AmEng-based conlang

A forum for all topics related to constructed languages
Post Reply
JuRi
rupestrian
rupestrian
Posts: 5
Joined: 01 Aug 2020 17:04

AmEng-based conlang

Post by JuRi »

Here is the phonology (IPA notation) of my conlang (with its orthography)

Consonants:

p /p/ b /b/ m /m/ ḿ /m̥/ fw /ɸ/ f /f/ t /t/ d /d/ n /n/ ń /n̥/ s /s/ z /z/ ł /ɬˠ/ r /ɾ/ ř /ɹ/ l /ɫ/ q /tʃ/ j /dʒ/ š /ʃ/ ž /ʒ/ ŕ /ɽ/ y /j/ c /k/ g /g/ w /w/

Vowels & Diphthongs

a /ä/ ä /a/ å /ɑ/ ää /æː/ aë /äɤ̯/ äë /aɤ̯/ åë /ɑɤ̯/ ai /äɪ̯/ au /äʊ̯/ äi /æɪ̯/ äö /aø̯/ åö /ɑø̯/ ääo /æːo̯/ e /e/ é /ɛ/ ë /ɤ/ eë /eɤ̯/ éë /ɛɤ̯/ ei /eɪ̯/ éi /ɛɪ̯/ ëu /ʌ̞ʊ̯/ ëëe /ʌ̞ːe̯/ i /i/ ï /ɯ/ ië /iɤ̯/ ïë /ɯɤ̯/ eö /eø̯/ éö /ɛø̯/ o /o/ ö /ø/ óó /ɔ̞ː/ ou /oʊ̯/ óu /ɔʊ̯/ oë /oɤ̯/ óë /ɔɤ̯/ oö /oø̯/ öü /øʏ̯/ ööü /øːʏ̯/ u /u/ ü /y/ uë /uɤ̯/ üë /yɤ̯/
sasasha
rupestrian
rupestrian
Posts: 23
Joined: 08 Jul 2020 14:18

Re: AmEng-based conlang

Post by sasasha »

Hi, I'm sure you're going to add to this, and I'm curious to see what you're going to do. [:)]

But I just want to query something: did I count 43 vowel segments? With 18 different quality distinctions? This is a lot! Quite possibly more than any natural language.

Of 564 natlangs surveyed by WALS, the highest number of attested vowel quality distinctions is German, with 14. Only two languages in the survey have 13 vowel quality distinctions. https://wals.info/chapter/2

Unless this language is supposed to be highly unnaturalistic, which from the title of the thread I'm guessing isn't the case, I would suggest maybe limiting your vowel inventory a bit more.
Khemehekis
mongolian
mongolian
Posts: 3884
Joined: 14 Aug 2010 09:36
Location: California über alles

Re: AmEng-based conlang

Post by Khemehekis »

The English words seat, sit, sate, set, sat, sot, soot, suit show that English has at least 8 different vowel sounds.
Is there some reason this list includes "sate" but not "site"? You could argue that /ai/ is a diphthong and shouldn't be counted, but /ei/ is also a diphthong, and they listed the word "sate".
♂♥♂♀

Squirrels chase koi . . . chase squirrels

My Kankonian-English dictionary: 86,336 words and counting

31,416: The number of the conlanging beast!
sasasha
rupestrian
rupestrian
Posts: 23
Joined: 08 Jul 2020 14:18

Re: AmEng-based conlang

Post by sasasha »

Khemehekis wrote: 03 Aug 2020 09:05
The English words seat, sit, sate, set, sat, sot, soot, suit show that English has at least 8 different vowel sounds.
Is there some reason this list includes "sate" but not "site"? You could argue that /ai/ is a diphthong and shouldn't be counted, but /ei/ is also a diphthong, and they listed the word "sate".
True; I guess it's not supposed to be an exhaustive list.

Looking elsewhere for information, it seems some varieties of Swedish can be said to have 17 vowel quality distinctions, and in some analyses Dinka has 336 distinct vowel segments (including tone and phonatory distinctions), so perhaps this isn't as out there as I first thought.
Salmoneus
MVP
MVP
Posts: 3031
Joined: 19 Sep 2011 19:37

Re: AmEng-based conlang

Post by Salmoneus »

The cases of /aI/ and /eI/ aren't equivalent, though.

The English diphthong [eI] is just one dialectical realisation of the phoneme [e] - much as in many dialects /i/ is more likely to be realised as [iI]. It can be treated as a distinct quality because a) it's phonemically equivalent to the [e] found in other dialects, and b) it contains a quality, [e], that is entirely distinct from any other phonemic vowel quality in the language.

It's not a cast-iron case, of course. On the one hand, some dialects have [a:] for /ai/ too. On the other, while [eI] can be phonemically distinguished from /i/ and /ai/, it can't be distinguished from a putative /Ei/ or /E\i/. So you could indeed argue that [eI] is just /E\/ (or /E/, for that matter, though only in cetain dialects) plus /i/.

However, the general cross-dialectical situation, intuition, and patterning concerns make it very attractive to consider English as having a vowel /e(:)/ - whereas there's no particularly compelling reason not to treat /ai/ in most dialects as being /{/ (or some other low unrounded vowel) + /i/.

---------

Anyway, FWIW, my English arguably has 17 vowel qualities, although you can reduce this somewhat if you analyse some of them as schwa-diphthongs or long vowels (eg, the vowel in CURE is a long mostly-monophthong with a very different quality from the short monophthong in PUT - but you could, if you needed to, call it the PUT vowel either with a length contrast or with a schwa offglide - although it's more monophthongal than, say, the vowel in PEAT), and if you refuse to consider schwa as phonemic. That's not counting the /ai/ and /oi/ diphthongs, the TIRE and TOWER triphthongs (incipiently adding another monophthong quality), and arguably the see-er and sue-er diphthongs.

However, this is clearly right at the edge of what's possible, I think, as shown by the ways the system is collapsing - diphthongisation, regularisation to a length contrast, merger of several qualities by many speakers.

-------------

Languages like Dinka, with tone and phonatory distinctions, aren't really relevant. When you add systemic suprasegmental features to a vowel system, you can very rapidly - exponentially! - multiply the number of 'vowel segments', but it's not that relevant to the number of vowel qualities distinguished.
sasasha
rupestrian
rupestrian
Posts: 23
Joined: 08 Jul 2020 14:18

Re: AmEng-based conlang

Post by sasasha »

Salmoneus wrote: 03 Aug 2020 15:30 The cases of /aI/ and /eI/ aren't equivalent, though.
True.
However, this is clearly right at the edge of what's possible, I think, as shown by the ways the system is collapsing - diphthongisation, regularisation to a length contrast, merger of several qualities by many speakers.

-------------

Languages like Dinka, with tone and phonatory distinctions, aren't really relevant. When you add systemic suprasegmental features to a vowel system, you can very rapidly - exponentially! - multiply the number of 'vowel segments', but it's not that relevant to the number of vowel qualities distinguished.
Yes. Well. Whether JuRi's language's vowel system has the most quality distinctions or not, it still has an unusually large number of them, and would probably benefit from being edited a little.
JuRi
rupestrian
rupestrian
Posts: 5
Joined: 01 Aug 2020 17:04

Re: AmEng-based conlang

Post by JuRi »

Some of the vowels qualities only occurr in diphthongs. As such, they can be considered allophones.
JuRi
rupestrian
rupestrian
Posts: 5
Joined: 01 Aug 2020 17:04

Re: AmEng-based conlang

Post by JuRi »

Also, a contrast between /f/ and /ɸ/ phonemes is cross-linguistically rare.
JuRi
rupestrian
rupestrian
Posts: 5
Joined: 01 Aug 2020 17:04

Re: AmEng-based conlang

Post by JuRi »

Here are the changes to the language:

θ→t ð→d h→∅ ʃ→s ʒ→z tw→p dw→b kj→tʃ gj→dʒ k[front vowel]→tʃ[front vowel] g[front vowel]→dʒ[front vowel] θj→ʃ ðj→ʒ θw→ɸ ðw→β sw→ɸ zw→β ɹt→ʈʂ→tʃ ɹd→ɖʐ→dʒ ɹɾ→ɽ ɹs→ʂ→ʃ ɹz→ʐ→ʒ sm→m̥ sn→n̥ sɹ→ʃ sɫ→ɬˠ tɹ̝̊→tʃ dɹ̝→dʒ stɹ̝̊→ʃ zdɹ̝→ʒ

æ→ɛ æˑ→æɪ̯ ɛə̯→jä ɛˑ→ɛɪ̯ eɪ̯→iː ɪ→e ɪˑ→eː→eɪ̯ i→ɯ iˑ→äɪ̯ ɑ→a ɑˑ→æʊ̯ ɔ→ɑ ɔˑ→ɔʊ̯ oʊ̯→u ʊ→o ʊˑ→oː→oʊ̯ u→y uˑ→äʊ̯ ʌ→ä ʌˑ→äɤ̯ äˑɪ̯→ʌ̞ːe̯ äˑʊ→æːo̯ ɐɪ→æː ɐʊ̯→ɔ̞ː ɔɪ̯→øʏ̯ ɔˑɪ→øːʏ̯ ɑɹ→ɑə̯→aø̯ ɚ→ə ə→ø (non-finally) ə→ä (finally) ɪɹ→eə̯→eø̯ ɔɹ→ɔə̯→ɑø̯ ʊɹ→oə̯→oø̯ jʊɹ→joə̯→jø ɫ→ɤ̯

The English it's based on isn't exclusively GenAm, it's instead a mixture of it, Canadian, Southern, Midland, and Inland North. Also, äˑɪ̯→ʌ̞ːe̯ and äˑʊ→æːo̯ precede iˑ→äɪ̯ and uˑ→äʊ̯. Some of these precursors have no short or half-long counterparts because of factors like Canadian raising. The i→ɯ change was to make it more exotic and give it a Turkic flavor.
JuRi
rupestrian
rupestrian
Posts: 5
Joined: 01 Aug 2020 17:04

Re: AmEng-based conlang

Post by JuRi »

Decided to turn the /øʏ̯/-/øːʏ̯/ distinction to a /œ/-/ø/ one to create symmetry with the front unrounded (and to a lesser extent the back rounded) series. As such, /ø/ and /øːʏ̯/ would merge into /ø/. /ɔ̞ː/ can be allophonically [ɔː], while /æː/ could be allophonically [ɛː]. [ɤ], [ʌ̞], and [ʌ̞ː] could be analyzed as a single vowel in complimentary distribution, and indeed they have the same graph (ë(ë)). It all depends on speaker choice.
Post Reply