Friday, July 26, 2002

I think at least one language should have the first person plural derived from 1sg+GEN+PL. I mean, 'me' and 'we' from separate roots is common enough, and 'me' and 'me+PL' too, but I'm want to see 'me' and 'mine'. Maybe in Trentish.

Muke Tever 8:18 AM
[archive/bookmark] [post/read] [Top]
[]

Wednesday, July 24, 2002

If I wrote a CRPG, I wouldn't do that "everyone in the world speaks the same language" thing.

Of course, I wouldn't go to Magic of Scheherazade's extreme, where the Peke Peke tribe have a wonderfully compact language where just about any sentence can be expressed as PEKE PEKE PEKEKE. unless you happen to be Supica the monkey, or the other guy who put him away.

We'd have a PC/NPC with a Translator job, and the engine would just generate (say) random Romance langs, whether by choosing a random set from a fixed list of rules, or just going hogwire. Yeah, I know the old-fashioned RPGs have had language problems modelled since the Elder Days, but I was just thinking what a set of Romance languages would look like if generated by random rulesets.

(Could a million monkeys at a million typewriters reproduce the work of Miguel de Cervantes? Would having to go back and put in all the tildes make it easier or harder?...)

Muke Tever 9:44 AM
[archive/bookmark] [post/read] [Top]
[]

Sunday, July 21, 2002

Some other attempts at Lighan ses Lion:

Muke Tever 1:35 AM
[archive/bookmark] [post/read] [Top]
[]

A fascinating conlang that could use more development is whatever Lighan ses Lion is written in.

>FAN DANCE
Se Dance wave seton leid fus zao pron va. Se seaty sititino poorvo fus se wovel leid puplen, lark se dorop xoprel e nerb pe paint noifi, calmao sikii kinj firpsa ve-shalok sesl bisylen.

It may look like a boryl calami pe zilfany, fire prel at first, but deciphering the majority of it is actually not too hard... although, as expected, the undecipherable words are the ones with the most meaning... what is a turn anyway?

Muke Tever 1:25 AM
[archive/bookmark] [post/read] [Top]
[]

Friday, July 12, 2002

I don't know what it is, but I feel like languages work better when they have a proto-stage to them. So yesterday I was working on a proto-lang for Trentish, especially since I was wondering where all those weird features came from (the rules like -NCh- > -NC'-, etc.)

Well, I didn't entirely solve that problem, but I do have a basic proto-lang.

Consonants - p t c k m n J N l ? s, with an h inserted to satisfy CV syllables.

Vowels - i e @ A o u, both short and long.

Diphthongs--ja je ji etc., wa we wi etc.

Syllable structure: (?)C(l)V(N)

Basic changes: *?C > C... *Cj > Ch... *Cw > Cw... CV > C'V... *tl > tl... *pl > S... *kl > x... *cl > tl... *lj > S... *lw > l... *pw > w... *?t > ? (initially)... *?t > t' (elsewhere)... *c > *S... *@ > V (stressed)... *@ > zero (unstressed)...

Stress was final. Isolating or agglutinating, not sure which yet.

Root structure is mono-, di-, and trisyllables, (?)C(l)V ... the final nasals were grammatical markers in proto-proto trentish (which is manifesting to me even as i speak....)

Muke Tever 1:29 PM
[archive/bookmark] [post/read] [Top]
[]

Wednesday, July 10, 2002

Well, I was reading more of Describing Morphosyntax, and I'm almost sure I read a bit about a language grammaticalizing "contrary to expectations" and "according to expectations" as verbal inflections, and not just verbs on their own.

But now I can't find it. I thought I saw it near the section on miratives, but it appears to have vanished entirely.

One point I did find was "languages tend to have about three distinct core grammatical-relation categories (usually subject, object, and indirect object)....there are two, and possibly three, categories necessary to keep participant roles distinct in normal human interaction without overburdening the mind." (7.0, p. 133)

That sounds like a challenge to me. I got to thinking about a less-related concept, about how language is a process that links the imaginary to the physical: to get an image of a tree from your head into someone else's, you say "a tree".

So, starting from the idea of a language that more acknowledges this, I came up with the idea of "FRAME" and "OBSERVER". Basically the observer is our fourth participant, in addition to the subject, object, and indirect object.

In English we can say things like "I see you're tossing the ball to John" or "I see your tossing of the ball to John". What we want is similar to and extended from this. In that sentence, I is (am?) the observer, you is (are?) the subject, the ball is the object, and John the indirect object. Let's see it in Obs-lang:

ter-tu-mi pet-a Ivan-ta
throw-2AGT-1OBS ball-PAT John-DAT
"I see you're throwing the ball to John."

We call it "observer" because in this language the default action is "see". "I" doesnt (dont?) have to be the only observer, anything can do:

vol-mu-ti Ivan-a
like-1AGT-2OBS John-PAT
"You see (that) I like John."

On a basic note these would probably just in normal use be the same as evidentials.

The beauty of this construct is that the observer can do more than just view the "frame" (i.e., the action: you throwing the ball to John, or me liking him)—we can combine it with other verbs for other effects:

Causative

par Ivan uchi-lu-ti ludi-zi-a
CAUS John learn-3AGT-2OBS play-INF-PAT
"You taught John to play" / "You caused John to learn to play"

Mirative

kata slan-tu-mi kwa azin!
MIR membrum-2SUBJ-1OBS like mule
"Whoa, you're hung like a mule!"

etc...

Ivan-ya naka Mari mansi-lu-li zan-y-a
John-OBS HATE Mary eat-3AGT-3OBS dog-PL-PAT
"John hates it when Mary eats dogs."

mere la faza-tu-mi prite preche
KNOW what do-2AGT-1OBS summer last
"I know what you did last summer."

How's that sound? Yeah, it's rough, but that's the general idea. Notice that the observer-actions don't take the morphology that the agent-actions do. In fact, they are probably adverbs.

This has been Yet Another attempt to reduce to morphology things more often found in syntax...

Muke Tever 9:05 AM
[archive/bookmark] [post/read] [Top]
[]

I know that grammatical gender in Ibran is not anywhere near as prominent as it is in other romlangs, but the article is always eluding me as to its proper form.

I vacillate between el and le. I wonder whether any feminine form *elhe or *lhe should exist. What do yall think?

I do know both of them would reduce to l' before vowel-initial words, if that makes a difference.

Muke Tever 7:27 AM
[archive/bookmark] [post/read] [Top]
[]

After posting "Le nàzeun sin leng êt un nàzeun sin cour" I realized that, even though it technically means "The nation without language is a nation without heart" that 'the nation' does not imply any specific reference as it does in the English.

(This is not a bug, this is a feature!)

I suppose if I wanted to say it about one language in particular I would throw ce in: le nàzeun sin leng c'êt...

Plus it'd be more idiomatic to say le nàzeun sin su leng..., but needless multiplication of revisions keeps the boss at work overtime.

Muke Tever 7:24 AM
[archive/bookmark] [post/read] [Top]
[]

Tuesday, July 09, 2002

And as you may be noticing, I dont really have a satisfactory manner of posting interlinears. Bah.

Muke Tever 7:18 PM
[archive/bookmark] [post/read] [Top]
[]

From a translation on the CONLANG list, the Ibran version I posted:
Nàzeunsinlengêtnàzeunsincour.
"n&tSEnzInlV~XE"n&tSEnzInkwV
"natS9nzInlE~XEh"natS9nzInku:r
nationwithoutlanguageisnationwithoutheart

"A nation without a language is a nation without a heart." Of course I realize now that's bad Ibran, it'd have to be Le nàzeun sin leng êt un nàzeun sin cour because an article or a quantifier is obligatory...

And it's one of those things, I dont know whether it should be leng or lengue. The pronunciation doesn't change, but it's just the spelling that gets me. This is probably why I have to work on a more solid "Old Ibran". Bah.

Muke Tever 7:14 PM
[archive/bookmark] [post/read] [Top]
[]

Monday, July 08, 2002

Tonight I thought Rami should have a word that means "to go contrary to expectations".

Reading about causatives in Describing Morphosyntax, and figured Rami, whose new version has a class of Adjectives/Qualities.. should give them all regular direct causatives, e.g.

  • "written" (adj) + CAUS = "write" (v)
  • "dead" (adj) + CAUS = "kill" (v)
  • "red" (adj) + CAUS = "paint red" (v)

This is pretty standard, but earlier I had in mind an idea that Rami would have no part-of-speech-change derivations, so this is somewhat of a change.

Muke Tever 10:44 AM
[archive/bookmark] [post/read] [Top]
[]

'Interpositions'

Anyway. The idea that set me off tonight was a certain kind of adposition.

You know, we have prepositions ("under the table") and postpositions ("a year ago"). But there's other options too... we have something like it in "water into wine"—I would call that something like an interposition.

So we have a hypothetical interposition-lang ("Iplang"). Here's what it's got.

Regular, "static", locative expressions could use ordinary adpositions:

syh pa dypma ihtan .
man be table under
"The man is under the table."

But more "dynamic" expressions would use interpositions:

syh tu   fydan-ehdu-feha.
man make water in   wine
"The man turned water into wine."

syh ku res -ehdu-lunhan.
man go REFL-in  -corner
"The man went into the corner." (lit: the man went himself into the corner)

These could be used to disambiguate what the semantics of the adposition-word doesn't:

syh bicr kenm lunhan ehdu.
man push girl corner in
"(The man pushed the girl) [and this happened] in the corner."

syh bicr kenm-ehdu-lunhan
man push girl-into-corner
"The man pushed (the girl in [into] the corner)

This assumes that, unlike English, the interpositional phrase is an indivisble unit, so you can have lunhan ehdu syh bicr kenm "in the corner the man pushed the girl" but not *ehdu-lunhan syh bicr kenm "into the corner the man pushed the girl" (it can only be kenm-ehdu-lunhan syh bicr.)

How's that sound?

Muke Tever 10:24 AM
[archive/bookmark] [post/read] [Top]
[]

I figured I should set up a blog for my conlang ideas since:

  • I never get around to implementing most of them anyway, so I might as well share;
  • It's easier to search a webpage than random notebooks;
  • and it's less hassle to update a blog than an ordinary website.

I could post all this stuff to CONLANG à la Abrigon, but I'd worry about stressing the list's posts-per-day limit... and besides, sending email is quirky since half the time I'm not allowed to send email through my email servers.

Hence this page. Share and enjoy.

Muke Tever 9:28 AM
[archive/bookmark] [post/read] [Top]
[]

This page is powered by Blogger. Isn't yours? I'm using blogBuddy.