Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: OntoPaedia [was Physical Grounding [was Foundation Ontology]]



John,

I agree.  Grounding the entirety of the set of symbols that are to be found in 
a mature natural human language like English is computationally intractable.   
Even worse would be to include all types of communication.

An easier task would be to ground an ontology for a specific domain.  
Here is a process to define a specific ontology:
Identify the purpose of the specific ontology.
Identify all the types of facts relevant to the purpose.
Identify the variables used in the fact types.
Identify all the possible items that can be used as values in each variable.
Define all the items.
Identify the terms that are used to define the items.
Identify the boundary terms.  (A boundary term is one that is used to define 
other terms, but is not itself defined by other terms in the ontology.)

In the weak sense of grounding an ontology, the set of boundary terms makes up 
the foundation of the ontology.  Such a foundation can be grounded in the Natural World.

James Nerney



On Sep 1, 2008, at 8:07 AM, John F. Sowa wrote:

James,


That is obvious:


> Isn't it fitting that the Ontological World, the location of

> fundamental concepts, be grounded in the Natural World, the source

> of all concepts?


Nobody objects to that principle.  But there are an open-ended number

of ways in which words or more formal terms are grounded.  Following

is a very brief summary that I sent to these lists last week:


JFS> ... there are three methods of grounding the symbols we use:

>

> 1. Direct experience with the referents by perception and action.

>

> 2. Indirect connections to experience by associations created by

>    patterns of words that are more directly grounded.

>

> 3. Communication by means of natural languages with other people

>    whose grounding for the symbols is more direct than ours.


Each of these three very broad methods can be further analyzed

into an enormous number of issues and details.  Point #1 gets into

all the issues of perception and action in every possible mode,

including every kind of extension to the sensory and motor mechanisms

by means of telescopes, microscopes, microphones, hearing aids, and

animals like seeing-eye dogs, etc.  They extend to every kind of

prosthetic device, and every kind of scientific and engineering

mechanism and methodology, each of which raises an enormous number

of idiosyncratic aspects, problems, and questions.


Point #2 gets into all possible issues about the syntax and

semantics of words and patterns of words.  Those topics have been

analyzed for many millennia by all the philosophers, linguists,

psychologists, lexicographers, logicians, and just ordinary

people who notice and comment on the ways they use language.


Point #3 gets into every mode of communication by any kind of

language or logic, formal or informal, natural or artificial,

transmitted by any means from sound, to gestures, to carrier pigeon,

to the most advanced electronics imaginable.  It extends to

communications by means of books and libraries from people who

lived thousands of years ago, and it includes artifacts dug up by

archaeologists that can tell us something about the experiences

of people long before there was any written form of language.


After you finish working through all that detail and you open an

arbitrary newspaper to an arbitrary page, column, and paragraph --

say page 17, column 1, paragraph 3 of the first one you pick up --

you will discover complexities that you hadn't thought of and,

if analyzed sufficiently far, will lead to open research questions

in philosophy, linguistics, etc.


Good luck,


John Sowa