🚀 go-pugleaf

In article <1992Jan14.212038.28608@beta.lanl.gov> jlg@cochiti.lanl.gov (Jim Giles)
writes:
>In article <1992Jan14.185453.20398@sunova.ssc.gov>, brian@helium.ssc.gov
(Brian Scipioni) writes:
>
>|> Statistical Mechanics does mention the concept of information. Take the H-theorem;
>|> H = S + I (the sum of entropy and macroscopic information of a closed system
>|> remains constant). DNA contains information - we even know the code. This means
>|> it represents a lower entropy state. [...]

Brian:
Would you be so kind as to expound on what this means?  What are your references?
Perhaps it is the original theorem?  How do you differentiate between S and I???
Your statements are extremely vague!  :-)

>You are confusing thermodynamic entropy with a similar concept (but 
>not realted to thermodynamics) in information theory.  Information
>is _NOT_ covered by the second law of thermodynamics.  There is 
>nothing in thermodynamics which prohibits "information" from arising 
>spontaneously.

Jim:
Oh cabbage dash!  :-)  If the states one is referring to (in the probabilities
of H or S) are the labeled as 'symbols', then there is a very close connection,
as I discuss in JTB 148:83 and 148:125.  Actually, it was von Neumann who
pointed this out first.

>Note also that for some reason, the definition of entropy that
>information theory usually uses has the opposite sign that the 
>definition in statistical mechanics does.

That's because you have made a mistake.  Information should always be taken for
the difference between two states.  In particular, it is the negative decrease
in uncertainty.  Get rid of this stupid "negentropy" concept!

Please join us over on bionet.info-theory for further discussions.

  Tom Schneider
  National Cancer Institute
  Laboratory of Mathematical Biology
  Frederick, Maryland  21702-1201
  toms@ncifcrf.gov