Re: SUO: RE: Model of Activity and Action in SUO ontology
>
>
>With only 62 primitives, simulation software could actively
>> interpret a significant coverage of English in terms of
>> these primitives, and possibly could provide an operational
>> semantic model of linguistic meanings. 62 objects is a
>> very manageable software undertaking, and with definitions
>> that can be organized relationally, composition of the 62
>> objects might result in active interpretation of most sentences
>> using the restricted vocabulary.
>>
>> Does this analysis seem feasible to you? If this is fairly
>> well known material to linguists, why hasn't someone already
>> done such a project? Or has someone tried and found a deeper
>> truth that nullifies these descriptions?
>
Problems: the semantics of the notation used in the NSM is not particularly
defined... This has been recognised/realised by the NSM-folks relatively
recently and they are moving / have moved? towards looking at the
"grammatical meanings" that their paraphrases rely on. The situation
is substantially more complex than it sounds: it is actually the
semantic configurations in the NSM-paraphrases that do a lot
of the work. Focusing on the word-like entities rather than the
where the real engine of the meaning-making system of language
lies--the grammar and the semantics--is usual with less
developed approaches to the phenomenon and complexities of language
(e.g., the ancient Greek).
>>The Word (CognitiveDescriptorClass) would have to be
>>expanded into subclasses that have linguistic meanings,
>>such as verb, noun, preposition, phrase, and so on as
>>needed to organize sentences, paragraphs and discourses.
>>
>>Then the individuals mentioned in sentences could be
>>represented as instances of one or more classes, so at
>>the prime level, reformulated sentences in pure prime
>>vocabulary plus individual names could be interpreted.
>
>Does anyone know of any similar research work that
>might have been done already?
>
You mean apart from the 30 odd years of work on interlingua in machine
translation?
John B.