[humanmarkup] Base Schema-emotion

[Rex Brooks <rexb@starbourne.com>]

Hi Everyone,

I'm moving on to another element not because I think we are done with 
culture, but because I don't have much else to say about it until I 
see what Sylvia and Len come up with, and because I don't expect to 
resolve it for the first draft until we revisit it in the run through 
of the first draft specification. I do, however, look forward with 
great anticipation to see what our resident semiotes think about it. 
Also it is a big topic and this one is not--at least not in itself. 
Thankfully.

emotion

This is a ComplexType with the attribute of abstract. It does not 
reference other elements. It belongs to the attribute group 
humldentifierAtts. It takes an attribute value of intensity.

It's description is: A basic set of primitive human emotions.

It is about as basic and atomistic an element as we have, and while 
we may have noodling to do with other elements, there is not much to 
say about this one. This does however beg the question of a Secondary 
Base Schema since a number of secondary schemata will need the 
emotion primitives, so I think we pretty much find ourselves 
requiring the Secondary Base Schema.

There is a point here which I have not brought up yet, but that I 
think we need to deal with now. Our Base Schema have not been defined 
as having attributes separate from the datatypes enumerated in the 
global attribute definitions and the reason I have not said boo about 
it is that I happen to agree wholeheartedly with a design principle 
that says one should not use attributes if we can accomplish what is 
needed with elements alone. However, with emotion here, and with 
several other elements we will be getting to soon, we will need, I 
think, the Secondary Base Schema to handle such things as the 
enumeration of attributes which James began to fill in for culture 
yesterday. Keeping attributes as secondary base elements allows us to 
disassociate such primitives as anger and resentment as types of 
emotion per se. That way they can take their own intensities rather 
than modifying an overall emotional state and allow for better 
computational efficiency.

I happen to be thinking in terms of how to get discrete numerical 
values for various primitives which an application author can then 
choose to implement in any way rather than specifying, for instance 
that resentment is always a modifier of anger.

I decided to use a simple element like emotion to point this out 
because I don't think we have much to quibble about with its 
description/definition.

Thoughts?

Ciao,
Rex
-- 
Rex Brooks
Starbourne Communications Design
1361-A Addison, Berkeley, CA 94702 *510-849-2309
http://www.starbourne.com * rexb@starbourne.com