Zompist Bboard Again

Posted: **Fri Mar 12, 2021 11:16 am**

Abbreviations with vowel changes would have to be ignored - if there's any process beyond clipping (and maybe affixation like AusEng -o), who knows what's going on.

One possibility for dealing with syllabification problems is to postulate some kind of automatic gemination process. This sounds to me like it's phonetically real in some BrEng (basically for ambisyllabicity), but something like it could also be used to deal with liquid syllabification in AmEng - e.g. "coronavirus" for me is [kɚ.(r)ʌwnəvajrəs] with a rhotic first vowel, but maybe we want either automatic gemination or cross-syllable coalescence (i.e. a formulation of the rule that doesn't care about syllables) to explain that.

(similarly, "conniving" seems like it starts with [kn̩-], and so on)

edit1: fix phoneposting
edit2: on reflection, "coronavirus" probably has [o̜:] or something - [ʌw] after /r/ doesn't seem quite right

Posted: **Fri Mar 12, 2021 3:59 pm**

I have:

coronavirus: [kʰʁ̩ˤˈʁˤʷõːɾ̃əːˌvəe̯ʁˤɘs]
conniving: [kʰn̩ːˈnaːe̯vɘ̃ːŋ]

So yeah, i have "gemination" of intervocalic consonants preceded by /ə/ that may become syllabic, with a syllabic consonant followed by a non-syllabic consonant (note that the labialization in coronavirus is due to preceding /oʊ/).

Posted: **Tue Mar 16, 2021 2:27 pm**

This is an intro text I wrote for the dictionary of my IAL. It's target audience is native English speakers who know nothing about linguistics. (I have only taken into consideration GA and RP, because otherwise this kind of thing becomes impossible to do.) I was wondering if you could take a look at it, and tell me if there are any details that need improvements. I'm not actually expecting anyone to read this but myself, but I'm doing this project just for the challenge of it, so that includes explaining how the language works in an easy to understand way.

Oh, and note that the underline might not be properly displaying under the letter g in the words thing and singer below.

Spelling and pronunciation
Omni-kan uses the Latin alphabet, A-Z. But only some of the letters are actually used in Omni-kan words, while some are only used in foreign names. The letters used in Omni-kan and their pronunciations are explained below.

Aa [a, ɑ] as in palm (not as in face nor as in comma.
Cc [tʃ, tɕ] as in cheap (not as in cut nor as in cent).
Ee [e] as in dress (not as in electric). Note that it is never silent as in home.
Ff [f] as in fish. Note that some speakers may pronounce this similar to P or V, but this is not recommended.
Gg [ŋ] as in thing or singer (not as in gut nor as in giraffe). Note that some speakers might pronounce this similar to N, but this is not recommended.
Hh [h] as in ham. Note that some speakers might omit this sound in their speech, but this is not recommended.
Ii [i] as in happy or kit (not as in price).
Kk [k] as in king. Note that some speakers might pronounce this similar to the G in gut, but this is not recommended.
Ll [l] as in left. (It is never silent as in balm.)
Mm [m] as in map.
Nn [n] as in thin. When it is followed by a consonant, it may be pronounced in the same place in the mouth as the following consonant. *Example words in Omni-kan.
Oo [o] as in British En. thought or American En. north (without the following r-sound). (It is not pronounced as in goat).
Pp [p] as in pit. Note that some speakers might pronounce this similar to F or B, but this is not recommended.
Rr [r] as in run or like the r-like sound some American En. dialects have in the middle of water. Note that some speakers might pronounce this similar to L, but this is not recommended.
Ss [s] as in sap. Note that some speakers might pronounce this similar to Z, or like the sh in mesh, but this is not recommended.
Tt [t] as in tin. Note that some speakers might pronounce this similar to D, but this is not recommended.
Uu [u] as in put (not as in universe nor as in strut).
Ww [ʋ, w, v] as in we (recommendable) or as in vat (acceptable). (It is not pronounced as in law, dew, nor lower.)
Yy [j] as in yes.

Other letters (B, D, J, Q, V, X, Z) are only found in foreign names, and are pronounced in accordance with those names.

Special vowel combinations
There are six different combinations of vowels that can appear in Omni-kan. The pronunciation of them is explained below. No other vowels may appear next to one another in the same word. When alphabetizing, these combinations are treated as two separate letters.

AI [ai, aj, ɑi, ɑj] as in price (not as in mail).
AU [au, aw, ɑu, ɑw] as in cow (not as in fault).
EI [ei, ej] as in eight (not as in heist nor as in field).
OI [oi, oj] as in choice.
OU [ou, ow] as in American En. soul (not as in mouth nor as in you).
UI [ui, uj] as in gooey (not as in guy nor as in guild).

Posted: **Tue Mar 16, 2021 2:55 pm**

Qwynegold wrote: ↑Tue Mar 16, 2021 2:27 pm This is an intro text I wrote for the dictionary of my IAL. It's target audience is native English speakers who know nothing about linguistics. (I have only taken into consideration GA and RP, because otherwise this kind of thing becomes impossible to do.) I was wondering if you could take a look at it, and tell me if there are any details that need improvements. I'm not actually expecting anyone to read this but myself, but I'm doing this project just for the challenge of it, so that includes explaining how the language works in an easy to understand way.

The "c g" as [tʃ ŋ], while hypothetically comprehensible, are going to be fighting conditioned readings for most languages using the Roman alphabet (not merely English, but also French and Hanyu Pinyin (none of the rest strike me as particularly problematic, except maybe "o" being very especially [ɔ] and not a range of /o/-like vowels); it is variably the Japanese pronunication, I understand); I also question the wisdom of including all of /l r tʃ ŋ/ in an auxiliary language, as opposed to just one of /l r/ and /tj/ or /s/, and /nk/, respectively. Same with all the diphthongs, which speakers of some languages will likely have trouble pronouncing.

Posted: **Tue Mar 16, 2021 3:23 pm**

Note that father is better than palm, because there are people like me who have /pɔːlm/ ([pʰɒ(ː)o̯m] for me) for palm (yes, a spelling pronunciation).

Posted: **Tue Mar 16, 2021 4:32 pm**

Sad to say, no matter what you write, I suspect you’ll still get people saying e.g. /əsˈkeɹɪk/ and /ˈsetɹəɡfiː/ for ESKERIC and SETREGFE.

Posted: **Tue Mar 16, 2021 6:24 pm**

Qwynegold wrote: ↑Tue Mar 16, 2021 2:27 pm Oo [o] as in British En. thought or American En. north (without the following r-sound). (It is not pronounced as in goat).

I think most linguistically-naive speakers of AmE will struggle to isolate the [ɔ] of north, and, if they have the cot-caught merger, won't understand what's supposed to be the difference between that and the GOAT vowel. (If they don't have the cot-caught merger then the example of thought works fine.)

OU [ou, ow] as in American En. soul (not as in mouth nor as in you).

Whatever phonetic nuance you're picking out by naming specifically the American pronunciation of specifically the word soul is going to be lost on pretty much everyone; you're better off just identifying OU with the GOAT set.

Aa [a, ɑ] as in palm (not as in face nor as in comma.)

I second Travis on palm—I have /l/ in that word too, and it messes a bit with the preceding vowel. For the same reason you might want to use walk rather than balm as your example of silent <L>.

Posted: **Tue Mar 16, 2021 6:29 pm**

Many people here wouldn't even know that GOAT is a diphthong, even though they commonly pronounce it as one in hello. Likewise, I bet many people here don't perceive a difference between the vowels of NORTH/FORCE and GOAT.

Posted: **Tue Mar 16, 2021 9:18 pm**

Rounin Ryuuji wrote: ↑Tue Mar 16, 2021 2:55 pm I also question the wisdom of including all of /l r tʃ ŋ/ in an auxiliary language, as opposed to just one of /l r/ and /tj/ or /s/, and /nk/, respectively. Same with all the diphthongs, which speakers of some languages will likely have trouble pronouncing.

I don't agree with the common doctrine that IALs ought to have ultra-simple phonologies without voicing contrasts, lateral/rhotic distinctions, etc. While a IAL shouldn't have a overly complicated phonology, simplifying things too much delivers diminishing returns and means that the a posteriori vocabulary which forms much of a IAL's backbone will be left unrecognisable. Additionally, speakers will have to adjust to the auxlang's phonological system no matter what; you can't design that problem away. Finally, if a IAL achieves widespread adoption, common phonemes that have been excluded will probably enter into the language through technical terminology, placenames, etc.; at first they may seem foreign, but eventually speakers will be unable to dispense with, say, /z/, rendering a ultra-minimalist phonology pointless.

As for the pronunciation examples, trying to have one system covering all accents of English is futile; for a NZE speaker, saying that <i> should be as in as in happy or kit would be puzzling, since those are two quite different sounds in NZE! It'd be better to systematically set up a system of correspondences for different English accents (probably AmE, BrE, and AuE/NZE, noting regional differences where applicable). Pronunciation guides for other major languages too would be good, such as Mandarin, French, Spanish, German, Russian, Arabic, Hindi, Malay/Indonesian, Swahili (think regional lingua francas), so someone who knows, say, English and Russian can "cross-check" the values of sounds (maybe I'm taking this too seriously tho)

Posted: **Wed Mar 17, 2021 6:46 am**

Sol717 wrote: ↑Tue Mar 16, 2021 9:18 pm
Rounin Ryuuji wrote: ↑Tue Mar 16, 2021 2:55 pm I also question the wisdom of including all of /l r tʃ ŋ/ in an auxiliary language, as opposed to just one of /l r/ and /tj/ or /s/, and /nk/, respectively. Same with all the diphthongs, which speakers of some languages will likely have trouble pronouncing.
I don't agree with the common doctrine that IALs ought to have ultra-simple phonologies without voicing contrasts, lateral/rhotic distinctions, etc. While a IAL shouldn't have a overly complicated phonology, simplifying things too much delivers diminishing returns and means that the a posteriori vocabulary which forms much of a IAL's backbone will be left unrecognisable. …

I don’t agree with this. As a proof by construction, here is an auxlang inventory I constructed a while ago, which is only 10 consonants but still manages to retain recognisability:

/m n/
/p t k/
/s h/
/l j w/

/a ə i u aː iː uː/

(+ all syllables (C)V(nasal), no hiatus allowed)

And some examples of random words being loaned to this phonology while not being distorted too extensively:

glottal → kəlutəl
ambidextrous → ampiti꞉kəsətələs
emergency → əməkənsi
Schadenfreude → sa꞉dənfəluyit
tkhelet → təhilət
Pǔtōnghuà → putunhuwa

Posted: **Wed Mar 17, 2021 7:41 am**

bradrn wrote: ↑Wed Mar 17, 2021 6:46 am /m n/
/p t k/
/s h/
/l j w/

/a ə i u aː iː uː/

Jeffrey Henning has you beat here... years ago he created Kalisise, whose phonemic inventory is

p t k
s l n
a e i u

Posted: **Wed Mar 17, 2021 8:42 am**

zompist wrote: ↑Wed Mar 17, 2021 7:41 am
bradrn wrote: ↑Wed Mar 17, 2021 6:46 am /m n/
/p t k/
/s h/
/l j w/

/a ə i u aː iː uː/
Jeffrey Henning has you beat here... years ago he created Kalisise, whose phonemic inventory is

p t k
s l n
a e i u

Sure, in terms of raw size, it’s possible to get plenty smaller than what I have. (Even amongst terrestrial languages, Proto–Lakes Plains had /p t k b d/.) But in the context of auxlangs, the smaller you go, the more you have to distort words if you want to borrow them. e.g. I see no easy way of borrowing Pǔtōnghuà into Kalisise in a recognisable form — the closest I can get is putunka, compared to my own putunhuwa.

Posted: **Wed Mar 17, 2021 10:17 am**

Qwynegold wrote: ↑Tue Mar 16, 2021 2:27 pm Ii [i] as in happy or kit (not as in price).

I assume both pronunciations are acceptable here?

Nn [n] as in thin. When it is followed by a consonant, it may be pronounced in the same place in the mouth as the following consonant.

I'm not sure this would be comprehensible to the average person, who I doubt is consciously aware of where they produce sounds at all. I think you would be better off either not having this note and letting allophony have her way, or give specific examples instead.

Oo [o] as in British En. thought or American En. north (without the following r-sound). (It is not pronounced as in goat).

Agreed with the others on the north/goat thing. You don't even need the cot-caught merger for this one to be confusing, either--in Inland North, my north and goat are also basically the same (or at least similar enough that a non-linguistically-minded person wouldn't pick up on the distinction).

Rr [r] as in run or like the r-like sound some American En. dialects have in the middle of water. Note that some speakers might pronounce this similar to L, but this is not recommended.

Not sure your average Engish speaker would actually pick up on a flap/tap as being "r-like". I'd say they're more likely to think of it as being similar to /d/, in fact--sometimes you'll see <water> written dialectically as <wadder> or something like that.

Tt [t] as in tin. Note that some speakers might pronounce this similar to D, but this is not recommended.

Given that there's no /d/ in the "native" orthography, it doesn't seem like it'd cause much confusion to allow this kind of variance.

Posted: **Wed Mar 17, 2021 7:48 pm**

bradrn wrote: ↑Wed Mar 17, 2021 6:46 amI don’t agree with this. As a proof by construction, here is an auxlang inventory I constructed a while ago, which is only 10 consonants but still manages to retain recognisability:

While those forms might be recognisable to us conlangers staring at them on a page, you'll need a higher degree of approximation in running speech for the words to be recognisable. Additionally, the lack of a voicing (or other phonation) contrast may be confusing for speakers of languages with one; there's a good chance that one will enter into the auxlang by loans anyway, frustrating attempts to eliminate one.

Posted: **Mon Mar 22, 2021 5:56 am**

Going back to my question about "not to" and "to not" in page 2, today I noticed that the problem can easily be avoided in "tend not to do" ~ "tend to not do" by negating "tend" instead (the third option):

1. They tend to not think that way.
2. They tend not to think that way.
3. They don't tend to think that way.

A phenomenon that, yes, zompist mentioned at the time, but the naturalness of the third option with its greater neg-hopping really jumped at me here.

Posted: **Sun Mar 28, 2021 2:11 am**

Rounin Ryuuji wrote: ↑Tue Mar 16, 2021 2:55 pm The "c g" as [tʃ ŋ], while hypothetically comprehensible, are going to be fighting conditioned readings for most languages using the Roman alphabet (not merely English, but also French and Hanyu Pinyin

Yeah, I know, but I don't want to use superfluous letters as in <ch>, <ng>. I've heard that many Chinese people don't even know Pinyin. In any case, [ts_h] would be a kinda acceptable pronunciation of C (see the next comment below).

Rounin Ryuuji wrote: ↑Tue Mar 16, 2021 2:55 pm(none of the rest strike me as particularly problematic, except maybe "o" being very especially [ɔ] and not a range of /o/-like vowels); it is variably the Japanese pronunication, I understand);

For most phonemes there are actually a few different realizations that are okay, but some realizations are more or less okay than others. But I didn't want to include that level of detail in this text. So I've focused on some kind of combination of the most desired realizations and what you realistically can expect from English speakers. In the case of /o/, mid high and mid realizations are the most desirable. Mid low is acceptable, but not as good, because some people might perceive it as /a/. But mid low is the best you can hope for when it comes to English speakers, so I've tried to exemplify that.

Rounin Ryuuji wrote: ↑Tue Mar 16, 2021 2:55 pmI also question the wisdom of including all of /l r tʃ ŋ/ in an auxiliary language, as opposed to just one of /l r/ and /tj/ or /s/, and /nk/, respectively. Same with all the diphthongs, which speakers of some languages will likely have trouble pronouncing.

Some of these "phonemes" don't actually form minimal pairs with one another, except in proper nouns and maybe interjections. Except for /tS/. Are there people that have such big problems with affricates, except for the Swedes?

Posted: **Sun Mar 28, 2021 2:12 am**

Travis B. wrote: ↑Tue Mar 16, 2021 3:23 pm Note that father is better than palm, because there are people like me who have /pɔːlm/ ([pʰɒ(ː)o̯m] for me) for palm (yes, a spelling pronunciation).

aporaporimos wrote: ↑Tue Mar 16, 2021 6:24 pm I second Travis on palm—I have /l/ in that word too, and it messes a bit with the preceding vowel. For the same reason you might want to use walk rather than balm as your example of silent <L>.

Okay, thanks. I'll change those examples.

Posted: **Sun Mar 28, 2021 2:18 am**

bradrn wrote: ↑Tue Mar 16, 2021 4:32 pm Sad to say, no matter what you write, I suspect you’ll still get people saying e.g. /əsˈkeɹɪk/ and /ˈsetɹəɡfiː/ for ESKERIC and SETREGFE.

Well, they will just have to learn it.

It's just two letters that have different values than in English. Compare that to natlangs that tend to have more unintuitive letters, and sometimes also complicated rules regarding those letters.

Posted: **Sun Mar 28, 2021 2:32 am**

aporaporimos wrote: ↑Tue Mar 16, 2021 6:24 pm
Qwynegold wrote: ↑Tue Mar 16, 2021 2:27 pm Oo [o] as in British En. thought or American En. north (without the following r-sound). (It is not pronounced as in goat).

I think most linguistically-naive speakers of AmE will struggle to isolate the [ɔ] of north, and, if they have the cot-caught merger, won't understand what's supposed to be the difference between that and the GOAT vowel. (If they don't have the cot-caught merger then the example of thought works fine.)

Travis B. wrote: ↑Tue Mar 16, 2021 6:29 pm Many people here wouldn't even know that GOAT is a diphthong, even though they commonly pronounce it as one in hello. Likewise, I bet many people here don't perceive a difference between the vowels of NORTH/FORCE and GOAT.

This part is very difficult, because I want to teach people to use a monophthong [o~O], and not the diphthong ([oU] or some such). Speakers of all other languages are capable of producing the monophthong.

So I don't know what to do.

aporaporimos wrote: ↑Tue Mar 16, 2021 6:24 pm
OU [ou, ow] as in American En. soul (not as in mouth nor as in you).

Whatever phonetic nuance you're picking out by naming specifically the American pronunciation of specifically the word soul is going to be lost on pretty much everyone; you're better off just identifying OU with the GOAT set.

According to Wiktionary it's [s@Ul] in RP, and that's not quite right.

I used this word as an example because it's actually spelled with <ou>, so I wanted people to associate <ou> with /ou/.

Posted: **Sun Mar 28, 2021 2:53 am**

Sol717 wrote: ↑Tue Mar 16, 2021 9:18 pm While a IAL shouldn't have a overly complicated phonology, simplifying things too much delivers diminishing returns and means that the a posteriori vocabulary which forms much of a IAL's backbone will be left unrecognisable.

THIS! Even with this phonology I have problems with unrecognizability.

Sol717 wrote: ↑Tue Mar 16, 2021 9:18 pmAs for the pronunciation examples, trying to have one system covering all accents of English is futile; for a NZE speaker, saying that <i> should be as in as in happy or kit would be puzzling, since those are two quite different sounds in NZE!

Yeah, that's why I only focused on GA and RP.

Sol717 wrote: ↑Tue Mar 16, 2021 9:18 pm It'd be better to systematically set up a system of correspondences for different English accents (probably AmE, BrE, and AuE/NZE, noting regional differences where applicable).

I'm not sure I understand what you mean.

Sol717 wrote: ↑Tue Mar 16, 2021 9:18 pm Pronunciation guides for other major languages too would be good, such as Mandarin, French, Spanish, German, Russian, Arabic, Hindi, Malay/Indonesian, Swahili (think regional lingua francas), so someone who knows, say, English and Russian can "cross-check" the values of sounds (maybe I'm taking this too seriously tho)

This text is from my dictionary, which is glossed in four languages, so this text also comes in four languages, with different example words of course. The grammar of this conlang has a big section on phonology. (It's not aimed at linguistically naive people though.) I intend the grammar to be translated into various major languages, but since I don't speak those languages, I'm not actually gonna do it.

Zompist Bboard Again

English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions

Re: English questions