Re: Information = Comprehension x Extension -- Discussion

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

ICE.  Discussion Note 11

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

HT = Hugh Trenchard

Re: ICE Discussion 9.  http://suo.ieee.org/ontology/msg05286.html
In: ICE Discussion.    http://suo.ieee.org/ontology/thrd1.html#05283

HT: I am guessing Dewey was the originator of the Dewey Decimal System?

no, that was melvil dewey (1851-1931):
http://www.mtsu.edu/~vvesper/dewey.html

HT: What you say about information as a measure of uncertainty accords
    with what I am reading about Boltzmann's measure of information as
    the equivalent of the entropy of a system.  The greater the entropy,
    the greater our ignorance of the locations and motions of specific
    molecules, and hence the lower the information content of the system.

yes, that's the way that it was taught to me a millennium ago in
my chemsitry and physics courses.  properly speaking, entropy is
a measure on distributions that attains its maximum on a uniform
distribution.  to talk about the information impact of a message --
here i will shift to semiotic-speak and dub anything of that ilk,
from signals to wffs to full-fledged theories, as a "sign" -- we
have to consider prior and posterior distributions as reflecting
an agent's "state of information" (SOI) before and after getting
a sign.  the "average uncertainty reduction on receiving a sign"
(AURORAS) then serves to measure the "capacity" of the so-called
"channel" that's defined by the statistical ensemble in question.

HT: As I understand it, Boltzmann was the first to identify
    a subjective element to information content (in physics).
    Prior to Boltzmann entropy and its information content
    was an expression of the number of possible ways of
    arranging the components in a system (W).

this is the place, of course, where a lot of fights usually break out.
readers who benefit from a reading of peirce will tend view the whole
melee from a slightly different angle.  i for one will try to reserve
the word "subject" and all its derivatives to its logical uses, as in
"subject and predicate", and i will usually describe this "relativity
of observation to the observer thereof" in terms that involve an idea
of these relationships among object, sign, and "interpretant" domains.

in that so-called "semiotic" or "sign-theoretic" framework,
it's a bit difficult to recall what all the fuss was about.

HT: von Baeyer, in his book 'Information' says:

    | ... W ... has a strange subjective quality ...
    | It is not just the number of ways a system can
    | be rearranged, but more specifically the number
    | of rearrangements consistent with with the known
    | properties of the system.  Known by whom?  Measured
    | by what observer?

yes, in formal terms one may associate this with the
so-called "reference class" of distributions that an
observer or theorist chooses to describe a domain of
phenomena.  there's nothing a priori that makes this
choice for you, it's a risky bit of abductive reason.

HT: von Baeyer gives the example of four clear flasks side by side,
    touching.  At the point of connection holes in the flasks allow gas
    input intitially into one container to escape into neighbouring container.
    As the gas escapes into the neighbouring flasks, we lose information about
    the system, because it is now more greatly distributed than it was before.
    But in this case the information was not just the objective number of ways
    of arranging the molecules of gas in the flask, but the subjective knowledge
    of the number and volumee of flasks through which the gas was flowing.

HT: von Baeyer also talks about the subjective element
    in probabilities, another measure of information,
    among other things.

HT: So, although, I am guessing that Dewey's discussion
    of "mental processes removing" uncertainty was made
    in a  different context than the entropy of gasses,
    it seems to me it is consistent with what Boltzmann
    said about the inclusion of a subjective element
    in information measurement;  consistent too with
    probability/information.

i think that he was trying to call our attention to both sides
of the coin, and to emphasize that the parameters of interest,
uncertainty and information, are grounded in the interactive
relationship of the agent and the environment, and are thus
not reducible to "pure objectivity" or "pure subjectivity".

just my current guess,

jon awbrey

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o
http://www.cs.bsu.edu/homepages/mighty/history.html
o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o