Re: [xml-dev] RE: Keep business-process-specific data separate?

[Frank Manola <fmanola@acm.org>]

On Feb 2, 2009, at 1:14 PM, Michael Kay wrote:

>>> I won't comment on the success or failure of RDF, but it
>> seems to me
>>> that it's highly abstract, not necessarily highly generic.
>>
>> Ok, I'll bite.  What's the difference?  (Seriously...)
>
> Excellent question. I'll hazard an answer.
>
> The class of "everything that moves" is generic (because it embraces  
> lots of
> different kinds of thing), but it's not abstract (because it's a  
> simple
> partitioning of all things based on a simple observable property).
> Abstraction demands more than dividing objects into categories; it  
> demands a
> creative leap, involving the discovery of new ways of looking at  
> things,
> yielding new insights.

It seems to me that the class "everything that moves" is both generic  
(because, as you say, it embraces lots of different kinds of thing)  
*and* abstract, because it took an act of abstraction to go from a lot  
of individual things that move to come up with the  (or at least one)  
property that characterizes all of them (at least in this particular  
class), namely movement.  I don't think the distinction has to do with  
*how* creative you have to be.  Another example would be numbers (one,  
two, etc.).  Numbers are abstractions.  We might not think of the idea  
of number as being particularly creative today, but it took some time  
for people to come up with the idea of "twoness" as distinct from two  
sheep, two cows, two goats, etc.

BTW, is this discussion abstract, generic, both, or neither?