In fact in large degree we already can, in terms of particle
interactions.
The trick is to show that a model consisting of particle interactions
leads to the actual laws, of qm and gtr, which we observe. Clearly such
a model can only be described statistically, and the manifold can be a
statistical prediction of position, as distinct from some kind of
continuous material thing. But to make this idea concrete we have to
show both that the correct statistical laws exhibit the wave behaviour
of qm, and that they lead to the field equation of gtr.
For some time I feel I have been able to produce large parts of the
demonstration of these properties, certainly enough to convince myself
of the correctness of the model. But I have now come to the conclusion
that to produce parts of it is not sufficient. Parts of a demonstration
always leave some unexplained issue on which the reader gets stuck, and
most unsatisfactorily they leave one trying to guess at how the bits fit
together (with consequent conceptual mistakes). I am working on a far
more comprehensive rewrite while continuing to improve standards of
mathematical rigour and the elimination of mistakes.
Regards
- --
Charles Francis
John Baez
a totally discrete model is basically an Automaton
(as is defined in computer science).
That means that a discrete model can be directly and exactly implemented
as a process by programming it on a computer.
What is meant by "programming it on a computer" ?
In particular, does "programming" refer to classical information theory,
or
does "programmming? refer to quanum computation ?
Unless I misunderstand the distinction,
quantum computation and quantum information follow rules
that are not necessarily those of classical computation and
classical Shannon information theory.
For instance:
In quantum information theory, Cerf and Adami have written a paper at
http://xxx.lanl.gov/abs/quant-ph/9512022
that describes virtual qubit-anti-qubit pairs
(they call them ebit-anti-ebit pairs) that are related to
negative conditional entropies for quantum entangled systems
and are similar to fermion particle-antiparticle pairs of particle physics.
Therefore, it seems to me that quantum computation has promise as
a realistic way to formulate particle physics.
For another instance,
consider the use of computational systems to describe
evolution (of physical systems, and perhaps more ambitiously
of biological and social systems) in terms of game theory.
The abstract of a paper by Iqbal and Toor at
http://xxx.lanl.gov/abs/quant-ph/0104091
states:
"... We consider a slightly modified version of the Rock-Scissors-Paper
(RSP) game from the point of view of evolutionary stability.
In its classical version the game has a mixed Nash equilibrium (NE)
not stable against mutants.
We find a quantized version of the RSP game
for which the classical mixed NE becomes stable as well. ...".
In light of such things, and in light of the recent paper of
Vandersypen, Steffen, Breyta, Yannoni, Sherwood, and Chuang at
http://xxx.lanl.gov/abs/quant-ph/0112176
entitled
Experimental realization of Shor's quantum factoring algorithm
using nuclear magnetic resonance
in which they "... report an implementation
of the simplest instance of Shor's algorithm:
factorization of N=15 (whose prime factors are 3 and 5). ...",
My present opinion is that quantum computation and game theory
will prove to be necessary and sufficient
for the construction of realistic Finite Nature models
of particle physics (and higher-level things).
However, quantum computation and game theory are just beginning
to be seriously studied, and it is by no means certain that
my opinion will be borne out as correct (although the process of
finding out what is correct will doubtless be fun).
Tony 28 Jan 2002
Tony Smith
Ed Fredkin said: "... Finite Nature ...
(as is defined in computer science).
That means that a discrete model can be directly and exactly implemented
as a process by programming it on a computer.
Tony 28 Jan 2002
[Moderator's note: if any of you saw a screwed-up version of
this post that looked like it was from me, sorry. I hate it
when that happens! - jb]
c.h.thompson
Charles Francis <cha...@clef.demon.co.uk> wrote
> <c.h.th...@pgen.net> writes:
> >> The photon will only be detected at one point.
> >How thoroughly has this been tested?
> It has been tested as thoroughly as it is physically possible to test. I
> don't know all the experimental details.
Might I suggest that you read, say:
Grangier, P, G Roger and A Aspect, "Experimental Evidence for a photon
anticorrelation effect on a beam splitter: a new light on single-photon
interferences", Europhysics Letters 1, 173-179(1986).
This is relevant to "photon" splitting, and can be explained alternatively
as in:
Marshall, T W and Santos, E, Europhysics Letters, 3, 293-6 (1987).
Have you read:
Tonomura, Akita, J Endo, T Matsuda and T Kawasaki, "Demonstration of
single-electron buildup of an interference pattern", American Journal of
Physics 57, 117 (1989)?
There was some discussion of this in sci.physics a year or so ago.
> Then the
> interference pattern on a photographic plate can be observed
> to build up, point by point, each photon marking a single point
> on the plate, but as more photons come through the statistical
> distribution of points builds up the interference pattern.
Which experiment do you have in mind?
> Likewise using a Geiger counter you may only know that a photon (in the
> form of gamma radiation) has entered the tube, not precisely where in
> the tube it is. But you always know that a single photon has caused a
> sudden rush of ionisation by the fact that the counter produces discrete
> clicks at random intervals.
Wouldn't the instrument make the same click if more than one had arrived?
> ... I think you have to devote your life to doing these kinds of
> experiment to get the experiments good enough and the analysis
> good enough to produce meaningful results.
An instrument click cannot tell you what is really happening, however many
experiments you do. The fact that convinced me that the "quantisation" is
(for low-intensity visible light at least, when measured with
photodetectors) almost entirely produced by the instrument is the Bell test
experiments. You have a choice, according to my analysis: either admit that
the photon can be split or accept that the quantum world allows non-local
effects.
Re CCD's, these operate on a different principle. Is it not the case,
though, that they require a long integration period? Accepted theory may
say that this is because they have to wait for the individual photon, but
might they not be simply accumulating sufficient wave information?
Urs Schreiber
news:8c7d34cb.02012...@posting.google.com...
[...]
> I can't remember the topic of your thesis but, if you want, tell me
> when it is about to be completed and I will see if I can find anything
> that might be relevant from the worlds of stochastic or noncommutative
> differential geometry.
I know a little about the basic ideas of noncommutative differential
geometry. But what exactly is it one does in stochastic differential
geometry?
Toby Bartels
>I think that you both got the trick wrong: C sends out a mixed state, say
>a spin up and down. It will remain so until A distorts the state say by
>measuring it. Then B measures it: If B finds a down (resp. up) state, B
>will know that A found an up (resp. down) state.
I accept this, but we all agree that no information is transmitted this way.
OTOH:
>If the scheme is changed so that A can force the mixed state measured to
>be up (resp. down), then that information is "teleported" to B which will
>measure down (resp. up), thus constituting the transport of information
>that does not rely on transportation of any physical quantity. It relies
>on the fact that the QM field is spread all over the universe
>simultaneously like wave.
How can A *force* the mixed state measured to be up?
In the usual EPR thing, A can make changes in their results
by changing *what*variable*they*measure*, but you're not doing that.
If A and B agree ahead of time on what variable to measure,
then A can't force a certain result but is stuck with the laws of chance.
(The alternative is that the mixed state happened to be
an eigenstate of this particular variable, being mixed in other ways,
but now you really can say that the information already existed at C.)
If I'm completely misinterpreting what your idea is about,
then you should probably give more detail on it ^_^.
>I still think that one should speak about the speed of "information"
>without explicitly defining which physical quantities it involves.
I assume that there was supposed to be "not" between "should" and "speak".
-- Toby
to...@math.ucr.edu
Maury Markowitz
> >faster than light. - jb]
>
> be up (resp. down), then that information is "teleported" to B which will
> measure down (reps. up)
QM wavefunction collapse is largely beyond the scope of the rest of the
thread, so it may be a trifle unfair to drop it in now. However there are
points that need to be considered:
1) consider the "normal" case with two non-entangled particles a and b
travelling to people at A and B. A only knows that they got particle a, and
that's that. This doesn't tell B anything at all, and the reception of b at
B doesn't either. No information has travelled faster than the speed of
light, would you agree with that?
Now of course A and B could agree beforehand to only send a's and b's, in
pairs. Then the reception of a b by B would indeed know that A got an a.
However this information travelled earlier at the speed of A and B, likely
at or far slower than light.
2) in the quantum case we have two spin-entangled particles a and b with
possible states u and d traveling to A and B.
Consider the reception of a by a person at A. This person measures a to be
in u, and thus B will measure b to be in p. Well, so? What if instead B
decides to measure the X of particle b, which destroys the u/p state? Well
in that case it's certain that B knows nothing of a. In fact, in order to
make any sort of quantum measure the experiment has to not only be agreed on
in advance, but to a degree that even the normal case doesn't - you might
have to send only at specific times so the axis of the planets are aligned
for instance. In these sorts of situations QM doesn't help a bit.
There are some ways to use this information even in these cases. SciAm had
a nice illustration of these sorts of setups last year some time, although I
believe it may have been in the context of quantum crypto. In all of these
cases the net information flow happens at c, the internal QM states are
forever beyond us and cannot be directly used for information. At least no
one's figured out a way yet, and it seems unlikely they will.
Maury
Brian J Flanagan
> Also, don't worry if wizards might seem a bit irratible because, after
> all, they are only wizards and not patient gods.
Also, please don't get overly invested in the the experts. Look up the
history of group theory for a nice cautionary tale.
nicolaa...@pandora.be
> >this photon ?
>
> As far away as you can place the apparatus.
> We can detect photons which have propagated
> from within a few seconds of the big bang.
How do you know that?
My interpretation of the Big Bang is that photons
originated from within the first 5 sceonds of the
big bang should be in the outer range/layer of the
Universe.
Those photons will still propagate outwards.
And because we are more or less in the centre
of the Universe we will not detect those.
The same problem I have with the interpretation
of the Cosmic Microwave Backgroud Radiation.
This represents a temperature of 2.7 degr K.
My understanding is also that suppose we would
measure the CMB Radiation 300000 years after the
Big Bang than we would have measured a different
radiation spectrum, representing a different
temperature.
Suppose we measure light from a protogalaxy with Z=5.
That light from that protogalaxy was emitted shortly
after the Big Bang suppose 300000 years.
When we measure the black body radiation from that region
why don't we measure a much higher temperature ?
(To claim that the whole explanation lies in space
expansion seems to easy to me)
Along that same line for Z=3 we should measure a
higher temperature, but lower as for Z=5
The true issue lies in the statement:
"However the black body radiation comes to us directly
from the decoupling era"
See The Big bang by Joseph Silk page 165.
How sure are we that the radiation/photons we measure
now are then created ?
Why not earlier ?
Why not later ?
The next sentence reads:
"The background radiation has a redshift far greater than
that of any known galaxy or quasar, probably as large as 1000"
Figure 8.3 does not seen to be in support of the statement:
> We can detect photons which have propagated
> from within a few seconds of the big bang.
The scale of that figure is a true issue.
Nick
c.h.thompson
zirkus <zir...@hotmail.com> wrote in message
news:8c7d34cb.02012...@posting.google.com...
>
> According to the following paper, it is not yet clear how to interpret
> the physical meaning of Bell's probabilistic assumptions, and thus
> local realism in QT might still be possible:
>
> http://arxiv.org/abs/quant-ph/0006016
Perhaps more to the point, there are known "loopholes" in all Bell test
experiments to date, so that it cannot be said that the experimental
evidence supports QM any better than "local hidden variable" theories.
The article that zirkus quotes (http://arxiv.org/abs/quant-ph/0006016)
opens:
--------
Last years became more evident that Bell's arguments (which imply Bell's
inequality [1] and its generalizations [2], [3]) are closely related to
foundations
of probability theory, see [4]- [7]. Despite the general viewpoint [1]-[3]
that
experimental violations of Bell's inequality [2] imply the impossibility to
use
the local realism in quantum theory, there are many publications [4]-[7] in
that it was pointed out that the derivation of Bell's inequality is based on
rather delicate probabilistic assumptions. If one of these assumptions is
not
justified in the probability description of the EPR experiment, [8],
[1]-[3],
then there would be no Bell's inequality at all (or it should be modified
[9]).
Experiments of Aspect et al. [2] may be interpreted not only as arguments
against the local realism, but also as experiments which examine the use
of Bell's probabilistic assumptions. Of course, Bell's probabilistic assump-
tions are not just mathematical postulates. They must have some physical
meaning. To find this meaning is an incredibly hard problem (at least on
the present level of experiments with quantum systems). Therefore careful
analysis of Bell's assumptions must be performed.
---------
The whole paper is theoretical, with no discussion of the experiments that I
could find -- just a listing of refs.
Perhaps a consideration of the models that become possible once you allow
for the loopholes would be worthwhile, rather than tackling this "incredibly
hard problem". Several papers have been published over the years claiming
to find fault with Bell's reasoning, but, when it comes to it, his only
fault may have been to fail to think at the outset about what would happen
in less than perfect conditions. Clauser, Horne, Shimony and Holt filled in
this gap in 1969, with their modified tests and a comprehensive study of the
subject in 1974. (Clauser, J F and Horne, M A, "Experimental consequences
of objective local theories", Physical Review D, 10, 526-35 (1974))
Actual experiments have always failed, in one way or another, to meet
Clauser and Hornes criteria as detailed in the 1974 paper. For discussion
of some of the consequences, see http://arxiv.org/abs/quant-ph/9611037
or http://arxiv.org/abs/quant-ph/9903066
It is possible that Bell's use of probabilities is NOT a problem and his
tests are perfectly valid. It is the experiments that are not valid
applications of them.
[I am sending a copy of this posting to the author of quant-ph/0006016 for
information]
Hans Aberg
<c.h.th...@pgen.net> wrote:
>> The wave behavior is a fundamental fact of the QM behavior of photons as
>> well. One experiment is to send the photons one-by-one onto a slit and
>> observe that the interference pattern remains, which would not happen if
>> photons were only particles.
>
>Can you give me a ref for the relevant experiment?
I only know what's in basic books, like the one by Cohen-Tannoudji et al
perhaps.
> The only ones that I
>have studied have not been convincing. There have always been other
>possible intrepretations...
It seems that every aspect of the basic physics laws are tested, re-tested
and re-interpreted. This is of course laudable, but the question is
whether one can get something out that gives better predictions than the
standard model.
So I leave it there, hoping that some experts on such experiments will continue.
>> One can do similar things with electrons, so
>> electrons are QM waves as well.
>
>Yes, but that's another story!
Not really, there are additional intrinsic properties that set them apart,
but as for the QM particle-wave behavior, electrons and photons are the
same; the same formulas apply.
John Baez
eric alan forgy <fo...@students.uiuc.edu> wrote:
>John Baez wrote:
>> I remember getting into an argument over
>> whether the world was fundamentally discrete. You may be pleased
>> to know that now I think it might be. But not a cellular automaton!
>OK, I'm sure you were aware of the subsequent fallout when you wrote this.
>:)
Yup. It was sufficiently vague and grandiose that everyone was bound
to have something to say about it. I normally avoid such remarks, to
set a good example. But they *do* keep certain people interested -
the poor folks whose eyes slowly close with weary incomprehension as
they read endless threads about S^7, Spin(7), Spin(8) and G2.
Good physics requires a careful mix of hot and cold. There's something
fundamentally bombastic about trying to figure out the laws of the
universe. Thinking about this quest in broad terms, one gets all heated
up and excited. At this point it's necessary to pour on some cold water:
rigorous mathematics and careful checking of theory against experiment.
Without this cold water, the enterprise devolves into pompous chatter
or even madness. But the heat is necessary too: without some grand
goals, it's hard to get up the energy to do good work! We must sail
between the Scylla of uninspired hack work and the Charybdis of crackpottery.
>Is there something in particular recently that has made you take this idea
>more seriously than perhaps you might have in the past that "the world was
>fundamentally discrete"? Does it have to do with smooth manifolds being
>used to set up the spin foam model, but after you have constructed the
>spin foam states, then the fact that they were built from smooth manifolds
>becomes immaterial and you could just as well have STARTED with a spin
>foam?
That's pretty much it, and it's not terribly recent. At first folks
only knew how to set up loop quantum gravity after choosing a smooth
manifold to use as "space". The combinatorial/algebraic flavor of
spin networks suggested that it might be possible to discard the
manifold, but only with the development of spin foam models, in the
summer of 1997, did people find a nice way to do this. Since then
I have alternated between optimism and pessimism over whether spin
foam models will really "work" - for example, at the bare minimum,
give some theory which reduces to general relativity plus matter in
the large-scale limit. Right now I'm feeling quite optimistic because
Dan Christensen and I have just found a spin foam model of Riemannian
quantum gravity that seems to avoid some of the problems which afflict
the DePietri-Freidel-Krasnov-Rovelli and Perez-Rovelli models. But
regardless of these short-run ups and downs, in the long run it seems
interesting to try to build spacetime out of some sort of "discrete",
"combinatorial" or "algebraic" structure at the Planck scale.
>I don't know many recognized physicists (Sorkin is a nice exception) that
>openly admit to this suspicion, but I bet that many of them do secretly
>think the world may be fundamentally discrete.
I don't know. Lots of physicists I talk to hold ideas along these lines
and are not afraid to publish papers about them. These are mainly
people working on quantum gravity. A famous example is 't Hooft.
He won the Nobel prize for his work on gauge theory, but more recently
he's been playing with discrete models of quantum gravity - and one
of the first people to think about the holographic hypothesis.
>The trouble with having a serious discussion about this topic is that the
>idea is so vague. What does it really mean for the world to be
>fundamentally discrete?
If I wanted to get serious, I probably wouldn't bother trying to define
"fundamentally discrete". Different people mean different things by it,
and I wouldn't want to get in an argument with all of them. Instead, I'd
prefer to talk about specific models that people have studied, their
properties, and their successes and failures in fitting the experimental
data. Cellular automata, causal sets, dynamical triangulations, matrix
models, spin foams, the Regge calculus - compare them all! I wasn't
trying to get into that, though - I was just chatting.
Patrick Van Esch
>
> An instrument click cannot tell you what is really happening, however many
> experiments you do. The fact that convinced me that the "quantisation" is
> (for low-intensity visible light at least, when measured with
> photodetectors) almost entirely produced by the instrument is the Bell test
> experiments.
The problem with all optical detection (optical is wavelengths about
around visible or longer) is that the photon energy is quite low, and
that there are no high-efficiency single photon detectors available.
The thing that comes closest is of course the photomultiplier, but efficiencies
are seldom higher than, say, 30%.
I know that this is the main reason of the "loopholes" in the statistics
of the Aspect-like experiments (and I know that you are one of the people
insisting on those loopholes).
The reason for this is that a certain event has to happen at the photocathode,
namely the emission of an electron (the photo-electric effect), and then
you have to capture that single electron, and send it into an E-field that
will accelerate it onto the first dynode. For low energy photons (pardon
me to talk a language you are contesting, but for most experimentalists
this is so common that it is hard to do otherwise) there is a lot that
can go wrong, some reasons are fundamental and others are of technological
origin, which result in the fact that you only get a click once every
three-five photons.
Integrating detectors such as CCD cameras can be much more "precise".
In fact, on a large sample of photons, there is an average ionisation
that occurs, and by dividing the total ionisation by the average ionisation
per photon, you get the number of photons (with some statistical error that
is completely mastered). It is EVIDENT that an integrating detector
such as a CCD cannot distinguish between classical radiation and "photonic
radiation": when "counting photons with a CCD" one ASSUMES that it has
been irradiated by a photon beam of single energy (and OF COURSE one
assumes that photons behave, well, as photons).
I could note that in our work of develloping neutron detectors we try
to avoid using integrating detectors exactly for that reason: as several high
energy photons, integrated, correspond approximately to the signal of
one single neutron in an integrating detector, we would not be able
to distinguish between the impact of gamma-photons and neutrons.
So we prefer to see the beasties one by one, even if this complicates
a lot the electronics.
In a single-particle detector, such as a proportional counter, things
happen a bit as in a photomultiplier. However, at higher photon energies,
the efficiencies of detection can be up to 99%! The pulse generated by
the detector has an amplitude proportional to the photon energy.
The difficulty is that we're now talking about X-rays. Doing optics
with X-rays is a lot harder (although not impossible).
I don't know if there are parametric down conversions known in the X-ray
regime so that one could do Aspect like experiments in that domain (I think
some people are working on that, remembering a talk, but I can't recall the
details).
You have a choice, according to my analysis: either admit that
> the photon can be split or accept that the quantum world allows non-local
> effects.
It is difficult to see how the "photon can be split" with non-integrating
detectors, as you can see them individually, with high efficiency, in
proportional counters...
>
> Re CCD's, these operate on a different principle. Is it not the case,
> though, that they require a long integration period? Accepted theory may
> say that this is because they have to wait for the individual photon, but
> might they not be simply accumulating sufficient wave information?
As I said, an integrating detector does by definition not make a difference
between "classical radiation" and "photon packetted radiation".
In fact, one uses these detectors in optics, under the hypothesis that
photons exist, because this is such a basic thing in radiation detection
that it is not considered anymore as part of the experiment, but part of
the technology, in about the same way that newtonian mechanics is considered
part of the technology and not part of the experiment when, say, measuring
vibrations.
In the X-ray domain, it is possible to use CCDs as "individual counters"
when going to very low intensities. However, you're at the limits of what
is possible concerning noise. I don't really know, but I would be surprised
that one can do the same with optical photons. Guess they're working at
liquid-helium temperature then !
cheers,
Patrick.
Hans Aberg
>>If the scheme is changed so that A can force the mixed state measured to
>>be up (resp. down), then that information is "teleported" to B which will
>>measure down (resp. up), thus constituting the transport of information
>>that does not rely on transportation of any physical quantity. It relies
>>on the fact that the QM field is spread all over the universe
>>simultaneously like wave.
>
>How can A *force* the mixed state measured to be up?
I do not know in the current set-up, that is the cath point, as you point out.
But there are quantum teleportation experiments performed, which I think
build on this set-up. This was discussed in this group, see for example:
From: greg...@netspace.zebra.net.au (Greg Egan)
Newsgroups: sci.physics.research
Subject: Re: Faster than light
Date: 7 Aug 2000 15:14:24 GMT
Message-ID: <gregegan-070...@dialup-m1-15.perth.netspace.net.au>
I recall I looked it up on the Internet, and I think one can even teleport
particles, not just information from A to B, but a particle of some sort.
But I found no URL to that on my computer.
>In the usual EPR thing, A can make changes in their results
>by changing *what*variable*they*measure*, but you're not doing that.
>If A and B agree ahead of time on what variable to measure,
>then A can't force a certain result but is stuck with the laws of chance.
>From the brief look I had at EPR in the FAQ, I got the impression that it
was different from the teleportation idea.
>>I still think that one should speak about the speed of "information"
>>without explicitly defining which physical quantities it involves.
>
>I assume that there was supposed to be "not" between "should" and "speak".
Sorry, a typo: I favour that one defines the physical quantities in the
physical statements. So even though that one perceives that knowledge must
be stored and transported via physically measurable quantities, I think
those should be explicitly defined at some point in the thought process.
Charles Francis
<nicolaa...@pandora.be> writes
[unnecessary quoted text deleted by s.p.r. moderator]
>The target of my experiments is to do the
>true two slit experiments with single photons
>but before I do that I first do a set of
>3 pre experiments to learn more.
>
>First I do an experiment without any screen
>with slits.
>I have only a single photon generator
>and CCD detectors of 100*100 CCD's
>I generate 100 single photons
>How many hits will I get?
>The place is not important.
>
>I expect 100. (or 99)
>Not more.
>(otherwise this are not single photons.)
My guess would be noticeably less, but I am a theorist not an
experimentalist. It depends on how directional the beam is from a photon
generator, and it depends on how much space is lost in the grid of CCD's
>Second I place a screen with two slits
>L and R between the generator and the CCD detectors
>However only the L slit is open
>The R slit is closed.
>I generate 100 single photons
>How many hits will I get?
>Accordingly to your reply much less then 100
>Assume we get 10 plus or min 1
Again, The slits need to be very narrow. I suspect, depending on beam
width from the photon generator, I think the actual number will be much
less. But that is not really significant to the question.
>Third exactly the same as above but now:
>only the R slit is open
>The L slit is closed.
>I generate 100 single photons
>How many hits will I get?
About the same as before.
>Accordingly to your reply much less then 100
>Assume we get 10 plus or min 1
>
>Fourth exactly the same as above but now:
> L and R slit are open.
>I generate 100 single photons
>How many hits will I get?
>20 plus or min 1 ?
>10 plus or min 1 ?
According to your above figures, I would think 20 plus or minus 2. But
this figure
>(We first have to agree about the numbers
>the position were comes later)
I don't see why you think the numbers are that important. It is
generally the distribution of hits on the grid that is considered
important.
>Did you ever do the experiment in this way
>considering those 4 possibilities ?
No. I would hope an experimentalist may be able to give you a better
answer.
Regards
--
Charles Francis
Charles Francis
<c.h.th...@pgen.net> writes
>You have a choice, according to my analysis: either admit that
>the photon can be split or accept that the quantum world allows non-local
>effects.
I have my own analysis and it does show that the mesoscopic idea of
locality does break down in the quantum world. However that is a
slightly different thing from saying the quantum world is non-local. It
is just that a different, perhaps more obscure, definition of locality
is appropriate. Locality does hold, in the sense that interaction
between two particles can only take place in the intersection of the
support of the particle wave functions.
Regards
--
Charles Francis
Ed Fredkin
> Ed Fredkin said: "... Finite Nature ...
> a totally discrete model is basically an Automaton
> (as is defined in computer science).
> That means that a discrete model can be directly and exactly implemented
> as a process by programming it on a computer.
> This is true for all totally discrete models. ...".
>
> What is meant by "programming it on a computer" ?
>
> In particular, does "programming" refer to classical information theory,
> or
Finite Nature is the assumption that all quantities in physics are
finite and discrete. There are no infinities, infinitesimals and no
truly random numbers. This model is much like what goes on in the
innards of an ordinary computer processing ordinary bits. The process
is strictly deterministic.
We have learned that such systems, even though extraordinarily simple,
are capable of exhibiting the most complex behavior imaginable.
"Digital Mechanics" (DM) refers to such systems (computers that are
Cellular Automatons) that model physics. The concept is that DM would
underlie QM. In essence, QM would not be a description of reality,
but rather DM would relegate QM to being a mathematical shortcut to
predict the probabilities of various outcomes of a deterministic
digital process, when we only have partial knowledge of the state of
the system.
>
> Unless I misunderstand the distinction,
> quantum computation and quantum information follow rules
> that are not necessarily those of classical computation and
> classical Shannon information theory.
That is true; consequently quantum computation has nothing to do with
finite nature or DM.
>
> For instance:
> In quantum information theory, Cerf and Adami have written a paper at
> http://xxx.lanl.gov/abs/quant-ph/9512022
> that describes virtual qubit-anti-qubit pairs
> (they call them ebit-anti-ebit pairs) that are related to
> negative conditional entropies for quantum entangled systems
> and are similar to fermion particle-antiparticle pairs of particle physics.
>
> Therefore, it seems to me that quantum computation has promise as
> a realistic way to formulate particle physics.
I believe that it may prove an interesting tool to model some QM
systems, but I doubt that it could ever be a useful or realistic way
to formulate particle physics.
>
> For another instance,
> consider the use of computational systems to describe
> evolution (of physical systems, and perhaps more ambitiously
> of biological and social systems) in terms of game theory.
> The abstract of a paper by Iqbal and Toor at
> http://xxx.lanl.gov/abs/quant-ph/0104091
> states:
> "... We consider a slightly modified version of the Rock-Scissors-Paper
> (RSP) game from the point of view of evolutionary stability.
> In its classical version the game has a mixed Nash equilibrium (NE)
> not stable against mutants.
> We find a quantized version of the RSP game
> for which the classical mixed NE becomes stable as well. ...".
>
> In light of such things, and in light of the recent paper of
> Vandersypen, Steffen, Breyta, Yannoni, Sherwood, and Chuang at
> http://xxx.lanl.gov/abs/quant-ph/0112176
> entitled
> Experimental realization of Shor's quantum factoring algorithm
> using nuclear magnetic resonance
> in which they "... report an implementation
> of the simplest instance of Shor's algorithm:
> factorization of N=15 (whose prime factors are 3 and 5). ...",
So far, no practical real problem, where a quantum computer is better
than an ordinary computer, has so far been proposed. While Shor's
insight, relating quantum mechanics to factoring and integer log was a
fantastic intellectual accomplishment, I doubt that there will ever be
a practical quantum computer doing something that can't be done better
on an ordinary computer. If, by some miracle, a quantum computer
could do factoring quickly that would be an amazing and wonderful
result. However, as a consequence there would no longer be a need to
do factoring quickly. (If factoring is easy, RSA is useless.)
Serious users of encryption are already moving away from the use of
RSA. The sorting examples given are all straw men, in that computers
never have to work on the kinds of problems that are touted as grist
for a quantum computer. Finally, quantum computers are analog
computers. The arguments in their favor remind one of the arguments
that were in vogue in the 1950's in favor of analog computers over
digital computers. The principles that allow us to build digital
computers with billions of parts, some of which are performing
billions of operations per second, for months and years with no
errors, are simply absent from Quantum Computers.
>
> My present opinion is that quantum computation and game theory
> will prove to be necessary and sufficient
> for the construction of realistic Finite Nature models
> of particle physics (and higher-level things).
Basically this is the exact opposite of my opinion. It reflects a
basic misunderstanding. Finite nature is a totally deterministic
system that is meant to underlies QM. It has nothing to do with game
theory. It can't be modeled by QM; it is designed to model the
processes in microscopic physics where QM is used as a means of
predicting probabilities and outcomes of proposed experiments.
>
> However, quantum computation and game theory are just beginning
> to be seriously studied, and it is by no means certain that
> my opinion will be borne out as correct (although the process of
> finding out what is correct will doubtless be fun).
Yes, it's all fun. As I mentioned in some other thread, shortly after
Shor wrote his first paper on the subject I predicted that after much
work, someone would factor 35 into 5 x 7. I over estimated the
progress as after much work we are only up to 15=5 x 3.
>
> Tony 28 Jan 2002
Ed F 30 Jan 2002
Charles Francis
<remove.haberg-3...@du142-226.ppp.su-anst.tninet.se>, Hans
Aberg <remove...@matematik.su.se> writes
>Someone not cited by Francis wrote:
>>Someone else not cited by Francis wrote:
>>>I still think that one should speak about the speed of "information"
>>>without explicitly defining which physical quantities it involves.
>>I assume that there was supposed to be "not" between "should" and "speak".
Funny, I thought Hans meant what he said, and that it was indicative of
a truly abstract mode of thought which I believe is necessary for the
correct treatment of both qm and relativity.
>Sorry, a typo: I favour that one defines the physical quantities in the
>physical statements.
On the other hand I do agree that one must define physical quantities.
>So even though that one perceives that knowledge must
>be stored and transported via physically measurable quantities, I think
>those should be explicitly defined at some point in the thought process.
But while the physical quantities must be defined, their may be a huge
range of possible definitions, and the actual definitions chosen must
not be important. There are numerous examples of this in physics. As I
recall temperature is defined according to a number of different
measurement techniques for different temperature ranges.
More pertinently it does not matter to relativity whether one defines
distance using a rule or by the radar method, or indeed whether one
defines cosmological distances by red shift. The important thing is that
any definition used must give the same results as any other definition.
Likewise with the maximum speed of information. We know that in practice
light is a carrier of information, and we also know that, to the best of
our ability to determine it, light travels at the maximum speed of
information. But the equations of special relativity do not depend upon
this being true for light. If the photon had (or has) a very small mass
light would not travel at the speed of light (sic), i.e. c, but instead
c would be a limiting velocity for very high energy light, exactly as it
is for other particles. Since the limit is the same for all particles it
makes sense to talk of this limiting velocity, the maximum speed of
information without specifying any special particle.
Regards
--
Charles Francis
Brian J Flanagan
> The trick is to show that a model consisting of particle interactions
> leads to the actual laws, of qm and gtr, which we observe. Clearly such
This begs the question stated in EPR's paper, viz., can the QM
description of reality be considered complete?
Ed Fredkin
> Tony Smith <tsm...@innerx.net> wrote:
>
> > In a post to s.p.r. from Eric Alan Forgy on the thread
> > World is fundamentally discrete? (was: Why is a photon called stable?)
> > EAF said:
> > over whether the world was fundamentally discrete.
> > You may be pleased to know that now I think it might be.
> > EAF continued, asking:
>
> Nothing actually,
> for such a statement is inherently not falsifiable,
> and therefore not scientific.
>
Yes, discrete physics is not inherently falsifiable. But to say
"...and therefore not scientific." is simply wrong. Discrete physics
is obviously verifiable. Discrete matter (particles and atoms) is
verifiable. Discrete spin state is verifiable. Discreted electrical
charges is verifiable.
There are a number of experimental tests for different models of
discrete physics. You might take a look at Chapter 37 of Intro to
Digital Philosophy at www.digitalphilosophy.org
While these test could prove that physics is discrete, they cannot
disprove discreteness or prove that space or time are continuous. So,
if you want to make accurate statements here are some possibilities:
Question: "What would is mean for the world to be inherently
continuous?"
Answer: "Nothing really, for such a statement is inherently
non-verifiable and therefore not scientific."
> It is always possible that below a 'fundamental' discrete theory
> there lies an even more 'fundamental' deeper continuous one,
> and visa versa.
>
> Jan
And here I agree completely with your "...visa versa."
Ed F
J. J. Lodder
> nos...@de-ster.demon.nl (J. J. Lodder) wrote in message
news:<1f6n4e7.bu5...@de-ster.demon.nl>...
> > Tony Smith <tsm...@innerx.net> wrote:
> >
> > > In a post to s.p.r. from Eric Alan Forgy on the thread
> > > World is fundamentally discrete? (was: Why is a photon called stable?)
> > > EAF said:
> > > "...John Baez wrote to Ed Fredkin ... I remember getting into an argument
> > > over whether the world was fundamentally discrete.
> > > You may be pleased to know that now I think it might be.
> > > But not a cellular automaton! ...".
> > > EAF continued, asking:
> > > "... What does it really mean for the world to be fundamentally discrete?
> >
> > Nothing actually,
> > for such a statement is inherently not falsifiable,
> > and therefore not scientific.
> >
>
> Yes, discrete physics is not inherently falsifiable. But to say
> "...and therefore not scientific." is simply wrong. Discrete physics
> is obviously verifiable. Discrete matter (particles and atoms) is
> verifiable. Discrete spin state is verifiable. Discreted electrical
> charges is verifiable.
How would you go about that?
How could you -verify- that the distribution of spin values really is
delta(x + 1/2) + delta(x - 1/2),
and not two Gaussians with a width too small to resolve?
You may have a theory that tells you so;
you may verify that said theory agrees with experiment,
but you cannot verify that the distribution really -is- discrete,
or that this theory must be the final word.
Jan
John Forkosh
: J. J. Lodder wrote
: > It is always possible that below a 'fundamental' discrete theory
: > there lies an even more 'fundamental' deeper continuous one,
: > Jan
: And here I agree completely with your "...vice versa."
: Ed F
This exactly follows the reasoning more thoroughly discussed
by Leo Kadanoff in this month's (February 2002) Reference Frame
column of Physics Today. Especially see the righthand column
om page 10 (start with the bottom paragraph of the middle column),
and the first few paragraphs on page 11.
You can exactly transpose his statistical vs. deterministic
discussion to the discrete vs. continuous "regime".
Hans Aberg
<cha...@clef.demon.co.uk> wrote:
>In article
><remove.haberg-3...@du142-226.ppp.su-anst.tninet.se>, Hans
>Aberg <remove...@matematik.su.se> writes
>>Someone not cited by Francis wrote:
>>>Someone else not cited by Francis wrote:
I did not write those things. -- The quoting business is somewhat funny
here. :-)
[Moderator's note: these were gentle reminders from the moderator
that people should cite the author of text they are quoting. - jb]
>>I favour that one defines the physical quantities in the
>>physical statements.
>On the other hand I do agree that one must define physical quantities.
>>So even though that one perceives that knowledge must
>>be stored and transported via physically measurable quantities, I think
>>those should be explicitly defined at some point in the thought process.
>But while the physical quantities must be defined, their may be a huge
>range of possible definitions, and the actual definitions chosen must
>not be important. There are numerous examples of this in physics. As I
>recall temperature is defined according to a number of different
>measurement techniques for different temperature ranges.
I think there is a difference between physical quantities and measurement
techniques. If one is using different measurement techniques for measuring
the same physical quantity, I figure it would be prudent to demonstrate
that these different measurement techniques measure the same physical
quantity, otherwise one might end up with some physical embarrassments.
>More pertinently it does not matter to relativity whether one defines
>distance using a rule or by the radar method, or indeed whether one
>defines cosmological distances by red shift. The important thing is that
>any definition used must give the same results as any other definition.
I gather that the equivalence of such difference measurements build upon
some theory. Then one must be aware of the conditions of that theory, so
it can be tested at will.
>Likewise with the maximum speed of information.
The problem is that "information" has not been defined as a physical quantity.
> We know that in practice
>light is a carrier of information,
Not really: If information is transport of knowledge, then we know that
light can be used to transport information, but what information? -- If
the light is a Morse signal or a picture beautiful woman, then that is
quite different kinds of information, but what formulas in physics allows
us to identify what type of information?
> and we also know that, to the best of
>our ability to determine it, light travels at the maximum speed of
>information.
But we also know that a common way of transporting information is not via
light, but via electrons, like for example this email. It also depends on
exploiting QM fields, in transistors.
>But the equations of special relativity do not depend upon
>this being true for light. If the photon had (or has) a very small mass
>light would not travel at the speed of light (sic), i.e. c, but instead
>c would be a limiting velocity for very high energy light, exactly as it
>is for other particles. Since the limit is the same for all particles it
>makes sense to talk of this limiting velocity, the maximum speed of
>information without specifying any special particle.
I do not object to that one may consider theories where the photons might
have a small mass, even though I do not believe that is so (but it is of
course important to design experiments for testing this).
But I object to that one should be able to claim that information must be
transported by photons, or that jumping to the conclusion that just
because this is the limit speed of information transported by photons,
that must be the limit of other ways of transporting information (say via
QM teleporting).
It seems me both simpler and safer to say that if photons do have a small
mass, then c is the limit speed of high energy photons: This avoids all
references to a physically fuzzy concept like "information".
Charles Francis
writes
>Charles Francis <cha...@clef.demon.co.uk> wrote:
>> The photon will only be detected at one point.
>But only if we have a detector of photons that returns information
>about some localised region of space in which the photon was detected.
>
>I could enclose an excited atom with a spherical photodetector
>which cannot tell me anything about where on it a photon
>might land.
Of course I agree with you that a photon is only detected at a point up
to the spacial resolution of the apparatus used to detect it. I am
slightly surprised that you thought I was being so naive.
>If the atom were to de-excite and emit a photon,
>saying
>> The photon will only be detected at one point.
>is clearly wrong.
It is of course trivial that you cannot detect a point more accurately
than the resolution of your apparatus allows. One should not use the
term point in physics to mean the mathematical idealisation used in
Euclidean geometry, but only that which we can detect in practice to
experimental accuracy. If I say point, I mean point up to the resolution
of the apparatus. I should not need to be corrected for not saying "up
to the resolution of the apparatus" in every statement I make about
physics, both because it would make reading tedious in the extreme, and
because no one else in physics qualifies every word they write like
that.
>Further, the only sensible description of the
>photon in this case would use (or a the very least include) a set of
>spherical modes centered on the atom.
That is of course what one does in a statistical analysis. However in
statistics the set of possibilities is just that - a set of
possibilities, one of which becomes actuality. The full description of
photon behaviour requires a statistical analysis, so there must be a set
of possibilities, and this set is somehow described by a wave function
instead of an ordinary probability distribution. The why's and
wherefore's of that are generally considered unresolved, but it is a big
jump from there to give the wave function a physical existence and claim
a degree of actuality for the members of the set of possibilities of how
the photon will behave.
>Of course to be properly
>non-localizing, the photon detector would operate by absorbing
>the photon and "converting" it into some some non-local excitation
>(random e.g. a surface plasmon), and this excitation would then
>be measured by some other process.
And indeed I mentioned in another post on this topic that in a Geiger
counter, which works not unlike this, we can only know that a high
energy photon has entered the tube. Nonetheless the photon is localised
sufficiently to produce a distinct click.
>>>2. Will this photon propagate in a cone ?
>>>(ie it can be detected only in certain directions)
>> Of course no practical light source can have perfect spherical symmetry,
>> so really the wave function propagates conically rather than
>> spherically, but that does not affect the fact that the photon will be
>> detected at a point, not spread out over an area.
>This is both misleading and wrong. Wrong in the case I described
>above,
To be fair, I think if you found it misleading it is, as much as
anything because you choose to be misled. If you read more of my posts
you would have found that I do not consider particles to be points in
the context of a Euclidean, or even curved space background. Nonetheless
the concept of a point, as defined in Euclid to be that which has
neither depth nor breadth, is a perfectly legitimate mathematical idea.
In quantum mechanics we either have to lose point-like entities or we
have to lose the notion of a Euclidean (Riemannian or semi-Riemannian)
background structure. As Dirac says:
In the general case we cannot speak of an observable having
a value for a particular state, but we can .... speak of the probability
of its having a specified value for the state, meaning the probability
of this specified value being obtained when one makes a measurement of
the observable
In the instance of position this means that particles do not have well
defined position, but it does not say that particles are not point-like.
It is equally possible to interpret it as meaning that particles are
point-like, but that one should lose the background structure R^n
As for the resolution of these issues there is no universally accepted
solution. But it is certainly unfair to take an explanation which I have
clearly simplified for the benefit of those unused to such concepts, and
call it wrong on the grounds of much deeper, and unresolved issues
concerning the fundamental nature of space-time, and the sense in which
it may be possible to call something a point.
>and misleading in suggesting there is anything intrisically
>"pointlike" about photons.
Now that is wrong, and misleading. A clear example of point like
behaviour is the locality condition in quantum field theory, that the
commutator
[A(x),dA(y)]
is zero outside the light cone. It says that if the matter the photon
interacts with is confined at x, then the interaction of the photon is
confined at x too. And it says that if the interaction is not confined,
then we only have statistical information about location. And it
definitely says that the statistical information we have about the
photon's location matches the statistical information we have about the
matter the photon interacts with, as illustrated by the example of
detection you gave, and in my book that is point-like behaviour.
Regards
--
Charles Francis
Charles Francis
In article <GqpLM...@research.att.com>, Peter Shor
<sh...@research.att.com> writes:
>I believe there's a real disconnect as to what people mean by
>fundamentally discrete here. I think Ed Fredkin wants to throw away
>all of quantum mechanics, and propose that the world is a classical
>discrete cellular automaton. Whereas I think what John Baez means by
>fundamentally discrete much more closely resembles quantum mechanics
>over a finite-dimensional Hilbert space.
....
>What's probably closer to John Baez's belief (and if
>he's reading this, he can correct me) is that any chunk of the
>universe of volume V can be represented (modulo boundary effects) by a
>finite-dimensional Hilbert space.
I might say something a bit like this, but in the discussions I have
had with John Baez he has argued strongly against it, principally on the
ground of the well known theorem that there are no finite unitary
representations of the Lorentz group, and he has produced Newton's
argument that it may be able to infer the existence of the continuum
even if it is not possible to measure it directly. Surprising that,
because Newton's argument was thoroughly and conclusively refuted by
Leibniz.
Actually I would have to qualify what you say, because in Copenhagen
and orthodox interpretations Hilbert space does not directly describe
the structure of the universe (or a part, V, of it), but describes the
information we have about the universe (or that part). Since the
information we have comes from bounded experiments at finite resolution
it is immediate that we should use finite dimensional Hilbert space.
But that does not tell us whether the underlying structure is
fundamentally discrete or not. It does tell us that wave effects are
simply the consequence of taking linear combinations of states in
Hilbert space, and hence it does tell us that wave effects can be put
down to information theory, not fundamental structure (see e.g. Adami &
Cerf's paper on lanl, but really information theoretic interpretations
are just another way of saying what Mermin, Von Neumann and others have
already said).
If you interpret qm like this then you can conclude that since the waves
appear only in information theory, and indeed background R^n appears
only as part of information space not as ontology, then there may well
be a fundamentally discrete structure of particles, described very well
as a network in which each line represents proper time and each node
represents interaction. These would not directly be Feynman diagrams but
would be related to them through the superposition principle which comes
from Hilbert space. A Feynman diagram only tells us the topology of
interactions, not proper time between interactions, and hence would be
the superposition of all proper time diagrams with the same topology.
Regards
--
Charles Francis
Charles Francis
Flanagan <sen...@yahoo.com> writes:
>Charles Francis wrote:
Quantum mechanics is best understood as a "black box" theory of
measurement, providing statistical correlations between "fore" and
"after" states, without necessarily saying anything about what goes on
between. As such it is not a complete theory of matter, but is a theory
which would hold whatever form matter takes.
Really the whole issue of interpretation should have moved from
interpretation of non-relativistic quantum mechanics to the
interpretation of relativistic quantum field theory, but unfortunately
those who are interested in interpretation usually stick to the
non-relativistic theory, while those who are well versed in relativistic
quantum field theory are usually well trained not to think too carefully
about interpretational issues.
To answer your question, in my view, when relativistic quantum field
theory is interpreted as a statistical description of a theory of
particle interactions using an information theoretic interpretation of
quantum mechanics, then I consider it is complete in the sense of EPR -
or if not complete at least not lacking in the very thing we were trying
to study in the first place, which was Einstein's complaint.
c.h.thompson
Brian J Flanagan <sen...@yahoo.com> wrote in message
news:c4858886.02013...@posting.google.com...
> Charles Francis wrote:
It also begs the question of what ARE the "actual laws"! A very important
doubt centres around the notion of quantum engtanglement, for instance. It
would be much easier to formulate a "complete" alternative to QM if the
phenomenon did not happen. And if you look at the actual evidence, maybe it
doesn't!
For an authoratitive account of the present state of play re entanglement,
see Prof Laloe's article:
Franck Laloë, 'Do we really understand quantum mechanics? Strange
correlations, paradoxes and theorems', American Journal of Physics, 69(6)
pp 655-701, June 2001.
Although Laloe expresses the belief that QM is correct, he admits that there
are loopholes and says:
"A fair summary of the situation is that no one has been able to disprove
quantum mechanics".
"Not disproven" is a far cry from "proven"!
Brian J Flanagan
> Quantum mechanics is best understood as a "black box" theory of
> measurement, providing statistical correlations between "fore" and
> "after" states, without necessarily saying anything about what goes on
> between.
How is this picture best? In practice, physics puts together highly
successful models of mechanisms thought to account for what goes on
inside the black box.
>As such it is not a complete theory of matter, but is a theory
> which would hold whatever form matter takes.
This is reminiscent of the S=matrix program; perhaps you know
Cushing's book on this approach? He seems to think it's been
discarded.
> Really the whole issue of interpretation should have moved from
> interpretation of non-relativistic quantum mechanics to the
> interpretation of relativistic quantum field theory, but unfortunately
> those who are interested in interpretation usually stick to the
> non-relativistic theory,
A fine exception is 'Philosophical Foundations of QFT' ed., Brown &
Harre (Oxford). Therein we find a kind of battle cry to the effect
that QFT ought to be the contemporary locus of research in technical
metaphysics. (This little volume also features an essay by Redfield on
nonlocality, Cao on gauge theory, and an essay by Saunders on QFT.)
>while those who are well versed in relativistic
> quantum field theory are usually well trained not to think too carefully
> about interpretational issues.
The "shut up and calculate" approach.
> To answer your question, in my view, when relativistic quantum field
> theory is interpreted as a statistical description of a theory of
> particle interactions using an information theoretic interpretation of
> quantum mechanics, then I consider it is complete in the sense of EPR -
How is the mathematics of information theory sufficiently rich to
capture QM?
> or if not complete at least not lacking in the very thing we were trying
> to study in the first place, which was Einstein's complaint.
What should we be trying to study? Isn't science all about correlating
observed phenomena?
Charles Francis
<c.h.th...@pgen.net> writes
>Brian J Flanagan <sen...@yahoo.com> wrote in message
>news:c4858886.02013...@posting.google.com...
>> Charles Francis wrote:
>>> The trick is to show that a model consisting of particle interactions
>>> leads to the actual laws, of qm and gtr, which we observe. Clearly such
>>> a model can only be described statistically (...)
>> This begs the question stated in EPR's paper, viz., can the QM
>> description of reality be considered complete?
>It also begs the question of what ARE the "actual laws"! A very important
>doubt centres around the notion of quantum engtanglement, for instance. It
>would be much easier to formulate a "complete" alternative to QM if the
>phenomenon did not happen. And if you look at the actual evidence, maybe it
>doesn't!
This is true, particularly in the sense that the moment you try and
measure it, it's not there. But then entanglement is only something
which occurs in a mathematical model which describes what we know of a
system, not what actually occurs in the system itself. We know it
happens in the mathematics, because mathematics is subject to proof, but
we should not be making a claim that it is a physical phenomenon on the
basis of anything in qm.
>For an authoratitive account of the present state of play re entanglement,
>see Prof Laloe's article:
>Franck Laloë, 'Do we really understand quantum mechanics? Strange
>correlations, paradoxes and theorems', American Journal of Physics, 69(6)
>pp 655-701, June 2001.
>
>Although Laloe expresses the belief that QM is correct, he admits that there
>are loopholes and says:
>"A fair summary of the situation is that no one has been able to disprove
>quantum mechanics".
>
>"Not disproven" is a far cry from "proven"!
The trouble is that for many people there is no such thing as a proven
scientific theory, so that the strongest thing they are able to say
about physical theory is that it is not disproven. If you have missed
all the debates on scientific truth, falsification, and Popper, you
should look up some of the old threads on the issue on Google.
Incidentally the status of this kind of belief (falsification, that it
is only possible to say scientific theories are not disproven) is pretty
shaky in philosophical circles these days. Popper started going out of
vogue about 30 years ago. It has taken quite a long time for this to
filter through to the wider public and there are still a lot of
newsgroup readers who hold to falsification, but over the time I have
been contributing the general view has changed, and whereas it would be
that most contributors took a falsificationist stance two or three years
ago, now most can see the faults in falsification.
But to return to the main point, notwithstanding Prof Laloe it is
possible to demonstrate the structure of quantum mechanics as a
universal mathematical structure applicable to *the information* we have
about measured states. This is why I find all this talk of loopholes so
strange. Qm comes about precisely because the best we can have is
imperfect information.
Regards
--
Charles Francis
Charles Francis
<e...@fredkin.com> writes:
>We have learned that such systems, even though extraordinarily simple,
>are capable of exhibiting the most complex behavior imaginable.
>"Digital Mechanics" (DM) refers to such systems (computers that are
>Cellular Automatons) that model physics. The concept is that DM would
>underlie QM. In essence, QM would not be a description of reality,
>but rather DM would relegate QM to being a mathematical shortcut to
>predict the probabilities of various outcomes of a deterministic
>digital process, when we only have partial knowledge of the state of
>the system.
Is this not just an attempt at developing a hidden variables theory, of
the sort which we already know does not work, but wrapping it up in
other language as though that would be a way around the problem?
Regards
--
Charles Francis
c.h.thompson
Charles Francis <cha...@clef.demon.co.uk> wrote in message
news:kHgQQTEg...@clef.demon.co.uk...
> In article <c4858886.02013...@posting.google.com>, Brian J
> Flanagan <sen...@yahoo.com> writes:
> >This begs the question stated in EPR's paper, viz., can the QM
> >description of reality be considered complete?
>
> Quantum mechanics is best understood as a "black box" theory of
> measurement, providing statistical correlations between "fore" and
> "after" states, without necessarily saying anything about what goes on
> between.
However, it is possible that Quantum Mechanics is not succeeding entirely in
this aim! If the Bell test loopholes prove impossible to close, this means
that the QM statistical formula may have jumped to a wrong answer in this
case. It may well be that all real experiments will for ever remain
explainable (as are all to date) using hidden variables. Actual experiments
have never been conducted in the perfect conditions in which Bell's test
provides a realist limit.
> As such it is not a complete theory of matter, but is a theory
> which would hold whatever form matter takes.
A replacement theory would not necessarily need to reproduce ALL of QM's
predictions. It is required only to reproduce experimental results.
Cheers
c.h.thompson
they have visited web sites mentioned in articles or that such web
sites conform to the sci.physics.research charter. -TB]
Nicolaas Vroom <nicolaa...@pandora.be> wrote in message
news:a34m86$a37$1...@news.state.mn.us...
>
> The target of my experiments is to do the
> true two slit experiments with single photons
> but before I do that I first do a set of
> 3 pre experiments to learn more.
Do have a look at:
www.aber.ac.uk/~cat/Suggestions/two_bs.htm
for one suggestion for an experiment on single photons.
c.h.thompson
Maury Markowitz <ma...@sympatico.ca> wrote in message
news:a3838b$5j6$1...@sue.its.caltech.edu...
> > >[Moderator's note: Right. That's why it doesn't transmit information
> > >faster than light. - jb]
> >
> > If the scheme is changed so that A can force the mixed state measured to
> > be up (resp. down), then that information is "teleported" to B which
will
> > measure down (reps. up)
>
> QM wavefunction collapse is largely beyond the scope of the rest of the
> thread, so it may be a trifle unfair to drop it in now. However there are
> points that need to be considered:
What you say is probably quite right, but I have it from the horse's mouth
that entanglement is not necessary for quantum teleportation! I happened to
hear Sam Braunstein talk about his paper with Furusawa:
Furusawa et al, Unconditional quantum teleportation, Science 282, 706-709
(1998)
Although they state that they use "EPR beams", and the text includes
statements about entangled beams having nonlocal correlations, Braunstein
told us that this is not the essence of teleportation. We had quite an
interesting conversation, as I felt that in this particular experiment he
was not even in his rights in saying that the two sides were independent.
One and the same pump laser powered both.
Charles Francis
<c.h.th...@pgen.net> writes
>Charles Francis <cha...@clef.demon.co.uk> wrote in message
>news:kHgQQTEg...@clef.demon.co.uk...
>> Quantum mechanics is best understood as a "black box" theory of
>> measurement, providing statistical correlations between "fore" and
>> "after" states, without necessarily saying anything about what goes on
>> between.
>However, it is possible that Quantum Mechanics is not succeeding entirely in
>this aim!
It can be shown that quantum mechanics succeeds perfectly in this aim.
Please stick to physics.
>If the Bell test loopholes prove impossible to close, this means
>that the QM statistical formula may have jumped to a wrong answer in this
>case. It may well be that all real experiments will for ever remain
>explainable (as are all to date) using hidden variables.
Please don't make claims which aren't true. The only hidden variables
which can even potentially explain anything is the Bohm theory, and that
is really just a convolution - as physical interpretation it leaves more
to be explained than it explains, and moreover has no relativistic
form.
> Actual experiments
>have never been conducted in the perfect conditions in which Bell's test
>provides a realist limit.
This is irrelevant to the mathematical form of quantum theory. Hilbert
space is just a way of doing a statistical analysis based on probability
theory without a random variable. You may call that an exotic
probability theory if you like, but QM works because the statistical
analysis is correct. You should hold your criticism at least until you
understand the analysis.
>> As such it is not a complete theory of matter, but is a theory
>> which would hold whatever form matter takes.
>A replacement theory would not necessarily need to reproduce ALL of QM's
>predictions. It is required only to reproduce experimental results.
It would have to reproduce the general theory of relationships between
in and out states, and if it did that it would actually be quantum
mechanics.
Ed Fredkin
> Ed Fredkin <e...@fredkin.com> wrote:
[unnecessary quoted text deleted by angry gods]
> > Yes, discrete physics is not inherently falsifiable. But to say
> > "...and therefore not scientific." is simply wrong. Discrete physics
> > is obviously verifiable. Discrete matter (particles and atoms) is
> > verifiable. Discrete spin state is verifiable. Discreted electrical
> > charges is verifiable.
> How would you go about that?
> How could you -verify- that the distribution of spin values really is
> delta(x + 1/2) + delta(x - 1/2),
> and not two Gaussians with a width too small to resolve?
>
> You may have a theory that tells you so;
> you may verify that said theory agrees with experiment,
> but you cannot verify that the distribution really -is- discrete,
> or that this theory must be the final word.
With regard to spin, I misspoke. What I was thinking about was that
insofar as measurement is concerned, spin is a 2 state system. By
"Discrete spin state is verifiable." I meant that a measurement of
that state is capable of being specified exactly by an integer; +1 or
-1 for up or down. If we have a discrete and finite model like
Digital Mechanics, where space and time are discrete, (a discrete
lattice where a space-time lattice coordinate can always be specified
by 4 integers and the state at each lattice point is a small integer)
then everything in such a model is ultimately discrete. While QM
tells us that spin can be in a superposition of states, in a correct
DM model, there would still be a finite, discrete representation of
the exact state.
As to our concepts of the ultimate reality that corresponds to every
correct model of physics, it must eventually boil down to Occam's
Razor.
There must be experimental tests that can eventually help us to decide
continuous vs discrete. DM predicts that both translational symmetry
and angular isotropy must both be violated, but the experimental
verification of that violation might be very similar to the violation
of CP symmetry by the decay of the neutral kaon.
Shades of Mach. He never saw an atom, concluded that no one would
ever see an atom and therefore saw no reason to believe in atoms.
Today, Occam would not side with Mach.
Ed F
Ed Fredkin
No. Conclusions about "…the sort which we already know does not
work…" rely on assumptions that are not applicable to discrete,
deterministic, universal models such as Digital Mechanics. Those who
pioneered the concept that there cannot be a hidden variable model
under QM didn't consider models like DM, which are not limited to what
can be done with mathematical models. This has nothing to do with
"…other language…" but it has to do with concepts that are totally
foreign to any of the mathematical models heretofore used in physics.
DM models share one quality in common with QM, there is no way, in
general, to calculate the exact future outcome of a given state by
means of any kind of analytical expression. Further, DM models have
the property that an observer will not be able to either measure or
determine any exact states.
One of the most amazing facts about a valid DM model (if there is such
a thing) would be the fact that all fundamental constants of physics
could be calculated from the model, with no need for any physical
experiments (other than those needed to verify the model, and we've
got those already). See my post "Baez's Dream".
Best regards
p.ki...@ic.ac.uk
> p.ki...@ic.ac.uk writes
>>Charles Francis <cha...@clef.demon.co.uk> wrote:
>>> The photon will only be detected at one point.
>>> The photon will only be detected at one point.
>>is clearly wrong.
> It is of course trivial that you cannot detect a point more accurately
> than the resolution of your apparatus allows.
I am saying that photons are not necessarily pointlike, which has little
to do with the resolution of some apparatus. My spherical photodetector
example was to show that a non-pointlike photon could be detected in
a consistent manner with a non-pointlike detector, with no theoretical,
physical, or philosophical requirement for the photon to be described
as having a particular direction.
Photons are mostly described as single quantum excitations of harmonic
oscillators inside modes of the EM field[1]. In a spherically symetric
case, is is sensible to use spherically symetric modes, in which case
the photons we are using to describe the situation are non-pointlike
and have no particular directionality. This has nothing to do with
some detector or other I may (or may not) chose to use.
If I swap photodetectors, and use one which covers only a small solid
angle of the "spherical photon", I am still unrestricted by my choice
of spherical photons as a basis, although the maths may become less
convenient. It is a matter of _choice_ as to which modes of the EM
field I use to construct my photons.
Some choices of photon modes might be so localised (wrt the rest of the
situation) as to be regarded as being pointlike, but to generalise
from that kind of special situation and make a statement that
"photons are pointlike", or, for example, as you did:
> The photon wave function will propagate in a sphere, but all that
> really does is tell us that the photon is equally likely to be found in
> any direction, not that the photon travels in more than one direction.
is, I insist, misleading and wrong. A photon in a spherically
propagating mode _is_ travelling in all directions at once. Just
bcause I might decide to interact it with a localised detector
and destroy or alter "the photon" does not mean it was (described
as) travelling in some particular direction all along.
A simplification like your is, IMO, worse for being _unecessary_
even in that case where you are trying to formulate your descriptions
to make them understood by a wide audience. What is so hard to
understand about "photons can travel in all directions at once"?
> [...]
> As for the resolution of these issues there is no universally accepted
> solution. But it is certainly unfair to take an explanation which I have
> clearly simplified for the benefit of those unused to such concepts, and
> call it wrong on the grounds of much deeper, and unresolved issues
> concerning the fundamental nature of space-time, and the sense in which
> it may be possible to call something a point.
My argument has nothing to do with any unresolved issues concerning the
fundamental nature of space-time, etc.
[1] e.g. Loudon, "The quantum theory of light", OUP.
--
---------------------------------+---------------------------------
Dr. Paul Kinsler
Department of Physics (QOLS)
Imperial College (ph) +44-20-75947520
Prince Consort Road (fax) +44-20-75947714
London SW7 2BW Dr.Paul...@physics.org
United Kingdom http://www.lsr.ph.ic.ac.uk/~kinsle/
Charles Francis
writes:
>Charles Francis <cha...@clef.demon.co.uk> wrote:
>> It is of course trivial that you cannot detect a point more accurately
>> than the resolution of your apparatus allows.
>I am saying that photons are not necessarily pointlike, which has little
>to do with the resolution of some apparatus. My spherical photodetector
>example was to show that a non-pointlike photon could be detected in
>a consistent manner with a non-pointlike detector, with no theoretical,
>physical, or philosophical requirement for the photon to be described
>as having a particular direction.
I consider that photons are necessarily pointlike, and that your example
did not demonstrate that position to be either misleading or wrong as I
was charged. The most you can say about an example like this is that it
leaves the issue unresolved.
>Photons are mostly described as single quantum excitations of harmonic
>oscillators inside modes of the EM field[1].
I do know how the standard descriptions go. But the question of how the
formulae should be interpreted is altogether different and more subtle.
>Some choices of photon modes might be so localised (wrt the rest of the
>situation) as to be regarded as being pointlike, but to generalise
>from that kind of special situation and make a statement that
>"photons are pointlike", or, for example, as you did:
Obviously there needs to be much more to my arguments than I can put
into every post.
>> The photon wave function will propagate in a sphere, but all that
>> really does is tell us that the photon is equally likely to be found in
>> any direction, not that the photon travels in more than one direction.
>is, I insist, misleading and wrong. A photon in a spherically
>propagating mode _is_ travelling in all directions at once.
No, that is misleading and wrong. The photon wave function travels in
all directions at once, but the photon travels in no defined direction.
The wave function is just a mathematical device used in the calculation
of probabilities. There is no way to say that there is an ontological
wave, or that anything actually travels in all directions.
> Just
>bcause I might decide to interact it with a localised detector
>and destroy or alter "the photon" does not mean it was (described
>as) travelling in some particular direction all along.
The photon was not travelling in some particular direction all along.
The concept of direction was ill defined, and it is quite impossible to
say that the photon had any direction of travel before you interacted
with it.
>A simplification like your is, IMO, worse for being _unecessary_
>even in that case where you are trying to formulate your descriptions
>to make them understood by a wide audience. What is so hard to
>understand about "photons can travel in all directions at once"?
And what is so hard to understand about basic quantum mechanics in the
orthodox interpretation as described by Dirac:
In the general case we cannot speak of an observable having
a value for a particular state, but we can .... speak of the probability
of its having a specified value for the state, meaning the probability
of this specified value being obtained when one makes a measurement of
the observable (Dirac)
Direction is an observable. What is so hard to understand about the
statement that "we cannot speak of the photon having a direction until a
measurement is done"?
It is quite wrong on this basis to claim that the photon can travel in
all directions. The photon does not even have the property of a
direction of travel until it is detected. As it has no direction it is
quite wrong to say it has all directions.
>My argument has nothing to do with any unresolved issues concerning the
>fundamental nature of space-time, etc.
I'm afraid it does, because you took a definite stance against a
statement which only makes sense in the context of issues concerning the
fundamental nature of space-time.
The statement that a photon is pointlike is not just a statement about a
photon. It is also a statement which uses the concept of a point, and a
point is a fundamental in the description of space-time. We do not have
to assume that space has a structure described by classical geometry, or
that the existence of points implies the existence in its own right of a
physical Riemannian manifold, and according to the orthodox
interpretation of qm it is actually wrong to make such an assumption.
Before saying that it is definitely wrong to call a photon pointlike you
should try to understand what that statement might actually mean, in
what sense it may be possible to talk of points, and you should
recognise that any discussion on the fundamental nature of a point is
also a discussion of the fundamental nature of space-time.
--
Charles Francis
Charles Francis
<e...@fredkin.com> writes
>Those who
>pioneered the concept that there cannot be a hidden variable model
>under QM didn't consider models like DM, which are not limited to what
>can be done with mathematical models.
You mean that DM is not limited to that which is can be described in a
logically consistent manner?.
(Oh, I know, you didn't mean to say that, but I could help quoting it
because I thought it was a classic. )
c.h.thompson
Charles Francis <cha...@clef.demon.co.uk> wrote in message
news:Atr5qRJI...@clef.demon.co.uk...
> In article <3c5a5...@news2.vip.uk.com>, c.h.thompson
> <c.h.th...@pgen.net> writes
> >It also begs the question of what ARE the "actual laws"!
> >A very important doubt centres around the notion of
> >quantum entanglement, for instance. It would be much
> >easier to formulate a "complete" alternative to QM if the
> >phenomenon did not happen. And if you look at the
> >actual evidence, maybe itdoesn't!
[skip]
Laloe:
> >"A fair summary of the situation is that no one has been
> >able to disprove quantum mechanics".
CHT:
> >"Not disproven" is a far cry from "proven"!
> But to return to the main point, notwithstanding Prof Laloe it is
> possible to demonstrate the structure of quantum mechanics as a
> universal mathematical structure applicable to *the information*
> we have about measured states.
This is where you are wrong, Charles! This is the whole point of the
derivation of Bell's inequalities. Notwithstanding the opinions of some of
the people who "apply" the inequalities, producing what has come to be known
as "Bell states", if QM were merely modelling a change in information then
Bell's inequalities would not be violated. The fact is that it cannot be
said that they ARE! All experiments to date have had "loopholes".
Now, of course, it might be said that ALL experiments ever conducted have
loopholes, but these Bell test ones are in a class on their own. The tests
are supposed to be checking a sensational claim: that the world at the
quantum level does not obey the same laws of local causality as we had been
assuming. People such as Newton, who had formulated laws that appeared to
model action at a distance, knew perfectly well that this was not the end of
the story -- that underlying his mathematical rules there had to be
something physical, causing his laws to happen. The twentieth century
appears to be different: theoretical physicists have accepted the
impossible, without, it seems, stopping to examine the evidence!
You are arguing, with some people working in the area (notably Prof Anton
Zeilinger), that the quantum mechanical formula is merely modelling the
information that we have. Presumably you are going along with Zeilinger in
believing that the observed "coincidence curves" merely reflect the results
of conditional measurements -- which is indeed the case. No experiment
conducted in Zeilinger's lab, though, or indeed anywhere in the world, has
violated a genuine Bell test.
Bell's reasoning, incorporating the assumption that there are hidden
variables causing the correlations and that the particle detectors are
acting independently, leads to his tests. Under perfect conditions, if the
tests are violated it means that something has happened that is against what
we had previously held to be laws of Nature. But the actual measurements
are made under conditions that do not satisfy the requirements of a true
Bell test! They are just ordinary measurements. Yes, they violate Bell's
inequalities, but this is of no fundamental significance because there are
these "loopholes"!
Have you read a letter of mine that was published in Physics World last
November? Do have a look (www.aber.ac.uk/~cat/Letters/PW_tangled.htm ).
In it I pointed out that the experts know about the problems. They are
fairly careful not to exaggerate the evidence. Most people only know about
Bell tests at second or third hand, though. They do not understand them.
Scientific journalists have been only too happy to make sensational stories
about the weird quantum world. The "official" journals such as Physical
Review Letters have been reluctant, to say the least, to publish articles
that explain the weaknesses. Articles that I myself have submitted have
been rejected, not on the grounds that I was wrong but mainly because the
referees thought that what I was saying was already well known!
How have we let this happen?
How has it come about that a belief that scientists have observed the
impossible is fast becoming embedded in our culture?
The experiments are not THAT hard to understand! Some of the faults are
very straightforward. Many experiments, for example, involve analyses
conducted after "subtraction of accidentals". Although the subtraction is
common practice and makes perfect sense in other contexts, it is not hard to
see that it invalidates the Bell tests. Other things being equal, the tests
apply to the raw coincidence rates, not the adjusted ones. My first paper
on this subject was, as it happens, quoted once by Wolfgang Tittel , of the
Geneva long distance Bell tests.
See:
C. H. Thompson, "Timing, 'Accidentals' and Other Artifacts in EPR
Experiments" (1997), <http://arXiv.org/abs/quant-ph/9711044> and
W. Tittel et al., "Violation of Bell inequalities by photons more than 10 km
apart", Physical Review Letters 81, 3563 (1998),
<http://xxx.lanl.gov/abs/quant-ph/9806043>
For more, do have a look at papers on my web site, under
www.aber.ac.uk/~cat/bibliography.htm . I'm working at present on a shorter
version of my first "Chaotic Ball" paper.
For some interesting opinions on QM, including those of Zeilinger, see:
Physics Today, February 1999, pp 11-15 and 89-92, correspondence re
Goldstein's article, Physics Today, March 1998, pp 42-46
> This is why I find all this talk of loopholes so
> strange. Qm comes about precisely because the best we
> can have is imperfect information.
I fear that the QM behind the entanglement prediction is based on some
strange assumptions. A model based on imperfect information is, I'm sure,
possible. This would be a "hidden variable" model of the kind that the Bell
tests are currently claimed to have proved can never be done. It would show
subtle differences from the QM model for separated particles.
The quantum computing people, incidentally, ought to be glad to be rid of
the "entanglement" idea! They can carry on getting essentially the same
correlations but will have a better understanding of their cause, and hence
better control of the situation, if they allow for the "local realist"
physics that lies behind them.
Nicolaas Vroom
Charles Francis wrote:
> >(We first have to agree about the numbers
> >the position were comes later)
>
> I don't see why you think the numbers are that important. It is
> generally the distribution of hits on the grid that is considered
> important.
I agree with you.
A single slit gives a normal distribution
A double slit gives interference pattern
See for example John Gribbin
"In search of schrodinger's cat" page 167
However before I want to explain those results
I first want to learn more.
(Also how difficult the experiment is)
If only one slit is open:
Q: How many hits are there ?
If I ^only^ move the one slit slightly
Q: How many hits are there ?
I can repeat this for many positions of the one slit.
You will then also get a normal distribution of hits
as a function of the position of the grid.
Q: How many hits are there in total.
If each run consists of 100 photons I expect more
than 100 hits in total.
(If that is the case a single photon is not so
"specific" exactly where a slit is.
The direction of the photon generator is not so
important in order that a photon goes through a slit)
> >Did you ever do the experiment in this way
> >considering those 4 possibilities ?
>
> No. I would hope an experimentalist may be able to give you a better
> answer.
I hope so too.
Thanks for your efforts.
> Regards
>
> --
> Charles Francis
Nicolaas Vroom
"c.h.thompson" wrote:
> Do have a look at:
> www.aber.ac.uk/~cat/Suggestions/two_bs.htm
> for one suggestion for an experiment on single photons.
I had a look, but more specific about the article:
http://arXiv.org/pdf/quant-ph/9811078
The main subject is about entanglement (between two beams)
The article mentions "entangled photon pairs"
I have also many (practical) questions related
to photon entanglement.
Q: How entangled are two photons ?
Maybe a better qustion is:
Q: How entangled can two photons be ?
To be more specific:
Q: If the spin of one photon is (x,y,z) = (10,10,10)
(assuming you can measure all the three coordinates)
what is than the chance that the other photon
has the coordinates (-10,-10,-10)
The article also mentions: teleportation
IMO you can only study teleportation if you
know all the ins and out (nitty-gritty details)
of photon entanglement experiments.
specific how difficult those experiments are
and how accurate they can be performed.
Nick
Brian J Flanagan
> >However, it is possible that Quantum Mechanics is not succeeding entirely in
> >this aim!
>
> It can be shown that quantum mechanics succeeds perfectly in this aim.
Why was QM superseded by QFT?
Brian J Flanagan
>
The concept is that DM would
> >underlie QM. In essence, QM would not be a description of reality,
> >but rather DM would relegate QM to being a mathematical shortcut to
> >predict the probabilities of various outcomes of a deterministic
> >digital process, when we only have partial knowledge of the state of
> >the system.
>
> Is this not just an attempt at developing a hidden variables theory, of
> the sort which we already know does not work, but wrapping it up in
> other language as though that would be a way around the problem?
Yes, this does sound like a hidden variables approach, but the notion
that we "know" that such approaches cannot work is highly dubious, and
more a matter of popular, and largely uninformed, opinion.
Charles Francis
In article
<remove.haberg-3...@du137-226.ppp.su-anst.tninet.se>, Hans
Aberg <remove...@matematik.su.se> writes:
>In article <EgnopSK9...@clef.demon.co.uk>, Charles Francis
><cha...@clef.demon.co.uk> wrote:
>>But while the physical quantities must be defined, their may be a huge
>>range of possible definitions, and the actual definitions chosen must
>>not be important. There are numerous examples of this in physics. As I
>>recall temperature is defined according to a number of different
>>measurement techniques for different temperature ranges.
>I think there is a difference between physical quantities and measurement
>techniques.
No there isn't. I cited temperature because I recall it as being
particularly obvious that there is no definition other than the
measurement process used to measure it. But the same is true of all
physical quantities. If you do not define the process used to measure
the quantity, you cannot say what the quantity is.
>If one is using different measurement techniques for measuring
>the same physical quantity, I figure it would be prudent to demonstrate
>that these different measurement techniques measure the same physical
>quantity, otherwise one might end up with some physical embarrassments.
It's called calibration. When we use another measurement technique we
must ensure it is calibrated to give the same result as the defining
measurement process.
>>Likewise with the maximum speed of information.
>The problem is that "information" has not been defined as a physical quantity.
Perhaps because information is not a physical quantity, but a mental
one. This makes it in many ways easier to define. We have experience of
the colour red, which gives us direct knowledge of what it is, but when
we try to understand what an electron is, we have no experience and our
description boils down to some equations, yet the electron is not an
equation. Still, no one said understanding the universe would be easy.
>> We know that in practice
>>light is a carrier of information,
>Not really: If information is transport of knowledge, then we know that
>light can be used to transport information, but what information? -- If
>the light is a Morse signal or a picture beautiful woman, then that is
>quite different kinds of information, but what formulas in physics allows
>us to identify what type of information?
The type of information is irrelevant. We are mathematising physics so
we intend to abstract (meaning draw out) the relevant and discard the
irrelevant, just as we do when we write 4+5=9, and don't identify
whether we are talking of apples or oranges.
>> and we also know that, to the best of
>>our ability to determine it, light travels at the maximum speed of
>>information.
>But we also know that a common way of transporting information is not via
>light, but via electrons, like for example this email. It also depends on
>exploiting QM fields, in transistors.
Yes, and the same constraint applies. There is a maximum speed at which
information can be transported in electric wires too. I have heard it
said that this is also equal to the speed of light, but I suspect that
depends on some idealisation of the structure of the metal used in the
wire.
>But I object to that one should be able to claim that information must be
>transported by photons, or that jumping to the conclusion that just
>because this is the limit speed of information transported by photons,
>that must be the limit of other ways of transporting information (say via
>QM teleporting).
I have not made either claim. Actually I said the opposite. That the
maximum speed of information is the same, irrespective of the manner in
which it is transported. I am talking of an absolute maximum speed, it
is convenient to measurement techniques that it happens that photons
actually travel at that absolute maximum. But as I say, the mathematical
form of relativity would be the same even if they did not. Which is why
it is strictly more accurate to call c the maximum speed of information
than it is to call it the speed of light.
>It seems me both simpler and safer to say that if photons do have a small
>mass, then c is the limit speed of high energy photons: This avoids all
>references to a physically fuzzy concept like "information".
But what if you were right and there was some quantum process that
transported information faster than the limiting speed of high energy
photons? If there were we could use this process to determine the
space-time manifold, and we would have to call this speed c. By defining c
as the maximum speed of information we sidestep the nasty issue of making
physical theory dependent on data (induction which has taxed
philosophers for centuries), and find general laws which must hold with
absolute veracity whatever the actual form and behaviour of matter.
Regards
--
Charles Francis
Charles Francis
In article <3c612...@news2.vip.uk.com>, c.h.thompson
<c.h.th...@pgen.net> writes:
>> But to return to the main point, notwithstanding Prof Laloe it is
>> possible to demonstrate the structure of quantum mechanics as a
>> universal mathematical structure applicable to *the information*
>> we have about measured states.
>This is where you are wrong, Charles! This is the whole point of the
>derivation of Bell's inequalities.
I am not wrong. You have to pay more heed to what "information" actually
means.
>Notwithstanding the opinions of some of
>the people who "apply" the inequalities, producing what has come to be known
>as "Bell states", if QM were merely modelling a change in information then
>Bell's inequalities would not be violated.
Bell's theorem assumes that the information exists but is not known.
That is not the situation described in any quantum mechanical
experiment, most of which are far easier than the Aspect experiment.
Quantum mechanics applies when definite information does not even exist,
and does not even make sense prior to measurement.
>Now, of course, it might be said that ALL experiments ever conducted have
>loopholes, but these Bell test ones are in a class on their own. The tests
>are supposed to be checking a sensational claim: that the world at the
>quantum level does not obey the same laws of local causality as we had been
>assuming.
Yes, we know that.
>People such as Newton, who had formulated laws that appeared to
>model action at a distance, knew perfectly well that this was not the end of
>the story
Quite.
> that underlying his mathematical rules there had to be
>something physical, causing his laws to happen.
And I agree, there must be.
>The twentieth century
>appears to be different: theoretical physicists have accepted the
>impossible, without, it seems, stopping to examine the evidence!
Not all theoretical physicists. It is not impossible, just very
difficult to understand. Many theoretical physicists do accept it
without further examination of the concepts, but I for one do not.
>How has it come about that a belief that scientists have observed the
>impossible is fast becoming embedded in our culture?
>
>The experiments are not THAT hard to understand!
The analysis of the situation in EPR is hard to understand. It is
extremely difficult to think about the implications of Leibniz' argument
that all measurement is relative or Descartes' argument that it makes no
sense to describe points in space in terms of his Cartesian coordinate
system, because location is a concept which can only correctly be
applied to an object when it is touching another object, and it is the
framework of contiguous matter which gives us our ideas of position, not
some fictitious empty space described by coordinates.
This is very difficult to understand indeed. Only a few of the most
intelligent men in history made any headway at all with the idea before
QM (I would cite Descartes, Leibniz, Gauss, Riemann, Einstein in the
sense that relativity of motion is another aspect of the same idea) and
even after QM very few have made much headway (notably Heisenberg
himself, Dirac and Von Neumann), and none of them completely sorted the
idea out. But once you do understand it you start to realise that not
only does qm make sense, but it is also just an inevitable body of
mathematical rules coming out of a correct statistical analysis.
>Some of the faults are
>very straightforward. Many experiments, for example, involve analyses
>conducted after "subtraction of accidentals". Although the subtraction is
>common practice and makes perfect sense in other contexts, it is not hard to
>see that it invalidates the Bell tests.
QM does not, in any case, hold or fail on the Bell tests. There is a huge
range of phenomena where the laws of QM are essential to correct
predictions.
>> This is why I find all this talk of loopholes so
>> strange. Qm comes about precisely because the best we
>> can have is imperfect information.
>I fear that the QM behind the entanglement prediction is based on some
>strange assumptions. A model based on imperfect information is, I'm sure,
>possible. This would be a "hidden variable" model of the kind that the Bell
>tests are currently claimed to have proved can never be done.
No. That is the point. It is not a hidden variables model. Actually
hidden variables models of this sort were eliminated by Von Neumann
simply on the basis that they are incompatible with the observed
statistical predictions of qm in conceptually much easier experiments
than Bell tests. In one way or another a hidden variables theory can be
modelled using classical probability theory based on one or many random
variables. In quantum mechanics we have to recognise that the random
variable is a meaningless idea, and develop a statistical analysis in
which there is no random variable unless you do an experiment and modify
the situation in such a way as to create one.
>It would show
>subtle differences from the QM model for separated particles.
>
>The quantum computing people, incidentally, ought to be glad to be rid of
>the "entanglement" idea! They can carry on getting essentially the same
>correlations but will have a better understanding of their cause, and hence
>better control of the situation, if they allow for the "local realist"
>physics that lies behind them.
Actually this is one case where I think "shut up and calculate" probably
is the easiest way to establish the correct circuitry.
--
Charles Francis
Eric Dennis
> Please don't make claims which aren't true. The only hidden variables
> which can even potentially explain anything is the Bohm theory, and that
There are others. Nelson's stochastic mechanics and spontaneous
localization models to name two. Sheldon Goldstein had a good overview
in _Physics Today_ a while back. See:
http://www.math.rutgers.edu/~oldstein/papers/qts/qts.html
> is really just a convolution - as physical interpretation it leaves more
> to be explained than it explains, and moreover has no relativistic
> form.
Not true. See "Beables for quantum field theory" in John Bell's
_Speakable and Unspeakable in QM_.
Ed Fredkin
DM is a universal process (a Universal Computer in the Turing sense),
not a set of equations. It is absolutely true that the set of
possibilities that can be described by the kinds of mathematical laws
we have in physics does not include all the possibilities of a
Universal Computer. This fact is well understood in Computer Science
(actually, Automata Theory). RUCAs (Reversible Universal Cellular
Automata) exhibit emergent behavior that is hard to believe! They
produce all kinds of surprising patterns and structures, apparent
randomness, particles, long range correlations that seem like action
at a distance, etc.
We all ought to be continually amazed at the power and scope of
mathematical models that capture aspects of our world from microscopic
to intergalactic. DM is something else, probably not useful for any
aspect of physics except for modeling one thing, the lowest level most
fundamental process. The nature of DM, which is not mathematical,
holds promise for allowing the analytical derivation of the
mathematical laws of physics.
DM suggests a number of tests, such as searching for violation of
angular isotropy, in a manner similar to tests for other violations of
symmetry, e.g. CP.
There are a number of very attractive aspects to such models (see my
recent post to "Baez's Dream) or look at www.digitalphilosophy.org
Regards,
Ed F
Hans Aberg
<cha...@clef.demon.co.uk> wrote:
> >I think there is a difference between physical quantities and measurement
> >techniques.
>No there isn't.
So you claim that if somebody is making lousy measurements of the Earth's
orbit, then it is going to be pretty unstable?
>I cited temperature because I recall it as being
>particularly obvious that there is no definition other than the
>measurement process used to measure it.
Isn't there a theoretical statistical definition, by which say the boiling
temperature of water can be computed?
> But the same is true of all
>physical quantities. If you do not define the process used to measure
>the quantity, you cannot say what the quantity is.
It is true that one must be able to measure measurable physical quantities
(sounds like a tautology to me), but that is not the same thing as saying
that the measurable physical quantity is the same thing as the particular
measurement technique: Often there are several different measurement
techniques to measure a quantity.
Take for example temperature that can be measured by the expansion of
material, or via changed electrical properties of semi-conductors. One
would be led to find a unified theoretical definition of what the concept
of temperature is, explaining the two different measurement techniques
available.
> >The problem is that "information" has not been defined as a physical
> >quantity.
>Perhaps because information is not a physical quantity, but a mental
>one.
I think this is the point of it: We use intuitive mental concepts and try
to project them to defining physical theories. However, once, one tries to
define the physical theory, it must be more accurate.
>This makes it in many ways easier to define. We have experience of
>the colour red, which gives us direct knowledge of what it is, but when
>we try to understand what an electron is, we have no experience and our
>description boils down to some equations, yet the electron is not an
>equation. Still, no one said understanding the universe would be easy.
> >Some uncited soul wrote:
> >> We know that in practice
> >>light is a carrier of information,
> >Not really: If information is transport of knowledge, then we know that
> >light can be used to transport information, but what information? -- If
> >the light is a Morse signal or a picture beautiful woman, then that is
> >quite different kinds of information, but what formulas in physics allows
> >us to identify what type of information?
>The type of information is irrelevant. We are mathematising physics so
>we intend to abstract (meaning draw out) the relevant and discard the
>irrelevant, just as we do when we write 4+5=9, and don't identify
>whether we are talking of apples or oranges.
In the process of transforming the intuitive mental concept into a
physical theory, one must be more exact, defining the physical quantities
accurately.
So when transforming an intuitive concept as "information" into physical
quantities, the type of information may or may not have any significance
in the physical model: But we must specify that.
For example, information must not be transported by photons, and not all
information can be said to be transportable by photons. (For example,
information residing in the human brain is as far as we know today not
always transportable by photons. If it is, we do not have any means of
verifying that.)
> >But we also know that a common way of transporting information is not via
> >light, but via electrons, like for example this email. It also depends on
> >exploiting QM fields, in transistors.
>Yes, and the same constraint applies. There is a maximum speed at which
>information can be transported in electric wires too. I have heard it
>said that this is also equal to the speed of light, but I suspect that
>depends on some idealisation of the structure of the metal used in the
>wire.
The problem here is your theory of "maximum speed of information" is
invalid until verified. Therefore it is safer speaking about the speed of
the physical quantities involved.
> >But I object to that one should be able to claim that information must be
> >transported by photons, or that jumping to the conclusion that just
> >because this is the limit speed of information transported by photons,
> >that must be the limit of other ways of transporting information (say via
> >QM teleporting).
>I have not made either claim. Actually I said the opposite. That the
>maximum speed of information is the same, irrespective of the manner in
>which it is transported. I am talking of an absolute maximum speed, it
>is convenient to measurement techniques that it happens that photons
>actually travel at that absolute maximum.
But what says that if say quantum teleportation is possible, that there is
a maximum speed of information. Or if it is, such an upper limit of it
agrees with the upper limit of the speed of photons?
> But as I say, the mathematical
>form of relativity would be the same even if they did not. Which is why
>it is strictly more accurate to call c the maximum speed of information
>than it is to call it the speed of light.
No. This is why one should say that c is the supremum of the photon speed
if one means that.
Then put the information stuff into a bag, until you have pinned it down
better from the physical point of view.
> >It seems me both simpler and safer to say that if photons do have a small
> >mass, then c is the limit speed of high energy photons: This avoids all
> >references to a physically fuzzy concept like "information".
>But what if you were right and there was some quantum process that
>transported information faster than the limiting speed of high energy
>photons? If there were we could use this process to determine the
>space-time manifold, and we would have to call this speed c.
Not really: All measurements actually made suggest that c is the supremum
of the photon speeds. (By Cerenkov counters etc).
So if that c_i := c_information is much different from c_gamma, then one
would have to introduce a new name for it, just as I just did. :-)
Hans Aberg * Anti-spam: remove "remove." from email address.
* Email: Hans Aberg <remove...@member.ams.org>
* Home Page: <http://www.matematik.su.se/~haberg/>
* AMS member listing: <http://www.ams.org/cml/>
Charles Francis
Flanagan <sen...@yahoo.com> writes:
>Charles Francis wrote:
>> Quantum mechanics is best understood as a "black box" theory of
>> measurement, providing statistical correlations between "fore" and
>> "after" states, without necessarily saying anything about what goes on
>> between.
>How is this picture best?
Suppose we have a measurement with possible results described by the set
X. Then we can *formally* describe the information we have about the
probability of a given result of measurement of any state by forming the
linear span of X and for two states |f> |g> and for any two numbers a b
we can interpret a |f> + b |g> as a weighted logical OR between the
possible results |f> and |g>. Then we are free to *define* an inner
product <x|f> whose square is the probability that a measurement will
yield x if we start with the information described by |f>.
In other words this Hilbert space structure can simply be formally
defined as a labelling system containing only the information we have
about possible measurement results, and because this is done by
definitional truism and tautology the validity of doing so cannot be
doubted.
>In practice, physics puts together highly
>successful models of mechanisms thought to account for what goes on
>inside the black box.
I distinguish the formal structure on in and out states described by
Hilbert space from the additional information we may have about the
evolution of a system, which generally comes into the description in the
form of either a Schrodinger equation or a Lagrangian. Hilbert space
only describes potential measurement results. It is when you put the
formal structure of Hilbert space together with the model of a mechanism
to describe a particular situation that we have highly successful
physical theory.
>>As such it is not a complete theory of matter, but is a theory
>> which would hold whatever form matter takes.
>This is reminiscent of the S=matrix program; perhaps you know
>Cushing's book on this approach? He seems to think it's been
>discarded.
The S-matrix approach certainly has not been discarded, and indeed
Scharf's book "Finite QED" took an S-matrix approach as recently as 1989
What I would say, however, is that the S-matrix is only part of the
story. We should understand the S-matrix because that will tell us what
is in the theory by definitional truism and mathematics. It is when we
go beyond the S-matrix that we start to study physics. It is good to
make this distinction because otherwise the borderline between physics
and mathematics can become very confused.
>> To answer your question, in my view, when relativistic quantum field
>> theory is interpreted as a statistical description of a theory of
>> particle interactions using an information theoretic interpretation of
>> quantum mechanics, then I consider it is complete in the sense of EPR -
>How is the mathematics of information theory sufficiently rich to
>capture QM?
As I say, information theory only gives us the formal structure of
Hilbert space and explains why we should use Hilbert space to describe
the information we have about measurement and to give probabilistic
predictions. We have to add some actual mechanics to find the form of
the Schrodinger equation of Lagrangian.. The important point about this
approach is that it enables us to make a clear distinction between
mathematics and the physics. It becomes clear for example that Fourier
transforms are just linear combinations of states and are strictly to do
with information not physics, and likewise it becomes clear that wave
behaviour and collapse are both strictly to do with information, not
physics.
>> or if not complete at least not lacking in the very thing we were trying
>> to study in the first place, which was Einstein's complaint.
>What should we be trying to study? Isn't science all about correlating
>observed phenomena?
It is not just about observing correlations, but also about interpreting
them.
Regards
--
Charles Francis
Charles Francis
Flanagan <sen...@yahoo.com> writes
>Charles Francis wrote:
>c.h.thompson writes:
Goodness, do you want a long answer, a short answer, or is this, as I
suspect, just a rhetorical question and you would really rather I did
not answer at all?
I'll give you a short answer, that no one has been able to construct
relativistic QFT from simple non-relativistic quantum mechanics. But
that does not mean that such a construction is impossible, just that it
has some quite unexpected attributes, or it may be complicated in ways
that are of no real benefit to physics.
For example I am quite convinced that there is no simple construction
based on a single Hilbert space. But on the other hand if one starts
with a C*-algebra it is possible to extract a number of Hilbert spaces
from it. If you had some reason to start with the correct collection of
Hilbert spaces it should be possible to work back to the C*-algebra. But
if your reason for starting with this collection of Hilbert spaces is
essentially philosophical it would not be the stuff of typical physics
journals.
Regards
--
Charles Francis
Brian J Flanagan
> Though failing to properly cite himself, it was Brian Flanagan who wrote:
> >Why was QM superseded by QFT?
> Goodness, do you want a long answer, a short answer, or is this, as I
> suspect, just a rhetorical question and you would really rather I did
> not answer at all?
Please answer as you see fit. Why would I trouble you with a
rhetorical question?
> I'll give you a short answer, that no one has been able to construct
> relativistic QFT from simple non-relativistic quantum mechanics.
Well, as Dyson says quite plainly in his old Scientific American
article on "Field Theory", QFT can be roughly understood as relativity
+ QM, so "relativistic QFT" would seem somewhat redundant. And then,
given the successes of QFT, and the simplicity of (say) the Dirac
equation, I'm not sure what you mean by the above statement, except
perhaps that no perfect theory has yet been devised, which seems true
enough.
> But
> that does not mean that such a construction is impossible, just that it
> has some quite unexpected attributes, or it may be complicated in ways
> that are of no real benefit to physics.
What attributes are you referencing? The infinities?
> For example I am quite convinced that there is no simple construction
> based on a single Hilbert space. But on the other hand if one starts
> with a C*-algebra it is possible to extract a number of Hilbert spaces
> from it. If you had some reason to start with the correct collection of
> Hilbert spaces it should be possible to work back to the C*-algebra. But
> if your reason for starting with this collection of Hilbert spaces is
> essentially philosophical it would not be the stuff of typical physics
> journals.
Until the middle of the 19th century, scientists were known as natural
philosophers. ('Dawn to Decadence' Jacques Barzun) Given the fractious
relations between contemporary (ahem) scientists and philosophers, it
would seem problematical to reach agreement between them as to what
constitutes an "essentially philosophical" reason for anything, much
less one's choice of Hilbert spaces. Having said that, I am reminded
of 'Structure & Interpretation of QM' by RIG Hughes (Harvard?) which
features a section entitled "Why Hilbert Spaces?" wrt QM; This work
has the distinction of being both scientifically accurate and
philosophically sound. As for typical physics journals ... who wants
to be thought typical?
Charles Francis
<remove.haberg-0...@du135-226.ppp.su-anst.tninet.se>, Hans
Aberg <remove...@matematik.su.se> writes
>In article <vOaCy5Ly...@clef.demon.co.uk>, Charles Francis
><cha...@clef.demon.co.uk> wrote:
To the effect that measurable quantities are defined by the procedures
used to measure them.
>So you claim that if somebody is making lousy measurements of the Earth's
>orbit, then it is going to be pretty unstable?
By definition, if someone is making lousy measurements they are not
making measurements according to the defining methodology.
I am not referring to the actual methodology used, but to the fact that
whatever measurement technique you use it is calibrated to give the same
answer (to limits of accuracy) as the defining measurement.
>>I cited temperature because I recall it as being
>>particularly obvious that there is no definition other than the
>>measurement process used to measure it.
>Isn't there a theoretical statistical definition, by which say the boiling
>temperature of water can be computed?
Although temperature is defined to give some reasonable approximation to
Boyle's law for an "ideal" gas the behaviour of real gases is more
complicated, and Boyle's law is only approximate. The behaviour of
liquids is more complicated again, especially a quirky one like water.
>>But the same is true of all
>>physical quantities. If you do not define the process used to measure
>>the quantity, you cannot say what the quantity is.
>It is true that one must be able to measure measurable physical quantities
>(sounds like a tautology to me)
When you understand what I am trying to say you will realise that it is
tautology. The really surprising thing is that so much real physics
actually depends on tautology.
>but that is not the same thing as saying
>that the measurable physical quantity is the same thing as the particular
>measurement technique: Often there are several different measurement
>techniques to measure a quantity.
Perhaps you could check through the internationally agreed definitions
to see that all physical processes, from the measurement of time
onwards, are defined, directly or indirectly, by specific measurements.
When we use another measurement technique we must ensure it is
calibrated to give the same result as the defining measurement process,
so that logically the different measurement technique is merely an
indirect way of measuring the quantity defined by some measurement
process described in the internationally agreed standards.
For example in the not so dim or distant past the standard meter was
defined as the distance between two lines on a *particular* platinum
iridium rod held at 0 deg C in a vault in Paris, and rarely touched
even to take measurements of it. They would calibrate other rules from
it, and then use these other rules to calibrate still more rules, so
that ultimately every measurement of a meter was calibrated to the
original rod in Paris. It did not matter how the calibration was done,
but only that every measurement was calibrated. So whatever actual
measurement technique used, ultimately every measurement of distance was
a comparison to that particular rod.
Now the definition has changed, but every measurement of a meter has to
be calibrated to a set number of wavelengths in vacuum of a particular
orange radiation of krypton 86. So that whenever you measure a meter,
whatever the actual measurement technique, the defining feature of the
measurement is that you are making a comparison to that wavelength of
radiation.
>Take for example temperature that can be measured by the expansion of
>material, or via changed electrical properties of semi-conductors. One
>would be led to find a unified theoretical definition of what the concept
>of temperature is, explaining the two different measurement techniques
>available.
When I was at school the temperature scale was defined by different
techniques for different temperature ranges, and I imagine this must
still be the case. Any measurement technique must still be calibrated to
the defining technique within the appropriate range.
>> We are mathematising physics so
>>we intend to abstract (meaning draw out) the relevant and discard the
>>irrelevant, just as we do when we write 4+5=9, and don't identify
>>whether we are talking of apples or oranges.
>In the process of transforming the intuitive mental concept into a
>physical theory, one must be more exact, defining the physical quantities
>accurately.
Yes, in the process of applying mathematics to physics. Of course 4+5=9
is still exact, but the point is that it remains exact whatever physical
objects one applies it to.
>So when transforming an intuitive concept as "information" into physical
>quantities, the type of information may or may not have any significance
>in the physical model: But we must specify that.
What we specify is that the statement of a maximum speed of information
remains exact irrespective of the type of information. If you must have
a physical definition of information, use the one from computer science,
that it is anything which can be stored as a finite array of 0's and
1's.
>For example, information must not be transported by photons, and not all
>information can be said to be transportable by photons. (For example,
>information residing in the human brain is as far as we know today not
>always transportable by photons. If it is, we do not have any means of
>verifying that.)
We don't know how to quantify what happens in the brain, so we can't be
scientific about it. I fear that information in the brain will have to
remain outside of the definition of information, although certainly
quite a lot of it can be stored in computers as 0's and 1's
>The problem here is your theory of "maximum speed of information" is
>invalid until verified.
How do you verify something which is actually true by tautology?
>But what says that if say quantum teleportation is possible, that there is
>a maximum speed of information. Or if it is, such an upper limit of it
>agrees with the upper limit of the speed of photons?
That is actually one of the main reasons why we have to use an abstract
concept like the speed of information, rather than a physical speed of
an object. Quantum teleportation is possible, and there is no upper
limit to the speed of individual photons. In relativistic quantum field
theory the probability amplitude for particle creation and annihilation
outside the light cone is non zero, so it can happen that a photon moves
faster than light. The relativistic condition (to do with the
commutator) actually says no observable effect may travel faster than c,
not that nothing may travel faster than c.
>> But as I say, the mathematical
>>form of relativity would be the same even if they did not. Which is why
>>it is strictly more accurate to call c the maximum speed of information
>>than it is to call it the speed of light.
>No. This is why one should say that c is the supremum of the photon speed
>if one means that.
As mentioned above there is no such thing. Again this is why it is
dangerous to be too dependent on specific assumed properties of matter
when making one's definitions. Matter has this awful habit of not
behaving how you thought it was going to in your assumptions, so that
your definitions break down when you get a bit deeper into it.
>Then put the information stuff into a bag, until you have pinned it down
>better from the physical point of view.
>>But what if you were right and there was some quantum process that
>>transported information faster than the limiting speed of high energy
>>photons? If there were we could use this process to determine the
>>space-time manifold, and we would have to call this speed c.
>Not really: All measurements actually made suggest that c is the supremum
>of the photon speeds. (By Cerenkov counters etc).
>
>So if that c_i := c_information is much different from c_gamma, then one
>would have to introduce a new name for it, just as I just did. :-)
Perhaps. But it is c_i, not c_gamma, which would appear in the
relativistic formulae.
Regards
--
Charles Francis
John Baez
Peter Shor <sh...@research.att.com> wrote:
>I believe there's a real disconnect as to what people mean by
>fundamentally discrete here. I think Ed Fredkin wants to throw away
>all of quantum mechanics, and propose that the world is a classical
>discrete cellular automaton. Whereas I think what John Baez means by
>fundamentally discrete much more closely resembles quantum mechanics
>over a finite-dimensional Hilbert space.
Something like that, but I wasn't trying to be very precise -
I was trying to vaguely include all sorts of obscure possibilities.
I was mainly hinting at the discrete spectrum of the area and
volume operators in loop quantum gravity, but I've also published
a paper with James Dolan - "From Finite Sets to Feynman Diagrams" -
which shows how the whole machinery of Fock space, annihilation/creation
operators, Feynman diagrams and the like arises naturally from
studying structures on finite sets - what combinatorists call
"species" or "structure types", or more generally, what we call
"stuff types". The complex numbers only show up as a kind of
afterthought! While it's way too soon to do actual physics using
this math, it makes me more optimistic that quantum theory has a
kind of combinatorial "skeleton" on which the complex numbers are
naturally somehow draped. Just don't anyone ask me what that means!
>Anyway, I think the continuous/discrete classification needs a new
>category: quantum.
That's certainly true - and when it comes to fundamental physics,
all I really care about are theories that take quantum theory into account.
>What's probably closer to John Baez's belief (and if
>he's reading this, he can correct me) is that any chunk of the
>universe of volume V can be represented (modulo boundary effects) by a
>finite-dimensional Hilbert space.
Yeah, something like that: the Bekenstein bound actually puts
an upper bound on the dimension of this Hilbert space in terms
of the *area* of the *boundary* of this chunk of the universe,
and while people have found violations to the original crude
statement of the Bekenstein bound, the more carefully stated
"Bousso bound" seems to be holding up. I guess my original
vague remark was also meant to include this!
Brian J Flanagan
> > This begs the question stated in EPR's paper, viz., can the QM
> > description of reality be considered complete?
>
> It also begs the question of what ARE the "actual laws"! A very important
> doubt centres around the notion of quantum engtanglement, for instance. It
> would be much easier to formulate a "complete" alternative to QM if the
> phenomenon did not happen. And if you look at the actual evidence, maybe it
> doesn't!
Well, what if the hidden variables are nonlocal?
> For an authoratitive account of the present state of play re entanglement,
> see Prof Laloe's article:
> Franck Laloë, 'Do we really understand quantum mechanics? Strange
> correlations, paradoxes and theorems', American Journal of Physics, 69(6)
> pp 655-701, June 2001.
Thanks for the reference.
> Although Laloe expresses the belief that QM is correct, he admits that there
> are loopholes and says:
> "A fair summary of the situation is that no one has been able to disprove
> quantum mechanics".
I believe that Einstein admitted that the theory was correct, but
incomplete, in that not all "elements of reality" were represented.
Christopher Simmons
> The analysis of the situation in EPR is hard to understand. It is
> extremely difficult to think about the implications of Leibniz' argument
> that all measurement is relative ...[snip]
I'm sure you can remove this seeming nonlocal behaviour by thinking
that measurements (and, thereby, all interactions) are relative. For
instance,
if two electons A & B were entangled so as to have opposite spin (if assume
spin has values logical 0 or 1), then instead of having
(A & !B) | (!A & B)
(ie a superposition of up/down and down/up)
we instead have the relative state
A = B+1 (mod 2)
So that A and B have no 'actual' value, only 'relative' values; relative to
one another.
I would argue that, in this model, no information is instantaneously
transmitted upon
measuring A. It simply follows immediately from the (relative = equational)
information
about the states of A and B.
Then whatever we decided to label A as being (up/down), then B would be
different.
Chris Simmons.
zirkus
> I know a little about the basic ideas of noncommutative differential
> geometry. But what exactly is it one does in stochastic differential
> geometry?
I don't know exactly what is done and there are different approaches.
See, for example, the book "Ordinary and stochastic differential
geometry as a tool for mathematical physics":
http://www.wkap.nl/prod/b/0-7923-4154-6
There are also various other researchers/papers on the topic such as:
p.ki...@ic.ac.uk
> In article <9qjo3a...@delillo.lsr.ph.ic.ac.uk>, p.ki...@ic.ac.uk
> writes:
>>Charles Francis <cha...@clef.demon.co.uk> wrote:
[unnecessary quoted text deleted]
>>My argument has nothing to do with any unresolved issues concerning the
>>fundamental nature of space-time, etc.
> I'm afraid it does, because you took a definite stance against a
> statement which only makes sense in the context of issues concerning the
> fundamental nature of space-time.
For me, the only fundamental-issues are along the lines of "what is the
model", and "how good is it", and "where does it break" (ie give wrong
answers). In the QED model, there are situations where a photon can be
ascribed a direction without having to measure what it's direction is.
These are not circumstances which need to break the applicability of QED.
Now, if I make a statement about photons, then I always will do it within the
context of some physical model, i.e. QED. In this I could define some
definition of point-like to which I could get almost all my colleagues
to agree on (at least in a for-the-sake-of-argument sense), and work
with that. This will most likely avoid any "unresolved issues concerning
the fundamental nature of space-time", because we will be interested
in making practical distinctions between point-like and non-point-like.
Disagreements about which chosen "point-like" is better, at least on
my side, will revolve around the results achieved from the chosen definition.
OK, so what is the point here? Given your fundamental-issues bias, and
my practical-distinctions one, we will continue debating past each other
and use a lot of time for no real progress: so I'm going to stop here and
"agree to disagree".
Eric Dennis
> Please don't make claims which aren't true. The only hidden variables
> which can even potentially explain anything is the Bohm theory, and that
> is really just a convolution - as physical interpretation it leaves more
> to be explained than it explains,
Just the opposite. Bohm's theory explains what actually goes on
between a system and a measuring aparatus when a measurement is made
and, in particular, how this results in the system being effectively
projected into an eigenstate of the operator "being measured". Thus
the projection postulate, with its strange distinction between
observer and observed, is obviated.
Moreover, Bohm's theory PREDICTS measurement probabilities going as
|psi|^2, rather than just postulating it. And it does so by including
normal point-particles into QM in a truly simple way, in fact maybe
even the most naive way possible.
I'm not sure how anyone could justify describing this as a
"convolution".
> and moreover [Bohm's theory] has no relativistic
> form.
There are relativistic versions of Bohm's theory. See John Bell's
"Beables for quantum field theory" in _Speakable and Unspeakable in
QM_.
Kevin A. Scaldeferri
QM was not superseded by QFT. QFT is a quantum mechanical theory (or,
class of theories). Relativistic QFT superseded non-relativistic
quantum mechanics because relativity is correct (so far as we know)
and the non-relativistic limit is just that.
--
======================================================================
Kevin Scaldeferri Calif. Institute of Technology
The INTJ's Prayer:
Lord keep me open to others' ideas, WRONG though they may be.
Hans Aberg
<cha...@clef.demon.co.uk> wrote:
>>So you claim that if somebody is making lousy measurements of the Earth's
>>orbit, then it is going to be pretty unstable?
>By definition, if someone is making lousy measurements they are not
>making measurements according to the defining methodology.
All measurements are more or less lousy: Even a highly accurate
measurement will make the Earth's orbit highly unstable by your
definition.
> The really surprising thing is that so much real physics
>actually depends on tautology.
I thought a tautology was a logical equivalence. If one already has a
logical equivalence, why bother about verifying it physically?
>>but that is not the same thing as saying
>>that the measurable physical quantity is the same thing as the particular
>>measurement technique: Often there are several different measurement
>>techniques to measure a quantity.
>Perhaps you could check through the internationally agreed definitions
>to see that all physical processes, from the measurement of time
>onwards, are defined, directly or indirectly, by specific measurements.
I thought that these conventions defines measurement units (second, meter,
etc), but not the measured quantities themselves (time, distance).
>When we use another measurement technique we must ensure it is
>calibrated to give the same result as the defining measurement process,
>so that logically the different measurement technique is merely an
>indirect way of measuring the quantity defined by some measurement
>process described in the internationally agreed standards.
You get into trouble with QM here, as the measurement disturbs the
measured quantity. Thus there is no perfect way to calibrate the way you
indicate, due to Heisenberg uncertainty.
>What we specify is that the statement of a maximum speed of information
>remains exact irrespective of the type of information.
If such a maximum exists: You haven't checked the quantum teleportation yet.
>If you must have
>a physical definition of information,
Well, if you haven't gotten that far how to define it physically, isn't it
hard to speak about a physical theory?
>use the one from computer science,
>that it is anything which can be stored as a finite array of 0's and
>1's.
Do you mean to exclude QM, by which information are held in QM fields?
>>(For example,
>>information residing in the human brain is as far as we know today not
>>always transportable by photons. If it is, we do not have any means of
>>verifying that.)
>We don't know how to quantify what happens in the brain, so we can't be
>scientific about it.
Not yet.
> I fear that information in the brain will have to
>remain outside of the definition of information, although certainly
>quite a lot of it can be stored in computers as 0's and 1's
So, in other words, I should disregard your posts, as emanating from your
brain, they cannot be judged constitute information?
>>The problem here is your theory of "maximum speed of information" is
>>invalid until verified.
>How do you verify something which is actually true by tautology?
A tautology is verified by logical reasoning. A physical theory is
"verified" by considering predictions and experiments, even though it is
never possible to cover up the full generality of the theory. Thus, one
must expect to revise existing physical theories, which is of course
constantly being done.
>>But what says that if say quantum teleportation is possible, that there is
>>a maximum speed of information. Or if it is, such an upper limit of it
>>agrees with the upper limit of the speed of photons?
>That is actually one of the main reasons why we have to use an abstract
>concept like the speed of information, rather than a physical speed of
>an object. Quantum teleportation is possible, and there is no upper
>limit to the speed of individual photons. In relativistic quantum field
>theory the probability amplitude for particle creation and annihilation
>outside the light cone is non zero, so it can happen that a photon moves
>faster than light. The relativistic condition (to do with the
>commutator) actually says no observable effect may travel faster than c,
>not that nothing may travel faster than c.
The quantum fields are not observables, and it is not the propagation of
quantum fields, but the collapse of the quantum field one considers. Which
QM axiom do you have in your mind when setting the limit to c in your
reasoning here? How is this derived from say QFT?
>>> But as I say, the mathematical
>>>form of relativity would be the same even if they did not. Which is why
>>>it is strictly more accurate to call c the maximum speed of information
>>>than it is to call it the speed of light.
>>No. This is why one should say that c is the supremum of the photon speed
>>if one means that.
>As mentioned above there is no such thing.
What's this: The supremum speed of the photon is c even if it has small
mass. If quantum teleportation can transport the photon faster than c, and
I want to see some ref's for experiments verifying this before I believe
it, then that is an effect similar to the slowdown of light in media,
where the light appears to move slower, but the photons still move at
speed c. Thus, the QM teleportation would make it appear that the light
moves faster than c, but the photons still move at speed c.
>Again this is why it is
>dangerous to be too dependent on specific assumed properties of matter
>when making one's definitions.
Well, if you do not define your theory in terms of physical quantities
(matter energy, whatever), then it is not a physical theory. Period.
> Matter has this awful habit of not
>behaving how you thought it was going to in your assumptions, so that
>your definitions break down when you get a bit deeper into it.
Well, the idea of a physical theory is to define the quantities that one
actually can predict, and not to do "just-in-case" definitions for future,
potential needs: If there are no accurate predictions, it it is
meaningless as a valid physical theory.
Charles Francis
writes:
>In article <GqpLM...@research.att.com>,
>Peter Shor <sh...@research.att.com> wrote:
>>What's probably closer to John Baez's belief (and if
>>he's reading this, he can correct me) is that any chunk of the
>>universe of volume V can be represented (modulo boundary effects) by a
>>finite-dimensional Hilbert space.
>Yeah, something like that:
So how come you always argued so strongly against this position?
Regards
--
Charles Francis
Paul Reilly
> The statement that a photon is pointlike is not just a statement about a
> photon. It is also a statement which uses the concept of a point, and a
> point is a fundamental in the description of space-time. We do not have
> to assume that space has a structure described by classical geometry, or
> that the existence of points implies the existence in its own right of a
> physical Riemannian manifold, and according to the orthodox
> interpretation of qm it is actually wrong to make such an assumption.
>
> Before saying that it is definitely wrong to call a photon pointlike you
> should try to understand what that statement might actually mean, in
> what sense it may be possible to talk of points, and you should
> recognise that any discussion on the fundamental nature of a point is
> also a discussion of the fundamental nature of space-time.
There is a good discussion by Bertrand Russell on "what is a point"
from his 1927 book "The Analysis of Matter". That argument needs to be
updated but basically, he notes that points in spacetime need to be
defined in terms of what is around them. He sees five partially
overlapping 'events' of finite size as defining a region of spacetime
smaller than any of the regions and a point as the extreme case where
you can't shrink any of the events without losing the point.
My own belief (shared by many physicists) is that there aren't
really any 'points' underlying physical spacetime. Without even
touching on string theory and the like, just consider what a point in
spacetime is: it's an idealization of a shrinking region. A region,
is, I believe, defined by the data on its boundary. If two regions
have identical data on their boundaries, they are indistinguishable as
far as the rest of the universe is concerned and should/must be
interchangeable. Data on the boundaries here includes physical fields
on null hypersurfaces or physical fields and their derivatives on non
null hypersurfaces. Take a closed region of spacetime; the only
things the rest of the universe can "know" about it are due to data
crossing the boundary of it, if locality holds. *What* is inside is
defined by looking at that data towards the inside, and *where/when*
the region is defined by looking at that same data from the inside
looking out. There is a symmetry between what is in the region and
where it is.
Now, if you shrink the boundary of the region, you have less and
less data to define what is in it and where it is, if you assume some
high frequency cutoff - for example you could take the area of the
boundary divided by the Planck area and you'd have a nice expression
for the possible upper limit on the amount of data you're talking
about. As you shrink that are, the region becomes indistinguishable
from more and more other regions and they all blend into each other -
you never get down to a point, unless you consider all of spacetime to
share ONE point as a basic building block, that point defined by the
data "the empty set" on its boundary.
I can try to explain this in more detail if people are interested, I
think about it a lot, but I'll just post this and see if crackpot
detectors go off.
Paul
Charles Francis
Flanagan <sen...@yahoo.com> writes:
>Charles Francis wrote:
>> Though failing to properly cite himself, it was Brian Flanagan who wrote:
>> >Why was QM superseded by QFT?
>> I'll give you a short answer, that no one has been able to construct
>> relativistic QFT from simple non-relativistic quantum mechanics.
>Well, as Dyson says quite plainly in his old Scientific American
>article on "Field Theory", QFT can be roughly understood as relativity
>+ QM,
Roughly yes. And when they produced it I think they expected it to
possible to make it exact given time to sort it out. I still do. But it
has proved very difficult, and many physicist think impossible. So
physicists in large numbers have given up on this aim.
>so "relativistic QFT" would seem somewhat redundant.
Field theory is more general than its application to elementary particle
physics, and you actually could have a non-relativistic quantum field
theory. In fact I think we do for the study of such things as
semi-conductors, though I have no personal experience of it. Anyway,
don't knock it. There's nothing like a bit of pedantry to a pedant. -:}
[Moderator's note: yes, condensed matter physics offers plenty of
reasons to study nonrelativistic quantum field theories involving
quasiparticles like "phonons", "excitons", "spinons", "magnons"
and "holes". - jb]
>And then,
>given the successes of QFT, and the simplicity of (say) the Dirac
>equation, I'm not sure what you mean by the above statement, except
>perhaps that no perfect theory has yet been devised, which seems true
>enough.
Yes, but also that most physicists now think it impossible. As it
happens I disagree.
>> But
>> that does not mean that such a construction is impossible, just that it
>> has some quite unexpected attributes, or it may be complicated in ways
>> that are of no real benefit to physics.
>What attributes are you referencing? The infinities?
Certainly that has been a stumbling block, though I regard the
infinities as essentially solved by correcting Wick's theorem as
described by Scharf in Finite QED. There is also the Landau pole, though
that may be just a breakdown of an iterative solution, or it may be an
indication that the theory needs modification on a very small scale, and
I would not think of that as a fatal problem, since it would only show
there has been an approximation which works on the scales on which we
use qed. I was really referencing the given example. Using all these
Hilbert spaces, each defined on a synchronous slice, the theory is not
manifestly covariant. I actually think that does not matter, so long as
the laws of physics are the same in each Hilbert space. But covariance
(Lorentz covariance or manifest covariance) is generally taken as a
requirement for this construction.
>> For example I am quite convinced that there is no simple construction
>> based on a single Hilbert space. But on the other hand if one starts
>> with a C*-algebra it is possible to extract a number of Hilbert spaces
>> from it. If you had some reason to start with the correct collection of
>> Hilbert spaces it should be possible to work back to the C*-algebra. But
>> if your reason for starting with this collection of Hilbert spaces is
>> essentially philosophical it would not be the stuff of typical physics
>> journals.
>Until the middle of the 19th century, scientists were known as natural
>philosophers. ('Dawn to Decadence' Jacques Barzun) Given the fractious
>relations between contemporary (ahem) scientists and philosophers, it
>would seem problematical to reach agreement between them as to what
>constitutes an "essentially philosophical" reason for anything, much
>less one's choice of Hilbert spaces.
Yes.
> Having said that, I am reminded
>of 'Structure & Interpretation of QM' by RIG Hughes (Harvard?) which
>features a section entitled "Why Hilbert Spaces?" wrt QM; This work
>has the distinction of being both scientifically accurate and
>philosophically sound. As for typical physics journals ... who wants
>to be thought typical?
How do you get published for not being typical?
Regards
--
Charles Francis