Theiling Online    Sitemap    Conlang Mailing List HQ   

THEORY: METAGRAM -- Pt.1

From:Caleb Hines <cph9fa@...>
Date:Wednesday, December 17, 2003, 7:17
I wanted to share a system I'm working on developing, tentatively called
METAGRAM. With the plethora of loglangs out there its probably unneccesary,
but I do it mostly for my own benefit. It's not really anything fancy, just
some notational conventions I try to use when analyzing sentences or
grammars. It's still very much developmental and any
comments/suggestions/criticisms are welcome. I was inspired by a
combination of Latin, AllNoun, Phrase-structure grammar, Backus-Naur form,
and alot of my own ideas. Hopefully, when I complete it, it will be a
powerful tool for analyzing the grammar of various languages, but I'm
primarily developing it as a sort of prototype for my conlang Akathanu's
grammar, but also as an inter-conlingual testbed for experimenting with
grammar. The lexicon would be entirely English, making this a "regram" of
that language, although any other lexicon would be acceptatble. The point
is to remove (as much as possible) the details of vocabulary from the
syntax. The main features of METAGRAM are these:

Remove all articles. There is no difference between relative and
non-relative clauses. All sentences are essentially noun clauses/phrases.
Each noun clause consists of a subject (or list of subjects), followed by a
list of states in which that subject is in. This definition is recursive.
You'll notice that as a result, all phrases are head-first
(right-branching). Also, as much as possible, I'm trying to merge
adjectives and verbs (and maybe eventually adverbs) into a single part of
speech called "states." Eventually, states may even become funcionally
equivalent to nouns. Active voice verbs should be converted to the
present-participle, and passive-voice verbs should be converted to the
past-participle. (Tenses, moods, and asspects may be added in a later
expansion pack). I'm not going to allnoun's extreme of eliminating
everything other than nouns. Words like prepositions are useful at the
present, because they allow nouns to become states (for example, the noun
"bed" becomes the state "in the bed"). Conjunctions are also useful,
although I don't currently have a rigorous way of dealing with them. I may
even keep a rudimentary verb system for conjugating states (the verb
expansion pack I mentioned). Having given a brief overview, let me go into
a bit more depth.

In BNF form, METAGRAM can be written in terms of the following rules:

Noun phrases:
01) N -> noun
02) N -> N {P}
03) N -> (N, N, ...)

Predicates (or states):
04) P -> N
05) P -> prep N
06) P -> adj  | adj {M}
07) P -> verb | verb {M}
08) P -> P, P, ...

Modifiers
09) M -> adv
10) M -> prep N
11) M -> M, M, ...

General translations:

N        = "A/an/the N" or "There is a/an N."

N{P}     = "N, which is (in the state of) P.",
           or "There is an N which is (in the state of) P."

(N1, N2) = "N1 and N2" or "There is an N1 and an N2."

N1 {N2}  = "N1 is equivalent to N2."

N {A}    = "N is in the state described by A." (or "N is A.")

N {V}    = "N is in the state of doing action V."

N {P1, P2, ...} = "N (which) is in the state P1, and in the state P2,
and..."


...and some examples:

I
"I (am)"

I {student}
"I am a student."

ball
"The ball..." or "There is a ball."

ball {red}
"The red ball...", "The ball is red.", or "There is a ball which is red."

ball {red{dark}}
"The dark red ball...", "(There is) a ball (which) is dark red.", etc...

ball {red, rolling}
"The ball which is red and which is rolling." or "The red ball is rolling."


ball {red, rolling{quickly}}
"The red ball (which) is rolling quickly."

window {broken}
"The window (which) is broken (by someone)."

*Notice the difference between the participle forms of the passive voice
and active voice verbs in the last three sentences. Also notice that so far
I've only used intransitive verbs. In this system, any transitive verbs
treat _all_ arguments (other than subject) as prepostional adverb phrases,
modifying the original verb. I'll show it below in a few progressive
examples:

wing {new, of library}
"The new wing of the library"

He {opening}
"He (which) is opening (something)."

opening {of wing{new}}
"The act-of-opening, which is of the wing which is new.", or
"The opening of the new wing"

In active voice, we have a transitive verb like this:
He {opening{now, of wing{new, of library}}}
"He which is opening now of the wing which is new and which is of the
library."
or "He is now opening the new wing of the library."

Or, in passive voice:
wing{new, of library, opened{now, by him}}
"The new wing of the library is opened by him."

And of course, the classic:
He {giving{of book, to John}}
"He is giving the book to John."


That's enough examples for now. I haven't shown everything about METAGRAM,
for instance quantifiers, conjuctions, interjections, conjugations, a
method for distinguishing between restrictive and non-restrictive clauses
(hint: apply rule 2 twice), and at least one known flaw (re: ambiguity of
predicates). But for now, I think this is a long enough introduction. I'll
post more later on it if anyone's interested, along with some more
translations. Let me know what you think, and don't hesitate if you have
any comments, suggestions, or criticisms. Also, chalange me with texts to
translate! As part of the "debugging" process, I want to find any possible
way this system can be broken!

Also, note that this system is very noun-centric. I have a vauge idea for a
verb-centric version of METAGRAM too, but I'm not too interested in it, at
the moment. I might post about it in the future. Nouns seem a more natural
way to go, though.[*]

Thanks,
~Caleb

[*] After typing this, I began skimming through the beginning of Fillmore's
"The Case For Case" in _Universals in Linguistic Theory_ from my Uni's
library, and it seems what he does is similar to my vauge concept of a
verb-centric METAGRAM. OTOH, E. Bach's following article about relative
noun phrases and an underlying "base class" encompassing nouns, adj's and
verbs, is much more similar to this current noun-centric system.

Reply

Gary Shannon <fiziwig@...>