List, Pedro, Bob:

A modestt proposal

Without a "gold standard" for the term "information" we keep going  
around in circles.

I would suggest that at least three distinct meanings of the term  
"information" are in current usage, probably more.

For convenience, let us take the Shannon view of information as it is  
exact in definition and application, it is purely an arbitrary term  
in this sense, as are all mathematical terms.  The symbols acquire  
meaning from the cultural context rather than emerging from  
biosemiotic origins.

Shannonesque information (if I can create such an adjective) is  
design to serve as the message bearer between two machines. But this  
is not an ordinary message bearer.  As a message bearer, Shannonesque  
information can include a rudimentary form of redundancy, an elegant  
form of symbolic repetition that allows machines to correct errors  
that the message bearer may acquire on it's journey from machine A to  
machine B. This design feature is constructed into the artificial  
encoding that creates the message.  From the antecedent redundancy of  
the code, the message bearer acquires a property that goes beyond the  
semantics of mathematical symbols when viewed from the perspective of  
enabling communication between machine A and machine B.  This  
consequence is not a property of all message bearers.

In general, "raw" physical information lacks the Shannonesque message  
bearer quality because physical events are unique.

In general, chemical information lacks the Shannonesque message  
bearer quality because chemical structures are unique.

In general, biosemiotic information, composed from chemical  
information, has the message bearing quality of Shannonesque  

In general, cultural information, composed from biosemiotic  
information, has the message bearing quality of Shannonesque  

At the most primitive levels of description, both biosemiotic and  
cultural information have a modicum of error correction capacity,  
metaphorical comparable to machine to machine communication.

So, where does this Peircian categorification of the kinds or sorts  
of information lead?

I suggest that Stan's usage of the term "valency" of information may  
be a useful name for the values of information in the respective  
systems that is being referred to. The polyvalency of biosemiotic and  
cultural information is already well-established in practice.

The concept of polyvalent information provides a reasonable term to  
describe the exactness of the reproduction of biological structures,  
of genetic inheritance.

The conundrums over the questions of symmetry and asymmetry remain  
open for description in logical terms of the valency of the  
symbolization of information used for communication.  For example,  
can practical communication be achieved with an infinitely  
polyvalent  "chunk" of information? Or, is this merely a useful  
metaphor? How does the Barwise metaphor of categorical information  
fit into the concept of polyvalency? Can one actually encode  
information into infinite groups or is this merely a mathematical  

Numerous other questions can be raised from the logical proposition  
that communication implicitly connects via valencies.

Happy Thanksgiving to one and all.



Research Professor
Krasnow Institute for Advanced Study

fis mailing list

Reply via email to