On Fri, 27 May 2011, Bryan Jurish wrote:
As far as language (or symbols) are concerned: yes, of course. But if I'm reading it right, there's nothing which says that the "features" (which may or may not "count" as information content) rely for their ontological status only on their use (or non-use) in a Shannon-esque "message"
Shannon's information theory assumes that there is a single stream of information that is all of à priori equal value except that unlikely outcomes are considered more remarkable. However, to explain human perception, you have to take into account pre-established notions of importance of information, and what's the flow of attention-span, among a lot more things.
(I wrote this message 6 months ago and forgot to click Send. Still trying to exorcise myself out of this habit. In the meantime, it makes filler for a slow week on pd-list. ;-)
| Mathieu Bouchard ---- tél: +1.514.383.3801 ---- Villeray, Montréal, QC