featural script

[qiihoskeh]

I'm also working on a new featural script, separate from the k/l language. The glyphs are composed as follows:
1. there's a vertical base part
2. and a modifier consisting of additional strokes to the sides (may be null)
3. a horizontal part across the top, representing the vowel
4. possible tone diacritics above the vowel (may be null)

1 and 2 represent the consonant, so each glyph reads as a CV(tone) syllable; 3 also includes a null vowel possibility, so syllables may be CV(T)C.

There are 7 possible bases and 7 possible modifiers (including null); of the 49 combinations 38~43 are usable. I'm trying to decide whether the base represents POA and the modifier MOA or vice versa and which POA/MOA goes with each base or modifier.

I also need to figure out what kind of language goes with the script.

Comments

[dedalvs.livejournal.com]

Your description (absent of any images) reminds me of dominoes. However, I was imagining an invariant base (like Hindi). I can't even begin to imagine 7 different bases. You should see what I'm coming up with in my mind... It looks like plants—specifically, some of the more plantish glyphs in the Bodoni Ornaments font (http://www.fonts.com/findfonts/detail.htm?productid=168246).

Myself, I'm not a fan of straight featural scripts, but if I were to guess (as a reader of the script), I'd assume the base was the manner, and the modifiers the place. This is because I take the base to be more important than the modifiers (visually), and to me, [s] sounds more different from [t] than [t] does from [p] (if you follow my logic).

[qiihoskeh]

Interesting; I've been leaning the other way: I agree with the base being more important, but using that as the reason to use the same base for the (approximately) same POA. For example /t/ with a different modifier becomes /d/.

I wish I had a scanner set up*; it (the script) doesn't look much like plants and less like dominoes. The simplest glyph looks like a T.

* and it will take a while to MSPaint the images -- there are ~40 base-modifier combinations.

[dedalvs.livejournal.com]

Have you ever tried Fontstruct (http://www.fontstruct.com/)? I just started playing with it the other day, and I really like it! (Especially for fonts that are grid-based.)

[qiihoskeh]

Yes, I've done all my fonts with Fontstruct. But I can't use it until I know what characters go with what glyphs! Unless I don't mind moving everything afterwards. I also have to decide if modifiers can be separated from bases or if they have to be combined; probably the latter, since one of the modifiers goes on whichever side is convenient.

[I suppose I should publish some of my Fonstruct fonts, so that they can be seen.]

[4pq1injbok.livejournal.com]

Huh. My intuition is the other way (matching Jeff's): PoA is clearly the one that I'd expect as the base, not manner. Perhaps because manner is the one which is more susceptible to sound change.

[dedalvs.livejournal.com]

Unless you're Hawai'ian. ;)