[humanmarkup-comment] Re: regarding stratified complexity systems andsemiotics

[Rex Brooks <rexb@starbourne.com>]

To the best of my knowledge, HumanMarkup has mostly been envisioned 
as a vocabulary that will enable many kinds of applications. We 
narrowed the mission down in scope to communication, and specifically 
for improving the accuracy or fidelity of human communication in 
digital information systems.

However, while that is the stated mission, any attempt to confine the 
uses to which HumanMarkup can be put would be an exercise 
metaphorically analagous to putting the proverbial genie back into 
the bottle.

For my own part, my interest is in providing for the representation 
of Human and Human-like behavior, which will almost always be 
intertwined with communications purposes within 3D environments. That 
includes facial expressions for which I usually term facial gestures 
to be consistent with bodily gestures within the vocabulary. One of 
the things I am working on right now is a kind of facial skeleton 
that will work with the H-Anim specification that is adjunct to 
VRML97, the current ISO standard and/or X3D which will be the next 
version of the ISO standard. By building in the possibility for 
numerical values for percentages of combinable emotional facial and 
bodily gestures which can be based on HumanMarkup, and using those 
values to drive H-Anim avatars be they representations of real human 
beings or software agents, a middleware can be built to provide 
automatic and semi-automatic animated representations of Human and 
Human-like behaviors.  But that is just my particular use.

The use I suggested earlier, that of forensic anthropolgy, is another 
kind of use for HumanMarkup.

These are by no means the most likely or widespread of the uses we 
have envisioned for HumanMarkup. Bear in mind that these uses are 
only the ones we have envisioned. Once HumanMarkup gains a stronger 
foothold in the computing environment, many more will surface.

I suspect that providing for in-depth personalization information 
will be the most widespread use of HumanMarkup. Making it possible 
for software to learn a user's preferences over time will become, I 
believe, almost a default application of the HumanMarkup. I also 
believe that HumanMarkup will be a de facto example of stratified 
complexity without putting that forward as a requirement, much as the 
process we are currently conducting is an example of descriptive 
enumeration, where the enumerations from various application areas 
are driving the development of the language/vocabulary.... We look to 
applications areas to inform us about what they need, and for 
vocabularies with which we must be consonant.

My response to Len's comment stems from some of our earliest attempts 
to frame a description of just what we wanted HumanMarkup to be and 
to do. I, myself, made the initial error of attempting to model Human 
Behavior with it, and so discovered the hard way, so to speak, that 
what we really wanted to do, and ALL that we really COULD do, was to 
make it possible to model Human Behavior, if someone wanted to. Or, 
to be more accurate still, to make it AS possible as we can for 
someone to model Human Behavior. It was the distinction between 
vocabulary and application that I was pointing toward.

There is no one single application or purpose for HumanMarkup. 
Developing Use Cases might be a good idea, and we have done a bit of 
that, which is available in the materials we make available from what 
we call Phase 0 of our effort from the time before we requested 
permission to join OASIS and later joined, on the OASIS HumanMarkup 
TC website. However, since we have done so much work, I would rather 
proceed onward toward a Primary Base Schema. We are making the 
Primary Base Schema adaptable so that as we learn more, we can add to 
it. Indeed, one of the issues before us is whether or not we want to 
have a Secondary Base Schema or Schemata in addition to the 
application area-specific schemata, or modules, such as cultural 
modules, which we have already settled on as the appropriate method 
for advancing HumanMarkup further.

Taking some time to study further is fine, but it doesn't require 
stopping and changing direction midstream. One of the developments 
Len is taking into consideration is that of processing. I applaud 
that and his intention to investigate stratified complexity as well. 
To use a trope of which I am not fond because it has been employed as 
slogan by the GOP is nevertheless apropo: We are building a big tent 
here, and there, of necessity, MUST be room for all, especially those 
whom we must fear and defend against due to the ill will they direct 
against us. Our aim is to enable better communication which we think 
may make better understanding, in turn, possible.

Ciao,
Rex

At 7:53 PM -0400 6/7/02, paul wrote:
>One must always make an apology for science when the principles are beyond
>what is desired by those who live in a illusion.
>
>This is as it is.  The mark-up of human behavior is what the Oasis
>Human-Mark-up standard will be about, when it is offered as a first draft?
>Yes?  Perhaps we need use-cases that demonstrate examples where the standard
>would be used.
>
>So, the reason why human mark-up must be framed in a stratified point of
>view is that any set of behaviors (for example, human learning behavior) is
>a composition of reflexes.  (One can think of the scholarly discussion about
>reflex arcs as introduced by Sherrington and then as part of behavior
>science).  Knowing the set of reflexes as a level of organization (like
>atoms) and the behaviors as a different level of organization (like chemical
>compounds), leads to a Peircean type model (semiotics) for agile composition
>of atoms into compounds.  Several of my students are working on issues
>relating this notion of composition of behavioral atoms to the
>OntologyStream Knowledge Operating System - but I just do not see how to
>involve them, or myself in the Oasis process. (Still listening.)
>
>The notion of complexity has to do with underdetermined "behavior" where the
>exact determination of the "behavior" (which must occur at some point) is a
>selection between possible outcomes (the collapse of the wave.)
>
>Rex, you ask about whether Len is talking about the combination of
>applications where the Mark-up, related to the XML - or ontology - internal
>to the application, is in domains of archeological/anthropological knowledge
>or geological knowledge.
>
>There is an aspect of terminology comparison in this problem of annotating
>knowledge structure.  This aspect does involve scope and viewpoint and thus
>is both complex and (I think) requiring of a stratified approach.
>
>Clearly we need to have a means to markup how humans learn in various
>domains.  Do you see this issue as relevant to the mission of the Human
>Mark-up committee?
>
>I am really trying to get a boundary around why one wants to have a Human
>mark up language/(set of symbols).


--