<< home  < Articles

No, but we know an awful lot and there's nothing in all our knowledge that suggests such a thing is possible. Remember, this violation of the Second Law only took place with a small number of particles (about 100) over short periods of time (a tenth of a second) under very contrived circumstances.

Any uncontrolled system the size of the Universe with billions of years running time will not show large-scale violations of the Second Law unless everything we know about thermodynamics is wrong. And while that's
possible, it's not very likely.

On the 27/7/2002, Chris Lawson wrote:

PhysicsWeb has a fascinating story on an experimental violation of the Second Law of Thermodynamics.

http://physicsweb.org/article/news/6/7/11

The research comes from Our Very Own Denis Evans of ANU (no, not Anu, or Ainu or whatever :-)

Those of us who are aware that the Second Law was a statistical law rather than a fundamental feature of reality won't be too surprised, especially given the very limited domain of the experiment, but it's still a great story and could have strange applications if nanotech ever gets off the ground.
 

At 17:36 26/07/02 +1000, Ray wrote:
>Chris, a big 'maybe' okay?
>
>Thinking of pre-solar super novae and subsequent coalescence as our present
>longer-lived, smaller massed Sun.... the whole 'condensation' and
>'expansion' basis to cosmological formation... and adding black holes and
>their kind of singularity... if the whole universe in space-time was a
>continuous cycle of destruction and construction, then the circle becomes
>self sustaining.

Of course this relies on our current understanding of the Universe, which may be wrong, but unless there are specific objections to our current understanding, then any further speculation is just fantasising. I've got nothing against that, btw. Some of my favourite Sf stories are pure speculation. But let's not fool ourselves that we're supported by science,
or even the reasonable hope of future science, when we delve into these areas.

Chris Lawson

Peter Macinnis added

I keep telling people about "What's new" -- here is the latest on several
things:

 
 "What's New" <whatsnew@aps.org>
Subject: What's New for Jul 26, 2002

WHAT'S NEW   Robert L. Park   Friday, 26 Jul 02   Washington, DC

1. BUBBLE FUSION: THE BUBBLE SEEMS TO HAVE COLLAPSED.  In March, against
the advice of physicists, Science published an article by Taleyarkan et al.
claiming to get fusion out of sonoluminescence (WN 1 Mar 02).  Two
experienced nuclear physicists, D. Shapira and M.J. Saltmarsh, using better
neutron detection in the same apparatus, said there was no evidence of
fusion.  Science refused to hold up publication of the Taleyarkan paper
until the Shapira and Saltmarsh findings could accompany it, or even add a
note warning that there were contrary results.  Not to worry!  The Shapira
and Saltmarsh paper is about to come out in Physical Review Letters, and is
expected to directly refute the Taleyarkan et al. paper.  And in this
week's Nature, a letter by Didenko and Suslick seemed to rule out bubble
fusion entirely.  Reactions of gases trapped inside bubbles soak up so much
energy that bubble temperatures could never get close to the threshold for
fusion.

2. ENTROPY: THE SECOND LAW OF THERMODYNAMICS STILL HOLDS.  Claims that the
Second Law of Thermodynamics has been violated are often found in fringe
journals.  This one is in Physical Review Letters
http://link.aps.org/abstract/PRL/v89/e050601.  The title: "Experimental
Demonstration of Violations of the Second Law of Thermodynamics in Small
Systems and Short Timescales," says it all.  The authors discovered that
when statistical laws are applied to systems that aren't statistically
significant, they don't work.  I experienced this myself.  As a boy I once
batted 1,000 for an entire day.  Ted Williams batted only 406, and for that
he has to hang upside down in liquid nitrogen until science figures out how
to revive him.  The statistics aren't promising.

3. MONEY: PHYSICS GETS A BREAK ON CAPITAL HILL.  After a dismal decade,
math and physical sciences got better news this week from Senate
appropriators, who increased NSF's MPS account by almost 15 percent.  DOE
also got some relief, as the House began work on Rep. Judith Biggert's
(R-IL) science authorization bill.  The goal of doubling the Office of
Science budget was supported by Nobelists Jerome Friedman and Richard
Smalley, who testified before the Energy Subcommittee.  Terrorism, the
possibility of war with Iraq, and a tanking stock market seem to have
persuaded Congress that it's time to support the physical sciences.

4. SECRECY: SELF-CENSORSHIP REPLACES GOVERNMENT INTIMIDATION.  In December,
WN heard that the White House was pushing the American Society for
Microbiology to develop guidelines for withholding information that could
help terrorists.  Today's NY Times says Ron Atlas, ASM President, is now
concerned that scientists may want to withhold information to keep others
from reproducing their results.  Atlas reportedly favors a full-disclosure
rule for all ASM journals.  We should all have such a rule.

THE UNIVERSITY OF MARYLAND and THE AMERICAN PHYSICAL SOCIETY Opinions are
the authors and are not necessarily shared by the University or the
American Physical Society, but they should be.
 

>Why is this needed for the hypothesis to be internally consistent?
>For internal consistency only the possibility need exist, not the
>actuality.  Completeness is not needed to validate the hypothesis.
>
>The newspaper is unlikely to fly up and hit the ceiling within the
>lifetime of the newspaper, the ceiling or the universe, but the
>possibility does exist.
 

But I don't understand the statement "completeness is not needed to
validate the hypothesis'. Doesn't a hypothesis without validation remains a
hypothesis?

Sue

Tony Morton replied on  30/7/2002

 > Of course, to be internally consistent, this hypothesis would imply the
 > repetition of everything that had ever happened previously—including
 > for  example, that someone named Chris Forbes-Ewan had typed this message  (and
 > had done so an infinite number of times) to an email discussion group  named
 > 'Science Matters'. And that someone with your name, and looking exactly  like
 > you, had read the message, and had done so an infinite number of times.

What you're talking about here, and what Zero called 'completeness' in his reply, is what statistical physicists call 'ergodicity'.  Basically, an ergodic system is one in which every possible state is visited and revisited infinitely often given an infinite amount of time.

Ergodicity for real physical systems is actually a big assumption, and one that nobody to my knowledge has so far proved.  Fortunately, although it used to be thought that ergodicity was required to justify the Second Law, this is no longer needed.

Using an information-theoretic notion of entropy, the thermodynamic states with the highest entropy are simply those that can be realised in the greatest number of ways.  So even if there were some particular state trajectory that turned out to be avoided for all time, or (like me typing this message) visited once and then avoided for all time, this would not invalidate the Second Law, being in a sense 'statistically insignificant'.

Cheers, Tony M.