"Information" from "nowhere"
Prof Weird
to do this or that come from ?' or 'there is no mechanism to increase
information' or 'information cannot be gained without an intelligence
behind it !'
Some recent articles on how 'information' can come from nowhere
(information, being a DESCRIPTION, not a physical entity, is not
subject to conservation laws, so nothing prevents it from appearing
from nowhere) :
"Structurally complex and highly active RNA ligases derived from
random RNA sequences", EH Ekland, JW Szostak, DP Bartel, Science 269:
364-370, 1995
Created a phage library of about 10^14 sequences; about 1 in 20
trillion (about 1 in 10^12) had activity - about 50 orders of
magnitude more often than standard misinformation theory would have us
believe.
"RNA catalyzed RNA polymerization : Accurate and general RNA-templated
primer extension", WK Johnson, PJ Unrau, MS Lawrence, MF Glasner, DP
Bartel, Science 292, 1319-1325 (18 May 2001)
Retrieved an active sequence from pool of 10^15 RANDOM RNA sequences
(started with the randomly generated RNA ligase backbone and added 70
or so random sequences to it).
"Functional proteins from a random sequence library", AD Keefe and JW
Szostak, Nature 410 : 5 April 2001
Were looking for proteins with ATP binding activity. Started with a
pool of 6 x 10^12 non-redundant random sequences, and ran 8
selection/amplification cycles (ran proteins through ATP-Sepharose
column; those that didn't stick well were washed through, while those
that bound well were retained, and made up a larger percentage of the
next population). Activity rose for 0.1% to 6.2% - so there ARE
sequences in there that bound to ATP (cannot amplify something that
does not exist).
24 clones in 4 families - unrelated to each other or to anything in
the database (apparently, there are many solutions to the 'problem' of
binding ATP).
3 mutagenic cycles (3.7%/amino acid), then 6 rounds of
selection/amplification INCREASES activity to 34%.
56/56 clones were all derived from one family. Had a specific,
invariant structure : Cys - X - X - Cys (the Zinc finger). The
researchers had overlooked the presence of zinc in their in vitro
transcription/translation system - but the phage expressing the random
protein libraries did not.
"Evolution of biological complexity", C Adami, C Ofria, TC Collier,
Proceedings of the National Academy of Sciences 97 #9, pg 4463-4468,
April 25, 2000
Self-replicating computer programs, when fitness determined only by
their code sequences, became MORE complex in a noisy system with
mutation and selection.
Useful mutations were selected for (enabled the program to get more
CPU time - the limited resource in the simulation) and retained -
generation of 'genes'. Since the variability (and uncertainty) of
some regions of the code were fixed by selection, the entropy dropped,
and so complexity/info must INCREASE by definition.
As to how a specific mutation in bacteria can 'appear' when needed :
"Genome-wide hpermutation in a sub-population of stationary-phase
cells underlies recombination-dependent adaptive mutation", JW
Torkelson, RS Harris, M-J Lombardi, N Nagendrau, C Thulin, SM
Rosenberg, EMBO J 16 (11), pg 3303-3337, June 2, 2000
http://emboj.oupjournals.org/cgi/content/full/16/11/3303
Seems like 1 in 10^5 of the cells become hypermutagenic when stressed
in stationary phase - rate goes up a few orders of magnitude (to 0.5 x
10^-2/day ), but is NOT targeted to any particular gene or DNA
(genome, plasmid, or F factor). This state is not permanent or
heritable - they revert back to standard rates as soon as they've
adapted.
Adam Marczyk
news:3b9c6c39.0112...@posting.google.com...
Would you be at all interested in turning this into a FAQ? It seems of that
caliber to me. Or POTM at least.
--
And I want to conquer the world,
give all the idiots a brand new religion,
put an end to poverty, uncleanliness and toil,
promote equality in all of my decisions...
--Bad Religion, "I Want to Conquer the World"
http://www.ebonmusings.org ICQ: 8777843
Tim Tyler
: Some recent articles on how 'information' can come from nowhere
: (information, being a DESCRIPTION, not a physical entity, is not
: subject to conservation laws, so nothing prevents it from appearing
: from nowhere)
How does being a description make something immune to conservation laws?
The reversibility of the laws of physics effectively prevents
the creation or destruction of information. If information is
created, it must also be destroyed. The destruction of information
would prevent reversibility - and has never been observed.
Information in biology does not "appear from nowhere".
It effectively comes from low-entropy sources - such as the sun.
--
__________
|im |yler Index of my domains: http://timtyler.org/ t...@iname.com
John Wilkins
> Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
>
> : Some recent articles on how 'information' can come from nowhere
> : (information, being a DESCRIPTION, not a physical entity, is not
> : subject to conservation laws, so nothing prevents it from appearing
> : from nowhere)
>
> How does being a description make something immune to conservation laws?
A description is a non-physical thing. It is an abstract series of
symbols. Abstract things like proper Turing machines are not subject to
the laws of physics, although any approximation of one must be. There is
also no direct relationship between information and energy.
Hence, information is not subject to the conservation of energy. The
distinction that matters here is between an abstract thing, which is not
temporal or spatial, and something that *instantiates* that abstract
object or property. Only objects that are spatiotemporal are subject to
conservation of energy restrictions.
Now you can describe things using a lot of energy or a little energy. In
other words (sorry about that pun) instantiating information is an
energy bound process, but the appropriate measure/s has/ve nothing to do
with thermodynamic properties of things.
>
> The reversibility of the laws of physics effectively prevents
> the creation or destruction of information. If information is
> created, it must also be destroyed. The destruction of information
> would prevent reversibility - and has never been observed.
I do not follow you here, Tim. How can the creation of information be
prevented by physical law? Physical things change conformation, bound
energy, and other properties all the time, and a relevant description of
those states must involve new information for any restricted system. If
a phsyical structure degrades, then information about that structure no
longer applies, and can be said to have been destroyed.
>
> Information in biology does not "appear from nowhere".
> It effectively comes from low-entropy sources - such as the sun.
I disagree - thermodynamic exergy comes from the sun, not information.
--
John Wilkins
Occasionally making sense for over 46 years
Howard Hershey
Prof Weird wrote:
>
> A seemingly common assault from cretos is 'where did the information
> to do this or that come from ?' or 'there is no mechanism to increase
> information' or 'information cannot be gained without an intelligence
> behind it !'
>
> Some recent articles on how 'information' can come from nowhere
> (information, being a DESCRIPTION, not a physical entity, is not
> subject to conservation laws, so nothing prevents it from appearing
> from nowhere) :
>
> "Structurally complex and highly active RNA ligases derived from
> random RNA sequences", EH Ekland, JW Szostak, DP Bartel, Science 269:
> 364-370, 1995
>
> Created a phage library of about 10^14 sequences; about 1 in 20
> trillion (about 1 in 10^12) had activity - about 50 orders of
> magnitude more often than standard misinformation theory would have us
> believe.
>
> "RNA catalyzed RNA polymerization : Accurate and general RNA-templated
> primer extension", WK Johnson, PJ Unrau, MS Lawrence, MF Glasner, DP
> Bartel, Science 292, 1319-1325 (18 May 2001)
>
> Retrieved an active sequence from pool of 10^15 RANDOM RNA sequences
> (started with the randomly generated RNA ligase backbone and added 70
> or so random sequences to it).
Remember that a mole contains 10^23 molecules. And, given the laws of
probability, every millimole will have not only the ligase activity but
also the other activities mentioned. A millimole is not a lot of
material. This *is* a crucial point in any assumptions about
abiogenesis. *If* such activities only arose by chance once in 10^50
molecules (more than the weight of the earth), then the probability of
generating a self-replicating *system* that required more than one such
activity would be vanishingly small. That the amounts needed to
generate these activities by pure chance is so (relatively) small is
consistent with a natural explanation for abiogenesis on the earth. If
large amounts were required, abiogenesis by this mechanism would be
highly unlikely.
Tim Tyler
:> : Some recent articles on how 'information' can come from nowhere
:> : (information, being a DESCRIPTION, not a physical entity, is not
:> : subject to conservation laws, so nothing prevents it from appearing
:> : from nowhere)
:>
:> How does being a description make something immune to conservation laws?
: A description is a non-physical thing. It is an abstract series of
: symbols. [...]
Those symbols are themselves subject to rules and regulations. Merely
observing that something is a description does not necessarily make it
immune to conservation laws.
: There is also no direct relationship between information and energy.
: Hence, information is not subject to the conservation of energy.
It does not follow from this that it is not conserved.
: The distinction that matters here is between an abstract thing, which
: is not temporal or spatial, and something that *instantiates* that
: abstract object or property. Only objects that are spatiotemporal are
: subject to conservation of energy restrictions.
I'd say that information is subject to conservation of information laws -
not conservation of energy ones.
:> The reversibility of the laws of physics effectively prevents
:> the creation or destruction of information. If information is
:> created, it must also be destroyed. The destruction of information
:> would prevent reversibility - and has never been observed.
: I do not follow you here, Tim. How can the creation of information be
: prevented by physical law? Physical things change conformation, bound
: energy, and other properties all the time, and a relevant description of
: those states must involve new information for any restricted system. If
: a phsyical structure degrades, then information about that structure no
: longer applies, and can be said to have been destroyed.
You are not considering a closed system. You refer to information
pertaining to a particular structure. However, this is a system subject
to inflows and outflows of information. Just as thermodynamic
conservation laws only apply directly to closed systems, so it is with
conservation of information.
I tried to explain above how information creation and destruction was
prevented by physical law. Were this not so, the universe would not
be microscopically reversible - and as far as we can tell - it is.
:> Information in biology does not "appear from nowhere".
:> It effectively comes from low-entropy sources - such as the sun.
: I disagree - thermodynamic energy comes from the sun, not information.
I would say that both thermodynamic energy and information come from the
sun.
The idea that information can come from nowhere is the hypothesis of
randomness. If "truly random" things exist, information *can* come from
nowhere. On the other hand - if the world is deterministic - you can't
get new information from anywhere - all you can do is shift it around.
If you are having difficulty understanding my position, you might
like to consider that I enjoy entertaining the finite nature hypothesis -
in which the universe is fundamentally informational. Information is
"spatiotemporal". Only spatiotemporal things can be said to exist at all.
One of my links pages about the subject: http://finitenature.com/links/
Adam Marczyk
> Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
>
> : Some recent articles on how 'information' can come from nowhere
> : (information, being a DESCRIPTION, not a physical entity, is not
> : subject to conservation laws, so nothing prevents it from appearing
> : from nowhere)
>
> How does being a description make something immune to conservation laws?
>
> The reversibility of the laws of physics effectively prevents
> the creation or destruction of information. If information is
> created, it must also be destroyed. The destruction of information
> would prevent reversibility - and has never been observed.
What about black holes?
[snip]
Jon Fleming
>John Wilkins <john.w...@bigpond.com> wrote:
>: Tim Tyler <t...@iname.com> wrote:
>:> Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
>
>:> : Some recent articles on how 'information' can come from nowhere
>:> : (information, being a DESCRIPTION, not a physical entity, is not
>:> : subject to conservation laws, so nothing prevents it from appearing
>:> : from nowhere)
>:>
>:> How does being a description make something immune to conservation laws?
>
>: A description is a non-physical thing. It is an abstract series of
>: symbols. [...]
>
>Those symbols are themselves subject to rules and regulations. Merely
>observing that something is a description does not necessarily make it
>immune to conservation laws.
>
>: There is also no direct relationship between information and energy.
>
>: Hence, information is not subject to the conservation of energy.
>
>It does not follow from this that it is not conserved.
>
>: The distinction that matters here is between an abstract thing, which
>: is not temporal or spatial, and something that *instantiates* that
>: abstract object or property. Only objects that are spatiotemporal are
>: subject to conservation of energy restrictions.
>
>I'd say that information is subject to conservation of information laws -
>not conservation of energy ones.
However, I haven't seen you or anyone state these conservation of
information laws, or provide an argument that establishes that they
actually exist.
OK, then since QM has to all intents and purposes proved that the
universe _is_ random, then there's no question; information can be
created.
dkomo
>
> I'd say that information is subject to conservation of information laws -
> not conservation of energy ones.
>
A compressed file is sent down a transmission line. Along the way a
burst of thermal noise occurs in the middle of the file, scrambling
many of the bits. When the file is received at the other end of the
line, it is too garbled to be decompressed. Effectively the whole
file is lost.
Please explain how information is conserved in this process.
Prof Weird
Sounds like a plan to me - what needs to be done for that ?
Dunk
<ebon...@hotmailNOTexcite.com> wrote:
<snip>
>Would you be at all interested in turning this into a FAQ?
Nice idea, but 'turn into' = a project.
Dunk
R. Baldwin
> John Wilkins <john.w...@bigpond.com> wrote:
> : Tim Tyler <t...@iname.com> wrote:
> :> Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
>
> :> : Some recent articles on how 'information' can come from nowhere
> :> : (information, being a DESCRIPTION, not a physical entity, is
not
> :> : subject to conservation laws, so nothing prevents it from
appearing
> :> : from nowhere)
> :>
> :> How does being a description make something immune to
conservation laws?
>
> : A description is a non-physical thing. It is an abstract series of
> : symbols. [...]
>
> Those symbols are themselves subject to rules and regulations.
Merely
> observing that something is a description does not necessarily make
it
> immune to conservation laws.
>
> : There is also no direct relationship between information and
energy.
>
> : Hence, information is not subject to the conservation of energy.
>
> It does not follow from this that it is not conserved.
The burden is on the advocates of this new law, "Conservation of
Information," to demonstrate how information is always conserved. Why
should we take your word for it?
>
> : The distinction that matters here is between an abstract thing,
which
> : is not temporal or spatial, and something that *instantiates* that
> : abstract object or property. Only objects that are spatiotemporal
are
> : subject to conservation of energy restrictions.
>
> I'd say that information is subject to conservation of information
laws -
> not conservation of energy ones.
A sample from the web site you reference herein, specifically page
http://digitalphilosophy.org/finite_nature.htm:
"Despite not having a good definition of what information is, we
postulate a new law; Conservation of Information. In this context
information is the stuff that enables a reversible system to, if time
were reversed, proceed backwards along the exact path that it took
going forwards. All reversible systems have a conserved quantity;
log2(n), where n is the number of different states that the system
progresses through in one cycle. (All finite state reversible systems
evolve along a closed cycle.) This number is constant, independent of
where the system is on that cycle and is equal to the quantity of
information conserved."
So this law is postulated without a good definition? It seems you and
your colleagues have much more work to do before putting this idea
forward. Perhaps you should start by defining "information." If you
want others to understand you, the commonly understood meaning of
"information" should be included in the definition.
The Principle of Conservation of Energy is a generalization of
experience, so far not contradicted by observation of nature. It has
years of research backing it up. What research provides the basis of
the "Law of Conservation of Information?" Show your work.
Consider the Harry Potter books burnt by Jack Brock's fundamentalist
church in New Mexico recently. By our common understanding of what
"information" means, some information was destroyed by this process.
For a more extreme example, suppose an artist creates a new original
work of art using lead as a medium. The artist works in a closed room
with no windows. Dissatisfied with the result, the artist tosses the
art in a melting pot. The artist then dies of a heart attack, losing
all memory of the creation. The electromagnetic energy that reflected
off the artwork while it existed is absorbed by the walls of the room.
How is information conserved?
Any change in molecular or atomic "information" caused by the
oxidation reduction of the paper had nothing to do with the shape of
the ink symbols on the page. The substrate holding the symbol pattern
is destroyed, so the symbol pattern is destroyed and the information
with it.
>
> :> Information in biology does not "appear from nowhere".
> :> It effectively comes from low-entropy sources - such as the sun.
>
> : I disagree - thermodynamic energy comes from the sun, not
information.
>
> I would say that both thermodynamic energy and information come from
the
> sun.
>
> The idea that information can come from nowhere is the hypothesis of
> randomness. If "truly random" things exist, information *can* come
from
> nowhere. On the other hand - if the world is deterministic - you
can't
> get new information from anywhere - all you can do is shift it
around.
>
How can you possibly avoid non-determinism? It requires more than one
atom to form the computational elements that model the behavior of an
atom. This makes it impossible to model everything.
dkomo
I'd like to see this also, but suggest targeting the FAQ to a general
audience. That means cutting down on the jargon (e.g.
"stationary-phase cells", "56/56 clones", "Cys - X - X - Cys (the Zinc
finger)", "8 selection/amplification cycles", "highly active RNA
ligases").
We're not all molecular biologists here. Also, if the cretinists are
to understand it, it needs to be at a 9th grade reading comprehension
level.
John Wilkins
with, so I'll confine myself to a few.
Tim Tyler <t...@iname.com> wrote:
> John Wilkins <john.w...@bigpond.com> wrote:
> : Tim Tyler <t...@iname.com> wrote:
> :> Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
>
> :> : Some recent articles on how 'information' can come from nowhere
> :> : (information, being a DESCRIPTION, not a physical entity, is not
> :> : subject to conservation laws, so nothing prevents it from appearing
> :> : from nowhere)
> :>
> :> How does being a description make something immune to conservation laws?
>
> : A description is a non-physical thing. It is an abstract series of
> : symbols. [...]
>
> Those symbols are themselves subject to rules and regulations. Merely
> observing that something is a description does not necessarily make it
> immune to conservation laws.
We seem to be lacking in some definitions here. Information has a cost
to propagate, and so it is not "immune" to conservation of energy laws.
But the laws of conservation of energy are not laws of information, and
information has no conservation law. I am using the term "information"
(or, as I so often mistype it, ifnormation, which somehow seems
appropriate) to mean Shannon-Weaver information or Kolmogorov-Chaitin
Algorithmic Information.
>
> : There is also no direct relationship between information and energy.
>
> : Hence, information is not subject to the conservation of energy.
>
> It does not follow from this that it is not conserved.
No, but it does follow that it is not subject to thermodynamic
conservation laws.
>
> : The distinction that matters here is between an abstract thing, which
> : is not temporal or spatial, and something that *instantiates* that
> : abstract object or property. Only objects that are spatiotemporal are
> : subject to conservation of energy restrictions.
>
> I'd say that information is subject to conservation of information laws -
> not conservation of energy ones.
OK, now about these putative laws of the conservation of information -
what are they? I have never come across a derivation from axioms, or a
theorem, or a theoretical demonstration that information is preserved.
>
> :> The reversibility of the laws of physics effectively prevents
> :> the creation or destruction of information. If information is
> :> created, it must also be destroyed. The destruction of information
> :> would prevent reversibility - and has never been observed.
>
> : I do not follow you here, Tim. How can the creation of information be
> : prevented by physical law? Physical things change conformation, bound
> : energy, and other properties all the time, and a relevant description of
> : those states must involve new information for any restricted system. If
> : a phsyical structure degrades, then information about that structure no
> : longer applies, and can be said to have been destroyed.
>
> You are not considering a closed system. You refer to information
> pertaining to a particular structure. However, this is a system subject
> to inflows and outflows of information. Just as thermodynamic
> conservation laws only apply directly to closed systems, so it is with
> conservation of information.
>
> I tried to explain above how information creation and destruction was
> prevented by physical law. Were this not so, the universe would not
> be microscopically reversible - and as far as we can tell - it is.
Why does the destruction or creation of information make physics
reversible? I thought *that* (irreversibility) was due to the 2LoT.
>
> :> Information in biology does not "appear from nowhere".
> :> It effectively comes from low-entropy sources - such as the sun.
>
> : I disagree - thermodynamic energy comes from the sun, not information.
>
> I would say that both thermodynamic energy and information come from the
> sun.
I know - I just do not understand why.
>
> The idea that information can come from nowhere is the hypothesis of
> randomness. If "truly random" things exist, information *can* come from
> nowhere. On the other hand - if the world is deterministic - you can't
> get new information from anywhere - all you can do is shift it around.
>
> If you are having difficulty understanding my position, you might
> like to consider that I enjoy entertaining the finite nature hypothesis -
> in which the universe is fundamentally informational. Information is
> "spatiotemporal". Only spatiotemporal things can be said to exist at all.
> One of my links pages about the subject: http://finitenature.com/links/
Well now we pass from information theory and physics to [gasp]
metaphysics. But I suppose we were there all along. Let me expound my
position.
I know of Wheeler's "it from bit" idea that physics is just information,
and I reject it. In fact, in my opinion, we derive "bit" from "it" by
abstraction. Abstract things exist, so far as they do (I am a bit of a
nominalist, and do not admit things other than spatiotemporal things
exist in the proper sense of "exist") outside time and space. At best
they can be instantiated in physical things. Information is not a causal
property - thermodynamic entropy is. So far as I admit, information is
an abstract property of abstract objects (symbols), and it is *not*
spatiotemporal. Wheeler's slogan is a cosmologist's conceit.
To demonstrate that information is lost, let me use an example of Elliot
Sober's - imagine you have a near-perfectly smooth spherical section
laying flat on a horizontal surface. Around the edge of the section is a
lip. At the bottom of the section is a marble. You may know that the
trajectory of the marble began at some point on the lip, but all
information as to what that trajectory was is now lost. You cannot
reconstruct the past perfectly, and as time progresses, less information
remains.
Now this sense of information is slightly different - here I am
referring to Fisher information, or exactness of measurement, but the
point remains for the other senses: a symbolic representation of the
entire system at t will not give us that trajectory, so Shannon
information is lost. And as time increases since the event, an
increasing number of possible trajectories could deliver that resting
position, and so AIT information is lost, because the number of
algorithms that could deliver the result in a model of the system
increases and so Omega is increasing.
At this point I would like you to give me the precis of your site - I'm
at home and I do not intend to use my limited ISP time to research
something that really ought to be read in a book, but I'd like to
understand the basic argument, if you would oblige. Who knows - perhaps
you'll convert me :-)
Steve Schaffner
> I tried to explain above how information creation and destruction was
> prevented by physical law. Were this not so, the universe would not
> be microscopically reversible - and as far as we can tell - it is.
Actually, as far as we can tell, the universe is _not_ microscopically
reversible. This conclusion is a consequence of the CPT theorem and
the observed violation of CP invariance.
--
Steve Schaffner s...@genome.wi.mit.edu
Immediate assurance is an excellent sign of probable lack of
insight into the topic. Josiah Royce
Dunk
Weird) wrote:
>> Would you be at all interested in turning this into a FAQ? It seems of that
>> caliber to me. Or POTM at least.
>
>Sounds like a plan to me - what needs to be done for that ?
A lot. Flesh it out so that a fairly general audience can follow it -
an introduction is needed to explain the point, and why someone might
be interested in the first place - yet you can't get by with analogies
alone, nor say anything that a specialist would wince at, nor can
there be any phrase that gives certain types an easy quibble.
No long sentences like the above :)
Standard sentence structure, well made paragraphs.
No funny words like ' cretos'.
And put it up for 'peer review' here, then again with a revised
edition (there are bound to be suggestions, some of them even cogent
:)
Takes a while, but it's worth while. You, too will understand the
whole business better when you're done.
Are you up to it?
Dunk
ps examples
http://www.talkorigins.org/faqs/moonrec.html
http://www.talkorigins.org/faqs/mutations.html
http://www.talkorigins.org/faqs/snrfab.html
and dozens more
Bobby D. Bryant
> I'd say that information is subject to conservation of information
> laws - not conservation of energy ones.
I toyed with the idea of a law of conservation of information for a
while, but thinking about various problems led me to reject the idea.
For instance, suppose your "information" is a string of bits.
o In what sense is information being "conserved" if you make a million
copies of that string?
o In what sense if information being "conserved" if you overwrite the
string with a random pattern of bits?
o If you encrypt the string with a one-time pad, where does the
"information" go? A naive answer would be that 'half' the information
is in the pad and 'half' is in the cyphertext, since you must have both
in order to reconstitute the original string. However, despite the
conventional name, a "one-time" pad can be used to encrypt a very large
number of message strings. But if 'half' the information goes into the
pad during the encryption process, its "information content" grows with
use -- even though the pad does not actually change! The same problem
arises if you assume *any* amount of the information is stored in the
pad, no matter how small a fraction of the "total information" you
specify. So if conserved, then the "information" must all go into the
cyphertext. But if you don't have the pad then the cyphertext is as
useless as a string of random bits. If you can't distinguish
cyphertext from a string of random bits, does it actually bear any
"information" in any meaningful sense? You can't get the "information"
out of it without the pad, but per above, the pad didn't inherit any of
the "information" when the encryption was done. So where did the
missing "information" go, if it is actually conserved?
I think there is a problem when you start thinking of information as a
peer of matter and energy, which seem to work fine in the absence of an
observer.
Bobby Bryant
Austin, Texas
Kim G. S. OEyhus
>
>You are not considering a closed system. You refer to information
>pertaining to a particular structure. However, this is a system subject
>to inflows and outflows of information. Just as thermodynamic
>conservation laws only apply directly to closed systems, so it is with
>conservation of information.
>
>I tried to explain above how information creation and destruction was
>prevented by physical law. Were this not so, the universe would not
>be microscopically reversible - and as far as we can tell - it is.
It seems you are using "entropy" and "information" as synonyms, which
they are only in some thermodynamic jargon. In more normal language,
"information" is data that can be used to predict other data. F.eks. a
TV program listing, which predicts the TV programs.
Thus, evolutions is a process that makes information in the genes of
how to survive.
John Bode
> Tim Tyler <t...@iname.com> wrote:
>
> > Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
> >
> > : Some recent articles on how 'information' can come from nowhere
> > : (information, being a DESCRIPTION, not a physical entity, is not
> > : subject to conservation laws, so nothing prevents it from appearing
> > : from nowhere)
> >
> > How does being a description make something immune to conservation laws?
>
> A description is a non-physical thing. It is an abstract series of
> symbols. Abstract things like proper Turing machines are not subject to
> the laws of physics, although any approximation of one must be. There is
> also no direct relationship between information and energy.
>
> Hence, information is not subject to the conservation of energy. The
> distinction that matters here is between an abstract thing, which is not
> temporal or spatial, and something that *instantiates* that abstract
> object or property. Only objects that are spatiotemporal are subject to
> conservation of energy restrictions.
>
I just had a couple of thoughts about this. Forgive me if they sound
a bit fuzzy; this is just off the top of my head.
I had never thought about the distinction between the abstract
property "information" and an instance of information (such as this
message); now that I see it, it makes perfect sense. It seems to me
that this same dichotomy can apply to matter and energy as well.
There is the abstract property "energy," which we define as the
ability to do work. There is also the *instance* of energy, such as
the energy bound up in the Sun. Similarly, there is the abstract
property of matter which we define mathematically, and then there is
the instance of matter, such as the Sun itself.
We cannot create new *instances* of matter or energy, but we can
create new *instances* of information at will (such as I am doing
right now, if this counts as information ;-). Conservation laws apply
to the instances, not the concepts, in which case information is not
and cannot be conserved.
This also leads me to think about the idea of information being an
intrinsic property of something. It's sort of like the old "if a tree
falls an no one's around to hear it, does it make a sound" question.
Yes, it creates disturbances in the air that propogate, but for us it
isn't "sound" until those disturbances have been converted to
electrochemical signals in our brains. In a sense, "sound" doesn't
exist unless something is present to perceive it as such.
Similarly, does information exist independent of our (or any other
organism's) perception of it? For example, we consider an atom's
position to be "information", but is that only because we are
interested in where the atom is? Does the atom itself contain
information, or do we create information independently of the atom?
Like I said, it's fuzzy. I haven't thought about this in great
detail, and I haven't studied information theory beyond what was
necessary to learn how Huffman compression works.
Tim Tyler
: Tim Tyler <t...@iname.com> wrote in message news:Gp9Ao...@bath.ac.uk...
:> Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
:> The reversibility of the laws of physics effectively prevents
:> the creation or destruction of information. If information is
:> created, it must also be destroyed. The destruction of information
:> would prevent reversibility - and has never been observed.
: What about black holes?
I would say not. Putting an object somewhere where you can't get at it
is not necessarily the same thing as destroying it.
Tim Tyler
: How can you possibly avoid non-determinism? It requires more than one
: atom to form the computational elements that model the behavior of an
: atom. This makes it impossible to model everything.
That is not what "determinism" means. It implies that the future of the
system is foreordained - not that entities embedded within the system can
predict that future in its entirety.
Tim Tyler
: Tim Tyler <t...@iname.com> writes:
:> I tried to explain above how information creation and destruction was
:> prevented by physical law. Were this not so, the universe would not
:> be microscopically reversible - and as far as we can tell - it is.
: Actually, as far as we can tell, the universe is _not_ microscopically
: reversible.
That is not correct - as far as is known. All our current best laws of
physics are symmetrical under a transformation involving reversing the
direction of time.
: This conclusion is a consequence of the CPT theorem and
: the observed violation of CP invariance.
That just means that running time backwards doesn't work.
Reversing time, parity and charge is the correct operation.
Tim Tyler
: On 1 Jan 2002 11:54:52 -0500, Tim Tyler <t...@iname.com> wrote:
:>John Wilkins <john.w...@bigpond.com> wrote:
:>: Tim Tyler <t...@iname.com> wrote:
:>The idea that information can come from nowhere is the hypothesis of
:>randomness. If "truly random" things exist, information *can* come from
:>nowhere. On the other hand - if the world is deterministic - you can't
:>get new information from anywhere - all you can do is shift it around.
: OK, then since QM has to all intents and purposes proved that the
: universe _is_ random, then there's no question; information can be
: created.
QM has /not/ proved "to all intents and purposes proved that the
universe is random".
Rather the equations of QM suggest that the laws of physics are
symmetrical with respect to an operation which involves inverting time -
which strongly suggests that *no* information is created - and that the
hypothesis of randomness is not correct.
Conventional interpretations of quantum physics that are deterministic
exist. E.g. http://www.hedweb.com/everett/everett.htm#deterministic
``THE EVERETT FAQ
``Q13 Is many-worlds a deterministic theory?
Yes, many-worlds is a deterministic theory[...]''
You may be thinking of the HUP - which says nothing about determinism.
Tim Tyler
: A compressed file is sent down a transmission line. Along the way a
: burst of thermal noise occurs in the middle of the file, scrambling
: many of the bits. When the file is received at the other end of the
: line, it is too garbled to be decompressed. Effectively the whole
: file is lost.
: Please explain how information is conserved in this process.
Such an operation does not demonstrate that the laws of physics do not
conserve information, or are not invertible.
What they demonstrate is that the universe started off in an ordered state
and is becoming disordered.
That phenomenon happens equally well in reversible universes.
E.g. Run the reversible gas simulation at:
http://hex.org.uk/diffusion forwards.
You will see an ordered state disintegrating into a chaotic mess.
Then hit the "reverse button". You will see that nothing was actually
lost - it merely appeared so.
So with your file.
Tim Tyler
: On 1 Jan 2002 11:54:52 -0500, Tim Tyler <t...@iname.com> wrote:
:>John Wilkins <john.w...@bigpond.com> wrote:
:>: Tim Tyler <t...@iname.com> wrote:
:>:> Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
:>:> : Some recent articles on how 'information' can come from nowhere
:>:> : (information, being a DESCRIPTION, not a physical entity, is not
:>:> : subject to conservation laws, so nothing prevents it from appearing
:>:> : from nowhere)
:>
:>I'd say that information is subject to conservation of information laws -
: However, I haven't seen you or anyone state these conservation of
: information laws, or provide an argument that establishes that they
: actually exist.
Someone else has quoted Fredkin stating them on:
http://digitalphilosophy.org/finite_nature.htm
We do not know whether they are correct or not for certain. However, they
have certainly not been rejected - and until they have been, asserting
that "information comes from nowhere" may prove to be mistaken.
Tim Tyler
: "Tim Tyler" <t...@iname.com> wrote in message
:> It does not follow from this that it is not conserved.
: The burden is on the advocates of this new law, "Conservation of
: Information," to demonstrate how information is always conserved. Why
: should we take your word for it?
You should not. However, I don't see why you think the burden lies with
one party.
There are two rival hypothesis in the area: randomness and determinism.
If the latter theory is true, to speak of information being created or
destroyed is suspect.
Currently, nobody knows which theory is true.
Those who advocate randomness have not proved their thesis - and nor
have those who thing the world is predictable. I don't see why the
burden of proof should lie with one side and not the other.
I am objecting to describing information as being created, since this
assumes the "randomness" hypothesis - which is not proven.
: A sample from the web site you reference herein, specifically page
: http://digitalphilosophy.org/finite_nature.htm:
: "Despite not having a good definition of what information is, we
: postulate a new law; Conservation of Information. In this context
: information is the stuff that enables a reversible system to, if time
: were reversed, proceed backwards along the exact path that it took
: going forwards. All reversible systems have a conserved quantity;
: log2(n), where n is the number of different states that the system
: progresses through in one cycle. (All finite state reversible systems
: evolve along a closed cycle.) This number is constant, independent of
: where the system is on that cycle and is equal to the quantity of
: information conserved."
: So this law is postulated without a good definition? It seems you and
: your colleagues have much more work to do before putting this idea
: forward.
Definitely not. That finite nature paper is ten years old. The
ideas are too important to wait until all the is have been dotted
and all the ts crossed.
Fredkin offered an informal definition:
``In this context information is the stuff that enables a reversible
system to, if time were reversed, proceed backwards along the exact
path that it took going forwards''
...and then a quantitative one:
``All reversible systems have a conserved quantity; log2(n), where n is
the number of different states that the system progresses through in
one cycle. (All finite state reversible systems evolve along a closed
cycle.) This number is constant, independent of where the system is on
that cycle and is equal to the quantity of information conserved.''
Is this unclear?
: Perhaps you should start by defining "information." If you
: want others to understand you, the commonly understood meaning of
: "information" should be included in the definition.
There is more than one meaning of "information" - e.g. see Shannon's
information, and Chaitin's algorithmic information.
: The Principle of Conservation of Energy is a generalization of
: experience, so far not contradicted by observation of nature. It has
: years of research backing it up. What research provides the basis of
: the "Law of Conservation of Information?" Show your work.
It is based on an unproven hypothesis. We do not know for certain whether
the law exists or not. The fact that the laws of physics are symmetrical
under an operation involving inverting time strongly suggests that no
information is created or destroyed as time passes.
My objection is that asserting that information is created *assumes* that
no such law exists - which has not been demonstrated.
: Consider the Harry Potter books burnt by Jack Brock's fundamentalist
: church in New Mexico recently. By our common understanding of what
: "information" means, some information was destroyed by this process.
Not really. If all the information necessary to recover the original
state is present, nothing has been destroyed. It merely /seems/ that way.
The apparent "destruction" is an illusion caused by the practical
difficulties in recovering the original state. However that doesn't mean
that recovering the original state is /impossible/, only that we find it
hard to do.
: For a more extreme example, suppose an artist creates a new original
: work of art using lead as a medium. The artist works in a closed room
: with no windows. Dissatisfied with the result, the artist tosses the
: art in a melting pot. The artist then dies of a heart attack, losing
: all memory of the creation. The electromagnetic energy that reflected
: off the artwork while it existed is absorbed by the walls of the room.
: How is information conserved?
You can make the example as complex as you like - but that doesn't alter
the reversibility of the laws of physics as we understand them.
If physics is reversible, information is conserved - if the laws were run
backwards you would recover the original state.
: Any change in molecular or atomic "information" caused by the
: oxidation reduction of the paper had nothing to do with the shape of
: the ink symbols on the page. The substrate holding the symbol pattern
: is destroyed, so the symbol pattern is destroyed and the information
: with it.
Not if the world is reversible. Physicists mathematical theories of the
world are reversible. As far as we know the world is reversible.
Tim Tyler
:> I tried to explain above how information creation and destruction was
:> prevented by physical law. Were this not so, the universe would not
:> be microscopically reversible - and as far as we can tell - it is.
: Why does the destruction or creation of information make physics
: reversible?
It doesn't. If it happened, it would make the laws of physics
*irreversible*.
: I thought *that* (irreversibility) was due to the 2LoT.
Yes, it is, which is a statistical law.
2LoT analogues are common in reversible systems - though in the long run
they are violated as much as they are obeyed in such environments.
:> The idea that information can come from nowhere is the hypothesis of
:> randomness. If "truly random" things exist, information *can* come from
:> nowhere. On the other hand - if the world is deterministic - you can't
:> get new information from anywhere - all you can do is shift it around.
:>
:> If you are having difficulty understanding my position, you might
:> like to consider that I enjoy entertaining the finite nature hypothesis -
:> in which the universe is fundamentally informational. Information is
:> "spatiotemporal". Only spatiotemporal things can be said to exist at all.
:> One of my links pages about the subject: http://finitenature.com/links/
: Well now we pass from information theory and physics to [gasp]
: metaphysics. But I suppose we were there all along.
I generally consider the finite nature hypothesis and digital physics
to be physics - rather than metaphysics. There are related experiments,
and it can be confirmed or rejected. Anyway...
: Let me expound my position.
: I know of Wheeler's "it from bit" idea that physics is just information,
: and I reject it. In fact, in my opinion, we derive "bit" from "it" by
: abstraction. Abstract things exist, so far as they do (I am a bit of a
: nominalist, and do not admit things other than spatiotemporal things
: exist in the proper sense of "exist") outside time and space. At best
: they can be instantiated in physical things. Information is not a causal
: property - thermodynamic entropy is. So far as I admit, information is
: an abstract property of abstract objects (symbols), and it is *not*
: spatiotemporal. Wheeler's slogan is a cosmologist's conceit.
It doesn't sound like we are destined to agree, then.
IMO, nobody can reject the idea that "physics is just information" at the
moment - except by making a leap of faith.
: To demonstrate that information is lost, let me use an example of Elliot
: Sober's - imagine you have a near-perfectly smooth spherical section
: laying flat on a horizontal surface. Around the edge of the section is a
: lip. At the bottom of the section is a marble. You may know that the
: trajectory of the marble began at some point on the lip, but all
: information as to what that trajectory was is now lost. You cannot
: reconstruct the past perfectly, and as time progresses, less information
: remains.
This experiment does not demonstrate what you appear to think it does :-(
: Now this sense of information is slightly different - here I am
: referring to Fisher information, or exactness of measurement, but the
: point remains for the other senses: a symbolic representation of the
: entire system at t will not give us that trajectory, so Shannon
: information is lost.
No - doesn't follow. You are not measuring the state of the system
properly. If you can't measure it exactly all bets about information are
off - since you don't know if you are measuring the same thing twice.
Again, this is like measuring heat in an open system - and concluding that
energy is not conserved, since your early reading does not match your
later one.
: And as time increases since the event, an increasing number of possible
: trajectories could deliver that resting position [...]
This is false in a reversible system. Physicists tell us that the world
is a reversible system - as far as they can tell.
: and so AIT information is lost, because the number of algorithms that
: could deliver the result in a model of the system increases and so
: Omega is increasing.
No. The conclusion is faulty because the premise is suspect - There is
not one resting state, but many - if you consider all the gas particles in
your system. The claim that many states lead to one final state is an
rests on this unproven assumption.
: At this point I would like you to give me the precis of your site - I'm
: at home and I do not intend to use my limited ISP time to research
: something that really ought to be read in a book, but I'd like to
: understand the basic argument, if you would oblige. Who knows - perhaps
: you'll convert me :-)
Well, my web site on the subject is mostly links to other sites at this
stage.
Here's the abstract of the finite nature paper:
``A fundamental question about time, space and the inhabitants thereof is
"Are things smooth or grainy?" Some things are obviously grainy
(matter, charge, angular momentum); for other things (space, time,
momentum, energy) the answers are not clear. Finite Nature is the
assumption that, at some scale, space and time are discrete and that the
number of possible states of every finite volume of space-time is
finite. In other words Finite Nature assumes that there is no thing
that is smooth or continuous and that there are no infinitesimals. If
finite nature is true, then there are certain consequences that are
independent of the scale.''
Many who accept this as a working hypothesis also assume "locality".
The result is the theory that the world is equivalent to some cellular
automaton.
Fredkin is the originator, and probably the prime mover of the theory.
http://digitalphilosophy.org/ is his web site.
Much follows from assuming finite nature. E.g. general relativity can't
hold on large scales.
Tim Tyler
: On Tue, 01 Jan 2002 10:54:52 -0600, Tim Tyler wrote:
:> I'd say that information is subject to conservation of information
:> laws - not conservation of energy ones.
: I toyed with the idea of a law of conservation of information for a
: while, but thinking about various problems led me to reject the idea.
Your problems definitely have solutions. If they did not, nobody would
be considering digital physics.
Tim Tyler
: In article <Gp9qz...@bath.ac.uk>, Tim Tyler <t...@iname.com> wrote:
:>I tried to explain above how information creation and destruction was
:>prevented by physical law. Were this not so, the universe would not
:>be microscopically reversible - and as far as we can tell - it is.
: It seems you are using "entropy" and "information" as synonyms [...]
They have different connotations - for example entropy can be defined
in terms of heat - while information cannot.
: In more normal language, "information" is data that can be used to
: predict other data. [...]
While a bit vague, that is consonant with my usage.
Jon Fleming
OK. However, I'm not a ware of anyone asserting that information
comes from nowhere ... instead, information appears to be generated by
a process.
Until conservation of information laws are demonstrated to have
real-world applicability and correspond to the "information" under
discussion, I don't see anything to worry about.
Jon Fleming
>Jon Fleming <jo...@fleming-nospam.com> wrote:
>: On 1 Jan 2002 11:54:52 -0500, Tim Tyler <t...@iname.com> wrote:
>
>:>John Wilkins <john.w...@bigpond.com> wrote:
>:>: Tim Tyler <t...@iname.com> wrote:
>
>
>:>The idea that information can come from nowhere is the hypothesis of
>:>randomness. If "truly random" things exist, information *can* come from
>:>nowhere. On the other hand - if the world is deterministic - you can't
>:>get new information from anywhere - all you can do is shift it around.
>
>: OK, then since QM has to all intents and purposes proved that the
>: universe _is_ random, then there's no question; information can be
>: created.
>
>QM has /not/ proved "to all intents and purposes proved that the
>universe is random".
>
>Rather the equations of QM suggest that the laws of physics are
>symmetrical with respect to an operation which involves inverting time -
>which strongly suggests that *no* information is created - and that the
>hypothesis of randomness is not correct.
>
>Conventional interpretations of quantum physics that are deterministic
>exist. E.g. http://www.hedweb.com/everett/everett.htm#deterministic
>
>``THE EVERETT FAQ
>
>``Q13 Is many-worlds a deterministic theory?
> Yes, many-worlds is a deterministic theory[...]''
Well, I don't buy the many-worlds theory, but I'm not enough of an
expert to discuss it in detail.
However, even if Everett's interpretation is true, for all intents and
purposes the (this) universe is random, until someone can demonstrate
an experiment that shows it isn't.
Tim Tyler
:>:>:> Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
:>:>:> : Some recent articles on how 'information' can come from nowhere
:>:>:> : (information, being a DESCRIPTION, not a physical entity, is not
:>:>:> : subject to conservation laws, so nothing prevents it from appearing
:>:>:> : from nowhere)
: OK. However, I'm not a ware of anyone asserting that information
: comes from nowhere ... [...]
I've left such an assertion quoted above. See also the thread title.
Tim Tyler
: However, even if Everett's interpretation is true, for all intents and
: purposes the (this) universe is random, until someone can demonstrate
: an experiment that shows it isn't.
It depends on what you mean by random. HUP prevents embedded observeres
accessing the information required to predict it - so if you equate
randomness with unpredictability by embedded observers, we have no
disagreement.
TomS
stated..."
A question about "conservation of information":
If information is conserved, is it possible, by Noether's
Theorem, to state a corresponding symmetry principle?
Tom S.
Jonathan Stone
>dkomo <dkomo...@cris.com> wrote:
>
>: A compressed file is sent down a transmission line. Along the way a
>: burst of thermal noise occurs in the middle of the file, scrambling
>: many of the bits. When the file is received at the other end of the
>: line, it is too garbled to be decompressed. Effectively the whole
>: file is lost.
>
>: Please explain how information is conserved in this process.
>
>Such an operation does not demonstrate that the laws of physics do not
>conserve information, or are not invertible.
Have you never heard of basic thermodynamics? Suppose I empties a
bottle of ink into the ocean. Now, physically reverse the process.
>What they demonstrate is that the universe started off in an ordered state
>and is becoming disordered.
>That phenomenon happens equally well in reversible universes.
>E.g. Run the reversible gas simulation at:
>http://hex.org.uk/diffusion forwards.
Reversible gases are an idealization. Real-world gases aren't
reversible. Computer games and computer-graphics simulations
aren't the same as reality.
>You will see an ordered state disintegrating into a chaotic mess.
>Then hit the "reverse button". You will see that nothing was actually
>lost - it merely appeared so.
>
>So with your file.
You haven't shown that at all. You simply *assume* that the process
is reversible -- which is the same thing as assuming the conclusion
you wish to hold.
Now, instead, suppose that the file is stored on magnetic media and
there are no other extant copies of the file. Suppose further that I
heat that magnetic media well past its Curie point. Suppose, in fact,
that we *smelt* the iron oxides in the magnetic media, melting them
in the process.
Standard thermodynamics says that the process is physically
irreversble and that the information is irretrievably lost.
If you want to claim that is not the case, the onus is on *you* to
show otherwise.
Honestly, Tim, this is no better than the YEC pseudo-science we
usually get here.
Jonathan Stone
>Jon Fleming <jo...@fleming-nospam.com> wrote:
>: On 1 Jan 2002 11:54:52 -0500, Tim Tyler <t...@iname.com> wrote:
>:>John Wilkins <john.w...@bigpond.com> wrote:
>:>: Tim Tyler <t...@iname.com> wrote:
>:>:> Prof Weird <pol...@msx.dept-med.pitt.edu> wrote:
>
>: However, I haven't seen you or anyone state these conservation of
>: information laws, or provide an argument that establishes that they
>: actually exist.
>
>Someone else has quoted Fredkin stating them on:
>http://digitalphilosophy.org/finite_nature.htm
That's quite simply wrong. the cited page quotes Fredkin as *postulating*
such a law. And the quotd postulation is simply handwaving.
Why handwaving? Because the cited page makes some fairly elementary
(undergraduate-level) mistakes. In particular, it confounds a notion
of a closed, recurring system -- one which cycles through the same set
of states -- with reversibility.
More to the point, the cited page doesn't show that
>We do not know whether they are correct or not for certain.
"For certain"? What d'you mean, "for certain"? Science doesn't give
us absolute certainty. Science is inductive, we never know any
scientific generalization _for certain_. But leaving aside sophistry
of the we-are-allbrains-in-vats sophistry, we know that the physical
universe has processes which are truly irreversible.
That's basic thermodynamics: high-school science (A levels, perhaps,
in the UK?); first-year univesirty chemistry for sure.
To coin a prhase, bringing up "certainty" like this suggests that
one's grasp of what science is, is lacking opposable thumbs.
> However, they
>have certainly not been rejected - and until they have been, asserting
>that "information comes from nowhere" may prove to be mistaken.
I dont see anyone defending the idea that ``information comes from
nowhere''. I think it's a strawman.
John Wilkins
You have it right, as I understand it.
Consider this: we can make turing machines, or can we? A turing machine
is a reader/writer head that can write or erase a mark on a cell on an
infinitely long piece of paper tape, and move left or right. A turing
machine can simulate anything that is computable. Is computable
information therefore physical?
No. A turing machine never fails in one of its components, or generates
heat, or gets coffee all over its keyboard. Only physical objects do
this. A turing machine has infinite resources (the tape is its working
store and its offline store simultaneously). As much as I like my Mac
with 1Gb RAM and 50-odd Gb storage, it is not even remotley like a
"real" turing machine.
You can make a turing machine instance (ie, an approximation of a turing
machine) out of a brass and mahogany steam engine (this is the Musgrave
Engine, suggested to me by Ian Musgrave). As a steam engine it converts
heat into work. Suppose it runs a simulation of a Shannon-Weaver
sender-channel-receiver system (call it SW). Now suppose I make a turing
machine approximation of SW using my HP Calculator. Now suppose I do the
same thing using a thousand slaves and blocks of stone. Each of these
runs SW. Each of these is *logically* (abstractly) equivalent. The
information content is the same. Take a snapshot time-series of each
state of each "turing machine" and play it at the same rate, and each
will have the same SW information content. But the thermodynamic
expenditure of each is vastly different.
Now, as to intrinsic information, no information is ever intrinsic.
Information is a property of the abstractions of things - the
symbolisation or algorithmic model of some physical system, or int he
case of Fisher Info, the measurement of some physical system. It
requires a relationship between the things modelled and the modelling
systems (the "observers"). Sans that measurement, or some prior mapping
relations between the abstract symbols and the things signified, there
is *no* information "present".
Atoms contain physical form or structure - but this doesn't become
"information" until it is abstracted in some way.
....
John Wilkins
> John Wilkins <john.w...@bigpond.com> wrote or quoted:
>
> :> I tried to explain above how information creation and destruction was
> :> prevented by physical law. Were this not so, the universe would not
> :> be microscopically reversible - and as far as we can tell - it is.
>
> : Why does the destruction or creation of information make physics
> : reversible?
>
> It doesn't. If it happened, it would make the laws of physics
> *irreversible*.
Typo. You have the correct reading.
>
> : I thought *that* (irreversibility) was due to the 2LoT.
>
> Yes, it is, which is a statistical law.
If statistical distributions are an irreducible aspect of world
ensembles, doesn't that make some aspects of the world randomised?
>
> 2LoT analogues are common in reversible systems - though in the long run
> they are violated as much as they are obeyed in such environments.
>
> :> The idea that information can come from nowhere is the hypothesis of
> :> randomness. If "truly random" things exist, information *can* come from
> :> nowhere. On the other hand - if the world is deterministic - you can't
> :> get new information from anywhere - all you can do is shift it around.
> :>
> :> If you are having difficulty understanding my position, you might
> :> like to consider that I enjoy entertaining the finite nature hypothesis -
> :> in which the universe is fundamentally informational. Information is
> :> "spatiotemporal". Only spatiotemporal things can be said to exist at all.
> :> One of my links pages about the subject: http://finitenature.com/links/
>
> : Well now we pass from information theory and physics to [gasp]
> : metaphysics. But I suppose we were there all along.
>
> I generally consider the finite nature hypothesis and digital physics
> to be physics - rather than metaphysics. There are related experiments,
> and it can be confirmed or rejected. Anyway...
You can experimentally disprove indeterminism?
>
> : Let me expound my position.
>
> : I know of Wheeler's "it from bit" idea that physics is just information,
> : and I reject it. In fact, in my opinion, we derive "bit" from "it" by
> : abstraction. Abstract things exist, so far as they do (I am a bit of a
> : nominalist, and do not admit things other than spatiotemporal things
> : exist in the proper sense of "exist") outside time and space. At best
> : they can be instantiated in physical things. Information is not a causal
> : property - thermodynamic entropy is. So far as I admit, information is
> : an abstract property of abstract objects (symbols), and it is *not*
> : spatiotemporal. Wheeler's slogan is a cosmologist's conceit.
>
> It doesn't sound like we are destined to agree, then.
>
> IMO, nobody can reject the idea that "physics is just information" at the
> moment - except by making a leap of faith.
Or, rephrasing, nobody can equate physics with information without
committing the ontological error - the fallacy of misplaced
concreteness, or thinking that a noun indicates a real object.
>
> : To demonstrate that information is lost, let me use an example of Elliot
> : Sober's - imagine you have a near-perfectly smooth spherical section
> : laying flat on a horizontal surface. Around the edge of the section is a
> : lip. At the bottom of the section is a marble. You may know that the
> : trajectory of the marble began at some point on the lip, but all
> : information as to what that trajectory was is now lost. You cannot
> : reconstruct the past perfectly, and as time progresses, less information
> : remains.
>
> This experiment does not demonstrate what you appear to think it does :-(
How so?
>
> : Now this sense of information is slightly different - here I am
> : referring to Fisher information, or exactness of measurement, but the
> : point remains for the other senses: a symbolic representation of the
> : entire system at t will not give us that trajectory, so Shannon
> : information is lost.
>
> No - doesn't follow. You are not measuring the state of the system
> properly. If you can't measure it exactly all bets about information are
> off - since you don't know if you are measuring the same thing twice.
Then except for the entire universe, which we cannot measure, any system
is uncertain and to all intents and purposes even if we grant Wheeler's
slogan, information is lost for any *observer*.
>
> Again, this is like measuring heat in an open system - and concluding that
> energy is not conserved, since your early reading does not match your
> later one.
>
> : And as time increases since the event, an increasing number of possible
> : trajectories could deliver that resting position [...]
>
> This is false in a reversible system. Physicists tell us that the world
> is a reversible system - as far as they can tell.
That's not what I understand, but I must concede to those who study
theoretical physics. I do not.
>
> : and so AIT information is lost, because the number of algorithms that
> : could deliver the result in a model of the system increases and so
> : Omega is increasing.
>
> No. The conclusion is faulty because the premise is suspect - There is
> not one resting state, but many - if you consider all the gas particles in
> your system. The claim that many states lead to one final state is an
> rests on this unproven assumption.
What gas particles? I said nothing about an atmosphere. A spherical
marble in a spherical section.
Is this the guy who has argued (well but not convincingly to me) that
the laws of physics are outworkings of Fisher Information? Friedkin? I
have a friend whose PhD is on him.
I do not think that the universe is grainy by necessity, and even if it
is this does not make macroscale information preserved (IMO).
Cheers
Wade Hines
Jonathan Stone wrote:
> In article <GpBvq...@bath.ac.uk>, Tim Tyler <t...@iname.com> wrote:
> >dkomo <dkomo...@cris.com> wrote:
> >: A compressed file is sent down a transmission line. Along the way a
> >: burst of thermal noise occurs in the middle of the file, scrambling
> >: many of the bits. When the file is received at the other end of the
> >: line, it is too garbled to be decompressed. Effectively the whole
> >: file is lost.
> >: Please explain how information is conserved in this process.
> >Such an operation does not demonstrate that the laws of physics do not
> >conserve information, or are not invertible.
> Have you never heard of basic thermodynamics? Suppose I empties a
> bottle of ink into the ocean. Now, physically reverse the process.
Tim has an unorthodox notion about microscopic reversibility.
As per your example, if one focused on a single molecule of dye in the
ink bottle, it can freely diffuse back into the bottle. The same is
true for each and every molecule that diffuses out. This does not mean
that all the ink can diffuse back into the bottle.
There are better examples of macroscopic irreversibility that don't
violate microscopic reversibility.
I can't make head or tail of the rest of his comments and keep thinking
he is simple confused based on a false start. Conservation of information
also seems like such a non starter that I have a hard time grasping
his perspective. This may be my loss.
For more on the break between microscopic reversibility and
macroscopic irreversibility, see:
http://www.mada.org.il/website/html/eng/2_1_1-13.htm
...
Jonathan Stone
[... snip preceding quoted text...]
>> It doesn't. If it happened, it would make the laws of physics
>> *irreversible*.
>
>Typo. You have the correct reading.
>> : I thought *that* (irreversibility) was due to the 2LoT.
>>
>> Yes, it is, which is a statistical law.
Tim, you missed the point. Classical thermo was (if we can exclude the
famous contradictions solved by QM, for argument's sake) derived in
accordance with both mathematical models *AND EXPERIMENT*.
Thermo is not *just* a statistical law. The statistical law is a part
of a *theory*. Boltzmann's theory can, like any theory, be viewed in
anumber either as a tool; or as a mathematical derivation within some
larger mathematical model; or as a explanation of many well-observed
phenomena.
Thermodyamics isn't just a statistical law. The underlying physics is
important, otherwise why would there be a difference between thermo
and Shannon information theory?
>If statistical distributions are an irreducible aspect of world
>ensembles, doesn't that make some aspects of the world randomised?
John: Happy new year. I'll see that, and raise you one posthumous
Popperian propensity theory. :-).
[...] link to http://finitenature.com/links/ ..]]
>> IMO, nobody can reject the idea that "physics is just information" at the
>> moment - except by making a leap of faith.
"Physics is just information?" John, how do you read this? Is this an
appeal to hidden-variable theories? Looks awfully like one, from
where I'm sitting.
>Or, rephrasing, nobody can equate physics with information without
>committing the ontological error - the fallacy of misplaced
>concreteness, or thinking that a noun indicates a real object.
Oh, get back in your cave....
>> : To demonstrate that information is lost, let me use an example of Elliot
>> : Sober's - imagine you have a near-perfectly smooth spherical section
>> : laying flat on a horizontal surface. Around the edge of the section is a
>> : lip. At the bottom of the section is a marble. You may know that the
>> : trajectory of the marble began at some point on the lip, but all
>> : information as to what that trajectory was is now lost. You cannot
>> : reconstruct the past perfectly, and as time progresses, less information
>> : remains.
>>
>> This experiment does not demonstrate what you appear to think it does :-(
>
>How so?
Tim doesn't understand the experiment.
>> This is false in a reversible system. Physicists tell us that the world
>> is a reversible system - as far as they can tell.
Pedant time: Physicists tell us that formulations of certain formal
systems are formally time-reversible. My understanding is that that's
*not* the same thing as saying that the real world is time-reversible.
But the map is not the territory. even if it's a damn good map.
>That's not what I understand, but I must concede to those who study
>theoretical physics. I do not.
The evidence indicates Tim is unwilling to defer even to those who
*have* studied theoretical physics :).
>> No. The conclusion is faulty because the premise is suspect - There is
>> not one resting state, but many - if you consider all the gas particles in
>> your system. The claim that many states lead to one final state is an
>> rests on this unproven assumption.
>
>What gas particles? I said nothing about an atmosphere. A spherical
>marble in a spherical section.
Even if one grants Tim the assumption of gas particles, for argument's
sake, Tim's argument is circular. He's *assuming* that the trajectory
of the system through time is an invertible function. The onus is on
him to demonstrate that.
At this point, I'm tempted to invoke Einstein and brownian motion, and
Feynman's comments on reversible modelling in the introduction to _The
Feynman Lectures_ (paint molecules evaporating from a chair, IRRC) but
I dont think Tim is capable of making the connection.
I _think_ Tim is assuming some Platonic world where *complete*
information about the physical is available, for free: which, as you
noted, is in violation of Heisenberg's uncertainty principle.
But I could well be wrong.
>> Much follows from assuming finite nature. E.g. general relativity can't
>> hold on large scales.
>
>Is this the guy who has argued (well but not convincingly to me) that
>the laws of physics are outworkings of Fisher Information? Friedkin? I
>have a friend whose PhD is on him.
>
>I do not think that the universe is grainy by necessity, and even if it
>is this does not make macroscale information preserved (IMO).
Or even quantum-scale, which makes the point more forcefully.
It's analagous to saying:
Here's an electron which just escaped from an undergrad physics Lab.
Now, tell me, did it come from a one-slit or a two-slit experiment?
Jonathan Stone
>> Have you never heard of basic thermodynamics? Suppose I empties a
>> bottle of ink into the ocean. Now, physically reverse the process.
>
>Tim has an unorthodox notion about microscopic reversibility.
>
>As per your example, if one focused on a single molecule of dye in the
>ink bottle, it can freely diffuse back into the bottle. The same is
>true for each and every molecule that diffuses out. This does not mean
>that all the ink can diffuse back into the bottle.
hi wade,
Well, it *could*, but the probabilities against it are astronomical.
I used the ink example because it made sense to me as a kid.
My experience of Tim is that quite a few of his notions are unorthox,
if not downright half-baked: a new-agey mishmash of popularizations
And, rather like some of the creationists we get here, he has a hard
time coming to grips with that. Tims' responses to criticism often
strike me as incoherent; also much as with the creationists.
But as you say, it may be my loss.
>There are better examples of macroscopic irreversibility that don't
>violate microscopic reversibility.
In another post, I suggested taking a hard disk drive, and heating the
magnetic oxide not just past its Curie point, but acutally smelting it
to iron. I wonder if that will ever get a coherent response.
>I can't make head or tail of the rest of his comments and keep thinking
>he is simple confused based on a false start. Conservation of information
>also seems like such a non starter that I have a hard time grasping
>his perspective. This may be my loss.
>For more on the break between microscopic reversibility and
>macroscopic irreversibility, see:
>http://www.mada.org.il/website/html/eng/2_1_1-13.htm
Arrow of time, huh? I'm not sure the old chestnut of filming a diver
into a pool and playing it backwards will help much.
The way I read it, Tim leaps for an almost Platonic idea of a state
vector *somewhere*, a state vector which can be used to reverse
computation of an irreverseible process, and thus (in Tim's view)
reverse the physical process. Something like that.
One can respond to that either with Heisenber and h-bar, or the kind
of response in the Feynman lectures. I think I alluded to both in this
thread already....
Wade Hines
Jonathan Stone wrote:
> In article <3C33CCA5...@rcn.com>, Wade Hines <wade....@rcn.com> wrote:
> >> Have you never heard of basic thermodynamics? Suppose I empties a
> >> bottle of ink into the ocean. Now, physically reverse the process.
> >Tim has an unorthodox notion about microscopic reversibility.
> >As per your example, if one focused on a single molecule of dye in the
> >ink bottle, it can freely diffuse back into the bottle. The same is
> >true for each and every molecule that diffuses out. This does not mean
> >that all the ink can diffuse back into the bottle.
> Well, it *could*, but the probabilities against it are astronomical.
> I used the ink example because it made sense to me as a kid.
...
> >There are better examples of macroscopic irreversibility that don't
> >violate microscopic reversibility.
> In another post, I suggested taking a hard disk drive, and heating the
> magnetic oxide not just past its Curie point, but acutally smelting it
> to iron. I wonder if that will ever get a coherent response.
>
> >I can't make head or tail of the rest of his comments and keep thinking
> >he is simple confused based on a false start. Conservation of information
> >also seems like such a non starter that I have a hard time grasping
> >his perspective. This may be my loss.
>
> >For more on the break between microscopic reversibility and
> >macroscopic irreversibility, see:
> >http://www.mada.org.il/website/html/eng/2_1_1-13.htm
>
> Arrow of time, huh? I'm not sure the old chestnut of filming a diver
> into a pool and playing it backwards will help much.
>
> The way I read it, Tim leaps for an almost Platonic idea of a state
> vector *somewhere*, a state vector which can be used to reverse
> computation of an irreverseible process, and thus (in Tim's view)
> reverse the physical process. Something like that.
>
> One can respond to that either with Heisenber and h-bar, or the kind
> of response in the Feynman lectures. I think I alluded to both in this
> thread already....
Things can dissolve to some debatable metaphysics. The uncertainty
principle does not require non determinancy so much as speak to
ignorance. Time reversibility in such cases is relegated to
thought experiments. Cross non determinancy::ignorance with
unheard trees falling in the woods and ask how information
exists (not even in a Coppenhagen sense) without any chance
of decifering it.
Apparently the angle is that the information is still there
in the reduced iron of your hard drive and it's radiated
afterglow but encrypted on the other side of the uncertainty
principles via cryptic determinancy. That or the rye bread
has gotten moldy again.
Jonathan Stone
>
>Things can dissolve to some debatable metaphysics. The uncertainty
>principle does not require non determinancy so much as speak to
>ignorance.
Well, uncertainty plus alleged demonstrations that hidden-variable
theories are untenable, i guess. I never followed von neumann's proof.
Time reversibility in such cases is relegated to
>thought experiments. Cross non determinancy::ignorance with
>unheard trees falling in the woods and ask how information
>exists (not even in a Coppenhagen sense) without any chance
>of decifering it.
>
>Apparently the angle is that the information is still there
>in the reduced iron of your hard drive and it's radiated
>afterglow but encrypted on the other side of the uncertainty
>principles via cryptic determinancy.
Which is when I asked Wilkins if this looked like hidden-variable
thoeires in newagey-clothes to him, too.
>That or the rye bread has gotten moldy again.
Discogito, Ergot sum? I was wondering what kind of drugs might
inspire Tim's worldview...
John Wilkins
*lol*
>
> [...] link to http://finitenature.com/links/ ..]]
>
>
> >> IMO, nobody can reject the idea that "physics is just information" at the
> >> moment - except by making a leap of faith.
>
> "Physics is just information?" John, how do you read this? Is this an
> appeal to hidden-variable theories? Looks awfully like one, from
> where I'm sitting.
I'm not up on that aspect of philphysics.
>
> >Or, rephrasing, nobody can equate physics with information without
> >committing the ontological error - the fallacy of misplaced
> >concreteness, or thinking that a noun indicates a real object.
>
> Oh, get back in your cave....
Them shadows are real, pardner.
>
>
>
> >> : To demonstrate that information is lost, let me use an example of Elliot
> >> : Sober's - imagine you have a near-perfectly smooth spherical section
> >> : laying flat on a horizontal surface. Around the edge of the section is a
> >> : lip. At the bottom of the section is a marble. You may know that the
> >> : trajectory of the marble began at some point on the lip, but all
> >> : information as to what that trajectory was is now lost. You cannot
> >> : reconstruct the past perfectly, and as time progresses, less information
> >> : remains.
> >>
> >> This experiment does not demonstrate what you appear to think it does :-(
> >
> >How so?
>
> Tim doesn't understand the experiment.
>
> >> This is false in a reversible system. Physicists tell us that the world
> >> is a reversible system - as far as they can tell.
>
> Pedant time: Physicists tell us that formulations of certain formal
> systems are formally time-reversible. My understanding is that that's
> *not* the same thing as saying that the real world is time-reversible.
>
> But the map is not the territory. even if it's a damn good map.
Unless we use the territory as its own map (apologies to Lewis Carroll).
>
> >That's not what I understand, but I must concede to those who study
> >theoretical physics. I do not.
>
> The evidence indicates Tim is unwilling to defer even to those who
> *have* studied theoretical physics :).
>
> >> No. The conclusion is faulty because the premise is suspect - There is
> >> not one resting state, but many - if you consider all the gas particles in
> >> your system. The claim that many states lead to one final state is an
> >> rests on this unproven assumption.
> >
> >What gas particles? I said nothing about an atmosphere. A spherical
> >marble in a spherical section.
>
> Even if one grants Tim the assumption of gas particles, for argument's
> sake, Tim's argument is circular. He's *assuming* that the trajectory
> of the system through time is an invertible function. The onus is on
> him to demonstrate that.
>
> At this point, I'm tempted to invoke Einstein and brownian motion, and
> Feynman's comments on reversible modelling in the introduction to _The
> Feynman Lectures_ (paint molecules evaporating from a chair, IRRC) but
> I dont think Tim is capable of making the connection.
>
> I _think_ Tim is assuming some Platonic world where *complete*
> information about the physical is available, for free: which, as you
> noted, is in violation of Heisenberg's uncertainty principle.
> But I could well be wrong.
No, I think you are right here. This is Platonism pure and simple.
That's why I said this was a metaphysical dispute.
>
>
> >> Much follows from assuming finite nature. E.g. general relativity can't
> >> hold on large scales.
> >
> >Is this the guy who has argued (well but not convincingly to me) that
> >the laws of physics are outworkings of Fisher Information? Friedkin? I
> >have a friend whose PhD is on him.
> >
> >I do not think that the universe is grainy by necessity, and even if it
> >is this does not make macroscale information preserved (IMO).
>
> Or even quantum-scale, which makes the point more forcefully.
> It's analagous to saying:
>
> Here's an electron which just escaped from an undergrad physics Lab.
> Now, tell me, did it come from a one-slit or a two-slit experiment?
Information in this case is retained. If it escaped an undergraduate,
then it will have escaped from the simpler of the two experiments - a
one slit. Now if it escaped from a graduate lab, then I'd not be able to
answer. In fact, I'd be uncertain...
R. Baldwin
> R. Baldwin <res0k7y...@verizon.net> wrote:
> : "Tim Tyler" <t...@iname.com> wrote in message
>
> :> It does not follow from this that it is not conserved.
>
> : The burden is on the advocates of this new law, "Conservation of
> : Information," to demonstrate how information is always conserved.
Why
> : should we take your word for it?
>
> You should not. However, I don't see why you think the burden lies
with
> one party.
The burden of demonstrating a new idea always lies with the proponent.
Well, don't expect to many people to agree with you if you can't
explain what you mean.
>
> Fredkin offered an informal definition:
>
> ``In this context information is the stuff that enables a reversible
> system to, if time were reversed, proceed backwards along the
exact
> path that it took going forwards''
>
> ...and then a quantitative one:
>
> ``All reversible systems have a conserved quantity; log2(n), where n
is
> the number of different states that the system progresses through
in
> one cycle. (All finite state reversible systems evolve along a
closed
> cycle.) This number is constant, independent of where the system
is on
> that cycle and is equal to the quantity of information
conserved.''
>
> Is this unclear?
The relevance is not clear. Why should we only be concerned about
reversible systems?
>
> : Perhaps you should start by defining "information." If you
> : want others to understand you, the commonly understood meaning of
> : "information" should be included in the definition.
>
> There is more than one meaning of "information" - e.g. see Shannon's
> information, and Chaitin's algorithmic information.
I didn't see that either of these was the basis of the principle you
propose.
By the way, I can destroy Shannon information. Let's say the
communication channel is a MOSFET memory device. The transmitter is
the initial write function, the receiver is the final read function.
In between I alternately write all ones then all zeros at least 1000
times. The initial data is now unrecoverable, even by statistical
techniques. It's final state is reliably either all ones or all zeros,
and can be represented by one bit.
>
> : The Principle of Conservation of Energy is a generalization of
> : experience, so far not contradicted by observation of nature. It
has
> : years of research backing it up. What research provides the basis
of
> : the "Law of Conservation of Information?" Show your work.
>
> It is based on an unproven hypothesis. We do not know for certain
whether
> the law exists or not. The fact that the laws of physics are
symmetrical
> under an operation involving inverting time strongly suggests that
no
> information is created or destroyed as time passes.
Not exactly. So far, Conservation of Energy has always been found to
agree with observed data. No evidence has ever been found to
contradict it. Though it has at times seemed to fail, subsequent
experiment has always found the missing energy. The Principle of
Conservation of Energy is a generalization of observation. As such it
can't be proven. It is a model that agrees with all the data we have
so far. There is no mathematical basis for a proof. It does not make
sense to discuss a proof. And we never know for certain whether any
scientific hypothesis is correct.
Conservation of Energy has a tremendous amount of experimental
observation backing it up, going back many years. What observations do
we have in favor of Conservation of Information?
>
> My objection is that asserting that information is created *assumes*
that
> no such law exists - which has not been demonstrated.
>
> : Consider the Harry Potter books burnt by Jack Brock's
fundamentalist
> : church in New Mexico recently. By our common understanding of what
> : "information" means, some information was destroyed by this
process.
>
> Not really. If all the information necessary to recover the
original
> state is present, nothing has been destroyed. It merely /seems/
that way.
How? I described an irreversible process. The pages of the book are
closed, so you don't have "fossil" electromagnetic energy telling you
what it used to say.
That's an awfully big if. There are many processes considered to be
irreversible. The example with the artist involves three: (a) melting
a solid object, thereby destroying its shape; (b) biological mortality
(followed obviously be decay); (c) absorbtion of electromagnetic
energy by (nonphotographic) walls (and subsequent blackbody radiation
having lost the visual phase information of the original
electromagnetic energy). Similarly, burning a book is an irreversible
process.
The Second Law of Thermodynamics tells us there ARE irreversible
processes. We have never observed any violation of this law. What
evidence do you have that all processes are reversible?
R. Baldwin
> R. Baldwin <res0k7y...@verizon.net> wrote:
>
> : How can you possibly avoid non-determinism? It requires more than
one
> : atom to form the computational elements that model the behavior of
an
> : atom. This makes it impossible to model everything.
>
> That is not what "determinism" means. It implies that the future of
the
> system is foreordained - not that entities embedded within the
system can
> predict that future in its entirety.
That is probably a matter of definition. The definition of
non-determinism I'm familiar with would be, "behaviour that cannot be
completely predicted through analysis."
David Jensen
>Jonathan Stone <jona...@DSG.Stanford.EDU> wrote:
>> John: Happy new year. I'll see that, and raise you one posthumous
>> Popperian propensity theory. :-).
>
>*lol*
Couldn't it be a posthumous Popperian propensity postulate, please?
Tim Tyler
: A question about "conservation of information":
: If information is conserved, is it possible, by Noether's
: Theorem, to state a corresponding symmetry principle?
The short answer is that I'm not certain. Noether´s theorem allows you to
obtain conservation laws from symmetries - but not the reverse AFAICS.
Tim Tyler
: "Tim Tyler" <t...@iname.com> wrote in message
:> R. Baldwin <res0k7y...@verizon.net> wrote:
:> : How can you possibly avoid non-determinism? It requires more than
:> : one atom to form the computational elements that model the behavior
:> : of an atom. This makes it impossible to model everything.
:>
:> That is not what "determinism" means. It implies that the future of
:> the system is foreordained - not that entities embedded within the
:> system can predict that future in its entirety.
: That is probably a matter of definition. The definition of
: non-determinism I'm familiar with would be, "behaviour
: that cannot be completely predicted through analysis."
``Determinism holds iff for every event that occurs, some subset of the
prior state of the universe is a sufficient cause of that event.''
- http://www.utas.edu.au/docs/humsoc/philosophy/ccc/slides/4h.html
Here's a summary of different definitions of determinism:
- http://www.utas.edu.au/docs/humsoc/philosophy/ccc/slides/5l.html
The third is concordant with yours - on the proviso that the agent
performing the analysis is "Laplace's Demon/Popper's superscientist".
This is important - *not* an embedded agent - an all-knowing entity.
I don't think I have encountered a definition of determinism that
allows determinism to be refuted by the difficulty of building a complete
model of the world within that world.
Tim Tyler
:> : The burden is on the advocates of this new law, "Conservation of
:> : Information," to demonstrate how information is always conserved.
:> : Why should we take your word for it?
:>
:> You should not. However, I don't see why you think the burden lies
:> with one party.
: The burden of demonstrating a new idea always lies with the proponent.
I do not know whether conservation of information holds or not. I
/suspect/ that it does - and consequently take exception to the
unproven assertion that information can be created from nothing.
:> : So this law is postulated without a good definition? It seems you
:> : and your colleagues have much more work to do before putting this idea
:> : forward.
:>
:> Definitely not. That finite nature paper is ten years old. The
:> ideas are too important to wait until all the is have been dotted
:> and all the ts crossed.
: Well, don't expect to many people to agree with you if you can't
: explain what you mean.
If you don't see what I am talking about, by all means ask for
clarification, and I'll try to clarify. The ideas are not terribly
difficult to grasp.
[snip Fredkin's definitions]
:> Is this unclear?
: The relevance is not clear. Why should we only be concerned about
: reversible systems?
We are concerned with reversible systems since physicists think the
universe is a reversible system. If the world proves /not/ to be a
reversible system, the law of conservation of information is almost
certainly toast.
:> : Perhaps you should start by defining "information." If you
:> : want others to understand you, the commonly understood meaning of
:> : "information" should be included in the definition.
:>
:> There is more than one meaning of "information" - e.g. see Shannon's
:> information, and Chaitin's algorithmic information.
: I didn't see that either of these was the basis of the principle you
: propose.
Either of them will do. They may differ quantitatively - but since we are
talking about qualities that do not change, that's not very relevant.
: By the way, I can destroy Shannon information. Let's say the
: communication channel is a MOSFET memory device. The transmitter is
: the initial write function, the receiver is the final read function.
: In between I alternately write all ones then all zeros at least 1000
: times. The initial data is now unrecoverable, even by statistical
: techniques. It's final state is reliably either all ones or all zeros,
: and can be represented by one bit.
That's like saying you can destroy energy by blowing up a building and
measuring the energy of the ashes. In no way does it demonstrate that you
can destroy Shannon information.
:> : The Principle of Conservation of Energy is a generalization of
:> : experience, so far not contradicted by observation of nature. It
:> : has years of research backing it up. What research provides the
:> : basis of the "Law of Conservation of Information?" Show your work.
:>
:> It is based on an unproven hypothesis. We do not know for certain
:> whether the law exists or not. The fact that the laws of physics are
:> symmetrical under an operation involving inverting time strongly
:> suggests that no information is created or destroyed as time passes.
: Not exactly. So far, Conservation of Energy has always been found to
: agree with observed data. No evidence has ever been found to
: contradict it. Though it has at times seemed to fail, subsequent
: experiment has always found the missing energy. The Principle of
: Conservation of Energy is a generalization of observation. As such it
: can't be proven. It is a model that agrees with all the data we have
: so far. There is no mathematical basis for a proof. It does not make
: sense to discuss a proof. And we never know for certain whether any
: scientific hypothesis is correct.
: Conservation of Energy has a tremendous amount of experimental
: observation backing it up, going back many years. What observations do
: we have in favor of Conservation of Information?
Didn't I just answer this above? The universe is reversible - as far as
decades of work by physicists can tell. Reversibility implies - and is
implied by - conservation of information.
:> My objection is that asserting that information is created *assumes*
:> that no such law exists - which has not been demonstrated.
:>
:> : Consider the Harry Potter books burnt by Jack Brock's
:> : fundamentalist church in New Mexico recently. By our common
:> : understanding of what "information" means, some information was
:> : destroyed by this process.
:>
:> Not really. If all the information necessary to recover the
:> original state is present, nothing has been destroyed. It merely
:> /seems/ that way.
: How? I described an irreversible process.
There are no irreversible processes: physics is reversible.
: The pages of the book are closed, so you don't have
: "fossil" electromagnetic energy telling you what it used to say.
If you are happy with throwing away energy and measuring the result,
I'm suprised you believe in the law of conservation of energy :-(
:> Not if the world is reversible. Physicists mathematical theories of
:> the world are reversible. As far as we know the world is reversible.
: That's an awfully big if. There are many processes considered to be
: irreversible.
No, there aren't any. In fact to be fair /some/ physicists think there
is. They think "wavefunction collapse" is irreversible. However this
process is not part of the laws of physics. Nowhere is it embodied in an
equation.
: The example with the artist involves three: (a) melting
: a solid object, thereby destroying its shape; (b) biological mortality
: (followed obviously be decay); (c) absorbtion of electromagnetic
: energy by (nonphotographic) walls (and subsequent blackbody radiation
: having lost the visual phase information of the original
: electromagnetic energy). Similarly, burning a book is an irreversible
: process.
These examples are ridiculous. Do you not understand how such processes
can occur in worlds that are microscopically reversible? The mere fact
that *you* can't see how to reverse them shows nothing.
: The Second Law of Thermodynamics tells us there ARE irreversible
: processes.
No it does not. Note that the second "law" is a statistical law, *not* a
law of physics. Analogues of the second law apply in reversible systems.
The existence of the 2LoT proves nothing about reversibility.
: We have never observed any violation of this law. What
: evidence do you have that all processes are reversible?
Quantum physics says it is. Newtonian physics said it was.
Physicists are in broad agreement: physics is microscopically reversible.
AFAIK, over the last hundred years, no reputable experiment has
seriously suggested otherwise.
Tim Tyler
:> : I thought *that* (irreversibility) was due to the 2LoT.
:>
:> Yes, it is, which is a statistical law.
: If statistical distributions are an irreducible aspect of world
: ensembles, doesn't that make some aspects of the world randomised?
Statistical laws of the behaviour of large numbers of objects say nothing
about whether the laws of physics are reversible. The 2LoT applies
equally in reversible system and irreversible ones. So no - the existence
of the 2LoT says nothing about reversibility.
:> :> The idea that information can come from nowhere is the hypothesis of
:> :> randomness. If "truly random" things exist, information *can* come from
:> :> nowhere. On the other hand - if the world is deterministic - you can't
:> :> get new information from anywhere - all you can do is shift it around.
:> :>
:> :> If you are having difficulty understanding my position, you might
:> :> like to consider that I enjoy entertaining the finite nature hypothesis -
:> :> in which the universe is fundamentally informational. Information is
:> :> "spatiotemporal". Only spatiotemporal things can be said to exist at all.
:> :> One of my links pages about the subject: http://finitenature.com/links/
:>
:> : Well now we pass from information theory and physics to [gasp]
:> : metaphysics. But I suppose we were there all along.
:>
:> I generally consider the finite nature hypothesis and digital physics
:> to be physics - rather than metaphysics. There are related experiments,
:> and it can be confirmed or rejected. Anyway...
: You can experimentally disprove indeterminism?
*I* can't - and no physics can ever be *completely* disproved.
However those who live in deterministic worlds can sometimes find out that
fact with as much certainty as they have about other scientific matters -
by performing experiments, and getting a good finite, discrete model of
it.
:> : Let me expound my position.
:>
:> : I know of Wheeler's "it from bit" idea that physics is just information,
:> : and I reject it. In fact, in my opinion, we derive "bit" from "it" by
:> : abstraction. Abstract things exist, so far as they do (I am a bit of a
:> : nominalist, and do not admit things other than spatiotemporal things
:> : exist in the proper sense of "exist") outside time and space. At best
:> : they can be instantiated in physical things. Information is not a causal
:> : property - thermodynamic entropy is. So far as I admit, information is
:> : an abstract property of abstract objects (symbols), and it is *not*
:> : spatiotemporal. Wheeler's slogan is a cosmologist's conceit.
:>
:> It doesn't sound like we are destined to agree, then.
:>
:> IMO, nobody can reject the idea that "physics is just information" at the
:> moment - except by making a leap of faith.
: Or, rephrasing, nobody can equate physics with information without
: committing the ontological error - the fallacy of misplaced
: concreteness, or thinking that a noun indicates a real object.
"Physics is information" is intended to convey the same idea as
"The world is exactly equivalent to some computer simulation",
which is a hypothesis, rather than an error.
Your ontology is not to my tatse - as I have said, only physical
things exist. Nothing can exist without physical basis - ideas,
information, whatever.
I'm sure you can define information in such a way this it is not
conserved - but I'm not sure whether such a definition is any
longer very relevant to real-world physics.
:> : To demonstrate that information is lost, let me use an example of Elliot
:> : Sober's - imagine you have a near-perfectly smooth spherical section
:> : laying flat on a horizontal surface. Around the edge of the section is a
:> : lip. At the bottom of the section is a marble. You may know that the
:> : trajectory of the marble began at some point on the lip, but all
:> : information as to what that trajectory was is now lost. You cannot
:> : reconstruct the past perfectly, and as time progresses, less information
:> : remains.
:>
:> This experiment does not demonstrate what you appear to think it does :-(
: How so?
It ammounts to an assertion that what you say is true. You /assert/ that
all information about the trajectory is lost, and go on from there,
but this is essentially what you are trying to prove.
:> : Now this sense of information is slightly different - here I am
:> : referring to Fisher information, or exactness of measurement, but the
:> : point remains for the other senses: a symbolic representation of the
:> : entire system at t will not give us that trajectory, so Shannon
:> : information is lost.
:>
:> No - doesn't follow. You are not measuring the state of the system
:> properly. If you can't measure it exactly all bets about information are
:> off - since you don't know if you are measuring the same thing twice.
: Then except for the entire universe, which we cannot measure, any system
: is uncertain and to all intents and purposes even if we grant Wheeler's
: slogan, information is lost for any *observer*.
An embedded observer, yes. Those folks are at the mercy of HUP, and can
never measure the information content of anything physical in the first
place.
:> : And as time increases since the event, an increasing number of possible
:> : trajectories could deliver that resting position [...]
:>
:> This is false in a reversible system. Physicists tell us that the world
:> is a reversible system - as far as they can tell.
: That's not what I understand, but I must concede to those who study
: theoretical physics. I do not.
The property is known as "time-reversal invariance".
"the laws of quantum physics are reversible in time".
- http://www.qtc.ecs.soton.ac.uk/lecture1/lecture1b.html
"both classical and quantum physics are reversible"
- http://www.cise.ufl.edu/~analla/phy_limits/Assign8.html
From a document entitled:
"Three Chronic Errors:Some Misconceptions concerning the Direction of Time":
``There is, nevertheless, an orthodoxy of sorts. For, despite their
differences over the origin or relevance of the Second Law of
Thermodynamics, all these authors (save Prigogine) agree on the
following premises as their point of departure: the direction of time is
not the direction in which events "come to be", since temporal becoming
is a subjective illusion. That is, the laws of physics do not
distinguish between past and future; [...]''
:> : and so AIT information is lost, because the number of algorithms that
:> : could deliver the result in a model of the system increases and so
:> : Omega is increasing.
:>
:> No. The conclusion is faulty because the premise is suspect - There is
:> not one resting state, but many - if you consider all the gas particles in
:> your system. The claim that many states lead to one final state is an
:> rests on this unproven assumption.
: What gas particles? I said nothing about an atmosphere. A spherical
: marble in a spherical section.
What causes the particle to come to rest? Friction. Where is the
information that you are failing to take into account? In the form
of heat, etc.
: Is this the guy who has argued (well but not convincingly to me) that
: the laws of physics are outworkings of Fisher Information? Friedkin?
"Fredkin".
: I have a friend whose PhD is on him.
He's at Carnegie Mellon, I believe.
: I do not think that the universe is grainy by necessity [...]
Nor do I - but I find discreteness a stimulating hypothesis - that I
feel is likely to bear fruit in the fulness of time.
Tim Tyler
: Jonathan Stone wrote:
:> Have you never heard of basic thermodynamics? [...]
: Tim has an unorthodox notion about microscopic reversibility.
My views about *reversibility* are quite orthodox as far as I am aware.
E.g. THE EVERETT FAQ - http://www.hedweb.com/everett/everett.htm
``The laws of physics are reversible (or CPT invariant, more
precisely) and fully compatible with the irreversibility of
thermodynamics''
That covers more than 50% of physicists - if Q1 from the FAQ is anything
to go by.
Where my views are unorthodox is that I like the "finite
nature" hypothesis. According to the originator:
``Today there are scattered throughout the world a very small number of
individuals who have some similar thoughts [...] but I know of no one in
the Physics Establishment. Many of my colleagues and students understand
my vision, but none that I know of admit to believing in it.''
[Quote from http://digitalphilosophy.org/digital_mechanics_book.htm p.1]
John Wilkins
> John Wilkins <john.w...@bigpond.com> wrote:
> : Tim Tyler <t...@iname.com> wrote:
> :> John Wilkins <john.w...@bigpond.com> wrote or quoted:
>
> :> : I thought *that* (irreversibility) was due to the 2LoT.
> :>
> :> Yes, it is, which is a statistical law.
>
> : If statistical distributions are an irreducible aspect of world
> : ensembles, doesn't that make some aspects of the world randomised?
>
> Statistical laws of the behaviour of large numbers of objects say nothing
> about whether the laws of physics are reversible. The 2LoT applies
> equally in reversible system and irreversible ones. So no - the existence
> of the 2LoT says nothing about reversibility.
I'm not competent to run this argument, but methinks you are wrong.
>
> :> :> The idea that information can come from nowhere is the hypothesis of
> :> :> randomness. If "truly random" things exist, information *can* come
> :> :> from nowhere. On the other hand - if the world is deterministic -
> :> :> you can't get new information from anywhere - all you can do is
> :> :> shift it around.
> :> :>
> :> :> If you are having difficulty understanding my position, you might
> :> :> like to consider that I enjoy entertaining the finite nature
> :> :> hypothesis - in which the universe is fundamentally informational.
> :> :> Information is "spatiotemporal". Only spatiotemporal things can be
> :> :> said to exist at all. One of my links pages about the subject:
> :> :> http://finitenature.com/links/
> :>
> :> : Well now we pass from information theory and physics to [gasp]
> :> : metaphysics. But I suppose we were there all along.
> :>
> :> I generally consider the finite nature hypothesis and digital physics
> :> to be physics - rather than metaphysics. There are related experiments,
> :> and it can be confirmed or rejected. Anyway...
>
> : You can experimentally disprove indeterminism?
>
> *I* can't - and no physics can ever be *completely* disproved.
Really? Aristotle's physics is as dead as a theory is capable of being.
Anyway, I wasn't asking for rigid Popperian falsification. Just a decent
disconfirmation.
>
> However those who live in deterministic worlds can sometimes find out that
> fact with as much certainty as they have about other scientific matters -
> by performing experiments, and getting a good finite, discrete model of
> it.
Have you read anything about the antirealist and internal realist
positions in the philosophy of science? If not, you should...
>
> :> : Let me expound my position.
> :>
> :> : I know of Wheeler's "it from bit" idea that physics is just information,
> :> : and I reject it. In fact, in my opinion, we derive "bit" from "it" by
> :> : abstraction. Abstract things exist, so far as they do (I am a bit of a
> :> : nominalist, and do not admit things other than spatiotemporal things
> :> : exist in the proper sense of "exist") outside time and space. At best
> :> : they can be instantiated in physical things. Information is not a causal
> :> : property - thermodynamic entropy is. So far as I admit, information is
> :> : an abstract property of abstract objects (symbols), and it is *not*
> :> : spatiotemporal. Wheeler's slogan is a cosmologist's conceit.
> :>
> :> It doesn't sound like we are destined to agree, then.
> :>
> :> IMO, nobody can reject the idea that "physics is just information" at the
> :> moment - except by making a leap of faith.
>
> : Or, rephrasing, nobody can equate physics with information without
> : committing the ontological error - the fallacy of misplaced
> : concreteness, or thinking that a noun indicates a real object.
>
> "Physics is information" is intended to convey the same idea as
> "The world is exactly equivalent to some computer simulation",
> which is a hypothesis, rather than an error.
In my understanding of logic, it is a category error. If true, like
brains in a vat, it is entirely untestable, and so it has no purchase.
>
> Your ontology is not to my tatse - as I have said, only physical
> things exist. Nothing can exist without physical basis - ideas,
> information, whatever.
This is a matter of philosophical debate. However, on this matter I
agree with you. For me, information is a semantic construct in a natural
language, but mathematicians and logicians will persist in thinking
their abstract symbols are real and that they use a fully formal
language. But nevertheless, all we end up with then is the claim that
information does not exist, because it is not a physical property of
anything. Stop Aristotleianising, you'll go blind :-)
>
> I'm sure you can define information in such a way this it is not
> conserved - but I'm not sure whether such a definition is any
> longer very relevant to real-world physics.
>
> :> : To demonstrate that information is lost, let me use an example of Elliot
> :> : Sober's - imagine you have a near-perfectly smooth spherical section
> :> : laying flat on a horizontal surface. Around the edge of the section is a
> :> : lip. At the bottom of the section is a marble. You may know that the
> :> : trajectory of the marble began at some point on the lip, but all
> :> : information as to what that trajectory was is now lost. You cannot
> :> : reconstruct the past perfectly, and as time progresses, less information
> :> : remains.
> :>
> :> This experiment does not demonstrate what you appear to think it does :-(
>
> : How so?
>
> It ammounts to an assertion that what you say is true. You /assert/ that
> all information about the trajectory is lost, and go on from there,
> but this is essentially what you are trying to prove.
Okay, that's a reasonable response. How could we retrieve that
information then? [I do assume that you cannot - but I'm open to
correction here.]
>
> :> : Now this sense of information is slightly different - here I am
> :> : referring to Fisher information, or exactness of measurement, but the
> :> : point remains for the other senses: a symbolic representation of the
> :> : entire system at t will not give us that trajectory, so Shannon
> :> : information is lost.
> :>
> :> No - doesn't follow. You are not measuring the state of the system
> :> properly. If you can't measure it exactly all bets about information are
> :> off - since you don't know if you are measuring the same thing twice.
>
> : Then except for the entire universe, which we cannot measure, any system
> : is uncertain and to all intents and purposes even if we grant Wheeler's
> : slogan, information is lost for any *observer*.
>
> An embedded observer, yes. Those folks are at the mercy of HUP, and can
> never measure the information content of anything physical in the first
> place.
Cool. Then operationally you are not taking a position that is testably
different from mine.
>
....
> :>
> :> No. The conclusion is faulty because the premise is suspect - There is
> :> not one resting state, but many - if you consider all the gas particles in
> :> your system. The claim that many states lead to one final state is an
> :> rests on this unproven assumption.
>
> : What gas particles? I said nothing about an atmosphere. A spherical
> : marble in a spherical section.
>
> What causes the particle to come to rest? Friction. Where is the
> information that you are failing to take into account? In the form
> of heat, etc.
Fine - but heat dissipates. How do you retrieve information? You cannot.
>
> : Is this the guy who has argued (well but not convincingly to me) that
> : the laws of physics are outworkings of Fisher Information? Friedkin?
>
> "Fredkin".
Sorry, I have the wrong name (but the guy I'm thinking of starts with F,
for what it's worth).
....
Tim Tyler
:> :> : I thought *that* (irreversibility) was due to the 2LoT.
:> :>
:> :> Yes, it is, which is a statistical law.
:>
:> : If statistical distributions are an irreducible aspect of world
:> : ensembles, doesn't that make some aspects of the world randomised?
:>
:> Statistical laws of the behaviour of large numbers of objects say nothing
:> about whether the laws of physics are reversible. The 2LoT applies
:> equally in reversible system and irreversible ones. So no - the existence
:> of the 2LoT says nothing about reversibility.
: I'm not competent to run this argument, but methinks you are wrong.
I'm pretty confident that I'm not.
In fact I'd go the other way - and say that the second law shows up *more*
often in reversible systems, if anything. Irreversible systems can
easily start off in disordered states and systematically become ordered -
no 2LoT there. Conway's game of life is a classic example of such an
irreversible system doing this.
:> :> I generally consider the finite nature hypothesis and digital physics
:> :> to be physics - rather than metaphysics. There are related experiments,
:> :> and it can be confirmed or rejected. Anyway...
:>
:> : You can experimentally disprove indeterminism?
:>
:> *I* can't - and no physics can ever be *completely* disproved.
: Really? Aristotle's physics is as dead as a theory is capable of being.
: Anyway, I wasn't asking for rigid Popperian falsification. Just a decent
: disconfirmation.
Well, I can't do that either. Reversibility "strongly suggests"
determinism - and there is much evidence for reversibility - but
I wouldn't say it was enough to cause most people to reject the random
hypothesis. Indeed, the random hypothesis seems to be quite popular.
:> However those who live in deterministic worlds can sometimes find out that
:> fact with as much certainty as they have about other scientific matters -
:> by performing experiments, and getting a good finite, discrete model of
:> it.
: Have you read anything about the antirealist and internal realist
: positions in the philosophy of science? [...]
I don't see much connection with the subject matter - but no, I have not.
:> : Or, rephrasing, nobody can equate physics with information without
:> : committing the ontological error - the fallacy of misplaced
:> : concreteness, or thinking that a noun indicates a real object.
:>
:> "Physics is information" is intended to convey the same idea as
:> "The world is exactly equivalent to some computer simulation",
:> which is a hypothesis, rather than an error.
: In my understanding of logic, it is a category error. If true, like
: brains in a vat, it is entirely untestable, and so it has no purchase.
How is it untestable?
ISTM that producing such a computer simulation would favour the theory -
and a proof or evidence that the universe can do things that no finite
computer could do would weigh in against it.
:> :> : To demonstrate that information is lost, let me use an example of Elliot
:> :> : Sober's - imagine you have a near-perfectly smooth spherical section
:> :> : laying flat on a horizontal surface. Around the edge of the section is a
:> :> : lip. At the bottom of the section is a marble. You may know that the
:> :> : trajectory of the marble began at some point on the lip, but all
:> :> : information as to what that trajectory was is now lost. You cannot
:> :> : reconstruct the past perfectly, and as time progresses, less information
:> :> : remains.
:> :>
:> :> This experiment does not demonstrate what you appear to think it does :-(
:>
:> : How so?
:>
:> It ammounts to an assertion that what you say is true. You /assert/ that
:> all information about the trajectory is lost, and go on from there,
:> but this is essentially what you are trying to prove.
: Okay, that's a reasonable response. How could we retrieve that
: information then? [I do assume that you cannot - but I'm open to
: correction here.]
*You* may not be able to - but that does not mean that it has been lost or
destroyed - it's a reflection of your capabilities.
:> :> : Now this sense of information is slightly different - here I am
:> :> : referring to Fisher information, or exactness of measurement, but the
:> :> : point remains for the other senses: a symbolic representation of the
:> :> : entire system at t will not give us that trajectory, so Shannon
:> :> : information is lost.
:> :>
:> :> No - doesn't follow. You are not measuring the state of the system
:> :> properly. [...]
:>
:> : Then except for the entire universe, which we cannot measure, any system
:> : is uncertain and to all intents and purposes even if we grant Wheeler's
:> : slogan, information is lost for any *observer*.
:>
:> An embedded observer, yes. Those folks are at the mercy of HUP, and can
:> never measure the information content of anything physical in the first
:> place.
: Cool. Then operationally you are not taking a position that is testably
: different from mine.
We can't figure out whether the universe is reversible by getting enough
data to reverse any section of it.
However, we can get a pretty good idea of whether it is reversible
by looking at the laws of physics.
:> :> No. The conclusion is faulty because the premise is suspect - There is
:> :> not one resting state, but many - if you consider all the gas particles in
:> :> your system. The claim that many states lead to one final state is an
:> :> rests on this unproven assumption.
:>
:> : What gas particles? I said nothing about an atmosphere. A spherical
:> : marble in a spherical section.
:>
:> What causes the particle to come to rest? Friction. Where is the
:> information that you are failing to take into account? In the form
:> of heat, etc.
: Fine - but heat dissipates. How do you retrieve information? You cannot.
I don't see that the inability of an embedded observed to get at something
means that it has been destroyed.
dkomo
>
> dkomo <dkomo...@cris.com> wrote:
>
> : A compressed file is sent down a transmission line. Along the way a
> : burst of thermal noise occurs in the middle of the file, scrambling
> : many of the bits. When the file is received at the other end of the
> : line, it is too garbled to be decompressed. Effectively the whole
> : file is lost.
>
> : Please explain how information is conserved in this process.
>
> Such an operation does not demonstrate that the laws of physics do not
> conserve information, or are not invertible.
It does indeed demonstrate that the laws of physics don't conserve
information. The fact that the laws of physics *are* invertible is
irrelevent.
>
> What they demonstrate is that the universe started off in an ordered state
> and is becoming disordered.
Indeed.
Information = ordered state = improbable configuration.
Entropy = lack of information = disordered state = probable
configuration.
Information tends to spontaneously decrease. This is another way of
stating the 2Lot.
You're using the term "information" in a completely counter-intutitive
sense almost opposite to its normal definition.
> That phenomenon happens equally well in reversible universes.
> E.g. Run the reversible gas simulation at:
> http://hex.org.uk/diffusion forwards.
>
> You will see an ordered state disintegrating into a chaotic mess.
> Then hit the "reverse button". You will see that nothing was actually
> lost - it merely appeared so.
>
Something was lost all right. The ordered state. The ordered state
distintegrated into the chaotic state, which is astronomically more
probable than the the ordered state. This means that information
(using its common sense definition) was irretreviably lost.
Information was therefore not conserved.
Yes, the process is reversable, but only in the sense that all the
equations of classical dynamics are time-reversible. The hooker is
the concept of "spontaneous". The ordered state cannot reappear
"spontaneously" from the disordered state. This is because of the
2Lot.
Hell, even we clever humans could not cause the ordered state to
reappear unspontaneously no matter how we try to direct the gas
molecules (computer simulations don't count).
So, because we can't *actually* reverse the gas molecule simulation,
we have lost information. This is how the real world works.
> So with your file.
And so with the file. There's no physical sequence of operations that
can be performed by us limited human beings that can get back the
information that was lost in that file.
Jonathan Stone
John Wilkins <john.w...@bigpond.com> wrote:
>Tim Tyler <t...@iname.com> wrote:
>
>> John Wilkins <john.w...@bigpond.com> wrote:
>> : Tim Tyler <t...@iname.com> wrote:
>> :> John Wilkins <john.w...@bigpond.com> wrote or quoted:
>>
>> :> : I thought *that* (irreversibility) was due to the 2LoT.
>> :>
>> :> Yes, it is, which is a statistical law.
>>
>> : If statistical distributions are an irreducible aspect of world
>> : ensembles, doesn't that make some aspects of the world randomised?
>>
>> Statistical laws of the behaviour of large numbers of objects say nothing
>> about whether the laws of physics are reversible. The 2LoT applies
>> equally in reversible system and irreversible ones. So no - the existence
>> of the 2LoT says nothing about reversibility.
>
>I'm not competent to run this argument, but methinks you are wrong.
I am too impatient with fools and liars to run this argument. But I
think you and I alreadyh agreed that Tim has an implicit appeal to
some Platonic ideal in which a complete total state-vector of the
Universe can be obtained, and manipulated, for free. (The physical
problems behind this assumption are, I think, far beyond Tim's comprehension.)
But I think here, Tim is confounding two distinct concepts:
a) the time-invertibility of the equations in physical theories,
often called "reversibility"
b) irreversible processes.
There really truly are irreversible physical processes, *despite* the
time-invariance of the physical equations.
Do we see time running backwards -- divers falling upwards, back onto
a springboard; as messy splashes of water leaping off the surroundings
and forming a flat surface on the pool? No, we don't.
Do bottles of ink emptied into an ocean suddenly un-dilute and
jump back into the bottle? No, they don't.
I cannot fathom why, but Tim's argument (and his refusal to accept
other, similar counterexamples) implies quite directly that,
in fact, this sort of thing does happen.
[...]
>This is a matter of philosophical debate. However, on this matter I
>agree with you. For me, information is a semantic construct in a natural
>language,
<splork> _Natural_ language?
>> It ammounts to an assertion that what you say is true. You /assert/ that
>> all information about the trajectory is lost, and go on from there,
>> but this is essentially what you are trying to prove.
>
>Okay, that's a reasonable response. How could we retrieve that
>information then? [I do assume that you cannot - but I'm open to
>correction here.]
Refer to previous posts, about loss of phase information as walls
absorb light and reradiate it as blackbody information.
Or heating magnetic media far past its Curie point.
In hindsight, perhaps Tim is such an fsckwit that he genuinely
*doesn't understand* the counter-examples that have been offered --
*why* it matters that phase infomration is lost, or that alignment of
electron orbital spins in an list of magnetic domains is lost.
I think that's a different kind of failure-to-understand than the
implicit Platonism which you and I see in Tim's postings; but as
before, I may well be wrong.
>Cool. Then operationally you are not taking a position that is testably
>different from mine.
Oh, yes he is. He asserts that there are no irreversible processes in
the real world. I'm tempted to extend (or proxy) Dr. Sokal's
invitation to Tim; but I'm not sure he would get the point even then.
>Fine - but heat dissipates. How do you retrieve information? You cannot.
I suspect that last sentence went beyond the event-horizon of Tim's
comprehension. If Tim responds at all, it will be an appeal to
time-invertibility, and he will not see why that is an inappropriate
response -- *DESPITE* Tim's prior citation of a page which gives a
popularized explaination of *why* it's not an appropriate response.
I generally go for Brazils and hazelnuts this time of year.
With the odd lump of crystallized ginger. How 'bout you?
[...]
Tim Tyler
:> : A compressed file is sent down a transmission line. Along the way a
:> : burst of thermal noise occurs in the middle of the file, scrambling
:> : many of the bits. When the file is received at the other end of the
:> : line, it is too garbled to be decompressed. Effectively the whole
:> : file is lost.
:>
:> : Please explain how information is conserved in this process.
:>
:> Such an operation does not demonstrate that the laws of physics do not
:> conserve information, or are not invertible.
: It does indeed demonstrate that the laws of physics don't conserve
: information. [...]
No, it does not.
: Indeed.
: Information = ordered state = improbable configuration.
: Entropy = lack of information = disordered state = probable
: configuration.
: Information tends to spontaneously decrease. This is another way of
: stating the 2Lot.
I think your logic is upside down. It is conventional to equate entropy
to information - not to a lack of information. This would mean
"increase", is substituted for "decrease" in the last sentence.
However the whole point is mistaken anyway, as far as I am concerned.
: You're using the term "information" in a completely counter-intutitive
: sense almost opposite to its normal definition.
No - the normal definition - i.e. Shannon's - suits me just fine.
:> That phenomenon happens equally well in reversible universes.
:> E.g. Run the reversible gas simulation at:
:> http://hex.org.uk/diffusion forwards.
:>
:> You will see an ordered state disintegrating into a chaotic mess.
:> Then hit the "reverse button". You will see that nothing was actually
:> lost - it merely appeared so.
: Something was lost all right. The ordered state. The ordered state
: distintegrated into the chaotic state, which is astronomically more
: probable than the the ordered state. This means that information
: (using its common sense definition) was irretreviably lost.
No. Did you neglect to run things backwards to recover the original
state. Nothing was lost. How can it have been?
: Yes, the process is reversable, but only in the sense that all the
: equations of classical dynamics are time-reversible.
: The ordered state cannot reappear "spontaneously" from the disordered
: state.
Yet it does *exactly* that if you run the system backwards or the system
is finite and you run it for long enough.
: This is because of the 2Lot.
That is not a law of physics. It does not completely prevent any event
from happening. You cannot invoke it and claim you have demonstrated that
something cannot happen.
: So, because we can't *actually* reverse the gas molecule simulation,
: we have lost information.
Did you hit the "reverse button" in the simulation? It should reverse
just fine.
:> So with your file.
: And so with the file. There's no physical sequence of operations that
: can be performed by us limited human beings that can get back the
: information that was lost in that file.
Ah, finally we are in agreement. Yes *you* can't do it - but that doesn't
mean anything. *You* could not measure the information in the system
initially - or finally. You are not in a good position to say whether
information has been lost or gained on the basis of actual measurements
of the system.
Steve Carlip
> Physicists mathematical theories of the
> world are reversible. As far as we know the world is reversible.
On the contrary, this is a highly controversial and unsettled
issue.
Consider the following process: matter in a pure quantum state
collapses to form a black hole, which then evaporates by Hawking
radiation. If, as all evidence suggests, Hawking radiation is thermal,
this represents an evolution from a pure state to a mixed state, an
irreversible process. Many attempts have been made to get around
this conclusion---subtle correlations within Hawking radiation,
entropy in ``remnants'' left after a black hole evaporates, even a
proposal that Lorentz invariance breaks down near black hole
horizons---but none has yet been especially successful. A good
many physicists (notably Hawking himself) believe that this
shows a fundamental irreversibility; a good many others think
there will be some other answer.
(Look up the ``black hole information loss paradox'' for more.)
> If physics is reversible, information is conserved - if the laws
> were run backwards you would recover the original state.
Well, lets assume that the black hole information loss problem
is resolved in a way that preserves reversibility. Since you're
talking about microscopic reversibility, I assume the definition
of ``information'' you're using is something like the von Neumann
entropy Trace[rho log(rho)], where rho is the density matrix. If
quantum evolution is unitary, this is indeed conserved. On the
other hand, if the Universe is in a pure state, its (conserved) value
is zero, so that doesn't give you a whole lot of information, so to
speak. If you want a useful measure of information, you need to
look at the *relative* entropy of a subsystem. That quantity is
typically nonzero, but is definitely *not* conserved (look up
``entanglement entropy'').
So while you can (perhaps, depending on black hole evaporation
and similar processes) define a conserved ``information'' for the
Universe as a whole, you do so only by losing any usefulness of
the concept. As soon as you try to define *local* information,
the information of a subsystem, you lose conservation.
Steve Carlip
Tim Tyler
: Tim Tyler <t...@iname.com> wrote:
:> Physicists mathematical theories of the
:> world are reversible. As far as we know the world is reversible.
: On the contrary, this is a highly controversial and unsettled
: issue.
: Consider the following process: matter in a pure quantum state
: collapses to form a black hole, which then evaporates by Hawking
: radiation. If, as all evidence suggests, Hawking radiation is thermal,
: this represents an evolution from a pure state to a mixed state, an
: irreversible process. Many attempts have been made to get around
: this conclusion---subtle correlations within Hawking radiation,
: entropy in ``remnants'' left after a black hole evaporates, even a
: proposal that Lorentz invariance breaks down near black hole
: horizons---but none has yet been especially successful.
My money would currently go on the thesis that the same volume of
information goes in as comes back out as the black hole evaporates by
Hawking radiation.
Anyway, these are extreme events on the edges of the laws of physics where
we can't perform experiments easily. It's easy to postulate all sorts of
nonsense at such locations, without much fear of being disproven.
A world that is reversible everywhere - except at such extreme points -
would not really make much difference to the general thesis that the
information that governs temporal evolution is preserved. They would
be the exceptions that "proved" the rule ;-)
: A good many physicists (notably Hawking himself) believe that this
: shows a fundamental irreversibility; [...]
This reminds me of Hawking's "arrow of time reversal" in "A brief
history...". IIRC, Hawking thought that - if the universe was closed and
began to collapse - we would start remembering the future and watch
entropy reversals arise. While - IIRC - he woke up and identified this
as one of his befuddlements - this is one area where he has shown that
his intuition is screwy - IMO.
: So while you can (perhaps, depending on black hole evaporation
: and similar processes) define a conserved ``information'' for the
: Universe as a whole, you do so only by losing any usefulness of
: the concept. As soon as you try to define *local* information,
: the information of a subsystem, you lose conservation.
The notion is still applicable to subsystems, though - as it is with
energy conservation.
Even in an open system, you can still apply conservation laws in a
specified volume - if you take into account the flux of the
conserved quantity through the surrounding surface.
Robert Parson
Jonathan Stone <jona...@DSG.Stanford.EDU> wrote:
>In article <3C33CCA5...@rcn.com>, Wade Hines
<wade....@rcn.com> wrote:
[Lots of snips]
>>Tim has an unorthodox notion about microscopic reversibility.
>>There are better examples of macroscopic irreversibility that don't
>>violate microscopic reversibility.
>
>>I can't make head or tail of the rest of his comments and keep thinking
>>he is simple confused based on a false start. Conservation of information
>>also seems like such a non starter that I have a hard time grasping
>>his perspective. This may be my loss.
>
>
>The way I read it, Tim leaps for an almost Platonic idea of a state
>vector *somewhere*, a state vector which can be used to reverse
>computation of an irreverseible process, and thus (in Tim's view)
>reverse the physical process. Something like that.
Y'all need to make an appointment with Dr. Gibbs. :-)
Tyler's "conservation of information" principle appears to me to
be identical with a well-known consequence of Liouville's Theorem
(conservation of phase space volume.) If so, it is almost trivially
correct, but not very useful.
Specifically, if you define
the entropy of a nonequilibrium system as the integral of
[\rho Log(\rho)] over phase space, where \rho is the phase space
density, then it is easy to show that this entropy is constant
in time. Gibbs explained this disturbing result by drawing an analogy
with the mixing of two incompressible fluids: as the mixture is stirred,
the two fluids stretch out into long and slender filaments which wrap
ever more finely about each other. If you look very closely you can
see the individual tendrils, but if your vision becomes ever so
slightly blurred you just see a uniform color. Gibbs therefore
argued that the phase-space density must be "coarse grained", i.e.
averaged over volume elements of the order of the resolution that
one would actually achieve in an experiment, before using it to
calculate the entropy. (For calculating _equilibrium_ properties
this doesn't matter - the coarse-grained and fine-grained entropies
yield the same answer.)
Von Neumann expressed the same concept
in quantum-mechanical language, where if anything the paradox is
even more striking - so much so that I can't resist a little math:
Let S = - Tr{rho Log(rho)}, where rho is the density matrix .
Then from the quantum Liouville equation:
d rho/dt = -i [rho, H]
where H is the Hamiltonian, you get dS/dt = -i Tr(rho [Log(rho),H]).
Expand the commutator and use the cyclic invariance of the trace,
and you get dS/dt = 0. As in the classical case, you escape from this
by summing rho over small groups of microscopic states before putting
it into the formula for S.
For a macroscopic system, the density of microstates is so large
(on the order of e to the power of Avogadro's number, a truly
colossal figure) that the grain size doesn't matter.
One of the great achievements of modern nonlinear dynamics (aka
"chaos theory") was to show that effectively irreversible behavior
can occur in rather small systems as well as in macroscopic
ones, and the techniques used to show this draw heavily upon ideas
that ultimately go back to Gibbs'arguments about mixing and
coarse-graining.
Stepping back from the technicalities: the fine-grained entropy, which
includes all of the details necessary to specify a system, is indeed
conserved. (Leaving aside cosmological questions, black holes, and
other phenomena beyond the scope of nonrelativistic classical or
quantum mechanics.) It's just not very useful. It is the coarse-grained
entropy that corresponds to the entropy of thermodynamics. N. G.
van Kampen puts this forcefully in his book on Stochastic Processes:
"To a chemist the entropy of a system is a macroscopic state function,
i.e. a function of the thermodynamic variables of the system. In
statistical mechanics entropy is a mesoscopic quantity, i.e. a functional
of the probability distribution. Never is it a microscopic quantity,
because on the microscopic level there is no irreversibility."
(_Stochastic Properties in Physics and Chemistry_ ,North Holland,
1980, p. 202).
------
Robert
Jonathan Stone
Robert Parson <rpa...@spot.colorado.edu> wrote:
>In article <a10kq5$41t$1...@Pescadero.DSG.Stanford.EDU>,
>Jonathan Stone <jona...@DSG.Stanford.EDU> wrote:
>>In article <3C33CCA5...@rcn.com>, Wade Hines
><wade....@rcn.com> wrote:
>
> [Lots of snips]
>
>>>Tim has an unorthodox notion about microscopic reversibility.
>
>>>There are better examples of macroscopic irreversibility that don't
>>>violate microscopic reversibility.
>>
[ deeply quoted text from , iirc, Wade ]
>>
>>The way I read it, Tim leaps for an almost Platonic idea of a state
>>vector *somewhere*, a state vector which can be used to reverse
>>computation of an irreverseible process, and thus (in Tim's view)
>>reverse the physical process. Something like that.
>
> Y'all need to make an appointment with Dr. Gibbs. :-)
I'm certain I mentioned free energy once or twice already. Tim
continues to ignore all counterexamples. It's getting reminiscent of
Karl Crawford. If you want to try banging your head against Tim's,
go for it.
> Tyler's "conservation of information" principle appears to me to
> be identical with a well-known consequence of Liouville's Theorem
> (conservation of phase space volume.) If so, it is almost trivially
> correct, but not very useful.
I don't disagree with the theroem. But TIm is explicilty sahing that
Shannon information is conserved. He's explicitly appealing to
running processes - any process, nevermind Dr. Gibbs or M. Carnot --
backwards. Not conservation of entropy over the phase space; of
reverting to some specified prior point in the phase space.
See Tim's response to the example of thermal noise in a communcations
channel clobbering a message transfer. What you are saying is more or
less that the total number of bits in the mesage is conserved. Tim,
OTOH, is saying that some *specific* message is conserved.
> Stepping back from the technicalities: the fine-grained entropy, which
> includes all of the details necessary to specify a system, is indeed
> conserved. (Leaving aside cosmological questions, black holes, and
> other phenomena beyond the scope of nonrelativistic classical or
> quantum mechanics.)
Uh, we've been talking sold-state semiconductors and ferromagnetic
domains (aligned spin), and one-slit vs two-slit experiments,
already...
> It's just not very useful. It is the coarse-grained
> entropy that corresponds to the entropy of thermodynamics. N. G.
> van Kampen puts this forcefully in his book on Stochastic Processes:
>
> "To a chemist the entropy of a system is a macroscopic state function,
> i.e. a function of the thermodynamic variables of the system. In
> statistical mechanics entropy is a mesoscopic quantity, i.e. a functional
> of the probability distribution. Never is it a microscopic quantity,
> because on the microscopic level there is no irreversibility."
>
> (_Stochastic Properties in Physics and Chemistry_ ,North Holland,
> 1980, p. 202).
Thanks for the quote. I can only hope Tim reads it. :-).
Tim Tyler
:Wade Hines <wade....@rcn.com> wrote:
:>>Tim has an unorthodox notion about microscopic reversibility.
: Tyler's "conservation of information" principle appears to me to
: be identical with a well-known consequence of Liouville's Theorem
: (conservation of phase space volume.) If so, it is almost trivially
: correct, but not very useful.
I don't know about the notion that the idea is not very useful:
For example it could have been used to predict Hawking radiation decades
ago. It is an immediate and trivially obvious consequence of microscopic
reversibility that - if things can fall into black holes making them
larger, other things /must/ be able to emerge, making them smaller.
Only because nobody was taking reversibility seriously did it take so
long for this rather obvious point to emerge.
Today - on similar grounds - we can predict that either singularities
also evaporate, or - much more likely IMO - gravity meets its
match and singularities are never formed in the first place.
Reversibility is - IMO - of critical importance to understanding physics.
Looking at this thread, I see that most participants seem to have no idea
that physics - as close as anyone can tell - is microscopically reversible.
Instead they think that it is irreversible, citing the 2LoT as
self-evident proof - or claiming that the difficultly people have
in actually reversing it supports their view :-(
R. Baldwin
I see what you're talking about, I just don't agree. You provide no
evidence in favor of your unusual position.
>
> [snip Fredkin's definitions]
>
> :> Is this unclear?
>
> : The relevance is not clear. Why should we only be concerned about
> : reversible systems?
>
> We are concerned with reversible systems since physicists think the
> universe is a reversible system. If the world proves /not/ to be a
> reversible system, the law of conservation of information is almost
> certainly toast.
Indeed it would be. And I suspect most of us think it is.
Hardly. A communications channel that always produces the same symbol
(such as a memory full of all zeros or all ones) can be represented by
one bit. You can't get much less information than that by Shannon's
definition. But if you previously had one of a very large number of
letters stored in ASCII in the memory, it had significantly more than
one bit of Shannon information. We started with more information and
ended with less. Seems quite simple.
No, you didn't. I was pointing out your error in stating that the
Principle of Conservation of Energy is "based on an unproven
hypothesis." Such a term is not meaningful in describing a
generalization of observation.
I do not agree with you that microscopic reversibility implies that
the universe is reversable.
>
> :> My objection is that asserting that information is created
*assumes*
> :> that no such law exists - which has not been demonstrated.
> :>
> :> : Consider the Harry Potter books burnt by Jack Brock's
> :> : fundamentalist church in New Mexico recently. By our common
> :> : understanding of what "information" means, some information was
> :> : destroyed by this process.
> :>
> :> Not really. If all the information necessary to recover the
> :> original state is present, nothing has been destroyed. It merely
> :> /seems/ that way.
>
> : How? I described an irreversible process.
>
> There are no irreversible processes: physics is reversible.
What is your evidence for this? We don't accept mathematical formulae
as fact without evidence backing it up. What experiments have shown
that on a large scale physics is reversible?
>
> : The pages of the book are closed, so you don't have
> : "fossil" electromagnetic energy telling you what it used to say.
>
> If you are happy with throwing away energy and measuring the result,
> I'm suprised you believe in the law of conservation of energy :-(
I did not propose throwing away energy and measuring the result, I was
showing that the lost information was not recoverable through
observation of electromagnetic energy.
Just because lead can go through a reversible phase change does not
mean that you can freeze a melted statue and recover the original
shape. Your proposition is ludicrous.
>
> : The Second Law of Thermodynamics tells us there ARE irreversible
> : processes.
>
> No it does not. Note that the second "law" is a statistical law,
*not* a
> law of physics. Analogues of the second law apply in reversible
systems.
> The existence of the 2LoT proves nothing about reversibility.
>
> : We have never observed any violation of this law. What
> : evidence do you have that all processes are reversible?
>
> Quantum physics says it is. Newtonian physics said it was.
> Physicists are in broad agreement: physics is microscopically
reversible.
There is a huge leap from "physics is microscopically reversible" to
"physics is reversible." You ask us to take it without any evidence in
favor, and despite much evidence to the contrary.
I'm going to take as a working hypothesis that "Conservation of
Information" is nonsense.
Tim Tyler
:> Didn't I just answer this above? The universe is reversible - as
:> far as decades of work by physicists can tell. Reversibility implies
:> - and is implied by - conservation of information.
: No, you didn't. I was pointing out your error in stating that the
: Principle of Conservation of Energy is "based on an unproven
: hypothesis." Such a term is not meaningful in describing a
: generalization of observation.
I don't recall stating that. Can you quote me doing so?
: I do not agree with you that microscopic reversibility implies that
: the universe is reversable.
Curious. How can things be reversible on one scale and irreversible on
another. Such a position is not self-consistent.
:> : The pages of the book are closed, so you don't have
:> : "fossil" electromagnetic energy telling you what it used to say.
:>
:> If you are happy with throwing away energy and measuring the result,
:> I'm suprised you believe in the law of conservation of energy :-(
: I did not propose throwing away energy and measuring the result, I was
: showing that the lost information was not recoverable through
: observation of electromagnetic energy.
I agree with that.
Our difference is in who we should consider to be measuring the
information in the universe.
I say, an agent capable of measuring particle positions.
Others seem to think it is best to talk about an embedded agent.
Therein lies the source of the different perspectives.
: Just because lead can go through a reversible phase change does not
: mean that you can freeze a melted statue and recover the original
: shape.
Of course not.
:> Quantum physics says it is. Newtonian physics said it was.
:> Physicists are in broad agreement: physics is microscopically
: reversible.
: There is a huge leap from "physics is microscopically reversible" to
: "physics is reversible."
No - one is a direct consequence of the other.
: You ask us to take it without any evidence in
: favor, and despite much evidence to the contrary.
It follows logically. If the whole system is irreversible some of its
components must be microscopically irreversible.
: I'm going to take as a working hypothesis that "Conservation of
: Information" is nonsense.
OK - you probably won't be wanting to engage me in discussion further
then - since you think my position makes no sense.
Steve Carlip
> Robert Parson <rpa...@spot.colorado.edu> wrote:
> : Tyler's "conservation of information" principle appears to me to
> : be identical with a well-known consequence of Liouville's Theorem
> : (conservation of phase space volume.) If so, it is almost trivially
> : correct, but not very useful.
> I don't know about the notion that the idea is not very useful:
> For example it could have been used to predict Hawking radiation
> decades ago. It is an immediate and trivially obvious consequence
> of microscopic reversibility that - if things can fall into black holes
> making them larger, other things /must/ be able to emerge, making
> them smaller.
A time-reversed black hole is a white hole. It looks nothing at all
like Hawking radiation.
> Only because nobody was taking reversibility seriously did it take
> so long for this rather obvious point to emerge.
This is wrong. The idea of white holes as time-reversed black holes
has been around for a very long time. It has nothing to do with
Hawking radiation. This is easy to see: a classical description of
an object falling into a black hole involves only general relativity,
and not quantum mechanics, so the time-reversed version must
also involve only general relativity. Hawking radiation, on the
other hand, is *fundamentally* quantum mechanical; it depends
explicitly on Planck' constant, which is not there in classical
general relativity at all.
(Moreover, as I pointed out earlier, Hawking radiation does not
appear to be time-reversal-invariant.)
Steve Carlip
Steve Carlip
> Steve Carlip <sjca...@ucdavis.edu> wrote:
> : So while you can (perhaps, depending on black hole evaporation
> : and similar processes) define a conserved ``information'' for the
> : Universe as a whole, you do so only by losing any usefulness of
> : the concept. As soon as you try to define *local* information,
> : the information of a subsystem, you lose conservation.
> The notion is still applicable to subsystems, though - as it is with
> energy conservation.
> Even in an open system, you can still apply conservation laws
> in aspecified volume - if you take into account the flux of the
> conserved quantity through the surrounding surface.
This is not true for the entropy of a subsystem.
It isn't actually even true for energy, at least in general relativity,
where a local energy density cannot be defined. (Gravitational
energy is nonlocalizable.) You might want to look at the relativity
FAQs,
http://www.math.ucr.edu/home/baez/physics/Relativity/GR/energy_gr.html
Steve Carlip
Jonathan Stone
>Robert Parson <rpa...@spot.colorado.edu> wrote:
>:Wade Hines <wade....@rcn.com> wrote:
>
>Looking at this thread, I see that most participants seem to have no idea
>that physics - as close as anyone can tell - is microscopically reversible.
If thats really what you think, then you have a reading comprehension
problem. As far as I can see, most particpants but you are (with the
same caveats as Robert Parson) fully aware at the microscopic level,
events are reversible.
The way I read this conversation, everybody *except you* is also aware
that macroscopic events *are* *not* *reversible*.
If you think macroscopic events are reversible, then I challenge you
to walk out Alan Sokal's window.
>Instead they think that it is irreversible, citing the 2LoT as
>self-evident proof - or claiming that the difficultly people have
>in actually reversing it supports their view :-(
Once again, you have a reading comprehension problem. May I suggest
you go read Chapter 26 of the Feynman lectures, including the section
with the standard (gedanken) experiment of a box with a divider
between two sections, one filled with (ideal-gas) white marbles, one
with black, and how the result of thermalized mixing is
*irreversible*, despite each individual collision being reversible?
To everyone else: I apologise for being so short with Tim. Mostly I'm
being so grumpy because Tim has cited Web pages in support of his
half-baked argumetns, even though those web pages state clearly that
macroscopic events really are irreversible.
Honestly, it makes me wonder if Tim genuinely has some subtle form of
brain-damage.
Tim Tyler
:> : Tyler's "conservation of information" principle appears to me to
:> : be identical with a well-known consequence of Liouville's Theorem
:> : (conservation of phase space volume.) If so, it is almost trivially
:> : correct, but not very useful.
:> I don't know about the notion that the idea is not very useful:
:> For example it could have been used to predict Hawking radiation
:> decades ago. It is an immediate and trivially obvious consequence
:> of microscopic reversibility that - if things can fall into black holes
:> making them larger, other things /must/ be able to emerge, making
:> them smaller.
: A time-reversed black hole is a white hole. It looks nothing at all
: like Hawking radiation.
You misunderstood my point. I am claiming that radiation from black holes
could have been predicted solely on the grounds of reversibility.
I am definitely not claiming that Hawking radition looks like a
time-reversed black hole - it looks nothing like it.
:> Only because nobody was taking reversibility seriously did it take
:> so long for this rather obvious point to emerge.
: This is wrong. The idea of white holes as time-reversed black holes
: has been around for a very long time. It has nothing to do with
: Hawking radiation.
Yes it does. Reversibility disctates that if things can go into a black
hole other things must be able to come out. Thus the existence of Hawking
radition - or something rather similar - could have been predicted solely
on the grounds of reversibility.
: This is easy to see: a classical description of an object falling
: into a black hole involves only general relativity, and not quantum
: mechanics, so the time-reversed version must also involve only general
: relativity. Hawking radiation, on the other hand, is *fundamentally*
: quantum mechanical; it depends explicitly on Planck' constant, which is
: not there in classical general relativity at all.
No - that doesn't demonstrate that you cannot predict from reversibility
that things will emerge from black holes. You can - and sure enough they
do.
All that matters to be able to make this prediction is reversibility.
The details of the laws of physics (be they QM or GR) dictate the form of
the emerged material - but not whether material emerges or not - that is
absolutely inevitable if the physics is reversible.
: (Moreover, as I pointed out earlier, Hawking radiation does not
: appear to be time-reversal-invariant.)
You pointed out that some folks were engaged in discussion of the issue.
These processes will prove to be reversible - as all the other laws
of physics are.
You described the black hole as low information state (or some such) -
perhaps on the grounds that you could only measure its mass, spin and
charge. However that is plainly false, - as Hawking effectively
demonstrated.
The idea that information is destroyed when matter enters the black hole
and is then created again when the black hole evaporates might be
effectively true for an embedded observer - but looks ridiculous when
considering the complete state.
Tim Tyler
:> : So while you can (perhaps, depending on black hole evaporation
:> : and similar processes) define a conserved ``information'' for the
:> : Universe as a whole, you do so only by losing any usefulness of
:> : the concept. As soon as you try to define *local* information,
:> : the information of a subsystem, you lose conservation.
:> The notion is still applicable to subsystems, though - as it is with
:> energy conservation.
:> Even in an open system, you can still apply conservation laws
:> in aspecified volume - if you take into account the flux of the
:> conserved quantity through the surrounding surface.
: This is not true for the entropy of a subsystem.
Thermodynamic entropy may not be conserved - *even* if you consider the
whole system.
I'll try another way of expressing my point:
You said that attempts to try and apply conservation of information
principles locally could only be done by "losing any usefulness of
the concept".
I don't see it that way at all. Conservation of information principles
place tight constraints on what physics is locally possible.
If they hold it means that you can rule out any behaviour that
results in creation or destruction of information - and leads to
irreversibility.
: It isn't actually even true for energy, at least in general relativity,
: where a local energy density cannot be defined. (Gravitational
: energy is nonlocalizable.) You might want to look at the relativity
: FAQs,
: http://www.math.ucr.edu/home/baez/physics/Relativity/GR/energy_gr.html
Complications. AFAICS, it doesn't fit your synopsis well. The nearest
thing I could find was:
``For these reasons, most physicists who work in general relativity do not
believe the pseudo-tensors give a good local definition of energy
density, although their integrals are sometimes useful as a measure of
total energy.''
FWIW, I don't think anything is "nonlocalizable", EPR notwithstanding.
John Wilkins
Yes. It seems to me that if the probability of a state being reversed is
X, then the probability of all states being reversed is the product of
all the Xs. So as the number of state reversals required increases, so
too does the probability diminish, approaching zero as we get to very
large numbers. For a *single* event of particles mixing (as in the gas
or ink cases) the probability of the state being reversed would tend to
exceed the number of state changes left in a finite universe.
But all this is a priori reasoning on my part, so I leave this to the
competent.
>
>
> [...]
>
>
> >This is a matter of philosophical debate. However, on this matter I
> >agree with you. For me, information is a semantic construct in a natural
> >language,
>
> <splork> _Natural_ language?
Yep (I'm glad you got this). The *existence* of information is
necessarily in a natural language. This is because no formal languages
actually exist. That is to say, you cannot (in my ontology) assert
Thereexists(x)(Lx)
Where Lx is "x is a formal language", because, as I never tire of
saying, abstract objects do not exist (in the physical universe). So if
information exists, it is in a natural language, while if it is a
property of a formal language, it does not exist. Such information as
exists is in a natural language :-)
It might help if we therefore distinguish two kinds of information: Real
Information, which is not real, and Actual Information, which is not
information. Or less risibly, Formal Information is never concrete,
while concrete information is not the property described by the various
Formal measures of Information.
To better understand what I mean, do a Google search on Ed Zalta
Principia Metaphysica. Zalta's "theory of abstract objects" has them as
all the objects (ie, predicable variables) that are not bound by
spacetime. Concrete objects are all those that *are* bounded in
spacetime. When I say that something "exists", I am asserting that it is
bounded in spacetime, that's all. But I like to be gnomic.
>
>
....
> I generally go for Brazils and hazelnuts this time of year.
> With the odd lump of crystallized ginger. How 'bout you?
>
> [...]
I avoid nuts after new year's, myself... But if I must - mmmm,
pistachio.
Steve Carlip
>I am claiming that radiation from black holes could have
> been predicted solely on the grounds of reversibility.
Well, you're clearly wrong, then. The time reversal of an
object falling into a black hole is *not* an object coming
out of a black hole; it's an object coming out of a white
hole. The classical Einstein field equations are time-
reversal invariant, and yet do not allow anything to
emerge from a black hole.
(Remember, if you have an interacting theory, you can't
just time-reverse *one* trajectory; you have to time-reverse
everything. When you do that, you don't have a black hole
any more. You need to keep in mind that the interior of
a black hole, inside the horizon, is not a static geometry;
time reversal makes the geometry run backwards, and
gives you something that is no longer a black hole.)
> Reversibility disctates that if things can go
> into a black hole other things must be able to come out.
No, it doesn't. You are misunderstanding some fairly basic
physics. Reversibility dictates that there is *some* solution
of the equations of motion that is a time reversal of an
object falling into a black hole. But that solution simply
isn't something coming out of a black hole. This is not
hard---just take the equations and actually do the t->-t
transformation, rather than guessing.
> You described the black hole as low information state
I certainly did not. A black hole has a very high entropy.
Steve Carlip
R. Baldwin
> R. Baldwin <res0k7y...@verizon.net> wrote or quoted:
>
> :> Didn't I just answer this above? The universe is reversible - as
> :> far as decades of work by physicists can tell. Reversibility
implies
> :> - and is implied by - conservation of information.
>
> : No, you didn't. I was pointing out your error in stating that the
> : Principle of Conservation of Energy is "based on an unproven
> : hypothesis." Such a term is not meaningful in describing a
> : generalization of observation.
>
> I don't recall stating that. Can you quote me doing so?
Gladly. In your post of Jan. 2:
My statement:
: The Principle of Conservation of Energy is a generalization of
: experience, so far not contradicted by observation of nature. It has
: years of research backing it up. What research provides the basis of
: the "Law of Conservation of Information?" Show your work.
Your reply:
It is based on an unproven hypothesis. We do not know for certain
whether
the law exists or not. The fact that the laws of physics are
symmetrical
under an operation involving inverting time strongly suggests that no
information is created or destroyed as time passes.
>
> : I do not agree with you that microscopic reversibility implies
that
> : the universe is reversable.
>
> Curious. How can things be reversible on one scale and irreversible
on
> another. Such a position is not self-consistent.
It is perfectly consistent for anyone who doesn't accept your
nonsensical notion of Conservation of Information. From reading all
the other posts in this thread, it seems all the others also disagree
with you about the existence of irreversible macroscopic processes.
Tim Tyler
: Tim Tyler <t...@iname.com> wrote:
:>I am claiming that radiation from black holes could have
:> been predicted solely on the grounds of reversibility.
: Well, you're clearly wrong, then. The time reversal of an
: object falling into a black hole is *not* an object coming
: out of a black hole; it's an object coming out of a white
: hole. The classical Einstein field equations are time-
: reversal invariant, and yet do not allow anything to
: emerge from a black hole.
It is not possible for reversible equations to model a system which
allows things in and does not allow them out. That is an irreversible
system - by definition. The point is really very simple. I have no idea
what your problem is in grasping it.
Perhaps considering a finite reversible system - that will traverse the
same path repeatedly forever will help you understand this - I don't know.
: (Remember, if you have an interacting theory, you can't
: just time-reverse *one* trajectory; you have to time-reverse
: everything. When you do that, you don't have a black hole
: any more.
To the extent that an anti-matter black hole which exudes material
(by effectively running backwards) is called a white hole, this is
true - but is irrelevant to my point.
:> Reversibility disctates that if things can go
:> into a black hole other things must be able to come out.
: No, it doesn't. [...]
It does, absolutely.
: You are misunderstanding some fairly basic physics.
: Reversibility dictates that there is *some* solution
: of the equations of motion that is a time reversal of an
: object falling into a black hole. But that solution simply
: isn't something coming out of a black hole. This is not
: hard---just take the equations and actually do the t->-t
: transformation, rather than guessing.
I am in complete agreement that a CPT reversal produces what is known
as a white hole - yet my point appears totally unaffected by this.
Holes do /not/ need to be made of anti-matter to exude material.
Run a system with a black hole in it for long enough and as much goes in
as comes out. This is an immediate and direct consequence of
reversibility.
Perhaps I'm beginning to understand why the phenomenon was not predicted
eariler: people don't intuitively understand what the consequences of
reversibility actually are.
:> You described the black hole as low information state
: I certainly did not. A black hole has a very high entropy.
OK - perhaps I misunderstood what you meant by "pure" in:
``consider the following process: matter in a pure quantum state
collapses to form a black hole, which then evaporates by Hawking
radiation. If, as all evidence suggests, Hawking radiation is thermal,
this represents an evolution from a pure state to a mixed state,''
Tim Tyler
: "Tim Tyler" <t...@iname.com> wrote in message
:> R. Baldwin <res0k7y...@verizon.net> wrote or quoted:
:> :> Didn't I just answer this above? The universe is reversible - as
:> :> far as decades of work by physicists can tell. Reversibility
:> :> implies - and is implied by - conservation of information.
:>
:> : No, you didn't. I was pointing out your error in stating that the
:> : Principle of Conservation of Energy is "based on an unproven
:> : hypothesis." Such a term is not meaningful in describing a
:> : generalization of observation.
:>
:> I don't recall stating that. Can you quote me doing so?
: Gladly. In your post of Jan. 2:
: My statement:
: : The Principle of Conservation of Energy is a generalization of
: : experience, so far not contradicted by observation of nature. It has
: : years of research backing it up. What research provides the basis of
: : the "Law of Conservation of Information?" Show your work.
: Your reply:
: It is based on an unproven hypothesis. We do not know for certain
: whether the law exists or not. The fact that the laws of physics are
: symmetrical under an operation involving inverting time strongly
: suggests that no information is created or destroyed as time passes.
I referred to the law of conservation of *information* being an unproved
hypothesis. Your interpretation was not what I intended at all.
We know that the laws of physics are *almost* exactly reversible - in
every case we have looked at. However, the law requires that they be
*exactly* reversible - in every single case. While there is room for
doubt I describe the hypothesis as "unproven".
:> : I do not agree with you that microscopic reversibility implies
:> : that the universe is reversable.
:>
:> Curious. How can things be reversible on one scale and irreversible
:> on another. Such a position is not self-consistent.
: It is perfectly consistent for anyone who doesn't accept your
: nonsensical notion of Conservation of Information.
No -scale is irrelevant. Either a system is reversible, or it is not.
: From reading all the other posts in this thread, it seems all the
: others also disagree with you about the existence of irreversible
: macroscopic processes.
Those other people seem confused over terminology. They mistake
processes that increase thermodynamic entropy for irreversible
processes.
Reversibility has a precise meaning when applied to dynamical systems.
That meaning is quite consistent with the 2LoT producing what appear to
be time-asymmetrical behaviour during evolution from a low entropy state.
Observation of such behaviour does not lead to the conclusion that the
dynamical system is irreversible.
Microsocopic reversibility, if true, implies that the universe is
reversable - it is exactly as simple as that.
Tim Tyler
: It seems to me that if the probability of a state being reversed is
: X, then the probability of all states being reversed is the product of
: all the Xs. So as the number of state reversals required increases, so
: too does the probability diminish, approaching zero as we get to very
: large numbers.
Yes - the probability of it happening in any time step gets smaller -
so you need more time to observe it happening...
: For a *single* event of particles mixing (as in the gas or ink cases)
: the probability of the state being reversed would tend to exceed the
: number of state changes left in a finite universe.
No - that would be an impossibility.
Consider a finite reversible system (e.g. the one you mentioned) and look
at the chance of it reaching the same state it was in ten minues
ago. That probability is precisely 1.
Exact reversal of state is *not* a miniscule probability - despite the
fact that your sums above still apply - rather, it is inevitable. All you
need is enough time.
R. Baldwin
>
> It is perfectly consistent for anyone who doesn't accept your
> nonsensical notion of Conservation of Information. From reading all
> the other posts in this thread, it seems all the others also
disagree
processes.
>
>
(i.e., the resist of us assert they exist.)
R. Baldwin
OK, I see what you meant by this now.
>
> We know that the laws of physics are *almost* exactly reversible -
in
> every case we have looked at. However, the law requires that they
be
> *exactly* reversible - in every single case. While there is room
for
> doubt I describe the hypothesis as "unproven".
>
> :> : I do not agree with you that microscopic reversibility implies
> :> : that the universe is reversable.
> :>
> :> Curious. How can things be reversible on one scale and
irreversible
> :> on another. Such a position is not self-consistent.
>
> : It is perfectly consistent for anyone who doesn't accept your
> : nonsensical notion of Conservation of Information.
>
> No -scale is irrelevant. Either a system is reversible, or it is
not.
>
> : From reading all the other posts in this thread, it seems all the
> : others also disagree with you about the existence of irreversible
> : macroscopic processes.
>
> Those other people seem confused over terminology. They mistake
> processes that increase thermodynamic entropy for irreversible
> processes.
It is you that appear confused over terminology. See exerpts below.
>
> Reversibility has a precise meaning when applied to dynamical
systems.
>
> That meaning is quite consistent with the 2LoT producing what appear
to
> be time-asymmetrical behaviour during evolution from a low entropy
state.
>
> Observation of such behaviour does not lead to the conclusion that
the
> dynamical system is irreversible.
>
> Microsocopic reversibility, if true, implies that the universe is
> reversable - it is exactly as simple as that.
> --
> __________
> |im |yler Index of my domains: http://timtyler.org/ t...@iname.com
>
No, microscopic reversibility does not imply the universe is
reversible if by that you mean destruction of individual macroscopic
structures must be reversible. It is not a simple matter of scale.
Regarding irreversibility, here are some quotations out of the
Halliday and Resnick college:
"The second law tells us that many processes are irreversible. For
example, Clausius' statement specifically rules out a simple reversal
of the process of heat transfer from hot body to cold body. Not only
will some processes not run backward by themselves, but no combination
of processes can undo the effect of an irreversible process without
causing another corresponding change elsewhere."
[Clausius stated: It is impossible for any cyclical machine to produce
no other effect than to convey heat continuously from one body to
another at a higher temperature.]
"1. Free expansion... let a gas double its volume by expanding into an
evacuated enclosure... The free expansion is certainly irreversible
because we lose control of the environment once we turn the
stopcock...
"2. Heat conduction. For another example consider two bodies that are
similar in every respect except that one is at a temperature T1 and
the other at temperature T2, where T1 > T2. If we put both objects in
contact inside a box with nonconducting walls, they will eventually
reach a temperature Tm, somewhere between T1 and T2. Like the free
expansion, the process is irreversible because we lose control of the
environment once we put the two bodies in the box...
"we assert that the second law is: A natural process that starts in
one equilibrium state and ends in another will go in the direction
that causes the entropy of the system plus environment to increase.
"The two experiments... are consistent with the second law. The
entropy of the system increased in each of these irreversible
processes. Note that the entropy of the environment in each of these
two cases remains unchanged because, both being carried out in
adiabatic enclosures, there was no interchange of heat with the
environment...
"In the form that we have written it the second law applies only to
irreversible processes because such processes have a "natural
direction." Indeed... the understanding of the natural directions of
such processes is the main concern of the second law. Reversible
processes can go equally well in either direction, however, and for
reversible processes the entropy of the system plus environment
remains unchanged."
One consequence is that the formula Sb - Sa = Integral from a to b of
dQ/T only applies for reversible processes. For an irreversible
process the calculation cannot be made.
Now if there are multiple paths from state a to state b, some
irreversible and some reversible, you can determine delta S for a
system by running the calculation on a reversible path. This is
because S is a state variable, and does not vary by the path it took.
On this topic, H&R reads:
"In each of these examples we must distinguish carefully between the
actual (irreversible) process (free expansion or heat conduction) and
the reversible process we introduce just so that we can calculate the
entropy change in the actual process. We can choose any reversible
process, as long as it connects the same initial and final state as
the actual process; all such reversible processes will yield the same
entropy change because this depends only on the initial and final
states and not on the process connecting them- be it reversible or
irreversible."
Note that this does NOT tell us that such reversible processes
necessarily exist between two states of a system, a and b. It would be
fair to state that all paths between a newly-created piece of Chiluly
glass and the same Chiluly glass shattered on the floor are
irreversible.
Tim Tyler
: "Tim Tyler" <t...@iname.com> wrote or quoted:
:> : From reading all the other posts in this thread, it seems all the
:> : others also disagree with you about the existence of irreversible
:> : macroscopic processes.
:>
:> Those other people seem confused over terminology. They mistake
:> processes that increase thermodynamic entropy for irreversible
:> processes.
: It is you that appear confused over terminology. See exerpts below.
[...]
: No, microscopic reversibility does not imply the universe is
: reversible if by that you mean destruction of individual macroscopic
: structures must be reversible. It is not a simple matter of scale.
An irreversible macroscopic phenomenon must depend for its existence
on irreversible microscopic phenomena. If there are none of the latter,
there can be none of the former.
: Regarding irreversibility, here are some quotations out of the
: Halliday and Resnick college:
: "The second law tells us that many processes are irreversible. For
: example, Clausius' statement specifically rules out a simple reversal
: of the process of heat transfer from hot body to cold body.
This is ridiculous. The second law is ***not*** a law of physics.
Indeed it is not an absolute law at all - since violations are quite
possible. It *can't* be used to rule out any behaviour.
It's sadenning to see this book dumbing things down so far that it gets
them wrong :-(
: [Clausius stated: It is impossible for any cyclical machine to produce
: no other effect than to convey heat continuously from one body to
: another at a higher temperature.]
How go you get an "impossibility" out of a statistical phenomenon?
What this should say is that "it is very unlikely to observe a cylical
machine producing no other effect than conveying heat continuously from
one body to another at a higher temperature." As it stands, it is a
straightforwards error.
: "1. Free expansion... let a gas double its volume by expanding into an
: evacuated enclosure... The free expansion is certainly irreversible [...]
No, it is likely to be exactly reversible - like every other physical
phenomenon that has ever been stuided.
: "2. Heat conduction. For another example consider two bodies that are
: similar in every respect except that one is at a temperature T1 and
: the other at temperature T2, where T1 > T2. If we put both objects in
: contact inside a box with nonconducting walls, they will eventually
: reach a temperature Tm, somewhere between T1 and T2. Like the free
: expansion, the process is irreversible [...]
No. What they /ought/ to say is that they can't reverse them - not that
they are irreversible.
: "In the form that we have written it the second law applies only to
: irreversible processes because such processes have a "natural
: direction."
This is garbage. The second law works just fine in reversible systems
with ordered initial states.
The second law has little to do with actual reversibility - and analogues
arise naturally in both reversible and irreversible systems.
I've previously mentioned an exactly-reversible system which clearly
exhibits the second law: http://hex.org.uk/diffusion/
Bigdakine
>From: Tim Tyler t...@iname.com
>Date: 1/4/02 8:33 PM Hawaiian Standard Time
>Message-id: <GpGCw...@bath.ac.uk>
Except that the reversibility of G.R breaks down at the singularity. You can't
use G.R. to say anything about what happens there.
Stuart
Dr. Stuart A. Weinstein
Ewa Beach Institute of Tectonics
"To err is human, but to really foul things up
requires a creationist"
Tim Tyler
:>:> Reversibility disctates that if things can go
:>:> into a black hole other things must be able to come out.
:>
:>: No, it doesn't. [...]
:>
:>It does, absolutely.
: Except that the reversibility of G.R breaks down at the singularity.
This is about the fact that Hawking radiation (or similar) can be
predicted on the grounds of reversibility. I.e. we are taking
reversibility for granted - and seeing what follows.
If finite nature is right, its likely that no singularities form
in the first place. They represent the equivalent of division by
zero in physics.
If finite nature is correct, GR /must/ break down on large scales.
We can be pretty certain GR is not the final theory on other
grounds - including its lack of integration with quantum theory.
Consequently I don't regard the question of whether reversibility
or GR breaks gown at singularities or not as being very interesting -
it's about on par with the angels on the head of the proverbial pin.
R. Baldwin
> R. Baldwin <res0k7y...@verizon.net> wrote:
> : "Tim Tyler" <t...@iname.com> wrote or quoted:
>
> :> : From reading all the other posts in this thread, it seems all
the
> :> : others also disagree with you about the existence of
irreversible
> :> : macroscopic processes.
> :>
> :> Those other people seem confused over terminology. They mistake
> :> processes that increase thermodynamic entropy for irreversible
> :> processes.
>
> : It is you that appear confused over terminology. See exerpts
below.
>
> [...]
>
> : No, microscopic reversibility does not imply the universe is
> : reversible if by that you mean destruction of individual
macroscopic
> : structures must be reversible. It is not a simple matter of scale.
>
> An irreversible macroscopic phenomenon must depend for its existence
> on irreversible microscopic phenomena. If there are none of the
latter,
> there can be none of the former.
No, an irreversible macroscopic process depends on a path between two
states that cannot be traced backward. For example, the path can be
through a series of nonequilibrium states.
You are also making false assumptions about the relationship between
macroscopic and microscopic objects. A macroscopic crystal goblet is
not made up of a collection of microscopic crystal goblets. The
microscopic components of the goblet have no properties that tend
toward gobletness.
>
> : Regarding irreversibility, here are some quotations out of the
> : Halliday and Resnick college:
>
> : "The second law tells us that many processes are irreversible. For
> : example, Clausius' statement specifically rules out a simple
reversal
> : of the process of heat transfer from hot body to cold body.
>
> This is ridiculous. The second law is ***not*** a law of physics.
What a curious statement. Thermodynamics is taught in physics,
chemistry, and engineering. It is common to all three subjects. The
historical origin of the second law lay with heat engines, which is
very much a concern of physics.
>
> Indeed it is not an absolute law at all - since violations are quite
> possible. It *can't* be used to rule out any behaviour.
Well, H&R does acknowledge this: "it is not absolutely certain that
the entropy increases in every spontaneous process. The entropy may
sometimes decrease. If we waited long enough, even the most improbable
states might occur: the water in a pond suddenly freezeing on a hot
summer day or a local vacuum occuring suddenly in a room. Although
such occurences are possible, the probability of their happening, when
computed, turns out to be incredibly small. Hence, the second law of
thermodynamics shows us the most probable course of events, not the
only possible ones. But its area of application is so broad and the
chance of nature's contradicting it so small that it occupies the
distinction of being one of the most useful and general laws in all
science."
>
> It's sadenning to see this book dumbing things down so far that it
gets
> them wrong :-(
It's sadenning to see you beating a dead horse by arguing for physical
laws that have no evidence backing them up and much to the contrary,
and against physical laws that have a tremendous amount of supporting
evidence but none to the contrary.
H&R is one of the most commonly used freshman/sophmore level physics
textbooks. If it were so dumbed down that it got things wrong, physics
professors around the world would not favor it.
>
> : [Clausius stated: It is impossible for any cyclical machine to
produce
> : no other effect than to convey heat continuously from one body to
> : another at a higher temperature.]
>
> How go you get an "impossibility" out of a statistical phenomenon?
See above. We would say that a pond suddenly freezing on a hot August
afternoon is impossible. We all know that it is conceivably possible
that all of the water molecules might suddenly align into ice, but the
odds of this occuring are so small that we rule it out.
Apparently you did not understand the quote. The point in H&R is that
entropy of system plus environment increases when the system's
processes are irreversible and remains constant when they are
reversible. It doesn't imply the second law does not work for
reversible systems; rather, that for reversible systems we will have
zero total entropy change in system plus environment. H&R stated the
second law in a form that assumes irreversible systems are
contemplated, then explained how to understand that phrasing in the
context of reversible systems.
>
> The second law has little to do with actual reversibility - and
analogues
> arise naturally in both reversible and irreversible systems.
You seem to be acknowledging the existence of irreversible systems,
which you previously denied existed.
The second law has everything to do with reversibility and
irreversibility.
Example of irreversible process in chemistry: primary voltaic cells.
Once the chemicals are consumed, they cannot be restored to their
initial state by application of an external electrical potential. In
other words, you can't return to the initial state from the final
state by following the same path backward.
>
> I've previously mentioned an exactly-reversible system which clearly
> exhibits the second law: http://hex.org.uk/diffusion/
I looked at this - I could find no explanation along with the
simulation.
Tim Tyler
: "Tim Tyler" <t...@iname.com> wrote in message
:> :> : From reading all the other posts in this thread, it seems all
:> :> : the others also disagree with you about the existence of
:> :> : irreversible macroscopic processes.
:> :>
:> :> Those other people seem confused over terminology. They mistake
:> :> processes that increase thermodynamic entropy for irreversible
:> :> processes.
:>
:> : It is you that appear confused over terminology. See exerpts
:> : below.
:>
:> [...]
:>
:> : No, microscopic reversibility does not imply the universe is
:> : reversible if by that you mean destruction of individual
: macroscopic
:> : structures must be reversible. It is not a simple matter of scale.
:>
:> An irreversible macroscopic phenomenon must depend for its existence
:> on irreversible microscopic phenomena. If there are none of the
:> latter, there can be none of the former.
: No, [...]
Yes.
: an irreversible macroscopic process depends on a path between two
: states that cannot be traced backward.
Indeed. You disgree and then say the same thing another way.
: You are also making false assumptions about the relationship between
: macroscopic and microscopic objects. A macroscopic crystal goblet is
: not made up of a collection of microscopic crystal goblets.
You concluded I thought that from what I had written how?
Reversibility of a dynamical system is contingent upon - and
an immediate consequence of - reversibility of all its components.
That is not remotely the same as saying that a goblet is made up of lots
of tiny goblets.
:> : Regarding irreversibility, here are some quotations out of the
:> : Halliday and Resnick college:
:>
:> : "The second law tells us that many processes are irreversible. For
:> : example, Clausius' statement specifically rules out a simple
:> : reversal of the process of heat transfer from hot body to cold body.
:>
:> This is ridiculous. The second law is ***not*** a law of physics.
: What a curious statement.
The second law is the outcome of statistics, not physical law.
It is not a "law" - in that it can be violated - while most other
physical laws cannot be. It might be described as a tendency -
rather than a law.
Other people have described this perspective:
``The fact that the Second Law of Thermodynamics applies only on the
macroscopic level ( unlike the First Law, conservation of mass-energy,
which applies on all levels ) suggests to me that the Second Law is a
principle of statistics, not a law of physics. I think that James Clerk
Maxwell clearly understood this [...] The so-called "Second
Law" is an expression of the observation that statistical systems tend
towards their most probable state.''
- http://csf.colorado.edu/forums/longwaves/jan00/msg00272.html
:> It's sadenning to see this book dumbing things down so far that it
:> gets them wrong :-(
: It's sadenning to see you beating a dead horse by arguing for physical
: laws that have no evidence backing them up
Not true. Evidence for reversibilty exists. We know that physics is
*extremely* close to exact reversiblity. No experimental results appear
to violate reversibility and the laws of physics reflect this by
exhibiting CPT symmetry.
These events did not just chance to be - they are the outcome of decades
of careful experimentation.
My position is not that the world is reversible. It is that we can't
say whether it is reversible or not. *Assuming* that it is irreversible -
and that information is created - is not a legitimate step in the light of
this.
: and much to the contrary [...]
No - no evidence as far as I am aware. Can you cite an experiment that
indicates that the world is not a reversible system?
: and against physical laws that have a tremendous amount of supporting
: evidence but none to the contrary.
What are you referring to? I am not in disagreement with any
well-established physical law, AFAIK - except perhaps general
relativity when travelling very close to the speed of light -
but physics there is hardly "well-established" or supported
by "a tremendous amount of" evidence.
: H&R is one of the most commonly used freshman/sophmore level physics
: textbooks. If it were so dumbed down that it got things wrong [...]
So - you argue for its correctness based on its sales, rather than on what
it actually said. Errors occur in almost every textbook - it is
practically inevitable.
What it said was not strictly correct - thermodynamically unlikely
behaviour can't be ruled out, nor can you conclude that a dynamical system
is reversible by observing that it started out in a low-entropy state.
To give it some of the benefit of the doubt, perhaps it was using the
term "rule out" loosely.
:> : [Clausius stated: It is impossible for any cyclical machine to
:> : produce no other effect than to convey heat continuously from one
:> : body to another at a higher temperature.]
:>
:> How go you get an "impossibility" out of a statistical phenomenon?
: See above. We would say that a pond suddenly freezing on a hot August
: afternoon is impossible. [...]
*You* might say that - but you would be incorrect.
: We all know that it is conceivably possible that all of the water
: molecules might suddenly align into ice, but the odds of this occuring
: are so small that we rule it out.
The odds of it happening in any time period are indeed small. However -
given long enough - the chances of it happening become overwhelming.
:> The second law has little to do with actual reversibility - and
:> analogues arise naturally in both reversible and irreversible systems.
: You seem to be acknowledging the existence of irreversible systems,
: which you previously denied existed.
They exist as conceptual entities. We can easily imagine a system that
destroys information, simulate it on computers and so forth.
Many natural systems are irreversible - ***IF*** you completely ignore
thermal noise, radiation, and so on.
: The second law has everything to do with reversibility and
: irreversibility.
: Example of irreversible process in chemistry: primary voltaic cells.
: Once the chemicals are consumed, they cannot be restored to their
: initial state by application of an external electrical potential.
Indeed not - though many batteries do charge up somewhat if you do this.
: In other words, you can't return to the initial state from the final
: state by following the same path backward.
No that is a completely different statement to the one above.
Following the same path backwards would restore the initial state.
You are simply failing in your crude attempt to drive the system backwards
by reversing the polarity across the terminals, since the rest of the
system is not reversed as well.
Instead of reversing the polarity, you need to turn the battery into
anti-matter, reverse the direction of motion of all its atoms, and
invert the charge on all the particles.
Presumably you are aware of this.
:> I've previously mentioned an exactly-reversible system which clearly
:> exhibits the second law: http://hex.org.uk/diffusion/
: I looked at this - I could find no explanation along with the
: simulation.
There is a description if you press "page down".
The important things to note are:
* The system is reversible, and van be run backwards at a touch;
* The system starts in an ordered state;
* The system exhibits the second law of thermodynamics - i.e. ordered
initial states rapidly disintegrate into what appears to be chaos;
* That chaos can be returned to order by running the system backwards.
Running the system backwards looks something like a miracle is occurring.
The system forms into a coherent structure from what appears to be total
chaos.
The behavious is directly analogous to the smashed vase forming into a
whole object and rising to its pedestal again.
R. Baldwin
information per second as javalog.txt under the windows/java
directory, at least in Win98.
Yesterday it used up all the available space on my hard drive. It took
hours to figure out what happened and recover.
Rather ironic, that Tim's web site creates that much information.
[SNIP]
R. Baldwin
alternate universe where the physical laws are different from our own.
"Tim Tyler" <t...@iname.com> wrote in message
news:GpIr8...@bath.ac.uk...
How?
>
> That is not remotely the same as saying that a goblet is made up of
lots
> of tiny goblets.
>
> :> : Regarding irreversibility, here are some quotations out of the
> :> : Halliday and Resnick college:
> :>
> :> : "The second law tells us that many processes are irreversible.
For
> :> : example, Clausius' statement specifically rules out a simple
> :> : reversal of the process of heat transfer from hot body to cold
body.
> :>
> :> This is ridiculous. The second law is ***not*** a law of
physics.
>
> : What a curious statement.
>
> The second law is the outcome of statistics, not physical law.
That hardly makes it of no utility to physics. Pressure is also an
outcome of statistics, but you could hardly teach physics without
discussing pressure.
>
> It is not a "law" - in that it can be violated - while most other
> physical laws cannot be. It might be described as a tendency -
> rather than a law.
>
> Other people have described this perspective:
>
> ``The fact that the Second Law of Thermodynamics applies only on the
> macroscopic level ( unlike the First Law, conservation of
mass-energy,
> which applies on all levels ) suggests to me that the Second Law
is a
> principle of statistics, not a law of physics. I think that James
Clerk
> Maxwell clearly understood this [...] The so-called "Second
> Law" is an expression of the observation that statistical systems
tend
> towards their most probable state.''
>
> - http://csf.colorado.edu/forums/longwaves/jan00/msg00272.html
So... at Verdun, there was some probability that the mustard and
phosgene gasses could implode back into their containers, which could
seal up, and of their own volition follow a trajectory back to the
artillery muzzles, in which the hot gases could then cool down and
become solid propellent. The bullets could likewise fling themselves
out of the ground, fly back through the rising, reanimating corpses
and return to the machine gun muzzles from whence they came. Fine.
Show us evidence that such things occur.
In other words, you follow a reversible path from the final state to
the initial state. There is a cycle. That doesn't get around the point
that the path taken from initial state to final state was
irreversible. You have a cycle that will require you to put work into
the system because entropy of the universe increased every time you
progressed one way down the irreversible path.
>
> :> I've previously mentioned an exactly-reversible system which
clearly
> :> exhibits the second law: http://hex.org.uk/diffusion/
>
> : I looked at this - I could find no explanation along with the
> : simulation.
>
> There is a description if you press "page down".
>
> The important things to note are:
>
> * The system is reversible, and van be run backwards at a touch;
> * The system starts in an ordered state;
> * The system exhibits the second law of thermodynamics - i.e.
ordered
> initial states rapidly disintegrate into what appears to be chaos;
> * That chaos can be returned to order by running the system
backwards.
>
> Running the system backwards looks something like a miracle is
occurring.
>
> The system forms into a coherent structure from what appears to be
total
> chaos.
>
> The behavious is directly analogous to the smashed vase forming into
a
> whole object and rising to its pedestal again.
How is it analogous? Did you model solid objects?
Jonathan Stone
Oh, much more than that. We defy Tim to take a walk out of Alan
Sokal's 20-floor window.
Jonathan Stone
Ah. Not just an ignoramus, a prize-winning ignoramus.
Could someone give Tim a gentle reminder to review the difference
between drunkards' walks in two dimensions and in three dimensions?
Jonathan Stone
John Wilkins <john.w...@bigpond.com> wrote:
>Jonathan Stone <jona...@DSG.Stanford.EDU> wrote:
>Yep (I'm glad you got this). The *existence* of information is
>necessarily in a natural language. This is because no formal languages
>actually exist. That is to say, you cannot (in my ontology) assert
>
>Thereexists(x)(Lx)
>
>Where Lx is "x is a formal language", because, as I never tire of
>saying, abstract objects do not exist (in the physical universe). So if
>information exists, it is in a natural language, while if it is a
>property of a formal language, it does not exist. Such information as
>exists is in a natural language :-)
John,
I can have no truck with this, er, this natural-language fascism.
Abstract objects exist in the abstract, and we can and do compute the
_abstract_ information in the _abstract_ object. It is, indeed, a
common exercise for a first course in computer networking. Nyquist,
and Shannon, and all that stuff.
I can also point to a data-storage device and describe its information
content as some fixed number of bits. For suitable storage devices, we
can (with suitable tools) go look at the individual solid-state
devices which comprise each bit of storage.
It's a fact of life (and of natural language) that the
natural-language units do not carefully distinguish between
data and information.
I think what you're after is an _interpretation_ of some bitstring in
some natural language. But then what about mathematical formulae?
Do you insist they have no informational content? Or do you backpedal
and say that mathematical notation (plus conventions of unbound
variables in some appropriate domain) is a ``natural language''?
That last one does huge violence to what linguists and other
philosophers take "natural language" to mean.
Anyways, there are well-deveopled mechanisms for finding the
"information" in non-natural languages. Just ask anyone who's done
research to detect cheating in undergrad programming assignments ;).
Uh -- didn't Chaitin use Lisp rather than UTMs for AIC? If so, I dont
see how you can advance your opinion with a straight face.
John Wilkins
Lisp *is* a natural language :-) At least, the one on my computer, and
on Chaitin's computer, is. Is there some universal abstract LISP out
"there" somewhere? Also, Chaitin's interpreter in LISP exists on his
website in two versions: Mathematica and C. Which is "the one true
LISP"? Here's what Chaitin says about it:" To me a formal axiomatic
system is like a black box that theorems come out of. My methods for
obtaining incompleteness theorems are so general that all I need to know
is that there's a proof-checking algorithm. ... The idea is to start
with a small number of powerful concepts, and get everything from that.
So LISP, the way I do it, is more like mathematics than computing." _The
Unknowable_ p29-30
At least Chaitin is aware of the distinction between a formal language
(maths) and a practical, natural, one (computing).
Oh, Okay, we have to distinguish between two things here (and this is
just rehashing the Platonist interpretation of maths versus the
nominalist one) - the formal standing of any object, and how it is
instantiated. This is an old, old debate. I am taking one of a number of
possible metaphysical positions on it - in this case, the nominalist
one.
An abstract object does not, in my view, actually exist. So I have the
problem of where to locate abstractions. I locate them in semantic
systems (natural languages). That's all. Nothing too abtruse about it.
We approximate abstract properties in our natural languages, and one of
those abstractions is formal language. Including the abstract LISP :-)
I always wanted to be a natural language fascist, although I prefer the
term "naturalist" (a la Quine, not Darwin).
--
John Wilkins
Occasionally being an NL fascist for over 46 years
Tim Tyler
: The web site cited below by Tim Tyler will create megabytes of
: information per second as javalog.txt under the windows/java
: directory, at least in Win98.
: Yesterday it used up all the available space on my hard drive. It took
: hours to figure out what happened and recover.
: Rather ironic, that Tim's web site creates that much information.
I believe this is a Microsoft Windows 9x JVM issue.
The problem is their thread support (or lack of a correct
implementation of it) - in conjunction with allowing applets
unlimited resource usage - exacerbated by allowing them to
persist in an active state once the web page they are on
has been closed.
I can (and have) worked around the problem.
Tim Tyler
: This is my last post on the topic. [...]
Yet you continue to ask me questions? Oh well, bye bye.
Kim G. S. OEyhus
>Kim G. S. OEyhus <k...@pvv.ntnu.no> wrote:
>: In article <Gp9qz...@bath.ac.uk>, Tim Tyler <t...@iname.com> wrote:
>
>:>I tried to explain above how information creation and destruction was
>:>prevented by physical law. Were this not so, the universe would not
>:>be microscopically reversible - and as far as we can tell - it is.
>
>: It seems you are using "entropy" and "information" as synonyms [...]
>
>They have different connotations - for example entropy can be defined
>in terms of heat - while information cannot.
Actually, informations can be defined in terms of heat, f.ex. through
algorithmic information theory, with turing machines composed of
zero-entropy friedkin gates, and stuff. But then one gets a theory of
the cost of computing measures of information.
Jonathan Stone
>Lisp *is* a natural language :-)
No, it isn't. This is one of those basic definitional things.
You are most welcome to invent your own private language, and even to
use it on Usenet. But don't try and pretend you're using English or
that you have any desire to communicate with the rest of humanity.
>At least, the one on my computer, and
>on Chaitin's computer, is. Is there some universal abstract LISP out
>"there" somewhere?
Boggle. Yes, of course there is. It's called lambda-calculus, or pure
lisp. Lambda calculus is Church's original formal mathematical
language. So-called "Pure lisp" or Scheme is nothing but syntactic
sugar for lambda-calculus.
One of the Cambridge-school textbooks must cover relationships between
lambda-calculus reduction, pure lisp, Landin-style SECDM machines, and
formal UTMs. (Where "Cambridge school" Is, I suppose, the
Church/Turing/Strachey/ Scott tradition, something like that.
I confess I cant recall a good citation at the moment.)
Be good, lest someone whack you upside th head with tutorials
on reduction order.
>At least Chaitin is aware of the distinction between a formal language
>(maths) and a practical, natural, one (computing).
Computing may be practical, but its not a *natural* language. (Ask
any linguist.) John, are you serious? Or deliberately trying one on?
>An abstract object does not, in my view, actually exist. So I have the
>problem of where to locate abstractions. I locate them in semantic
>systems (natural languages). That's all. Nothing too abtruse about it.
>We approximate abstract properties in our natural languages, and one of
>those abstractions is formal language. Including the abstract LISP :-)
Er... I can only surmise this is one of the days of the forty-odd
years, when you deliberately choose to make no sense at all?
Robert Parson
No, Tim is right here. We are not talking about random walks here,
we are talking about deterministic Hamiltonian dynamical systems
(classical or quantum mechanical.) Poincare's recurrence theorem
(another consequence of Liouville's theorem)
states that given enough time, a bounded Hamiltonian system will
approach arbitrarily near to its initial state. Pictorially, you
start out with a neat, compact blob in phase space. This evolves
into a horrendously complicated structure with whorls and tendrils
all over the place. But if you wait 'long enough' it eventually condenses
to a blob again. The kicker, of course, is that the time scales tend
to be fantastically long for all but the simplest systems - not just
lifetime-of-the-universe, more like -exponentials of the lifetime of
the universe sorts of scales. And roughly speaking, the system is not
likely to approach near to its initial state
until it's had a chance to visit almost all of the rest of phase space.
(In fact, there's a theorem somewhere that expresses that sentiment
in precise language. ) Even for a system with only three degrees of
freedom, these time scales go through the ceiling.
Caveat: I have no idea what happens to these theorems when general
relativity has to be taken into account, although I suppose they
must break down in the case of an open universe, at least.
There's an interesting bit of history here. When Boltzmann first
attempted to prove from kinetic theory that the entropy of an isolated
gas must increase monotonically with time (the "Boltzmann H Theorem"),
he was confronted with two celebrated objections. Boltzmann's friend
Loschmidt argued that the monotonic increase of entropy contradicted the
reversibility of the microscopic equations of motion, while Zermelo argued
that it contradicted Poincare's recurrence theorem. In the process of
answering these two objections, Boltzmann was led to reinterpret
the "H theorem" as a statistical law rather than a mechanical one.
------
Robert
"In Phase Space, no one can hear you scream."
Robert Parson
Jonathan Stone <jona...@DSG.Stanford.EDU> wrote:
>In article <3C33DB94...@rcn.com>, Wade Hines <wade....@rcn.com> wrote:
>>Things can dissolve to some debatable metaphysics. The uncertainty
>>principle does not require non determinancy so much as speak to
>>ignorance.
The evolution of the _wave function_ is strictly deterministic, and
reversible (modulo a complex conjugate) .
Given Psi(t), I can calculate Psi(t'), for t' > t or t' < t.
However, it is not possible to determine Psi(t) by means of
measurements on a single system. A large number of measurements on
identically prepared systems is necessary. The time evolution of
the observable properties of single systems is not deterministic.
No one can predict when the Uranium nucleus will emit an alpha
particle, although we know very well how many alpha particles
will be emitted in unit time, on the average.
It's easy to get caught in linguistic tangles here. The Schrodinger
equation is a deterministic equation for nondeterministic phenomena.
>Well, uncertainty plus alleged demonstrations that hidden-variable
>theories are untenable, i guess. I never followed von neumann's proof.
Just as well since it's wrong. :-) That's the point of Bell's first
paper. (The 2nd paper, which fortuitously appeared first, is the one
that demonstrates that a hidden variable theory must be _nonlocal_.)
------
Robert
Richard Harter
Stone) wrote:
[snip much]
You inability to recognize smilies in the text that you read is duly
noted.
Richard Harter, c...@tiac.net,
http://www.tiac.net/users/cri, http://www.varinoma.com
Love, no matter how pure, is the most selfish of gifts.
For that reason it is the one gift that must be given.
Jonathan Stone
Richard Harter <c...@tiac.net> wrote:
>On 7 Jan 2002 17:22:59 -0500, jona...@DSG.Stanford.EDU (Jonathan
>Stone) wrote:
>
>[snip much]
>
>You inability to recognize smilies in the text that you read is duly
>noted.
I asked John a completely serious question about his use of the phrase
"natural language". John's substantive answer is that he seriously
claims information resides in natural languages, and that
(smileys aside) he does say LISP *is* a natural language.
I sincerely can't make any sense of John's distinction between LISP,
pure lisp, and `mathematical' constructs like lambda-calculus. Can you?
Jonathan Stone
Robert Parson <rpa...@spot.colorado.edu> wrote:
>In article <a10nb3$4bi$1...@Pescadero.DSG.Stanford.EDU>,
>Jonathan Stone <jona...@DSG.Stanford.EDU> wrote:
>>In article <3C33DB94...@rcn.com>, Wade Hines <wade....@rcn.com> wrote:
>>>Things can dissolve to some debatable metaphysics. The uncertainty
>>>principle does not require non determinancy so much as speak to
>>>ignorance.
>
> The evolution of the _wave function_ is strictly deterministic, and
> reversible (modulo a complex conjugate) .
> Given Psi(t), I can calculate Psi(t'), for t' > t or t' < t.
>
> However, it is not possible to determine Psi(t) by means of
> measurements on a single system. A large number of measurements on
> identically prepared systems is necessary. The time evolution of
> the observable properties of single systems is not deterministic.
Did that come before or after the interpretation of the amplitude of
wave-function as the square of the probability of finding an electron
at that point in any given single system?
>>Well, uncertainty plus alleged demonstrations that hidden-variable
>>theories are untenable, i guess. I never followed von neumann's proof.
>
> Just as well since it's wrong. :-)
Drat, and I was hoping to feed a straight-line about having
the wrong Aspect on the problem.
>That's the point of Bell's first
> paper. (The 2nd paper, which fortuitously appeared first, is the one
> that demonstrates that a hidden variable theory must be _nonlocal_.)
Oh? Thanks for the history lesson; I didn't recall that.
Wade Hines
Robert Parson wrote:
> Jonathan Stone <jona...@DSG.Stanford.EDU> wrote:
> >In article <3C33DB94...@rcn.com>, Wade Hines <wade....@rcn.com> wrote:
> >>Things can dissolve to some debatable metaphysics. The uncertainty
> >>principle does not require non determinancy so much as speak to
> >>ignorance.
Robert always makes me wonder if my hair just got parted differently
until I realize it's all too far over my head to matter.
> The evolution of the _wave function_ is strictly deterministic, and
> reversible (modulo a complex conjugate) .
> Given Psi(t), I can calculate Psi(t'), for t' > t or t' < t.
>
> However, it is not possible to determine Psi(t) by means of
> measurements on a single system. A large number of measurements on
> identically prepared systems is necessary. The time evolution of
> the observable properties of single systems is not deterministic.
> No one can predict when the Uranium nucleus will emit an alpha
> particle, although we know very well how many alpha particles
> will be emitted in unit time, on the average.
>
> It's easy to get caught in linguistic tangles here. The Schrodinger
> equation is a deterministic equation for nondeterministic phenomena.
My point, such as it was, is that the limits implied by the Schrodinger
equation may not be actual limits. You're free to argue hidden variables
with Jonathan if you think anything productive could come of that.
My further point was angling towards more practical notions of
irreversibility in macrostates despite reversibility of microstates.
The mixture of schlock philosophy and and profundity in that
nexus will only make me thirsty for more scotch.
Dunk
Parson) wrote:
>In article <a10nb3$4bi$1...@Pescadero.DSG.Stanford.EDU>,
>Jonathan Stone <jona...@DSG.Stanford.EDU> wrote:
>>In article <3C33DB94...@rcn.com>, Wade Hines <wade....@rcn.com> wrote:
>>>Things can dissolve to some debatable metaphysics. The uncertainty
>>>principle does not require non determinancy so much as speak to
>>>ignorance.
>
> The evolution of the _wave function_ is strictly deterministic, and
> reversible (modulo a complex conjugate) .
> Given Psi(t), I can calculate Psi(t'), for t' > t or t' < t.
>
> However, it is not possible to determine Psi(t) by means of
> measurements on a single system. A large number of measurements on
> identically prepared systems is necessary. The time evolution of
> the observable properties of single systems is not deterministic.
> No one can predict when the Uranium nucleus will emit an alpha
> particle, although we know very well how many alpha particles
> will be emitted in unit time, on the average.
>
> It's easy to get caught in linguistic tangles here. The Schrodinger
> equation is a deterministic equation for nondeterministic phenomena.
Agreed.. The wave function is a function. Plug in x, and the
function 'determines' an answer f(x).
It's deterministic; it determines probabilities. God's little joke.
Dunk
Tim Tyler
: Just as well since it's wrong. :-) That's the point of Bell's first
: paper. (The 2nd paper, which fortuitously appeared first, is the one
: that demonstrates that a hidden variable theory must be _nonlocal_.)
Strictly speaking it didn't demonstrate that...
I'll quote from the MWI FAQ:
``Q12 Is many-worlds a local theory? [...]
the many-worlds metatheory is a local theory [...]''
``Q13 Is many-worlds a deterministic theory?
Yes, many-worlds is a deterministic theory [...]''
``Many-worlds is local and deterministic.
``So where did Bell and Eberhard go wrong? They thought that all theories
that reproduced the standard predictions must be non-local. It has been
pointed out by both Albert and Cramer (who both support
different interpretations of QM) that Bell and Eberhard had implicity
assumed that every possible measurement - even if not performed - would
have yielded a single definite result. This assumption is called
contra-factual definiteness or CFD. What Bell and Eberhard really
proved was that every quantum theory must either violate locality or
CFD. Many-worlds with its multiplicity of results in different worlds
violates CFD, of course, and thus can be local.''
- http://www.hedweb.com/everett/everett.htm
Acording to "Q1 Who believes in many-worlds?" 58% of "leading cosmologists
and other quantum field theorists" thought that MWI was true, in a survey.
The only "No, I don't accept MWI" voter who left their name was Penrose.