/moacng/

A forum for all topics related to constructed languages
Post Reply
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

/moacng/

Post by eldin raigmore »

/moacng/ is a sketchlang.
Max Koelbl and I have been working on it.
We haven’t settled on a phoneme inventory but I’ll mention one we’ve been thinking of.

I read the sticky post about rules for posting a conlang here, and I think this one passes muster.
However a mod or administrator might disagree.
If one of them moves it I’ll accept their decision;
I do request they explain it to me, and/or tell me where they move it to?

This /moacng/ all centers around a purely phonological way to construct noun-classifier-words and anaphor-roots (or should I say pronoun-roots?) from nouns.

The constructed words all take the form
/mV1V2C1C2/
where
V1 is the first vowel of the root of the antecedent or coreferent noun
C1 is the first consonant of the root of the antecedent or coreferent noun
V2 is the last vowel of the root of the antecedent or coreferent noun
C2 is the last consonant of the root of the antecedent or coreferent noun.

If V1 and V2 are the same the nucleus will be a long monophthong instead of a diphthong.
If C1 and C2 are the same the coda will be a long or geminated consonant instead of a two-consonant cluster.

One phoneme-inventory we’ve considered and are considering is the set of nine consonants and three vowels that PHOIBLE.org says are included in >66.667% of their phoneme-inventories;
/m k j p w n t l s/ and /i u a/.
Right now, though, that doesn’t make it easy to form the classifier-word for “conlang”, which is what /moacng/ is supposed to be.
So maybe we’ll have to come up with a native word for “conlang” and change the name of the sketchlang;
or maybe we’ll have to change the phoneme-inventory.
(Or the romanization, I suppose? Maybe?)

These words are used as numeral classifier words and as possessive classifier words.
So they optionally inflect for grammatical number and/or optionally inflect for grammatical case or state (viz “construct state”).
They inflect by taking a suffix, which may fuse number and case-or-state.

They are also used as anaphora and/or pronouns, or as the roots of anaphora or pronouns if those must be inflected.

They can classify proper nouns as well as common nouns.
But ordinarily if they classify proper nouns they co-refer to a group of two or more entities which share a proper noun, such as
/maown/ “the Washingtons”.

A native noun-root with at most one vowel and at most two consonants, or with at most two vowels and at most one consonant,
is considered “a stone-age noun” and is its own classifier-word and “its own pronoun”.

A loan word with at most one vowel and three or more consonants, or with at most one consonant and three or more vowels, is altered to adjust to /moacng/ phonology.

We assume so far that no other morpheme begins with a syllable-body whose onset is /m/ and whose nucleus is a long vowel or a diphthong;
nor does any other morpheme end with a syllable-rime whose nucleus is a long vowel or a diphthong and whose coda is a two-consonant cluster or a geminated or long consonant.
Any other way a ...mVVCC... string can arise in a word must be sandhi across a morpheme boundary and across a syllable boundary.

____________________

With the nine-consonant three-vowel inventory just proposed there would be 729 possible classifier-words and pronoun-roots.

Real-life natlangs with noun-classifier words instead of genders tend not to have many more than around 300 or 400 classes.

Any phonology in which there are 18 or fewer CV syllables and 18 or fewer VC syllables would, if the language used this scheme, produce 324 or fewer possible noun-classifier-words or pronoun-roots.
But with 24 or more CV or VC collocations up to 576 or more candidate classifier-words might be possible.

____________________

Maybe these things could apply to adjectives too; I haven’t thought about it enough, I haven’t even mentioned it to Max yet.

We don’t have any plans to make verbs.
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: /moacng/

Post by eldin raigmore »

I’ve been thinking about this some more.
I’m currently toying with the following ideas.

Most (ie nearly all) one-syllable words are (C)V(C).
Most word-initial syllables are (C)V.
Most word-final syllables are CV(C).
Most word-medial syllables are CV.

There are up to 6561 CVCVC words, and almost all quotidian speech can be conducted with few words longer than that, or otherwise outside that set.
There are up to 59,049 VCVCVCV words, and most expert-level discourse can be accomplished without much use of words longer than that, or otherwise outside that set and the previous set.
There are up to 177,147 CVCVCVC roots. Nearly all nouns and adjectives and verbs have roots in that set. All root-morphemes for those parts-of-speech are that long or shorter or contained in one of the previously mentioned sets. So /moacng/ could have as many roots as the OED says English has.
There could theoretically be up to 1,594,323 words shaped like VCVCVCVCV. I think that’s as many one-word lexicon-entries as there are in the OED.

Longer words could arise through affixation (probably inflexion) and/or compounding, I suppose; I think it’s probable.

The actual maximum syllable structure is probably going to be something like (C)V(V)(C)(C).
(Here VV also includes V: (long monophthongs) and CC also includes C: (geminated or long consonants).)
But diphthongs, long vowels, coda clusters, and geminated codas, should be rare in the lexicon, even if some words containing such features occur frequently in speech. Such phonological features should signal that the word or morpheme that contains it belongs to some small closed class.

….

I wonder whether that’s logically consistent?

Maybe I should post a “Yay or Nay” question about it.
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: /moacng/. Maybe it should be /muakn/?

Post by eldin raigmore »

Should it be /muakn/, maybe?
And the word for “conlang” be /kunlan/?
….
There potentially could be up to
300 (C)V(C) syllables, morphemes, or words;
8,100 (C)VCV(C) morphemes or words;
218,700 (C)VCVCV(C) morphemes or words; and up to
5,904,900 (C)VCVCVCV(C) words, which would be a very large lexicon that would probably imply either a lot of compounds or two-word phrasal lexemes or even polysynthetic morphology.
Post Reply