RE: [humanmarkup-comment] Re: Economics Conference Example

[Rex Brooks <rexb@starbourne.com>]

Title: RE: [humanmarkup-comment] Re: Economics Conference Exa

Hi Folks,

Having chewed through Paul's posts and Len's replies and the url mentioned below, which is part of of original research material, albeit within the overall topic area of semiotics, I came to the conclusion that, while it is needful for us to maintain an awareness that this is a possible application area, our charter is to make such applications as Paul mentions doable with HumanMarkup insofar as we are able. The Human Markup Language itself cannot make the kind of distinctions between languages/cultures as are brought up in this discussion, much less attempt the finer grained distinctions that theoreticians might like to draw, if indeed, such distinctions can reliably be determined.

However, the kind of broad, immediate feedback techniques Susan's discussion suggests, ought to be very doable with, perhaps, even the basic schema we will develop, and certainly to a much more sophisticated degree with an extended schema, or a combination of such extended schemata, provided we can elicit the participation of experts in these fields to build applications with HumanML specifications.

Our task at this time is to delve into such potential target application areas and draw out the requirements we need to meet to make this possible, so I would hope that we can get much such discussions going, as well as probing deeper into this one and pulling out requirements from what we have been introduced here.

However, one particular element of this discussion that interests me and I think extends beyond the scope of this scenario to include AI in particular is the notion of tokenizing information sets in such a way that useable data can be extracted. A common set of tokens for various kinds of languages might be a key to developing applications for translations, as well as applications for learning by verification of hypotheses, distilling meaning or pattern from noise, etc. In this particular case, I am thinking that some measure of chronemic variation in response time would be useful for, as an example, participants in a conference with electronic response pickups or keyboard/mouse responses to such real time translation as simultaneous subtitling for real time audience feedback analysis.

Comments?

Ciao,

Rex

At 1:22 PM -0600 1/24/02, Bullard, Claude L (Len) wrote:

> psp writes:  "(the Whorf Hypothesis) should
> remind us that knowledge mapping in communities of practice is NOT the
> same as machine knowing of data invariance and patterns expressed via
> computer programs."
>  
> Sapir-Whorf?  http://www.aber.ac.uk/media/Documents/short/whorf.html
>  
> "Whorf distanced himself from the behaviourist
> stance that thinking is entirely linguistic (Whorf 1956, p. 66). In
> its most extreme version 'the Sapir-Whorf hypothesis' can be described
> as consisting of two associated principles. According to the first,
> linguistic determinism, our thinking is determined by language.
> According to the second, linguistic relativity, people who
> speak different languages perceive and think about the world quite
> differently...Moderate Whorfianism differs from extreme
> Whorfianism in these ways:

• the emphasis is on the potential for thinking to be 'influenced' rather than unavoidably 'determined' by language;
• it is a two-way process, so that 'the kind of language we use' is also influenced by 'the way we see the world';
• any influence is ascribed not to 'Language' as such or to one language compared with another, but to the use within a language of one variety rather than another (typically a sociolect - the language used primarily by members of a particular social group);
• emphasis is given to the social context of language use rather than to purely linguistic considerations, such as the social pressure in particular contexts to use language in one way rather than another. "

> Byte analysis is not the same as analysis of
> observable gestures, true.  Trying to solve that purely
> mechanically, and probably even with AI, is
> likely to be very difficult.   However,
> human translators do a reasonable job and a reasonable job may be all
> that is required.  For Susan's
> problem, real-time analysis is not required.  A
> human observer system that enables rapid acquisition and
> classification to provide near
> real-time response formulation is probably adequate. 
> Micro-analysis is best done after game time when one has the luxury to
> sit back with
> a beer and the remote to shuttle back and forth over
> the tape.  What the WTO participants need is a means to sort
> faster and formulate
> a response by the next day.   Elimination of
> miscommunication is not possible and probably not desirable given that
> it is a tactic to
> achieve a strategic objective.  Much like radar
> guns are met with radar detectors, the simply humans work harder to
> outwit the system.  The
> best approach is a clear long term policy for goals
> that remain consistent with the tactics used to achieve it insofar as
> these goals are achievable.
> For this, a means to gather macro level and events and
> get an analyis of how these amplify or interfere with the current
> tactics and overall goals
> is likely to produce satisfactory results with respect
> to the latency between the call for service and the dispatch of the
> right assets.
>  
> len
>

-- 

Rex Brooks
GeoAddress: 1361-A Addison, Berkeley, CA, 94702 USA, Earth
W3Address: http://www.starbourne.com
Email: rexb@starbourne.com

Tel: 510-849-2309

Fax: By Request