Dŭhog (Collaborative Project), Now: Nominal Morphology, etc.

[clawgrip]

Example:

Thrice's proposed orthography:

クㇼ kru
コㇼ kro

My proposed orthography:

クㇽ kru
クㇿ kro

クㇼ and コㇼ should be kur and kor.

[loglorn]

Small り adds an intersyllabic /r/, so a CV sign plus small り results in CrV. This is not how small signs work in hiragana/katakana for Japanese, Okinawan or Ainu, which is why I recommended against it.

Example:

Thrice's proposed orthography:

クㇼ kru
コㇼ kro

My proposed orthography:

クㇽ kru
クㇿ kro

I got that. I was puzzled about the small "i", my bet right now is that Cŏ+i = Co.

Anyway, it needs to be able to contrast 'kro' and 'kor' since both happen.

Actually there's a bunch of orthographic minutia that needs to be discussed. For example, for marking lone consonants there's a devoweling diacritic, but does it get applied to a particular standard CV glyph (likely Cu) or is it more etymological?

[clawgrip]

This should be sufficient, I think (using /k/ as an example, but the same pattern can be applied to all the others):



I'm making the small letters 80% size and dropping them slightly below the text line, which is what MS Mincho does. If it is unclear, I suppose I could exaggerate it a bit. But I also did this by hand in Photoshop, not with a font or anything, so it is not uniformly aligned or anything.

[Thrice Xandvii]

I'm a fan of Clawgrip's system... except for one thing. I'm not sure how I feel about using <ŭ> to mark the <Cu> syllables.

I took to using the small <i> to write <o> and <e> because it serves a function similar to i-umlaut, noting a rise in the vowel qualities. That same thing doesn't hold true for <ŭ>, nor does it make as much sense for /əə/ to represent /ɯ/. As such, I chose to use the long mark to indicate what used to be <ū> in Japanese, whereas short <u> got changed to a schwa in many cases. I'm open to suggestions here.

Further, to be consistent, I think I will have the vowel killer mark only apply to <Cŭ> since it will then better fit the system that already exists of ignoring the <ŭ>s that will be in those above combinations.

---

(I'll be editing in a step by step here of the transformation from Clawgrip's "name" to the version in Dŭhog soon, here.)

[clawgrip]

As a general rule, ー is not used with hiragana, so I wouldn't bother with it at all. Long u is written with a Cu sign followed by う. As for which small vowel letter is used to make which compound vowel I don't really have any opinion and was only going by the old image from the other thread. My post was mainly focused on the /r/ of CrV and CVr.

[Thrice Xandvii]

As a general rule, ー is not used with hiragana, so I wouldn't bother with it at all. Long u is written with a Cu sign followed by う. As for which small vowel letter is used to make which compound vowel I don't really have any opinion and was only going by the old image from the other thread. My post was mainly focused on the /r/ of CrV and CVr.

Yeah, I was aware of that. Maybe we should go with <ŏ>? Does that make any more sense? Now I'm unsure.

Also, does that mean that technically the <ćV> sequences which are currently formed by <kV>+ small <yV> should all change to <kŭ> + small <yV>?

(I'll be editing in a step by step here of the transformation from Clawgrip's "name" to the version in Dŭhog soon, here.)

Now that I look at it again... I don't think that will be necessary. I did two things that make my changes align much more fully with yours: I am now notating <CC> in my source document with <Cː> and I moved the change that deletes long consonants to BEFORE the <Cj> segments get altered. Between those two things, I think everything aligns much more closely. (Including getting the same krokrŭb as you did.)

---

What would the differences in the numbers be? Preferably coming with the sound change log. If i had to guess, i'd say it's got to do with the treatment of geminates.

These are the numbers I got:
100 hyaku → hak
200 nihyaku → dŭhak
300 sanbyaku → cŭtdak

• sanbjaku → sãnbjaku ( V → V[+nasal] / _N )
  • → sãdbjaku ( n → d / V[+nasal]_ )
  • → sədbjaku ( ã → ə / _ )
  • → səddaku ( bj → d / _ )
  • → səddak ( u → ∅ / _# )
  • → ɕəddak ( s → ɕ / _ )
  • → ɕətdak ( CC → [−voiced][+voiced] / _ )

400 yonhyaku → yotgak
500 gohyaku → gŏhak
600 ropːyaku → rŏtak
700 nanahyaku → dŭdahak
800 hapːyaku → hatak
900 kyuhyaku → ćuhak
1,000 sen → ci
2,000 nisen → dŭci
3,000 sanzen → cŭtzi

Did you get rid of the final devoicing? I liked that, and it made sense that all coda were devoiced.

I didn't get rid of it... but it does ignore <r> since I had toyed with the idea of making vowels rhoticized instead of actually pronouncing the /ɾ~ɺ/. I figured it wasn't toooo far-fetched to have a rule like that ignore one class of consonant (namely rhotics/liquids.)

1. Keep it ɦ. That ɦ probably wouldn't be phonemic but an allophone of /h/ after consonants. If we do that, i think the voicing of ʔ̬ should be ɦ too.

2. Get rid of it. Either ɦ -> ∅ or ɦ -> g. I find ɦ -> g to be more plausible if ʔ̬ → g is indeed installed.

Hmmm. Personally, I like the simplicity of #1... but /g/ is a favorite phoneme of mine... *goes to mess with sound changes a bit* Okay, I'm back. I think we'll go with the ɦ → g version of #2. (Which includes using ʔ̬ → g.)

I liked that. Not sure about plausibility and whatnot, but i like it.

Yeah, I have no idea about the plausibility either, but –̈ći does have a certain appeal... but, it also removes the need for disambiguating in any situation with -da, which could be good or bad. I'm unsure.

And, I've messed around with a few things with the dĕhrŭkda... again.

Edit: Editing in my most recent version of the Chart! I redrew a few of them, and made some minor alterations to others. <Qe> and maybe <he> are likely the biggest changes. For <qe> I based it on <ce> and the main swoop of <ka>. As for <he> I found another character on that webpage that used 阝 as a secondary element and used that. I've also been toying with the idea of something like the below image as a further simplification of <co>.

[clawgrip]

Here's my take on the sans serif version:

[Thrice Xandvii]

Krokrŭp, that looks pretty amazing! Thanks for making that!


(I am a little unsure about <ba> and <pi> though.)

[clawgrip]

Thanks. I was least happy with po, personally, and was not sure what to do. In the Mincho style one, you can see that the bottom horizontal is wider than the top horizontal, but in the sans serif one, they are about equal width. I don't like this, but the reason I kept it this way is because I don't want the triangular hole in the top right to get any smaller than it already is, meaning I can't pull the bottom one any further to the right. I suppose I could lose the curve at the bottom left, but I'm also not so fond of mostly curveless hiragana. Seems off somehow.

Also u should have the same top curve as e, but I accidentally gave it the same one as ra. It's slightly out of proportion. (EDIT: I fixed this)

On another topic, have you touched on grammar at all yet (besides the particles-turned-case endings)?

[Thrice Xandvii]

Yeah, I noticed the thing with <û>, but its a fairly minor issue in my book.

That's about as far as we have gotten as far as grammar. We were/are still assuring we have the sound changes down and from there we will stay on the same page throughout and can mess around with stuff more readily.

On that topic, any idea what a copula-fused case ending might mean?

[clawgrip]

What's a copula-fused case ending? The copula becomes a suffix?

[Thrice Xandvii]

I believe so, it was an idea loglorn was tossing around up-thread.

[clawgrip]

I really don't know what it could mean. I don't see much detail on it.

[gestaltist]

What's a copula-fused case ending? The copula becomes a suffix?

Copula often fuses with verb stems to create new tense forms, so maybe something analogous but with nouns?

[loglorn]

As a general rule, ー is not used with hiragana, so I wouldn't bother with it at all. Long u is written with a Cu sign followed by う. As for which small vowel letter is used to make which compound vowel I don't really have any opinion and was only going by the old image from the other thread. My post was mainly focused on the /r/ of CrV and CVr.

Yeah, I was aware of that. Maybe we should go with <ŏ>? Does that make any more sense? Now I'm unsure.

Also, does that mean that technically the <ćV> sequences which are currently formed by <kV>+ small <yV> should all change to <kŭ> + small <yV>?

What's most in line with modern Japanese hiragana usage would be having <ćV> as <Ki-yV> (Large ki, small yV, as it might not have been clear)

---

As for the compound vowels, i'd have <Ŏu> for o, <Ĕi> for e and <Ŭu> for u. Because /o/, /e/ and /u/ stem mostly directly from ō, ē, ū, which are written in modern hiragana as Cou, Cei, Cuu. Doesn't make all that sense syncronically indeed, but etymologically it makes the most sense.

What would the differences in the numbers be? Preferably coming with the sound change log. If i had to guess, i'd say it's got to do with the treatment of geminates.

These are the numbers I got:
100 hyaku → hak
200 nihyaku → dŭhak
300 sanbyaku → cŭtdak

• sanbjaku → sãnbjaku ( V → V[+nasal] / _N )
  • → sãdbjaku ( n → d / V[+nasal]_ )
  • → sədbjaku ( ã → ə / _ )
  • → səddaku ( bj → d / _ )
  • → səddak ( u → ∅ / _# )
  • → ɕəddak ( s → ɕ / _ )
  • → ɕətdak ( CC → [−voiced][+voiced] / _ )

400 yonhyaku → yotgak
500 gohyaku → gŏhak
600 ropːyaku → rŏtak
700 nanahyaku → dŭdahak
800 hapːyaku → hatak
900 kyuhyaku → ćuhak
1,000 sen → ci
2,000 nisen → dŭci
3,000 sanzen → cŭtzi

sanbyaku -> cŭtdak is a clear input issue. Inputting sambyaku will probably work.

yonhyaku and sanzen worry me more.

For me i have:

saɴzeɴ -> sãzẽ -> səze -> səzi -> ɕəʑi

Which is clearly not what's happening in yours.

[Thrice Xandvii]

What's most in line with modern Japanese hiragana usage would be having <ćV> as <Ki-yV> (Large ki, small yV, as it might not have been clear)

---

As for the compound vowels, i'd have <Ŏu> for o, <Ĕi> for e and <Ŭu> for u. Because /o/, /e/ and /u/ stem mostly directly from ō, ē, ū, which are written in modern hiragana as Cou, Cei, Cuu. Doesn't make all that sense syncronically indeed, but etymologically it makes the most sense.

I actually used that system originally... but I turned away from it when I realized using schwa for those functions was somewhat illogical. However, going back to it might be the best solution overall. Make it official.

As for <KI>+<yV>, I can definitely be okay with that. Plus, spelling from Japanese is retained in other places... like using a <na> glyph in places where now it is a <da>, but using a <ta"> when it isn't from a nasal (" being my shorthand for a dakuten).

sanbyaku -> cŭtdak is a clear input issue. Inputting sambyaku will probably work.
yonhyaku and sanzen worry me more.

For me i have:
saɴzeɴ -> sãzẽ -> səze -> səzi -> ɕəʑi

Which is clearly not what's happening in yours.

Clearly you are better at inputting Japanese than I! I suspect that all these issues go back to me being lame at rendering Japanese properly in my source file for the sound changes. I mean, I'm never quite sure when to use /ɴ/ and when it becomes some other nasal, for instance, so as a result I usually just use <n>.

I think boning up on some of those things will be my project for tonight.

[loglorn]

What's most in line with modern Japanese hiragana usage would be having <ćV> as <Ki-yV> (Large ki, small yV, as it might not have been clear)

---

As for the compound vowels, i'd have <Ŏu> for o, <Ĕi> for e and <Ŭu> for u. Because /o/, /e/ and /u/ stem mostly directly from ō, ē, ū, which are written in modern hiragana as Cou, Cei, Cuu. Doesn't make all that sense syncronically indeed, but etymologically it makes the most sense.

I actually used that system originally... but I turned away from it when I realized using schwa for those functions was somewhat illogical. However, going back to it might be the best solution overall. Make it official.

As for <KI>+<yV>, I can definitely be okay with that. Plus, spelling from Japanese is retained in other places... like using a <na> glyph in places where now it is a <da>, but using a <ta"> when it isn't from a nasal (" being my shorthand for a dakuten).

sanbyaku -> cŭtdak is a clear input issue. Inputting sambyaku will probably work.
yonhyaku and sanzen worry me more.

For me i have:
saɴzeɴ -> sãzẽ -> səze -> səzi -> ɕəʑi

Which is clearly not what's happening in yours.

Clearly you are better at inputting Japanese than I! I suspect that all these issues go back to me being lame at rendering Japanese properly in my source file for the sound changes. I mean, I'm never quite sure when to use /ɴ/ and when it becomes some other nasal, for instance, so as I result I usually just use <n>.

I think boning up on some of those things will be my project for tonight.

My sound changes incorporate ɴ so stuff like <ren'ai> doesn't get parsed as re.nai by the sound change applier.

But back to the point i was trying to make, your changes display N → ∅ / Ṽ_#, while mine have N → ∅ / Ṽ_[#C], which effectively, along with the other changes, remove all coda nasals. That's a detail that explains some slightly different outputs we're getting.

[Thrice Xandvii]

Okay.

I'll just have to edit in "N → ∅ / Ṽ_C" just after where the rest of the nasals delete and we should be golden!

[loglorn]

Okay.

I'll just have to edit in "N → ∅ / Ṽ_C" just after where the rest of the nasals delete and we should be golden!

Probably. One last test: Run 交通事故 kōtsūjiko (traffic accident) and tell me what you've got.

[Thrice Xandvii]

Add one with some funky nasal placement, too... Then I can give you an answer tonight.