RE: Re [Fis] Sustainable use of resources

From: Loet Leydesdorff <[email protected]>
Date: Thu 11 Dec 2003 - 22:05:07 CET

Dear Pedro and FIS colleagues,

In my opinion, the confusion is created because one does not carefully
distinguish between the mathematical definition of information and the
meaning of the information with reference to a system under study. The
mathematical definition of information (Shannon, 1948) in terms of bits
of information is yet content-free (Theil, 1972). The formalism (H = -
Sigma p log(p)) was chosen by Shannon so that it corresponds with the
formalism for thermodynamic entropy, but this correspondence is only
formal. (Thermodynamic entropy is not expressed in terms of bits, but in
terms of Watts/Kelvin.)

By specifying the system of reference, the information is provided with
meaning by the researcher. The information can, for example, be
contained in the distribution of molecules or, in a completely different
system, by the distributions of signals on a telephone line. The system
of reference determines also what can be considered as meaningful
information and what as noise because the system contains the selection
mechanism. For example, it determines also the maximum information
content (and therefore the redundancy). The information content,
however, remains a property of the distribution, i.e., a mathematical
measure.

For example, the distribution 1/2 + 1/2 contains one bit of information,
independent of whether it is about coin flipping or anything else.
However, this tells us nothing but a mathematical formalism. Once we are
talking about heads and tails we have specified a system of reference.
Similarly, we may wish to specify a molecular system as a system of
reference and express the changes in the distributions in terms of the
probabilistic entropy generated. We can provide the result of the
calculation then with an interpretation because we have specified the
system of reference.

I hope that this is helpful.

With kind regards,

Loet

Reference:
Henri Theil, Statistical Decomposition Analysis, Amsterdam:
North-Holland, 1972.

> -----Original Message-----
> From: fis-bounces@listas.unizar.es
> [mailto:fis-bounces@listas.unizar.es] On Behalf Of Pedro C. Mariju�n
> Sent: Thursday, December 11, 2003 2:52 PM
> To: fis-listas.unizar.es
> Subject: Re: Re [Fis] Sustainable use of resources
>
>
> Dear FIS colleagues,
>
> Most of recent messages transpires a strong parallel between energy &
> information. Biologically (ecosystems included) that parallel
> can be quite
> misleading, particularly if it implies --perhaps inadvertently-- the
> 'interconversion' between any forms of information, like
> happens among the
> forms of energy. In the functionality of the living there is
> no such thing
> as a global information content to interrelate with a
> corresponding energy
> global. Actually, along past FIS discussions we have
> developed a brute
> taxonomy of info categories at the molecular level, but the
> dynamics is not
> well solved yet.
>
> It is interesting that in the perspective of information science
> (tremendously incomplete at it is nowadays), themes that
> disciplinarily are
> worlds apart become quite close here. Somehow, information
> propitiates an
> original realignment of themes, for instance: meaning,
> knowledge, fitness,
> value,... Regarding ecological economics and sustainable
> development, the
> exploration of such alignment could provide (who knows) a genuine
> 'informational theory of value'. And it could help to point
> at the blind
> spots of the 'invisible hand' of market economies --beyond
> the guidance of
> the Second Law.
>
> There are many laws of nature that are also crucial for the
> organization of
> the living (and not only the Second Law): Pauli's exclusion
> (so we may
> have complex atoms), or De Broglie�s and Schrodinger�s wave-particle
> complementarity (so we may have bonds), Maupertuis &
> Hamilton's minimal
> action, Gibbs' free energy, etc. Fair interdisciplinary (or
> multidisciplinary) practices imply that one cannot 'reduce' all the
> complexity looking only under one disciplinary box and
> disregarding the
> rest. It is a pity that 'reductionism' methods have been discussed ad
> nauseam--but there is almost nothing about 'integrationism'.
>
> greetings to all
>
> Pedro
>
>
> _______________________________________________
> fis mailing list
> [email protected] http://webmail.unizar.es/mailman/listinfo/fis
>

_______________________________________________
fis mailing list
fis@listas.unizar.es
http://webmail.unizar.es/mailman/listinfo/fis
Received on Thu Dec 11 22:06:47 2003

This archive was generated by hypermail 2.1.8 : Mon 07 Mar 2005 - 10:24:46 CET