That durned î, re-revisited

[Sir C. M. Siervicül]

So, a few clarifying questions about the proposal to restore î. This list is based on discussions I've had with Lord Hooligan, who identified a number of practical issues that will have to be addressed in deciding how to implement this proposal.

1. Would all instances of î be restored as of 1997, or only some?  Sir Tomás proposed, and said that he and Ben had agreed, that it should be eliminated where it appeared under the influence of a following n or m, but not where it existed as an independent sound. Dame Miestra, do you recall any further details of what he proposed on the Republican Witt?
 
2. If the answer to î is "only some," should we apply some black and white rule like "î before n or m becomes i, î elsewhere stays î," or something more nuanced? For example, should î remain î in some cases before n or m on etymological or other grounds?

3. If î goes to i before n and m (either generally or with exceptions), is this because the î pronunciation is now recognised as a regular allophone of i in those environments?

4. What about the words where the phoneme was spelled â (e.g. quând and sânc)? Should the î pronunciation be restored there too, and the words respelled with whatever grapheme is settled on for î?
 
5. What about the present participle ending -înd (-ind under A2007), which does not have the î sound (it's pronounced as though spelled -ant)?

6. Now about orthography. Should ï be the new grapheme for î, so that the umlaut consistently represents a difference in vowel quality? If so, are there acceptable alternatives for marking diaeresis in all cases where ï is currently used under A2007? And what will the rules for î/ï be in terms of effect on c and diphthong formation? For example, does pre-A2007 cînt ("hundred"), currently chint, become chïnt or cïnt?

[Ián Tamorán S.H.]

There have been, elsewhere, comments that î represents (or represented) a vowel which it is difficult for English speakers to pronounce. I, personally, disagree with this: I am a (native!) speaker of English, and I can say this vowel almost perfectly. I say "almost" because I am not an anything-like-native Russian speaker (in fact, I'm terrible at it!), but did study Russian for over a year. When I listen to recordings of Russians saying î (ы) and my saying it, they sound the same to me. I leave it to others to describe the requisite tongue position properly: personally I do it simply by copying the sound - but then, I am a musician!

[Magniloqueu Épiqeu da Lhiun]

To be honest, I would love a proper Skype discussion with all those concerned. The main reason I never reply is that I just do not want to write around about all these proposals. I do want to discuss them as personally close as possible, though.

[Alèx Soleighlfred]

I believe this is more of a psychological difficulty than anything else. That's because Ы doesn't exist in English and sounds alien. I'm sure that trying to repeat after a native speaker solves the problem well enough!

(for the record we aren't aliens. Just so that we escape any misunderstanding, I said "sounds alien" as a characteristic adjective of the sound itself, not as a hint on extraterrestrial origin of any of the Slavic nations. Certainly not a hint. Wink wink)

[Marcel Eðo Pairescu Tafial]

That's because Ы doesn't exist in English and sounds alien.

Couldn't the same be said about the umlauted vowels? Or some consonants like C'H [x], GH [ɣ], GLH [ʎ], GNH [ɲ], L [ɬ] before T and the like?

[Alèx Soleighlfred]

Jan 7, 2018 11:46:47 GMT -6 Marcel Eðo Pairescu Tafial said:

That's because Ы doesn't exist in English and sounds alien.

Couldn't the same be said about the umlauted vowels? Or some consonants like C'H [x], GH [ɣ], GLH [ʎ], GNH [ɲ], L [ɬ] before T and the like?

Well, I think people are (mostly) more aware of pronetical trategies of other European languages, rather than those of Russian. I mean, most people won't have a problem figuring out how the Spanish pronounce "j", or how they pronounce [ñ], or Germans their [ä], and from that they'll derive some of that pronounciation. On the contrary, saying "pronounce it like Russian ы" (or for that matter korean 'eu', turkic 'ı' - am I right that this is the same sound, Epic?) - they don't eaily figure out what that means.

That's just my personal opinion, though. As a Russian, I have easy time understanding how English people pronounce their 'R', because it's common knowledge, though it never appears in Russian. I had a harder time when a Hindu man tried to explain me the difference between two variants of [t] in Hindi.

[Sir C. M. Siervicül]

Bump! Does anyone have any opinions on the questions in my original post? This seems to be the biggest outstanding issue in reaching a broad consensus (or much broader than we've seen to date, at any rate) on Talossan orthography.

[Marcel Eðo Pairescu Tafial]

I will respond, why not.

1. Would all instances of î be restored as of 1997, or only some? Sir Tomás proposed, and said that he and Ben had agreed, that it should be eliminated where it appeared under the influence of a following n or m, but not where it existed as an independent sound. Dame Miestra, do you recall any further details of what he proposed on the Republican Witt?

Only when not followed by a nasal (so [m], [n] and also [ɲ], since pîgñhetâ exists.)

2. If the answer to î is "only some," should we apply some black and white rule like "î before n or m becomes i, î elsewhere stays î," or something more nuanced? For example, should î remain î in some cases before n or m on etymological or other grounds?

See above. Out of curiosity, what etymological reasons would there be to leave î intact before a nasal?

3. If î goes to i before n and m (either generally or with exceptions), is this because the î pronunciation is now recognised as a regular allophone of i in those environments?

Yes.

4. What about the words where the phoneme was spelled â (e.g. quând and sânc)? Should the î pronunciation be restored there too, and the words respelled with whatever grapheme is settled on for î?

Ah, hm. Really not sure about this one.

5. What about the present participle ending -înd (-ind under A2007), which does not have the î sound (it's pronounced as though spelled -ant)?

Personally, I'd just spell it -ind and keep the irregular pronunciation. Also, I'm in favour of keeping the rule that -ind becomes -ant- when other suffixes are applied, y'know which one I'm talking about.

6. Now about orthography. Should ï be the new grapheme for î, so that the umlaut consistently represents a difference in vowel quality? If so, are there acceptable alternatives for marking diaeresis in all cases where ï is currently used under A2007? And what will the rules for î/ï be in terms of effect on c and diphthong formation? For example, does pre-A2007 cînt ("hundred"), currently chint, become chïnt or cïnt?

For this to work, you'd have to prohibit using ë and ï to mark hiatus, which I'm really not a fan of. Speaking of C though, I'd say that C [k] before I, regardless of diacritic, should be spelled as CH. So, acîrd > achîrd and Cîvián > Chîvian (L'Översteir currently has it respelled as "Civian", I'm guessing this is a typo?)

[Miestrâ Schivâ, UrN]

I endorse cxhn. Taifal's reasoning.

[Sir C. M. Siervicül]

Jul 12, 2018 9:12:43 GMT -6 Marcel Eðo Pairescu Tafial said:

Out of curiosity, what etymological reasons would there be to leave î intact before a nasal?

Well, some î are the result of i being conditioned by a following nasal, but others are the result of a merger of various other vowels in certain contexts (generally before n or r). We might regard the former as allophonic î and the latter as phonemic î. Also, theoretically there could be cases (although I'm not sure any actually exist) where î exists independently in a root but then finds itself in front of a nasal through affixation.

Maybe the question is this: if we're restoring î because we view it as a Talossan phoneme, but we remove the circumflex from all instances of î-nasal, does that imply that the î phoneme never occurs before a nasal, or that the î phoneme is never orthographically marked in that position?

Jul 12, 2018 9:12:43 GMT -6 Marcel Eðo Pairescu Tafial said:

For this to work, you'd have to prohibit using ë and ï to mark hiatus, which I'm really not a fan of. Speaking of C though, I'd say that C [k] before I, regardless of diacritic, should be spelled as CH. So, acîrd > achîrd and Cîvián > Chîvian (L'Översteir currently has it respelled as "Civian", I'm guessing this is a typo?)

True about ë/ï. There are alternatives for marking diæresis/hiatus (such as the circumpunct or apostrophe), but those are less attractive for vowels. But what about irregular stress? I suppose any words with irregular stress on î would just have to be added to the list of words with irregular spelling (I doubt there are more than can be counted on one hand at present). I like the idea of continuing to spell "hard c" as ch before i even when the circumflex is restored. And I think you're right about Chivian, good catch.

Putting everything you suggest together, things would actually look quite a bit like they do under A2007. We'd add î to fewer than 100 Treisour entries (many of which are compounds including vîrt, îrt, acîrd, or vîrb), about 92% of the entries that lost î would continue to not have î (though some of them were respelled with other vowels on etymological grounds--would we change those back to a plain i?), and the cî words that gained an h would keep it.

[Marcel Eðo Pairescu Tafial]

Jul 13, 2018 4:35:03 GMT -6 Sir C. M. Siervicül said:

Maybe the question is this: if we're restoring î because we view it as a Talossan phoneme, but we remove the circumflex from all instances of î-nasal, does that imply that the î phoneme never occurs before a nasal, or that the î phoneme is never orthographically marked in that position?

In my opinion, both analyses are equivalent, but the way I see it, the phoneme is /i/, which is always [ɨ] before nasals.

But what about irregular stress? I suppose any words with irregular stress on î would just have to be added to the list of words with irregular spelling (I doubt there are more than can be counted on one hand at present).

I've looked through the list of words that would retain î, and I havent found any word where î would have to be stressmarked. Of course, should such a word show up in future, you could either mark it with î́ (î with a stress mark on top) or add the word in question to the list of words with irregular spellings, as you said.

Putting everything you suggest together, things would actually look quite a bit like they do under A2007. We'd add î to fewer than 100 Treisour entries (many of which are compounds including vîrt, îrt, acîrd, or vîrb), about 92% of the entries that lost î would continue to not have î (though some of them were respelled with other vowels on etymological grounds--would we change those back to a plain i?), and the cî words that gained an h would keep it.

Oh, I'm well aware. I'm a big fan of the ortho reform in the 2007 Arestada, just this "abolition" of î was untenable. That said, the rationale behind declaring [ɨ] and allophone of /i/ before nasals, and thus respelling î as i before nasals, was a way of resolving the issue where the distribution of îm- / în or im- / in- was completely random. Another way of solving this wouldve been to declare [i​] and allophone of /ɨ/ before nasals and therefore respell i as î in these environments, which accomplishes the same thing but with way more clutter, just like you couldve changed all instances of rd/dr to rð/ðr instead of vice versa.

EDIT: "would we change those back to a plain i?" -- Yes.

[Sir C. M. Siervicül]

Jul 13, 2018 6:20:19 GMT -6 Marcel Eðo Pairescu Tafial said:

In my opinion, both analyses are equivalent, but the way I see it, the phoneme is /i/, which is always [ɨ] before nasals.

I'm not sure about that. Take English phonemes /t/ and /d/. In many American English varieties, at least, they share an allophone (alveolar flap or tap) in intervocalic position, but speakers grok that the underlying form is different. In a phonemic orthography, the graphemes for different phonemes should be different even when the surface representation is the same. It would be a poor orthography for English (even American English), for example, that spelled "bat" and "bad" differently but "badder" (nonstandard but lots of people use it) and "batter" the same.

I've looked through the list of words that would retain î, and I havent found any word where î would have to be stressmarked.

I saw one or two (at least one has a stressmark in the A2007 version), but I don't have a problem with your solution.

[Marcel Eðo Pairescu Tafial]

I've been thinking about the allophonic vs phonemic î thing.

If we wanted to restore "phonemic" î, we'd first have to define what counts as such (maybe non-i vowels merging to î before liquids?), and then we'd have to go through every word that used to have î and check on a case-by-case basis which ones fall under our definition of phonemic î (as an example, two out of three î's in *împîrtînt* > *impîrtînt* would fit) and change all other î's to i as I proposed earlier.

Good news is that this also resolves the în/îm/in/im thing AND we could change quând, sânc. etc to quînd and sînc etc. since theyre phonemic, bad news is that it's more time-consuming to pull off.

Another perhaps unintended consequence is that -înd would most likely retain its î, which makes the -ind > -ant- rule obsolete.

Thoughts?

[Sir C. M. Siervicül]

Jul 13, 2018 17:11:05 GMT -6 Marcel Eðo Pairescu Tafial said:

If we wanted to restore "phonemic" î, we'd first have to define what counts as such (maybe non-i vowels merging to î before liquids?), and then we'd have to go through every word that used to have î and check on a case-by-case basis which ones fall under our definition of phonemic î (as an example, two out of three î's in *împîrtînt* > *impîrtînt* would fit) and change all other î's to i as I proposed earlier.

Makes sense to me. You seem to suggest that â meant a phonemic î, which would help explain why many cases of â before n exist. I've made lists of unique (removing compounds and derived forms, except for whatever I missed) words that feature â (when pronounced î, not the schwa â) and words that have î not before m or n. These lists may provide clues as to how to recognise î as a phoneme.

PORTRAIT portrâteu portrateu (masculine noun)
PRODUCT producât producat (masculine noun)
WHEN quând quand (adverb)
AS FOR quânt à quant à (preposition)
COMEBACK, RETURN revând revand (masculine noun)
ROMANIA Româniâ Romania (feminine placename)
ROMANIAN Româniesc Romaniesc (adjective)
WITHOUT sânc sanc (preposition)
SIGNIFICANT sigñhificând signhificand (adjective)
ACROSS, THROUGH, VIA trâns

ACHOO! (SNEEZE) aciîc'h! aciic'h! (interjection)
ACCORD (AGREEMENT), INFORMAL AGREEMENT acîrd acurd (masculine noun)
ADSORB adsîrpar adsurparh (verb)
ADULT adîlt adült (masculine noun)
ALERT alîrt alert (adjective)
ART îrt art (masculine noun)
ABSURD ausîrd ausürd (adjective)
ABSORBTION, MERGER ausîrvaziun ausurvaziun (feminine noun)
ABSORB ausîrvar ausurvarh (verb)
WARN avîrtar avertarh (verb)
THROW AWAY, DISCARD azvîrlar azvirlarh (verb)
COAL TAR bîgâ biga (feminine noun)
BILL (TO PAY) bîlt bilt (masculine noun)
PIGEON bîschcù bischcù (feminine noun)
BUZZING SOUND bîz! biz! (interjection)
BIZARRE bîzarreu bizarreu (adjective)
BAGUETTE (FRENCH BREAD) blangîtâ blangita (feminine noun)
BOILED bîltescù bültescu (adjective declined as a past participle)
PULP carnît carnit (masculine noun)
KEVIN Cîvián Civian (masculine proper name)
COMFORT cunfîrt cunfurt (masculine noun)
FINGER, TOE dît dit (masculine noun)
EXPERT, SPECIALIST expîrt expert (masculine noun)
FLIRT flîrtar flirtarh (verb)
HANDLE grîp grip (masculine noun)
IMPORTANT împîrtînt impirtint (adjective)
IMPORTANTLY împîrtîntamînt impirtintamint (adverb)
INVIDIOUS învîziös inviziös (adjective)
GRAIN (SMALL PIECE, UNIT) îrd ird (masculine noun)
INTEGRATE îtzegrar itzegrarh (verb)
KYRGYZSTANI Kîrgîz Kirgiz (masculine noun)
BILL OF LADING ladînbîlt ladinbilt (masculine noun)
AWKWARD maladrît maladrit (adjective)
MELT mîltar multarh (verb)
PAUSE piararîftar piarariftarh (verb)
WRIST pîgñhetâ pignheta (feminine noun)
DITCH pît pit (masculine noun)
BOARD (WOOD) plîc plac (masculine noun)
TRANSCEND prevîschar previscharh (verb)
PRIMROSE primavîrâ primavira (feminine noun)
GET USED TO privîcar privicarh (verb)
QUARTER (ONE-FOURTH), SHIFT (WORKING HOURS) quîrt quart (masculine noun)
ROWAN quîrn quirn (masculine noun)
LAUGH rîd rid (masculine noun)
SLATTERN salîp salip (masculine noun)
COMMIT (AN ACT) savîrschar savirscharh (verb)
BRUSH schîtâ schita (feminine noun)
LATTICE, LATTICEWORK sîreu sireu (masculine noun)
HANDMILL tasîrt tasirt (masculine noun)
BEND tîrnar tirnarh (verb)
HARM tîrt tirt (masculine noun)
OPPORTUNITY îportunità uportunità (feminine noun)
VERB vîrb verb (masculine noun)
GREEN vîrt virt (adjective)
OUTSPOKEN vîscasînd viscasind (adjective)
EXHIBITION vîstaucâ vistauca (feminine noun)
VORTEX vîrteschâ vurtescha (feminine noun)
ZIP UP zîpiçar zipiçarh (verb)

[Marcel Eðo Pairescu Tafial]

Jul 17, 2018 15:37:20 GMT -6 Sir C. M. Siervicül said:

You seem to suggest that â meant a phonemic î, which would help explain why many cases of â before n exist.

Yes, but not only â. Occurences like pîgñhetâ (< *pugnitta), tîmp (< tempus) and dînt (< dent-) would, in my opinion, also potentially qualify as phonemic î, since a non-i vowel became î before a nasal, while înalt (< in alto) would not. This is also why the -înd ending would keep its î (< -ant-).

If we wanted to be completely consistent with this approach, we would have to check the etymology of every word in the dictionary that has î before a nasal, to see if the î is just representing an allophone of i before nasals or rather if its a former non-i vowel that later became phonemic î. A lot, if not all, of the words that got their î replaced by an etymologically fitting vowel would qualify as having phonemic î.

In the list you posted, every instance of â [ɨ] and every î (except învîziös or ladînbîlt) would be phonemic î.