[Fis] [External Email] Re: defining information

Terrence W. DEACON deacon at berkeley.edu
Fri Mar 6 20:18:43 CET 2020


RE: DEFINING ‘INFORMATION’

I am constantly surprised how this question keeps resurfacing in the FIS
discussions.

Diverse unmarked divergent uses of the term 'information' continue to haunt
these discussions

and lead to pointless arguments about what is THE ONE CORRECT definition,
and who is right about how the term ‘information' SHOULD be used.

Both in FIS online discussion and in IS4SI meetings I have argued that the
term is used with three different but hierarchically dependent meanings
without careful marking which is implied. The often unmarked distinction
forms what Stan would describe as a subsumptive hierarchy of definitions
and uses. Let me explain what I mean:

These three distinctions can indirectly be traced to C S Peirce's
prescinded Categories as they were re-characterized (some would say
mischaracterized) by Charles Morris in linguistic terms as syntactic,
semantic, and pragmatic information.

This triadic usage can, for example, be found in numerous writings in the
philosophy of information.

I prefer  Structural (IStruc),  Referential (IRef), and  Normative (INorm)
information, however, in order to avoid the linguistic framing so that the
concept can be applied to nonlinguistic uses without confusion.*

Their relationship to one another is therefore:

IStruc = Shannon's and related uses (e.g. data, signal properties,
statistical “entropy” of a communication medium irrespective of any
referential or normative relations)

IRef = most semiotic uses where reference to an object is invoked
irrespective of its factual correspondence, as in Frege’s *Sinn* (sense,
intension), such that IRef assumes IStruc.

INorm = functional, teleological, correct/incorrect, true/false,
useful/unuseful, etc., property of IRef (which in turn assumes IStruc) and
can roughly be understood in terms of Frege’s concept of *Bedeutung*
(correspondence with a specific actual physical entity or state of affairs).

This could be diagrammed [ INorm [ IRef [ IStruc ]]] or INorm > IRef  > IStruc
or with simpler subscripts as In, Ir, and Is. Of course, there has been
considerable philosophical debate concerning both In and Ir. But this does
not undermine the need to avoid this equivocation of meanings, because this
asymmetric dependency leads to confusions when using the term 'information’
in a way that ignores one of its implicitly subsumed meanings.

This can be avoided by attending to this asymmetric dependency in a
particular usage. For example:

1. An encrypted message prior to being decrypted can be unequivocally
analyzed in terms of Is irrespective of the other levels of analysis. And
the character string “Vladimir Putin is the president of the United States”
can be analyzed irrespective of its referential interpretation.

2. In addition, the referential information (Ir) of the sentence “Vladimir
Putin is the president of the United States” is unambiguous as is its Is
irrespective of whether the sentence is deemed true, false, or merely being
used for ironic emphasis.

3. Whereas that sentence’s actual correspondence (literal truth) with the
current political situation is a normative (In) assessment that depends on
a prior analysis of its referential relation Ir.

So my plea to the contributors of FIS is that rather than assuming that
every reader is interpreting the term ‘information’ in the same way, or
assuming that the way one is currently using it is the ONLY correct usage,
perhaps we could be more specific; either by explicitly framing the context
of usage or by marking each usage in a way that is analogous to what I have
done here. I think that this could avoid considerable miscommunication and
needless disagreement. Of course, this preliminary categorization is
subject to further analysis, and I welcome discussion of these
relationships.


* This three-part distinction has also been explicitly presented in a
submitted paper and IS4SI presentation delivered at the June 2019 meeting
by Liqian Zhou—both of which reference my work—though there is a slight
difference in how we distinguish between In and Ir.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listas.unizar.es/pipermail/fis/attachments/20200306/228ff301/attachment.html>


More information about the Fis mailing list