Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!sunybcs!dmark
From: dmark@cs.Buffalo.EDU (David Mark)
Newsgroups: comp.cog-eng
Subject: Re: Cross-linguistic issues in the design of Icons
Keywords: interfaces, icons, cross-linguistic issues, Spanish
Message-ID: <9472@cs.Buffalo.EDU>
Date: 22 Aug 89 02:37:25 GMT
References: <9268@cs.Buffalo.EDU> <1985@softway.oz> <6531@stiatl.UUCP>
Reply-To: dmark@sunybcs.UUCP (David Mark)
Organization: SUNY/Buffalo Geography
Lines: 43

In article <6531@stiatl.UUCP> tom@stiatl.UUCP (Tom Wiencko) writes:
>
>I contend that from a "theory of language" point of view that the concept
>of a "universal icon" cannot exist.  Here's why.
>
>An icon is nothing more or less than a symbol for a concept.  As we are well
>aware from the study of language (verbal and otherwise) a symbol can only 
>represent the concept which the user applies to it.  If two people who do not
>agree on the referant for a particular symbol attempt to communicate using
>that symbol, communication becomes impossible.
>
>Even symbols representing supposedly universal concepts (as can be found in
>physics and mathematics) are still symbols, and their use pre-supposes that
>both the sender and receiver agree on the referant (remember, the symbol is
>not the referant).

  My impression is that the theory of language as arbitrary symbol manipulation
  is passe.  I certainly susbscribe to the Rosch-Lakoff-Johnson model of
  categories, congition, and language.  The idea is that our perception and
  cognition is highly influenced by a fairly small set of cognitive image-
  schemata that provide 'top-down' structure to sesory inputs.  As I said in an
  earlier posting, my interpretation of Lakoff & Johnson's 'experiential
  realism' is that meanings arise largely out of the ways our bodies and
  senses interact with the real world.  If this is so, then we could expect
  at least a few concepts, such as the container schema (inside/outside) and
  the up-is-more schema, to be universal or nearly so.  Now to design icons
  that represent and/or invoke/trigger the same image schemata as some target
  concept in our image-schemata for the real world-- there's a challenge!!

>
>Without first finding a language independent "map" with which to map a 
>a set of universal symbols to their appropriate referants, I do not believe
>that a "universal symbol" can exist.
>
>Comments?
>
>Tom

  The cognitive image-schemata discussed by Johnson and Lakoff seem like
  possible candidates for this 'map'.

David Mark
dmark@cs.buffalo.edu