To: [email protected]
Subject: Re: [NGL] Proposed Revision of Vocabulary Rules for Public Posting
From: Jack Durst
Date: Sat, 28 Aug 1999 23:08:15 -0700 (PDT)

From: Jack Durst

Good work, Stephen.

As the body of this proposal was debated in full the last time and all of the major modifications were debated as they arose, or, at the very least, garnered no objection and have since been the accepted practice, I see no need for a through examination here. Current practice and existing rules are re-stated correctly.

As for the few nits I did find:

Modules: The word "morpheme" is, by definition the smallest meaningfull part of a word; if a morpheme could be generated by a module, then the parts from which it were derived would be the morphemes, not the larger unit (which, if it had any special meaning beyond the morphemes, would be a lexeme or a word) Words and grammar features, not morphemes, are generated by modules.

I prefer the view of core grammar as generics plus pick a module statements as opposed to the set of modules ratified for a reason: it makes the theoretical structure of the language much simpler, and has practical ramifications. In order to speak grammatical NGL, one really need know only one of the three word orders, even though all are official and part of the same module; likewise, to speak grammatically, one need not know number at all, as generics can be used throughout. Even something so general as general derivation is modular, as the most minimal possible definition of the language need not include it. Indeed, modules do not become official until ratified, but they remain modules, outside of the pure most minimal description of the language; which can have consiquences as to both their description, ratification, and use.

I, personally, would be more comfortable defining a "modular morpheme" as follows:

A morpheme to which a special rule of grammar is attached in its use within a module; leading to a different lexical meaning.

The definition is vague, this definition would provide a bright-line rule as to weather a morpheme belonged to a module.

The module rules state that modular morphemes can (and indeed strongly implies that it is desirable to) be shared between several modules. (Number, for example, is carried over as a whole from nouns, to verb and adjective agreement) so there is no restriction on the cross-borrowing of parts from other modules provided that the meaning is the same, generics are preserved, and the rules applied don't interfere with each other.

Outside of modularity, I see nothing wrong with the proposed revision.

Sincerely,     Jack Durst
[email protected]
[this posting written in Net English]

 
 

back


 
Hosted by www.Geocities.ws

1