[Fis] Shannonian Mechanics?

Hector Zenil hzenilc at gmail.com
Wed Jun 29 15:37:12 CEST 2016


I think complaining about Shannon entropy as a measure of information is
completely justified because it is steam-engine physics unfortunately still
widely used despite its many flaws and limitations.

But to think that Shannon entropy is at the front-end in the mathematical
discussion of information is a mistake and this, and other groups, have
perpetually been entrapped in a 60s and 70s discussion on a fake ancient
theory of information that not even Shannon himself thought was worth to be
used for anything meaningful in information but for communication measuring
purposes only.

Indeed, Shannon entropy is nothing else but a counting function of
states/symbols, at best it is a measure of diversity, a bound on
information transfer. The technical and philosophical discussion here and
everywhere else should be (and has been among those at the scientific
front) focused on what has been done in the last 50 years to leave Shannon
entropy behind, but nobody here (and almost nowhere else) are people
discuss about algorithmic randomness, Levin's universal distribution,
measures of sophistication, etc. but prefer to be in a continuous state of
pre 60s Shannon entropy discussion.

Shannon entropy should not even be mentioned any longer in serious
discussions about information, we moved on a long time ago (unfortunately
not even many physicists have done)

Trying to be constructive. All best,

- Hector
http://www.hectorzenil.net/


On Wed, Jun 29, 2016 at 3:16 PM, joe.brenner <joe.brenner en bluewin.ch> wrote:

> Dear Loet,
> The way you have asked it, I think the answer to your question is known:
> both order and disorder are universals, linked  dialectically. Never one
> without the other, as for symmetry and asymmetry, except in trivially
> simple cases.
> Cheers,
> Joseph
>
>
> Sent from Samsung Mobile.
>
>
> -------- Original message --------
> From: Loet Leydesdorff
> Date:29/06/2016 14:40 (GMT+01:00)
> To: "'Pedro C. Marijuan'" , fis en listas.unizar.es
> Subject: Re: [Fis] Shannonian Mechanics?
>
> Dear Pedro and colleagues,
>
>
>
> The figure from Weaver in Loet's excellent posting leaves a few aspects
> outside. The why, the what, the how long, the with whom, and other aspects
> of the information phenomenon do not enter. By doing that we have
> streamlined the phenomenon... and have left it ready for applying a highly
> successful theory, in the technological and in many other realms
> (linguistics, artif. intelligence, neurodynamics, molec. networks, ecol.
> networks, applied soc. metrics, etc). Pretty big and impressive, but is it
> enough? Shouldn't we try to go beyond?
>
> In my opinion, “The why, the what, the how long, the with whom, and other
> aspects …” are subject to substantive theorizing. The type of answers will
> be very different when studying biological or other systems of reference.
> But then the information is provided with meaning by these theories and we
> discuss “meaningful information” as different from Shannon-type
> information. There will in this case a dimension to the information.
>
>
>
> For example, when particles collide, there is exchange of momenta and
> energy. The dissipation is then dimensioned as Joule/Kelvin (S = k H). In
> chemistry one assumes a mass balance and thus a redistribution of atoms
> over molecules, etc. The dimensionality of interhuman communication is
> hitherto not specified.
>
>
> I wonder whether a far wider "phenomenology of information" is needed
> (reminding what Maxine argued months ago about the whole contemplation of
> our own movement, or Plamen about the "war on cancer"?). If that inquiry is
> successful we could find for instance that:
>
> This is not successful. It does not lead to a research program, but to
> “philosophie spontanée des savant” (Althusser) as your comprehensive
> question for “The why, the what, the how long, the with whom, and other
> aspects” illustrates. The hidden program is biologistic:
>
>
> 2. Those UNIVERSALS are SPECIES' SPECIFIC.
>
>
>
> “ESSENTIAL CORES” are discipline specific!
>
>
> 3. Those UNIVERSALS would be organized, wrapped, around an ESSENTIAL CORE.
> It would consist in the tight ingraining of self-production and
> communication (almost inseparable, and both information based!). In the
> human special case, it is the whole advancement of our own lives what
> propels us to engage in endless communication --about the universals of our
> own species-- but with the terrific advantage of an open-ended
> communication system, language.
>
> 4. Those UNIVERSALS would have been streamlined in very different ways and
> taken as "principles" or starting points for a number of
> disciplines--remembering the discussion about the four Great Domains of
> Science. A renewed Information Science should nucleate one of those
> domains.
>
> “Should” is an expression of uneasiness? In my opinion, the assumption of
> an origin is problematic: order is not given (ex ante) and then branching,
> but emerging (ex post) from disorder (entropy). Is “disorder” perhaps a
> universal? In which specific system? (I would have a provisional answer/
> hypothesis; but this is my second penny for this week.)
>
>
>
> Best,
>
> Loet
>
>
> Best regards to all,
> (and particular greetings to the new parties joined for this discussion)
> --Pedro
>
>
> El 27/06/2016 a las 12:43, Marcus Abundis escribió:
>
>
>
> Dear Loet,
>
>
>
>     I hoped to reply to your posts sooner as of all the voices on FIS I
> often sense a general kinship with your views. But I also confess I have
> difficulty in precisely grasping your views – the reason for my delay.
>
>
>
> >[while Shannon’s] concept of information (uncertainty) <
>
> > is counter-intuitive. It enables us among other things <
>
> > to distinguish between "information" and "meaningful <
>
> > information". <
>
> • Easily agreed; *how* to distinguish a presumed meaning (or
> meaningless-ness) then becomes the remaining issue.
>
>
>
> > Providing . . . meaning presumes the specification <
>
> > of a system of reference; for example, an observer.<
>
> • It is telling for me (in viewing our differences and likenesses) that
> you suggest an observer. My “system of relating“ accommodates but does not
> require an observer (okay – observer, defined how?), as shown immediately
> below.
>
>
>
> >Different[ly] . . . expected information is dimensionless<
>
> > ("a priori"). <
>
> • I suggest the act of “expectation“ already infers minimal dimensions –
> for example, who/what/how is doing the expecting? Thus, in my view, this is
> not truly a priori. A “readiness“ or a compelling functional need innate to
> any “system of relating“ has bearing. For example, a single Oxygen atom has
> a compelling/innate need to react with other elements, just as any agent is
> compelled to react to “nutrients.“ Both imply dimensional expectations, no?
> (obviously – of different orders/types).
>
>
>
> > In my opinion, a "real theory of meaning" should enable <
>
> > us to specify/measure meaning as redundancy / reduction <
>
> > of uncertainty given . . . I took this further in . . . <
>
> > The Self-Organization of Meaning and the Reflexive . . .<
>
> • My weak grasp of the concepts in this paper leads me to think you are
> actually modeling the “processing of meaning,“ related-to-but-distinct-from
> “generating meaning“ that I target. I also vaguely recall(?) in an offline
> exchange I asked you if you saw this paper as presenting a “theory of
> meaning“ and you answered “No.“
>
>
>
> • In your later response to Pedro, I found your citation matrix a
> interesting example of your thinking, but still too “high-order“ for my
> reductive-but-meaningful aim. Your matrix (for me) presents an essential
> complexity of high-order views, but in itself it is too simple to detail
> *how* a citation is *meaningfully used.* Still, an intriguing concept that
> might be meaningfully expanded? Perhaps there are more useful details in
> the additional papers you list, which I have not had a chance to explore.
>
>
>
> • Your last post then reinforces my sense you are actually exploring the
> processing of meaning, rather than the generation of meaning. Diverse
> “systems of relating“ you name seem to be “on point“ and
>
> > can be considered as a semantic domain (Maturana,1978)<
>
> But I find this unsatisfying as exactly *what(s)* is being related, and
> exactly *how* it is being related, does not seem to be covered. It is in
> precisely naming those “whats“ and “hows“ that true a priori models become
> possible. For example, a *system of relating* between “a hominid and a
> rock“ affords certain types of meaning, equally a *system of relating*
> between “the same rock and an ant“ affords wholly different types of
> meaning – all in regards to an identical (autonomous) rock.
>
> > the same information is delineated differently and <
>
> > considered from a different perspective <
>
> arguing for essential subjectivity? This seems to point to my use of delta
> O and delta S in the video.
>
>
>
> • I am unsure if we are in: radical agreement, radical disagreement, or if
> we just name subtle differences. . . but I thought I should at least
> attempt a reply to your posts and see what ensues.
>
> > In my opinion, the task is to specify mechanisms which <
>
> > generate redundancy <
>
> This leads me to believe we essentially agree but focus on different
> levels of operation and complexity. Any thoughts you have to share are
> appreciated.
>
>
>
> Sincerely,
>
>
>
> Marcus
>
>
>
>
>
> --
>
> -------------------------------------------------
>
> Pedro C. Marijuán
>
> Grupo de Bioinformación / Bioinformation Group
>
> Instituto Aragonés de Ciencias de la Salud
>
> Centro de Investigación Biomédica de Aragón (CIBA)
>
> Avda. San Juan Bosco, 13, planta X
>
> 50009 Zaragoza, Spain
>
> Tfno. +34 976 71 3526 (& 6818)
>
> pcmarijuan.iacs en aragon.es
>
> http://sites.google.com/site/pedrocmarijuan/
>
> -------------------------------------------------
>
>
> _______________________________________________
> Fis mailing list
> Fis en listas.unizar.es
> http://listas.unizar.es/cgi-bin/mailman/listinfo/fis
>
>
------------ pr�xima parte ------------
Se ha borrado un adjunto en formato HTML...
URL: <http://listas.unizar.es/pipermail/fis/attachments/20160629/8bbe5b64/attachment-0001.html>


More information about the Fis mailing list