[YG Conlang Archives] > [katanda group] > messages [Date Index] [Thread Index] >


[Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: [katanda] Design Philsophy



 > saluton, kapitano,

Kaj saluton al vi.

 i guess ram's observation was that while modeling a language, there
 are very often arguments that are very close to the verb, so that it
 seems obvious to call them core arguments... i also was wondering if
 we could use core arguments with special ... rolmontriloj ...
 "prepositions", but that's just a question of implementation (a
 question of how you want to represent core arguments).

 i mean: core arguments exists, don't you think?

Yes, but I'm not sure that they're necessary. And I'm not sure
that the core argument structure of verbs in the source language
should be retained in the translation interlanguage. Partly because
the 'equivilent' verb in the target language may have a different
core structure. And partly because I think it makes the
interlanguage unnecessarily rigid, which may cause problems
when translating unusual or highly idiomatic sentences.

The sentence...

'The child silently put the large battered book on the dusty table.'

...can be analysed like this:

Verb(1): Put
Verb(2): Silent
Agent: Child
Patient(1): Book
Patient(2): Battered
Patient(3): Large
Destination(1): Table
Destination(2): Dusty
Instrument: [none]
Location: [none]
Beneficiary: [none]
Tense: Past
Aspect: Completed

(Note that I'm not (at the moment) making a destinction between
Patient and Focus.

Note also that some of the roles can be filled
with more than one term - not necessarily with one of the terms
being the head and the others being modifiers. This goes also
for the verbs, which can be regarded as being contained within
their own prepositional phrase.

Finally, some of the arguments are left blank.)

In English, the example sentence 'The child silently put the large
battered book on the dusty table.' actually has *six* core
arguments:
The Agent
The Patient
The Destination
The Tense
The Aspect
The *Verb*

And yet we have no qualms about turning tense and aspect
into oblique arguments in an interlanguage. Indeed, any
subset of those six could happily be made oblique or
omitted, and we would be left with a stable propositional
structure. We may no longer be obliged to specify *where*
the book was placed, or who placed it there, but the
result is still meaningful. It still conveys information.

I realise that I've made a switch from lexical semantics to
sentential semantics And that I'm treating the verb as an
argument of the sentence, rather than the nouns as aruments
of the (main) verb.

what is a sentence without verb? i wouldn't have tried and made the verb an argument of the sentence. please explain.

the rest i think to understand quite well. and i think you're right. thank you!
but the consequence is that the core arguments should remain, and that all the other stuff like tense should be added to it (which would simply mean that everything is an argument that can be reduced to middle-voiced, passive-voiced or simply not mentioned).

This is obviously a different design philosophy from that of
Katanda, but I suspect that some problems Rick has
identified in Katanda are the result of beginning with
morphemes and building them into sentences, as opposed
to beginning with sentences and breaking them down into
semantic units.

I'm quite prepared to be proven wrong on this. Perhaps
Rick could say why he views morphemes rather than
sentences as the 'units' of communication.

stefo,
sts.